Hydrogen Production from Ammonia Borane over PtNi Alloy Nanoparticles Immobilized on Graphite Carbon Nitride

Graphite carbon nitride (g-C3N4) supported PtNi alloy nanoparticles (NPs) were fabricated via a facile and simple impregnation and chemical reduction method and explored their catalytic performance towards hydrogen evolution from ammonia borane (AB) hydrolysis dehydrogenation. Interestingly, the resultant Pt0.5Ni0.5/g-C3N4 catalyst affords superior performance, including 100% conversion, 100% H2 selectivity, yielding the extraordinary initial total turnover frequency (TOF) of 250.8 molH2 min−1 (molPt)−1 for hydrogen evolution from AB at 10 ◦C, a relatively low activation energy of 38.09 kJ mol−1, and a remarkable reusability (at least 10 times), which surpass most of the noble metal heterogeneous catalysts. This notably improved activity is attributed to the charge interaction between PtNi NPs and g-C3N4 support. Especially, the nitrogen-containing functional groups on g-C3N4, serving as the anchoring sites for PtNi NPs, may be beneficial for becoming a uniform distribution and decreasing the particle size for the NPs. Our work not only provides a cost-effective route for constructing high-performance catalysts towards the hydrogen evolution of AB but also prompts the utilization of g-C3N4 in energy fields.


Introduction
With the ever-growing consumption of fossil energy, accompanied with serious environmental issues, searching for green, sustainable, abundant, and alternative energy sources is of burgeoning urgency [1,2].Hydrogen, as a clean energy source, has attracted significant research interest owing to its distinctive merits such as producing only water as a by-product and possessing more energy density than that of fossil fuels [3][4][5].However, hydrogen, possessing the feature of low density, makes it difficult to liquefy and compress, thus hindering the large-scale applications [4][5][6].Therefore, the exploration and seek for hydrogen storage materials with outstanding performance remains a challenging issue.
Tremendous efforts, in the past decades, have been made to explore and design hydrogen storage materials such as hydrazine, formic acid, and ammonia borane and so forth [7][8][9][10][11][12][13].Among various hydrogen storage materials conducted, ammonia borane (AB), a white solid with excellent stability at room temperature and high hydrogen content (19.6 wt%), has aroused considerable interest as a promising hydrogen storage material [14,15].There are two main approaches for AB to release hydrogen: (i) thermal dehydrogenation and (ii) hydrolytic dehydrogenation [14][15][16].However, compared with thermal dehydrogenation, proceeding under high temperature, hydrolysis dehydrogenation process is easier to accomplish the industrial application under mild conditions [15][16][17].With the aid of the suitable catalyst, 1 mol of AB can be controlled to release 3 mol of hydrogen under moderate conditions.The catalytic hydrolysis reaction can be described in detail as follows [18][19][20][21][22]: In recent years, a large number of supported metal catalysts including noble metal and non-noble metal, have been constructed for the hydrolysis of AB, among which Pt-based noble catalysts, as one of the most studied catalysts both in industrial and scientific fields, demonstrate more extraordinary catalytic performance than that of other metal catalysts for hydrolysis of AB [20][21][22][23][24][25][26][27].However, the features of Pt with its high prices and limited reserves severely restrict its extensive utilization of Pt metal as the catalyst [27][28][29].Currently, to actually reduce the utilization amount of Pt, assembling with the enhancement of catalysts, the fabrication of Pt alloy nanomaterials, especially coupling with the noble-free metals, such as Fe, Co, Ni, etc., has been identified to be an effective approach due to its structural and electronic effects [30][31][32][33].Previous studies have revealed that Pt-M bimetallic catalysts could display higher catalytic activity than that of their single counterparts as well as lower cost [32][33][34][35][36].For instance, Han et al. applied a chemical reduction route to synthesize amino modified SiO 2 nanospheres supported CoPt-Co hybrid at 278 K, affording a turnover frequency (TOF) value of 25.59 mol H2 min −1 mol M −1 , almost two-fold as high as that of unsupported Pt 0.1 Co 0.9 nanoparticles (NPs) [37].Xu et al. fabricated hierarchical nanoporous PtCu alloy nanoflowers by means of the two-step dealloying method with a maximum TOF value up to 108 mol H2 min −1 mol M −1 [38].Lu et al. reported the synthesis of PtNi/NiO clusters coated by hollow silica using atomic sacrifice method, reaching a TOF of 1240.3 mol H2 min −1 mol Pt −1 [39].However, Pt-M NPs are readily tending to aggregate to form a bigger particle, resulting in a momentous decrease in the catalytic performance including activity, stability, and efficiency.Therefore, in order to obtain high catalytic performance for hydrolytic dehydrogenation of AB, considerable effort has been devoted to fabricate high efficiency catalysts involving many types of supporting materials.It is noted that assembling alloy with appropriate supports, such as metal oxides, carbon nanotubes, mesoporous carbon nitride, graphene, etc., emerges as one of the novel encouraging approaches to further improve the catalytic activities and stability [39][40][41][42][43][44][45][46].It is well known that the supporting materials play a critical role in enhancing its catalytic properties due to the synergistic effects between alloy and supports.Graphitic carbon nitride (g-C 3 N 4 ), a promising two-dimensional non-metal material, is regarded as a promising candidate owing to its attractive electronic structure, high nitrogen content, excellent chemical and thermal stability, and environment friendliness [45][46][47][48] Encouraged by these achievements, the construction of PtNi alloy anchored into g-C 3 N 4 for hydrogen production from the hydrolysis of AB is of paramount significance and has rarely been reported.Additionally, investigating the construction of bi-metal NPs in g-C 3 N 4 as well as the exploration of the synergistic effect between them are also of outmost importance.Hence, in this work, we focused on the construction of PtNi with different ratios immobilized on g-C 3 N 4 , preparing via the direct pyrolysis of melamine under nitrogen atmosphere [51,52], through a simple impregnation and co-reduction method under an ambient atmosphere.The resulting materials were evaluated as catalysts towards the hydrogen production from AB hydrolysis dehydrogenation under mild conditions.The influence of some parameters, such as metal concentration, catalyst concentration, AB concentration, and reaction temperature, on the catalytic performance of PtNi/g-C 3 N 4 were conducted in detail.Especially, the Pt 0.5 Ni 0.5 /g-C 3 N 4 demonstrates optimal catalytic performance for the hydrolysis of AB compared with the samples of other molar ratios of PtNi.Furthermore, the stability of the optimum catalyst was investigated as well.

