Teaming up main group metals with metallic iron to boost hydrogenation catalysis

Hydrogenation is a foundational transformation in both heterogeneous and homogeneous catalysis, historically dominated by late noble metals such as Pt, Pd, Rh, Ir. Current efforts seek sustainable alternatives using abundant 3d or main group metals. Early main group (alkaline earth, Ae) catalysis has advanced via soluble hydride clusters formed from amide precursors, enabling alkene and imine hydrogenation through monohydride cycles under relatively mild conditions. Recently, metallic barium activated by metal vapor synthesis (MVS-Ba) showed unexpectedly strong hydrogenation activity, proposed to arise from a dual-site mechanism at Ba⁰/BaH₂ interfaces where Ba⁰ activates unsaturated substrates and polar Ba–H species deliver hydride. Building on this, the study asks whether combining reactive main group hydrides with a transition metal surface—specifically abundant, biocompatible Fe⁰ known to activate unsaturated bonds—can synergistically boost hydrogenation catalysis and expand scope to the most challenging substrates, including non-activated arenes.

• Heterogeneous hydrogenation traditionally employs Pt/Al2O3, Pd/C, Raney-Ni; robust under forcing conditions for challenging substrates such as arenes. Homogeneous systems include Wilkinson’s Rh and Crabtree’s Ir catalysts, the latter capable of tetrasubstituted alkene reductions.
• Mechanistic paradigms in transition-metal hydrogenation include dihydride and monohydride cycles; modern developments: early/late bimetallic cooperativity, ligand–metal cooperativity, cooperative H-atom transfer (cHAT), photoactivation, and metal-free FLP hydrogenation.
• Early main group catalysis emerged with soluble Ca hydride complexes leading to monohydride hydrogenation cycles. Amide precursors AeNʺ2 (Nʺ = N(SiMe3)2) generate thermally robust amide–hydride clusters (Ae_nH_xNʺ_y) under H2, active at 120–140 °C; activity increases Ca < Sr < Ba. Bulky amides increase reactivity; Ba[N(Si^iPr3)2]2 extended scope to tetrasubstituted alkenes and even slow benzene hydrogenation.
• Metallic barium (MVS-Ba⁰) alone is an effective hydrogenation catalyst; prior work proposed concurrent Ba⁰/BaH2 cycles: (i) soluble Ba hydride clusters effect insertion/hydrogenolysis; (ii) Ba⁰ oxidative addition to activated alkenes/imines followed by H2 addition. High activity hypothesized at Ba⁰/BaH2 interfaces (dual-site), consistent with Wright and Weller’s early Ae/AeH2 studies and evidence for Ba⁰ surface adsorption and possible d→π* backbonding in heavier Ae systems.
• Iron is abundant and prominent in homogeneous and heterogeneous hydrogenations and in enzymatic hydrogenations, motivating its selection as the surface partner in cooperative catalysis.

• Catalyst preparation: MVS-activated Ba⁰ prepared by co-condensation of barium metal with n-heptane; MVS-activated Fe⁰ prepared by co-condensation of Fe with toluene to give Fe(toluene) intermediates that decompose above −60 °C to a fine black Fe⁰ powder. Both powders are uncapped, extremely air-sensitive, and pyrophoric.
• Characterization: Elemental analysis and powder X-ray diffraction (sealed capillaries) confirmed metallic state; nanoparticle size ~5 nm. MVS-Fe⁰ shows bcc α-Fe; MVS-Ba⁰ shows fcc β-Ba (metastable). SEM/TEM indicate porous agglomerates of sub-10 nm particles; XPS shows partial surface oxidation from handling.
• Catalytic testing: Hydrogenations of alkenes (terminal, internal, cyclic, tri-/tetrasubstituted), arenes (benzene, toluene, p-xylene), polycyclic aromatics (naphthalene, biphenyl, anthracene, acenaphthylene), alkynes, imines, and heteroaromatics (quinoline) were conducted varying catalyst loadings, H2 pressures (6–50 bar), temperatures (RT–150 °C), and times. Activities compared for Fe⁰ alone, Ba⁰ alone, and mixed Ba⁰/Fe⁰ (ground combination noted). TOFs reported at (near) full conversion; additional data in Supplementary Tables.
• Operational parameters: Assessed dependence on H2 pressure (including constant low 6 bar) and substrate concentration; catalyst loading down to 0.05 mol%. Optimized Ba:Fe ratios and effect of mechanical grinding (mortar/pestle) on activity.
• Recycling: Post-reaction, the heterogeneous catalyst was magnetically separated from the reaction mixture, washed, and reused; activity monitored across cycles.
• Variations: Explored other Ae metals (Mg, Ca, Sr) as metals with Fe; tested commercially available AeH2 (BaH2, CaH2, SrH2) with Fe⁰; tested Ae amides AeNʺ2 with Fe⁰; evaluated alkali metal amides (MNʺ, M = Li, Na, K); used a defined soluble Mg hydride complex [(BDI)MgH]2 with Fe⁰ to probe solid–solution cooperativity; specific challenging substrate Me2C=CMe2 tested with MgNʺ2/Fe⁰.
• Mechanistic probes: EDX mapping of fresh and spent BaFe mixtures (surface elemental ratios), p-XRD and XPS of fresh/spent catalysts, NMR of mother liquors to detect soluble species (and rule out paramagnetic Fe), mercury poisoning tests, assessment of rate order with respect to H2 and substrate.

