2. Formic Acid Dehydrogenation
Formic acid (HCOOH) is a promising liquid chemical hydrogen material owing to its advantages of low toxicity, high hydrogen content (4.4 wt.%), high stability, and easy availability through biomass processing and CO
2 hydrogenation
[11][12][13]. The stored hydrogen in formic acid can be released through two pathways, dehydrogenation (1) and dehydration (2)
[14]:
Given that dehydration (2) not only reduces hydrogen production but also generates carbon monoxide (which is toxic to dehydrogenation and fuel cell catalysts), the key to the practical application of formic acid dehydrogenation lies in the development of high-performance MNP catalysts. These catalysts can be used for gas-phase or liquid-phase dehydrogenation and can be easily separated from the reaction mixture as compared to dissolved metal complexes.
2.1. Monometallic Catalysts
Au bulk catalysts are inactive for the formic acid dehydrogenation. Surprisingly, Au NPs anchored on specific supports are highly active for this reaction, due to its surface electronic structure change resulting from the strong metal–support interaction. Xu et al. investigated the relationship between Au NPs and spherical SiO
2, and found that Au NPs coated with amine-functionalized SiO
2 spheres showed high dehydrogenation rates, releasing CO-free hydrogen at 363 K
[15]. Au NPs coated with SiO
2 spheres without amine modification possessed much poor catalytic activity. Cao et al. anchored ~1.8 nm Au NPs on the acid-resistant ZrO
2 by regulating the deposition and activation processes
[16]. The resultant Au/ZrO
2 catalyst was highly active for the dehydrogenation of formic acid/triethylamine mixture, achieving an initial turnover frequency (TOF) value of 1590 h
−1 at 323 K, without any traces of CO. The TOF value maintained at 1200 h
−1 even after 100 h of continuous operation at 323 K. Zhang et al. loaded Au NPs on the Schiff base modified SiO
2 surface, which showed fast dehydrogenation of high concentrations of formic acid in the absence of additives, owing to the excellent C-H bond activation performance of the interfaces between the protonated Schiff base and Au atoms
[17]. The initial TOF value reached as high as 4368 h
−1 at 323 K and the formic acid concentration of 10 mol L
−1.
For the gas-phase reaction over catalysts with approximately the same mean size of the Au particles (2.4–3 nm) supported on different oxide supports, Zacharska et al. also demonstrated that the hydrogen production from formic acid strongly depends on the nature of the support
[18]. An Au/Al
2O
3 catalyst showed a better performance in the reaction. The catalytic properties of this catalyst could be further improved by doping with potassium ions
[19]. Recently, Bulushev et al. showed that the doping of an Au catalyst supported on carbon nanotubes with cucurbituril led to a significant decrease of the reaction temperature by 110 K due to the hydrogen abstraction from formic acid by carbonyl portals of cucurbituril
[20].
Compared to Au catalysts, Pd NPs were explored extensively for the formic acid dehydrogenation. Yu et al. encapsulated ultrafine Pd NPs (~1.5 nm) inside the pores of Silicalite (MFI), which were highly active for the dehydrogenation of formic acid/sodium formate mixture, and the TOF value reached 3027 h
−1 at 323 K with 100% hydrogen selectivity
[21]. Yu et al. further found that the simultaneous encapsulation of Pd NPs and Ni(OH)
2 within zeolite S-1 boosted the catalytic activity of Pd NPs significantly
[22]. The initial TOF value reached 5803 h
−1 at 333 K, when the optimal molar ratio of Pd and Ni(OH)
2 was 4/1.
