Electrochemical hydrogen separation and recovery by non-platinum group metal catalyst

Abstract

Electrochemical hydrogen separation and recovery by a non-platinum group metal catalyst are disclosed and more particularly to a mesoporous molybdenum oxide catalyst structure formed at a low pH of about 5 to about 4.5 that includes hexagonal and orthorhombic crystal structures. The catalyst layer containing the mesoporous molybdenum oxide can be formed on a gas diffusion layer of a membrane electrode assembly of a high temperature proton exchange membrane fuel cell.

Description

ELECTROCHEMICAL HYDROGEN SEPARATION AND RECOVERY BY NON- PLATINUM GROUP METAL CATALYST CROSS REFERENCE TO RELATED APPLICATIONS This application claims the benefit of US Application No.63 / 588,468, filed on October 6, 2023, which is incorporated herein by reference in its entirety. BACKGROUND

[0001] The present disclosure generally relates to electrochemical hydrogen separation andrecovery by a non-platinum group metal catalyst, and more particularly to a mesoporous molybdenum oxide catalyst formed at a low pH of about 4 to 4.5 that includes hexagonal and orthorhombic crystal structures.

[0002] The goal of electrochemical hydrogen separation and compression (EHSC) technologyis to separate (recover) hydrogen gas from what would otherwise be considered waste streams of mixed composition. High-temperature proton exchange membrane fuel cell (HT-PEMFC) technology is considered a promising, clean, and efficient technology for hydrogen recovery from waste streams such as: plasma pyrolysis assembly (PPA) effluent, steam methane reformer processes, and metal heat-treatment processes. Platinum group metal (PGM) catalysts such as Pt black, Pt / C, Pt-alloy, Pt-Ru, and the like are routinely used in proton exchange membrane (PEM) cells to separate the hydrogen gas from waste streams.

[0003] Although useful, PGM catalysts suffer from a high cost due to scarcity of the basematerials. Indeed, the U.S. Department of Energy estimates that PGM catalysts may constitute upwards of 40% the cost of fuel cells incorporating them. Further, efficiency of PGM catalystssuffers in waste streams where CO, CO2, and unsaturated hydrocarbons are significantcontaminants. The problem of unsaturated hydrocarbon contamination can be further reduced by operation of PEMs at higher temperatures. Thus, there is potential for increased efficiency of hydrogen recovery.

[0004] Additionally, many alternative catalysts such as those made from Fe-N-C, Pd-CeO2 / C,showed decreased performance; either losing reactivity, having a low power density, or increased fabrication complexity.

[0005] Thus, there is a need for development of a HT-PEMFC suitable catalyst with decreasedexpense, increased recovery rate of hydrogen, greater tolerance to contaminant gases, increased durability; and, greater power densities and mass activities. BRIEF SUMMARY

[0006] Disclosed herein are electrochemical systems including a mesoporousmolybdenum oxide catalyst layer and processes for making the mesoporous molybdenum catalyst and membrane electrode assemblies including the same. The mesoporous molybdenum oxide catalyst composition includes hexagonal and orthorhombic crystal structures. This disclosure further describes a method of use for hydrogen recovery in a high-temperature proton exchange membrane fuel cell.

[0007] In one or more embodiments, a membrane electrode assembly comprises ananode with an anodic gas diffusion layer coupled to the anode. An anodic catalyst layer is on the anodic gas diffusion layer. The anodic catalyst layer is a mesoporous molybdenum oxide of the formula MoO3. The membrane electrode assembly further comprises a cathode. A cathodic gas diffusion layer is coupled to the cathode. A cathodic catalyst layer is on the cathodic gas diffusion layer. A proton exchange membrane is sandwiched between the anodic catalyst layer and that cathodic catalyst layer. In certain embodiments the proton exchange membrane of the membrane electrode assembly is a sulfonated tetrafluoroethylene-based fluoropolymer-copolymer. In still other embodiments of the membrane electrode assembly, the anodic catalyst layer has a loading in a range from 1.5 mg cm-2to 2.6 mg cm-2. In still other embodiments, the cathodic catalyst comprises Pt / C. In still other embodiments the membrane electrode assembly is a component of a proton-exchange membrane fuel cell. In certain embodiments, the mesoporous molybdenum oxide has a hexagonal and an orthorhombic crystal structure. In still other embodiments, the mesoporous molybdenum oxide has an agglomerated nanorod-shaped morphology. In still other embodiments of the membrane electrode assembly, the mesoporous molybdenum oxide comprises a slit / wedge mesoporous structure.

[0008] In one or more additional embodiments there is presented a method of preparinga catalyst. A first solution is formed by dissolving ammonium heptamolybdenum tetrahydrate in water at an elevated temperature. A second solution is formed by dissolving [poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol) with tetrahydrofuransolution. Water is statically added to the second solution when the second solution is clear. The first solution is then added to the second solution forming a third solution. At room temperature the pH of the third solution is lowered to about 4 to about 4.5 to form a precipitate. The precipitate is heated to form a blue-colored powder. The powder is collected and calcined under heating cycles. The calcined powder is mixed with an alcohol / water solution and a water based sulfonated tetrafluoroethylene-based fluoropolymer-copolymer to form a homogenous solution to result in the catalyst. In some embodiments of the method the heating cycles are 150 °C for 12 h at 2°C min-1, 250 °C for 4 h at 1 °C min-1, 300 °C for 3 h at 1°C min-1, and 400 °C for 2 hat a 1 °C min-1 ramp rates. In another embodiment of the method of preparing a catalyst,lowering the pH comprises adding acetic acid.

