Introduction

Nowadays, global energy demand is increasing dramatically; nonetheless, the problems associated with scarcity of fossil energy and environmental pollution have been unable to meet the requirements of human development [1,2,3]. Recently, attention has been turned toward the development and utilization of new energy sources, such as wind, tidal, and nuclear energy, to sustain energy demand. Hydrogen energy, as a new energy carrier, is considered to be one of the effective substitutes for fossil fuels. Hydrogen has the advantages of rich reserves, high energy density, and zero emissions, thus making it the globally accepted cleanest energy carrier [4, 5]. However, difficult storage, requirement of efficient transportation methods of hydrogen, and controlled release of hydrogen limit the development and utilization of hydrogen energy. Compared to the currently used high-pressure and low-temperature liquid hydrogen storage technologies, hydrogen storage materials with higher hydrogen content can better meet the requirements of mobile applications due to their stability and light weight [6, 7]. Although porous materials (e.g., carbon materials, metal–organic frameworks (MOFs), and organic complexes) can physically adsorb hydrogen through weak van der Waals forces at extremely low operating temperatures to form hydrogen storage materials, they are not conducive to practical applications [8, 9]. In contrast, chemical hydrogen storage materials (hydrogen storage alloys, light metal hydrides, B–N hydrides, etc.) usually have a bulk hydrogen density that exceeds that of liquid hydrogen and high hydrogen release purity, thus making them very promising as hydrogen source for fuel cells [5, 10].

Ammonia borane (NH3BH3, AB) is the simplest type of B–N hydride, with a molecular weight of only 30.7 g mol−1 and a hydrogen storage capacity of up to 19.6 wt.%. Moreover, AB is non-toxic, stable, and environmentally friendly, making it a highly promising candidate for hydrogen storage [11]. At present, AB mainly releases hydrogen in the following two ways: pyrolysis and hydrolysis. Pyrolysis involves the opening of the B–H and N–H bonds in the AB molecule through high temperature, and then the H atoms recombine with each other to release hydrogen. However, this process requires higher temperature (to completely release three equivalents of hydrogen, temperature of above 1200 °C is required), consumes large energy, and generates gaseous by-products which reduce hydrogen purity [12]. Therefore, milder hydrolysis of AB is a more promising way to produce hydrogen. In the hydrolysis reaction, the proton H in water reacts with the ionic H connected to the B atom in AB to release three equivalents of hydrogen, and the hydrogen production is 8.9 wt.% of the raw materials AB and H2O [13]. The reaction equation is represented as follows:

$${\text{NH}}_{3} {\text{BH}}_{{3}} \, + \,2{\text{H}}_{2} {\text{O}}\, \to \,{\text{NH}}_{4}^{ + } \, + \,{\text{BO}}_{2}^{ - } \, + \,3{\text{H}}_{2}$$

AB can exist stably in water; therefore, the hydrolysis reaction requires the involvement of a suitable catalyst to be carried out under ambient conditions.

Noble metal-based catalysts (Pt, Rh, and Ru) were found to have excellent catalytic activity for the hydrolysis of AB [14,15,16,17]. However, due to the high cost and limited resources of noble metals, their practical application is greatly limited. Researchers found that non-noble metals such as Fe, Co, and Ni also show catalytic activity toward hydrolysis of AB [18,19,20]. Compared to noble metals, the non-noble metals offer reduced cost and higher earth abundance. Besides, non-noble metals are abundant in the types of compounds that can serve as potential catalysts. Although many studies proved that using non-noble metals as dopants or for formation of alloys or heterostructures with noble metals can improve the catalytic activity of the catalyst; nonetheless, the cost of catalysts still remains high [21,22,23,24,25,26,27,28]. The non-noble metal has a much lower catalytic activity than the noble metal, and its problems of easy agglomeration and oxidation are also harmful for efficient catalytic activity. Therefore, improving the activity and durability of the non-noble metal-based catalyst is the focus of current research. So far, significant progress has been made in this field.

Status and development of hydrolysis of AB catalyzed with noble metals have been summarized [29]. At the same time, the research on non-noble metal-based catalysts for catalytic hydrolysis of AB has been developing rapidly, and many types of catalysts have emerged. Moreover, the catalytic activity of non-noble metal-based catalysts has been catching up with that of noble metal-based catalysts. Therefore, it is important to summarize the current research status of non-noble metal-based catalysts, which can guide future researchers. In this review, we intend to summarize the research progress of non-noble metal-based catalysts for the hydrolysis of AB to hydrogen, in the past five years. Research on non-noble metals has focused on the following three types of materials: pure non-noble metal-based materials, non-noble metal-based compounds, and their composite catalysts. This review mainly includes a summary of the catalyst composition, structure, and activity improvement. Furthermore, the advantages and challenges of non-noble metal-based catalysts for the hydrolysis of AB to produce hydrogen are discussed to cope with their future development.