In Xu’s group, a series of highly catalytically active Pd NPs were synthesized for the dehydrogenation of formic acid: (a) by using an NaOH-assisted NaBH
4 reduction strategy, highly dispersed Pd NPs were immobilized on porous carbon MSC-30, which possessed a TOF value of 2623 h
−1 at 323 K and 100% hydrogen selectivity
[23]; (b) by using anhydrous methanol as a surface capping agent and reducing agent, ultrafine Pd NPs (~1.4 nm) were loaded on porous carbon Vulcan XC-72R, which showed an initial TOF value of 4452 h
−1 at 323 K
[24]; (c) by functionalizing graphene oxide with
p-phenylenediamine (PDA), highly dispersed Pd NPs (~1.5 nm) were immobilized on the surface of graphene, which showed a TOF value of 3810 h
−1 at 323 K; even if pure formic acid was used, this catalyst maintained high catalytic activity, achieving a TOF value of 1500 h
−1 at the same temperature
[25]; (d) by functionalizing porous carbon MSC-30 with abundant N sites through the pyrolysis of urea at low-temperature, ultrafine Pd NPs (~1.4 nm) were anchored on the surface of MSC-30, which exhibited a record TOF value of 8414 h
−1 at 333 K and there was no change in activity and selectivity after 15 cycles
[26]; and (e) highly dispersed Pd NPs were immobilized on Fe-N-C support, which produced a high TOF value of 7361 h
−1 at 323 K without any traces of CO, owing to the strong interaction between Pd and Fe-N sites
[27]. Cao et al. immobilized Au, Ir, Pt, Ru, Rh, Pd and other monometallic NPs on the surface of reduced graphite oxide (rGO) for the dehydrogenation of potassium formate solution and found that only Pd NPs showed extremely high initial TOF, up to 11,299 h
−1 at 353 K—the other particles were inactive
[28]. The lattice mismatch between the Pd nanoparticles and the surface of rGO was considered to be the main reason for the enhanced catalytic activity. Cao et al. also used the strong interaction between pyridine nitrogen sites and Pd to anchor Pd NPs on the surface of N-doped carbon
[29]. The obtained Pd/N-doped carbon catalyst efficiently catalyzed the dehydrogenation of additive-free formic acid at room temperature, achieving a TOF of 5530 h
−1 without any traces of CO.
For the gas-phase reaction, Bulushev et al. prepared supported Pd catalysts with single Pd atoms using N-doped carbon
[30][31] and found that single-atom Pd sites showed a TOF value three-to-four times higher than that of Pd atoms in N-free carbon-supported Pd particles and support-free Pd catalyst. Jia et al. found that doping of dispersed Pd catalysts on different supports with potassium ions leads to an increase of the TOF values by a factor of up to 33
[32]. The TOF value reached 3600 h
−1 at 353 K for the K-doped Pd/C catalyst with a mean particle size of 3.6 nm. Bulushev and Bulusheva recently discussed the role of supported single metal atoms in the reaction and showed that they can sometimes perform better than NPs
[33]. This could also be valid for some Ru or Ir immobilized complexes
[34].
Notably, the non-noble metal catalysts are more promising for commercial applications owing to their cost. However, the non-noble metal catalysts often exhibit poor catalytic activity in the dehydrogenation of formic acid, as they can be etched easily by formic acid during the dehydrogenation reaction. By pyrolysis of the cobalt acetate, phenanthroline and Vulcan XC-72R mixture at high temperature, Beller et al. synthesized carbon-supported Co active sites with good resistance to acid corrosion, which possessed excellent catalytic activity and stability for the dehydrogenation of formic acid at 371 K, achieving an average H
2 production rate of 180 mL g
cat−1 h
−1 after continuous operation for 6 days
[35], owing to the highly dispersed CoN
x active centers. Furthermore, Beller et al. synthesized the N-doped carbon-supported Co single-atom catalysts (Co-N-C) through the pyrolysis of ZnCo-ZIFs at high temperature, which exhibited high activity in the catalytic hydrogen production from formic acid in propylene carbonate solvent
[36], achieving an average H
2 production rate of 319.2 mL g
cat−1 h
−1 (initial TOF=357 h
-1) much higher than that of Co NPs prepared by the same process.