[0009] Some embodiments of the disclosure herein are present as a method of forminga membrane electrode assembly. A catalyst ink comprising mesoporous MoO3 is coated ontoa gas diffusion layer to form a catalyst layer thereon. The catalyst layer is hot pressed to aproton exchange membrane. In certain embodiments of the method of forming a membrane electrode assembly the proton exchange membrane comprises a sulfonated tetrafluoroethylene- based fluoropolymer-copolymer. In still other embodiments, the gas diffusion layer comprises a carbon paper or carbon cloth. In still other embodiments, the catalyst ink is comprised of 1- propanol, water, tetrafluoroethylene-based fluoropolymer-copolymer, and the embodiments of the catalyst as described herein.

[0010] The disclosure may be understood more readily by reference to the followingdetailed description of the various features of the disclosure and the examples included therein. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Features and advantages of embodiments of the present invention will becomeapparent on reading the detailed description below with reference to the drawings, which are illustrative but non-limiting, wherein the like elements are numbered alike:

[0012] FIG. 1 illustrates a schematic of a membrane electrode assembly (MEA)fabrication technique with a schematic representation of a proton exchange membrane fuel cell (PEMFC) with all required layers in a single cell in accordance with one or more embodiments of the present disclosure;

[0013] FIG. 2 presents an example experimental apparatus and equipment layout toperform electrochemical studies for the evaluation and validation of the catalyst in accordance with one or more embodiments of the present disclosure;

[0014] FIG. 3 presents a powder X-Ray Diffraction (PXRD) pattern of molybdenumoxide and transmission electron microscopy (TEM) images, FIGs 3a and 3b, of MoO3catalyst according to an embodiment of the present disclosure;

[0015] FIG. 4 (a-d) presents scanning electron microscopy (SEM) images of calcinedMoO3catalysts according to embodiments of the present disclosure;

[0016] FIG. 5 presents the Raman spectra of MoO3 calcined samples according toembodiments of the present disclosure;

[0017] FIG.6 illustrates the X-ray photoelectron spectroscopy (XPS) wide-scan surveyspectra of -MoO3nanoflakes in accordance with one or more embodiments of the present disclosure;

[0018] FIG. 7 illustrates Brunauer-Emmett-Teller (BET) isotherm and Barrett-Joyner-Halenda (BJH) pore size distribution curves of MoO3 samples at 300 and 400 °C in accordance with one or more embodiments of the present disclosure;

[0019] FIG. 8 presents the results of impedance spectroscopy of single MEAs with twodifferent loading amounts and at different voltages in accordance with one or more embodiments of the present disclosure;

[0020] FIG. 9 illustrates the fuel cell equivalent circuit used to fit data in accordancewith one or more embodiments of the present disclosure;

[0021] FIG. 10 presents MEA performance durability testing at different voltages witha constant cell temperature in accordance with one or more embodiments of the present disclosure;

[0022] FIG. 11 presents a comparison of power densities with current densities at aconstant voltage in accordance with one or more embodiments of the present disclosure;

[0023] FIG. 12 illustrates a 70 hour stability test of MoO3-B MEA comparing a fuelcell polarization curve at a constant voltage and cell temperature in accordance with one or more embodiments of the present disclosure; and

[0024] FIG. 13A and B presents representative gas chromatograph traces for analysesperformed on embodiments of the invention with different feed gases and voltages. DETAILED DESCRIPTION

[0025] Disclosed herein is a mesoporous molybdenum oxide catalyst suitable for usagein a high-temperature membrane electrode assembly (HT-MEA) of a high-temperature proton exchange membrane fuel cell (HT-PEMFC). The mesoporous molybdenum oxide catalyst can be formed at a relatively low pH of less than about 4-4.5 using basic media interaction with a block-copolymer (BCP). The resulting mesoporous molybdenum oxide includes both a hexagonal and orthorhombic crystalline structure and exhibits a slit / wedge porous structure. The performance and stability of the material was quantified in an HT-PEMFC as the anodic catalyst layer of the high-temperature membrane electrode assembly (HT-MEA or MEA). Advantageously, the mesoporous molybdenum oxide catalyst demonstrated superior performance than the significantly more expensive platinum group catalysts such as platinum on carbon (Pt / C) used in comparable PEMFCs.

[0026] Presented in further detail below, by way of general example, the anodic catalystlayer including the mesoporous MoO3was prepared by a gas diffusion layer (GDL) method, which generally included coating the anodic catalyst on an anodic GDL and hot pressing ontoa sulfonated tetrafluoroethylene-based fluoropolymer-copolymer membrane commerciallyavailable under the tradename Nafion®. The GDL is generally carbon-based material, e.g., acarbon cloth or carbon paper, or a metal mesh structure formed of stainless steel, nickel ortitanium. Hot pressing of the catalyst onto the GDL was done without affecting the porosity of the anodic catalyst or the GDL. Electrochemical impedance spectroscopy (EIS) characterized the performance of the HT-PEMFC incorporating the anodic catalyst tested at loading rates of 1.5 or 2.6 mg cm-2. To simulate a plasma pyrolysis assembly (PPA), waste effluent gases tests were performed with H2, H2 / CO, and H2 / CO / C2H2 feed gases at variable flow rates. A comparison to a commercially available Pt / C catalyst demonstrated that the mesoporous molybdenum oxide catalyst provided herein demonstrated increased CO- tolerance and electrochemical performance over a 70-hour (h) timeframe.

[0027] In the following examples, the chemicals, surfactants, and gas cylinders wereused as received from commercial suppliers. Stock materials included: ammonium heptamolybdenum tetrahydrate; P123 [poly(ethylene glycol)-block-poly(propylene glycol)- block-poly(ethylene glycol), molar mass, 5400 g mol-1]; tetrahydrofuran; nanopure water; acetic acid; 1-propanol; Nafion®solution-1100 MW; hydrogen (UHP grade); hydrogen and carbon monoxide mixture (H2-98%, CO-2%, Research grade); acetylene (Standard grade); nitrogen (Zero grade); and, 13X molecular sieve.

