Cu_0.4Co_0.6MoO₄ Nanorods Supported on Graphitic Carbon Nitride as a Highly Active Catalyst for the Hydrolytic Dehydrogenation of Ammonia Borane

Abstract

As a typical chemical hydride, ammonia borane (AB) has received extensive attention because of its safety and high hydrogen storage capacity. The aim of this work was to develop a cost-efficient and highly reactive catalyst for hydrolyzing AB. Herein, we synthesized a series of Cu_xCo_1–xMoO₄ dispersed on graphitic carbon nitride (g-C₃N₄) to dehydrogenate AB. Among those Cu_xCo_1–xMoO₄/g-C₃N₄ catalysts, Cu_0.4Co_0.6MoO₄/g-C₃N₄ exhibited the highest site time yield (STY) value of 75.7 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ with a low activation energy of 14.46 kJ mol⁻¹. The STY value for Cu_0.4Co_0.6MoO₄/g-C₃N₄ was about 4.3 times as high as that for the unsupported Cu_0.4Co_0.6MoO₄, indicating that the g-C₃N₄ support plays a crucial role in improving the catalytic activity. Considering its low cost and high catalytic activity, our Cu_0.4Co_0.6MoO₄/g-C₃N₄ catalyst is a strong candidate for AB hydrolysis for hydrogen production in practical applications.

1. Introduction

With the rapid development of social economy and technology, the consumption of fossil fuels keeps growing, leading to the destruction of the environment and ecology [1,2]. Hydrogen energy, generating H₂O after reaction, can effectively solve the contradiction in the current energy structure between developing need and the environment [3]. To realize the vision of hydrogen power popularization, problems of hydrogen storage and acquisition must be overcome [4]. Chemical hydrogen storage, one of the most popular approaches, is the process by which hydrogen is trapped in a liquid or solid and can be extracted in due course. Ammonia borane (AB), with its high hydrogen density (19.6%), low molar molecular weight (30.7 g mol⁻¹), and high stability, is a high-profile hydrogen storage material [5,6,7,8]. When appropriate catalysts are used, the activation energy of the reaction is reduced, and the AB produces a large amount of hydrogen under mild conditions according to the following reaction [9,10]:

NH₃BH₃ + 2H₂O → NH₄BO₂ + 3H₂

In the literature, great efforts have been made to develop noble-metal-based catalysts [11]. However, their large-scale applications are remarkably limited by their high costs. Although the costs of bimetallic catalysts composed of both noble and nonnoble metals, such as Au–Ni, Ru–Ni, etc., are reduced in contrast to those of the noble metal catalysts, they are still high [12,13]. Thus, it is important for us to further reduce the cost of catalysts by developing noble-metal-free catalysts. On the other hand, the introduction of supports is one of the most efficient strategies for restraining the agglomeration of nanoparticles during a catalytic reaction. Due to its unique structure, inexpensiveness, and high specific surface area, carbon nitride (g-C₃N₄) has proved to be an ideal support in the field of heterogeneous catalysis, such as in photocatalysis, electrocatalysis, and the oxidation of toluene [14,15,16]. In recent years, there have been several successful examples of the application of g-C₃N₄ as a catalyst support in AB dehydrogenation. For example, Guo et al. designed g-C₃N₄-supported Au–Co nanoparticles with a site time yield (STY) value of 28.4 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ [17]. In contrast, the STY value of Au–Co nanoparticles without g-C₃N₄ support was only ca. 14 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_{\mathrm{2}}}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$. Fan et al. synthesized Rh nanoparticles with an average size of 3.1 nm which dispersed onto g-C₃N₄. The STY value for the resultant catalysts was 969 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ [18]. Hamza et al. reported that the STY value of AgPd/g-C₃N₄ was 94.1 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ [19]. Jia et al. synthesized Pb(0) nanoparticles anchored into g-C₃N₄ with chitosan, and their STY value was 27.7 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ at 30 °C [20]. Fan et al. synthesized AgCo bimetallic nanoparticles supported on g-C₃N₄, and the STY value of Ag_0.1Co_0.9/g-C₃N₄ was 249.02 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ [21]. It should be noted that the applications of g-C₃N₄ as a support in AB dehydrogenation are still rarely reported. In addition, the active components were noble metal or noble-metal-containing alloys in those reports. Therefore, it is important to explore some other types of active components supported on g-C₃N₄ and investigate their activity in AB dehydrogenation.

