Publication date: 21st July 2025
The design of nanocatalysts that combine stability, activity, and selectivity under mild conditions is a key challenge in thermal and photo-assisted hydrogenation reactions. Here, we present a hybrid catalytic environment consisting of Pd nanoparticles coordinated with ionophilic phosphine ligands and confined in an ionic liquid. This architecture resembles a “quasi nano-frustrated Lewis pair,” where electron donation from phosphine increases Pd electron density while the ionic liquid forms dynamic cages around the nanoparticles, modulating mass transport and reactivity. Solid-state NMR and XPS confirm strong Pd–phosphine coordination, while DFT calculations reveal that one phenyl ring aligns nearly parallel to the Pd facets, interacting through its π system. These hybrid features translate into remarkable catalytic activity in selective semi-hydrogenation, achieving TOFs of 3.85 s⁻¹ for phenylacetylene, 0.8 s⁻¹ for 2-cyclohexen-1-one, and 12.82 s⁻¹ for 1,3-cyclohexadiene at 40 °C and 2–4 bar H₂ in BMIm.NTF₂. The enhanced performance arises from (i) ionic liquid cages acting as catalytic boundaries that regulate diffusion, and (ii) steric hindrance from Pd–P coordination directing substrate access. Importantly, this metal–organic–ionic hybrid interface provides a transferable platform for the design of photo-assisted catalytic systems, where light can further couple with electronic and steric effects to tune reactivity. This study highlights how hybrid nanocatalysts can bridge thermal and light-driven transformations, offering new insights into materials, mechanisms, and structure–activity relationships relevant to photochemistry.
G.C-R Acknowledges to FGCSIC ComFuturo/Horizon 2020 program (Marie Skłodowska-Curie grant agreement No. 10103426) by a fellowship of the “ComFuturo program” at the Institute of Chemical Technology from UPV-CSIC (Spain). G.A acknowledges to FONDECYT 1230991 and ECOS-ANID (ECOS230008) projects for financial supporting. H.C.B.O thanks financial support by LaMCAD/UFG. R.B.P. acknowledges CNPq (grant 309599/2021-0) for financial support as well as CENAPAD-SP, supercomputer SDumont/LNCC-MCTI and Cluster Euler/CeMEAI for providing the computational resources. J.D Acknowledges to CAPES (001), FAPERGS (22/2551-0000386-9 and 18/2551-0000561-4) and CNPq (406260/2018-4) for financial support.