In Situ Construction of Interface with Photothermal and Mutual Catalytic Effect for Efficient Solar‐Driven Reversible Hydrogen Storage of MgH2

Abstract Hydrogen storage in MgH2 is an ideal solution for realizing the safe storage of hydrogen. High operating temperature, however, is required for hydrogen storage of MgH2 induced by high thermodynamic stability and kinetic barrier. Herein, flower‐like microspheres uniformly constructed by N‐doped TiO2 nanosheets coated with TiN nanoparticles are fabricated to integrate the light absorber and thermo‐chemical catalysts at a nanometer scale for driving hydrogen storage of MgH2 using solar energy. N‐doped TiO2 is in situ transformed into TiNxOy and Ti/TiH2 uniformly distributed inside of TiN matrix during cycling, in which TiN and Ti/TiHx pairs serve as light absorbers that exhibit strong localized surface plasmon resonance effect with full‐spectrum light absorbance capability. On the other hand, it is theoretically and experimentally demonstrated that the intimate interface between TiH2 and MgH2 can not only thermodynamically and kinetically promote H2 desorption from MgH2 but also simultaneously weaken Ti─H bonds and hence in turn improve H2 desorption from the combination of weakened Ti─H and Ti─H bonds. The uniform integration of photothermal and catalytic effect leads to the direct action of localized heat generated from TiN on initiating the catalytic effect in realizing hydrogen storage of MgH2 with a capacity of 6.1 wt.% under 27 sun.

samples were pre-grinded and calcined at 500 °C for 3 h to obtain the final products as flowerliked TiO2 microspheres.Synthesis of carbon coated TiN@N-TiO2-x (TNTx) The as-synthesized TiO2 was mixed with melamine (C3H6N6, CP, ≥ 99.0%) at a mass ratio of 1:x by facile grinding and then calcined at 750 °C for 3 h with a heating rate of 3 °C min -1 at a flow of 5% H2/N2.The thus-obtained dark black products were denoted as TiN@N-TiO2-x (TNTx, x = 3, 5, 20).

Preparation of MgH2 catalyzed by TiO2, TNTx, and TiN
The TiN@N-TiO2-x catalyst with a mass ratio of 10 wt.% was mixed with MgH2 (Alfa Aesar, 98%) by mechanically milling for 12 h under a H2 pressure of 50 atm.The ball-to-powder ratio was controlled to be 100:1 and the milling speed was kept at 500 rpm during the whole process.
The resulting composite was denoted as MgH2-TNTx.For comparison, as-synthesized TiO2 and commercial TiN (Aladdin, ≥ 99.90%) were also mixed with MgH2 under identical conditions.All the operations were carried out in an Ar-filled glovebox with the water and oxygen contents below 0.01 ppm.

Characterizations
The phase composition was determined by X-ray diffraction (XRD, D8 Advance, Bruker AXS) with Cu Kα radiation (λ = 1.542Å).The samples were protected by amorphous tapes (with a broad peak around 2θ ≈ 20°) in an Ar-filled glovebox to avoid oxidation.The morphologies and element distribution were characterized by scanning electron microscopy (SEM, JEOL 7500FA, Japan) and transmission electron microscopy (TEM, JEOL JEM-2100F, Japan) equipped with an energy-dispersive X-ray spectroscopy (EDS) analysis unit.The X-ray photoelectron spectroscopy (XPS) was carried out on a Thermo Scientific K-Alpha+ system equipped with dual X-ray sources, adopting an Al Kα anode with a hemispherical energy analyzer.The UV-vis-NIR Spectrophotometer (Lambda 35, Perkin-Elmer, USA) was taken to determine the absorption spectra of samples in the wavelength range of 200~2000 nm.A 300 W Xe lamp (CEL-HXF300-T3, Beijing China Education Au-light Co., Ltd.) was used as the light source, and an optical power meter (PL-MW2000, Perfect Light Co., Ltd.) was adopted to measure the light intensity.A short-wave infrared thermometer (DGE44N, range: 75~650 °C, DIAS, Germany) was used to measure the surface temperature of the composites under light irradiation with a recording time step of 1 second.The light reactor cell (manufactured by Beijing Century Senlong experimental apparatus Co., Ltd.) was equipped with a Sapphire window with a light transmittance of ~90-95% on the top surface and several fiber glass filter on the ground to achieve effective thermal management.