Herein, we report a series of Cu_xCo_1–xMoO₄ supported on g-C₃N₄ for ammonia borane hydrolysis. As far as we know, a noble-metal-free catalyst supported on g-C₃N₄ towards AB hydrolysis has not been reported previously. In this work, we found that the catalytic activity of the catalysts was closely related with the molar ratio of Co to Cu in Cu_xCo_1–xMoO₄/g-C₃N₄. It was discovered that Cu_0.4Co_0.6MoO₄/g-C₃N₄ shows the best catalytic activity as its STY value can reach 75.7 mol_H2 mol_cat⁻¹ min⁻¹, which is better than the properties of most non-precious metals reported in the literature. In addition, Cu_0.4Co_0.6MoO₄/g-C₃N₄ exhibited higher catalytic activity in contrast to the unsupported counterpart, Cu_0.4Co_0.6MoO₄.

2. Results and Discussion

Figure 1 shows the Powder X-ray diffraction (XRD) patterns of g-C₃N₄ before loading and Cu_0.4Co_0.6MoO₄/g-C₃N₄. Evidently, the peaks of g-C₃N₄, CuMoO₄, and CoMoO₄ can be found in Cu_0.4Co_0.6MoO₄/g-C₃N₄. The diffraction peak at 27.7° can be observed in both XRD patterns and is the characteristic peak of g-C₃N₄. The diffraction peaks of Cu_0.4Co_0.6MoO₄/g-C₃N₄ at 23.4°, 36.5°, 38.6°, 43.1°, and 52.9° correspond to the (110), (002), (200), ($\overline{1}\overline{1}$2), and (1$\overline{3}$0) planes of CuMoO₄ (PDF#26-0546). The peaks at 26.5°, 40.2°, and 46.3° correspond to the (002), (003), and (403) planes of CoMoO₄ (PDF#21-0868). No other peaks of impurities can be observed in the XRD pattern, indicating that the composites mainly contain C₃N₄, CoMoO₄, and CuMoO₄. For comparison, a series of Cu_xCo_1–xMoO₄/g-C₃N₄ were synthesized, and their XRD patterns are shown in Figure S1.

The Fourier transform infrared (FTIR) spectra of Cu_0.4Co_0.6MoO₄, Cu_0.4Co_0.6MoO₄/g-C₃N₄, and pure g-C₃N₄ are shown in Figure 2. For pure g-C₃N₄ powder, the main peaks at 815, 1240, 1316, 1400, 1458, 1573, and 1640 cm⁻¹ can be seen, due to the stretching vibration modes of C=N and C–N, which is similar to previous reports [20,22,23]. For Cu_0.4Co_0.6MoO₄/g-C₃N₄, the bands in the range of 1240–1640 cm⁻¹ correspond to the bonds of pure g-C₃N₄. In contrast, the bands at 937, 615, and 475 cm⁻¹ are assigned to the stretching vibrations of the t-MoO₄ group, Mo=O stretching, and Mo–O–Mo bending vibrations [24,25,26]. This result further confirms that Cu_0.4Co_0.6MoO₄ nanorods were successfully loaded onto g-C₃N₄. The FTIR spectra of different Cu_xCo_1–xMoO₄/g-C₃N₄ samples are shown in Figure S2 and are similar to that of Cu_0.4Co_0.6MoO₄/g-C₃N₄.

The field-emission scanning electron microscope (FESEM) images and Energy Dispersive Spectrometer (EDS) spectra of the catalysts are shown in Figure 3 and Figure S3, respectively. The morphology of g-C₃N₄ in Figure 3a is similar to that in a previous report [6]. The unsupported Cu_0.4Co_0.6MoO₄ is shown in Figure 3b,c, which shows aggregated nanorods with a diameter of ca. 500 nm. We note that most of the rods agglomerate together and form large aggregates, leading to a reduction in the contact area of the catalyst and substrate in the catalysis. To disperse Cu_0.4Co_0.6MoO₄ well and increase the number of exposed active sites, we introduced g-C₃N₄ into solution during the synthesis process and successfully obtained the Cu_0.4Co_0.6MoO₄/g-C₃N₄ material (Figure 3d,e). The rods could be easily found on the surface of g-C₃N₄. The diameter of the rods was still ca. 500 nm, and the length was ca. 5–10 μm. This observation indicates that the introduction of the g-C₃N₄ into the synthetic system did not change the morphology and size of the Cu_0.4Co_0.6MoO₄ nanorods. However, their dispersion was remarkably improved. EDS analysis was carried out on the nanorods, and the results are shown in Figure 3f. The elements C, N, O, Mo, Co, and Cu were detected. Evidently, the elements C and N come from the g-C₃N₄ material, while the elements O, Mo, Co, and Cu result from the Cu_0.4Co_0.6MoO₄. As displayed in Figure S3, the relative peak heights of Co to Cu decrease as x increases. The molar ratios of the different elements are shown in Table S1, indicating that the actual ratios of Co, Cu, and Mo in Cu_xCo_1–xMoO₄/g-C₃N₄ are close to the theoretical values.