• Strong synergistic catalysis: Equimolar Ba⁰/Fe⁰ mixtures are up to three orders of magnitude more active than Ba⁰ alone. For 1-hexene at only 0.05 mol% BaFe, full conversion in 15 min with TOF ≈ 8000 h⁻¹ vs Ba⁰ alone TOF ≈ 7 h⁻¹.
• Substrate scope expansion: Efficient hydrogenation of internal alkenes following expected order (cyclic > cis > trans > trisubstituted). Tetrasubstituted alkene Ph2C=CPh2 reduced quantitatively within ~1 h; slight arene ring reduction indicates strong arene hydrogenation capability.
• Arene hydrogenation: Fe-only catalysts are inactive; Ba⁰ alone requires ~10 mol% and ≥6 days for benzene. Under similar conditions, BaFe (3 mol%) achieves quantitative benzene-to-cyclohexane conversion within 0.5 h at 50 bar H2 and 150 °C. Electron-rich arenes (toluene, p-xylene) fully hydrogenated, with rates decreasing with increased alkylation.
• Polycyclic aromatics: BaFe hydrogenates naphthalene and biphenyl fully; alters selectivity in anthracene (terminal rings reduced catalytically vs Ba⁰ central ring stoichiometric reduction). Acenaphthylene fully converted in minutes; stepwise three-ring reduction controllable by conditions.
• Other functionalities: Alkynes and imines hydrogenated (under harsher conditions, phenyl substituents on imines also reduced). Quinoline ring fully hydrogenated within 12 h. Ketones were not hydrogenated under tested conditions.
• Catalyst robustness and operation: BaFe operates up to 150 °C; with higher loading/longer time, full conversion at room temperature. Effective at constant low H2 pressure (6 bar); rates independent of H2 pressure and substrate concentration.
• Composition and processing effects: Activity in benzene reduction increases with Fe content, optimal near Ba:Fe ~1:1. Mechanical grinding increases TOFs by ~10× and renders the mixture fully magnetic (indicating intimate contact). Ba and Fe do not form alloys (confirmed by p-XRD/XPS). Commercial Fe powder inactive; pyrophoric Fe from Fe-oxalate decomposition boosts Ba⁰ but is inferior to MVS-Fe⁰.
• Recycling: Catalyst magnetically separated and reused without significant loss of activity; no dissolved salts/complexes detected in mother liquor; 1H NMR shows no paramagnetic line broadening (excludes soluble Fe species at detectable levels).
• Generality across main group components: Mixed Fe⁰ with Mg⁰, Ca⁰, Sr⁰ also active within similar magnitude to BaFe. With Fe⁰, BaH2 becomes catalytically competent (29% conversion at 1.5 mol% in benzene hydrogenation over 2 h, TOF ≈ 10 h⁻¹), while CaH2 and SrH2 remain essentially inactive. Ae amides with Fe⁰ are effective; BaNʺ2/Fe⁰ shows 86% conversion (1.5 mol%, 2 h), MgNʺ2/Fe⁰ achieves full reduction of the highly challenging Me2C=CMe2 (3 mol%, 150 °C, 50 bar, 6 h, TOF ≈ 6 h⁻¹). Alkali amides with Fe⁰ show increasing activity Li < Na < K but are less active than Ae amides.
• Soluble hydride partner efficacy: Soluble [(BDI)MgH]2 alone is inactive for benzene hydrogenation but with Fe⁰ achieves >99% conversion under standard conditions (1.5 mol%, 150 °C, 50 bar, 2 h), comparable to Ba⁰/Fe⁰. After magnetic Fe⁰ removal, only [(BDI)MgH]2 is detectable by 1H NMR.
• Mechanistic indicators: EDX mapping shows Ba surface enrichment (Ba/Fe ~2/1) in spent catalyst vs ~1/1 initially, consistent with Ba solubilization during catalysis and re-precipitation post-reaction. Hg poisoning suppresses activity, supporting a heterogeneous component. Rate independence from H2 and substrate aligns with surface site-limited (heterogeneous) kinetics. Spent catalysts contain Ba⁰ and hydride functions (p-XRD/XPS).

