Reversible Hydrogen Storage in Solid‐State Reaction Derived Core‐Shell NaBH4@Ni Nanocubes

The method to synthesize high‐capacity complex hydride nanostructures, such as borohydrides, decorated with metallic shells has surfaced as an attractive approach for enabling reversible hydrogen storage. However, the current solvent‐based synthesis methods of such core‐shell nanostructures are limited by solvents/hydrides compatibility issues and the low solubility of the shell precursor in such solvents. Herein, for the first time, an alternative solid‐state method to prepare core‐shell‐like nanostructures is reported. Simply, by mixing and heating vanadium (V)‐doped sodium borohydride (V‐NaBH4) cores and nickelocene (as nickel precursor) at 150 °C, it is possible to decorate V‐NaBH4 with Ni, which shows an improved hydrogen release (≈8 mass% H2) at 350 °C and a net reversible hydrogen capacity of 2 mass%. Detailed structural investigations reveal that the in‐situ formed VxBy and NixBy are responsible for superior hydrogen absorption in the core‐shell material, where these boride species around the shell/interfaces suppress the loss of Na or B and the formation of B12H12 during hydrogen release/uptake. This work opens solvent‐free pathways to design and control the chemical composition of core–shell (boro)hydrides for practical hydrogen storage.


Introduction
Hydrogen is an appealing carbon-free and sustainable energy vector.It is the most abundant element in the universe and is being used in many industrial applications. [1]Currently, a plethora of methods are available to produce hydrogen such as steammethane reforming, [2] thermochemical and biochemical conversion of biomass, [3] and/or water-splitting. [4]However, no method exists to safely store and transport hydrogen, and this has remained the biggest issue in the implementation of the hydrogen economy. [5]Traditionally, hydrogen is stored and transported either in a gaseous state in pressurized tanks at 700 bar or in a liquid state in cryogenic containers. [6]However, these methods are associated with several drawbacks such as cost, safety issues, and/or energy penalty.On the contrary, materials-based (solid-state) hydrogen storage has been viewed as an attractive alternative. [6,7]Solidstate hydrogen storage involves the interaction of hydrogen gas with a solid host material (usually a hydride or porous scaffold) at a certain temperature and pressure.Among the various high-capacity complex hydrides such as alanates, [8] amides, [9] and borohydrides; [10] borohydrides have gathered significant attention for solid-state hydrogen storage owing to their highgravimetric hydrogen contents. [11]Within the borohydride family, sodium borohydride (NaBH 4 ) has widely been explored as a potential candidate for hydrogen storage, however, it requires a very high temperature (%500 °C) to release hydrogen, and hydrogen reversibility is impractical due to the thermodynamic barriers and structural instability (e.g., loss of active Na and B). [12]Hydrolysis is another potential route for hydrogen production from borohydrides (e.g., NaBH 4 ), [11b,13] however, due to low hydrogen capacity, high cost, inefficient hydrolysis, and issue in regenerating NaBO 2 (hydrolysis by-product), US.DOE Hydrogen Program panel unanimously issued a "No-Go" decision for NaBH 4 hydrolysis for onboard hydrogen storage. [14]onfinement of borohydrides inside porous scaffolds has been regarded as an attractive approach to stabilize nanosized borohydride particles, reduce the hydrogen desorption temperature, and suppress the loss of active species (Na and B) essential for reversibility. [15]The well-known nanoconfinement systems for NaBH 4 are nanoporous carbons, [16] carbon hollow spheres. [17]esoporous silica, [18] and CuS hollow nanospheres. [19]However, these material systems have inadequate control over borohydride nanostructure and low hydrogen storage capacities (0.6-0.8 mass %) owing to the dead mass/volume of the scaffolds.Recently, the development of core-shell (nano)hydrides in an appropriate solvent has surfaced as a potential technique to get size-and shapecontrolled borohydride nanostructures within stable metallic shells, which can achieve hydrogen reversibility to some extent. [20]or example, our group developed the core-shell NaBH 4 @M (M = Co, Cu, Ni, Sn) structures by reducing metal precursors directly on the surface of the ligand-stabilized NaBH 4 in a solvent. [21]Owing to a higher hydrogen permeability of Ni, the core-shell NaBH 4 @Ni demonstrated a reversible hydrogen capacity of 5 mass % at 350 °C, 4 MPa H 2 pressure. [20]12a,22] We also demonstrated excellent hydrogen production rates (i.e., 6.1 L min À1 g À1 at 39 °C) via hydrolysis using the core-shell NaBH 4 @Ni nanostructures. [23]These studies show the potential of a solvent-based approach to design reversible coreshell borohydride nanostructures for hydrogen storage.However, these methods have several limitations, such as: 1) a limited choice of the organic solvents in which NaBH 4 can be dispersed for coating; 2) excess amine ligand complexed with Ni 2þ , which results in an inevitable consumption or dissolution of active NaBH 4 during the core-shell synthesis; 3) inevitable self-nucleation of Ni particles in solution; and/or 4) lack of scalability.To overcome these limitations, an alternative scalable solvent-free method to produce core-shell structures is needed.
