Membrane electrode assembly for hydrogen production

Abstract

The present disclosure relates to a membrane electrode assembly for hydrogen production and a method of producing hydrogen using the membrane electrode assembly

Description

[0001] Description

[0002] MEMBRANE ELECTRODE ASSEMBLY FOR HYDROGEN PRODUCTION

[0003] Technical Field

[0004] The present invention generally relates to a novel construction of a membrane electrode assembly (MEA), which may be used for hydrogen production.

[0005] Background

[0006] Hydrogen is a clean, efficient and sustainable energy carrier, and is regarded as an ideal alternative to fossil fuels for addressing the global warming crisis. To date, most of the hydrogen is produced via traditional methods, such as methane steam reforming, oil reforming and coal gasification, which consume a huge amount of energy and release undesirable levels of greenhouse gases (GHG). Additionally, due to logistical difficulties in transporting and storing hydrogen, on-site and on- demand hydrogen generation is generally preferred. Electrochemical hydrogen production from water splitting has been studied for GHG-free hydrogen production. The electrolysis process splits water into hydrogen and oxygen. The electric energy consumption for hydrogen production through water electrolysis is correlated with the multiple of the cell voltage, the sum of both anodic and cathodic reactions (including oxygen evolution reaction (OER) at anode and hydrogen generation at cathode, respectively), and the charge transfer. One of the problems associated with water electrolysis is the high standard electrical potential required for the OER.

[0007] Accordingly, it has been proposed to replace OER with a thermodynamically more favorable reaction. The glycerol oxidation reaction (GOR) has a comparatively lower standard potential (0.003 V) than that of the oxygen evolution reaction. Glycerol is a readily available by-product obtainable from biomass conversion and is produced on a large scale with stable increments annually. Many of the glycerol oxidation products such as formic acid, oxalic acid, glyceraldehyde and etc, also serve as useful intermediates or valuable fine chemicals to the pharmaceutical, cosmetics, polymer, and food industries. Accordingly, utilizing GOR as the anode reaction for hydrogen electrolysis-production represents a promising technical solution for hydrogen production.

[0008] However, currently available electrochemical cells for glycerol oxidation are mostly operated at a current density lower than 200 mA cm-2, which reduces overall efficiency when paired with a H2 evolution reaction (HER), which requires high current density. Even with the use of highly catalytically active electrocatalysts, electrochemical cells which are capable of achieving current densities greater than 200 mA cm-2at low cell voltages remain elusive. Besides, in typical GOR MEA electrolyzers, the crossover of liquid productions via anion exchange membrane (AEM) reduces the hydrogen evolution reaction (HER) activity at the cathode, thereby increasing the hydrogen production energy consumption, requiring further product separation / purification processes and resulting in overall higher costs. Moreover, the tendency of C-C bond cleavage at high cell voltages leads to lower C3 products selectivity and more carbon gas emission, which compromise the profitability of GOR considering the higher value of C3 products than C2 and C1 products.

[0009] Accordingly, there is a need to provide a novel electrochemical cell which ameliorates or overcomes one or more of the above drawbacks described above.

[0010] Summary of Invention

[0011] In one aspect, the present disclosure provides a membrane electrode assembly for hydrogen production, comprising: an anode compartment having an inlet for receiving an anolyte comprising an alcohol and an alkali and an outlet for discharging reaction products of an alcohol oxidation reaction, the anode compartment having an electrically conductive layer coated with at least one anode catalyst; a cathode compartment having an inlet for receiving a catholyte comprising an acid and an outlet for discharging hydrogen, the cathode compartment having an electrically conductive layer coated with at least one cathode catalyst; an ion exchange membrane interposed between the anode compartment and the cathode compartment, wherein the ion exchange membrane comprises opposite surfaces respectively in contact with the anode catalyst and the cathode catalyst, thereby forming a continuous structure with the anode compartment and the cathode compartment.

[0012] Due to the asymmetric, alkali anolyte-acid catholyte configuration, the membrane electrode assembly as disclosed herein advantageously requires less cell voltage for hydrogen production when compared to conventional H-cell or flow cells or other MEA configurations.

[0013] Additionally, the membrane electrode assembly as disclosed herein is capable of reducing, suppressing or completely avoiding liquid product crossover when the electrochemical oxidation and reduction reactions are taking place. This effect may be due to the pH difference between the anolyte and catholyte, which facilitates hydroxide anion diffusion from anolyte to catholyte. Such hydroxide anions diffusion along with electromigration of ion species under an external voltage supply from catholyte to anolyte may partially or completely prevent or suppress the crossover of anions other than hydroxide ions from anolyte to catholyte due to limited ion exchange capacity of the ion exchange membrane. This useful effect may be further complemented by selecting ion exchange membranes which block the exchange of oxidation products.

[0014] Moreover, the membrane electrode assembly as disclosed herein may result in the selective oxidation of the fuel (e.g., glycerol) into higher-value liquid C3 products (such as tartronate, glycerate, etc.) over lower value C2 and / or C1 products (e.g. formate, acetate, oxalate and glycolate) and carbon-containing gas. This oxidation selectivity may be caused by the suppression of C-C bond cleavage in the anolyte due to the pH difference between anolyte and catholyte.

[0015] In another aspect of the invention, there is provided a method of producing hydrogen via the membrane electrode assembly as disclosed herein, the method comprising: (a) electrochemically oxidizing the alcohol in the anode compartment in the presence of at least one alkali, and (b) concurrently electrochemically reducing protons in the catholyte housed in the cathode compartment to thereby evolve hydrogen gas.

[0016] In yet another aspect of the invention, there is provided an electrochemical cell comprising the membrane electrode assembly as disclosed herein.

[0017] Definitions

[0018] The following words and terms used herein shall have the meaning indicated:

[0019] The word “substantially” does not exclude “completely” e.g. a composition which is “substantially free” from Y may be completely free from Y. Where necessary, the word “substantially” may be omitted from the definition of the invention.

[0020] Unless specified otherwise, the terms "comprising" and "comprise", and grammatical variants thereof, are intended to represent "open" or "inclusive" language such that they include recited elements but also permit inclusion of additional, unrecited elements.

[0021] As used herein, the term "about", in the context of concentrations of components of the formulations, typically means + / - 5% of the stated value, more typically + / - 4% of the stated value, more typically + / - 3% of the stated value, more typically, + / - 2% of the stated value, even more typically + / - 1% of the stated value, and even more typically + / - 0.5% of the stated value.

[0022] Throughout this disclosure, certain embodiments may be disclosed in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the disclosed ranges. Accordingly, the description of a range should be considered to have specifically disclosed all the possible sub-ranges as well as individual numerical values within that range. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed sub-ranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1 , 2, 3, 4, 5, and 6. This applies regardless of the breadth of the range.

[0023] Certain embodiments may also be described broadly and generically herein. Each of the narrower species and subgeneric groupings falling within the generic disclosure also form part of the disclosure. This includes the generic description of the embodiments with a proviso or negative limitation removing any subject matter from the genus, regardless of whether or not the excised material is specifically recited herein.

[0024] Detailed Disclosure of Embodiments

[0025] The following detailed description is merely exemplary in nature and is not intended to limit the invention or the application and uses of the invention. Exemplary, nonlimiting embodiments of a membrane electrode assembly will now be disclosed.

[0026] In embodiments, the membrane electrode assembly (MEA) as disclosed herein may comprise an anode compartment having an inlet for receiving an anolyte comprising an alcohol and an alkali and an outlet for discharging reaction products of an alcohol oxidation reaction, the anode compartment having an electrically conductive layer coated with at least one anode catalyst; a cathode compartment having an inlet for receiving a catholyte comprising an acid and an outlet for discharging hydrogen, the cathode compartment having an electrically conductive layer coated with at least one cathode catalyst; an ion exchange membrane interposed between the anode compartment and the cathode compartment, wherein the ion exchange membrane comprises opposite surfaces respectively in contact with the anode catalyst and the cathode catalyst, thereby forming a continuous structure with the anode compartment and the cathode compartment.

[0027] The ion exchange membrane may be selected from anion exchange membrane (AEM), proton exchange membrane (PEM), bipolar membrane (BPM) or mixtures thereof. The ion exchange membrane may be selected from anion exchange membrane (AEM), proton exchange membrane (PEM), or combinations thereof. The ionic exchange membrane serves to separate the anode and cathode compartments. The ionic exchange membrane may be configured to be in fluid communication with both the anode compartment and the cathode compartment but restricting the species which can be exchanged between the anode and cathode compartments. For example, where the ionic exchange membrane is an AEM, only the exchange of anions between the cathode and anode compartments may be permitted whereas the passages of gases, electrons and cations are blocked. Alternatively, when the ionic exchange membrane is a PEM, only the exchange of protons between the cathode and anode compartments is allowed whereas the passages of gases, electrons and anions are restricted. Further alternatively, where the ionic exchange membrane is a BPM, the exchange of both anions and cations is permitted while the passages of gases and electrons are blocked.

[0028] In some embodiments, the ionic exchange membrane is an AEM. The AEM may comprise a copolymer of poly(4-vinylbenzyl chloride-co-styrene) (polymer backbone) functionalized with 1-benzyl-2,3,4,5-tetramethyl-imidazolium groups (anion exchange sites), poly(phenylene oxide) polymer backbone grafted with imidazolium, ammonium groups, poly(arylene ether) polymer backbone with quaternary ammonium groups, poly(aryl piperidinium) polymer backbone with piperidinium cation (quaternary ammonium in a piperidine ring) or mixtures thereof. The AEM may be selected from Sustainion AEM, Fumasep AEM, PiperlON AEM, Alkymer AEM or combinations thereof. More preferably, the AEM may be Tokuyama A201 , Fumasep FAA-3 or Sustainion X37- 50. In embodiments, the AEM is Sustainion X-37-50, which comprises a copolymer of poly(4-vinylbenzyl chloride-co-styrene) (polymer backbone) functionalized with 1 -benzyl- 2,3,4,5-tetramethyl-imidazolium groups (anion exchange sites) and porous polytetrafluoroethylene (PTFE) layers on both sides of the membrane for mechanical reinforcements. In some embodiments, the AEM used in the membrane electrode assembly may be converted to its hydroxide form prior to use. In some embodiments, the ionic exchange membrane is PEM. The PEM may comprise perfluorosulfonic acid (PFSA) polymers having PTFE backbone functionalized with side chains terminated with sulfonic acid groups. The PEM may be selected from Nation PFSA PEM, 3M PFSA PEM, Aquivion PFSA PEM, Fumasep PFSA PEM or combinations thereof. Preferably, the PEM may be Nation PEM.

[0029] When in use, the ionic exchange membranes may have a limited ion exchange capacity (IEC) depending on factors such as the density of the active sites for ion exchange on the membrane. In practice, when the active sites are occupied with ions for exchange, further exchange of ion species from the catholyte and / or anolyte will be hampered.