Non-noble metal-based catalyst

Metal

Single metal

Initially, in 2006, inspired by the catalytic activity of noble metals, Xu et al. found that Co and Mo could also catalyze the hydrolysis of AB, and additionally, supported Co, Ni, and Cu also exhibited excellent catalytic activity [30, 31]. Subsequently, more designs were carried out for pure metal-based catalysts, microstructure design, ligand stabilization, and addition of support to obtain more active sites of the metal to enhance the catalytic activity. In recent years, in addition to the use of NaBH4 or LiBH4 to reduce metal salts, single metal component catalysts could also be obtained by other methods. For instance, Nozaki et al. used amorphous alloys to prepare skeletal Ni and Cu, respectively [32, 33]. The higher specific surface area (SSA) of the skeletal structure catalyst allowed the reactants to be in more contact. The activity of the catalyst thus improved, and reduction in catalytic activity was not observed even after three cycles of use. Fang et al. encapsulated ultra-small Co nanoclusters in soluble porous coordination cages (Co NCs@PCC), wherein the particles size of Co NCs was only 2.5 nm and PCC could very well stabilize and disperse Co NCs [34]. This design could obtain small-sized metals and solve the problem of the ligand covering the active sites. Therefore, the turnover frequency (TOF) value of catalytic hydrolysis of Co NCs@PCC was 90.1 min−1, which could reach the level of noble metals.

Currently, various support materials are used to fabricate well-dispersed metal nanoparticles (NPs), which can control size and dispersion of these NPs. The strong metal–support interactions also make an important reason for the increased activity. A certain degree of electron transfer occurs between the metal and the support; therefore, the active metal components have a more optimized electronic structure. This metal obtains a more optimal adsorption capacity and aggressiveness to the reactants and intermediates, thereby enhancing the activity of the catalyst. Carbon supports (activated carbon, graphene, and carbon nanotubes (CNTs)) are most concerned catalyst candidates [35,36,37,38,39]. Zhang et al. used atomic layer deposition to uniformly disperse Ni on CNTs, and Ni/CNTs showed excellent hydrolysis activity [39]. Zhang et al. used porous carbon prepared by direct carbonization of Co/Zn-MOFs to support Co NPs (Co-HPC), wherein the Co particle size was only 12.1 ± 4.8 nm [40]. Carbon supports prepared with MOFs exhibited a higher SSA, thus making them more conducive to dispersion of metals, and could more effectively contact with the reactants during catalysis. Besides, the initial activity of Co-HPC was still largely maintained after the catalyst was reused 12 times, indicating its good durability. Zhang et al. prepared a highly nitrogen-doped and uniformly embedded Co-based mesoporous carbon (Co@NMC) catalyst with a high SSA (1044 m2  g−1) [41]. Extended X-ray absorption fine structure (EXAFS) and X-ray photoelectron spectroscopy (XPS) were used to characterize the surface compositions and chemical states of Co@NMC, indicating the decrease in the valence electrons of the Co, and the presence of the strong Co-Nx effect and the Mott–Schottky effect. The Co NPs were stably fixed on the support, showing excellent catalytic activity for hydrolysis. For the modification of the support, Li’s group reported a three-dimensional (3D) structure composed of polyethyleneimine (PEI) and graphene oxide (GO) as support of Co NPs [42]. It was found that the spatial distribution and surface-active sites of Co could be controlled. Owing to the stability of the 3D structure and the synergy between the amine group and Co NPs, the Co/PEI-GO exhibited high AB hydrolysis activity and cycle durability.

The high SSA and adjustable pore size of the MOFs can make the NPs have better dispersibility and more number of exposed active sites, thus making them efficient supports [43, 44]. For example, Wang et al. used ZIF-8 to highly disperse Ni NPs with a particle size of only 2.7 nm, which showed excellent catalytic activity (TOF = 85.7 min−1). Besides, according to the kinetic isotope effects (KIEs) design experiment, it was proved that the rate-determining step (RDS) of the hydrolysis reaction was the cleavage of the O–H bond in H2O [19]. Furthermore, Liu et al. used MIL-101 as a support to study the catalytic activity of Co NPs in amorphous and crystalline phases with different sizes for AB hydrolysis (Fig. 1). The results showed that the amorphous Co/MIL-101 catalyst with small size (1.6–2.6 nm) showed the best performance, the TOF value could reach 51.4 min−1, and the activity did not decrease significantly even after 30 cycles [45].

