Preparation of Rh / Ni Bimetallic Nanoparticles and Their Catalytic Activities for Hydrogen Generation from Hydrolysis of KBH 4

ISOBAM-104 protected Rh/Ni bimetallic nanoparticles (BNPs) of 3.1 nm in diameter were synthesized by a co-reduction method with a rapid injection of KBH4 solution. The catalytic activities of as-prepared BNPs for hydrogen generation from hydrolysis of a basic KBH4 solution were evaluated. Ultraviolet-visible spectrophotometry (UV-Vis), transmission electron microscopy (TEM), and high-resolution transmission electron microscopy (HRTEM) were employed to characterize the structure, particle size, and chemical composition of the resultant BNPs. Catalytic activities for hydrolysis of KBH4 and catalytic kinetics of prepared BNPs were also investigated. It was shown that Rh/Ni BNPs displayed much higher catalytic activities than that of Rh or Ni monometallic nanoparticles (MNPs), and the prepared Rh10Ni90 BNPs possessed the highest catalytic activities, with a value of 11,580 mol-H2·h·mol-Rh. The high catalytic activities of Rh/Ni BNPs could be attributed to the electron transfer effect between Rh and Ni atoms, which was confirmed by a density functional theory (DFT) calculation. The apparent activation energy for hydrogen generation of the prepared Rh10Ni90 BNPs was about 47.2 ± 2.1 kJ/mol, according to a kinetic study.


Introduction
Hydrogen is one of many potential alternatives to replace nonrenewable fuel sources that are used nowadays, as it is an environmentally-friendly and renewable energy carrier.However, the technique for hydrogen storage is still a large problem which hinders the application of hydrogen.To date, an extensive body of research has been published on hydrogen storage including liquid hydrogen storage, high pressure gaseous hydrogen storage, adsorption hydrogen storage, metal hydride hydrogen storage, organic compounds hydrogen storage, and liquid phase chemical hydrogen storage [1][2][3][4].Among these methods, liquid phase chemical hydrogen storage attracted considerable attention due to their high hydrogen content, high hydrogen purity, and easy control of the hydrogen generation rate [5,6].Compared with other chemical hydrogen storage materials (such as hydrazine hydrate, ammonia borane and formic acid), potassium/sodium borohydride (KBH 4 /NaBH 4 ) is more competitive because of several advantages, including safe production process, convenience of transportation, and environmentally benign hydrolysis product NaBO 2 .However, the rate of hydrogen generation from hydrolysis of KBH 4 /NaBH 4 aqueous solution is usually low at elevated pH, especially under alkaline conditions, such as pH = 12 [7][8][9].Therefore, catalysts are important for hydrolysis of KBH 4 , as shown in formula (1): It is well-known that nanoparticles (NPs), especially noble metal NPs, can act as high-efficiency catalysts for hydrogen generation from hydrolysis of KBH 4 /NaBH 4 .Moreover, combinations of non-noble metal can reduce the use of noble metals and lower the cost of catalysts.Further, it can also improve the catalytic activities of the catalysts due to the synergistic effect between different metal atoms [9][10][11][12][13][14][15][16][17][18][19].Our previous work has indicated that alloy-structured Au 50 Ni 50 bimetallic nanoparticles (BNPs) exhibited catalytic activities several times higher for hydrogen generation from hydrolysis of NaBH 4 aqueous solution, compared with that of Au and Ni monometallic nanoparticles (MNPs) [12].Au/Co BNPs also displayed a much higher catalytic activity for hydrogen generation than that of Au and Co MNPs [13].
Rh NPs have attracted considerable attention in the catalysis area because they are active for many chemical reactions [20].For example, Rh NPs supported on silica-coated magnetite showed significant hydrogenation activity of benzene and cyclohexene, and the catalytic activity remained unchangeable for up to 20 cycles [21].Poly(N-vinyl-2-pyrrolidone)-protected Ru/Rh BNPs can act as highly efficient catalysts in the hydrolysis of ammonia borane for hydrogen generation [22].Ni@(RhNi-alloy) nanocomposites supported on NiAl-layered double hydroxides (NiAl-LDHs) were reported to be highly efficient catalysts towards hydrogen generation in the hydrolysis of N 2 H 4 BH 3 [23].The addition of Rh metals has greatly improved the catalytic activity of Co-based catalysts in the ethanol stream reforming reaction, indicating the second metal could fundamentally influence the properties of the catalyst [24,25].Nevertheless, the high cost of Rh hinders its wide industrial application.Thus, a combination of non-noble metal with Rh is a promising strategy for the development of Rh-based catalyst for hydrogen generation [23,26].
In the present paper, a series of ISOBAM-104 (poly (isobutylene-alt-maleic anhydride) (C 8 H 10 O 3 )m(C 8 H 16 O 3 N 2 ) i , designed as ISOBAM-104) protected Rh/Ni BNPs were prepared by a facile method, and the relationship between compositions and structures of the BNPs on their catalytic activities for hydrogen generation were also investigated.ISOBAM-104 is expected to protect the metal NPs from agglomeration because it has numerous of functional groups and can act as chelant.The apparent activation energy of Rh 10 Ni 90 BNPs for hydrolysis of KBH 4 aqueous solution was calculated by the Arrhenius method.Moreover, the correlation between catalytic activities of Rh/Ni BNPs and their electronic properties was established based on a density functional theory (DFT) calculation.