[0030] In certain embodiments, the anolyte may contain a fuel for the oxidation reaction. The fuel may be an alcohol selected from glycerol, methanol, ethanol, ethylene glycol, isopropanol, propylene glycol, butanol or mixtures thereof. Preferably, the alcohol may be selected from glycerol, methanol, ethanol, ethylene glycol or mixtures thereof. More preferably, the alcohol may be selected from glycerol, methanol, ethanol, ethylene glycol, and in particular, glycerol. The alcohol may be selected to have an oxidation potential below that of oxygen evolution reaction (OER), thus requiring lower energy consumption when compared to OER. Advantageously, the MEA of the present invention can be used for oxidation of a variety of organic molecules with less cell voltage required for attaining an industrially acceptable current density such as 100 or 200 mA / cm2. It is to be understood that other molecules having an oxidation potential below OER may be used and the invention is not to be limited to the above recited embodiments.

[0031] The catholyte acid may be selected from an inorganic acid. The inorganic acid may be selected from H2SO4, HCI, HNO3, H3PO4 or mixtures thereof. In embodiments, the catholyte comprises H2SO4.

[0032] The anolyte alkali may be selected from KOH, NaOH, LiOH, CsOH, or mixtures thereof. In embodiments, the anolyte alkali comprises KOH. In one embodiment of the invention, the catholyte and anolyte are selected to result in a pH difference between the anolyte and catholyte. The pH difference (ApH) between the anolyte and catholyte may range from about 8 to about 16, from about 9 to about 16, from about 10 to about 16, from about 11 to about 16, from about 1 to about 16, from about 12 to about 15, from about 13 to about 15, from about 14 to about 15, or about 14. The ApH may preferably range from about 12 to about 15. In one embodiment, the ApH is 14. Preferably, the pH difference is sufficient to cause a negative shift of onset potential of alcohol oxidation and / or oxygen evolution reaction, which advantageously leads to reduced cell voltage required at current densities below 100 or 200mA / cm2. For example, the membrane electrode assembly as described herein can achieve a current density of up to 200 mA / cm2at a cell voltage of 0.377V or less, whereas membrane electrode assemblies having no pH difference between catholyte and anolyte may require a cell voltage around 0.83V to reach comparable current densities. Moreover, the pH difference between the anolyte and catholyte may suppress or eliminate crossover of products from anolyte to catholyte. In some embodiments, the crossover to catholyte for the membrane electrode assembly as described herein may be 4.1% or less at 10 mA cm2. The product crossover may also be completely suppressed at current densities of 50 mA crrr2or greater.

[0033] The pH difference between the anolyte and cathode may also direct the oxidation reaction at the anode selectively towards the formation of liquid C3 carbon products and not liquid C2, C1 products, or gaseous products such as CO2. In some embodiments, the membrane electrode assembly of the present invention may exhibit a high faradic efficiency (FE) of all liquid products from oxidation such as formic acid, oxalic acid, glyceric acid, etc. of up to 100%, and a high FE toward C3 products such as glyceric acid, tartronic acid, lactic acid, etc. of up to 76% at current densities of up to 500 mA crrr2. The FE towards C3 and all liquid products may also be maintained at least 55% and 84% even under a high current density of 1000 mA / cm2, respectively. Advantageously, the selective formation of these liquid products may provide desirable raw materials for use in pharmaceuticals, cosmetics, food industry, etc.

[0034] The hydroxide ion concentration or the concentration of the anolyte alkali may be in a range of from about 0.1 M to about 6 M, from about 0.1 M to about 5.5 M, from about 0.1 M to about 5 M, from about 0.1 M to about 4 M, from about 0.1 M to about 4.5 M, from about 0.1 M to about 3.5 M, from about 0.1 M to about 3 M, from about 0.1 M to about 2.5 M, from about 0.1 M to about 2 M, from about 0.1 M to about 1 .75 M, from about 0.1 M to about 1 .5 M, from about 0.1 M to about 1 .25 M, from about 0.1 M to about 1 M, from about 0.1 M to about 0.75 M, from about 0.1 M to about 0.5 M, from about 0.75 M to 1 M, from about 0.5 M to 1 M, from about 0.75 M to 1 .5 M, from about 0.1 M to about 0.25 M, from about 0.5 M to about 1 .5 M, from about 0.2 M to about 6 M, from about 0.25 M to about 6 M, from about 0.5 M to about 6 M, from about 0.75 M to about 6 M, from about 1 M to about 6 M, from about 1 .25 M to about 6 M, from about 1 .5 M to about 6 M, from about 2 M to about 6 M, from about 2.5 M to about 6 M, from about 3 M to about 6 M, from about 3.5 M to about 6 M, from about 4 M to about 6 M, from about 4.5 M to about 6 M, from about 5 M to about 6 M, from about 5.5 M to about 6 M; or at most about 6 M; or about 0.1 M, about 0.2 M, about 0.25 M, about 0.5 M, about 0.75 M, about 1 M, about 1 .25 M, about 1.5 M, about 1.75 M, about 2 M, about 2.5 M, about 3 M, about 3.5 M, about 4 M, about 4.5 M, about 5 M, about 5.5 M, about 6 M, or any ranges or values therebetween. The increase of the hydroxide ion concentration or the concentration of the anolyte alkali may be expected to increase the activities of the electrochemical reactions within the MEA and / or increase the power output when the MEA is used as an electricity generator. The hydroxide ion concentration or the concentration of the anolyte alkali may be selected based on the ion exchange capacity of the ionic exchange membrane. In embodiments, the concentration of the anolyte alkali is 1 M.

[0035] The anolyte comprising the alkali may have a pH in the range of from about 11 to about 14.6, from about 1 1 to about 14.5, from about 1 1 to about 14, from about 1 1 to about

[0036] 13.5, from about 1 1 to about 13, from about 11 to about 12.5, from about 11 to about 12, from about 1 1 to about 1 1.5; or from about 1 1 .5 to about 14.6, from about 12 to about

[0037] 14.6, from about 12.5 to about 14.6, from about 13 to about 14.6, from about 13.5 to about 14.6, from about 14 to about 14.6, from about 14.5 to about 14.6, from about 1 1 to about 14.5; or at most about 14.6; or about 1 1 , about 11 .5, about 12, about 12.5, about 13, about 13.5, about 14, about 14.5, about 14.6. In embodiments, the anolyte comprising the alkali may have a pH of 14.

[0038] The catholyte may contain a proton concentration in a range from about 0.1 M to about 6 M, from about 0.1 M to about 5.5 M, from about 0.1 M to about 5 M, from about 0.1 M to about 4 M, from about 0.1 M to about 4.5 M, from about 0.1 M to about 3.5 M, from about 0.1 M to about 3 M, from about 0.1 M to about 2.5 M, from about 0.1 M to about 2 M, from about 0.1 M to about 1.75 M, from about 0.1 M to about 1.5 M, from about 0.1 M to about 1.25 M, from about 0.1 M to about 1 M, from about 0.1 M to about 0.75 M, from about 0.1 M to about 0.5 M, from about 0.75 M to 1 M, from about 0.5 M to 1 M, from about 0.75 M to 1 .5 M, from about 0.1 M to about 0.25 M, from about 0.5 M to about 1 .5 M, from about 0.2 M to about 6 M, from about 0.25 M to about 6 M, from about 0.5 M to about 6 M, from about 0.75 M to about 6 M, from about 1 M to about 6 M, from about 1 .25 M to about 6 M, from about 1 .5 M to about 6 M, from about 2 M to about 6 M, from about 2.5 M to about 6 M, from about 3 M to about 6 M, from about 3.5 M to about 6 M, from about 4 M to about 6 M, from about 4.5 M to about 6 M, from about 5 M to about 6 M, from about 5.5 M to about 6 M; or at most about 6 M; or about 0.1 M, about 0.2 M, about 0.25 M, about 0.5 M, about 0.75 M, about 1 M, about 1 .25 M, about 1 .5 M, about 1 .75 M, about 2 M, about 2.5 M, about 3 M, about 3.5 M, about 4 M, about 4.5 M, about 5 M, about 5.5 M, about 6 M. The increase of the proton concentration or the concentration of the catholyte acid may be expected to increase the activities of the electrochemical reactions within the MEA and / or increase the power output when the MEA is used as an electricity generator. The proton concentration or the concentration of the catholyte acid may be selected based on the ion exchange capacity of the ionic exchange membrane. In embodiments, the proton concentration of the catholyte acid is 1 M.

[0039] The catholyte comprising the acid may have a pH in the range of from about -0.8 to about 1 .0, from about -0.8 to about 0.9, from about -0.8 to about 0.8, from about -0.8 to about 0.7, from about -0.8 to about 0.6, from about -0.8 to about 0.5, from about -0.8 to about 0.4, about -0.8 to about 0.3, about -0.8 to about 0.2, about -0.8 to about 0.1 , from about -0.8 to about 0, from about -0.78 to about 1 , from about -0.6 to about 1 , from about -0.5 to about 1 , from about -0.4 to about 1 , from about -0.3 to about 1 , from about -0.3 to about 0.5, from about -0.6 to about 0, from about -0.6 to about -0.5; from about -0.4 to about 0, from about -0.2 to about 0, from about 0 to about 1 , from about 0.5 to about 1 , from about -0.2 to about 0.5, from about -0.2 to about 0.8, from about -0.2 to about 0.2; about 0.78, about -0.6, about -0.5, about 0, about 0.5, or about 1. In embodiments, the catholyte comprising the acid has a pH of 0.

[0040] The anode catalyst and / or cathode catalyst which are coated on the electrical conductive layer of the cathode and / or anode compartment may independently comprise a metal selected from Pt, Pd or alloys thereof. The anode catalyst may be the same or different from the cathode catalyst. In embodiments, the catalysts may independently comprise from about 20 to about 80 wt.% of the metal and optionally one or more catalyst supports. The catalyst support may comprise materials suitable for dispersing the metal and / or for providing mechanical flexibility during MEA fabrication. In one embodiment, the catalyst support comprises carbon black, which may be selected from Ketjenblack, Vulcan XC- 72 or mixtures thereof.

[0041] The catalysts may further independently comprise a binder. In some embodiments, the binder is an ionomer, such as an anion exchange ionomer selected from poly(arylene ether sulfone) polymer backbone grafted with quaternary ammonium, sulfonated tetrafluoroethylene-based fluoropolymer-copolymer or mixtures thereof. Preferably, the ionomer may be selected from Fumion FAA-3-SOLUT-10 ionomer comprising poly(arylene ether sulfone) polymer backbone grafted with quaternary ammonium, Nation D-520 ionomer comprising sulfonated tetrafluoroethylene-based fluoropolymer- copolymer or mixtures thereof. The binder may be selected for binding the catalyst particles and preventing the catalyst from delamination when in use.