[0028] Synthesis of catalyst material in the form of mesoporous molybdenum oxideshaving pores with diameters between 2 nanometers (nm) and 50 nm was completed as follows: A precursor for molybdenum (Mo), ammonium heptamolybdenum tetrahydrate powder (AMT) (0.003 mol) was transferred to a 15 milliliter (mL) vial and dissolved in 10 mL of H2O at 60°C. P123 (0.0004 moles) was added in (0.0375 moles) tetrahydrofuran (THF) solution (concentration ratio of 5:95, THF: H2O) with constant stirring at room temperature in a 100 mL beaker. When the solution was clear, H2O (0.72 mol) was added slowly at a static position to avoid froth. The mixture was stirred for 1 hour, and the hot AMT mixture was added dropwise with constant stirring at room temperature. After 30 min of stirring, acetic acid (0.03 mol) was added dropwise to the clear gel solution at room temperature. Precipitation occurred with the addition of acid. The resultant solution was stirred overnight to complete the interaction between the reactant and surfactant. The solid final product was transferred to an oven at 100 °C for 5 h to obtain a blue-colored powder. The solid product was collected and calcined under different heating cycles of 150 °C for 12 h at 2°C min-1, 250 °C for 4 h at 1 °C min-1, 300 °C for 3 h at 1°C min-1, and 400 °C for 2 h at a 1 °C min-1ramp rate. These cycles were used to remove all surfactants and control the mesoporosity of the material.

[0029] As noted above, the mesoporous MoO3 catalyst layer was only used for theanode side. However, the mesoporous molybdenum oxide catalyst could also be used on the cathode side. In these experiments, 0.5 mg cm-2Pt / C coated GDL was used on the cathode side.

[0030] The catalyst ink preparation method is as follows. First, 141 microliters ( L) of1-propanol and 108.9 L of Nanopure water were mixed well in a 2 mL sample tube. Further, 250 L of water-based Nafion®(20wt.%) solution was added to the same sample tube and stirred for 10 min to form a homogenous solution. Next, 50 mg of MoO3 catalyst, created as above described, was added to prepare a 50 wt.% solution with respect to Nafion®solution.Nafion®is a sulfonated tetrafluoroethylene-based fluoropolymer-copolymer with CASNumber 66796-30-3 and a chemical formula of C7HF13O5S .C2F4. Further, using Parafilm®(a semi-transparent, flexible film composed of a blend of waxes and polyolefins), the sample vial was closed airtight and stirred thoroughly on a magnetic stirrer overnight to obtain a homogenous dispersion of catalyst ink. The next day, before coating, the solution was stirred for 30 seconds on a vertigo stirrer to form a thick ink so that it could be easily coated on the GDL by brush.

[0031] FIG. 1 illustrates a schematic of a MEA fabrication technique with a schematicrepresentation of a PEMFC with all required layers in a single cell. Catalyst, prepared as above-described, is mixed with a Nafion® solution (also above-described) to form a catalyst ink. The catalyst ink is applied via brush to a first gas diffusion layer (GDL) forming a catalyst layer. The catalyst layer is hot pressed to a solid electrolyte on a first side with a second GDL on the opposite side of the solid electrolyte also hot-pressed into place forming a MEA. The MEA is further sandwiched between two gaskets and placed between two opposing graphite plates. The graphite plates, in turn, are in contact with respective cathode and anode side current collector plates with respective current collector terminals; which, when screwed together, seal the PEMFC assembly. A plurality of input and output pipes feed and remove gas flows into, through, and out from the system. Examples include: hydrogen input, input nitrogen, excess hydrogen output, and product gases output pipes. Alternatively, in other embodiments, there may be cathode and anode inlet and outlet pipes.

[0032] Nafion 212®, proton conductivity electrolyte membranes, were used for allexperiments. The graphite electrode used for analysis had a 5 square centimeter (cm2) active area. The overall size of the electrode plate was 8 x 8 cm. After cutting the Nafion®membrane to match the electrode size, there was no need for protonation in any solvent system. The Nafion®membrane was used as a solid electrolyte between the anode and cathode. For the anode electrode preparation, the MoO3catalyst was coated on the anodic GDL. The reason for using GDL was that the alcohol solution of the catalyst ink, when directly used on Nafion®film, resulted in membrane swelling. The GDL was cut to the electrode size, and uncoated GDL was regularly weighed until a constant weight was seen, which could be used for coating calculation.

[0033] To achieve the desired thickness of the anodic catalyst layer with a certainamount of catalyst loading (mg cm-2), an ink solution was added drop-wise onto the GDLusing a micropipette. The GDL coating was prepared by standard GDL brush coatingmethodology, which generally includes coating a mixture of carbon black powder, ahydrophobic dispersion agent, and solvents, onto carbon paper or woven carbon cloth substrates (both have macro-pores) to form a uniform thin micro-porous layer on top of the macroporous layer provided by the substrate. A drop of catalyst ink, prepared as above- described, was then uniformly spread on the GDL, and the GDL was dried on a hot plate at 50°C. GDL was weighed and coated until the desired catalyst loading was achieved. The hot plate temperature was kept constant at 50°C to keep the GDL heating stable and to allow fast and uniform drying of alcohol and water-based ink. GDL was kept for 30 min on the hot plate, cooled down at room temperature, weighed again, and further coated in case the loading amount was not appropriate. The GDL was kept overnight for drying at room temperature toavoid cracking of the GDL and swelling of the Nafion® film due to wet ink. The next day,GDL was weighed to determine the precise coating amount. Next, the temperature of a hotpress machine was adjusted to 120 °C and a Nafion 212® membrane was pressed betweenanode and cathode GDL for 90 sec, resulting in a completed MEA. The MEA was installedin a graphite single cell with a Teflon® gasket loading of 15%. After packing the cell at apressure of 40 Newton-meter (Nm), the gas flow inlet and outlet were connected and the cell was attached to a DC power supply.