Figure 1
figure 1

Schematic illustration for the synthesis of four types of MIL-101-supported Co NPs (a) and plots of time versus volume of generated H2 from AB aqueous solution over catalyst (b): (a) Co/MIL-101-1-U; (b) Co/MIL-101-1; (c) Co/MIL-101-2-U; (d) Co/MIL-101-2. Inset: corresponding TOF values of the catalysts. Long-time durability characterization of Co/MIL-101–1-U for the generation of H2 from AB aqueous solution (c). Reproduced with permission from ref [45]. Copyright 2017 American Chemical Society

Full size image

Semiconductor materials with unique light response characteristics are also used as supports. On the one hand, semiconductor materials can interact with metals by electron transport [46]; on the other hand, these materials can use light energy, and the excited state of photogenerated electrons and holes (e−/h+) on the reactants is more conducive to the hydrolysis reaction [47,48,49]. The post-modified C3N4-supported Co NPs exhibited excellent catalytic activity toward hydrolysis of AB under visible light irradiation [50]. The modified C3N4 showed an increased SSA and extended light absorption. The enhanced e−/h+ separation of C3N4-supported Co NPs efficiency ensures an electron transfer process under light irradiation, leads to an increase in the electron density of Co NPs, and thus improves the catalytic efficiency. Song et al. synthesized a series of porous two-dimensional V2O5 nanosheets with oxygen vacancies loaded with Co NPs and performed catalytic hydrolysis under visible light [51]. The band structure of porous V2O5 nanosheets with specific oxygen vacancies became controllable, leading to a higher separation of e−/h+ than the original V2O5. The separated h+ easily reacted with H2O molecules to generate hydroxyl radicals (·OH). Besides, the electron density of Co NPs in Co/V2O5 increased, indicating the interaction between the metal and support. ·OH and the electrons enriched in Co could favorably attack AB to produce hydrogen. Furthermore, this group also prepared C3N4 nanosheet-supported Co and Ni NPs for catalytic hydrogen generation, which showed excellent catalytic activity, with the TOF value of Co-based catalysts reaching 123.2 min−1 under visible light irradiation [52]. Besides, BN, SiO2, and Al2O3 have also been used to support metal NPs, and their catalytic activity was found to be excellent. Although some supports cannot optimize the electronic structure of the metal, the stable structure of the support can fix the metal to make it more dispersed and stable. Table 1 summarizes the literature results of metal-based catalysts for AB hydrolysis.

Table 1 Summary of the catalytic performance of different pure non-noble metal-based materials for the hydrolysis of AB
Full size table

Bi- and multi-metallic

However, the catalytic activity of a single metal is finite. For the first time, in 2009, Xu et al. in situ synthesized FeNi alloy for AB hydrolysis. The prepared nanoalloys possessed Pt-like high catalytic activity [53]. Later studies have shown that the bi- or multi-metallic composite catalyst through component regulation allows the regulation of the electronic structure of catalyst due to the synergistic effect among the different metals, thus providing the optimized activity of the catalyst [54,55,56]. Zhang et al. prepared Cu@Ni cubic-cage microstructure formed by coating a fine Cu cube with Ni nanospheres. The XPS characterization confirmed that the electrons were transferred from Cu to Ni, and the increase in Cu content facilitated the transfer of more charges to Ni [57]. The synergistic effect between the metal components resulted in an increase in the number of active sites on the catalyst surface. Our previous study explored the relationship between the electronic structure and the catalytic activity of different Co-based alloys toward hydrolysis of AB [58]. The result indicated that the hydrogen production rate of the alloy exhibited a volcanic trend with the center of the d-band, which could affect the absorption of water and AB molecules when the electronic structure was appropriate to obtain excellent activity.

Although synergistic interaction occurs between metals to regulate the electronic structure, some metal elements also as dopants can act as atomic diffusion layers and hinder grain growth. For instance, Wang et al. reported that Cu@MoCo catalysts exhibited significantly higher activity than CuCo catalysts [59]. On the one hand, the addition of Mo could reduce the size of NPs and lead to an increase in active sites; on the other hand, Mo acted as an electron donor, and the synergy between the metal components effectively optimized the electronic structure of the catalyst, thereby reducing the potential barrier of the hydrolysis reaction. In order to overcome the problem of easy agglomeration of metal particles, Sang et al. used dendritic CoCu and NiCu catalysts synthesized with starch, and their catalytic activity was much higher than that of the catalyst without starch [60]. The oleylamine-dispersed CuNi NPs catalyst also showed the effect of enhancing activity [61]. These special dispersants can stabilize metal particles and expose more active sites, which improves metal utilization.