[0042] The MEA may further comprise a pair of bipolar plates disposed on either side of the MEA, wherein each bipolar plate is respectively configured to contact or electrically communicate with at least a surface of the cathode compartment or the anode compartment to thereby form a sandwich structure whereby the MEA is disposed between the bipolar plates. The bipolar plates may be configured to apply an electrical bias from the cathode compartment to the anode compartment. The bipolar plate may be composed of electrically conductive materials that are chemically resistant or inert to the anolyte and / or catholyte. In embodiments, the bipolar plate may be composed of titanium (Ti), iron (Fe) or alloys thereof. In another embodiment, the bipolar plate may comprise stainless steel. It is to be understood that any conductive plate which is capable of applying an electrical bias between the cathode and anode may be used in place of the bipolar plates.

[0043] The electrically conductive layer of the cathode and / or anode compartment may further comprise a gas diffusion layer (GDL). The GDL may be coated with the anode catalyst and / or the cathode catalyst. The GDL may have a porous structure to allow the distribution of reactants and removal of generated gases. For example, when hydrogen is generated at the cathode catalyst, the GDL may facilitate the passage of the evolved hydrogen gas from the catalyst site to a volume present in the cathode compartment to prevent accumulation of gases at the catalyst site. The GDL may also facilitate the transport of liquid reactants to the catalysts site for reaction. The GDL may also be electrically conductive to an extent that efficiently moves electrons from the bipolar plate to the catalyst.

[0044] In embodiments, the GDL may be composed of a material selected from the group consisting of carbon paper, carbon cloth, titanium foam, nickel foam, Pt-coated titanium foam, titanium felt, Pt-coated titanium felt, and nickel felt. Preferably, the GDL at the anode may be selected to be resistant to high anodic potentials. For example, the anode GDL may preferably comprise titanium foam, nickel foam, Pt-coated titanium foam, titanium felt, Pt-coated titanium felt, or nickel felt.

[0045] In some embodiments, the anolyte may comprise an alcohol concentration present in a range of from about 0.1 M to about 5M, from about 0.1 M to about 4M, from about 0.1 M to about 3M, from about 0.5M to about 5M, from about 0.5M to about 4M, from about 0.5M to about 3M, from about 0.5M to about 2M, from about 1 M to about 5M, from about 1 M to about 4M, from about 1 M to about 3M, from about 2M to about 3M, from about 3M to about 4M, from about 4M to about 5M, from about 3M to about 5M, from about 2M to about 5M, from about 0.1 M to about 1 M, from about 0.1 M to about 0.5M, from about 0.1 M to about 0.75M, from about 0.05 M to about 0.1 M, from about 0.1 M to about 0.25 M, from about 0.25 M to about 0.5 M, from about 0.5 M to about 0.75 M, from about 0.75 M to about 1 M, from about 0.1 M to about 0.15 M, from about 0.15 M to about 0.2 M, from about 0.2 M to about 0.3 M, from about 0.3 M to about 0.4 M, from about 0.4 M to about 0.5 M, from about 0.5 M to about 0.6 M, from about 0.6 M to about 0.7 M, from about 0.7 M to about 0.8 M, from about 0.8 M to about 0.9 M, from about 0.9 M to about 1 M, from about 0.2 M to about 0.25 M, from about 0.25 M to about 0.3 M, from about 0.3 M to about 0.35 M, from about 0.35 M to about 0.4 M, from about 0.4 M to about 0.45 M, from about 0.45 M to about 0.5 M, from about 0.5 M to about 0.55 M, from about 0.55 M to about 0.6 M, from about 0.6 M to about 0.65 M, from about 0.65 M to about 0.7 M, from about 0.7 M to about 0.75 M, from about 0.75 M to about 0.8 M, from about 0.8 M to about 0.85 M, from about 0.85 M to about 0.9 M, from about 0.9 M to about 0.95 M, or from about 0.95 M to about 1 M. Preferably, the anolyte may comprise no more than 1 M or no more than 2 M alcohol. The alcohol concentration in the anolyte may be chosen to be sufficient for oxidation while not impeding the mass transfer of hydroxide ions, oxidation products or blocking the active sites of the anode catalyst. In one embodiment, the anolyte may comprise a glycerol concentration of up to 5M.

[0046] The present disclosure further provides a method of producing hydrogen via the membrane electrode assembly as described herein, the method comprising: (a) electrochemically oxidizing the alcohol in the anode compartment in the presence of at least one alkali, and (b) concurrently electrochemically reducing protons in the catholyte housed in the cathode compartment to thereby evolve hydrogen gas. In some embodiments, step (a) is a glycerol oxidation reaction.

[0047] The protons in the catholyte may be produced by the dissociation of an acid in the catholyte. The acid is selected from H2SO4, HCI, HNO3, H3PO4, or mixtures thereof. In embodiments, the acid is H2SO4.

[0048] The electrochemical oxidation step (a) or the electrochemical reduction step (b) may be facilitated by an ion exchange membrane that is in fluid communication with both the catholyte and the anolyte.

[0049] The ion exchange membrane may be bipolar membrane, anionic exchange membrane, proton exchange membrane or combinations thereof. The ion exchange membrane may be anionic exchange membrane, proton exchange membrane or combinations thereof. Preferably, the ion exchange membrane comprises an anionic exchange membrane.

[0050] When the membrane electrode assembly is in use, the method may comprise conveying the anolyte comprising a mixture of the alcohol and alkali at a predetermined flow rate into the anode compartment which ensures that convection of the alcohol and alkali dominates over its diffusion. This chosen flow rate may advantageously prevent the formation of thickening diffusion layers within the anolyte compartment. In embodiments, the flow rate is from about 0.5 to about 10 ml / min. In preferred embodiments, the flow rate is no less than 5 ml / min. The method may further comprise conveying the catholyte comprising an acid at a flow rate which is identical to the flow rate of the anolyte into the catholyte compartment.

[0051] When the membrane electrode assembly is in operation, the alcohol at the anolyte consumes hydroxide ions and loses electrons to form the oxidized products while the protons at the catholyte obtain electrons and are reduced to hydrogen. For example, when the alcohol is glycerol, the following reactions may take place:

[0052] At the cathode:

[0053] CsHgOs + xOH~ -» oxidized products + ne"

[0054] The oxidation products may include a mixture of glyceraldehyde, glyceric acid, formic acid, oxalic acid etc., whereas x and n are integers corresponding to the number of electrons extracted from glycerol and oxidation level of glycerol.

[0055] At the anode:

[0056] 2H++ 2e“ — > H2

[0057] When replacing glycerol with other alcohols, the HER remains the same at the anode while at the cathode, the oxidation products may vary depending on the types of alcohols used.

[0058] The oxidation products which are discharged from the outlet of the anode compartment may be recovered for further processing. In embodiments, the discharged oxidation products may be treated with an acid to neutralize the same for further separation of the oxidation products from the anolyte. In one embodiment, the acid used for neutralization may be derived from the catholyte, which advantageously does not require the use of additional acids. The neutralized product stream may then be subject to distillation to separate and purify oxidation products.

[0059] The above reactions may be undertaken at a temperature no lower than 15 °C; or from about 20 °C to about 90 °C, from about 25°C to about 90°C, from about 20 °C to about 80 °C, from about 25°C to about 80°C, from about 20 °C to about 70 °C, from about 25°C to about 70°C, from about 20 °C to about 60 °C, from about 25 °C to about 60 °C, from about 20 °C to about 50 °C, from about 25 °C to about 50 °C, from about 20 °C to about 40 °C, from about 25 °C to about 40 °C, from about 20 °C to about 30 °C, from about 20 °C to about 25 °C, from about 25 °C to about 30 °C, from about 50 °C to about 90 °C, from about 60°C to about 90 °C, from about 30 °C to about 90 °C, from about 40 °C to about 90 °C, from about 50 °C to about 90 °C, from about 60 °C to about 90 °C, from about 70 °C to about 90 °C, from about 80 °C to about 90 °C, or at most about 90 °C. In some embodiments, the method is undertaken at about 80 °C.

[0060] Advantageously, increasing the reaction temperature may reduce the cell voltage required for hydrogen production. In one embodiment, the reactions are undertaken at a temperature from 25°C to 80°C, which is observed to decrease the cell voltage at 100 mA / cm2by around 65%.

[0061] Brief Description of Drawings

[0062] The accompanying drawings illustrate disclosed embodiments and serve to explain the principles of the disclosed embodiment. It is to be understood, however, that the drawings are designed for purposes of illustration only, and not as a definition of the limits of the invention.

[0063] Figure 1 shows a cross-sectional schematic illustration of the MEA for GOR and hydrogen production according to the present invention, wherein only anions or protons are allowed to cross the membrane when AEM or PEM is used. When BPM is used, water is dissociated into protons and hydroxide ions, catalyzed by TiOa, and the protons and hydroxide ions then enter the cathode and anode through GEM and AEM, respectively.

[0064] Figure 2 shows the GOR performance in an AEM-based alkali-alkali MEA electrolyzer.

[0065] Figure 2a shows the l-V curves of GOR in an AEM-based alkali-alkali MEA electrolyzer using 1 M KOH & 0.5 M glycerol as an anolyte and 1 M KOH as a catholyte at various flow rates of the anolyte and catholyte.

[0066] Figure 2b shows the l-V curves of OER in an AEM-based alkali-alkali MEA electrolyzer using 1 M KOH as an anolyte and GOR in an AEM-based alkali-alkali MEA electrolyzer using 1 M KOH with 0.1 M, 0.5 M, 1 M, 1 .5 M or 2 M glycerol as an anolyte and 1 M KOH as a catholyte with flow rate of 5 ml min1.

[0067] Figure 2c shows the total faradaic efficiencies of GOR toward liquid products in the anolyte and catholyte using chronopotentiometry tests at various current densities, in an AEM-based alkali-alkali MEA electrolyzer at 80 °C. Figure 2d is a comparison of GOR activities in MEA electrolyzer and H-cell measured using a two-electrode system at room temperature (RT) and 80 °C, the MEA and H-cell are both AEM-based alkali-alkali.

[0068] Figure 2e shows the LSV curves of HER on Pt / C in 1 M KOH with and without formic acid (FA) measured using three-electrode system at room temperature (RT), 85% system resistance is used for IR compensation.

[0069] Figure 2f is the in-situ Attenuated Total Reflectance-Fourier Transform Infrared spectroscopy (ATR-FTIR) of glycerol oxidation on Pt / C catalysts in 1 M KOH & 0.5 M glycerol solution with the applied potential range of 0.2 - 1.1 V vs. RHE.

[0070] Figure 2g is the in-situ Raman spectra of glycerol oxidation on Pt / C catalysts in 1 M KOH & 0.5 M glycerol solution with the applied potential range of 0.2 - 1.1 V vs. RHE. The terms “AEM-MEA, alkali-alkali” or “AEM alkali-alkali” used in these figures refer to AEM-based MEA electrolyzers using 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte.