[0034] Testing included first purging the cell with N2 for 15 minutes. Hydrogen flowwas then started to the anode side. The anode side was operated with humidified H2(T= 95°C), and dry N2 flowed from the cathode at 95 °C with ambient pressure on both sides.Electrochemical impedance spectroscopy (EIS) of the cell was first operated at 25°C at 10mHz-100 KHz frequency to check cell performance. The EIS setup was purchased from Bio- Logic SAS power supply which was coupled to a potentiostat-Galvanometer system connected to electronic load for online experiments. The 20 kW fuel cell test station used for all the experiments was built by Fuel Cell Technologies, Inc, USA. All the impedance spectra were measured between the cathode (counter and reference electrode) and anode (working and sense electrode) of the fuel cell.

[0035] Different instruments characterized the chemical and structural properties of thematerials. Catalytic preparation solution pH was checked by a Thermo Scientific® pH meter (Orion Versa Star Pro). The powder X-ray diffraction (XRD) analysis was done on the RigakuUltima IV diffractometer (Cu K radiation, = 1.5406 Angstroms), where the beam voltagewas 40 kV and the current was 44 mA. Catalyst samples were scanned from 5-75 degree- 2value at a rate of 2 degrees per minute. Microporosity of the materials was analyzed usingnitrogen sorption on Quantachrome Autosorb iQ®, an automated gas sorption analyzer. The surface area and pore diameter were calculated by Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) methods. The pore size distribution information came from analysis of the desorption branch of the isotherm. The Raman analysis was recorded with a Renishaw system 2000 microscope, equipped with a 514 nm wavelength laser. The oxygen vacancy and oxidation states of the material were determined using X-ray photoelectron spectroscopy (XPS). Analysis was performed on a ThermoFisher K-Alpha X-rayphotoelectron spectrometer with Al K radiation ( = 1486.6 eV) as the radiation source. Passenergies of 200.00 eV and 50.00 eV were used to obtain survey and high-resolution spectra, respectively. All spectra were normalized to adventitious carbon (BE = 284.8 eV) and fitted using CasaXPS® fitting software (version 2.3.18PR1.0). The particle morphology was observed using an FEI Verios 460L field-emission scanning electron microscope (FE-SEM) equipped with an Everhart-Thornley detector (ETD). For MEA pressing, a hot-pressing machine was used (Maxx clam heat press by Stahl’s). Transmission electron micrographs were obtained using a Thermo Scientific Talos 200 S / TEM®. The outlet gases from the anode and cathode were sent to an SRI 8610C-MG-1 online gas chromatography instrument with a thermal conductivity detector (TCD) for analysis, and argon was used as a mobile phase.

[0036] FIG. 2 presents an example experimental apparatus and equipment layout toperform the electrochemical studies for evaluation and validation of the mesoporous MoO3 catalyst herein disclosed. The system was engineered to take advantage of different mixtures of gases with the gas flows metered by mass flow controllers (MFCs). Experiments were performed under various feed gases, including hydrogen, nitrogen, acetylene, and H2+2%CO, named H2CO. A 13X molecular sieve column was used to remove acetone from the acetylene gas. Nitrogen gas was used to purge the anode and cathode inlet of the cell before the reaction. After a few minutes of N2flow, H2flow was initiated and passed through a humidifier to add water content to the gas stream. The humidifier was inside the 1.2 kW test station, whose gas stream was attached to a 5 cm2cell. The cell operating temperature was increased from 25 to 150 °C; subsequently, H2 / N2 gases were passed through it with 100% relative humidity. During cell heating and testing of the MEA, N2was flowed from the cathode (at 30 sccm). Outlet gases were passed through a cold trap to a gas chromatograph (GC) for the anode outlet or directly to vent for the cathode outlet. The cold trap was used to trap the liquid by-productof the reaction and to send gases without moisture to GC for analysis. A mass flow meter(MFM) was attached before the GC to check the outlet gas flow rate. All the system metaltubing was built by using a 1 / 8-inch (0.125 mm) outer diameter (OD) steel tubing. A vacuum pump was attached for purging the metal tubing with inert gases before and at the end of the experiment using an anode sample valve. Cell stabilization was achieved by slowly ramping down the hydrogen gas feed (from 30 to 9 sccm) with a constant potential of 0.1 V. Hydrogen flow is considerably smaller than nitrogen, and the capacity of hydrogen to carry water is alsoconsiderably smaller; hence to overcome the drying issue and humidify the catalyst layer, 0.1V was applied over 8 h. Further, testing experiments were done with different voltages (0.15, 0.30, and 0.40 V) and reached stable current densities in 15-20 min by keeping a stable current. Finally, all gases were vented into fume hoods and vented to the atmosphere with all necessary precautions.

[0037] As an example embodiment of the present disclosure, mesoporous molybdenumoxide catalyst prepared in aqueous media at 4.56 pH conditions resulted in catalytically active species. When characterized by X-ray diffraction, the PXRD pattern of the molybdenum complex showed an orthorhombic crystal system (Pbnm space group) as shown in FIG. 3, wherein the left side of the figure illustrates the PXRD pattern of molybdenum oxide at 300 and 400 °C calcination temperatures. The molybdenum oxide catalyst exhibited semi- crystalline peaks at 300 °C, which was converted into a crystalline phase upon increasing the calcination temperature to 400 °C. The diffraction peaks observed at 400 °C belong to crystalline MoO3. The high-intensity peaks correspond to the three planes (110), (040), and (021) and were attributed to the formation of the -MoO3 phase. The average crystallite size was calculated using the Scherrer equation for small single crystallites in bulk samples. Calculated via the Scherrer equation, the crystallite size of MoO3 is about 16 nm.