Structure and Catalytic Activities of Rh/Ni Bimetallic Nanoparticles (BNPs)
UV-Vis spectra of as-prepared colloidal dispersion are shown in Figure 1.There is no surface plasmon resonance (SPR) peak of Rh, Ni MNPs or Rh/Ni BNPs in measuring range, which is consistent with previous reports [9,18].The spectra of aqueous dispersed Rh/Ni BNPs displays featureless absorbance that monotonically increase toward a higher Rh content.The absorbance spectra of all BNPs lie between the spectrum of Rh and Ni MNPs, and the obvious differences in absorbance between as-prepared BNPs with varied Rh content suggest that alloy-structured Rh/Ni BNPs were formed.
Figure 2  analysis are about 1.9 ± 0.9 nm, 3.5 ± 1.9 nm, 3.7 ± 1.9 nm, 2.7 ± 0.8 nm, 2.7 ± 0.9 nm and 3.5 ± 1.2 nm, respectively.The elemental ratio of Rh 70 Ni 30 BNPs at the selected square in Figure 3 was measured by mapping energy dispersion X-ray spectroscopy (EDS), and it indicates that the compositions of as-prepared Rh 70 Ni 30 BNPs are similar to their feeding ratio.
Catalysts 2017, 7, x FOR PEER REVIEW 3 of 11 Figure 2 presents a set of transmission electron microscopy (TEM) micrographs of the prepared Rh, Ni MNPs and Rh/Ni BNPs.The individual NPs appear to be separated uniformly without obvious agglomeration.The average particle sizes of Rh, Rh90Ni10, Rh70Ni30, Rh30Ni70, Rh10Ni90 and Ni NPs based on size distribution analysis are about 1.9 ± 0.9 nm, 3.5 ± 1.9 nm, 3.7 ± 1.9 nm, 2.7 ± 0.8 nm, 2.7 ± 0.9 nm and 3.5 ± 1.2 nm, respectively.The elemental ratio of Rh70Ni30 BNPs at the selected square in Figure 3 was measured by mapping energy dispersion X-ray spectroscopy (EDS), and it indicates that the compositions of as-prepared Rh70Ni30 BNPs are similar to their feeding ratio.In order to further verify the formation of an alloyed structure in the as-prepared BNPs, a lattice fringes analysis was also carried out, based on HRTEM images of the Rh 70 Ni 30 colloidal dispersions.The particles exhibit an obvious crystalline structure, as revealed in Figure 4.The interplanar distances of three individual randomly-chosen particles of Rh 70 Ni 30 BNPs were respectively measured to be 0.212 nm (particle-1), 0.218 nm (particle-2), and 0.216 nm (particle-3), as labeled in Figure 4. Comparing the results of Figure 4 with the theoretically interplanar spacing of Rh and Ni (based on XRD standard card, as shown in Table 1), the formation of individual Rh and Ni MNPs in the as-prepared samples can be ruled out.This is due to the mismatch of interplanar distances between these BNPs and Rh, or Ni MNPs.However, it should be noted that the measured interplanar distances lie between the interplanar spacing of Rh (111) (0.2196 nm), and that of Ni (111) (0.2034 nm), as shown in Table 2.This suggests that alloyed structures are formed in the particles and the interplanar spacing can be assigned to (111) of the alloy-structured Rh/Ni BNPs.To the best of our knowledge, this is the first report on the preparation of ISOBAM-104 protected alloy-structured Rh/Ni BNPs by using such a facile co-reduction method.
to the mismatch of interplanar distances between these BNPs and Rh, or Ni MNPs.However, it should be noted that the measured interplanar distances lie between the interplanar spacing of Rh (111) (0.2196 nm), and that of Ni (111) (0.2034 nm), as shown in Table 2.This suggests that alloyed structures are formed in the particles and the interplanar spacing can be assigned to (111) of the alloystructured Rh/Ni BNPs.To the best of our knowledge, this is the first report on the preparation of ISOBAM-104 protected alloy-structured Rh/Ni BNPs by using such a facile co-reduction method.Catalytic activities of ISOBAM-104 protected RhxNi(100-x) (x = 0, 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100) NPs for H2 generation from hydrolysis of alkaline KBH4 aqueous solution at 303 K, are illustrated in Figure 5.The activities of BNPs were normalized to mol-H2•h −1 •mol-Rh −1 since the catalytic activity of Ni MNPs is very low, showing that most of the Rh/Ni BNPs exhibit higher catalytic activities than that of Rh or Ni MNPs.Moreover, Rh10Ni90 BNPs possess the highest catalytic activities with a value of 11,580 mol-H2•h −1 •mol-Rh −1 for hydrogen generation, which is respectively about 3 and 37 times higher than that of Rh (3560 mol-H2•h −1 •mol-Rh −1 ), and Ni MNPs (310 mol-H2•h −1 •mol-Ni −1 ).

