From gangue to the fuel-cells application

The study addresses the challenge of safe, efficient hydrogen storage for a future hydrogen economy. Conventional methods such as high-pressure gas (up to 700 bar) and liquid hydrogen require heavy, costly systems with safety and energy penalties, while cryogenic pressure vessels combine high pressure with low temperature but remain complex. Solid-state storage via metal hydrides offers an alternative, with magnesium hydride (MgH2) being attractive due to its abundance, low weight, and high theoretical capacity (~7.6 wt%), but limited by sluggish sorption kinetics and relatively high thermal stability (high decomposition temperatures and activation energies). Prior work has improved MgH2 via severe plastic deformation (e.g., cold rolling, ECAP, HPT), nanostructuring by high-energy ball milling, and catalytic doping with transition metals and alloys. Amorphous (metallic glass) catalysts have shown promise but are less explored. This work proposes integrating cold rolling (CR), reactive ball milling (RBM), and doping with amorphous Zr2Cu derived from solid-waste metals to enhance MgH2 kinetics and cyclability, and demonstrates practical utility by powering a PEM fuel cell.

• Conventional hydrogen storage: compressed gas tanks (low density, safety concerns) and liquid hydrogen (large tank volume, cryogenic complexity and liquefaction energy cost ~25–30% of stored energy). A cryogenic pressure vessel concept combines high pressure (~350 bar) with T < 20 K to increase density.
• Solid-state hydrides: Since Graham’s discovery with Pd, many metals store hydrogen via physisorption/chemisorption; Mg is highlighted for high capacity (7.6 wt%, 0.11 kg H2/L) but with slow kinetics and high stability.
• Mechanical processing to destabilize MgH2 and refine grains: high-energy ball milling, ECAP, HPT, cold rolling. Formation of metastable MgH2 phases (orthorhombic, fcc, γ-MgH2) improves kinetics.
• Catalytic doping: Transition metals (Ti, Zr, V, Nb, Fe, Co, Ni, Mn) and alloys (e.g., TiV, TiMn, VTiCr, ZrNi5) reduce sorption temperatures and accelerate kinetics. Nanocomposites (e.g., MgH2/Ni/Nb2O5; MgH2/TiC; MgH2/big-cube Zr2Ni) show superior behavior.
• Amorphous catalysts: Limited reports (e.g., amorphous Zr70Ni20Pd10) indicate significant kinetics improvement; further exploration is needed.
• BMAS (ball milling with aerosol spraying) methods have produced LiBH4/MgH2 nanocomposites with ~5.7 wt% reversible H2 at ~265 °C, highlighting alternative processing strategies. This study builds on these by using solid-waste-derived amorphous Zr2Cu as a hard, thermally stable catalytic medium combined with severe plastic deformation and RBM.

Feedstock preparation: Solid-waste metals were cleaned (acetone sonication, ethanol rinse, 180 °C drying). Mg chips and Cu plates were melt-cast under purified Ar; Mg ingot purity ~99.88 wt% (O2 ~0.11 wt%), Cu ~99.994 wt%. Zr scrap was refined by arc melting with Ti-getter under Ar to ~98.7 wt% purity.

Mg ribbon fabrication and mechanical processing: Mg ingot pieces were melt-spun (PA 500, He atmosphere, 5000 rpm Cu wheel) into ribbons (~1.3 cm width, ~3 m length, ~0.5 cm thickness). Ribbons were cold rolled (CR) for up to 150 passes with intermittent warm pressing (150 °C) to reduce brittleness, achieving ~112% elongation and ~108 µm thickness; ribbons were cut into shots.

Reactive ball milling (RBM) to form MgH2: 50 g Mg shots plus tool-steel balls (ball-to-powder 36:1) were sealed in a tool-steel vial under He, evacuated (10^-6 bar), pressurized with 15 bar H2, and milled in a tumbling mill at 250 rpm at room temperature. Milling times: 3–50 h with intermediate sampling to track phase evolution.