2.2. Bimetallic and Multimetallic Catalysts
Given that the intrinsic catalytic capability of metals can be enhanced through the synergistic effect of different metal elements, the bimetallic or multimetallic catalysts are a promising choice for the highly efficient dehydrogenation of formic acid. As early as in 2008, Xing et al. prepared AgPd/C and AuPd/C for the dehydrogenation of a formic acid/sodium formate mixture, which showed much higher catalytic activities than those of Ag/C, Au/C and Pd/C
[37]. By introducing CeO
2 as a co-catalyst, the TOF of AuPd/C was further improved, reaching 832 h
−1 at 373 K. This group further synthesized the AuPd@Cu/C core-shell catalyst based on the weak CO adsorption characteristic of the Au surface, which showed an average H
2 production rate of 90 mL g
−1 min
−1 at 365 K, with a small trace of CO
[38]. Tsang et al. synthesized various core-shell catalysts by depositing Pd on the surface of MNPs, such as Ag, Rh, Au, Ru and Pt, respectively
[39]. Compared with the monometallic catalysts, the catalytic capabilities of the core-shell catalysts were enhanced significantly, and Ag@Pd showed a TOF value of 252 h
−1 at 323 K, the highest activity among all the core-shell catalysts. The enhanced catalytic capability was attributed to the strong electronic interaction between core metal and shell metal. Chen et al. synthesized AuPd NPs on porous carbon by using a Mg
2+ ion-assisted low-temperature reduction strategy, which possessed an initial TOF value of 1120 h
−1 for the dehydrogenation of formic acid without a trace of CO
[40].
The support can control the size, distribution and metal–support interaction of MNPs, and then boost their catalytic capabilities. In an early study, Xu et al. encapsulated highly dispersed AuPd NPs inside the ethylene diamine-modified ED-MIL-101, which showed enhanced catalytic activity for the dehydrogenation of formic acid with good resistance to CO poisoning, in comparison with AuPd NPs inside MIL-101
[41]. Since then, large numbers of bimetallic NPs were reported. For example, Yan et al. synthesized AuPd and MnO
x NPs on a composite support of ZIF-8/graphene, which showed a high activity for the catalytic hydrogen production from pure formic acid, achieving an initial TOF value of 328 h
−1 at room temperature
[42]. This group further immobilized AuPd NPs (~2.4 nm) on amine-modified N-doped graphene, which were highly active for the dehydrogenation of formic acid, achieving an initial TOF value as high as 4445.6 h
−1 at room temperature
[43]. The enhanced catalytic activity was attributed to the low adsorption capacity of Pd surface after alloying. Xu et al. immobilized AgPd NPs (~2.5 nm) on a ZrO
2/C/rGO support, which showed a high TOF value of 4500 h
−1 at 323 K for the dehydrogenation of the formic acid/sodium formate mixture
[44].
New strategies have also been developed to control the size, distribution and metal-support interaction of bimetallic NPs, and then improved their catalytic performance. Xu et al. developed a non-noble metal sacrificial approach, based on the different solubility between noble (Ag/Au/Pd) and non-noble (Co) metals in acid solution (H
3PO
4), to immobilize highly dispersed AgPd and AuPd NPs on reduced graphene oxide (rGO), respectively
[45][46]. The obtained AgPd/rGO and AuPd/rGO exhibited high activity and selectivity for the dehydrogenation of the formic acid/sodium formate mixture, achieving TOF values of 2739 h
−1 and 4840 h
−1 at 323 K without the releasing of CO, respectively. In addition, Xu et al. developed a solid-ligand-assisted approach, based on the strong binding ability of N sites to Au
3+ and Rh
3+ ions, to immobilize the bulk immiscible AuRh NPs (~1.6 nm) on the surface of porous carbon with abundant N and O groups
[47]. The pure Au or Rh NPs obtained by using the similar approach were inactive in the dehydrogenation of the formic acid/sodium formate mixture. Surprisingly, the AuRh NPs showed outstanding activity and the TOF value reached 2297 h
−1 at 333 K.