[0038] Catalyst samples were analyzed by transmission electron microscopy (TEM) toinvestigate the MoO3nanostructure and its crystallinity, further illustrated in FIG. 3 at the right side which illustrates (a) TEM images of porous MoO3 catalyst calcined at 300 °C; and, (b) TEM images of porous MoO3catalyst calcined at 400 °C. The TEM micrographs illustrate an agglomerated nanorod-shaped morphology. The 400 °C calcined sample in FIG. 3(b) displays the distribution of MoO3 nanorods, which confirms the highly crystalline nature of MoO3nanorods. Based on XRD results where the (110) plane showed high intensity, likewise, TEM images show a lattice spacing of 0.38 nm, which matches the plane of MoO3.

[0039] However, agglomerated particles showed mixed morphology, clearly seen inthe SEM images of FIG. 4, which presents images of calcined MoO3 catalysts according to embodiments of the present disclosure at 300 °C (a and b) and 400 °C (c and d). The calcination temperature was a major deciding factor for material crystallite size. As shown in FIG.4, nanoflake nanoparticles of -MoO3 were around 1 m in size. The temperature of 400°C was responsible for reduced particle size, and the formation of uniform MoO3nanoparticles, as shown in FIG. 4 (c & d). The nanoflake morphology resolved the issue of volume expansion during the electrochemical reaction and significantly enhanced electrical conductivity.

[0040] Raman spectra of MoO3 calcined samples as shown in FIG. 5 displayed peaksat 668 cm-1, 820 cm-1, and 993 cm-1, which were well matched with the reported -MoO3 and corresponds to stretching vibrations of (O-Mo). The peak at 993 cm-1represented the (Mo6+= O) asymmetric stretching due to unshared oxygen. On the other side of the Raman shifts, 820 cm-1 showed bridging oxygen (Mo-O-Mo) coordination in a stretching mode. Similarly, 663 cm-1determined the stretching mode of triply coordinated oxygen (O-Mo3). Along with that, the strong intensity peaks of 993 and 820 cm-1also represented terminal Mo=O which describes oxygen vacancy, which was also confirmed by XPS.

[0041] FIG. 6 displays an XPS wide-scan survey spectra of the -MoO3 nanoflakes,indicating the presence of Mo, C, and O elements. The Mo 3d and O 1s regions show thedeconvoluted Gaussian peaks, which correspond to the oxidation state and oxygen vacancy in Mo. Deconvoluted 3d peaks at 233.1 and 233.9 eV were for the 3d5 / 2 shift, whereas 236.2 and 236.7 were for the 3d3 / 2energy shift. The catalyst preparation method and individual doublet of 3d5 / 2 and 3d3 / 2 of Mo6+higher oxidation state were in accordance with molybdenum oxo- species. The binding energy in 3d5 / 2-3d3 / 2doublet had a higher difference that explains thetwo types of Mo6+ species. Monomeric species were observed at lower values, and thedifference of higher values showed polymeric species35. The higher binding energy doublet (i.e., 233.9 and 236.7 eV) represented the presence of strong electronic interaction of Mo6+oxo-species. Along with that, Mo5 / 2lower binding energy was reported as a peak of -MoO. The deconvoluted O 1s fitted curve showed three components due to mesoporosity and oxygen bonding in oxide formation. The band observed at 530.8 eV corresponds to the lattice oxygen (O2-) of transition metal oxides. In the remaining two oxygen species, the peak at 531.6 eV could be from chemisorbed water, C=O, or hydroxide groups attached to the metal center.Peak 533.3 eV corresponds to oxygen absorption species or bound water. This energy doublet and oxygen species enhanced the activation for ion adsorption and desorption in Mo (VI). The high surface area and porous structure of catalysts in electrodes provide additional features for charge exchange at the electrode-electrolyte interface.

[0042] N2 adsorption and desorption isotherm studies were conducted to examinesurface area, pore volume, and (inset) pore diameter of MoO3 samples. FIG.7 illustrates BET isotherm and BJH pore size distribution curves of MoO3samples at 300 and 400 °C. The isotherm curve of MoO3 nanoparticles illustrated a type IV isotherm. Type-IV is sub- categorized as H1, H2, H3, and H4, which corresponds to different shapes and sizes of pores. On studying the MoO3 N2 isotherm hysteresis loop, the pores were consistent with similarities to H3 category, which contains slit / wedge-shaped pores. The surface area of the material decreased with increasing calcination temperature; hence samples at 300 °C observed 29 m2g-1, 0.3 cc g-1pore volume, and at 400 °C showed 16 m2g-1, 0.1 cc g-1pore volume. However, MoO3porous structure at both temperatures remained constant, which is 3.8 nm pore size, as shown in FIG. 7. Overall, average pore size distribution and slit / wedge mesoporous structure provide good physical contact between nanoparticles, and improved ionic and electronic conductivity.

[0043] Electrochemical impedance spectroscopy (EIS) is a non-destructive methodapplicable for fuel cells that provides valuable information about cell performance without interfering with the system equilibrium. EIS studies make it easy to distinguish the individual contribution of the interfacial charge transfer and the mass transport resistance between the gas diffusion layer (GDL) and the catalyst layer. When studying a single fuel cell with a lower surface area, stacking parameters also play a significant part. A single cell contains different components that are clamped together, such as the MEA, GDL, bipolar plates, current collectors, endplates, and sealing jackets. While sealing the cell, it is vital to apply appropriate clamping torque and use a suitable thickness of the gasket to obtain uniform pressure across the cell. As exemplified in Figure 1, a single cell was packed using an applied torque of 40 N·m. ESI results were collected at room temperature (25 °C). Clamping torque, pressure uniformity, and the temperature had a substantial effect on proton transfer through the cell in which the MEA was packed.