Amorphous Zr2Cu (a-Zr2Cu) catalyst preparation: A master alloy Zr67Cu33 (nominal Zr2Cu) was arc-melted under Ar with multiple remelts to ensure homogeneity (final composition ~Zr 74.08 wt%, Cu 24.96 wt%). The alloy was disintegrated and mechanically disordered (planetary mill, tool-steel vial, 250 rpm, ball-to-powder 20:1) for up to 50 h to convert crystalline Zr2Cu to amorphous; confirmation by XRD and HRTEM.

Nanocomposite formation: As-RBM MgH2 powders (50 h) were doped with 5 wt% a-Zr2Cu, mixed under He, and further RBM in planetary mill (PULVERISETTE 6) using 11 mm tool-steel balls (ball-to-powder 40:1) with gas-temperature monitoring (evico magnetic). Additional compositions (3, 7, 10 wt% a-Zr2Cu) were also prepared for kinetics comparisons.

Characterization: Crystal structure by XRD (Cu Kα, SmartLab-Rigaku 9 kW). Microstructure by 200 kV FE-HRTEM/STEM (JEOL-2100F) with EDS; CR ribbon TEM prepared by Cryo Ion Slicer. Morphology and elemental mapping by FE-SEM (JEOL JSM-7800F) with EDS. Thermal analysis by DSC (Setaram) for MgH2 decomposition and a-Zr2Cu glass transition/crystallization; melting behavior by DTA. Kinetics evaluated via Sieverts method (PCT-Pro2000) at 250 °C under 10 bar H2 (absorption) and 400 mbar (desorption). Activation energies determined from DSC peak shifts at multiple heating rates using Arrhenius analysis.

Fuel-cell demonstration: 300 g of MgH2/5 wt% a-Zr2Cu (50 h RBM) were loaded into a custom tool-steel hydrogen reactor tank equipped with gauge and high-pressure valve, evacuated to ~10^-2 bar, pressurized to 35 bar H2 for activation, then heated to ~250 °C to release H2. Gas was pumped (rotary pump) at ~130 ml/min to a PEM fuel cell unit (rated 500 W/15 A) to charge a cell phone battery and operate a 3 V prototype car via a 5 V regulator. Voltage and power output were logged.

• Phase formation and nanostructure: RBM of CR Mg under 15 bar H2 produced nanocrystalline β- and γ-MgH2; after 50 h, Mg peaks vanished and MgH2 peaks broadened (nanocrystallinity; grains <20 nm). With 5 wt% a-Zr2Cu and additional RBM, MgH2 grains were refined to ~6.5–8 nm; a-Zr2Cu was homogeneously distributed as nano-lenses (~200 nm clusters) across MgH2 matrix.
• Thermal stability (decomposition temperature, Tdec): CR alone reduced Tdec from 434 °C (0 passes) to 360 °C (150 passes). CR + RBM (150-pass ribbons) further reduced Tdec to 324, 317, 306, 292, and 281 °C at RBM times 6, 12.5, 25, 37.5, and 50 h, respectively. Adding 5 wt% a-Zr2Cu lowered Tdec further to 265, 256, 249, 239, and ~237 °C at the same times, saturating near 237 °C at 50–75 h.
• Activation energy (Ea) for dehydrogenation (Arrhenius from DSC): as-cast Mg ~162.37 kJ/mol; melt-spun (MS) ~153.61 kJ/mol; CR 150 passes ~124.67 kJ/mol; CR + RBM 50 h ~96.82 kJ/mol; CR + RBM + 5 wt% a-Zr2Cu 50 h ~88.75 kJ/mol.
• Kinetics at 250 °C (MgH2/5 wt% a-Zr2Cu): Absorption under 10 bar H2 reached 2.00, 3.15, 3.16 wt% in 30 s for 6, 25, 50 h RBM; at 1 min: 3.5, 4.0, 4.48 wt%; at 5 min: 5.53, 6.35, 6.25 wt%; at 10 min: 6.14, 6.57, 6.56 wt% H2. Desorption under 400 mbar achieved ~6.66 wt% H2 released within 10 min for all three RBM times; at 5 min: ~4.56 (6 h), 6.77 (25 h), 6.78 (50 h) wt% released.
• Catalyst loading effect (3–10 wt% a-Zr2Cu): Higher a-Zr2Cu loadings accelerated sorption/desorption but slightly reduced reversible capacity; 10 wt% a-Zr2Cu released ~6.63 wt% H2 within 20 min at 250 °C, lower loadings were slower.
• Cyclability: CR-only samples showed capacity fade from ~4.52 to 3.83 wt% over 180 h at 275 °C (~96 cycles). CR + RBM improved cycling to ~113 cycles/180 h. The optimized MgH2/5 wt% a-Zr2Cu sustained continuous cycling for ~1100 h: first ~800 h at 250 °C (10 bar/400 mbar) with stable reversible capacity ~6.5–6.6 wt%; subsequent 300 h at 175 °C retained ~6.45 wt% capacity but with slower kinetics.
• Microstructural evolution on cycling: After extended cycling, nanosized fcc-ZrH2 (~11 nm) and fcc-Cu formed via partial devitrification/phase separation of a-Zr2Cu, coexisting with β-MgH2/hcp-Mg and residual amorphous Zr2Cu. Amorphous Zr2Cu formed a nanoscale “shield” around Mg particles, acting as grain growth inhibitor and enhancing hydrogen transport via “nano-hydrogen gates.”
• Fuel-cell demonstration: At ~250 °C, H2 release raised reactor pressure to ~25 bar within ~500 s; using a rotary pump, a steady H2 flow ~130 ml/min delivered to a PEM FC produced stable 4–8 V and ~5 W over ~10,000 s, sufficient to charge a phone battery and power a prototype car.