Kinetic Study on Rh10Ni90 BNPs
The effects of pH and reaction temperature on the catalytic activities of the as-prepared BNPs were also investigated using Rh10Ni90 as model catalysts.It shows that the final hydrogen productivity of the BNPs decreases from 80% to 45%, with pH increasing from 12 to 14, as shown in Figure 6.

Kinetic Study on Rh 10 Ni 90 BNPs
The effects of pH and reaction temperature on the catalytic activities of the as-prepared BNPs were also investigated using Rh 10 Ni 90 as model catalysts.It shows that the final hydrogen productivity of the BNPs decreases from 80% to 45%, with pH increasing from 12 to 14, as shown in Figure 6.

Kinetic Study on Rh10Ni90 BNPs
The effects of pH and reaction temperature on the catalytic activities of the as-prepared BNPs were also investigated using Rh10Ni90 as model catalysts.It shows that the final hydrogen productivity of the BNPs decreases from 80% to 45%, with pH increasing from 12 to 14, as shown in Figure 6.The apparent activation energy (Ea) of the BNPs for hydrogen generation from the hydrolysis of alkaline KBH4 solution was calculated by using the Arrhenius method [9,17,27].The catalytic activities of Rh10Ni90 were enhanced with the increasing reaction temperature and a linear dependence between catalytic rates (in a logarithmic scale, ln k) and the reciprocals of temperature was observed (as shown in Figure 7).According to the Arrhenius equation, the slope of the linear plot is −Ea/R, where R represents the universal gas constant.Within the temperatures ranging from 303 to 323 K, Ea was calculated to be 47.2 ± 2.1 kJ/mol for Rh10Ni90 BNPs.These results suggest that the as-prepared Rh10Ni90 BNPs are excellent catalysts for the hydrolysis of KBH4 because of their lower apparent activation energy compared with other reported catalysts, such as 51.2 kJ/mol for Co-La-Zr-B NPs [28], 52.0 kJ/mol for Co-αAl2O3-Cu catalysts [29], 55.6 kJ/mol for Co/alginate hydrogels [30], and 48.8 kJ/mol for Mo incorporated Co-Ru-B catalysts [31], etc.The apparent activation energy (E a ) of the BNPs for hydrogen generation from the hydrolysis of alkaline KBH 4 solution was calculated by using the Arrhenius method [9,17,27].The catalytic activities of Rh 10 Ni 90 were enhanced with the increasing reaction temperature and a linear dependence between catalytic rates (in a logarithmic scale, ln k) and the reciprocals of temperature was observed (as shown in Figure 7).According to the Arrhenius equation, the slope of the linear plot is −E a /R, where R represents the universal gas constant.Within the temperatures ranging from 303 to 323 K, E a was calculated to be 47.2 ± 2.1 kJ/mol for Rh 10 Ni 90 BNPs.These results suggest that the as-prepared Rh 10 Ni 90 BNPs are excellent catalysts for the hydrolysis of KBH 4 because of their lower apparent activation energy compared with other reported catalysts, such as 51.2 kJ/mol for Co-La-Zr-B NPs [28], 52.0 kJ/mol for Co-αAl 2 O 3 -Cu catalysts [29], 55.6 kJ/mol for Co/alginate hydrogels [30], and 48.8 kJ/mol for Mo incorporated Co-Ru-B catalysts [31], etc.