Similarly, the support is advantageous for the dispersion of bi- and multi-metallic catalysts. Bimetallic CuCo alloy NPs supported on activated carbon (CoCu/C) were reported by Bulut et al. [62]. The size of alloy NPs was only 1.8 nm, which resulted in significant increase in its active site. The catalyst showed TOF value of up to 28.67 min−1 and good recycling performance (> 95% activity after 10 cycles). Our group also successively studied CuNi alloy NPs supported on TiO2-based materials, including TiO2(B) NTs, TiO2(B)/anatase nanofibers (NFs), and TiO2-CdS [47, 48, 63]. The stabilization of the support made the metal NPs smaller in size and resulted in their better dispersion. The synergy between Cu and Ni makes the catalyst much more efficient than single metal. Besides, the improvement of the support through these studies makes the catalyst’s absorption of light to expand the visible light region, and the utilization efficiency becomes higher. All these effects endow the catalyst with higher hydrolysis activity. This series of work provides better ideas for the conversion of light energy to hydrogen energy. Yang et al. studied CuCo/BN NFs to catalyze the activity of AB. A series of first-principle calculations of the catalyst showed that when the CuCo/BN catalyst adsorbs AB molecules (Fig. 2), the interaction between B and N atoms increases, while the bond length is shortened, indicating that adsorption can inhibit the B–N bond from breaking. Moreover, the B–H bond becomes longer, which is conducive to the release of hydrogen [64]. This also shows the beneficial effect of the synergy between the bimetals toward the hydrolysis reaction. Additional studies on bi- and multi-metallic catalytic hydrolysis are listed in Table 1 for activity comparison. Most bi- and multi- metallic catalysts exhibit significantly improved catalytic activity compared to single metals and to a certain extent reduces the use of specific metal components, which is more conducive to long-term development.

Figure 2
figure 2

The geometrically optimized most stable adsorption structures for AB on Cu/h-BN (a), Co/h-BN (b), and CuCo/h-BN (c), the bond lengths (Å) of adsorbed AB and pristine AB are also indicated. Reproduced with permission from ref [64]. Copyright 2019 Elsevier

Full size image

Thus, the above-mentioned description indicates that the current enhancement strategies for non-noble metal-based catalysts are mainly divided into the following four aspects (Fig. 3): multi-component, micromorphology control, caped ligands, and added support. This can provide a certain reference for the design of the non-noble catalysts in the future.

Figure 3
figure 3

Strategies for enhancing the activity of non-noble metal catalysts and corresponding examples

Full size image

Metal-based compounds

Metal borides

Among metal borides, cobalt boride (Co–B) was first widely studied in 1953 as a catalyst for the hydrogen production reaction via NaBH4 hydrolysis [95]. Co–B has also been used in subsequent studies to catalyze the hydrolysis of AB. Owing to its high catalytic activity, its preparation and catalytic performance have been extensively studied and reported. Wang et al. synthesized a cotton-like CoB alloy (CoBssR) catalyst through a novel room-temperature solid-state reaction. CoBssR acquired a large SSA (222.4 m2 g−1) and a rich porous structure, making its TOF value twice that of conventional CoB alloys with good durability [96]. CoB films have been investigated for catalytic hydrolysis of AB as they prevent aggregation to a certain extent and have better stability compared to powders [97]. Use of binary transition metal boride is also an important strategy to improve catalytic activity [98].

The substrate stabilizes the catalyst, and the doped elements improve the electronic structure so that the catalytic activity gets enhanced. Based on the research on Co–B thin films, Wang’s group synthesized improved Co–B thin-film catalysts doped with Mo or W on different substrates (Ni foam, foam sponge, and Cu foil) [99,100,101,102]. Men et al. doped P element into CoB to synthesize CoBP NPs supported on 3D nitrogen-doped graphene hydrogel (CoBP/NGH). Owing to the electron transfer among Co, B, and P and the metal–support interaction, the electrons on Co are enriched, which is beneficial to attacking AB molecules. The hydrolysis of AB obtained by CoBP/NGH showed excellent catalytic performance (TOF = 32.8 min−1), which was better than that of most Co–B-based catalysts reported [103]. However, compared to other types of catalysts (Table 2), the activity of metal borides used for the hydrolysis of AB was still at a relatively low level.

Table 2 Summary of the catalytic performance of different non-noble metal boride/phosphide/oxide-based materials for the hydrolysis of AB
Full size table

Metal phosphides

Transition metal phosphides (such as Ni2P and CoP) exhibit the catalytic behavior similar to that of noble metals due to their similar d electron state density. Moreover, metal phosphides are inexpensive; thus, they have been extensively studied in the hydrolysis of AB [104]. In terms of microstructure regulation, CoP and Ni2P nanosheet arrays prepared by Tang et al. were studied to have the advantages of easy separation and good stability [105, 106].