Hydrogen Storage Measurements
Hydrogen storage performance of the samples was evaluated by a home-built high-pressure gas sorption apparatus (HPSA-auto), which was carefully calibrated by adopting LaNi5 as a reference sample in terms of hydrogen storage capacity and guaranteed an accuracy of ±1%.
The hydrogen gravimetric capacity was calculated on the total mass of test sample.
In the term of electric heating test, approximately 25 mg of test samples were heated from room temperature to 450 °C at different heating rate under an initial pressure below 0.0001 bar in the non-isothermal desorption process.The isothermal H2 desorption and absorption test was handled by rapidly heating up to the target temperature followed by keeping at the preset temperature.In the isothermal H2 absorption experiment, approximately 80-90 mg of samples were used for hydrogenation under the hydrogen pressure of 50 bar.
In the term of light irradiation test, 25 mg of samples were pressed into pieces with a diameter of 15 mm under a pressure of 5 tons to avoid powder spattering and ensure the stability of the test and the comparability of the experimental results.The top of the light reactor is a sapphire window with a light transmittance of more than 95%, and the light reactor was connected to the HPSA-auto to record real-time hydrogen capacity.The light intensity was adjusted by tuning the current of the light source with a fixed distance between the light source and the reactor, which could avoid the measure error caused by the different irradiation area due to the different distance.To reduce the heat loss, a layer of SiO2 glass fiber with low thermal conductivity was added to the bottom of the device and a tailor-made SiO2 glass fiber ring with an inner diameter of 16mm was introduced to restrict the pellet's movement under light irradiation in the vacuum, thereby preventing heat dissipation through the sidewall.The cycling test was carried out at a light intensity of 35 sun with an initial desorption pressure below 0.0001 bar and a H2 pressure of 50 bar for adsorption.

FDTD Simulation
To clarify the mechanism of plasmon-enhanced photothermal properties, the electric field distributions of the TiN@N-TiO2 and TiO2 were simulated by the finite-difference time-domain (FDTD) simulation performed on FDTD solutions 8.6 (Lumerical Solutions Inc., Vancouver, Canada).The profiles of electric fields (described by |E| 2 /|E0| 2 ) were obtained when the incident light was perpendicular to one of the facets [1] .The simulation structure of TiN@N-TiO2 was composed of N-TiO2 core (diameter: 40 nm), periodical TiN nanoparticles on the surface (diameter: 10 nm) and the top layer covered with amorphous carbon coating (thickness: 2 nm).
The dielectric constant of TiN, and C was obtained from the FDTD database taken from the Palik model [2] .The permittivity of N-TiO2 was obtained from previous literatures, and the dielectric functions with the minimum root-mean-square error (RMSE) of experimental results was calculated with Eq. (S1) as follow [3] using the model parameters of ellipsoid fitting: The simulation regions (x, y, z) = (-50:50, -50:50, 0) nm was divided into uniform Yee cells with perfectly matched layer boundary (PML) in x-, y-, and z-directions.An incident plane wave was propagated from the top along z axis and polarized along the x axis, and a wavelength of 540 and 1200 nm light was used for calculation.The process was performed at 1 nm mesh resolution and 500 fs time.A 2D electric field monitor located at y = 0 was used to investigate the electric field distribution across the 3D core-shell structure.