Results and Discussion
The synthesis process of PtNi/g-C 3 N 4 is schematically emerged in Scheme 1.Typically, g-C 3 N 4 was prepared via a direct pyrolysis route employing melamine as precursor.Then, the Pt and Ni precursor with different contents immobilized on g-C 3 N 4 was fabricated, using a simple impregnation and reduction process, where sodium borohydride is employed as a reducing agent, and then applied as catalysts towards hydrogen production of AB.The accurate composition of PtNi/g-C 3 N 4 was analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES), which were close to their desired contents (Table S1).Elemental analysis revealed that the composition of C, H, and N of g-C 3 N 4 was 34.71 wt%, 1.89 wt%, and 61.47 wt%, respectively.As displayed in Table S1, Figure S1, the Brunner−Emmet−Teller (BET) specific areas of PtNi/g-C 3 N 4 was around 10 m 2 /g, which was close to the values reported in the literature [51,52].Figure 1 shows the XRD patterns of PtNi/g-C 3 N 4 with different contents.optimal catalytic performance for the hydrolysis of AB compared with the samples of other molar ratios of PtNi.Furthermore, the stability of the optimum catalyst was investigated as well.

Results and Discussion
The synthesis process of PtNi/g-C3N4 is schematically emerged in Scheme 1.Typically, g-C3N4 was prepared via a direct pyrolysis route employing melamine as precursor.Then, the Pt and Ni precursor with different contents immobilized on g-C3N4 was fabricated, using a simple impregnation and reduction process, where sodium borohydride is employed as a reducing agent, and then applied as catalysts towards hydrogen production of AB.The accurate composition of PtNi/g-C3N4 was analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES), which were close to their desired contents (Table S1).Elemental analysis revealed that the composition of C, H, and N of g-C3N4 was 34.71 wt%, 1.89 wt%, and 61.47 wt%, respectively.As displayed in Table S1, Figure S1, the Brunner−Emmet−Teller (BET) specific areas of PtNi/g-C3N4 was around 10 m 2 /g, which was close to the values reported in the literature [51,52].Figure 1 shows the XRD patterns of PtNi/g-C3N4 with different contents.Catalysts 2019, 9, x FOR PEER REVIEW 3 of 15 optimal catalytic performance for the hydrolysis of AB compared with the samples of other molar ratios of PtNi.Furthermore, the stability of the optimum catalyst was investigated as well.