Correlation between Catalytic Activities of the Rh/Ni BNPs and Their Electronic Properties
Figure 5 showed that most of the as-prepared BNPs have higher activity than that of Rh and Ni MNPs, and that Rh10Ni90 BNPs possessed the highest catalytic activity for the hydrolysis of KBH4 among all prepared NPs.According to previous investigations [19,[32][33][34], it is reasonable to suggest that the element components and electronic properties of Rh and Ni atoms affect the catalytic activities of the prepared BNPs.To confirm the existence of electron donation between Rh atoms and Ni atoms, DFT calculations were carried out to study the electron transfer of the BNPs, and Rh6Ni49 BNP were calculated as a model-the calculation results show that there is indeed an electron charge transfer effect between Rh and Ni atoms.The electron transfers from Rh atoms to Ni atoms owing to the relatively higher electron negativity value of Rh (2.28) than that of Ni (1.91), leading to the

Correlation between Catalytic Activities of the Rh/Ni BNPs and Their Electronic Properties
Figure 5 showed that most of the as-prepared BNPs have higher activity than that of Rh and Ni MNPs, and that Rh 10 Ni 90 BNPs possessed the highest catalytic activity for the hydrolysis of KBH 4 among all prepared NPs.According to previous investigations [19,[32][33][34], it is reasonable to suggest that the element components and electronic properties of Rh and Ni atoms affect the catalytic activities of the prepared BNPs.To confirm the existence of electron donation between Rh atoms and Ni atoms, DFT calculations were carried out to study the electron transfer of the BNPs, and Rh 6 Ni 49 BNP were calculated as a model-the calculation results show that there is indeed an electron charge transfer effect between Rh and Ni atoms.The electron transfers from Rh atoms to Ni atoms owing to the relatively higher electron negativity value of Rh (2.28) than that of Ni (1.91), leading to the presence of negatively-charged Rh atoms and positively-charged Ni atoms in the Rh 6 Ni 49 BNPs (Figure 8).It is believed that the charged Rh and Ni atoms can act as catalytically active sites, and can then enhance the catalytic activities for the hydrogen generation from hydrolysis of KBH 4 aqueous solution [35][36][37][38].