[0071] Figure 3 shows the GOR performance of a PEM-based alkali-alkali / acid-alkali MEA electrolyzer.

[0072] Figure 3a is a plot of the l-V curves of the GOR in a PEM-based MEA electrolyzer using 1 M KOH and 1 M glycerol as an anolyte and 1 M KOH or 0.5 M H2SO4as a catholyte. The performance is compared with the l-V curve of a PEM-based MEA electrolyzer using 1 M KOH as an anolyte and 1 M KOH or 0.5 M H2SO4 as a catholyte.

[0073] Figure 3b shows the total faradaic efficiencies of GOR toward liquid products in the anolyte and catholyte using chronopotentiometry tests at various current densities in a PEM-based acid-alkali MEA electrolyzer at 80 °C.

[0074] Figure 3c is the LSV curve of GOR in a H- cell using 1 M KOH & 1 M glycerol anolyte with 1 M KOH or 0.5 M H2SO4 catholyte at room temperature.

[0075] Figure 3d is a Bode plot of GOR in a H-cell using 1 M KOH & 1 M glycerol as an anolyte with 1 M KOH as a catholyte at room temperature.

[0076] Figure 3e is a Bode plot of GOR in a H-cell using 1 M KOH & 1 M glycerol as an anolyte with 0.5 M H2SO4 as a catholyte at room temperature.

[0077] Figure 3f is a schematic representation of the PEM-based alkali-alkali and acid-alkali MEA electrolyzers. PEM-based MEA electrolyzers using 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte are denoted as “PEM-MEA, alkali-alkali”, while PEM-based MEA electrolyzers using 1 M KOH & 1 M glycerol anolyte and 0.5 M H2SO4 catholyte are denoted as “PEM-MEA, acid-alkali”. Similarly, H-cells using 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte are labelled as “H-cell alkali-alkali”, and H-cells using 1 M KOH & 1 M glycerol anolyte and 0.5 M H2SO4 catholyte are labelled as “H-cell acid- alkali”.

[0078] Figure 4 shows the GOR performance in an AEM-based alkali / acid-alkali MEA electrolyzer.

[0079] Figure 4a shows the l-V curves of GOR in an AEM-based MEA electrolyzer with 1 M KOH & 1 M glycerol anolyte and 1 M KOH or 0.5 M H2SO4 catholyte.

[0080] Figure 4b shows the total faradaic efficiencies of GOR toward liquid products in anolyte and catholyte using chronopotentiometry tests at various current densities in an AEM- based acid-alkali MEA electrolyzer.

[0081] Figure 4c compares the product crossover ratios of acid-alkaline and alkali-alkali AEM- based MEA electrolyzers.

[0082] Figure 4d shows the total Faradaic Efficiency (FE) towards C3 products in AEM-based alkali-alkali and acid-alkali MEA systems at 80 °C.

[0083] Figure 4e shows the FE towards liquid products in AEM-based alkali-alkali and acid- alkali MEA systems at 80 °C.

[0084] A comparison of cell voltages of electrolyzers which couple GOR in alkaline media with HER (in alkaline media, unless otherwise denoted. HER in acidic media is denoted as “acid-alkali”) at the current density of 100 mA cm2using various catalysts is shown in Figure 4f.

[0085] Figure 4g shows the molar selectivity towards C3 products using the electrolyzers described herein, as compared to the molar selectivity for C3 products using other systems known in the art. AEM-based MEA electrolyzers using 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte are denoted as “AEM-MEA, alkali-alkali”, while AEM- based MEA electrolyzers using 1 M KOH & 1 M glycerol anolyte and 0.5 M H2SO4 catholyte are denoted as “AEM-MEA, acid-alkali”.

[0086] Figure 4h shows schematic illustrations of acid-alkaline and alkali-alkali AEM-based MEA electrolyzers. Figure 5a shows a graphical illustration of AEM-based alkali-alkali MEA GOR-coupled hydrogen production.

[0087] Figure 5b shows a graphical illustration of PEM-based alkali-alkali MEA GOR-coupled hydrogen production.

[0088] Figure 5c shows a graphical illustration of PEM-based acid-alkali GOR-coupled hydrogen production.

[0089] Figure 5d shows a graphical illustration of BPM-based acid-alkali GOR-coupled hydrogen production.

[0090] Figure 5e shows a graphical illustration of AEM-based acid-alkali MEA for GOR-coupled hydrogen production.

[0091] Figure 5f shows the corresponding GOR l-V curves of figures 5a, b, c, d and e, respectively. MEA using 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte are denoted as “alkali-alkali MEA”. MEA using 1 M KOH & 1 M glycerol anolyte and 0.5 M H2SO4 catholyte are denoted as “acid-alkali MEA”.

[0092] Figure 6 shows the GOR performance in H-cells.

[0093] Figure 6a shows a linear sweep voltammetry (LSV) of GOR on Pt / C in anolytes having 1 M or 2M KOH with 0.5M or 1 M glycerol as well as the LSV of OER using 1 M KOH only as an anolyte in a H-cell measured by a three-electrode system. The catholyte is KOH having the same concentration as the anolyte KOH. The ion exchange membrane is Nation 212.

[0094] Figure 6b shows a graph of l-V curves of GOR on Pt / C in an alkali-alkali and acid-alkali H-cell with various electrolyte concentrations measured by a two-electrode system. The alkali-alkali H-cells use 1 M or 2M KOH with 1 M glycerol as anolyte and 1 M or 2M KOH as catholyte. The acid-alkali H-cells use 1 M or 2M KOH with 1 M glycerol as anolyte and 0.5M, 1 M or 2M H2SO4 as catholyte.

[0095] Figures 6c to 6e show the cell performance as an electricity generator. Figure 6c show the polarization and power density curves of an acid-alkali H-cell having 2 M KOH and 1 M glycerol as anolyte and 1 M H2SO4 as catholyte.

[0096] Figure 6d shows the polarization and power density curves of an acid-alkali H-cell having 2 M KOH and 1 M glycerol as anolyte and 2 M H2SO4 as catholyte. Figure 6e shows the polarization and power density curves of an acid-alkali H-cell having 1 M KOH and 1 M glycerol as anolyte and 0.5 M H2SO4 as catholyte.

[0097] Figure 6f shows the comparison of the l-V curves of GOR in acid-alkali and alkali-alkali H-cells and AEM-based electrolyzers (at room temperature and 80 °C).

[0098] Figure 6g shows l-V curves at RT using two-electrode system in an H-cell having 1 M KOH & 1 M glycerol anolyte with 1 M KOH or 0.5 M H2SO4 catholyte. Figure 6h shows the Tafel curves of GOR in the H-cell of figure 6g. Figure 6i shows LSV curves of GOR in a three-electrode system in an H-cell having 1 M KOH & 1 M glycerol as anolyte and 1 M KOH as catholyte measured by (85% system resistance used for iR compensation). Figure 6j shows the Tafel plots of the HER in 1 M KOH catholyte of the H-cell of figure 6i (85% system resistance used for iR compensation).

[0099] Figure 7 shows l-V curves of GOR in an AEM-based MEA having 1 M KOH & 1 M glycerol as anolyte and 1 M KOH as catholyte in which both catholyte and anolyte were conveyed in a flow rate of 5 ml min-1at 80 °C, the catalyst on both cathode and anode is either 40 wt% Pt on high surface area Ketjenblack or on Vulcan XC-72. The 40 wt% Pt on Ketjenblack shows higher GOR activity compared to Vulcan XC-72.

[0100] Figure 8a shows l-V curves of GOR in AEM- and PEM-based MEAs having 1 M KOH & 1 M glycerol as anolyte and 1 M KOH or 0.5 M H2SO4 as catholyte in at 80 °C.

[0101] Figure 8b shows Tafel curves of GOR in the MEAs of figure 8a.

[0102] Figure 9 shows the performance of the cell as an electricity generator. Figure 9a is a plot of the polarization curve and power density curve of a PEM-based acid-alkali MEA electrolyzer at 80 °C. Figure 9b is a plot of the polarization curve and power density curve of an AEM-based acid-alkali MEA electrolyzer at 80 °C. PEM and AEM-based MEA electrolyzers using 1 M KOH & 1 M glycerol anolyte and 0.5 M H2SO4 catholyte are denoted herein as “PEM acid-alkali” and “AEM acid-alkali”, respectively.

[0103] Figure 10 shows HPLC curves of catholyte and anolyte of after GOR in AEM alkali-alkali, AEM acid-alkali, and PEM acid-alkali cells having 1 M KOH & 1 M GOR anolyte, 1 M KOH or 0.5 M H2SO4 as catholyte pumped into the cell at 5 ml min1at 80 °C.

[0104] Figure 11a shows Faradic Efficiency (FE) of each GOR product in an AEM-based alkaline-alkaline MEA having 1 M KOH & 1 M glycerol anolyte and 1 M KOH as catholyte, wherein the anolyte and catholyte were both pumped at 5 ml min-1at 80 °C. Figure 11b shows Faradic efficiency (FE) of each GOR product in PEM-based acid- alkaline MEA having 1 M KOH & 1 M glycerol anolyte and 0.5M H2SO4 as catholyte wherein the anolyte and catholyte were both pumped at 5 ml min-1at 80 °C.

[0105] Figure 11c shows Faradic efficiency (FE) of each GOR product in AEM-based acid- alkaline MEA having 1 M KOH & 1 M glycerol anolyte and 0.5M H2SO4 as catholyte and both the anolyte and catholyte were pumped at 5 ml min1at 80 °C (OA: oxalate; TA: tartronate; GLA: glycerate; GA: glycolate; LA: lactate; FA: formate; AA: acetate).

[0106] Figure 12 shows concentrations of C3 (TA, GLA and LA) crossed from anolyte to catholyte at open circuit voltage (OCV) with time in AEM-based alkali-alkali and acid- alkali MEAs having 1 M KOH & 1 M glycerol & 0.1 M potassium tartronate & 0.1 M potassium glycerate & 0.1 M potassium lactate as the anolyte and 1 M KOH or 0.5 M H2SO4 as the catholyte, both anolyte and catholyte were pumped at 5 ml min1at 80 °C. The volume of anolyte / catholyte was 50 ml, and the electrolytes were pumped and cycled to measure the effects of increased anolyte-catholyte pH difference on the product anion crossover rate to support the hypothesis that increased hydroxide crossover rate could suppress the product anion crossover.

[0107] Figure 13a shows a graph measuring the product FEs at current densities of 200 mA cm2over a period of 2 hours in an AEM-based acid-alkali MEA electrolyzer having 1 M KOH & 0.5 M glycerol as the anolyte and 0.5 M H2SO4 as the catholyte, which were both pumped at 5 ml min1at 80 °C.