In addition to bimetallic NPs, multimetallic NPs have exhibited much high catalytic activity and selectivity for the dehydrogenation of formic acid. Jiang et al. immobilized CoPdAu NPs and NiPdAu NPs on porous carbon Vulcan XC-72R by using a wet-chemical reduction method, respectively
[48][49]. Since the second and third components modified the electronic structure of Pd, the CoPdAu/C and NiPdAu/C catalysts showed good catalytic activities, and the TOF values reached 80 and 12.4 h
−1 at room temperature, respectively. Zahmakiran et al. prepared a PdNiAg/C catalyst with a particle size of ~5.6 nm, which exhibited a TOF value of 85 h
−1 at 323 K in the dehydrogenation of formic acid/sodium formate mixture
[50]. Luo et al. synthesized CoAgPd NPs (~2.8 nm) by oleic acid as a reducing agent and stabilizer
[51]. Such ternary NPs achieved a TOF value of 110 h
−1 for the dehydrogenation of the formic acid/sodium formate mixture.
3. Hydrazine Hydrate Dehydrogenation
Hydrazine hydrate (N
2H
4·H
2O) is often used as a reducing agent, anti-oxidant, or fuel. Given its high hydrogen content (8.0 wt.%), outstanding thermal stability and liquid state at room temperature, hydrazine hydrate can be considered as a promising hydrogen storage material
[52][53]. At the same time, this compound is toxic and it is difficult to produce hydrazine hydrate back from the by-product N
2.
If researchers use a catalyst, the hydrogen stored in hydrazine hydrate will release through two pathways: complete decomposition (3) or incomplete decomposition (4):
Complete decomposition (3) is an ideal route for producing hydrogen. Incomplete decomposition (4) produces ammonia, which not only reduces the yield of hydrogen but also poisons the fuel cell catalysts and even the environment. In order to achieve a commercial application of hydrazine hydrate as a chemical hydrogen material, the high-performance catalysts for its complete decomposition are highly desirable.
3.1. Monometallic Catalysts
In an earlier study, Xu et al. prepared a series of noble (Rh, Ru, Ir, Pt, Pd) and non-noble (Co, Cu, Ni, Fe) MNPs by using cetyltrimethylammonium bromide (CTAB) as a capping agent for the decomposition of hydrazine hydrate
[54]. The experimental results indicated that Rh NPs have the highest hydrogen selectivity, up to 44%; Ru, Ir, Co NPs have a certain activity with the selectivity of <7%; other NPs, such as Pt, Pd, Cu, Ni, and Fe, were inactive. The researchers proposed that Ru, Ir, and Co NPs tend to break N-N bonds rather than N-H bonds, and thus promote the occurrence of incomplete decomposition. Zhang et al. prepared an Ni/Al
2O
3 nanocatalyst by using co-precipitation method combined with high-temperature reduction and Ni-Al hydrotalcite as a raw material
[55]. The Ni/Al
2O
3 catalyst exhibited a selectivity of 93% for the complete decomposition of hydrazine hydrate at 303 K, and this selectivity was increased to 96% through the introduction of KOH. This group further used amorphous CeO
2 as a support to anchor Ni NPs (Ni/CeO
2)
[56]. When the molar ratio of Ni to CeO
2 was 1/0.08, the Ni/CeO
2 catalyst exhibited good catalytic activity (TOF = 56.1 h
−1, 303 K) and selectivity (99%). Similarly, Varma et al. prepared a Ni/CeO
2 nanocatalyst by a solution combustion method. When the loading of Ni reached 6 wt.%, Ni/CeO
2 exhibited good catalytic activity (TOF = 34.0 h
−1, 323 K) and H
2 selectivity (100%)
[57].
Notably, although monometallic NPs often show poor catalytic performance for the complete decomposition of hydrazine hydrate, their catalytic performance can be improved by doping, increasing the number of basic sites and strengthening the metal–support interactions
[58][59]. Wang et al. encapsulated Ni NPs inside titanate nanotubes, which exhibited a TOF value of 96.0 h
−1 at 333 K (1 M NaOH) and 100% H
2 selectivity, owing to the high dispersion of Ni NPs and the enhanced interaction between Ni and titanate
[58]. Lu et al. confined Ni-CeO
2 nanowires on SiO
2 by the reverse micelle method, which possessed high catalytic activity (TOF = 219.5 h
−1, 343 K) and H
2 selectivity (100%), due to the enhanced electronic synergistic effect between Ni NPs and the abundant oxygen vacancies of CeO
2 [59].