Solid-state syntheses have been widely explored in fabricating metals and metal oxide-based materials.For example, the solidstate reaction between Al-Li-Sc alloy at 450 °C produced the Al 3 Sc phase, which upon further heat treatment at 190 °C formed Al 3 Sc(core)@Al 3 Li (shell) particles. [24]Nalluri et al. also reported a large-scale solvent-less synthesis of core-shell Fe 3 O 4 @M (where M = Au, Ag, and Au-Ag alloy) nanostructures. [25]For borohydrides, solid-gas reactions for the synthesis of borohydrides from the respective metal hydride precursors have been reported. [26]Herein, for the first time, we report on a simple solid-state method to fabricate core-shell borohydride nanostructures for reversible hydrogen storage.To show the potential of this approach, V-doped NaBH 4 particles were used and then coated with Ni to generate V x B y within the core and Ni x B y species within the shell/interfaces to suppress B 2 H 6 emissions and enhance hydrogen reversibility.

Solid-State Synthesis and Structural Evaluation
The synthesis of core-shell V-NaBH 4 @Ni nanostructures involves a two-step process where the cube-like NaBH 4 particles were first synthesized using our previously reported method in the presence of octadecylamine (ODA) as a stabilizing ligand (Figure 1a). [22]Briefly, in the present work, a set amount of high-purity NaBH 4 and VCl 3 were separately dissolved in IPA at room temperature (RT), and the latter was added dropwise to the NaBH 4 solution.After overnight homogenization, ODA/DDT as stabilizing ligands were added to the mixture to get cube-like V-NaBH 4 particles.After removing IPA, the as-prepared cube-like V-NaBH 4 particles were then mixed with different concentrations (i.e., 20, 30, or 40 mass%) of NiCp 2 as the nickel precursor in an agate mortar at RT.The mixture was packed in a stainless steel vessel under argon and heated at 150 °C for 2 h to get Ni-decorated V-NaBH 4 particles.After cooling, the resulting material was washed with toluene to remove excess ligands and dried under a vacuum.