The valence states of the elements in Cu_0.4Co_0.6MoO₄/g-C₃N₄ were measured by X-ray photoelectron spectrometer (XPS), and the results are shown in Figure 4. The peaks at 284.4 eV and 287.7 eV correspond to C 1s, assigned to carbon atoms on the surface and coordination between carbon atoms and nitrogen atoms, respectively [23]. The N 1s spectrum separated into two peaks at 398.4 eV and 400.1 eV, which were assigned to sp²-hybridized nitrogen in triazine rings (C–N=C) and tertiary nitrogen (N–(C)₃) groups [27]. The Co 2p spectrum shows two peaks at 780.9 eV and 796.1 eV of the Co 2p_3/2 and Co 2p_1/2 energy levels, demonstrating that Co is divalent [28]. At the same time, the peaks of Cu 2p at 931.8 eV and 951.7 eV imply the existence of Cu²⁺. Figure 4e is the XPS spectrum of Mo 3d. Two peaks at 232.2 eV and 235.3 eV imply that the element Mo is present as Mo⁶⁺ [29]. The peak at 531.3 eV in the O 1s spectrum illustrates the existence of O²⁻ [30]. Considering the XRD, SEM, and XPS results together, we can conclude that our obtained product is g-C₃N₄-supported Cu_0.4Co_0.6MoO₄.

The catalytic activity levels of g-C₃N₄, Cu_0.4Co_0.6MoO₄, and Cu_0.4Co_0.6MoO₄/g-C₃N₄ in the hydrolysis of ammonia borane were tested at 293 K. It can be seen in Figure 5 that g-C₃N₄ had no catalytic activity in the hydrolytic reaction. In addition, the catalytic activity of unsupported Cu_0.4Co_0.6MoO₄ was significantly lower than that of Cu_0.4Co_0.6MoO₄/g-C₃N₄. The STY values were 75.7 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ for Cu_0.4Co_0.6MoO₄/g-C₃N₄ and 17.6 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$ for the unsupported Cu_0.4Co_0.6MoO₄. As discussed above, the introduction of g-C₃N₄ into the catalysts can help to disperse active materials and thereby improve the catalytic property tremendously. Notably, the molar ratio of hydrogen to ammonia borane reached 3, demonstrating that complete hydrogen release can be achieved when Cu_0.4Co_0.6MoO₄/g-C₃N₄ acts as a catalyst. Figure 6 illustrates catalytic hydrolysis on the unsupported Cu_0.4Co_0.6MoO₄ and Cu_0.4Co_0.6MoO₄/g-C₃N₄ clearly. We also tried to synthesize ZnMoO₄ by a similar strategy to measure the catalysis activity of other molybdates. The result that there was no catalysis activity in ZnMoO₄ illustrates that molybdates have no reactivity. However, according to the literature, molybdates can serve as a Lewis acid and conduce the adsorption of OH⁻ on the catalyst surface, which would be of benefit to the hydrolysis reaction [31].