Many researchers have found that the appropriate concentration of NaOH significantly enhances the activity of the catalyst; nonetheless, the effect of OH− on the catalyst has not been clearly studied [19, 107]. Fu et al. studied the effect of CoP NPs on the induction period of AB hydrolysis in the presence of anions (OH−, F−, and Cl−). They found that OH− was the most obvious to reduce the induction period of hydrolysis reaction [108]. Characterization of the electronic structure by XPS indicated that in the absence of anions, CoP exhibited the electrons transfer from Co to P, forming electron-rich Pδ− and electron-deficient Coδ+, which is consistent with the results provided by other researchers. The OH− coordinated with the catalyst, the valence electrons of metallic Co increased, and different valence electron states changed, indicating that the introduction of OH− on the surface improved the electronic properties of the catalyst. Density functional theory (DFT) calculation of OH− adsorption on the surface of CoP further proved that the interaction between the metal center and the anion leads to the enrichment of the charge in the CoP, thus enhancing the catalytic activity. This study explains the enhanced catalytic reaction of OH− and metal phosphides mainly due to the fact that the increase of valence electrons of metal ions is supported by experimental and theoretical calculations.

Similar to previously mentioned metals and metal borides, use of binary metal phosphides also constitutes an effective strategy to enhance catalytic activity. Incorporation of the second metal into the phosphide can optimize the electronic structure and enhance the interaction between the catalyst and the reactants. Hou et al. changed the ratio of Ni/Co in Ni2−xCoxP catalysts and studied the changes in electronic structures (Figs. 4a–c) [109]. Characterization by XPS and X-ray absorption spectroscopy (XAS) revealed that with the incorporation of Co, the electrons were transferred from the metal to the P center, and change was observed in the coordination structure of the Ni and Co. The phase-shift-free Fourier transform (FT) also showed that Co was successfully doped into the lattice of Ni2P and partially occupied the initial position of Ni atoms, forming Co–P bonds. The DFT calculation of the electronic structure of Ni2−xCoxP confirmed that the electron transfer from the metal center to the P center was conducive to the strong adsorption of AB molecules. Therefore, the catalyst could attack the B–N bond more effectively and reduce the reaction barrier. These results also indicate that the incorporation of the second metal element has an enhanced effect on the electronic structure and catalytic activity of the material. NiCoP/GO nanohybrid material was also used in AB hydrolysis catalysis with TOF values of up to 153.9 min−1, even higher than those of noble metal catalysts.

Figure 4
figure 4

The total TOF values (a) and the corresponding FT curves of a series of catalyst (b). Energy profiles of AB hydrolytic dehydrogenation on Ni3P2(0001) and NiCo2P2(0001) surfaces with the same initial hydroxyl concentration (c). Reproduced with permission from ref [109]. Copyright 2017 the Royal Society of Chemistry. Band structure of phosphides and TiO2 (d), and AB hydrogen evolution rate of NiCoP/TiO2 under different light (e). Reproduced with permission from ref [110]. Copyright 2020 Elsevier

Full size image

TiO2 is a good electron-accepting metal oxide. Wang et al. used NiCoP as a sensitizer to design a solar sensitization system, which constituted the state of electron imbalance on the surface of NiCoP/TiO2 and produced hydrogen by photocatalytic hydrolysis of AB [110]. NiCoP showed wide light absorption over the entire ultraviolet–visible (UV–Vis) range, ensuring full use of visible light by the catalyst. Furthermore, the band position of the catalyst was roughly estimated by UV–Vis diffuse absorption spectra (Figs. 4d and e). Unlike traditional photocatalytic water splitting, photogenerated electrons during the reaction do not directly participate in the redox reaction, but rather induce a highly imbalanced charge on NiCoP. The difference in surface charge can significantly promote the activation of B–H and O–H bonds, leading to an increase in the rate of hydrogen release. This study provides new insights and evidence for the mechanism of photocatalytic hydrolysis of AB and indicates that excellent hydrolysis activity is more conducive to the design of subsequent catalysts. Furthermore, other metal phosphides show excellent catalytic activity in the hydrolysis reaction, which are listed in Table 2.

Metal oxides

It is well known that the problem of easy oxidation of transition metals seriously reduces their activity and stability. Although the use of boride and phosphide solves this problem to some extent, the oxidation of transition metals inevitably occurs in these catalysts, which causes a slight decrease in catalytic activity. In 2008, Kalidindi et al. carried out Cu-catalyzed AB hydrolysis and found that Cu2O was catalytically active, and it was proposed that the reduction of Cu2O–Cu(0) was due to the release of H2 facilitated by the attack of water on a transient Cu–H [111]. Some specific non-noble metal oxides, such as Co, Cu, CuCo, NiCo, and CuFe oxides, have been reported to have good catalytic activity [112,113,114,115,116,117,118]; thus, more research is conducted in this regard.