3.2. Bimetallic Catalysts
Compared to the studies with monometallic NPs, the studies of Ni-based alloy NPs, especially RhNi and PtNi catalysts, have contributed too much to the area of decomposition of hydrazine hydrate. In 2009, Xu et al. synthesized ~3 nm RhNi NPs by using the surface capping agent CTAB, which showed 100% H
2 selectivity when the molar ratio of Rh to Ni reached 4/1
[60]. In addition, their H
2 selectivity relied significantly on the reaction temperature: the H
2 selectivity was only 28% at room temperature, however, it increased to 100% when the temperature reached 323 K
[61]. In 2010, this group found that PtNi alloy NPs are highly active for the complete decomposition of hydrazine hydrate, though the Ni and Pt NPs are often inactive
[62]. Since then, highly dispersed PtNi NPs (~2.4 nm) were synthesized by NaOH-assisted NaBH
4 reduction process, which exhibited a TOF value of 150 h
−1 [63]; ultrafine PtNi NPs were immobilized on 3-dimensional (3D) N-doped graphene frameworks, which achieved 100% H
2 selectivity and a recorded TOF value of 943 h
−1 at 303 K, and such 3D N-doped graphene frameworks were synthesized by the self-assembly of graphene layers with simultaneous nitrogen doping via crosslinking of graphene oxide with melamine formaldehyde resin under hydrothermal conditions followed by carbonization
[64]; PtNi NPs were anchored on ZrO
2/porous carbon/graphene, which showed 100% H
2 selectivity and a high TOF value of 1920 h
−1 at 323 K
[65].
The strong metal–support interaction can significantly enhance the catalytic activity and selectivity of bimetallic NPs for the complete decomposition of hydrazine hydrate. In one example, Lu et al. encapsulated ultrafine PtNi NPs inside MIL-101 and found that the confinement of pores can significantly enhance the interaction between PtNi and MIL-101, thus boosting their catalytic activity (TOF = 140 h
−1, 303 K), selectivity (100%) and stability (not less than 20 cycles)
[66]. In another example, this group immobilized PtNi NPs on the surface of TiO
2-modified Ti
3C
2T
x nanosheets, which exhibited a high TOF value of 1220 h
−1 at 323 K, due to the enhanced synergistic effect between PtNi atoms and the abundant functional groups on Ti
3C
2T
x nanosheet
[67]. In addition, Wang et al. prepared N-doped carbon spheres by co-pyrolysis of SiO
2 spheres and dopamine for immobilizing PtNi NPs
[68]. The obtained PtNi/N-doped carbon showed a high TOF value of 1602 h
−1 at 323 K. N-doped graphene hydrogels with abundant N sites were also used for anchoring PtNi-CeO
x NPs
[69]. The PtNi-CeO
x/N-doped graphene hydrogels achieved a high TOF value of 3064 h
−1 at 323 K for the dehydrogenation of hydrazine hydrate.
PdNi and IrNi alloys are another choice for the efficient catalytic hydrogen production of hydrazine hydrate. Xu et al. found that Pd
0.4Ni
0.6 NPs exhibited the best catalytic activity and selectivity in all their metal ratios, achieving a hydrogen selectivity of 82% at 323 K
[70]. Lu et al. anchored IrNi NPs on the surface of lanthanum oxycarbonate, which showed a high TOF value of 1250 h
−1 at 323 K, as the abundant basic sites on the surface of lanthanum oxycarbonate facilitated the cleavage of N-H bonds
[71].