[0108] Figure 13b shows the change of glycerol conversion ratio and pH difference between anolyte and catholyte over a period of 2 hours in AEM-based acid-alkali MEA electrolyzer with 1 M KOH & 0.5 M glycerol anolyte and 0.5 M H2SO4 catholyte pumped at 5 ml min'1at 80 °C.

[0109] Figure 13c shows the change of pH difference between anolyte and catholyte under OCV and 200 mA cm-2in an AEM-based acid-alkali MEA electrolyzer with 1 M KOH & 0.5 M glycerol anolyte and 0.5 M H2SO4catholyte pumped at 5 ml min-1at 80 °C.

[0110] Figure 13d shows the sulfate concentration in anolyte via crossover under OCV and 200 mA cm2with time in AEM-based acid-alkali MEA electrolyzer with 1 M KOH & 0.5 M glycerol anolyte and 0.5 M H2SO4 catholyte pumped at 5 ml min1at 80 °C. The volume of anolyte / catholyte was 50 ml, and the electrolytes were pumped and cycled to measure the change in product selectivity, pH and activity with consumption of glycerol. Figure 14a shows the stability of the AEM-based acid-alkali MEA electrolyzer having an anolyte of 1 M KOH and 1 M glycerol and 0.5 M H2SO4 catholyte pumped at 5 ml min1at 80 °C as a hydrogen production electricity generator at current densities of 10 mA cm'2, 25 mA cm2, 50 mA cm2, 75 mA cm2, 100 mA cm2; while Figure 14b illustrates the stability of the AEM-based acid-alkali MEA electrolyzer with an anolyte of 1 M KOH and 1 M glycerol and 0.5 M H2SO4 catholyte pumped at 5 ml min1at 80 °C at current densities of 100 mA crrr2, 200 mA cm-2, 500 mA cm'2and 1000 mA cm'2.

[0111] Figures 14c to 14f show the corresponding faradaic efficiencies (FEs) towards various liquid products (tartronate acid (TA), glycolic acid (GA), formic acid (FA), oxalic acid (OA), glyceric acid (GLA), lactic acid (LA), acetate (AA)) and the total FE values in an AEM- based acid-alkali MEA electrolyzer with an anolyte of 1 M KOH and 1 M glycerol and a catholyte of 0.5 M H2SO4 pumped at 5 ml min1at 80 °C;

[0112] Figure 14c shows the FEs toward liquid products of the AEM-based acid-alkali MEA electrolyzer at a current density of 100 mA cm2.

[0113] Figure 14d shows the FEs toward liquid products of the AEM-based acid-alkali MEA electrolyzer at a current density of 200 mA cm'2

[0114] Figure 14e shows the FEs toward liquid products of the AEM-based acid-alkali MEA electrolyzer at a current density of 500 mA cm2

[0115] Figure 14f shows the FEs toward liquid products in the AEM-based acid-alkali MEA electrolyzer at a current density of 1000 mA cm2.

[0116] Figure 15a shows the l-V curves of electrochemical oxidation of methanol (MeOH) in alkali-alkali and acid-alkali MEAs.

[0117] Figure 15b shows polarization and power density curves of an acid-alkali MEA having 1 M KOH and 1 M MeOH as anolyte and 0.5M H2SO4 catholyte as an electricity generator using MeOH as fuels at 80 °C.

[0118] Figure 15c shows the l-V curves of electrochemical oxidation of ethanol (EtOH) in alkalialkali and acid-alkali MEA.

[0119] Figure 15d shows polarization and power density curves of an acid-alkali MEA having 1 M KOH and 1 M EtOH as anolyte and 0.5M H2SO4 catholyte as an electricity generator using EtOH as fuels at 80 °C. Figure 15e shows the l-V curves of electrochemical oxidation of ethylene glycol (EG) in alkali-alkali and acid-alkali MEA.

[0120] Figure 15f shows polarization and power density curves of an AEM-based acid-alkali MEA having 1 M KOH and 1 M EG as anolyte and 0.5M H2SO4 catholyte as an electricity generator using EG as fuels at 80 °C.

[0121] Figure 16a shows the optimized adsorption geometry of species before and after C-C bond cleavage on Pt (111 ) surface (CH2OHC*OC*Oads — > CH2OHC*Oads + COads).

[0122] Figure 16b is a plot of the calculated free energy change for the C-C bond cleavage at 0.3V, 0.5 V, 0.7 V and 0.9V vs. RHE against pH change based on free energy values of step 1 & 2, as shown in Figure 16c, and the free energy changes at various potentials vs. SHE, as shown in Figure 16d.

[0123] Examples

[0124] The invention will be further described in greater detail by reference to specific nonlimiting Examples, which should not be construed as in any way limiting the scope of the invention.

[0125] Materials

[0126] The Toray 090 carbon paper, FAA-3 ionomer solution, Nation D-520 solution and Nation 212 PEM were purchased from Fuel Cell Store, US. The Ni fiber paper and Sustainion X37-50 Grade T AEM were purchased from Dioxide Materials, US. The 40% Pt on high surface area Ketjenblack was purchased from Premetek, US. The SS316I bipolar plates were fabricated by Axis Synergy, Singapore, and the Ti bipolar plates were fabricated by Baoji Yinggao, China. All other chemicals were purchased from Sigma Aldrich and used without purification. Deionized water with resistance of 18.2 MQ cm was used for all experiments.

[0127] Preparation of H-shaped electrochemical cell

[0128] The electrochemical glycerol oxidation coupling (EOR) with hydrogen evolution reaction (HER) was tested in a H-shaped cell (H-cell) with two chambers separated by an alkaline- pretreated Nation 212 membrane (The membrane was immersed in the 1 M KOH solution overnight before the test.). The two-electrode setup employed a glassy carbon electrode (5 mm diameter) dropped cast with 1 mgpterm240 wt% Pt / Ketjenblack (from catalyst ink of 5 mg ml’140 wt% Pt / Ketjenblack in IPA : deionized water : 5 wt% Nation ionomer solution = 4 : 1 : 0.0486 (volume ratio) to form 10 wt% Nation ionomer in the catalyst layer) and Pt plates (2 cm * 2 cm) as the working and counter electrode, respectively. The three-electrode setup used an additional Hg / HgO (1 M KOH) reference electrode.

[0129] The anode and cathode chambers both with alkaline electrolyte were tested with both three-electrode and two-electrode systems (1 & 2 M KOH with 0.5 & 1 M glycerol anolyte, 1 & 2 M KOH catholyte) to measure both the catalyst activity and whole cell voltage (Figures 6a and 6b).

[0130] The alkaline anolyte (1 & 2 M KOH with 0.5 & 1 M glycerol) and acidic catholyte (0.5 & 1 M H2SO4) tests were also measured with the two-electrode H-cell system to study the influence of the electrochemical neutralization energy (ENE) on the whole cell voltage (Figures 6c to 6f).

[0131] For the HER tests (figure 6j), a glassy carbon electrode (5 mm diameter) dropped cast with 0.2 mgpt cm240 wt% Pt / Ketjenblack (from catalyst ink of 5 mg ml-140 wt% Pt / Ketjenblack in IPA : deionized water : 5 wt% Nation ionomer solution provided at a volume ratio of 4 : 1 : 0.0486, to form 10 wt% Nation ionomer in the catalyst layer), Hg / HgO (1 M KOH) electrode. A Ni mesh (2 cm * 2 cm) was employed with the working, reference and counter electrodes, respectively.

[0132] Preparation of membrane electrode assembly

[0133] 50 mg of 40wt% Pt / C catalyst with 15 mL isopropanol, 5 mL deionized water and 0.24 mL 10wt% FAA-3 ionomer solution was mixed and ultrasonicated for about 1 hour for preparation of the catalyst ink. Nation D-520 solution was used at a concentration of 5 wt.% in place of FAA-3-ionomer when acidic catholyte is used. The catalyst ink was airsprayed onto a nickel foam with a Pt loading of 2 mg / cm2to provide an anode. The catalyst ink was air sprayed onto carbon paper with a Pt loading of 1 mg / cm2to provide a cathode. The AEM was immersed in 1 M KOH solution for 24 hours to convert it to hydroxide form. The MEA was fabricated using a sandwich arrangement of cathode AEM | anode. Stainless steel 3161 blocks with a single serpentine channel were used as bipolar plates (titanium bipolar plate was used when acid electrolyte was used), in which two bipolar plate are disposed on either side of the MEA to sandwich the cathode | AEM anode structure and the serpentine channel is configured to convey the flows of catholyte and anolyte into the anode and cathode compartments, respectively. Viton gaskets with suitable thickness were also placed to prevent the liquid / gas from leaking. A torque was applied to the bipolar plate | cathode | AEM | anode | bipolar plate to assemble the cell at 8.5 Nm and the active area of the MEA was 4 cm2.

[0134] Table 1. Examples of MEAs In-situ ATR-FTIR tests

[0135] The in-situ ATR-FTIR tests were conducted using a Bruker Vertex 80 connected with an electrochemical VeeMax III apparatus from PIKE. A mercury-cadmium-telluride (MCT) detector with liquid nitrogen cooling was used. A Si prism evaporated with a 5 nm Ti layer and a 25 nm Au layer by an HHV 306 e-beam metal evaporator was used to reflect the signal at 60-degree angle. The catalysts were drop-casted onto the surface of the prism to form a catalyst layer of about 0.3 mgptcm240 wt% Pt / 'C and used as working electrodes. An Ag / AgCI (saturated KCI) electrode and a graphite rod were used as reference and counter electrodes, respectively. An electrolyte of 1 M KOH + 0.5 M glycerol was used. The in-situ signals under chronoamperometry tests at different potentials were captured.

[0136] In-situ Raman

[0137] The in-situ Raman tests were carried out using an i-Raman Plus 785H Raman spectrometer with a laser wavelength of 785 nm. The catalysts were drop-casted onto a glassy carbon electrode as the working electrode (1 mgPtcm240 wt% Pt / C), of which the plane was set to perpendicular to the incident laser. An Ag / AgCI (saturated KCI) and a Pt wire were used as reference and counter electrodes, respectively. The anolyte and catholyte was separated by a Nation 212 membrane. An anolyte of 1 M KOH + 0.5 M glycerol and catholyte of 1 M KOH were used. The spectrum was acquired under chronoamperometry tests at different potentials.

[0138] MEA cell tests

[0139] The cell temperature was maintained at 80 °C by two electric heating plates and measured by a thermocouple placed beside the two bipolar plates. The electrolyte was pumped into the cell by peristaltic pump and preheated in a heating coil before being pumped into the cell. The electrochemical tests were conducted with an Autolab potentiostat with a 20-A booster (Metrohm). The l-V tests were conducted from -0.6 to 1 .3 V with a scan rate of 10 mV s1to avoid serious carbon corrosion. The polarization curves were obtained by running chronopotentiometry test at various current densities for 10 min and the outlet product electrolyte was collected and later analyzed by high performance liquid chromatography (HPLC). All experiments were conducted at least three times to draw the error bars to show reliability and repeatability.