The microstructure has an important influence on the catalytic performance. For example, Li et al. studied NiCuO with different micromorphologies to catalyze AB and found that the bracelet-like Ni0.4Cu0.6O nanoplate exhibited the best catalytic activity, with a TOF value of 33.43 min−1. CuO and NiO showed a good synergistic effect in the hydrolysis process [119]. The support was introduced into the catalytic system, and the interaction between the oxide and the support provided powerful help for the catalysis. Feng et al. prepared the hybrids of CuxCo1−xO NPs on GO. The optimized hybrid (CuxCo1−xO/GO) showed superior catalytic performance in the hydrolysis of AB [120]. XAS characterization revealed that the interfacial interaction between CuO-GO was weaker than that in CoO-GO; however, CoO-GO exhibited no catalytic activity during catalysis; thus, it was shown that the addition of Co element could enhance the bonding ability, leading to the enhancement in the interaction between the active component and the support. A new hybrid electronic state was formed in this interaction, which could help improve the catalytic process, thereby enhancing the performance of CuxCo1xO/GO in AB hydrolysis. This group also studied the cubic CuCoO nanostructures supported on the reduced GO (rGO) based on the study on CuxCo1−xO/GO [121]. The Cu in the catalyst can activate H2O and then cooperate with Co to anchor the AB molecule. The interaction between the metal and support results in a high activity of Cu0.5Co0.5O/rGO. Furthermore, the conversion of the oxidized state to the metallic state in the catalyst after recycling was studied. The cubic structure of the oxide was broken and the catalyst activity was not restored, a phenomenon that further illustrates the peculiarities of metal oxides. Besides, Feng’s group also expanded this research to study Ni0.5Co0.5O NPs hybrid material on nitric acid-treated carbon nitride for the hydrolysis of AB [122].

Inspired by the high performance of oxides in supercapacitors [123], Li and his group developed a series of MCo2O4 (M=Cu, Mn, Ni) catalysts for AB hydrolysis [124,125,126,127]. Li et al. compared the catalytic activity of CuCo2O4 catalyst with three different micromorphologies: nanoplates, nanosheets, and unsupported NPs. Among them, CuCo2O4 nanoplates/Ti showed the best catalytic activity compared to the other two catalytic systems (Fig. 5) [125]. On the one hand, the presence of the Ti substrate can prevent the loss of activity caused by the aggregation of the nanofilm; thus, the catalyst can still maintain 95% of the original activity after eight cycles. On the other hand, the special morphology is expected to have more edges and corners in contrast to the spherical NPs, which can provide more active sites for the catalytic reaction. Besides, the oxidized Co and Cu ions were reduced to metallic Co and Cu after the catalytic reaction. The synergistic effect between Cu and Co could effectively improve the catalytic performance toward hydrolytic dehydrogenation of AB, leading to the TOF value of up to 44 min−1. Similarly, in the study of the catalytic performance of CuCo2O4 nanoplates, Li’s group also proposed that the interaction between the reduced metal state and the support CuCo2O4 played a positive role in improving the catalytic activity [124]. This group also conducted extensive research on the introduction of the second transition metal such as Cu, Ni, or Co in the MCo2O4 catalyst [128,129,130]. The synergistic effect of the metal elements in the oxide was found to be beneficial to enhancing the catalytic activity, which is generally superior to that of the single metal component MCo2O4, and their TOF values are listed in Table 2. Besides, Mo6+ has Lewis acid characteristic, which is beneficial to the adsorption of ·OH; thus, it can enhance the hydrolysis process. Furthermore, extensive research efforts have also been devoted to the study on the incorporation of Mo into the catalyst, and these metal oxide catalysts exhibited superior hydrolytic activity [131,132,133,134,135]. Many of these studies on oxide believe that the reduced metal is the hydrolytic active component, and the remaining oxide serves as the support. Although the metal component has high activity, the interaction between the metal and support is an excellent activity to provide assurance. However, the studies still lack the verification of this interaction, which undeniably requires further research efforts.

Figure 5
figure 5

SEM image of CuCo2O4(nanoplates)/Ti (a). The hydrogen evolution when different CuCo2O4 catalysts were used (b). XPS spectra of CuCo2O4(nanoplates)/Ti: Cu (2p3/2) for the fresh and used catalyst (c–d), Co (2p3/2) for the fresh and used catalyst (e–f). Reproduced with permission from ref [125]. Copyright 2017 Elsevier

Full size image

MoO3 has the characteristic of localized surface plasmon resonances (LSPRS); thus, it was used by researchers as a catalyst for AB hydrolysis [136,137,138]. Yin et al. synthesized MoO3−x structure with a relatively high SSA (30 m2 g−1) and strong LSPRS in the visible light region. Under visible light irradiation, the catalytic activity of the MoO3−x was much higher than that of commercial MoO3 for hydrolysis of AB [137]. In addition to the strong LSPRS, MoO3−x forms a new band at the bottom of the original conduction band due to the presence of Mo5+ center and oxygen vacancies; thereby, the recombination of e− and h+ gets reduced. During the hydrolysis reaction, e− accumulated on the surface of the catalyst reacts with dissolved O2 to generate O2−, and h+ reacts with absorbed H2O to generate ·OH. Subsequently, these groups and H2O attack MoO3−x-AB complex species to dissociate the B–N bond in AB to form BH3, thereby leading to the enhancement in the hydrolysis process (Fig. 6). Based on this research, Yin’s group incorporated WO3 into the MoO3−x structure to obtain a stronger LSPRS and optimized its catalytic activity [136]. However, in the study on the photocatalytic hydrolysis of AB, researchers have different views on the mechanism of e−/h+ on the catalytic reaction; thus, this topic is worth exploring in future research.