In addition to Ni-based alloy, Co-based alloys also showed high catalytic performance. Yan et al. prepared Co
0.65Pt
0.30(CeOx)
0.05 nanocatalysts by a wet-chemical method, which showed H
2 selectivity of 72.1% at room temperature and a TOF value of 194.8 h
−1 [72]. Lu et al. prepared PtCo/La(OH)
3 catalyst, based on the outstanding synergistic effect between PtCo and the abundant basic sites on La(OH)
3, which showed 100% H
2 selectivity and a high TOF value of 2400 h
−1 at 323 K
[73].
Heteroatom doping can change the electronic structure of active metals and then improve their catalytic performance. Zhang et al. introduced B element into RhNi to form a RhNiB catalyst, which possessed an eight times higher catalytic activity than that of RhNi at room temperature and alkaline conditions, and 100% H
2 selectivity
[74]. Luo et al. synthesized graphene-supported RhNiP NPs, which showed a five times higher catalytic activity than that of graphene-supported RhNi NPs, achieving a TOF value of 471 h
−1 at 323 K and alkaline conditions
[75]. Liu et al. prepared graphene-supported PtNiP NPs by using sodium hypophosphite as the P source
[76]. Due to the synergistic effect resulting from P doping and the strong metal–graphene interaction, the Pt
0.3Ni
0.7P/rGO catalyst showed high catalytic activity for the dehydrogenation of hydrazine hydrate, demonstrating a TOF value of 742 h
−1 at 323 K.
In order to obtain low-cost catalysts, the researchers have made great efforts to design and prepare alloy catalysts containing two non-noble elements. Lu et al. prepared Cr
2O
3-modified NiFe NPs for the dehydrogenation of hydrazine hydrate to produce hydrogen
[77]. Compared to monometallic Fe and Ni catalysts, NiFe-Cr
2O
3 showed higher catalytic activity, and the TOF value and H
2 selectivity reached 893.5 h
−1 and 100% at 343 K, respectively. The presence of Cr
2O
3 not only restricted the aggregation of NiFe NPs but also enhanced synergistic effect between NiFe and Cr
2O
3. Varma et al. prepared a NiCu/CeO
2 catalyst by one-step solution combustion method for the dehydrogenation of hydrazine hydrate
[78]. When the alkali concentration was 6 M and the temperature was 323 K, the TOF value reached 1450 h
−1—much higher than that of Ni/CeO
2. Wen et al. immobilized highly dispersed NiCo NPs (~4 nm) on the surface of ultrathin layered NiO-CoO
x for the catalytic hydrogen production from hydrazine hydrate
[79]. The NiCo/NiO-CoO
x exhibited the TOF of 5.49 h
−1 and 100% H
2 selectivity. The interaction between metals and the high density of basic sites was considered as the main reason for its enhanced catalytic performance.
4. Ammonia Borane Dehydrogenation
Ammonia borane (NH3BH3) is another promising chemical hydrogen storage material owing to its high hydrogen content (19.6 wt.%), low molecular weight (30.9 g mol−1), and good solubility in water, and other solvents. Ammonia borane can release hydrogen through thermal decomposition, hydrolysis (5) or methanolysis (6) with the use of an appropriate catalyst. In the research, researchers do not consider thermal decomposition. The key to promoting the practical application of ammonia borane as hydrogen source lies in the development of efficient and stable catalysts.
For practical utilization of water solutions of ammonia borane for hydrogen production, it should be considered that it is difficult to convert the by-product NH4BO2 back to ammonia borane due to the problems of converting the B-O bonds into the B-H bonds.