[0044] FIG. 8 presents the EIS results of single MEAs with two different loadingamounts (1.5 mg cm-2for MoO3-A and 2.6 mg cm-2for MoO3-B) and at different voltages, allEIS arcs were initiated and ended at common points, demonstrating that each tested single cell was clamped at similar torque and run under similar pressure. The results showed sufficient pressure was applied over the area and did not show leakage around the perimeter of thelayered assembly; similarly, it did not affect the porosity of the GDL or damage the MEA,which was not affected by ohmic resistance and increased the efficiency of the fuel cell.

[0045] Forming the mesoporous molybdenum oxide catalyst layer on the GDL and hotpressing to form the MEA could reduce the electrical conductivity of the GDL in case of over pressing and annealing, as this increases the ohmic resistance which would decrease the power density of the MEA due to a damaged GDL or Nafion®membrane obstructing the electronic pathway. In contrast, MoO3coated GDL pressed on the membranes with different catalystlayer thicknesses showed higher electronic conductivity. Even though MoO3-A of FIG. 8exhibited the highest ohmic resistance i.e. 15 ·cm-2 and MoO3-B exhibited the lowest cellresistance of approximately 5 ·cm-2, MoO3-B showed a lower kinetic loop since high catalystloading makes the electrode relatively thick.

[0046] FIG. 9 illustrates an equivalent-circuit model representing the data fit, whereinL represents the inductive element likely attributed to lead wires. The resistance Re describes the ohmic resistance due to the GDL electrolytes. MoO3catalyst layer in the planar form isrepresented by the parallel combination of charge transfer resistance Rct1 and Rct2 with theconstant phase element Q, respectively. This was the simplest equivalent circuit model that fits well with the data. In FIG.8, the semi-circles of the Nyquist impedance responses (starting with the high-frequency intercept) explains the ionic resistance of the membrane, and the arc ends with low-frequency impedance, which represents the sum of the ionic resistance of the membrane and charge transfer resistance of the catalyst layer. All impedance studies were measured by flowing H2at the anode and N2at the cathode and checking the anode (working electrode) electrochemical performance by blocking the cathode reactions and simplifying the impedance response of the electrode.

[0047] For certain embodiments of the present disclosure, the context for developmentwas increased recovery of H2 from a waste stream for provision to a Sabatier process. Hence,fits of the impedance arcs of MoO3-B show a shorter kinetic loop diameter than MoO3-A,which reveals the decrement of the charge transfer resistance (Rct). An increment in catalyst loading with an increased thickness of the electrode limits the ionic transportation betweenthe catalyst layer and the membrane. All the MoO3-B arcs show a 45 slope at high-frequencyimpedance due to the thick electrode, which resulted in non-homogeneous transport properties of the mesoporous molybdenum oxide catalyst layer. On the other hand, MoO3-A catalyst layer had a low loading, which made a thin electrode that was responsible for homogenousion transport, and as such it did not show any 45 straight line. Along with that, the hydrationof the membrane was found to be another factor that limits the fuel cell efficiency. Hence, hydrogen feed stream was passed through a wash bottle to become humidified before EIS for a long time, known as the breaking-in period.

[0048] The temperature was gradually increased over the break-in period to allow thefuel cell to achieve consistent current output at a specific voltage. As shown in Tables 1 and 2 below, applied voltages of 0.15, 0.30, and 0.40 V showed an excellent kinetic loop with agood Rct loop and less mass transfer through the catalyst layer in both MoO3-A and MoO3-Bcatalyst layers. The mass transport slightly increased in MoO3-B at 0.40 V potential due to oxygen transport limitation at lower frequency impedance. Overall, the real-axis intercepts assist in understanding anodic polarization resistance values, charge transport phenomenon, and diffusion. These values provide a valid confirmation with resistance values obtained fromequivalent-circuit fits.Viewed against the activities of the variety of platinum group metal (PGM) catalysts, two important metrics for comparison are the specific activity and metric activity of the catalysts. Specific activity (mA cm-2) provides catalytic activity and electronic properties of the catalyst to understand the exchange current density through redox reaction. From an industrial or commercialization viewpoint, mass activity of the catalyst is considered as surface activity and utilization of surface atoms which is nothing but the direct transformation as a cost of a fuel cell.

[0049] Although PGM catalysts such as Pt black, Pt / C, Pt-alloy, Pt-Ru, etc., are viewedas reliable, commercially viable, and durable catalysts for PEMFC; as above noted, per US Department of Energy estimates, PGM catalysts lead to a high cost, which constitutes upwards of 40% of fuel cell cost at large- scale production. Alternative methods of non-PGM catalysts have also suffered from performance and durability issues.

[0050] As noted above, tests of embodiments of the present disclosure controlled forfactors such as catalyst loading, catalyst morphology and activity, and the thickness of the catalyst layer. EIS observations, as above detailed, indicated that the “thick” electrode demonstrated shorter charge resistance which affected the cell performance.