[0140] Product analysis

[0141] Chromatographic determination of glycerol oxidation products was carried out by an Agilent 1260 Infinity II HPLC (Agilent Technologies). The column used was an Aminex HPX87-H (Bio-Rad) and the eluent used was 5 mM sulfuric acid. During the test, 25 pL pre-mixed solution of 0.6 ml 0.5 M H2SO4 and 0.5 ml sample solution were injected into the column and the temperature polar of the column was kept at 60 °C. The flow rate was 0.5 mL / min. The separated compounds were then detected with a refractive index detector (RID) and a multiple wavelength detector (MWD). The expected products were also analyzed by HPLC to perform a standard calibration curve. Ion-chromatography (IQ measurement

[0142] The anolyte in the sulfate crossover tests was collected and diluted for IC measurement. The IC tests were performed by an ICS-1000 Ion Chromatography System using a Dionex lonPac AS11 with EGC III KOH as the eluent. Faradaic Efficiency (FE) calculations

[0143] The FE of GOR products is calculated based on the following equation: where n is the number of electrons transferred to produce the product, C is the concentration of the product (M), V is the volume of the electrolyte (L), F is the Faradaic constant (96485 C mol'1), and Q is the total charge passed during the electrolysis (C).

[0144] The value for n for each reaction is calculated based on the reaction pathways for glycerol oxidation in alkaline solution as shown in the scheme below (Reaction pathways for glycerol oxidation) in alkaline solution and listed in Table 2.

[0145] Table 2. The number of electron transfer for each GOR liquid product (product in salt form as GOR was conducted in alkaline electrolyte in this work)

[0146] Product Reaction n

[0147] Tartronate CH2OHCHOHCH2OH 10OH — OOCCHOHCOO + 8H2O 8

[0148] (TA) + 8e_

[0149] Glycolate CH2OHCHOHCH2OH 6.5OH' — > 1.5CH2OHCOO- + 5H2O 10 / 3 (GA) + 5e

[0150] Oxalate CH2OHCHOHCH2OH 14OH' — 1 .5 OOCCOO' + 1 1 H2O + 22 / 3

[0151] (OA) 11 e

[0152] Energy consumption calculation for electrochemical hydrogen production

[0153] The energy to obtain hydrogen is the multiple of the charge transfer (Q) and the cell voltage (UECO))- Hence, the energy consumption can be expressed as: where, 144 is energy consumption for hydrogen production (kWh (kg H2)'1) and UECis the electrochemical cell voltage (V). To produce 1 kg hydrogen, Q = n x F x 1000 / MH2(Mwjs the molar mass of hydrogen (2 g mol'1); F is the Faradaic constant (96485 C mol'1)). The equation to obtain 1 kg of hydrogen can therefore be expressed as:

[0154] Calculation of the GOR standard potential

[0155] For the glycerol oxidation reaction to formic acid (take formic acid but not carbon dioxide as an example product due to the negligible FE towards carbon dioxide at lower potentials), the anode and cathode reactions can be expressed as:

[0156] Anode: CH2OHCHOHCH2OH (1) + 5 H2O (1) -> 3 HCOOH (1) + 8 H++ 8 e"

[0157] Cathode: 8 H++ 8 e“ 4 H2(g~)

[0158] Overall: CH2OHCHOHCH2OH (1) - 3 ll2O (I) 3 HCOOH (1) + 4 H2(#)

[0159] At 298 K and 1 atm, the Gibbs free energy can be express as: where n is the number of electrons transferred (n = 2), and F is the Faraday constant (96485 C mol1). The standard potential for glycerol oxidation to formic acid at 298 K and 1 atm is therefore 0.0004 V. Meanwhile, the standard potential for oxygen evolution reaction (OER) is 1 .229 V.

[0160] The standard potentials vs. SHE (standard hydrogen electrode) of OER, GOR and HER reactions at 298 K and 1 atm are: where R is the universal gas constant, V is the temperature in Kelvins, n is the number of electrons transferred, and FIs the Faraday constant; pH refers to the pH of the solution where the reaction takes place; p^Jp0refers to the partial pressure of hydrogen relative to standard atmospheric pressure (i.e. 101.325kPa); p0., / parefers to the partial pressure of oxygen relative to standard atmospheric pressure.

[0161] Consequently, for the electrochemical water splitting, when pH values of anolyte and catholyte are equal, the theoretical potential of OER is 1 .229 V. However, if the pH values of anolyte and catholyte are 14 and 0, respectively, the theoretical potential of OER is reduced to only 0.400 V. The same theoretical potential reduction assisted by ENE can also be applied to the GOR-coupled HER, as the glycerol oxidation reaction is also a proton-coupled electron transfer reaction, and the reduced cell voltage can be up to 829 mV theoretically. In addition, according to the Nernst equation, there is a positive relationship between the ENE and temperature. Therefore, the reduction in the cell voltage via ENE harvesting is theoretically more significant at higher temperatures.

[0162] Mass transport in the MEA electrolyzer

[0163] The current density, ji tdepends on the flux of the species i:

[0164] (S5) where A(x) is the flux of species i at distance x from surface (mol s1cm2). The mass transport in the MEA electrolyzer is composed of diffusion, migration and convection. The total flux can be expressed by the Nernst-Planck equation: where Dtis the diffusion coefficient (cm2s~1),dc^ is the concentration gradient along distance x, Zj and C, are the charge (dimensionless) and concentration (mol cm3) of species is the potential gradient along distrance x, v(x) is the velocity profile of the solution. (-£), diffusion; migration; Qv(x): convection).

[0165] Example 1 - GOR in the AEM-based alkali-alkali MEA

[0166] Glycerol oxidation reaction (GOR) was investigated in the alkali-alkali AEM-based MEA electrolyzer using an anolyte of 1 M KOH with various concentrations of glycerol, and a catholyte with 1 M KOH. The reaction was carried out at varying flow rates at both chambers.

[0167] Figure 2a shows the l-V curves of GOR in 1 M KOH & 0.5 M glycerol with various flow rates of catholyte and anolyte in an AEM electrolyzer at 80 °C. Current density around 0.5 V increases as the flow rate rises from 0.5 to 5 ml min1. As the flow rate increases, convection becomes more pronounced, leading to augmented overall mass transfer of the reactants according to the Nernst-Planck equation (S6). As a result, the corresponding current density, which is proportional to the flux of electro-active species (Equation (S5)), is elevated, especially at the higher current density region. Moreover, a noticeable drop in current density was observed after the peak current density is achieved at around 0.5 V for flow rates of up to 2 ml min1. This phenomenon is postulated to arise from the thicker diffusion layer induced by polarization, as a result from the dominant diffusion limitation and lack of convection. The thick diffusion layer leads to a decrease in the diffusion layer concentration gradient and thus reduce the mass transfer.

[0168] However, this current density drop was absent when the flow rate is maintained at 5 ml min1or higher, ensuring robust convection that prevents the diffusion layer from thickening significantly. It should also be noted that a positive shift in the curve occurs at lower current density region (up to 0.4 A cm2) when the flow rate increased from 5 to 10 ml min-1. A possible explanation is that the increased flow rate leads to enhanced convective heat transfer, thus lowering the temperature. As a result, the ionic conductivity decreased, compromising the current density. Meanwhile, the peak current density at 1 .0 V increased due to the enhanced convection, as the mass transfer becomes more dominant. Therefore, a flow rate of 5 ml min1was employed for the subsequent tests.

[0169] Figure 2b shows the l-V curves of GOR in 1 M KOH with glycerol concentrations of 0.1 M, 0.5 M, 1 M, 1.5 M and 2 M, conducted at a flow rate of 5 ml min1. Significant peak current density increment at about 0.5 V can be observed when the glycerol concentration increased from 0.1 to 1 M. This suggests that the mass transfer of glycerol from bulk to the catalyst governs the reaction rate. Conversely, when the glycerol concentration is more than 1 M, the change in current density became minimal. This suggests that the reaction kinetics is the limiting factor when the glycerol concentration is more than 1 M. A slight increment in the onset potential was observed with increasing glycerol concentration, which could be attributed to the mass transfer of all relevant species (glycerol, hydroxyl, and the formed oxidation products that must diffuse away from the electrode) being hampered at high glycerol concentration. This issue arises from the heightened viscosity and a potential overabundance of glycerol molecules covering the catalyst's active sites. Hence, 1 M KOH & 1 M glycerol as anolyte was used for the subsequent tests.

[0170] Under these optimized operation conditions (i.e. 1 M KOH & 1 M glycerol as anolyte, 1 M KOH as catholyte, both at a flow rate of 5 ml min1), a current density of 0.53 A cm2was obtained at a cell voltage of 0.67 V. The optimization in device operation reduces over 60% of the energy consumption for hydrogen production, as compared with majority of AEM-based water electrolysis at the same current density.

[0171] Figure 2d shows a comparison between the whole cell voltage of GOR in a H-cell and an AEM-based MEA electrolyzer, both with alkali-alkali electrolyte. The efficiency improvement is significant, as evidenced by the reduction in the whole cell voltage to 0.65 V in an AEM-based MEA from 1.19V in H-cell at 100 mA cm2at room temperature (RT), and an increment in temperature to 80°C further reduced the cell voltage to 0.28 V.

[0172] The products of GOR in an AEM-based MEA electrolyzer were analyzed, using 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte across various current densities. As shown in Figure 2c, a cell voltage of only 0.49 V was required to achieve a stable industrial-level current density of 200 mA cm-2. Notably, a substantial liquid product crossover effect was discernible in the catholyte in this system, with the maximum total product faradaic efficiency (FE) in the catholyte reaching 30%. This outcome arises from the inevitable concentration gradient-driven diffusion through the AEM. The collection of products in catholyte can increase separation / purification cost of products, and the existence of products in catholyte can be detrimental for the HER at cathode. For example, Figure 2e shows that the addition of formic acid (i.e. one of the GOR liquid products) dramatically decreased the HER activity on Pt / C catalyst.

[0173] Even though the crossover of product / intermediate anions from catholyte to anolyte have been widely investigated, and the crossover has even been utilized for carbon capture. The crossover of anodic product anions to catholyte in MEA electrolyzers are often overlooked. Different from the crossover from catholyte to anolyte, the anion transfer from anolyte to catholyte is diffusion dominated since the electromigration inherently promotes anion transfer from cathode to anode. Consequently, the product crossover ratio increased from 28% at 10 mA crrr2to 30% at 25 mA cm-2 due to the increase of product concentration gradient along the AEM.