Correlation between Catalytic Activities of the Rh/Ni BNPs and Their Electronic Properties
Figure 5 showed that most of the as-prepared BNPs have higher activity than that of Rh and Ni MNPs, and that Rh10Ni90 BNPs possessed the highest catalytic activity for the hydrolysis of KBH4 among all prepared NPs.According to previous investigations [19,[32][33][34], it is reasonable to suggest that the element components and electronic properties of Rh and Ni atoms affect the catalytic activities of the prepared BNPs.To confirm the existence of electron donation between Rh atoms and Ni atoms, DFT calculations were carried out to study the electron transfer of the BNPs, and Rh6Ni49 BNP were calculated as a model-the calculation results show that there is indeed an electron charge transfer effect between Rh and Ni atoms.The electron transfers from Rh atoms to Ni atoms owing to the relatively higher electron negativity value of Rh (2.28) than that of Ni (1.91), leading to the presence of negatively-charged Rh atoms and positively-charged Ni atoms in the Rh6Ni49 BNPs (Figure 8).It is believed that the charged Rh and Ni atoms can act as catalytically active sites, and can then enhance the catalytic activities for the hydrogen generation from hydrolysis of KBH4 aqueous solution [35][36][37][38].
Catalysts 2017, 7, x FOR PEER REVIEW 8 of 11 chemical structure is shown in Figure 9) was purchased from KURARAY company, Japan.Water was purified by a water distiller system.

Experiments
A series of Rh(100-x)Nix (x = 0, 10, 20, 30, 60, 80, 90, 95 and 100) BNPs were synthesized by changing the addition content of RhCl3 and NiCl2 with the total metal concentration kept at 0.66 mM.Rh/Ni BNPs were prepared through co-reduction method under 273 K in N2 atmosphere.For example, Rh50Ni50 BNPs were prepared as follows: 25 mL RhCl3 solution (0.66 mM) and 25 mL NiCl2 solution (0.66 mM) were firstly mixed homogeneously in a two-neck flask under vigorous stirring, and then 50 mL ISOBAM-104 (66 mM) was added into the flask and stirred for another 30 min.Then 10 mL KBH4 (16.5 mM) was injected into the aqueous solution within 5 s in an ice-water bath [39][40][41].The color of the mixed solution slowly changed from to black, which represents the formation of the Rh/Ni BNPs.Finally, colloidal dispersions Rh50Ni50BNPs were obtained after another 1 h of mixing.

Experiments
A series of Rh (100-x) Ni x (x = 0, 10, 20, 30, 60, 80, 90, 95 and 100) BNPs were synthesized by changing the addition content of RhCl 3 and NiCl 2 with the total metal concentration kept at 0.66 mM.Rh/Ni BNPs were prepared through co-reduction method under 273 K in N 2 atmosphere.For example, Rh 50 Ni 50 BNPs were prepared as follows: 25 mL RhCl 3 solution (0.66 mM) and 25 mL NiCl 2 solution (0.66 mM) were firstly mixed homogeneously in a two-neck flask under vigorous stirring, and then 50 mL ISOBAM-104 (66 mM) was added into the flask and stirred for another 30 min.Then 10 mL KBH 4 (16.5 mM) was injected into the aqueous solution within 5 s in an ice-water bath [39][40][41].The color of the mixed solution slowly changed from transparent to black, which represents the formation of the Rh/Ni BNPs.Finally, colloidal dispersions Rh 50 Ni 50 BNPs were obtained after another 1 h of mixing.

Characterization of Nanoparticles
UV-Vis absorption spectra were measured at 200-800 nm by a Shimadzu UV-2550 (Shimadzu company, Kobe, Japan) recording spectrophotometer.TEM images were taken with a FEI Tecnai G2 50-S-TWIN TEM (FEI company, Hillsboro, OR, USA) at the accelerated voltage of 80 kV.The specimens were prepared by placing two or three drops of the prepared colloidal aqueous solution onto a copper microgrid, which was covered with a thin amorphous carbon film, and drying it in air at an ambient temperature.Generally, to evaluate the mean diameter, at least 200 particles from different locations on the grid were selected for each sample.HRTEM images were observed at the accelerated voltage of 200 kV using a JEM-2100F (JEOL company, Tokyo, Japan) Field Emission High-resolution TEM.The EDS measurement was performed with a NORAN UTW type Si (Li) semiconducting detector attached to the HRTEM equipment.