Figure 1b,c shows the TEM images of representative cube-like V-NaBH 4 (size %250 nm) and their core-shell particles obtained after coating with 30 mass% of NiCp 2 .After Ni coating, the particle started to decompose under the electron beam, which is an indication of a low-temperature H 2 release from the core-shell structures (Figure 1c).With the increase in the NiCp 2 loading from 20 to 40 mass%, the contrast (darker Ni coating over lighter grey NaBH 4 particles) of the cube-like core-shell particles also increased (Figure S1, Supporting Information).This suggests the deposition and reduction/conversion of NiCp 2 to Ni metallic or Ni x B y onto the NaBH 4 surface.After Ni coating, cube-like particles became bigger with edge lengths of %415 nm (V-NaBH 4 @Ni-20), %730 nm (V-NaBH 4 @Ni-30), and %630 nm (V-NaBH 4 @Ni-40) and morphology of particles changed to irregular shapes, and this may be due to some particle agglomeration during heat treatment (Figure 1b and S1, Supporting Information).12a,27] Since NaBH 4 was doped with VCl 3 and coated with NiCp 2 , one can imagine the presence of the destabilized borohydride core inside the Ni shell.One should also expect the formation of a hollow structure after decomposing the hydride core (leaving the shell intact), which can be a direct proof for the formation of a core-shell structure.To test this hypothesis, in our preliminary investigations, we exposed a section of a cubic particle to the focused electron beam under TEM (operating at 200 kV) for a longer duration (>2 min).The particle under the electron beam evolved and decomposed (lighter material corresponding to the doped-borohydride core) while the decomposed product was inside the darker shell (Figure S2a-c, Supporting Information).In contrast, an uncoated particle lost its morphology without leaving a clear shell (Figure S2d, Supporting Information).12a,27] X-ray diffraction (XRD) was further performed to investigate the phase composition of the as-synthesized materials.The XRD patterns showed NaBH 4 and NaCl crystalline phases before and after NiCp 2 (Figure 2a).The major phase identified was a cubic α-NaBH 4 whereas the formation of NaCl is primarily due to the metathesis reaction (precipitating NaCl as a by-product) between VCl 3 and NaBH 4 .The most intense diffraction peaks corresponding to the cubic α-NaBH 4 phase occurred around %25.11, 29.07, 41.57, 49.11, 51.52, 60.3, 66.24, 68.27, and 75.87°, and the peaks corresponding to NaCl occurred around %31.82, 45.62, 31.89,45.73, 75.61, and 56.69°(Figure 2a).In both V-NaBH 4 and V-NaBH 4 @Ni, the peaks corresponding to the NaBO 2 phase were not observed, which shows that the as-prepared materials were not oxidized.It is also worth mentioning that due to the dissolution of NaCl or Cl À substitution into NaBH 4 a discernible shift in the diffraction peaks of NaBH 4 should be observed. [28]owever, no appreciable shift in the diffraction patterns of NaBH 4 was observed in agreement with our previous findings on V-NaBH 4 and V-NaBH 4 @Ni prepared via solvent-based method. [27]12a,27]

B-H Bonding and Surface Chemical Composition
Fourier transform infrared (FTIR) analysis was further carried out to investigate the B-H bonding and identify any amorphous phases in V-NaBH 4 and V-NaBH 4 @Ni-30 (Figure 2b and S4, Supporting Information).For V-NaBH 4 , the FTIR spectra display the typical B-H stretching and bending vibrations between 2353-2223 and at 1120 cm À1 , respectively (Figure 2b).For V-NaBH 4 @Ni-30, the B-H stretching vibrations slightly shifted toward lower wavenumbers and appeared between 2348-2221 cm À1 while the B-H bending vibration was observed at 1120 cm À1 (Figure 2b).The shift in the wavenumber corresponding to the B-H stretching vibrations could be due to structural changes and/or weakening of the B-H bonding after Ni coating.It is important to note that after Ni coating the relative peak intensity of the B-H vibrations at 1118, 2221, 2288, and 2348 cm À1 was reduced (i.e., the screening effect from the shell shown in Figure S4, Supporting Information) compared to V-NaBH 4 , which may suggest the formation of a core-shell structure in agreement with the observations made on the Mg(BH 4 ) 2 @MgCl 2 core-shell. [29]The bending vibration at 1465 cm À1 was assigned to -NH 2 vibration, which suggests traces of ODA remained on the surface of particles. [30]he surface chemical composition and the electronic states of the elements were investigated by XPS (Figure 3).For V-NaBH 4 , in the B1s spectra, the typical B1s peak corresponding to the B-H species occurred at 187.8 eV (Figure S5a, Supporting Information).The peaks at 522.0 and 513.6 eV in the V2p spectra suggest the presence of V-B species (Figure S5b, Supporting Information).The presence of N and S was further evidenced by their typical N1s and S2p spectra (Figure S5c,d, Supporting Information).Particularly, in the S2p spectra, the peaks at 164.5 eV and 163.4 eV display the terminal SH groups from a thiol ligand (DDT). [31]The peaks at 162.9 eV and 161.6 eV indicate chemisorbed sulfur species and a metal-sulfur bond, respectively. [31,32]or V-NaBH 4 @Ni-30, in the B1s spectra, the peak at 188.1 eV strongly suggests the formation of V-B or Ni-B (after NiCp 2 coating) species consistent with the previous observations (Figure 3a). [33]In the V2p spectra, the peaks centered at 522.6 eV and 513.5 eV suggest the presence of V x B y species. [34]The N and S species were also evidenced from the typical N1s and S2p spectra (Figure S6, Supporting Information).Compared to V-NaBH 4 , the peaks at 400.5 eV and 164.5 eV corresponding to the -NH and -SH species in V-NaBH 4 @Ni-30, respectively, disappeared suggesting the formation of a Ni shell (Figure S5 and S6, Supporting Information).