[0174] Product crossover was slightly attenuated when the current density / cell voltage is further increased, from ~ 30% at 10 mA cm-2to ~ 17% at 200 mA cm'2, and further declined to 4.3% at 400 mA cm2. This is because the electromigration tends to move the anions from cathode to anode, which can offset partial diffusion transfer of anions via AEM, and eventually reduce the product crossover with increasing current density / cell voltage. Despite the existence of the liquid crossover effect, the sum of the anolyte and catholyte FEs was close to 100% at the current densities up to 200 mA cm2. This finding indicates exclusive liquid products generated via GOR, without carbon dioxide or oxygen gas production. When the current density exceeds 200 mA cm'2but remains below 400 mA cm2, the applied cell voltage has yet to initiate the OER, which indicates the loss of total liquid product FE was attributed to the emergence of undesirable CO2 production. With a further increase in current density, the cell voltage enters the OER region, leading to further decrease in total liquid product FE to below 30%. To better understand the GOR mechanism on the Pt / C catalyst in alkaline, the in-situ attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR) was conducted in the potential window of 0.2 - 1 .1 V vs. RHE in 1 M KOH & 0.5 M glycerol. As shown in Figure 2f , the band at ca. 2043 cm1is assigned to the adsorption of linearly bonded COads- The CO band is only found in the potential range of 0.2 - 0.8 V vs. RHE, this suggests the C-C bond cleavage taking place from glycerol dissociative adsorption, and thus results in the accumulation of catalyst poisoning. The poisoning effect, in addition to the mass transfer limitation mentioned above, may lead to the current drop or plateau in the positive l-V scan after the first peak at ca. 0.5 V. Meanwhile, the absence of COads band at higher potentials suggests the stripping of COadS, and thus a CO oxidation peak was observed at ~ 0.8 - 1 V in the voltammogram.

[0175] Moreover, the CO2, the oxidation product of COads, is produced, which is consistent with the HPLC results where significant CO2 production can be measured when the cell voltage reaches ca. 0.8 V. Additionally, the band at approximately 1040 and 1106 cm-1is attributed to the v(C-O) of glycerol. The band at ca. 1081 cm-1is related to the v(C-O) of two C3 aldehyde / ketone intermediates, glyceraldehyde and dihydroxyacetone, which are not stable in alkaline and thus are not detected in HPLC measurement. The band located at ca. 1330 cm1is assigned to the symmetric v(O-C- O) of tartronate and dihydroxyacetone. The band at ca. 1225 and 1351 cm-1is assigned to the v(C-O) and symmetric v(O-C-O) of hydroxypyruvate, which is not detected in HPLC measurement most likely due to the low selectivity towards it.

[0176] The other C3 product glycerate contributed to three bands, the symmetric v(O-C-O) at ca. 1380 cm-1, the asymmetric v(O-C-O) at ca. 1433 cm1. The intensive band around 1589 cm-1cannot be assigned to a single product because most of carboxylate products possess carboxylic group band in this region. Meanwhile, the band at ca. 1380 cm-1 is attributed to the symmetric v(O-C-O) of C1 product formate. The band at ca. 1307 cm-1is attributed to the symmetric v(O-C-O) of C2 product oxalate. The band at ca. 1433 cm'1is assigned to the asymmetric v(O-C-O) of the other C2 product glycolate. The band at ca. 1668 cm-1 is related to v(C=O) of intermediate enols.

[0177] Overall, these results are consistent with the HPLC results and previous reports on GOR on Pt / C catalysts in alkaline media. The in-situ Raman spectroscopy tests was also conducted to observe the catalyst surface evolution in the presence of glycerol in alkaline at the potential range of 0.2 - 1 .1 V vs. RHE. As shown in Figure 2g, a broad band at ~ 600 cm2is believed to be the formation of PtOx. The formation of PtjO4at 709 cm1, Pt5Oe at 644 and 674 cm-1, and PtO at 657 cm-1, was observed. Example 2- GOR with the PEM-based alkali-alkali MEA

[0178] GOR was carried out using a PEM-based MEA electrolyzer, with 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte.

[0179] Replacing the AEM by a PEM is expected to effectively suppress the anion crossover. In the present experiment, GOR was carried out in PEM-based MEA electrolyzer, with 1 M KOH & 1 M glycerol anolyte and 1 M KOH catholyte. However, distinct from the H-cell configurations, in MEA, the anode catalyst is sandwiched between gas diffusion layer (GDL) and PEM, which leads to the insufficient supply of hydroxide ions from the PEM side. Consequently, as shown in Figure 3a, the current density experienced notable reduction, with a maximal current density of only 0.1 A cm-2 obtained at a cell voltage of 0.80 V.

[0180] Example 3 - GOR with the PEM-based acid-alkali MEA

[0181] GOR was carried out using a PEM-based MEA electrolyzer, with 1 M KOH & 1 M glycerol anolyte and 0.5 M H2SO4 catholyte.

[0182] The ENE was introduced by replacing the 1 M KOH catholyte in the MEA with 0.5 M H2SO4 catholyte to improve the performance of the PEM-based MEA. The ENE harvesting was first tested in an H-cell.

[0183] As shown in Figure 3c, the l-V curve moved negatively when the catholyte changed from 1 KOH to 0.5 M H2SO4. Specifically, the onset potential at 10 mA cm-2shifted 0.80 V negatively, from 0.50 V to -0.30 V, which is close to the theoretical value of 0.81 V (pH difference of 13.7).

[0184] In addition, in-situ EIS tests at various potentials have been used to investigate the catalytic kinetics during the small-molecule valorization. The Bode phase plot depicts the trend of phase angle variation with frequency, and its peak at low frequency range is regarded as a symbol of interface reaction charge transfer, and therefore detected the voltage where the redox reaction start to occur. As shown in Figure 3d&e, the oxidation onset potential drops from 0.2 - 0.3 V to -0.6 - -0.5 V when the catholyte changes from 1 M KOH to 0.5 M H2SO4, which is consistent with the onset potential change in the linear sweep voltammetry (LSV) tests of Figure 3c.

[0185] Further measurements were conducted on the PEM-based acid-alkali MEA device. As shown in Figure 3a, a current density of 0.1 A cm-2was obtained at a cell voltage of only 0.05 V for the PEM-based acid-alkali MEA, indicative of a significant reduction of cell voltage due to the ENE utilization. Although the limiting current density before the cell voltage enters the OER region remains unaffected due to the OH- supply constraint, the decrease in cell voltage greatly diminishes the energy consumption for hydrogen production at lower current densities. GOR product analysis in PEM-based acid-alkali MEA was performed using the chronopotentiometric method.

[0186] As depicted in Figure 3b, the overall FE of liquid products in the anolyte approaches 100% when the current density remains below 100 mA cm-2, corresponding to a cell voltage of 0.53 V. No liquid product was detected in the catholyte owing to PEM obstruction. Once beyond the plateau current density as shown in Figure 3a (0 - 0.6 V), the cell voltage increases dramatically accompanied by a notable decrease in total FE of the liquid products. Due to the reduced potential requirement for the concurrent OER, the decrease in FEs of liquid products above a current density of 0.15 A cm-2could originate from CO2 production or OER.

[0187] Additionally, facilitated by ENE, the PEM-based acid-alkali MEA electrolyzer also functions as an electricity supply, yielding electricity output and simultaneously cogenerating GOR products and hydrogen. Figure 9a shows the polarization and power density curves attained from the l-V tests. An open circuit voltage (OCV) of 0.43 V is evident, achieving a peak power density of 10.2 mW cm2at a current density of 51 .4 mA cm2.

[0188] Example 4 - GOR with the AEM-based acid-alkali MEA

[0189] Glycerol oxidation reaction (GOR) was further investigated in an acid-alkali AEM-based MEA, using an anolyte of 1 M KOH with various concentrations of glycerol, and a catholyte of 0.5M H2SO4

[0190] The ENE effect was also introduced and investigated on the AEM-based MEA electrolyzer for GOR (Figure 4).

[0191] As shown in the Figure 4a, a negative shift of the onset potential was observed for both OER and GOR compared to Figure 2b, indicative of the effects of ENE harvesting. Similar to the PEM-based acid-alkali GOR, a current density plateau below 200 mA cm-2was observed, resulting from limited hydroxide ion supply from acid catholyte via AEM. In contrast to the AEM-based alkali-alkali cell, where the plateau is governed by the constant concentration gradient through the diffusion layer, the prohibited or insufficient supply of hydroxide ion, essential to facilitate GOR, from acidic catholyte through the ion exchange membrane reduced the plateau current density from exceeding 500 mA cm-2 to less than 200 mA cm-2. It may also be noted that the GOR in an acid-alkali H-cell is not suffered from the lack of hydroxide ion supply. This is because, unlike in MEA, the electrode / catalyst is enclosed by the anolyte and not in direct contact with the membrane in an H-cell. Consequently, it may be found out that the slope of the kinetic dominant region in the l-V curves decreased when the catholyte was changed from 1 M KOH to 0.5M H2SO4 in the AEM-based MEA (in Figure 4a), with the Tafel slope increased from 24.7 to 50.3 mV dec-1(in Figure 8b); however, the slope was almost the same after the change of the catholyte in the H-cell (in Figure 3c), with Tafel slopes ~ 39 mV dec-1in both acid-alkali and alkali-alkali cells (in Figure 6h). Moreover, as shown in Figure 3a, the slope of the kinetic dominant region in PEM-based MEA demonstrates no slope change between two types of catholytes, with Tafel slopes ~59 mV dec1, because that the PEM prohibits the exchange of anions regardless of the types of catholyte.

[0192] It may be noted that the kinetics should be determined by the GOR rate due to its slower kinetics compared with HER, proved by the comparison of Tafel slope values as shown in Figure 6i-j (~38 and 28 mV dec-1for GOR and HER, respectively). These results suggest that the sufficient hydroxide ion supply from catholyte to anolyte via the ion exchange membrane is crucial in MEA devices. When the cell voltage is above 0.3 V (figure 4a), the current density started to increase dramatically, which is a phenomenon not observed in PEM-based acid-alkali system. This is because that the electromigration- driven hydroxide transfer from catholyte to anolyte via AEM, which was prohibited via PEM, increased the hydroxide supply more effectively when the cell was more polarized. In Figure 6f, it is clear that GOR performance in the acid-alkali MEA electrolyzer is much improved than that in the H-cell.

[0193] Figure 4b shows the overall FE of liquid products in anolyte and catholyte. It is noteworthy that the crossover of liquid products to catholyte has been greatly suppressed even when using the AEM within the acid-alkali system. This may be attributable to the improved hydroxide anion concentration gradient between the anolyte and catholyte, which greatly improves the hydroxide anion diffusion from anolyte to catholyte via AEM according to Fick’s law, therefore alleviating the crossover of other anions as the ion exchange capacity (IEC) of the AEM is fixed. To support this hypothesis, the crossover rates of C3 products, TA, GLA and LA, in alkali-alkali and acid-alkali MEA cells were compared at OCV under experimental conditions. As shown in Figure 1 , the crossover of C3 anions was greatly suppressed in acid-alkali cell compared with that in alkali-alkali cell, which is consistent with the hypothesis. Furthermore, the overall FE towards liquid products remains close to 100% up to the current density of 500 mA cm2, corresponding to the cell voltage at which the OER has been initiated. Nevertheless, the GOR still dominates over OER, maintaining the total FE of liquid products at more than 80% at current density up to 1 A cm2. This is a dramatic improvement compared to the AEM-based alkali-alkali MEA electrolyzer, especially at high current densities.