In consideration of variation of the molar ratio of cobalt to copper, we synthesized a series of Cu_xCo_1–xMoO₄/g-C₃N₄ (x = 0, 0.2, 0.4, 0.6, 0.8, 1) samples to test their catalytic performance, shown in Figure 7. No matter the value of x, Cu_xCo_1–xMoO₄/g-C₃N₄ samples were active in the hydrolysis of ammonia borane. CuMoO₄/g-C₃N₄, without Co element, showed the lowest catalytic activity with a STY value of 11.7 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$. When Co element was present, the activities increased remarkably, and the best molar ratio of Cu to Co was 4:6 with a STY value of 75.7 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$. It is worth noting that the STY of CoMoO₄/g-C₃N₄ was 35.35 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$, which is lower than those of Cu_0.2Co_0.8MoO₄/g-C₃N₄, Cu_0.4Co_0.6MoO₄/g-C₃N₄, and Cu_0.6Co_0.4MoO₄/g-C₃N₄, indicating that the coexistence of cobalt and copper in the sample is helpful to improving the catalytic activity. Notably, when CoMoO₄ served as a catalyst, there was no hydrogen released from the AB solution within the first minute. After then, hydrogen was constantly produced. This observation indicates that there is an induction period for the CoMoO₄ catalyst. In contrast, all the Cu-containing catalysts in this study had no induction period in AB hydrolysis, illustrating that copper element can shorten the induction period of catalysis. These conclusions are in a good agreement with those in previous reports [22,32]. We also synthesized CoMoO₄ and CuMoO₄, which had STY values of 25.5 and 3.9 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$, respectively; these STY values are lower than those of CoMoO₄/g-C₃N₄ and CuMoO₄/g-C₃N₄, further confirming that the introduction of g-C₃N₄ enhanced the catalytic activity.

To understand the relationship between catalyst dosage and hydrogen production rate, the hydrogen evolution curves for different amounts of Cu_0.4Co_0.6MoO₄/g-C₃N₄ are shown in Figure 8. As the amount of Cu_0.4Co_0.6MoO₄/g-C₃N₄ increased, the rate of hydrogen production also increased. By fitting the curve of ln(catalyst) vs. ln(rate), the relationship between the catalyst dosage and the hydrogen production rate was clarified. As shown in Figure 8b, the slope of the fitted curve was 0.901, indicating that the hydrolytic process is a pseudo-first-order reaction in the initial stage. The hydrogen production rates of the catalysts decreased gradually due to the diffusion rate limitation of AB and catalyst poisoning. This result is consistent with those in previous literature [5]. Therefore, hydrogen production can be controlled easily by adjusting the catalyst dosage in practice.

Figure 9a shows the relationship between the amount of ammonia borane and the volume of hydrogen production. It is interesting to note that the hydrogen release rates were almost the same at the initial stage of the hydrolytic reaction, demonstrating that the AB concentration has no pronounced effect on the initial reaction rate of hydrogen production. As shown in Figure 9b, the slope of ln(rate) vs. ln(NH₃BH₃) was 0.0222, which is close to zero, indicating that the catalytic reaction rate is independent of the AB concentration due to a zero-order reaction.

As we know, the temperature exerts a significant influence on catalytic activity of a catalyst. We also explored the relationship between the rate of hydrogen generation and catalytic temperature in the range 298 to 313 K (Figure 10). With increasing temperature, the slope of the hydrogen evolution curve became larger, and the time taken for complete hydrogen release became shorter. According to the Arrhenius formula, the apparent activation energy (E_a) can be calculated after fitting the curve of ln k vs. 1/T, where the k value is the slope of hydrogen generation vs. time. The E_a value of Cu_0.4Co_0.6MoO₄/g-C₃N₄ was as low as 14.46 kJ mol⁻¹. In general, E_a can be used to roughly assess the catalytic activity of a catalyst. The low activation energy indicates that its energy barrier is low in the catalytic process, hinting that the catalysts possess high catalytic activity.

Table 1. Comparison of the STY value and apparent activation energy of our Cu_0.4Co_0.6MoO₄/g-C₃N₄ catalyst with those of some other representative noble-metal-free catalysts in the literature.

Catalysts	STY (molH2 molcat−1 min−1)	STY without support (molH2 molcat−1 min−1)	Ea (kJ mol−1)	Reference
Ni/ZIF-8	85.7	/	28	[33]
Cu0.5Co0.5O-rGO	81.7	/	−45.26	[34]
Cu0.4Co0.6MoO4/g-C3N4	75.7	17.6	14.46	This work
CoP	72.2	/	/	[35]
Cu0.8Co0.2O-GO	70.0	/	45.53	[36]
Ni0.9Mo0.1/graphene	66.7	2.3	/	[37]
Co0.8Cu0.2MoO4	55	/	/	[32]
Cu0.49Co0.51/C	45	/	51.9	[38]
Cu0.8Ni0.2	41.9	/	40.53	[39]
CuCo/graphene	41	12	54.89	[40]
NiP	40.4	/	44.6	[5]
Ni91P9/rGO	13.3	/	34.7	[41]
Ni/g-C3N4	18.7	/	36	[6]

lists the STY values and activation energy data of some representative noble-metal-free catalysts in the literature and those of our Cu_0.4Co_0.6MoO₄/g-C₃N₄. As can be seen, the STY value of Cu_0.4Co_0.6MoO₄/g-C₃N₄ is higher than those of most noble-metal-free catalysts in the literature. Therefore, Cu_0.4Co_0.6MoO₄/g-C₃N₄ is one of the best potential catalysts for AB hydrolysis.