Theoretical calculations
The density functional theory (DFT) calculations were performed using the VASP code [4] .
The electron-ion interaction was modeled using the projector-augmented wave (PAW) method [5] .The exchange-correlation functional was described using the Perdew-Burke-Ernzerhof (PBE) generalized gradient approximation (GGA) [6] .A plane-wave energy cutoff of 400 eV was employed for all calculations.To minimize interactions between neighboring layers, a large vacuum slab (~26 Å) was used along the z-direction.A Gaussian smearing technique with a small SIGMA value of 0.05 was applied in all the calculations.The Wulff-like models for TiH2 and MgH2 nanoparticles were generated using the bulk cut nanoparticle model (BCN-M) computational tool [7] .Geometric optimization was performed by relaxing all atoms until the residual force was less than 0.02 eV/Å.Ab initio molecular dynamics (AIMD) simulations were carried out using the Nose-Hoover method to solve the ionic motion equations, with a time step of 1.0 fs.The system was equilibrated at 1200 K for 60 ps in the canonical ensemble (NVT).The climbing image nudged elastic bond (CI-NEB) method [8] was utilized to investigate the pathways and energy barriers associated with hydrogen release in the given system.Table S1.N atom content of TiN@N-TiO2-x samples.. Figure S9.Analysis of absorption spectra of the as-synthesized TiO2 and TNT3 for the indirect electronic transition (αhν) 1/2 as a function of the photon energy (eV).Spectral analysis is conducted using a Tauc plot and the base lines are used to determine Eg [9] .Table S2.The apparent activation energy (Ea) of MgH2 catalyzed by TiO2 ,TNTx, and TiN calculated by the Kissinger's equation [10] .Figure S24.Energy profile and Schematic illustration of H2 desorption pathway of MgH2 under the catalysis of TiO2 and TiNO [11] .

Figure S1 .
Figure S1.Schematic illustration of the preparation process of TiN@N-TiO2

Figure S4 .
Figure S4.HRTEM images and SAED patterns of the as-synthesized TiO2.

Figure S5 .
Figure S5.Elements mapping images of Ti, N, O and C in TNT5.

Figure S6 .
Figure S6.The relative elemental content analysis of TNT5.

Figure S7 .
Figure S7.Measured TG curves as a function of temperature for TNT5.

Figure S10 .
Figure S10.Theoretical FDTD simulated localized electric field enhancement profiles obtained under the wavelength of 1200 nm light, corresponding to the NIR region.

Figure
Figure S15.(a) The stable temperatures at 20 min and (b) the corresponding H2 desorption curves of MgH2 under the catalysis of TNT5 and TiO2 under various light intensities.(c, d) Isothermal H2 desorption curves of MgH2 catalyzed by TNT5 and TiO2 at various light intensity irradiation.

Figure S16 .
Figure S16.Comparison of surface temperature and dehydrogenation kinetics of MgH2 catalyzed by TNT5 under solar irradiation between the 1st and 15th cycles.

Figure
Figure S17.(a) UV-vis-NIR absorption spectra of MgH2 catalyzed by TiO2 and TNT5 after cycling compared with ball-milled state.(b) UV-vis-NIR absorption spectra of MgH2 catalyzed by Ti, TiH2 and TiO2.(c, d) The surface temperature curves and solar-driven H2 desorption curves of MgH2 catalyzed by Ti, TiH2 and TiO2 at the light intensity of 27.5 sun.

Figure S20 .
Figure S20.High-resolution (a) Ti 2p and (b) N 1s XPS spectra of MgH2 catalyzed by TNT5 after cycling under light irradiation.

Figure S25 .
Figure S25.HRTEM and elements mapping images of Mg, Ti, and N in MgH2 catalyzed by TNT5 after cycling.

Figure S27 .
Figure S27.Comparison of the temperature curves of TiN nanoparticles and MgH2 catalyzed by TNT5 over time under the various light irradiation conditions.

Table S3 .
Comparison for solar-driven reversible H2 desorption of MgH2 catalyzed by various