In the case of V-NaBH 4 @Ni-30, the peaks at 871.4 eV and 853.3 eV in the Ni2p spectra suggest the formation of Ni metallic, while the peak at 856.5 eV may be due to Ni x B y (Figure 3c). [35]12a,27] Also, the surface content (at%) of B-H/B-O and V-B/V-O reduced after Ni-coating, suggesting a screening effect from Ni or Ni-B species (Table S1, Supporting Information).Overall, these results show a successful reduction of NiCp 2 to Ni metallic and the formation of V x B y /Ni x B y species on the surface of NaBH 4 .

Hydrogen Desorption and Reversibility
The hydrogen properties of the materials were investigated by TGA/DSC coupled with MS.The as-synthesized V-NaBH 4 showed a two-step decomposition between 25-400 and 400-500 °C with %13 and 3 mass% losses, and a concomitant release of hydrogen (visibly) at 100 °C with major peaks at 330 and 420 °C, respectively (Figure 4a,b).12a] In contrast, unmodified NaBH 4 showed no mass loss and H 2 release until 500 °C (Figure S7, Supporting Information).
For V-NaBH 4 @Ni-20, a mass loss of 6 and 9 mass% was observed between 25-350 °C and 350-410 °C, respectively (Figure S8a,b, Supporting Information).The endothermic peaks at 350 and 410 °C on the DSC curve may indicate melting of the borohydride phase and/or H 2 release from V-NaBH 4 @Ni in good agreement with the H 2 release peaks observed by MS (Figure S8a,b, Supporting Information).For V-NaBH 4 @Ni-30, a total mass loss of 14 mass% was observed between 25-415 °C and H 2 release occurring from 50 °C with major peaks at 355 °C and 410 °C (Figure 4c,d).The DSC curve shows two discernible endothermic peaks at 355 and 415 °C consistent with the H 2 release peaks observed by MS (Figure 4c,d).Notably, V-NaBH 4 @Ni-30 started releasing H 2 at 50 °C, which is comparable to V-NaBH 4 @Ni-20 (Figure 4d and S8b, Supporting Information).Similar trends were observed for V-NaBH 4 @Ni-40 (Figure S8c,d, Supporting Information).It should also be noted that by increasing the NiCp 2 loading an increased mass loss was observed and the hydrogen release temperature slightly shifted to higher temperatures.For example, a mass loss of 6 mass% in V-NaBH 4 @Ni-30 increased to 8 mass% for V-NaBH 4 @Ni-40, but hydrogen release was observed at 355 °C and 360 °C, respectively (Figure 4d and S8d, Supporting Information).Once again, for all the core-shell materials, a slightly greater mass loss than the theoretical H 2 content (10.6 mass%) of NaBH 4 and V-doped NaBH 4 (7.8 mass%) materials could be due to ODA/DDT decomposition around 350 °C (Figure 4 and S8, Supporting Information).Amongst these, only a negligible release of B 2 H 6 was observed in the MS at 300 °C after Ni coating, which we believe is due to the destabilization of the borohydride phase in the presence of Ni (Figure 4b,d, and S8b,d, Supporting Information).Importantly, NaBH 4 is the only hydrogen-releasing material in the composite as no hydrogen was observed from ODA and DDT (Figure S9 and S10, Supporting Information).
