Modiﬁed Nimo Nanoparticles for E ﬃ cient Catalytic Hydrogen Generation from Hydrous Hydrazine

: Precious metal-free NiMoM (M = Pr 2 O 3 , Cu 2 O) catalysts have been synthesized through a simple coreduction method, without any surfactant or support material, and characterized using X-ray di ﬀ raction (XRD), transmission electron microscopy (TEM), and X-ray photoelectron spectroscopy (XPS). The resultant Pr 2 O 3 - or Cu 2 O-modiﬁed NiMo catalysts exhibit di ﬀ erent structures, which is due to a di ﬀ erence in the synergistic e ﬀ ects of NiMo and the modifying elements. NiMoPr 2 O 3 has an amorphous structure, with low crystallinity and uniform particle dispersion, while NiMo@Cu 2 O adopts the core–shell structure, where the core and shell are synergistic with each other to promote electron transfer e ﬃ ciency. The support material-free nanocatalysts Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O are both highly e ﬃ cient compared with bimetallic NiMo catalysts, in terms of hydrogen generation from hydrous hydrazine (N 2 H 4 · H 2 O) at 343 K, with total turnover frequencies (TOFs) of 62 h − 1 and 71.4 h − 1 , respectively. Their corresponding activation energies (Ea) were determined to be 43.24 kJ mol − 1 and 46.47 kJ mol − 1 , respectively. This is the ﬁrst report on the use of Pr-modiﬁed NiMo and core–shell NiMo@Cu 2 O catalysts, and these results may be used to promote the e ﬀ ective application of noble metal-free nanocatalysts for hydrogen production from hydrous hydrazine.


Introduction
It is expected that energy consumption and associated environmental issues will continue to grow dramatically in the near future. As an environmentally friendly fuel, hydrogen has the advantages of high combustion heat and high combustion speed. In addition to generating water and a little hydrogen azide during the combustion process, it will not produce substances that are harmful to the environment such as carbon oxides and hydrocarbons compounds. Hydrogen has been proposed as crucial to ensuring secure and sustainable energy development [1][2][3]. However, the search for effective hydrogen storage materials and methods for hydrogen generation remains a challenge.
In recent years, chemical hydrogen storage has become a popular approach to overcoming the barriers of hydrogen delivery. Hydrous hydrazine (N 2 H 4 ·H 2 O), which is a liquid at room temperature, is a promising hydrogen carrier due to its relatively low cost, high hydrogen density (8.0 wt. %), low molecular weight (50.06 g mol −1 ) [4,5], and the advantages of only producing H 2 and N 2 in its decomposition reactions (pathway 1: N 2 H 4 → N 2 + 2H 2 ). However, to maximize the usability of N 2 H 4 ·H 2 O in hydrogen storage, we should avoid the undesirable generation of ammonia (pathway 2: 3N 2 H 4 → N 2 + 4NH 3 ). Although ammonia can be burned, not only generates less heat than that of H 2 as fuel, but also produces harmful gases like NO. Catalysis plays a critical role in the design of efficient processes and systems able to exploit the advantages of the starting materials, while minimizing waste generation and energy requirements [6]. Several recent studies found that Ni-based catalysts promoted by noble metals, such as Ni-Rh [7][8][9], Ni-Pt [10][11][12][13], Ni-Pd [14][15][16], and Ni-Ir exhibit superior catalytic performances in the decomposition of hydrous hydrazine, with more than 90% H 2 selectivity [17,18]. In fact, although nanocatalysts formed from noble metals and nickel exhibit good catalytic performance, high associated costs limit their use. Other studies have therefore employed nonprecious metals instead of precious metals. He et al. [19] modified Ni nanoparticles with a small amount of CeO 2 (8.0 mol %), resulting in significant enhancement of the turnover frequency (TOF) and H 2 selectivity. Men et al. [20] synthesized catalysts using CeO x -modified NiFe deposited on reduced graphene oxide (rGO), which exhibited good catalytic characteristics. Manukyan et al. [21] selected copper nanoparticles as the support material, and synthesized a NiFe/Cu catalyst. Copper may be an efficient support material for nickel in the water-gas shift reaction, as it shows high activity and stability. Wang et al. [22] successfully synthesized a Cu@Fe 5 Ni 5 catalyst, which exhibited high activity in terms of complete N 2 H 4 ·H 2 O decomposition. This is due to the electronic coupling between the core and shell metals. Noble metal-free modified Ni-based catalysts not only facilitate the use of N 2 H 4 as a chemical hydrogen-producing material, but also provide a potential development path for more rare earth-doped catalysts.
Yang et al. [23] reported that a NiMo catalyst catalyzed the hydrogen production of ammonium borane, and that the electronic interactions between the NiMo enable the catalyst to achieve complete decomposition of the ammonia borane within 2 min. We report a simple coreduction method for preparing NiMo catalysts, which are modified with rare earth elements and nonprecious metals under room temperature, without any support material or surfactant. Pr and Ce are the same as lanthanides elements, with similar chemical structures and properties, although most studies employ Ce as the raw material. In order to further evaluate the value of rare earth elements and their wide application, we employed a Pr-modified NiMo catalyst. As a transition metal, Cu is often the main active element in catalysis. A Cu-modified NiMo catalyst is used to synergize the elements to further improve catalytic performance, and the catalytic properties of the Ni 1-x Mo x (Pr 2 O 3 ) y and Ni 1-x Mo x @(Cu 2 O) y materials are investigated and compared. The effects of reaction temperature, concentrations of NaOH, and the kinetics of N 2 H 4 ·H 2 O decomposition are also evaluated in this work.