[0051] To detail H2 recovery and cell performance, embodiments of the MoO3 catalystlayer were coated on GDL as above described with two different loading rates similar to those of commercially purchased Pt / C catalysts. Current density, power density, and stability ordurability with contaminant molecules (e.g., CO and C2H2) were compared. As seen in FIG.8, the first MoO3loading amount was 1.5 mg cm-2and termed MoO3-A with the second loadingdose of 2.6 mg cm-2 termed MoO3-B. Tests were compared against a commercialized Pt / Cmembrane control which was coated on both sides with a loading of 1.0 mg cm-2of Pt / C on each side (i.e., anode and cathode). A second control test was performed with an empty GDL on the anode side and coated GDL with Pt / C with 0.5 mg cm-2(60% of Pt / C) used as a cathode. The second control test served to account for the effect of MEA preparation using a hot-press. Hot pressing is not an ideal method for the commercial Pt / C membranes which had previously undergone heating and other treatments as part of their fabrication. For embodiments of the invention tested, the GDL coating method with hot-pressing at 120 C for 90 s performed well as an MEA fabrication procedure, resulting in better performance than the hot-pressed Pt / C membrane.

[0052] As demonstrated in Tables 1 and 2 above, and in FIG. 10, control studies ofcommercial Pt / C showed poor performance compared to the hot-pressed MEA including the mesoporous molybdenum oxide catalyst. Without subscribing to, or being bound by, a particular theory; it is believed that hot pressing conditions caused deep embedding of the conductive layer and catalyst into the electrolyte membrane thus accounting for the enhanced MEA performance and increased electron flow. As seen in Table 3, 90 s of hot pressing resulted in the highest power density (44 mW cm-2).

[0053] Embodiments of the catalyst tested at a high temperature of 150 C showedhigher performance and better power density compared to the PGM catalyst. Embodiments of the present disclosure as tested demonstrated: 1. increased electrochemical reaction kinetics of hydrogen oxidation reaction on the surface of the MoO3 catalyst layer; 2. decreased CO adsorption on the catalyst (resulting in more hydrogen production); 3. easier balancing of water transportation in the membrane, catalyst layers, and diffusion layers (also increased the chance of more catalyst active surface area exposure and the ability of reactants to diffuse to the reaction layer); and, 4. MoO3enhanced electrode kinetics more than Pt / C. These traits will advantageously reduce the overall total cost of a PEMFC incorporating embodiments of the invention.

[0054] Flowing nitrogen on the cathode as above described yielded the additionalbenefit of reduced H2O2 intermediate formation, which typically occurs at high oxygen concentrations. Thus, the electrochemical studies with the MoO3 catalyst layer under H2 / N2 and H2CO / N2performed with a constant voltage drop across a bipolar plate (0.15, 0.30, and 0.40 V) resulted in different current densities as seen in FIG. 10. Without subscription to, or binding to any particular theory, it is believed that in operation a bipolar plate is advantageous as it works in a corrosive environment with low pH and one side exposed to the reductive gas (H2). In embodiments of the present disclosure and the studies described herein, a nonporous graphite carbon served as the bipolar plate material. Experiments conducted withempty GDL at the anode and 0.5 mg of Pt / C cm-2 at the cathode utilized a Nafion 212®membrane due to better observed performance.

[0055] Presented in FIG. 11, the current density of MoO3-A and MoO3-B at 0.40 Vshowed 112 mA cm-2 with a peak power density of 44 mW cm-2, recorded when H2CO gaswas fed to MEA. The resultant power density was higher than others, even when the MoO3catalyst layer was fed with poisonous gas (CO). The electrode showed a highly orderedconfiguration and robust channeling for charge transfer due to two factors: ionomer and the thickness of the MoO3 catalyst layer. The kinetic performance is enhanced with 50 wt% of Nafion®, which limits the mass transport region and becomes more significant. The thickness of MoO3-B catalyst layer exposed more CO and adsorbed more amount, which lowers thecurrent density at 0.40 V, i.e.81 mA cm-2 with 32 mW cm-2 power density. The gas feed ratioof H2, H2CO, and C2H2 on the cell with two different total flow rates such as H2 / C2H2-15sccm and H2CO / C2H2-12 sccm. TEM showed that the (110) active plane of MoO3was ableto avoid hydrogenation reaction and recovered more percentage of hydrogen, shown in Table 3 above.

[0056] In addition, feed gas H2CO with the addition of C2H2 increased the poison levelto anode inlet gas feed on MoO3-A and decreased the current density to 110 mA cm-2. The mass activity of the MoO3-A MEA reached 73 mA mg-1(0.07 A mg-1). Per a literature survey, the highest reported mass activity of a PGM catalyst was for Pt3Ni at 0.7 A mg-1at 0.60 V at high pressure and low temperature. Even though fuel cell testing changes from lab to lab, MoO3-A as a non-PGM catalyst displays promising mass activity compared to a large cell size with PGM catalyst. On the other hand, a non-PGM Fe-N-C catalyst reached 35 mA cm-2current density at 0.9 VIR-freein H2 / air gas feed, while MoO3-A in H2CO poisonous feed gas displays 112 mA / cm current density at 0.4 V. Referencing the Table 3 values, it seems thelower thickness of the MoO3 catalyst layer had more tolerance for CO and C2H2 than a thickone formed in a comparable NiPCy2-functionalized multi-wall carbon nanotubes and carbon microfiber electrodes at the anode and Pt / C at the cathode. To make a more direct comparison, NiPCy2obtained higher hydrogen oxidation at 0.30 V to reach up to 40.1 mAcm-2 current density at 85 °C while flowing H2 on a 15 cm2 electrode, whereas MoO3electrode size of 5 cm2showed 71.8 mA cm-2current density at 150 °C in H2gas feed.