It is also clear from the aforementioned results that the as-synthesized materials released hydrogen without a noticeable B m H n or NH x contamination (Figure 4b,d).Although we removed the ligands by multiple washings with toluene, the N and S species remained on the surface of V-NaBH 4 @Ni-30 may also promote hydrogen release (Figure 2b, S5, S6 and Table S1, Supporting Information). [22]For example, in the case of N species, the proximity of -NH 2 to BH 4 (i.e., H δþ /H δÀ interactions) has been known to promote hydrogen release at lower temperatures (<350 °C). [36]In the case of S-containing species, they may also trigger hydrogen release from NaBH 4 and surprisingly assist in the formation of borides as observed by Zhang et al. [13] However, the exact mechanism is unclear.Considering the content of Ni and V x B y /Ni x B y species on/at the surface/interfaces compared to N or S species (Table S1, Supporting Information), boride species are expected to play a key in improving hydrogen release and uptake.
To further investigate the reversibility of the core-shell materials, firstly, we fully desorbed V-NaBH 4 @Ni-30 by heat treating it at 350 °C for 4 h under 0.01 MPa and then subjected the same material to H 2 pressure (10 MPa) at 350 °C. Figure 5a shows that the material was dehydrogenated after heat treatment at 350 °C in 4 h.After absorption, H 2 release started at 50 °C and peaked at 390 °C with a total mass loss of 2% (Figure 5b).This highlights an improved H 2 uptake in V-NaBH 4 @Ni-30 compared to unmodified NaBH 4 , which does not absorb H 2 under these conditions.The improvement in hydrogen uptake could be due to the core-shell structure and/or presence of Ni x B y /V x B y species in good agreement with our investigations on core-shell V-NaBH 4 @Ni. [20,27]

Mechanism
To gain insights into absorption/desorption mechanisms, XRD and FTIR analyses were conducted on the cycled materials (Figure 6).According to XRD, after desorption at 350 °C, all the peaks corresponding to the NaBH 4 phase disappeared indicating a complete decomposition of NaBH 4 (Figure 2a and 6a).After absorption, the diffraction peaks appeared at 27.04, 29.89, 31.20,43.57, 50.38, 53.30, 64.59, and 70.20°, which suggests that the decomposed material partially regenerated to NaBH 4 (Figure 6a).It is noteworthy that after hydrogen uptake the diffraction peaks were shifted to higher Braggs angles compared to the reference diffraction of NaBH 4 (Figure 2a and 6a), which may indicate some Cl À substitution in the NaBH 4 lattice. [28,37]herefore, the regenerated borohydride is expected to be of composition Na(BH 4 ) 1Àx Cl x after cycling, where the substitution of Cl À into BH À is inevitable at 350 °C.For the sake of simplicity, we still denote the substituted borohydride as NaBH 4 .In the FTIR spectra, after desorption, the B-H stretching and bending vibrations completely disappeared, which suggested the decomposition of NaBH 4 consistent with XRD analysis (Figure 6b).After absorption, the B-H stretching and bending vibrations appeared between 2439-2256 and at 1130 cm À1 , respectively (Figure 6b).The shifts in the B-H stretching and bending vibrations could be due to an altered structural environment (e.g., substituted borohydride) in the cycled materials compared to the parent borohydride.Overall, the appearance of the B-H vibrations after absorption strongly suggests that the decomposed products were regenerated to the borohydride phase (i.e., NaBH 4 or Na(BH 4 ) 1Àx Cl x ) in agreement with our XRD analysis and the previous studies. [16,20,38]The regeneration of NaBH 4 could only be possible without the formation of B 12 H 12 , as the B 12 H 12 phases act as thermodynamic sinks and preclude the rehydrogenation of the decomposed material to BH 4 À . [16]t should be noted that the decomposition of the destabilized NaBH 4 in a core-shell structure can lead to the formation of B 2 H 6 , and possibly, the catalytic dissociation of B 2 H 6 into B and H 2 is favored by Ni and/or Ni x B y /V x B y species at the core-shell's surface/interfaces (Figure 7). [39]During hydrogen absorption, the Ni and/or Ni x B y /V x B y phases may also facilitate the dissociation of H 2 and its recombination with Na and B present inside the core--shell structure.Thus, we believe that the regeneration in the present study could be because of the Ni shell and/or presence of V x B y /Ni x B y phases, which can accelerate H 2 release/uptake of NaBH 4 as illustrated in Figure 7.