Results and Discussion
The synthesis process of PtNi/g-C3N4 is schematically emerged in Scheme 1.Typically, g-C3N4 was prepared via a direct pyrolysis route employing melamine as precursor.Then, the Pt and Ni precursor with different contents immobilized on g-C3N4 was fabricated, using a simple impregnation and reduction process, where sodium borohydride is employed as a reducing agent, and then applied as catalysts towards hydrogen production of AB.The accurate composition of PtNi/g-C3N4 was analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES), which were close to their desired contents (Table S1).Elemental analysis revealed that the composition of C, H, and N of g-C3N4 was 34.71 wt%, 1.89 wt%, and 61.47 wt%, respectively.As displayed in Table S1, Figure S1, the Brunner−Emmet−Teller (BET) specific areas of PtNi/g-C3N4 was around 10 m 2 /g, which was close to the values reported in the literature [51,52].Figure 1 shows the XRD patterns of PtNi/g-C3N4 with different contents.It is evident that a strong peak at 27.5 • for all samples was attributed to the (200) plane of g-C 3 N 4 .For Ni/g-C 3 N 4 , other than the characteristic peak of g-C 3 N 4 , there were three peaks centered at 44.5 • , 51.8 • , and 76.4 • , respectively, which could be ascribed to the (111), (200), and (220) planes of Ni (JCPDS no: 65-0380).For Pt/g-C 3 N 4 , three peaks at 39.8 • , 46.2 • , and 67.5 • were assigned to the (111), (200), and (220) planes of Pt (JCPDS no: 65-2868).Furthermore, it was evident that a shift to a higher angle compared to that of the Pt (111) peak in Pt/g-C 3 N 4 , indicating that PtNi existed in the form of an alloy, which was consistent with the previous literatures yet reported [39,[53][54][55].The chemical structures of g-C 3 N 4 , Pt/g-C 3 N 4 , Ni/g-C 3 N 4 , and Pt 0.5 Ni 0.5 /g-C 3 N 4 were further investigated by Fourier transform infrared (FTIR), as presented in Figure 2. A similar FT-IR spectra could be observed for all samples.An obvious absorption peak centered at 810 cm −1 for all samples was ascribed to the bending vibration of the s-triazine ring.A series of peaks detected in 1200-1600 cm −1 were identified as the stretching modes of aromatic CN heterocycles.In addition, the broad absorption ranging from 3000 to 3400 cm −1 was attributed to the stretching mode of O-H (adsorbed water molecules) and N-H (amino groups) [31][32][33][34][35][36]51,52].
It is evident that a strong peak at 27.5° for all samples was attributed to the (200) plane of g-C3N4.For Ni/g-C3N4, other than the characteristic peak of g-C3N4, there were three peaks centered at 44.5°, 51.8°, and 76.4°, respectively, which could be ascribed to the (111), (200), and (220) planes of Ni (JCPDS no: 65-0380).For Pt/g-C3N4, three peaks at 39.8°, 46.2°, and 67.5° were assigned to the (111), (200), and (220) planes of Pt (JCPDS no: 65-2868).Furthermore, it was evident that a shift to a higher angle compared to that of the Pt (111) peak in Pt/g-C3N4, indicating that PtNi existed in the form of an alloy, which was consistent with the previous literatures yet reported [39,[53][54][55].The chemical structures of g-C3N4, Pt/g-C3N4, Ni/g-C3N4, and Pt0.5Ni0.5/g-C3N4were further investigated by Fourier transform infrared (FTIR), as presented in Figure 2. A similar FT-IR spectra could be observed for all samples.An obvious absorption peak centered at 810 cm −1 for all samples was ascribed to the bending vibration of the s-triazine ring.A series of peaks detected in 1200-1600 cm −1 were identified as the stretching modes of aromatic CN heterocycles.In addition, the broad absorption ranging from 3000 to 3400 cm −1 was attributed to the stretching mode of O-H (adsorbed water molecules) and N-H (amino groups) [31][32][33][34][35][36]51,52].To gain the valence state of Pt and Ni, X-ray photoelectron spectroscopy (XPS) analysis was employed for Pt0.5Ni0.5/g-C3N4,and the result is illustrated in Figure 3, which presents the characteristic peaks for Pt and Ni, thus implying the coexistence of both metals.The binding energy (BE) of C and N in Pt0.5Ni0.5/g-C3N4were in accordance with the previous reported literatures about g-C3N4 (Figure S2) [50][51][52].As given in Figure 3a, the BE observed at 852.5 and 869.9 eV were ascribed to the Ni 2p3/2 and Ni 2p1/2 of metallic Ni, respectively.The oxidized Ni centered at 858.1 and 876.5 eV were detected, which might likely attribute to the partial oxidation of Ni during the sample treatment route for the XPS measurements, as revealed before [56][57][58].The BE of 71.4 and 74.6 eV in Figure 3b corresponded to the Pt 4f7/2 and Pt 4f5/2 of metallic Pt.In addition, there was a positive shift toward 0.1 eV for the BE of Pt in Pt0.5Ni0.5/g-C3N4compared with that of pure Pt, while a negative shift with 0.4 eV could be observed for Ni in Pt0.5Ni0.5/g-C3N4 in comparison with that of pure Ni, thereby confirming the formation of the PtNi alloy, which was in good agreement with those of alloys previously reported [56][57][58][59][60][61].4000 3500 3000 2500 2000 1500 1000 500 Wavenumber(cm -1 ) To gain the valence state of Pt and Ni, X-ray photoelectron spectroscopy (XPS) analysis was employed for Pt 0.5 Ni 0.5 /g-C 3 N 4 , and the result is illustrated in Figure 3, which presents the characteristic peaks for Pt and Ni, thus implying the coexistence of both metals.The binding energy (BE) of C and N in Pt 0.5 Ni 0.5 /g-C 3 N 4 were in accordance with the previous reported literatures about g-C 3 N 4 (Figure S2) [50][51][52].As given in Figure 3a, the BE observed at 852.5 and 869.9 eV were ascribed to the Ni 2p 3/2 and Ni 2p 1/2 of metallic Ni, respectively.The oxidized Ni centered at 858.1 and 876.5 eV were detected, which might likely attribute to the partial oxidation of Ni during the sample treatment route for the XPS measurements, as revealed before [56][57][58].The BE of 71.4 and 74.6 eV in Figure 3b corresponded to the Pt 4f 7/2 and Pt 4f 5/2 of metallic Pt.In addition, there was a positive shift toward 0.1 eV for the BE of Pt in Pt 0.5 Ni 0.5 /g-C 3 N 4 compared with that of pure Pt, while a negative shift with 0.4 eV could be observed for Ni in Pt 0.5 Ni 0.5 /g-C 3 N 4 in comparison with that of pure Ni, thereby confirming the formation of the PtNi alloy, which was in good agreement with those of alloys previously reported [56][57][58][59][60][61].