4.1. Hydrolysis of Ammonia Borane
4.1.1. Monometallic Catalysts
In 2006, Xu et al. reported for the first time that Pt, Ru and Rh NPs possessed high catalytic activity for the hydrolysis of ammonia borane under ambient conditions
[80], which attracted much attention in the scientific community on the application of ammonia borane as a practical chemical hydrogen storage material. Since then, various high-surface-area solid materials, such as porous carbon, mesoporous SiO
2, metal oxides, and MOFs, have been explored as supports to anchor highly dispersed MNPs for the catalytic hydrogen production from ammonia borane. For example, a double-solvents approach was developed to encapsulate small and highly dispersed Pt NPs (~1.8 nm) inside the pores of MIL-101, which possessed high catalytic activity for the ammonia borane hydrolysis, achieving a TOF value of 414 min
−1 [81]. Ultrasmall and uniformly dispersed Ru NPs were immobilized on the graphene, which showed a high initial TOF value (600 min
−1) at room temperature
[82]. Pt and Rh NPs were immobilized on the surface of activated carbon nanotubes, which efficiently catalyzed the hydrolysis of ammonia borane to produce hydrogen, achieving high TOF values of 567 min
−1 and 760 min
−1, respectively
[83][84]. Highly dispersed Pt NPs were anchored on the surface of graphene by a non-noble metal sacrificial approach
[85], which showed a high TOF value of 284 min
−1 at 298 K. The MOF-derived Ru/Cu@C catalysts exhibited a TOF value of 97 min
−1 at room temperature for ammonia borane hydrolysis
[86]. Ultrasmall Pt clusters were immobilized within the charged ionic organic molecular cage I-cage-Cl based on the electrostatic repulsion effect, and the obtained Pt@I-cage-Cl catalyst showed a unique on–off effect on the hydrolysis of ammonia borane with a TOF value of 115 min
−1 at 300 K
[87].
Notably, non-noble metal NPs can exhibit good catalytic activity in the hydrolysis of ammonia borane, which is of great importance for practical applications. Lu et al. found that the catalytic activity of Co NPs on polyethyleneimine-modified graphene oxide was promoted by the synergistic effect of Co and amino groups, showing a TOF value of 39.9 min
−1 at room temperature
[88]. Shaabani et al. used 3D graphene framework-supported Ni NPs to perform hydrolytic ammonia borane decomposition, and achieved a TOF value of 41.7 min
−1 at room temperature
[89]. However, non-noble metals often show unstable catalytic activities, as they are easily oxidized in air. In this regard, the researchers synthesized CoB
[90], CoP
[91], and Ni
2P
[92] catalysts, which showed excellent anti-oxidation ability and catalytic activity in the catalytic hydrogen production from ammonia borane.
4.1.2. Bimetallic Catalysts
Compared to monometallic catalysts, bimetallic catalysts often show better catalytic activity and selectivity in the hydrolysis of ammonia borane
[93][94][95][96][97]. For example, Rakap et al. reported that the polyvinylpyrrolidone (PVP)-stablized PdRh catalyst possessed a TOF value of 1333 min
−1 for the hydrolysis of ammonia borane at room temperature, which was much higher than that of monometallic Pd or Rh
[94]. Cho et al. synthesized octahedral Au@Pt core-shell NPs by using polydiallyldimethylammonium (PDDA) as stabilizer through polyol reduction method
[95]. When Pt/Au molar ratio was optimized to be 1/1, the Au@Pt nanocatalyst exhibited the best catalytic activity with an activation energy of 44.28 kJ mol
−1. Yu et al. encapsulated ultrafine Rh
0.8Ru
0.2 NPs inside Si-OH-modified MFI zeolites, which showed high catalytic activity for the hydrolysis of ammonia borane, and the TOF value reached 1006 min
−1 at room temperature
[96].
The introduction of a non-noble metal into a noble metal catalyst can enhance the catalytic performance, while reducing the catalyst cost. Li et al. immobilized RuCu NPs (~5.4 nm) on the surface of TiO
2@C-N for the hydrolysis of ammonia borane, which showed a high TOF value of 626 min
−1 at room temperature (based on Ru metal), when the molar ratio of Ru/Cu was 3/2
[97]. Zhang et al. immobilized Pt
0.4Ni
0.6 NPs (~2.6 nm) on the graphene for the hydrolysis of potassium borohydride/ammonia borane mixture, which showed a TOF value of 408 min
−1 at 308 K
[98].