We carried out a detailed study on different solid-state reaction derived core-shell-like structures to understand the impact of anionic substitution and the boride species on reversibility.28a,37] In addition, we also investigated the effect of fluoride (F À ) ion substitution to weaken the B-H bonds and facilitate H 2 uptake, prevent foaming and phase segregation, and the loss of active species (Na and B). [40]Indeed, substituting F À into NaBH 4 and the core-shell materials showed improved H 2 release compared to bulk NaBH 4 , however, the reversibility was again negligible mainly because of B 2 H 6 emissions, which then reacted with BH 4 À species and led to the formation of stable B 12 H 12 phases (Figure S15-S18, Supporting Information).It is noteworthy here that besides B 2 H 6 emissions, F À substitution also led to the formation of NaBF 4 , which severely hampered the reversibility (Figure S15-S18, Supporting Information).Previous approaches to improving reversibility by destabilizing borohydride and mitigating the formation of B 12 H 12 phases include: 1) the use of metal hydrides (e.g., MgH 2 ), [41] or reactive hydrides (e.g., Mg 2 NiH 4 ), [42] 2) transition metal compounds (e.g., TiF 3 , TiCl 3 , etc.), [43] 3) metal borides (e.g., Co x B y ), [44] and/or 4) nanoconfinement (e.g., in mesoporous carbons). [16]Among these, transition metal doping leading to the formation of boride species is an attractive approach for destabilizing borohydrides and improving their hydrogen reversibility.We also demonstrated that the presence of boride species (V x B y and Ni x B y ) is essential to destabilize borohydride and improve its hydrogen release and uptake. [27]The present study once again provides convincing evidence for improved hydrogen release and uptake of V-doped core-shell borohydride structures.These results also emphasize that compared to the destabilization achieved by Cl À , Br À , or F À substitutions, V doping leads to the formation of V x B y species, accelerates the H 2 release/uptake in the core-shell structure and bypasses the formation of B 12 H 12 .12a,27] Considering the potential of our method, a better understanding of the underpinning processes in improving reversible hydrogen storage is likely to be the focus of future research.

Conclusions
In summary, a simple solid-state method has been proposed to prepare core-shell NaBH 4 @Ni structures, which showed improved hydrogen release properties compared to the uncoated or bulk counterparts.
Synthesis of Cube-Like V-NaBH 4 : To a solution of NaBH 4 (100 mM in IPA), a set concentration of VCl 3 (10 mM final) was added while stirring the mixture at 500 rpm and 25 °C.After stirring for 16 h, the desired concentration (5 mM) of ODA was added to the homogenized mixture and additionally stirred for 1 h.The mixture was dried at 2 mbar and 30 °C.The as-synthesized V-doped cube-like NaBH 4 material is denoted as V-NaBH 4 .
Synthesis of Core-Shell V-NaBH 4 @Ni: In a typical synthesis, ODA/DDT stabilized V-doped NaBH 4 obtained via the solvent evaporation method was mixed with a varying concentration of NiCp 2 (i.e., 20, 30, and 40 mass%) in an agate mortar, and then heat-treated in a sealed stainless-steel vial at 150 °C for 2 h.The materials were washed with toluene several times before drying under the vacuum.The core-shell materials obtained with 20, 30, and 40 mass% NiCp 2 are denoted as V-NaBH 4 @Ni-X (where X stands for the concentration of NiCp 2 ).