[0057] A 70-hour fuel cell performance test at a constant cell voltage of 0.40 Vdemonstrated the stability and performance of embodiments of the invention for the recovery of H2 at the anode in highly toxic gases. The durability test was performed on MoO3-B (2.6 mg cm-2) since current density for MoO3-A (1.5 mg cm-2) remained constant outside of testparameters. As seen in FIG. 12 left side, the current density of MoO3-B MEA remainedconstant for a few minutes and decreased slowly over time. Power density at the start was 32 mW cm-2and declined more than 40% in the first 24 hours to 18 mW cm-2. Fuel cell performance durability with H2CO / C2H2feed gas mixture at a constant 150 C temperature was affected slowly and increased carbon corrosion, which provides additional channels for mass transfer.

[0058] The catalyst active phase of -MoO3 stabilized over the measured time, losing~16% of the activity in the next 24 hours at 0.40 V, and showed a power density of 15 mW cm-2. Further, the catalyst layer showed CO tolerance and did not lose stability, declining to 13 mW cm-2power density on the 70thhour. Typically, high hydrogen concentration leads toalmost complete hydrogenation of acetylene to ethylene and ethane byproducts, but GC data, shown in FIG. 13, indicated that no byproducts were formed. During the MEA testing with both MoO3-A and MoO3-B, either hydrogen gas (FIG. 13A) or hydrogen with carbon monoxide gas (FIG. 13B) mixture flowed and mixed with acetylene before feeding to the MoO3catalyst at the anode side. The GC data demonstrates that the catalyst has active sites for hydrogen separation from acetylene and recovers 42% of H2for a longer time. In the case of the longer durability test, on the 70thhour, 12% of hydrogen gas was recovered; a higher percentage of hydrogen recovery compared with Pt / C catalyst recovery.

[0059] Finally, the written description uses examples to disclose the invention,including the best mode, and also to enable any person skilled in the art to practice the invention, including making and using any devices or systems and performing any incorporated methods. The patentable scope of the invention is defined by the claims, and may include other examples that occur to those skilled in the art. Such other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal language of the claims, or if they include equivalent structural elements with insubstantial differences from the literal language of the claims.

[0060] As used herein, an element or step recited in the singular and proceeded withthe word "a" or "an" should be understood as not excluding plural of said elements or steps, unless such exclusion is explicitly stated. Furthermore, references to "one embodiment" of the present invention are not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features. Moreover, unless explicitly stated to the contrary, embodiments "comprising," "including," or "having" an element or a plurality of elements having a particular property may include additional such elements not having that property.

[0061] Since certain changes may be made in the above-described invention, withoutdeparting from the spirit and scope of the invention herein involved, it is intended that all the subject matter of the above description shown in the accompanying drawings shall be interpreted merely as examples illustrating the inventive concept herein and shall not be construed as limiting the invention.

Claims

CLAIMS 1. A membrane electrode assembly, comprising: an anode; an anodic gas diffusion layer coupled to the anode; an anodic catalyst layer on the anodic gas diffusion layer, wherein the anodic catalyst layer is a mesoporous molybdenum oxide of the formula MoO3; a cathode a cathodic gas diffusion layer coupled to the cathode; a cathodic catalyst layer on the cathodic gas diffusion layer; and a proton exchange membrane sandwiched between the anodic catalyst layer and the cathodic catalyst layer.

2. The membrane electrode assembly according to claim 1, wherein the proton exchange membrane is a sulfonated tetrafluoroethylene-based fluoropolymer-copolymer.

3. The membrane electrode assembly of claim 1, wherein the anodic catalyst layer has aloading in a range from 1.5 mg cm-2to 2.6 mg cm-2.

4. The membrane electrode assembly of claim 1, wherein the cathodic catalyst comprisesPt / C.

5. The membrane electrode assembly of claim 1, wherein the membrane electrodeassembly is a component of a proton-exchange membrane fuel cell.

6. The membrane electrode assembly of claim 1, wherein the mesoporous molybdenumoxide has a hexagonal and an orthorhombic crystal structure.

7. The membrane electrode assembly of claim 1, wherein the mesoporous molybdenumoxide has an agglomerated nanorod-shaped morphology.

8. The membrane electrode assembly of claim 1, wherein the mesoporous molybdenumoxide comprises a slit / wedge mesoporous structure.

9. A method of preparing a catalyst, comprising:forming a first solution by dissolving ammonium heptamolybdenum tetrahydrate in water at an elevated temperature; forming a second solution by dissolving [poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol) with tetrahydrofuran solution ; adding water statically to the second solution when the second solution is clear; adding the first solution to the second solution forming a third solution; lowering pH of the third solution at room temperature to a pH of about 4 to about 4.5 to form a precipitate;heating the precipitate to create a blue-colored powder; collecting the powder and calcining under heating cycles; and mixing the calcined power with an alcohol / water solution and a water based sulfonated tetrafluoroethylene-based fluoropolymer-copolymer to form a homogenous solution resulting in the catalyst.

10. The method of preparing the catalyst of claim 9, wherein the heating cycles are 150 °C for 12 h at 2°C min-1, 250 °C for 4 h at 1 °C min-1, 300 °C for 3 h at 1°C min-1, and 400 °C for 2 h at a 1 °C min-1ramp rates.

11. The method of preparing the catalyst of claim 9, wherein lowering the pH comprises adding acetic acid.

12. A method of forming a membrane electrode assembly, the method comprising: coating a catalyst ink comprising mesoporous MoO3onto a gas diffusion layer to form a catalyst layer thereon; and hot pressing the catalyst layer to a proton exchange membrane.

13. The method of claim 12, wherein the proton exchange membrane comprises a sulfonated tetrafluoroethylene-based fluoropolymer-copolymer.

14. The method of claim 12, wherein the gas diffusion layer comprises a carbon paper or carbon cloth.

15. The method of claim 12, wherein the catalyst ink is comprised of 1-promanol, water, tetrafluoroethylne-based fluroropolymer-copolymer, and the catalyst of claim 1.

Images