To further explore the morphology and microstructure of the Pt 0.5 Ni 0.5 /g-C 3 N 4 catalyst, we conducted TEM, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), and energy-dispersive X-ray spectroscopic (EDS) measurements, as depicted in Figure 4a-i.The TEM images of Ni/g-C 3 N 4 revealed that almost all the nanoparticles were uniformly dispersed into the surface of g-C 3 N 4 with an average size of 4.3 nm (Figure S3a,b).Furthermore, for Pt/g-C 3 N 4 , there were many homogeneous distribution of nanoparticles with their size ranging from 3.6 to 6.0 nm, as depicted in Figure S3d,e.However, TEM images revealed that the PtNi nanoparticles in Pt 0.5 Ni 0.5 /g-C 3 N 4 catalyst had a uniform diameter distribution with a mean diameter of 3.2 nm and were homogeneously dispersed on the surface of g-C 3 N 4 (Figure 4a-c), which might be due to the alloy effect.Figure 4b presents that the d-spacing of the particle was 0.211 nm, which differed from the (111) plane of Pt (0.227 nm; Figure S3f) and Ni (0.204 nm; Figure S3c), further revealing that Pt 0.5 Ni 0.5 was in the form of the alloy state [54][55][56][58][59][60][61].The energy dispersive X-ray (EDX) result (Figure S4 To further explore the morphology and microstructure of the Pt0.5Ni0.5/g-C3N4catalyst, we conducted TEM, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), and energy-dispersive X-ray spectroscopic (EDS) measurements, as depicted in Figure 4a-i.The TEM images of Ni/g-C3N4 revealed that almost all the nanoparticles were uniformly dispersed into the surface of g-C3N4 with an average size of 4.3 nm (Figure S3a,b).Furthermore, for Pt/g-C3N4, there were many homogeneous distribution of nanoparticles with their size ranging from 3.6 to 6.0 nm, as depicted in Figure S3d,e.However, TEM images revealed that the PtNi nanoparticles in Pt0.5Ni0.5/g-C3N4catalyst had a uniform diameter distribution with a mean diameter of 3.2 nm and were homogeneously dispersed on the surface of g-C3N4 (Figure 4a-c), which might be due to the alloy effect.Figure 4b presents that the d-spacing of the particle was 0.211 nm, which differed from the (111) plane of Pt (0.227 nm; Figure S3f) and Ni (0.204 nm; Figure S3c), further revealing that Pt0.5Ni0.5 was in the form of the alloy state [54][55][56][58][59][60][61].The energy dispersive X-ray (EDX) result (Figure S4) further verified the coexistence of Pt and Ni in the Pt0.5Ni0.5/g-C3N4.A representative highangle annular dark-field scanning TEM (HAADF-STEM) image confirmed the homogeneous distribution of Pt and Ni elements on the g-C3N4 catalyst surface at the same position (Figure 4d-i), which confirmed again the formation of a PtNi alloy.Figure 5a presents the hydrogen release from AB catalyzed by PtxNi1-x/g-C3N4 catalyst with different components.Among all the PtxNi1-x/g-C3N4 conducted, Pt0.5Ni0.5/g-C3N4revealed optimum catalytic performance towards the dehydrogenation of AB in comparison with that of another molar ratio of PtxNi1-x/g-C3N4.The hydrogen production from AB hydrolysis dehydrogenation over Pt0.5Ni0.5/g-C3N4could be completed within only 1.5 min, giving a TOF value of 250.8 molH2 min −1 (molPt) −1 (Figure 5b), which was the highest as compared with other as-prepared PtNi alloy catalysts, Figure 5a presents the hydrogen release from AB catalyzed by Pt x Ni 1-x /g-C 3 N 4 catalyst with different components.Among all the Pt x Ni 1-x /g-C 3 N 4 conducted, Pt 0.5 Ni 0.5 /g-C 3 N 4 revealed optimum catalytic performance towards the dehydrogenation of AB in comparison with that of another molar ratio of Pt x Ni 1-x /g-C 3 N 4 .The hydrogen production from AB hydrolysis dehydrogenation over Pt 0.5 Ni 0.5 /g-C 3 N 4 could be completed within only 1.5 min, giving a TOF value of 250.8 mol H2 min −1 (mol Pt ) −1 (Figure 5b), which was the highest as compared with other as-prepared PtNi alloy catalysts, as well higher than most reported Pt-based or other noble metal-based catalysts towards the hydrogen evolution of AB hydrolysis dehydrogenation as displayed in Table 1 [38,40,[62][63][64][65][66][67][68].The results showed that the platinum coupling with nickel could significantly improve the hydrolytic activity of AB.The superior catalytic performance of the Pt 0.5 Ni 0.5 /g-C 3 N 4 can be ascribed to the synergetic effect between Pt and Ni atoms as well as the enhanced interaction between PtNi NPs and g-C 3 N 4 .Given that Pt0.5Ni0.5/g-C3N4demonstrated the best catalytic performance towards the dehydrogenation of AB in this work, Pt0.5Ni0.5/g-C3N4,as the representative, was selected to further explore its dehydrogenation kinetics.As illustrated in Figure 6a, the catalyst concentrationsdependent test of Pt0.5Ni0.5 NPs was investigated at 10 °C by modifying the catalyst concentration ranging from 6.25 to 25 mM.The hydrogen generation rate demonstrated an obvious upward tendency with the increasing of catalyst concentrations.The relation between logarithmic plots of dehydrogenation rate and catalyst concentrations is illustrated in Figure 6b.The slope of straight line was estimated to be 1.08, implying that the dehydrogenation reaction was first-order in terms of the catalyst concentration.Furthermore, to explore the influence of AB concentration on the dehydrogenation of AB, a series of experiments with different concentrations of AB were performed, where the AB concentrations were modulated from 250 to 1000 mM, as shown in Figure 6a, and Pt0.5Ni0.5/g-C3N4NPs was kept at 0.100g.Figure 7a shows corresponding hydrogen amount versus time.As the concentration of AB increased, the dehydrogenation reaction time increased clearly from 0.8 to 1.5 min.Figure 7b displays logarithmic plots of dehydrogenation rate versus AB concentrations,  Given that Pt 0.5 Ni 0.5 /g-C 3 N 4 demonstrated the best catalytic performance towards the dehydrogenation of AB in this work, Pt 0.5 Ni 0.5 /g-C 3 N 4 , as the representative, was selected to further explore its dehydrogenation kinetics.As illustrated in Figure 6a, the catalyst concentrations-dependent test of Pt 0.5 Ni 0.5 NPs was investigated at 10 • C by modifying the catalyst concentration ranging from 6.25 to 25 mM.The hydrogen generation rate demonstrated an obvious upward tendency with the increasing of catalyst concentrations.The relation between logarithmic plots of dehydrogenation rate and catalyst concentrations is illustrated in Figure 6b.The slope of straight line was estimated to be 1.08, implying that the dehydrogenation reaction was first-order in terms of the catalyst concentration.Furthermore, to explore the influence of AB concentration on the dehydrogenation of AB, a series of experiments with different concentrations of AB were performed, where the AB concentrations were modulated from 250 to 1000 mM, as shown in Figure 6a, and Pt 0.5 Ni 0.5 /g-C 3 N 4 NPs was kept at 0.100 g. Figure 7a shows corresponding hydrogen amount versus time.As the concentration of AB increased, the dehydrogenation reaction time increased clearly from 0.8 to 1.5 min.Figure 7b displays logarithmic plots of dehydrogenation rate versus AB concentrations, where the line slope was 0.05, suggesting that the dehydrogenation reaction of AB presents zero-order regarding the catalyst concentrations.To explore