The integrated processing strategy effectively addresses MgH2’s intrinsic kinetic and thermodynamic limitations. Cold rolling and RBM generate extensive lattice defects and nanoscale grains, shortening hydrogen diffusion paths and facilitating phase transformations (β ↔ γ-MgH2), thereby lowering decomposition temperatures and activation energy. Introducing hard, thermally stable amorphous Zr2Cu nanoparticles enhances milling efficiency (further grain refinement), inhibits grain growth during cycling, and improves thermal conductivity of the composite, collectively accelerating sorption/desorption. Partial devitrification during prolonged cycling yields nanoscale fcc-ZrH2 and fcc-Cu, which likely modify the electronic environment and increase heat transfer within the MgH2 matrix, further promoting kinetics without sacrificing capacity. The composite maintains high reversible capacity (~6.6 wt%) and exhibits exceptional long-term stability (1100 h), demonstrating feasibility for practical application. The successful operation of a PEM fuel-cell system using the released hydrogen validates the material’s applicability as a solid-state hydrogen source.

This work demonstrates a solid-waste-to-energy approach that converts waste Mg, Zr, and Cu into a high-performance MgH2/5 wt% amorphous Zr2Cu nanocomposite. The combination of cold rolling, reactive ball milling, and amorphous Zr2Cu doping produces ultrafine MgH2 grains and stable catalyst distribution, achieving: (i) low decomposition temperature (~237 °C), (ii) reduced activation energy (~88.75 kJ/mol), (iii) fast sorption/desorption kinetics (~6.6 wt% within ~10 min at 250 °C), and (iv) outstanding cyclability over 1100 h. Partial formation of nanoscale ZrH2 and Cu during cycling further contributes to kinetics and stability by enhancing thermal transport and inhibiting grain growth. The composite successfully powered a PEM fuel cell for device demonstrations. Future work should optimize catalyst loading and distribution, push operating temperatures lower while maintaining kinetics, scale up processing with energy-efficient milling protocols, and investigate long-term phase stability (especially ZrH2) and potential oxygen contamination control for practical storage systems.

• Operating temperature remains moderate (typically 250 °C) for optimal kinetics; significant slowing occurs at 175 °C despite good capacity.
• Small MgO formation observed during ex situ handling indicates sensitivity to oxidation; rigorous inert handling is required.
• Partial devitrification of amorphous Zr2Cu to ZrH2 and Cu occurs over long cycling; while beneficial to kinetics, long-term structural evolution and its impact require further study.
• Demonstrated fuel-cell output (~5 W) was a proof-of-concept with 300 g powder; scaling to higher power-density systems and on-board applications requires engineering optimization.
• Energy and cost implications of extended high-energy milling were not quantified.
• Exact role and stability of γ-MgH2 during extended cycling warrants deeper in situ analysis.