Physical Characteristics of Ni 1-x Mo x (Pr 2 O 3 ) y and Ni 1-x Mo x @(Cu 2 O) y Nanocatalysts
The one-pot coreduction process provides simple, effective, and readily scalable preparation of nanocatalysts. In the present study, we employ this method to prepare a series of Ni 1-x Mo x (Pr 2 O 3 ) y and Ni 1-x Mo x @(Cu 2 O) y catalysts, where x represents the molar portion of Mo and y represents the molar portion of the third material (Cu 2 O or Pr 2 O 3 ). It should be noted that in the preparation of catalysts, the total masses of Ni and Mo are controlled, to a certain extent. After the modification of the NiMo catalysts, Ni 9 Mo 1 (Pr 2 O 3 ) y , and Ni 8 Mo 2 @(Cu 2 O) y refer to the Ni 0.9 Mo 0.1 (Pr 2 O 3 ) y and Ni 0.8 Mo 0.2 @(Cu 2 O) y nanocatalysts.
X-ray diffraction (XRD) is used to analyze the crystalline structures of the prepared catalysts. Compared with the Ni 9 Mo 1 catalyst, Ni 9 Mo 1 (Pr 2 O 3 ) y showed much weaker Ni peaks, with the (111) peak intensities of the Ni 9 Mo 1 (Pr 2 O 3 ) y samples decreasing gradually with increasing Pr 2 O 3 content. This suggests that the incorporation of Pr into the catalyst may perturb the crystal growth of the Ni nanocrystallite. When the doping ratio of Pr reaches 0.375, the diffraction peaks of Ni are no longer weakened, which indicates that the catalysts of Ni 9 Mo 1 (Pr 2 O 3 ) y can reach a point where there is a coexistence of crystalline and amorphous states, under optimal conditions [21,22]. However, as shown in Figure 1b  The TEM images of NiMo, Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 , and Ni 4 Mo@Cu 2 O, together with the corresponding particle size distributions, are presented in Figure 2. It can be seen from Figure 2a,g that the NiMo catalyst is irregularly dispersed, with a particle size range between 6.5 and 6.9 nm, which is significantly higher than those of the Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O catalysts (5.5−5.9 nm and 6.0-6.3 nm, respectively). The regions with lattice distribution and amorphous states can be clearly seen in high-resolution TEM (HRTEM) images ( Figure 2b). According to fast Fourier-transform (FFT) analysis, the lattice spacing of 0.203 nm is consistent with that of the Ni(111) crystal surface, indicating that the crystalline domain is Ni and the amorphous part is formed through NiMo codoping. The Pr 2 O 3 -doped catalyst still has an irregular granular structure after doping, as shown in Figure 2c. According to HRTEM (Figure 2d) observations and FFT analysis, the Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 nanocatalyst has a crystalline structure with a lattice fringe distance of 0.208 nm, which is slightly larger than that of the (111) plane of FCC Ni (0.203 nm), indicating that the presence of the atom of Pr in the crystal lattice of Ni increases the lattice spacing. This is consistent with XRD analysis showing the weakening of the diffraction peak after the addition of Pr. At the same time, the Ni lattice expands and distorts, creating more oxygen defects that trap more electrons at the bottom of the conduction band. It can be seen from the Figure 2e that the Cu-modified NiMo catalyst has a transparent coating at the edge and an amorphous structure inside. The corresponding HRTEM image (Figure 2f), suggests that the lattice spacing of the surrounding transparent particles is 0.19 nm, consistent with that of the crystal surface of Cu 2 O(111), which indicates that the shell formed in the catalyst is comprised of Cu 2 O. No lattice fringes were observed at the core, indicating that the core is amorphous, which is consistent with the XRD data showing only the diffraction peak of Cu 2 O. This implies that a core-shell structure is formed, with NiMo as the nucleus and Cu 2 O as the shell. Based on the above analyses, the various modified NiMo nanocatalysts form different structures, which is due to the different configurations of extranuclear electrons in the Pr and Cu elements. The 5d empty orbit of Pr is a candidate for electron transfer in the synthesized catalyst, which disturbs the Ni electron configuration and causes the crystal shape to change [20,21]. While Ni, Cu, and Mo are transition metals, their different oxidation abilities play an important role in the formation of the structure [25].    Figure 3. From the XPS spectra, it is apparent that Ni (Figure 3a) is present in the metallic and oxidation state as NiO and Ni(OH) 2 [20,26]. The main Ni peaks for Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O are positively shifted to higher values, compared to that of undoped NiMo. This suggests that Ni plays an important role in the electron transfer process. Doping with Cu leads to a positive shift in the binding energies of the metallic Ni (Ni 2p3/2 852.6 eV), which may be due to a decrease in the electron density and an increase in metal center d-band vacancies [27,28]. However, after Pr doping, the binding energy of Ni 2p3/2 was slightly shifted from 855.  Figure 3e, the prominent peak at 932.9 eV is assigned to Pr (III) 3d5/2, indicating that NaBH 4 did not reduce Pr 3+ to Pr in Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 during catalyst preparation, and the three elements and their oxides are doped with each other to form irregular particles [29,30]. According to the electrode potential, Ni 2+ can be readily reduced to Ni by NaBH 4 , but it is relatively difficult to reduce Mo 6+ to Mo, due to the lower reduction potentials of Mo 6+ /Mo. After the addition of Ni 2+ and Mo 6+ , Ni 2+ is first reduced by NaBH 4 due to its higher potential. However, the oxidizing ability of Cu 2+ is stronger than that of Ni 2+ , and Cu 2+ was reduced first during the reduction, shown in the XPS results in Figure 3f