the influence of temperature, the AB concentration was maintained at 1000 mM and the Pt0.5Ni0.5/g-C3N4was kept at 0.100 g.Plot of the molar ratio of H2/AB versus time for hydrogen evolution from AB catalyzed by Pt0.5Ni0.5/g-C3N4catalyst at different temperatures are presented in Figure 8a.As the reaction temperature changed from −5 to 10 °C, the dehydrogenation rate significantly increased.According to the slop of the straight line in Figure 8b, the activation energy was estimated to be 38.09kJ/mol, which was lower than most of the reported Ea values of many different Pt-based and some other catalysts [62][63][64][65][66][67][68].To explore the influence of temperature, the AB concentration was maintained at 1000 mM and the Pt0.5Ni0.5/g-C3N4was kept at 0.100 g.Plot of the molar ratio of H2/AB versus time for hydrogen evolution from AB catalyzed by Pt0.5Ni0.5/g-C3N4catalyst at different temperatures are presented in Figure 8a.As the reaction temperature changed from −5 to 10 °C, the dehydrogenation rate significantly increased.According to the slop of the straight line in Figure 8b, the activation energy was estimated to be 38.09kJ/mol, which was lower than most of the reported Ea values of many different Pt-based and some other catalysts [62][63][64][65][66][67][68].To explore the influence of temperature, the AB concentration was maintained at 1000 mM and the Pt 0.5 Ni 0.5 /g-C 3 N 4 was kept at 0.100 g.Plot of the molar ratio of H 2 /AB versus time for hydrogen evolution from AB catalyzed by Pt 0.5 Ni 0.5 /g-C 3 N 4 catalyst at different temperatures are presented in Figure 8a.As the reaction temperature changed from −5 to 10 • C, the dehydrogenation rate significantly increased.According to the slop of the straight line in Figure 8b, the activation energy was estimated to be 38.09kJ/mol, which was lower than most of the reported E a values of many different Pt-based and some other catalysts [62][63][64][65][66][67][68].
Combining the XRD, FT-IR, XPS, and TEM, a possible mechanism for the hydrogen evolution from ammonia borane catalyzed by Pt 0.5 Ni 0.5 /g-C 3 N 4 can be proposed.The obtained results endow persuasive and obvious evidence regarding the synergistic effects between Pt and Ni in the Pt 0.5 Ni 0.5 NPs and the synergistic electronic effects between Pt 0.5 Ni 0.5 nanoparticles and g-C 3 N 4 , which could efficiently activate the B-N bonds in AB, thereby lowering the reaction energy barrier and prompting hydrogen evolution, as verified by Table S1.As shown in Scheme 2, firstly, both H 2 O and AB were adsorbed on the surface of the catalyst, and then, the B-H bond of AB was broken to form H 3 NBH 2 −OH by attacking BH 3 * group in H 3 NBH 3 using OH*, the OH* further attacked other B-H to dissociate hydrogen atoms.At last, the dissociated hydrogen atoms in the surface of the catalysts could combine to release hydrogen gas [69][70][71][72].In addition, the active energy for the dehydrogenation catalyzed by Pt 0.5 Ni 0.5 /g-C 3 N 4 was calculated to as low as 38.09 kJ mol −1 and its kinetics were also obviously improved.To explore the influence of temperature, the AB concentration was maintained at 1000 mM and the Pt0.5Ni0.5/g-C3N4was kept at 0.100 g.Plot of the molar ratio of H2/AB versus time for hydrogen evolution from AB catalyzed by Pt0.5Ni0.5/g-C3N4catalyst at different temperatures are presented in Figure 8a.As the reaction temperature changed from −5 to 10 °C, the dehydrogenation rate significantly increased.According to the slop of the straight line in Figure 8b, the activation energy was estimated to be 38.09kJ/mol, which was lower than most of the reported Ea values of many different Pt-based and some other catalysts [62][63][64][65][66][67][68].Combining the XRD, FT-IR, XPS, and TEM, a possible mechanism for the hydrogen evolution from ammonia borane catalyzed by Pt0.5Ni0.5/g-C3N4can be proposed.The obtained results endow persuasive and obvious evidence regarding the synergistic effects between Pt and Ni in the Pt0.5Ni0.5 NPs and the synergistic electronic effects between Pt0.5Ni0.5 nanoparticles and g-C3N4, which could efficiently activate the B-N bonds in AB, thereby lowering the reaction energy barrier and prompting hydrogen evolution, as verified by Table S1.As shown in Scheme 2, firstly, both H2O and AB were adsorbed on the surface of the catalyst, and then, the B-H bond of AB was broken to form H3NBH2−OH by attacking BH3* group in H3NBH3 using OH*, the OH* further attacked other B-H to dissociate hydrogen atoms.At last, the dissociated hydrogen atoms in the surface of the catalysts could combine to release hydrogen gas [69][70][71][72].In addition, the active energy for the dehydrogenation catalyzed by Pt0.5Ni0.5/g-C3N4was calculated to as low as 38.09 kJ mol −1 and its kinetics were also obviously improved.The stability, as a significant issue for the large-scale application of one catalyst, of the Pt0.5Ni0.5/g-C3N4 in the hydrolytic dehydrogenation of AB was studied and displayed in Figure 9a.However, after 10 successive cycles, the hydrolysis of AB can also be completed within 3 min, generating a TOF value of 136.8 molH2 min −1 (molPt) −1 , which is only 55% of its initial catalytic activity.To further explore the reason for the decreased dehydrogenation performance, the morphology of Pt0.5Ni0.5/g-C3N4after the cycle test was characterized by TEM, as presented in Figure 9b.The slight aggregation can be observed, which may be attributed to the activity decay of Pt0.5Ni0.5/g-C3N4.The stability, as a significant issue for the large-scale application of one catalyst, of the Pt 0.5 Ni 0.5 /g-C 3 N 4 in the hydrolytic dehydrogenation of AB was studied and displayed in Figure 9a.However, after 10 successive cycles, the hydrolysis of AB can also be completed within 3 min, generating a TOF value of 136.8 mol H2 min −1 (mol Pt ) −1 , which is only 55% of its initial catalytic activity.To further explore the reason for the decreased dehydrogenation performance, the morphology of Pt 0.5 Ni 0.5 /g-C 3 N 4 after the cycle test was characterized by TEM, as presented in Figure 9b.The slight aggregation can be observed, which may be attributed to the activity decay of Pt 0.5 Ni 0.5 /g-C 3 N 4 .
C3N4 in the hydrolytic dehydrogenation of AB was studied and displayed in Figure 9a.However, after 10 successive cycles, the hydrolysis of AB can also be completed within 3 min, generating a TOF value of 136.8 molH2 min −1 (molPt) −1 , which is only 55% of its initial catalytic activity.To further explore the reason for the decreased dehydrogenation performance, the morphology of Pt0.5Ni0.5/g-C3N4after the cycle test was characterized by TEM, as presented in Figure 9b.The slight aggregation can be observed, which may be attributed to the activity decay of Pt0.5Ni0.5/g-C3N4.