3. Materials and Methods

3.1. Synthesis of Catalysts

All chemical reagents were analytically pure and had not been further purified. Deionized water was used during the experiment. To synthesize g-C₃N₄, 25 g urea was calcined at 550 °C for 6 hours in air with a heating rate of 2.5 °C min⁻¹. A resultant yellow soft and porous powder was obtained. For the purpose of synthesizing 10 wt % Cu_xCo_1–xMoO₄/g-C₃N₄, 0.4 g g-C₃N₄ was dispersed in 40 mL ethanol and 20 mL deionized water. Then, 0.18 mmol copper chloride and cobalt chloride mixture solution was dropped into the g-C₃N₄ suspension under magnetic stirring for 2 hours. After that, 0.18 mmol sodium molybdate was added into the mixture solution. The molar ratios of Cu to Co to Mo were set at x:(1–x):1 (x = 0, 0.2, 0.4, 0.6, 0.8, 1). Then, 0.1 M ammonium hydroxide was used to adjust the pH of the suspension to 7. Finally, the suspension was transferred into a Teflon-lined stainless reaction vessel, which was placed in an oven and heated at 160 °C for 8 hours. The obtained samples were centrifugally separated, washed, dried, and finally calcined at 350 °C for 2 hours. For comparison, unsupported Cu_0.4Co_0.6MoO₄ was also prepared in a similar process to that described above, except that no g-C₃N₄ was used.

3.2. Characterizations

Powder X-ray diffraction (XRD) was measured to obtain crystallographic information using a Bruker D8 Discover X-ray diffractometer with Cu Kα radiation (λ = 1.5406Å) in the range of 2θ = 10°–80° (Bruker, Billerica, MA, USA). The Fourier transform infrared spectrum (FTIR) was measured to acquire functional group information using a Nicolet 6700 FTIR spectrometer in the range of 450–3000 cm⁻¹ (ThermoFisher Scientific, Waltham, MA, USA). Microstructures and the EDS spectrum were analyzed using a JEOL-7100F field-emission scanning electron microscope (FESEM) (Jeol Ltd., Akishima, Japan). The elements and oxidation states were analyzed using a Kratos Axis Ultra DLD X-ray photoelectron spectrometer (XPS) (Kratos Analytical, New York, NY, USA). The molar ratios of Cu, Co, and Mo in different samples were determined using an Agilent 7800 inductively coupled plasma mass spectrometer (ICP-MASS) (Agilent Technologies Inc., Santa Clara, CA, USA).

3.3. Catalytic Experiments

Due to the significant influence of temperature on catalytic performance, the catalyst properties were tested at 298 K. Typically, 5.0 mg of active substance, was dispersed in 5 mL deionized water. Then, 15 mL of mixture solution of 3 mmol ammonia borane and 20 mmol NaOH was quickly added to the Cu_xCo_1–xMoO₄/g-C₃N₄ solution. The volume of the generated hydrogen was measured by a drainage method, and the hydrogen production was recorded at intervals.

4. Conclusions

In summary, we synthesized a series of Cu_xCo_1–xMoO₄/g-C₃N₄ (x = 0, 0.2, 0.4, 0.6, 0.8, 1) and unsupported Cu_0.4Co_0.6MoO₄ for the hydrolysis of ammonia borane. As a representative catalyst, Cu_0.4Co_0.6MoO₄/g-C₃N₄ was characterized by XRD, FTIR, SEM, and XPS. The catalytic activity of Cu_0.4Co_0.6MoO₄/g-C₃N₄ was measured under different temperatures, different catalyst dosages, and different amounts of ammonia borane. It was found that Cu_0.4Co_0.6MoO₄/g-C₃N₄ showed the best catalytic activity with a STY value of 75.7 $\mathrm{m}\mathrm{o}{\mathrm{l}}_{{\mathrm{H}}_2}\mathrm{m}\mathrm{o}{\mathrm{l}}_{\mathrm{c}\mathrm{a}\mathrm{t}}{}^{-1}\mathrm{m}\mathrm{i}{\mathrm{n}}^{-1}$. The well-controlled composition and the well-dispersed active substances contributed to the improvement of the catalytic performance.