Page 72 - Read Online

Meng et al. J. Mater. Inf. 2025, 5, 3 https://dx.doi.org/10.20517/jmi.2024.74 Page 13 of 25

Chen et al. investigated the NRR activity of SACs, DACs, and TACs supported on heterogeneous
graphylene and graphene, denoted as M -GDY/Gra (where M = Mn, Fe, Co, and Ni; x = 1, 2, 3). Their
x
findings showed that TACs exhibited better stability and catalytic performance than SACs and DACs, with
Fe -GDY/Gra with a theoretical mass loading of 35.8 wt% achieving a particularly low U of -0.26 V. This
3
L
enhanced activity was attributed to the M active site, which not only provided additional electrons for N
2
3
[103]
activation but also displayed weaker adsorption of the product, facilitating easier release .
The reaction mechanisms and U of DACs and TACs are summarized in Table 3; however, TACs are not
L
always superior to DACs. Luo et al. demonstrated that Fe /MoS exhibits high catalytic activity for NRR.
2
2
Their findings suggest that, compared to Fe/MoS , the presence of adjacent Fe atoms in Fe /MoS enables a
2
2
2
side-on adsorption configuration for N , which is more favorable for effective activation. However, the
2
addition of a third Fe atom (Fe /MoS ) reduces electron sharing between Fe atoms, thereby inhibiting N
3
2
2
adsorption .
[110]
In summary, the factors influencing NRR activity and selectivity are complex, and the catalytic performance
of multi-atom clusters in NRR depends on the specific atomic configuration. Further research is needed to
understand the underlying characteristics.
TMFCs
2
3
TMFCs, such as boron-based catalysts, have shown promise for the NRR. Due to the sp or sp electron
configuration, boron atoms can provide empty orbitals that facilitate N activation. For example, Anis et al.
2
investigated single, double, and triple boron atoms supported on a GDY monolayer (B@GDY, B @GDY,
2
and B @GDY) for NRR, concluding that B @GDY demonstrates outstanding catalytic performance and
3
3
[111]
effectively suppresses HER [Figure 7A] . Similarly, Wang et al. proposed that a g-C N monolayer
3
4
embedded with double B atoms (B @C N ) enhances N activation [Figure 7B]. Their findings suggest
3
2
4
2
strong hybridization between N -2p orbitals and B-sp orbitals, which accounts for the high catalytic
3
2
efficiency of B @C N 4 [112] . In a related study, Rasool et al. explored NRR on single- and double-boron-doped
3
2
configurations across five different substrates, demonstrating that g-C N is particularly effective .
[113]
3
4
These studies highlight boron-based TMFCs as efficient and selective NRR catalysts, offering a viable
alternative to TMs by utilizing unique electronic interactions that enhance N activation while suppressing
2
HER. This approach broadens the materials landscape for sustainable, TM-free catalysis. More efforts are
encouraged to explore such catalysts, especially with elements beyond boron.
HIGH-THROUGHPUT CALCULATIONS AND EMERGING MACHINE LEARNING TOWARDS
NRR
The search for materials with specific properties is challenging, as traditional trial-and-error methods are
often inefficient. However, combining high-throughput computing with machine learning (ML) offers a
powerful approach for materials prediction and design, addressing long development cycles and high costs,
and accelerating the discovery of novel catalysts .
[114]
High-throughput DFT calculations
TM elements, coordination environments, and substrates significantly influence NRR efficiency, and the
numerous possible combinations can greatly benefit from high-throughput DFT calculations to streamline
the search for optimal configurations .
[115]