Synthesis of g-C 3 N 4
Pure g-C 3 N 4 was prepared by the direct pyrolysis of melamine, as reported by previous literatures [51,52].Specifically, 2.00 g of melamine, placed into an alumina crucible with a cover, was annealed at 520 • C in a muffle furnace for 2 h at a ramp rate of 3 • C/min.After cooling to room temperature naturally, the resulting yellow-colored material was g-C 3 N 4 .

Synthesis of PtNi/g-C 3 N 4
PtNi/g-C 3 N 4 was fabricated via a simple and facile impregnation and chemical reduction method.In a typical procedure, for Pt 0.5 Ni 0.5 /g-C 3 N 4 , 200 mg of g-C 3 N 4 and 0.4 mmol of metal precursors (0.2 mmol of H 2 PtCl 6 and 0.2 mmol of NiCl 2 ) were firstly mixed under ultrasonic for 30 min, and then the mixture was continuously stirred for 24 h at room temperature.After that, a mixture solution containing 0.08 g of NaBH 4 was added the above-mentioned solution at −3 • C, and then stirring for 5 h.Finally, the resulting mixture was centrifuged and dried at 80 • C in a vacuum oven.The obtained powder was denoted as Pt 0.5 Ni 0.5 /g-C 3 N 4 .The g-C 3 N 4 supported Pt/Ni with other molar ratios could be labeled as Pt x Ni 1-x /g-C 3 N 4 (x = 0, 0.2, 0.4, 0.6, 0.8, and 1).