The rationale behind using ODA (amine) and DDT (thiol) is: 1) to obtain size and shape-controlled NaBH 4 , 2) that NiCp 2 would complex with the -SH groups of DDT on the surface of the stabilized NaBH 4 cores according to the reaction in Equation ( 1), [45] and 3) that the melting of the ODA/DDT ligands at around %50 °C would supply a sufficient wetting on the NaBH 4 surface where the Ni 2þ ions complex with the ligands, be reduced by the BH 4 À species and form a uniform Ni coating over the NaBH 4 cores.
Preparation of Ni-Hexamine: In a typical synthesis, %500 mg of NiCl 2 •6H 2 O was dissolved in water at RT. Ammonia solution was then added dropwise to displace H 2 O molecules while stirring.The solution was refrigerated (%4 °C) for at least 4 h and then mixed with absolute ethanol (40 mL) to precipitate the product.The precipitate was separated and washed with ethanol several times before drying under a vacuum.
Synthesis of Core-Shell Using ODA/TBAB Stabilized NaBH 4 and Ni-Hexamine: In a typical synthesis, ODA (2.5 mM)/TBAB (20 mM) stabilized NaBH 4 were mixed with 20-30 mass% of Ni-hexamine and heattreated in a sealed stainless-steel vial at 200 °C for 2 h.The materials were washed with toluene several times before drying under a high vacuum.
Synthesis of ODA/TBAB or ODA/TBAF/DDT Stabilized NaBH 4 : In a typical synthesis, 100 mM of NaBH 4 was dissolved in IPA followed by dropwise addition of ODA and TBAF to achieve a final concentration of 2.5 and 20 mM, respectively.After stirring for 1 h, the homogenized mixture was dried at 2 mbar, 30 °C.The same method was repeated to get ODA (2.5 mM)/TBAF (20 mM)/DDT(1 mM) stabilized NaBH 4 .
Synthesis of Core-Shell Using ODA/TBAF/DDT Stabilized NaBH 4 and NiCp 2 : In a typical synthesis, ODA/TBAF/DDT (ODA: 2.5 mM, TBAF: 20 mM, and DDT: 1 mM) stabilized NaBH 4 was mixed with a varying concentration of NiCp 2 (i.e., 10, 20, and 30 mass%) in an agate mortar and then heat-treated in a sealed stainless-steel vessel at 150 °C for 2 h.The materials were washed with toluene several times before drying under a high vacuum.
Transmission Electron Microscopy (TEM): Microscopic analysis was done by TEM on Philips CM 200 (Eindhoven, the Netherlands) operated at 200 kV.HAADF-STEM images were taken by field emission gun (FEG)scanning TEM (FEG-STEM, JEOL, JEM-F200 Multipurpose FEG-STEM, Tokyo, Japan) at 200 kV at a camera length of around 120 mm and an angle of AE15°.The materials were dispersed in toluene followed by short ultrasonication and then dropped onto a carbon-coated copper grid.The grids were enclosed in an argon-filled bottle to minimize air exposure and then transferred to the microscope.
X-ray Diffraction (XRD): Crystalline phases were determined by XRD on an X'pert Multipurpose XRD system operated at 40 mA and 45 kV with a monochromated Cu K α radiation (λ = 1.541Å) from 10 to 80°.The materials were protected against oxidation in the air by a Kapton foil.
Fourier Transform Infrared (FTIR) Spectroscopy: FTIR analysis was conducted on a Bruker Vertex 70 V.The materials were mixed with dry KBr and loaded in an air-tight chamber fitted on a Harrick-scientific praying mantis diffuse reflectance infrared fourier transform spectroscopy accessory.The spectra were collected at RT from 600 to 3000 cm À1 over 124 scans with a resolution of 1 cm À1 .