By controlling the size, distribution and metal–support interaction, the catalytic performance of the bimetallic catalysts can be improved significantly. Ren et al. immobilized AuCo NPs on carbon nitride (CN) by impregnation for the catalytic hydrolysis of ammonia borane
[99]. With light radiation, the AuCo/CN catalysts showed excellent catalytic activity (TOF = 2897 h
−1) at room temperature. The synergistic effect between Au and Co, and the Motty–Schottky effect at the metal-semiconductor interface were responsible for the enhanced catalytic performance. By using the double-solvents approach combined with a high-concentration NaBH
4 reduction strategy, Xu et al. encapsulated AuNi and AuCo NPs inside the pores of MIL-101, respectively
[100][101]. Owing to the small size effect and high dispersion, AuNi@MIL-101 and AuCo@MIL-101 showed catalytic activities for the hydrolysis of ammonia borane. The optimal TOF values reached 66.2 and 23.5 min
−1 at room temperature, respectively. Astruc et al. encapsulated ultrafine Ni
2Pt NPs inside the pores of ZIF-8 by a deposition-precipitation process, which exhibited excellent catalytic activity for the hydrolysis of ammonia borane, owing to the enhanced small size effect, distribution, and metal–support interaction
[102]. The optimal TOF value reached 2222 min
−1. Ma et al. immobilized highly dispersed CoNi NPs on the surface of α-MoC through a strong metal–support interaction, which possessed a high catalytic activity for the hydrolysis of ammonia borane
[103]. The TOF value reached 321.1 min
−1 at room temperature; one of the best TOF value showed by non-noble metals.
4.2. Methanolysis of Ammonia Borane
In recent years, researchers have explored a series of MNPs for the highly efficient methanolysis of ammonia borane, which provides a new method for the practical application of chemical hydrogen storage. In one case, Wang et al. loaded Ru NPs (~5 nm) on the surface of montmorillonite (MMT) by an ion-exchange method with hydrogen reduction, which showed an outstanding H
2 release rate up to 29 L min
−1 g
−1 for the methanolysis of ammonia borane
[104]. Such a catalyst showed good durability, as Ru NPs were firmly stabilized inside the montmorillonite interlayers. Xu et al. used the imine molecular cage CC3R to immobilize Rh NPs, which showed excellent catalytic activity and selectivity for the methanolysis of ammonia borane, achieving a high TOF value of 215.3 min
−1 at room temperature
[105]. The enhanced catalytic activity of Rh NPs was attributed to the good solubility of CC3 cages, which facilitated the homogenization of the heterogeneous Rh NPs with high dispersibility in reaction solvents. By N
2 flowing combined with water or methanol washing, the catalysts were easy to recycle after use.
Given the hydrophobicity of organic molecular cages, Xu et al. developed a reverse double-solvents approach to encapsulate ultrafine metal clusters inside the organic molecular cage RCC3
[106]. This approach involved three steps: (a) the hydrophobic RCC3 cages were dispersed into H
2O solution firstly, and then a small amount of hydrophobic solvent (CH
2Cl
2) containing metal precursors was slowly dropped; (b) the RCC3 cages adsorbed the CH
2Cl
2 containing metal precursors into their hydrophobic cavities; and (c) by reducing with the high-concentration NaBH
4, the metal precursors inside RCC3 were converted to metal clusters. The Pd@RCC3 catalyst showed a high TOF value of 176 min
−1 at 303 K for the methanolysis of ammonia borane; one of the best activities among Pd catalysts. This group further encapsulated ultrastable Pd clusters inside the reticular Schiff base nanospaces, which showed a similar TOF value of 177 min
−1 at 303 K for the methanolysis of ammonia borane
[107]. The clean surface of metal clusters and the synergistic effect between Schiff base sites and metal atoms were responsible for the enhanced catalytic performance. In addition, Sun et al. encapsulated metal clusters inside ionic molecular cages and racemic molecular cages, which also showed excellent catalytic activities and selectivity for the catalytic methanolysis of ammonia borane
[108][109].