Characterization
The crystal structure of the catalysts was recorded at a scanning rate of 4 • min −1 using a X-ray diffraction (XRD, Bruker, Berlin, Germany) D8-Advance, Cu Ka radiation source (λ = 0.154178 nm)).The Fourier transformed infrared spectra (FTIR) was investigated by a Tensor 27 spectrometer (Bruker, Berlin, Germany).The metal composition of the catalysts was conducted by employing inductively coupled plasma-atomic emission spectrometer (ICP-AES, Thermo iCAP6300, Washington, USA).The specific surface area was measured using Micromeritics ASAP2020 (Washington, USA) according to adsorption/desorption nitrogen isotherms.The electronic states of the surface of the as-obtained catalyst were measured by X-ray photoelectron spectroscopy (XPS, Thermo scientific Escalab 250Xi, Washington, USA).The morphologies and sizes of the catalysts were obtained using a transmission electron microscope (TEM, JEOL JEM 2100F, Freising, Germany) equipped with an energy dispersive X-ray detector (EDX, Freising, Germany) at a working voltage of 200 kV.The composition of the generated gas was evaluated through a Hiden QIC-20 quadruple mass spectrometer (Warrington, UK) using Ar as carrying gas.

Catalytic Activity Measurement
In a typical experiment, 0.1 g of the as-obtained PtNi/g-C 3 N 4 was dispersed into 2 mL water in a two-necked round-bottom flask.A solution containing 2 mL of AB (2 M) was injected through one neck with a syringe under stirring, and the other was connected with a gas burette to estimate the volume of gas.Upon the injection of AB aqueous solution, the amount of H 2 production was estimated via water-displacement approach by measuring the volume of drained water.The reaction time was recorded as the first bubble appeared.The reaction was over when there was no gas released.The reaction was carried out under designed temperature, controlling by cryogenic bath.The turnover frequency (TOF, mol H2 mol Pt −1 min −1 ) was calculated according to the linear relationship between volume and time in AB hydrolysis.
In order to explore the influence of catalysts concentration, AB concentration over the dehydrogenation of AB, AB hydrolysis for the Pt 0.5 Ni 0.5 /g-C 3 N 4 was evaluated in the same way as described above, except that the parameters, such as catalysts concentration, AB concentration, were changed, respectively.In addition, to acquire the activation energy (E a ) value of the hydrogen production of AB over Pt 0.5 Ni 0.5 /g-C 3 N 4 catalyst, the hydrolysis of AB was conducted at a series of temperatures including −5 • C, 0 • C, 5 • C, and 10 • C.

