Highly-efficient RuNi single-atom alloy catalysts toward chemoselective hydrogenation of nitroarenes

The design and exploitation of high-performance catalysts have gained considerable attention in selective hydrogenation reactions, but remain a huge challenge. Herein, we report a RuNi single atom alloy (SAA) in which Ru single atoms are anchored onto Ni nanoparticle surface via Ru–Ni coordination accompanied with electron transfer from sub-surface Ni to Ru. The optimal catalyst 0.4% RuNi SAA exhibits simultaneously improved activity (TOF value: 4293 h–1) and chemoselectivity toward selective hydrogenation of 4-nitrostyrene to 4-aminostyrene (yield: >99%), which is, to the best of our knowledge, the highest level compared with reported heterogeneous catalysts. In situ experiments and theoretical calculations reveal that the Ru–Ni interfacial sites as intrinsic active centers facilitate the preferential cleavage of N–O bond with a decreased energy barrier by 0.28 eV. In addition, the Ru–Ni synergistic catalysis promotes the formation of intermediates (C8H7NO* and C8H7NOH*) and accelerates the rate-determining step (hydrogenation of C8H7NOH*).

S3 C·min -1 ) for 1 h, followed by cooling to the room temperature in a high purity He stream, and then the background signal was collected. Afterwards, the CO/He (1/19, v/v; 30 mL•min −1 ) was purged into the cell, and DRIFTS spectra were collected until the adsorption spectrums kept unchanged. Finally, the gas flow was switched to a pure He stream to collect CO chemisorption spectra. In situ FT-IR measurements of 4-NS adsorption and surface reaction were performed using a transmission reactor. The sample (20 mg) was pressed into self-supporting wafer with a diameter of 13 mm, followed by a pretreatment under the same conditions. After the sample was cooled down to 60 °C in He stream, 4-NS was introduced into the reactor for 30 min. Subsequently, He was purged to remove the physically adsorbed molecule followed by collection of IR signals.
Finally, the spectra for hydrogenation process were collected per 60 s after the introduction of H2 (flow rate: 30 mL·min −1 ).

Computational details
We performed first-principle calculation within the DFT methodology by using the Vienna ab initio simulation package (VASP 5.4) 3,4 . The exchange and correlation energy were calculated with the generalized gradient approximation (GGA) of the density functional theory. The core electrons are described with the projector augmented wave (PAW) method 5,6 . Lattice parameters of bulk Ni were optimized by the PBE-D3 7,8 , PBEsol 9 , PBE 10  convergence criterion for the total energy self-consistent iterations was 10 −4 eV, and the geometry optimization stopped when the total force on the system was less than 0.05 eV/Å. The energy barriers were determined by the climbing image nudged elastic band (CI-NEB) and Dimer method 13,14 .
According to the HRTEM characterization, the Ni(111) surface was modeled with a threelayer slab in a p(6 × 6) surface unit cell consisting of 108 atoms; the bottom one slab was fixed while the top two slabs were relaxed. A vacuum space of 15 Å ensures no interaction between the periodically repeated slabs or adsorbates in the direction normal to the slab. As Ni is not a nonmagnetic metal, we performed spin polarization calculation for the Ni(111) surface. The RuNi (111) surface was built by substituting one Ni atom on the topmost layer with one Ru atom.
The activation energy (Ea) and adsorption energy (Eads) are calculated by the following equations: where E TS , E IS , E A/M , E A and E M represent the energies of the TS (transition state), IS (initial state), adsorbed system, substrate, and adsorbate species, respectively.
The Bader Charge analysis was performed via Henkelman programme based on near-grid algorithm with refine-edge method 15−17 . The charge density differential analysis was performed by using visualization software VESTA 18 , with the isosurface level set as 0.005 e/Å 3 .