Catalytic Activities of Ni 1-x Mo x (Pr 2 O 3 ) y and Ni 1-x Mo x @(Cu 2 O) y Nanocatalysts
The catalytic performances of the as-synthesized nanocatalysts, in terms of hydrogen generation from N 2 H 4 ·H 2 O, are shown in Figure 4. Reactions were initiated by introducing N 2 H 4 ·H 2 O into the reaction flask with vigorous shaking, at a specified temperature. The catalytic performances of the catalysts were evaluated in a typical water-filled gas burette system. As show in Figure 4a, elemental nickel and elemental molybdenum have no catalytic activity, and with the increase of NiMo ratio, the catalytic activity appears to initially increase, and then decrease. A NiMo ratio of 9:1 produces the best catalytic effect, with n(N 2 +H 2 )/n(N 2 H 4 ) at 1.9, but the requirement for the highly efficient catalytic hydrogen production of hydrazine hydrate is not achieved. It can be seen from Figure 4b   catalysts, the optimal alkali concentration was determined to be 6 mM. As shown in Figure 5a, when the concentration of NaOH in the system was 6 mM, the TOF achieved over the Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 catalyst was almost ten-fold higher than that obtained in the absence of NaOH, whereas an increase of H2 selectivity from 25% to 92% was also observed. H 2 selectivity from 25% to 92%, compared with catalytic system with no added NaOH. Similarly, it can be seen from Figure 5b that the Ni 4 Mo@Cu 2 O catalyst also showed better performance when 6 mM NaOH was added, where the H 2 selectivity increased to 97% and the TOF increased from 4 h -1 to 71.4 h -1 . The hydrogen production rate and H 2 selectivity show no further change with further increases in the Na OH concentration. We surmise that the promoting effects of alkali on both catalytic activity and selectivity might be associated with the enhancement of the reducibility of N 2 H 4 . As seen in Equations (1) and (2), increasing OHconcentration not only enhances conversion from N 2 H 5 + to N 2 H 4 , but is also responsible for decreasing the basic byproduct NH 3 , enhancing the selectivity of H 2 [22,32,35]. Furthermore, high concentration of OHenhanced the hydrolysis of N 2 H 4 ·H 2 O, while the decomposition rate of catalytic substrate is improved. The N 2 H 4 ·H 2 O gas production curve and the Arrhenius diagram of ln(TOF) and 1/T from 298 to 343 K are shown in Figure 6. It can be seen from Figure 6a,c that the catalysts show poor catalytic activity at room temperature, and the gas production rate and hydrogen production of N 2 H 4 ·H 2 O increase gradually with increasing reaction temperature. When the reaction temperature is 80 • C, the reaction rate and gas production are not changed, and so a reaction temperature of 70 • C is used in the experiments described in this paper. The Arrhenius plot of ln(TOF) vs. 1/T for our catalysts are plotted in Figure 6b   High durability and stability are crucial to the practical application of a catalyst. Therefore, in the present study, durability tests were performed on Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O for the same decomposition reaction at 343 K, by adding aqueous hydrous hydrazine (2 mM) to the catalyst after the last reaction round is completed. Time-course plots for the decomposition of N 2 H 4 ·H 2 O that was catalyzed by Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 are shown in Figure 7a, while Figure 7b shows the corresponding plots for Ni 4 Mo@Cu 2 O. As shown in Figure 7, the catalytic activity of the Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 catalyst exhibits no significant decline even after four rounds of decomposition reactions, indicating that the catalyst is highly stable. However, the Ni 4 Mo@Cu 2 O catalyst exhibits obvious activity attenuation throughout cyclic usage. We may speculate that this observed activity decay could be attributed to the following possible mechanisms. First, the structure and morphology of the catalysts may have changed because as the number of cycles increases, both the partial agglomeration of metal nanoparticles and amount of exposed active sites decrease. Second, the decay could be caused by the surface oxidation of the catalysts, although as the catalyst is not cleaned by centrifugal separation during the catalytic process, which reduces the area and time of contact of the catalyst with air, this is less likely than the first potential mechanism [37,38].