The findings demonstrate that coupling a reactive main group hydride component with a heterogeneous Fe⁰ surface creates a cooperative hydrogenation system far outperforming either component alone. The Fe surface activates H2 and unsaturated substrates (including arenes) via d→π* backbonding and dissociative adsorption of H2, while the main group component (Ba or soluble Mg hydride) serves as a potent hydride donor operating through insertion and hydrogenolysis steps typical of monohydride cycles. The dramatic TOF enhancements, expanded substrate scope to non-activated arenes, and altered selectivities in polycyclic systems address the central hypothesis that main group catalysts can be boosted by transition metal surfaces. Kinetic insensitivity to H2 pressure and substrate concentration, Hg poisoning, and catalyst recyclability indicate a heterogeneous rate-determining step at the surface. Concurrent evidence (EDX surface enrichment, effective catalysis with soluble hydride complexes in the presence of Fe⁰, NMR detection of soluble hydrides but not Fe species) supports a solid–solution interface mechanism in which the main group hydride operates in solution in concert with the Fe surface. The proposed synergy includes multiple roles for Fe⁰: facilitating H–H cleavage and BaH2 formation, π-backbonding activation of substrates, aiding hydrogenolysis via surface H radicals, enhancing H2 desorption from BaH2, and potential electronic promotion by Ba⁰ (analogous to K- or Cs-promotion in Fe-catalyzed N2 activation).

A simple physical mixture of Ba⁰ and Fe⁰ functions as a highly potent, recyclable hydrogenation catalyst with activities surpassing each individual component, expanding the scope to demanding substrates such as non-activated arenes and multi-substituted alkenes. The strategy generalizes across alkaline earth and related systems: Mg⁰/Ca⁰/Sr⁰ with Fe⁰ show activity; BaH2/Fe⁰ and Ae amides/Fe⁰ combinations become competent; and a well-defined soluble Mg hydride complex [(BDI)MgH]2 with Fe⁰ efficiently hydrogenates benzene. The data support a cooperative mechanism at the interface between a homogeneous main group hydride and a heterogeneous Fe⁰ surface, with Fe enabling H2 and substrate activation and Ba providing reactive hydride equivalents and possible electronic promotion. Future work should further elucidate interfacial mechanisms, explore broader metal pairings and ligand-stabilized hydrides, and develop fully homogeneous analogues to disentangle mechanistic steps and enhance catalyst design for sustainable hydrogenation based on abundant metals.

• Mechanistic ambiguity inherent to heterogeneous/homogeneous interfaces: the exact nature of active sites and intermediates cannot be fully resolved; trace soluble Fe species cannot be completely excluded.
• The powders are extremely pyrophoric and partially oxidize during handling, complicating detailed surface characterization and microscopy.
• Activity relies on finely activated Fe⁰; commercial Fe powders are inactive, and non-MVS pyrophoric Fe is less effective than MVS-Fe⁰.
• Substrate limitations remain: ketones were not hydrogenated; synergy is smaller for substrates containing heteroatoms compared to alkenes/arenes.
• Commercial CaH2 and SrH2 with Fe⁰ are essentially inactive, likely due to lattice energy/solubility issues, indicating dependence on forming soluble hydride species.
• The system is compositionally heterogeneous; Ba and Fe do not alloy, and post-reaction redistribution (Ba enrichment at surface) suggests dynamic speciation difficult to control or quantify.