Figure 6
figure 6

Proposed reaction mechanism of hydrolysis of aqueous AB solution over plasmonic MoO3-x under visible light. Reproduced with permission from ref [137]. Copyright 2017 the Royal Society of Chemistry

Full size image

Composite materials of metals and metal-based compounds

Both metals and some specific metal-based compounds have the activity of catalyzing AB. Many studies have found that the interaction between the two active components is very beneficial to the improvement of catalytic activity. For example, Lin et al. designed a carbon-supported Ni/Ni2P hybrid system [152]. A strong interaction was observed between NPs and carbon in the hybrid system, and electron transfer occurred from Ni to P, which was characterized by XAFS and XPS. P not only caused the surface disorder of NPs, but also changed the chemical environment of Ni in C-Ni/Ni2P. DFT calculations further indicated that Ni/Ni2P hybrid system showed higher adsorption energy and lower dissociation energy for H2O molecules than metal Ni. Besides, carbon-supported Ni NPs could facilitate desorption of ·H in the catalytic H2 evolution process. The different AB hydrolysis performance obtained by adjusting the hybrid structure of Ni/Ni2P ratio was more illustrative of the synergistic effect among C, Ni, and Ni/Ni2P, which promoted the hydrolysis process at the same time.

Li et al. designed a hybrid nanocatalyst composed of metal Cu, CuCoOx, and GO support (Cu@CuCoOx@GO) to promote the room temperature hydrolysis of AB [153]. The CuCoOx in the catalyst promoted the adsorption of ·OH. At the same time, surface of the Cu metal showed an electron-rich region, which was kinetically favorable to the decomposition of H–OH to ·H. The conductive GO not only dispersed Cu@CuCoOx to expose more catalytically active sites, but also contributed to electron and substance transfer during the hydrolysis process. These advantages in Cu@CuCoOx@GO synergistically improved the room temperature catalytic performance toward hydrolysis of AB. Zhang et al. prepared Co–N-doped carbon spherical particles with MOFs as the precursor and then controlled the oxidation to synthesize Co-CoOx@NCS catalysts of the “birdcage” model (Fig. 7) [154]. Co–CoOx NPs are surrounded by a porous carbon shell. During the hydrolysis reaction, AB molecules interact with the core NPs. The Co–Co bond activates the B–H bond in AB, and the Co–O bond activates the H–O bond in H2O; thus, two activated H atoms produce a H2 molecule. Novel active sites for higher activity in catalytic reactions can be provided by synergistic interactions between metallic Co and CoOx, and recombination of electronic orbitals between various oxidation-state Co species. Noteworthy, carbon shell limits the growth, aggregation, and loss of active components, while the diffusion of reactants and products can still be conducted through it. The coordination of the entire system makes the catalyst acquire higher reactivity. Besides, Cu@Cu2O [155], Ni-CeOx/graphene [156], and Co-CeOx/NGH [157] have also been studied as AB hydrolysis catalysts (Table 3).

Figure 7
figure 7

Structural evolution of Co@NCS and Co-CoOx@NCS (a). XRD patterns of various catalysts (b). Hydrogen generation catalyzed by various catalysts (c). Catalytic mechanism in the reaction system for hydrogen generation of AB (d). Reproduced with permission from ref [154]. Copyright 2019 American Chemical Society

Full size image
Table 3 Summary of the catalytic performance of different non-noble composite materials of metals and metal-based compounds for the hydrolysis of AB
Full size table

Advantages and challenges

The hydrogen production process of hydrolyzing AB significantly reduces the energy consumption compared to pyrolysis and comparatively can be carried out under extremely milder conditions. Researches on hydrolysis catalysts in recent years have shown that non-noble metal-based catalysts can obtain excellent catalytic activity at a lower cost. Non-noble metal catalysts that can reach the activity level of noble metals lay the foundation for the sustainable development of hydrogen energy. These advantages are shown in Fig. 8, although there are still several problems related to AB hydrolysis.