X-ray Photoelectron Spectroscopy (XPS): The chemical properties of the surface of the nanomaterials were characterized by XPS using a Thermo Scientific ESCALAB250Xi, UK spectrometer (base pressure below 2 Â 10 À6 Pa).Powder materials were pressed on high-quality indium substrates, placed in a container filled with argon and transferred to the spectrometer to minimize air exposure.The XPS spectra were collected using a monochromatic Al K α (1486.7 keV) X-ray source at 150 W power. Survey scans were collected at 100 eV pass energy with an energy step of 0.5 eV, while high-resolution spectra were acquired at the 20 eV pass energy and 0.1 eV energy step.The data were analyzed and processed using the Avantage and CasaXPS packages.
Hydrogen Desorption: The hydrogen release properties of the materials were evaluated by thermal gravimetric analysis/differential scanning calorimetry (TGA/DSC) using a Mettler Toledo DSC 3þ coupled to a Pfeiffer Omnistar mass spectrometer (MS).Alumina pans were used, and the TGA/DSC measurements were performed under a flow of argon of 20 mL min À1 from 20 °C to 500 °C with a heating rate of 10 °C min À1 .The MS was used to acquire the hydrogen desorption profiles of the materials.
Hydrogen Cycling: Hydrogen desorption kinetics were characterized by using a homemade Sievert apparatus.For desorption, about 50 mg of the material was loaded into a stainless steel vessel and subjected to a pressure of 0.01 MPa and heated at 350 °C for 4 h.For absorption, the dehydrogenated material was subjected to a hydrogen pressure of 10 MPa at 350 °C for 12 h.The materials were then evaluated by TGA/DSC-MS measurements to check any gas emissions after each desorption or absorption cycle.The hydrogen storage capacity was estimated by TGA.

Figure 3 .
Figure 3. High-resolution XPS spectra.a) B1s, b) V2p, and c) Ni2p of V-NaBH 4 @Ni-30.In the B1s and V2p spectra, the presence of B-O and V-O suggests some surface oxidation while transferring the materials in air to the XPS instrument.The N1s and S2p spectra are given in Figure S5, Supporting Information.For comparison, the XPS spectrum of V-NaBH 4 is given in Figure S6, Supporting Information.

Figure 4 .
Figure 4. TGA/DSC-MS of a,b) V-NaBH 4 and c,d) V-NaBH 4 @Ni-30.The onset of H 2 release is indicated by a red arrow in (b and d).In (b) and (d), negligible mass fragments corresponding to B 2 H 6 were observed, indicating pure H 2 release.For comparison, V-NaBH 4 @Ni-20 and V-NaBH 4 @Ni-40 are given in Figure S8, Supporting Information.

Figure 7 .
Figure 7. Schematic illustration of the proposed reaction pathways during desorption-absorption for core-shell borohydrides.During desorption (left panel), B 2 H 6 may form inside the core-shell structure due to the destabilized borohydride core and then split into B and H in the presence of Ni shell and/or V x B y /Ni x B y species.During absorption, hydrogen permeates through the Ni shell and combines with the Na and B species retained inside the core-shell structure, leading to the formation of NaBH 4 .
The solid-state reaction derived core-shell materials (in the presence of Ni, V x B y /Ni x B y species) bypassed the formation of B 12 H 12 by suppressing B 2 H 6 emission, and this improved the hydrogen reversibility of NaBH 4 .Notably, a net reversible capacity of 2 mass% H 2 was achieved at 350 °C and 10 MPa H 2 pressure.A high reversible hydrogen capacity can be achieved by removing the byproducts (e.g., chlorides).Detailed investigations on the Br À , Cl À , and/or F À substituted materials (without vanadium) indicated that these species either strongly interacted or reacted with NaBH 4 to produce stable byproducts such as NaCl, NaF, and/or NaBF 4 , which impaired the H 2 uptake properties of NaBH 4 .The poor hydrogen absorption in these materials was also found to be due to B 2 H 6 emissions and the formation of B 12 H 12 phases during cycling.Overall, this work discloses a new method and provides a guide to designing practical hydrogen storage materials.Future work is expected to be toward improving the hydrogen release and uptake properties of solid-state reaction derived hydride nanostructures.