Catalytic Properties
The catalytic performance of Rh/Ni BNPs was evaluated by the hydrogen generation from hydrolysis of alkaline KBH 4 aqueous solution.The reaction was started when the alkaline KBH 4 aqueous solution was added into the colloidal catalyst under continuous stirring.Hydrogen was bubbled through the suspension, and its volume was obtained with a water drainage method.At the same time, plots of hydrogen volume vs. reaction time with an interval of 2 s were collected by a computer.The turnover frequency (TOF) was calculated through the slope of a fitted straight line using H 2 volume vs. reaction time curve.The initial specific activities (mol-H 2 •h −1 •mol-Rh −1 ) related to the noble metal content of the catalysts were calculated for comparison.Every experiment was repeated at least twice, and the mean value of the measuring results was used for calculating the value of TOF.The catalytic kinetics were investigated at varied pH (12, 13 and 14, 303 K) and different temperatures (303 K, 308 K, 313 K, 318 K and 323 K, pH = 12,), using Rh 10 Ni 90 as model catalysts.

DFT Calculation
DFT calculations were carried out using spin-polarization DFT/GGA with the PBE exchange-correlation functional [42], as implemented in the DMol 3 package [43] (BIOVIA company, San Diego, CA, USA).Double numerical basis set and polarization functions were carried out to describe the valence electrons, and an electron relativistic core treatment was used to perform full optimization of the investigated cluster model of Rh 6 Ni 49 BNP without symmetry constraint.The convergence criteria were set to medium quality with a tolerance for self-consistent field (SCF), optimization energy, maximum force, and maximum displacement of 10 −5 Ha, 2 × 10 −5 Ha, 0.004 Ha/Å and 0.005 Å, respectively.Charge analysis was performed on the basis of the Mulliken population distribution scheme [44,45].

Conclusions
ISOBAM-104 protected alloy-structured Rh/Ni BNPs were prepared by a co-reduction method and characterized by UV-Vis, TEM, EDS and HRTEM.The catalytic activities and kinetic study for KBH 4 hydrolysis reaction were also investigated.The as-prepared Rh/Ni BNPs possessed high catalytic activities, and the activities of the Rh 10 Ni 90 BNPs with an average size of 3 nm were higher than that of Ni MNPs.They were also higher than that of the Rh MNPs, even though the latter has a much smaller size of 1.9 nm.The apparent activation energy was calculated to be 47.2 ± 2.1 kJ/mol for Rh 10 Ni 90 BNPs, which is lower than that of most reported catalysts, suggesting that Rh/Ni BNPs with low Rh loading were excellent catalysts for the hydrolysis of KBH 4 .The high catalytic activities of Rh/Ni BNPs could be attributed to the existence of the electron transfer effects between Rh and presents a set of TEM micrographs of the prepared Rh, Ni MNPs and Rh/Ni BNPs.The individual NPs appear to be separated uniformly without obvious agglomeration.The average particle sizes of Rh, Rh 90 Ni 10 , Rh 70 Ni 30 , Rh 30 Ni 70 , Rh 10 Ni 90 and Ni NPs based on size distribution

11 Figure 2 .
Figure 2. Transmission electron microscopy (TEM) images and size distribution histograms of

Figure 6 .
Figure 6.Evaluation of H 2 productivity with reaction time of Rh 10 Ni 90 BNPs at different pH values (R ISO = 40, C Metal = 0.66 mM, reduced under ice-water bath for 1 h; 30 • C).

Figure 8 .
Figure 8.The density functional theory (DFT) calculated Mulliken charge on selected Rh atom and Ni atoms (green, Rh; and purple, Ni).

Figure 8 .
Figure 8.The density functional theory (DFT) calculated Mulliken charge on selected Rh atom and Ni atoms (green, Rh; and purple, Ni).

Table 1 .
Lattice spacing (nm) and indexed reflection planes of Rh and Ni.

Table 2 .
Lattice spacing (nm) and indexed reflection planes of Rh70Ni30 BNPs determined by HRTEM in Figure4.

Table 1 .
Lattice spacing (nm) and indexed reflection planes of Rh and Ni.

Table 2 .
Lattice spacing (nm) and indexed reflection planes of Rh 70 Ni 30 BNPs determined by HRTEM in Figure 4.