[0194] The effect of electromigration force was also studied. Figure 13c and 13d shows that the crossover of hydroxide from anolyte to catholyte was supressed, and sulfate from catholyte to anolyte was accelerated, when a current density of 200 mA cm2is applied.

[0195] Figure 4c compares the percentage of products in catholyte to better demonstrate the crossover effect between AEM-based alkali-alkali MEA and acid-alkali MEA. The highest crossover ratio dropped from 30% in the AEM alkali-alkali system, to 4.1% in the AEM acid-alkali system, at a current density of 10 mA cm2. With the assistance from the electromigration, the crossover ratio rapidly decreased to lower than 1% at a current density of 50 mA cm2, and further declined to close to 0% with increasing current density. Accordingly, compared to the addition of high-concentration salt to suppress the liquid product crossover, the acid-base system does not require further salt usages for supressing crossover of liquid products. Moreover, the resulting acidic solution discharged from the system can also be used for the product separation and purification process, contributing to further cost reduction. Typical process in industrial production involves the neutralization of alkaline product electrolyte by sulfuric acid before conducting organic acid product distillation for separation and purification. Therefore, even the negligible amount of product in catholyte can be collected while the sulfuric acid catholyte is used for neutralization. Figure 4d showed that the overall liquid FE improved from 30% to 97% at the current density of 0.45 A cm2.

[0196] To further investigate the change brought by the catholyte replacement, the FE values towards all liquid products including C3 products (e.g. tartronate, glycerate), C2 (e.g. acetate, oxalate and glycolate) and C1 (e.g. formate) products, are provided in Figure 4e. For both AEM-based alkali-alkali MEA and acid-alkali MEA, C3 products are dominant at lower current densities. However, the C3 FE dropped to only 7.1% at 0.45 A cm2in the AEM-based alkali-alkali MEA electrolyzer, whereas the value obtained in the AEM-based acid-alkali MEA was 55% even at the current density of 1 A cm2. This observation is consistent with the report where higher KOH concentration facilitates the C-C bond cleavage, whereas the limited hydroxide ion supply from cathode via AEM- based acid-alkali MEA suppresses the C-C bond cleavage. The influence is more prominent at higher current densities due to the faster hydroxide ion depletion and lack of hydroxide supply through AEM. Such a configuration thus promotes the generation of higher value C3 products from GOR and obstructs the C-C bond cleavage to reduce the carbon emission at high current densities.

[0197] The density functional theory (DFT) calculations were performed. The free energy change in the C-C bond cleavage step, CH2OHC*OC*OadS- CH2OHC*OadS+ COadS, was investigated on Pt (11 1) surface (shown in figure 16a). Figure 16b exhibits the free energy changes at various potentials (vs. RHE) against pH at the anode surface. In the potential range of GOR, the values of free energy change all decrease with increasing pH, implying a reduced barrier for C-C bond cleavage with increasing pH at the anode.

[0198] Additionally, similar to the PEM-based acid-alkali electrolyzer, the AEM-based acid-alkali electrolyzer may work under electricity generation mode, with an OCV of 0.43 V, and a maximal power density of 13.4 mW cm2at the current density of 71 .4 mA cm2(Figure 9b). Although the increment in the KOH and H2SO4 concentration increases the power output as demonstrated in the H-cell (as shown in the Figure 6c-e), due to the limitations of the Sustainion AEM in this experiment, KOH electrolyte is limited at a concentration of 1 M. This may be resolved by replacing Sustainion AEM with other AEM of higher ion exchange capacity. Additionally, as shown in Figures 13a to 13d and 14a to 14f, the stability tests were conducted on the AEM-based acid-alkali electrolyzer as both a hydrogen and electricity co-generator and a GOR / HER electrolyzer at industrial-level current densities, further highlights the potential of the innovative MEA for practical applications.

[0199] In summary, five different MEA configurations to perform GOR with various ion-exchange membranes and catholyte have been investigated. As shown in Table 3, the AEM-based acid-alkali MEA outperforms all other configurations. This AEM-based acid-alkali MEA introduces convection-enhanced mass transfer, ENE harvesting and reduced resistance to simultaneously boost GOR activity. The product crossover was also impeded as the IEC was occupied by the increased concentration gradient-driven diffusion hydroxide ion in addition to the electromigration force directing anions transfer from catholyte to anolyte. The C-C bond cleavage suppression arised from the decrease of the local pH, which led to a high selectivity toward valuable C3 oxidation products and low carbon emission. Consequently, a cell voltage of only 0.377 V was needed to obtain a current density of 200 mA cm2, meanwhile the FE towards C3 products was 55%, and the product crossover ratio was close to zero at the current density of 1000 mA cm2. Additionally, the outlet sulfuric acid catholyte is promising for liquid product separation and purification process to further cut the expense. It is also impressive to find that AEM- based acid-alkali MEA may act as an electricity generator. The present invention can provide a potential prototype device for the energy-efficient, environmentally friendly and decentralized co-generation of hydrogen and valorized products without additional energy input.

[0200] Table 3. Advantages and disadvantages of glycerol reformer with various configurations

[0201] Density functional theory (DFT) calculations

[0202] DFT was used within the Vienna ab initio simulation package (VASP) to calculate the free energies. A 3 x 3 x 1 supercell cell of Pt( 111 ) slab with four layers (bottom two layer fixed) is constructed as substrate electrocatalysts and 4 x 4 x 1 Monkhorst- ack k-point grids are used. The vacuum layer was larger than 15 A to prevent the interaction between periodical slabs. Grimme’s D3 method was employed to consider the van der Waals (vdW) interactions. The cut-off energy was set as 500 eV. The systems were optimized until energy and force are less than 10-5 eV and 0.01 eV / A. The constant potential method was used to calculate the potential-dependent energy of the systems. The VASPsol code is used to model the aqueous environment with a relative permittivity of 80. The effective surface tension parameter and Debye screening length were set to 0 and the 3.0 A in VASPsol, separately. The charges for each system were added from - 1 .5 e to +1 .5 e in steps of 0.5 e to clarify the electrode potential function. The pH effect for electrode potential is set as UKHE= USHE+ kBTln(10)pH / e.

[0203] Industrial Applicability

[0204] The present invention relates to a membrane electrode assembly which may be useful for hydrogen production. The membrane electrode assembly according to the present invention is capable of selectively cracking an alcohol to form valuable liquid products including formic acid, etc., while concurrently generating hydrogen for energy storage and consumption. Therefore, the industrial applicability of the present invention is self-evident.

[0205] It will be noted that various other modifications and adaptations of the invention will be apparent to the person skilled in the art after reading the foregoing disclosure without departing from the spirit and scope of the invention and it is intended that all such modifications and adaptations come within the scope of the appended claims.

Claims

Claims1 . A membrane electrode assembly for hydrogen production, comprising: an anode compartment having an inlet for receiving an anolyte comprising an alcohol and an alkali and an outlet for discharging reaction products of an alcohol oxidation reaction, the anode compartment having an electrically conductive layer coated with at least one anode catalyst; a cathode compartment having an inlet for receiving a catholyte comprising an acid and an outlet for discharging hydrogen, the cathode compartment having an electrically conductive layer coated with at least one cathode catalyst; an ion exchange membrane interposed between the anode compartment and the cathode compartment, wherein the ion exchange membrane comprises opposite surfaces respectively in contact with the anode catalyst and the cathode catalyst, thereby forming a continuous structure with the anode compartment and the cathode compartment.

2. The membrane electrode assembly of claim 1 , wherein the ion exchange membrane is selected from anion exchange membrane (AEM), proton exchange membrane (PEM), or combinations thereof.

3. The membrane electrode assembly of claim 1 , wherein the ion exchange membrane comprises an anion exchange membrane (AEM).

4. The membrane electrode assembly of claim 1 , wherein the alcohol is selected from glycerol, methanol, ethanol, ethylene glycol, isopropanol, propylene glycol, butanol or mixtures thereof.

5. The membrane electrode assembly of claim 4, wherein the alcohol is selected from glycerol, methanol, ethanol, ethylene glycol or mixtures thereof.

6. The membrane electrode assembly of claim 1 , wherein the anode catalyst and / or cathode catalyst comprises a metal selected from Pt, Pd or alloys thereof.

7. The membrane electrode assembly of claim 6, wherein the anode catalyst and / or cathode catalyst comprises 20 to 80 wt.% of the metal.

8. The membrane electrode assembly of claim 1 , wherein the catholyte comprises an acid selected from H2SO4, HCI, HNOs, H3PO4 or mixtures thereof.

9. The membrane electrode assembly of claim 1 , wherein the anolyte comprises an alkali selected from KOH, CsOH, NaOH, LIOH or mixtures thereof.

10. The membrane electrode assembly of any one of claims 1-9, wherein the electrically conductive layer comprises a gas diffusion layer (GDL) in contact with the anode catalyst or the cathode catalyst.1 1. The membrane electrode assembly of claim 10, wherein the GDL comprises carbon paper, carbon cloth, titanium foam, nickel foam, Pt-coated titanium foam, titanium felt, Pt-coated titanium felt, or nickel felt.

12. A method of producing hydrogen via the membrane electrode assembly as defined in claim 1 , the method comprising: (a) electrochemically oxidizing the alcohol in the anode compartment in the presence of at least one alkali, and (b) concurrently electrochemically reducing protons in the catholyte housed in the cathode compartment to thereby evolve hydrogen gas.

13. The method of claim 12, wherein the alcohol is selected from glycerol, methanol, ethanol, ethylene glycol, or mixtures thereof.

14. The method of claim 13, wherein the alcohol is glycerol and wherein step (a) is a glycerol oxidation reaction (GOR).

15. The method of claim 12, wherein the protons are produced by the dissociation of an acid selected from H2SO4, HCI, HNO3, H3PO4, or mixtures thereof in the catholyte.

16. The method of any one of claims 12 to 15, wherein the acid is H2SO4.

17. The method of claim 16, wherein step (a) and step (b) are facilitated by an ion exchange membrane in fluid communication with both the catholyte and anolyte.

18. The method of claim 17, wherein the ion exchange membrane is an anionic exchange membrane.

19. The method of claim 17, wherein step (a) and / or step (b) are independently performed at a temperature of from 25 °C to 90 °C.

20. An electrochemical cell comprising the MEA according to any one of claims 1 to 1 1.

Images