Figure 8
figure 8

Advantages and challenges in the field of hydrogen production by hydrolysis of ammonia borane

Full size image

Recently, the mechanism of the catalyst catalyzing the hydrolysis of AB has been proposed. (1) Similar to the mechanism of NaBH4 hydrolysis, the M–H intermediate is proposed as an important step in the AB hydrolysis reaction [16, 111]. AB molecules are adsorbed on the surface of the catalyst, the H of B–H bond and the metal form an activated intermediate M–H, and then H2O molecules attack the intermediate. The ·H produced by M–H combines with the ·H produced by H2O molecules to generate H2. Therefore, the formation of the M–H intermediate has a great influence on the rate of hydrolysis. (2) Attack of H2O molecules on the B–N bond in the AB molecule is an important step [120, 158, 159]. First, the reactant molecules are adsorbed on the catalyst surface. When they are in proximity, the O and H atoms in the H2O molecule are close to the B and N atoms in AB, respectively, thus attacking the B–N bond to make it grow and break to form the transition state B…O and N…H2. The second H2O molecule continues to be adsorbed on the surface of the catalyst and attacks -BH3. The ·H on the BH3 combines with the ·H in the H2O molecule to form H2 and escapes. (3) The breaking of the O–H bond in the H2O molecule is the RDS of the hydrolysis reaction [19, 38, 109]. In terms of experiments, the researchers used the KIE to prove that the bond which breaks in the hydrolysis reaction corresponds to the RDS of the reaction. The ·OH generated after the breaking of the O–H bond in the H2O molecule can attack the BH3 group to form a BH3OH· intermediate, and then ·H is generated, which combines with the ·H generated from the H2O to eventually form H2. Most recent researches can only detect and prove a certain step or intermediate of the hydrolysis process through experimental [38] and theoretical simulation calculations [160], but they cannot accurately provide comprehensive understanding of the entire hydrolysis process. Therefore, a more comprehensive study on the mechanism of the catalyst catalyzing the hydrolysis of AB is required.

Nowadays, the durability of the catalyst is also an important issue for AB catalytic hydrolysis. Although the durability of non-noble metal-based catalysts has been significantly improved through numerous research efforts, most of the current studies only evaluate the catalytic activity of catalysts that can be reused within 10 times. However, in practical applications, it is more desirable for the catalyst to have long-lasting usability in order to reduce the waste of resources, thus making it a challenge task in future research.

Besides, the regeneration strategy after AB hydrolysis is still a major issue. Borate is the main product after the hydrolysis reaction and is very stable. The B–O bond in the by-product is very strong, and it is difficult to convert it into a B–H bond; as a result, expensive strong reducing agents are usually required. Ramachandran et al. regenerated AB from the by-product NH4B(OMe)4 by methanolysis in tetrahydrofuran (THF); LiAlH4 was required as a reducing agent [161]. Liu and his group converted H3BO3 into B(OCH3)3 through methanol esterification and then reacted with NaH to form NaBH4. Then ammonia sulfate reacted with NaBH4 in THF to obtain AB [162]. Regeneration of AB was inseparable from NaBH4 and other hydrogen storage materials, which resulted in increased cost. Besides, regeneration process required a complex process and a large amount of energy. Despite great efforts, the realization of large-scale regeneration of spent fuel still faces major challenges.

Summary and outlook

In this review, recent developments in the use of non-noble metal-based material catalysts for the hydrolysis of AB are summarized. Furthermore, the great improvements made in pure non-noble metal-based materials, metal-based compounds (borides, phosphides, and oxides), and metal/metal-based heterogeneous structural materials, have been discussed. Pure metal-based catalysts are based on the original structure, with significantly improved activity and durability due to microstructural regulation, metal component regulation, and support modification. Therefore, the problem of exposing more active sites to metals and preventing their oxidation and agglomeration is the focus of subsequent research. For the development of metal compounds, non-noble metal-based materials for hydrolysis of AB have been diversified and the consumption of metal resources has been reduced. Metal borides enhance the active surface area, and doping leads to enhancement in activity as indicated by several studies. Metal phosphides and oxides have been receiving increasing attention in recent years for their excellent activity and stability. The heterogeneous structure of metals and metal-based compounds has received significant attention due to their unique internal structure interactions that provide superior capabilities. First, various chemical states of non-noble metals can synergistically enhance the performance of the catalyst. Second, the wide range of compound types offers more possibilities for non-noble metals, which can be exploited for further research. As a whole, non-noble metal-based catalysts have proven to be economical, efficient, and stable catalysts that can be favorable substitutes for noble metals.

Although the catalyst has excellent performance, the study provides more insights into the mechanism of hydrolysis. However, the impact of catalysts on the overall hydrolysis process is not yet known. Moreover, the durability of the catalyst is still an important issue in practical applications. Undeniably, a lot more systematic explorations are still demanded to investigate the issue related to regeneration of AB.

Nonetheless, the development of new non-noble metal-based catalysts can help to further reveal the catalytic mechanism based on obtaining high performance. The study of catalytic mechanism can provide guidance for the design of the efficient catalyst. The design and development of simulation calculation and in situ characterization testing technology is also very beneficial to the study of the structure and chemical state of the catalyst during the reaction process. We look forward to the further development of non-noble metal-based catalysts to meet the requirements of practical applications in the near future.