Characterization
Powder XRD data were obtained using an X-ray diffractometer with Cu kα radiation, at an operational voltage and current maintained at 40 kV and 40 mA, respectively. The scanning range for the spectra was 2θ = 10-90 • , at a scanning rate of 8 • /min. The sample morphologies and sizes were analyzed using a JEM-2001F field emission transmission electron microscope, manufactured by JEOL Ltd., Tokyo, Japan. XPS measurements were conducted on an ESCALABMKLL 250Xi spectrometer using an aluminum kα source. The energy of the radiation was hν = 1486.6 eV, while the spectrometer was operated at 12.5 kV and 16 mA. The Ar sputtering experiments were performed under a background vacuum of 8 × 10 −10 Pa. The binding energies were calibrated using the C 1 s peak (284.6 eV) of adventitious carbon, and curve fitting was performed using the XPS PEAK 4.1 software. Particle size distribution was measured using a dynamic light scattering nanometer laser particle size analyzer (GS90). The catalyst powder was evenly dispersed in the ultra-pure water by ultrasonic waves before the particle size distribution of the solution was analyzed. It is of note that Cu must be added before the Na 2 MoO 4 ·2H 2 O, where the Na 2 MoO 4 ·2H 2 O was alkaline, Cu 2+ was easily precipitated when it encountered an alkaline solution, and so after adding the Na 2 MoO 4 ·2H 2 O, NaBH 4 was quickly added to avoid precipitation. Ni 1-x Mo x (Pr 2 O 3 ) y and Ni 1-x Mo x @(Cu 2 O) y catalysts with other molar ratios were prepared using the same procedure by adjusting the relative amounts of precursors, and these catalysts were used for characterization tests after vacuum drying.