Stability Tests
The catalyst was recovered by centrifugation, washed with ethanol and water, and dried in an oven after the completion of the dehydrogenation reaction.Then the recovered catalyst was set into a two-necked round-bottom flask according to the above-mentioned dehydrogenation procedure for stability tests.A similar operation was repeated ten times.

Conclusions
In summary, we reported a simple and facile impregnation and chemical reduction approach to synthesize PtNi/g-C 3 N 4 nanoparticles as a catalyst for boosting the hydrogen generation of AB.Among the Pt x Ni 1-x /g-C 3 N 4 catalysts, the resultant Pt 0.5 Ni 0.5 /g-C 3 N 4 catalyst demonstrated outstanding performance, including 100% conversion, 100% H 2 selectivity, yielding the extraordinary initial total turnover frequency (TOF) of 250.8 mol H2 min −1 (mol Pt ) −1 for hydrogen production from AB at 10 • C, a relatively low activation energy of 38.09 kJ mol −1 , and a remarkable reusability (at least 10 times), which outperformed most of the noble metal heterogeneous catalysts.This notably improved activity was attributed to the charge interaction between PtNi NPs and g-C 3 N 4 support.Especially, the nitrogen-containing functional groups on g-C 3 N 4 , serving as the anchoring sites for PtNi NPs, might be beneficial for forming a uniform distribution and decreasing the particle size for the NPs.Moreover, the fabrication of g-C 3 N 4 -based catalysts might prompt the utilization of g-C 3 N 4 in energy fields.

Catalysts 2019, 9 , 15 Figure 5 .
Figure 5. (a) Time plots of catalytic dehydrogenation of AB (4 mmol) by PtNi/g-C3N4 catalyst with different ratios at 10 °C; and (b) time needed to complete the reaction and turnover frequency (TOF) values (obtained based on the overall Pt moles) of the PtNi/g-C3N4 catalysts.

Figure 5 .
Figure 5. (a) Time plots of catalytic dehydrogenation of AB (4 mmol) by PtNi/g-C 3 N 4 catalyst with different ratios at 10 • C; and (b) time needed to complete the reaction and turnover frequency (TOF) values (obtained based on the overall Pt moles) of the PtNi/g-C 3 N 4 catalysts.

Figure 6 .Figure 7 .
Figure 6.(a) Plots of moles of H2 per mole of AB versus time for the hydrolysis of AB (4 mmol) in the presence of Pt0.5Ni0.5/g-C3N4 at different catalyst concentrations at 10 °C, and (b) the logarithmic plot of hydrogen evolution rate versus PtNi concentrations.

Figure 6 .Figure 6 .Figure 7 .
Figure 6.(a) Plots of moles of H 2 per mole of AB versus time for the hydrolysis of AB (4 mmol) in the presence of Pt 0.5 Ni 0.5 /g-C 3 N 4 at different catalyst concentrations at 10 C, and (b) the logarithmic plot of hydrogen evolution rate versus PtNi concentrations.

Figure 7 .
Figure 7. (a) Plot of H 2 mole versus time for hydrogen generation from AB hydrolysis catalyzed by Pt 0.5 Ni 0.5 /g-C 3 N 4 catalyst at different AB concentrations ([Pt 0.5 Ni 0.5 /g-C 3 N 4 ] = 0.100 g, T = 10 • C), and (b) the logarithmic plot of hydrogen evolution rate versus AB concentrations.

Scheme 2 .
Scheme 2. Possible mechanism of hydrogen evolution from NH 3 BH 3 in aqueous solution over Pt 0.5 Ni 0.5 /g-C 3 N 4 .

Figure 9 .
Figure 9. (a) Durability test of Pt 0.5 Ni 0.5 /g-C 3 N 4 in ten runs for hydrogen generation from hydrolysis of aqueous AB solution, and (b) TEM image of Pt 0.5 Ni 0.5 /g-C 3 N 4 after ten cycles.

Table 1 .
Catalytic activities and the activation energy of catalysts in the hydrolysis of AB.

Table 1 .
Catalytic activities and the activation energy of catalysts in the hydrolysis of AB.