Catalysis of Hydrazine Hydrate to Produce Hydrogen
The catalytic reaction of hydrous hydrazine to produce hydrogen by Ni 1-x Mo x (Pr 2 O 3 ) y and Ni 1-x Mo x @(Cu 2 O) y were tested at 343 K with vigorous stirring. The catalysts were kept in a 50-mL two-neck round-bottomed flask, one neck was connected to a gas burette to monitor the volume of H 2 released from hydrous hydrazine hydrolysis, while the other neck was used for the introduction of hydrous hydrazine (0.2 mL, 10 mol L −1 ). In order to maintain the reaction temperature, the flask was placed in a thermostat equipped with a water circulating system, and the reaction was determined to be complete when there was no further gas generation. The molar ratios of n(metal)/n(N 2 H 4 ) were theoretically set to 0.11 and 0.14 for the catalytic reactions of Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O, respectively. The selectivity towards H 2 generation (X) can be calculated using Equation (3), while the TOF was calculated using Equation (4).
where TOF refers to the initial turnover frequency, n(N 2 H 4 ) 50% is the amount of hydrous hydrazine when it is half dissolved, n (nanocatalysts) is the mole amount of the metal, and t 50% is the reaction time when conversion reached 50%.

Durability Testing for the Catalysts
In order to test the durability of the Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O catalysts, after the hydrogen generation reaction was completed the first time, 0.2 mL of hydrazine monohydrate was subsequently added into the reaction flask. Such cycles for testing the N 2 H 4 ·H 2 O decomposition of the catalyst were conducted four times at 343 K.

Conclusions
In summary, we demonstrated a facile coreduction approach to preparing Pr 2 O 3 and Cu 2 O-doped NiMo catalysts, without the need for any surfactant or support material, under ambient conditions. The resulting Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O catalysts exhibit good catalytic performances in terms of N 2 H 4 ·H 2 O decomposition, which surpassed that of NiMo catalysts that were produced without Pr 2 O 3 or Cu 2 O doping. The effects of material ratios, NaOH concentration, and reaction temperature on the catalytic efficiency of the catalysts were studied through characterization and catalytic performance evaluation, and compared with unmodified NiMo catalysts. The results show that in the process of modifying NiMo, the NiMoPr 2 O 3 is amorphous in structure, and characterized by low crystallinity and an increased number of oxygen vacancies that are beneficial to active site exposure and particle dispersion. NiMo@Cu 2 O exhibited a core-shell structure, in which Cu 2 O is the shell and NiMo is the core, and the electrons in the core and shell are coupled to each other, with an electron-transfer efficiency that improves catalytic efficiency. Among all the prepared catalysts, Ni 9 Mo 1 (Pr 2 O 3 ) 0.375 and Ni 4 Mo@Cu 2 O showed the highest H 2 selectivity in the presence of 0.2 mL of 10 M N 2 H 4 ·H 2 O and 6 mM NaOH at 343 K, with TOFs of 62 h −1 and 71.4 h −1 , respectively. These values are higher than the majority of reported results for similar systems The diameters of these well-dispersed ultrafine catalysts were approximately 5.5-5.9 nm and 6.0-6.3 nm, respectively, lower than that of NiMo catalyst particles (approximately 6.7 nm), which led to an increase in the contact area between N 2 H 4 ·H 2 O and the catalyst, and promoted decomposition. This remarkable improvement in the catalytic performance of precious metal-free nanocatalysts demonstrates a promising strategy for future hydrogen production from hydrous hydrazine.

Conflicts of Interest:
The authors declare no conflicts of interest.