Keywords

1 Introduction

The burning of fossil energy is inevitably accompanied by the production of large amounts of carbon dioxide, which is the main culprit of the greenhouse effect. To achieve carbon neutrality and carbon peaking, it is necessary to get rid of the dependence on fossil energy. As an emerging clean energy source, hydrogen energy is considered the most promising alternative energy source due to its many advantages [1]. In the hydrogen energy cycle, hydrogen is used as a carrier and contains three main components: preparation, storage and transportation, and application [2]. In contrast to hydrogen application and preparation technologies, hydrogen storage and transportation technologies are relatively underdeveloped. Among the hydrogen storage methods, solid-state hydrogen storage is a hot research topic in this field. And Magnesium hydride (MgH2), as a popular research object among solid-state hydrogen storage materials, possesses high gravimetric hydrogen density (about 7.6 wt%) and volumetric hydrogen density (about 110 g/L), good reversibility and low cost characteristics [3,4,5].

However, poor dehydrogenation kinetics and thermodynamic properties have hindered its commercial application. In practical applications, its operating temperature is 450–550 °C [6]. To enhance the dehydrogenation performance of MgH2, researchers have taken a series of measures, including alloying [7,8,9,10], nanostructuring [11], catalyzing [12,13,14,15], and construction of composite systems [16,17,18]. Among them, catalyst doping is considered as a promising improvement strategy. Catalysts can enhance the hydrogen release thermodynamic properties of MgH2 by reducing the activation energy required for the chemical reaction. Common catalysts such as transition metals, metal oxides and carbon nanomaterials can be used to enhance the dehydrogenation performance of MgH2. However, with the addition of catalysts also makes the gravimetric hydrogen storage density of the system significantly lower. The emergence of metal–carbon composites has emerged as one of the most promising solutions. Carbon-based single-atom catalysts offer the maximum atom utilization while allowing further enhancement of the gravimetric hydrogen storage density.

Single atom catalysts show excellent catalytic activity in many fields, and also have exceptional performance in promoting metal hydride dehydrogenation. Both experiments and theoretical calculations have found that single atom catalysts can enhance the dehydrogenation performance of MgH2. However, the catalytic mechanism of single atom catalysts for the MgH2 system is still unknown. In the experiment, Huang et al. successfully prepared highly dispersed metal catalysts synthesized on nitrogen doped carbon (M–N–C) [23]. It was found that the addition of the nickel based catalysts in MgH2 reduced the activation energy to 87.2 ± 5.4 kJ mol−1. Previous work found the “burst effect” of layer-by-layer dehydrogenation on the surface of MgH2(110) by DFT calculations, where the surface dehydrogenation energy barrier is maximum and then decreases layer by layer [19]. Then we constructed heterojunctions of nine single-atom catalysts with MgH2(110) and found that the surface dehydrogenation energy barrier of MgH2 decreased a lot in the presence of catalysts [20]. In addition, Sun et al. [21] found that graphene doped with N, P, and S was also able to reduce the surface dehydrogenation energy barrier of MgH2. Deng et al. [22]constructed a MgH2/graphene heterojunction and introduced precious metals (Pd and Pt) to further improve dehydrogenation performance. Obviously, we can find that single atom catalysts have an ideal effect in improving the dehydrogenation performance of MgH2. At the same time, it can also ensure the gravimetric hydrogen storage density of solid hydrogen storage materials.

Motivated by the above work, in this work, we constructed nine heterojunction structures. These heterojunction structures are composed of a single atom catalyst with graphene doped with three nitrogen atoms loaded with transition metal atoms as the substrate and a MgH2(110) surface. Firstly, we found the optimal interlayer distance for different heterojunctions, and then conducted structural optimization calculations for all heterojunctions. After calculation, it was found that the catalyst-loaded metal atoms moved further toward the MgH2 surface. At the same time, the H-Mg bond length on the MgH2 surface slightly increased. Moreover, the degree of increase in H-Mg bonds varies slightly among different species. To further explore the intrinsic reasons for this phenomenon, electron localization function, charge density difference and crystal orbital Hamiltonian population analyses were performed on these heterojunction systems. These analyses were used to further improve the knowledge of the mechanism of catalyst-promoted dehydrogenation of MgH2.

2 Computational Methods

In this work, the Projector Augmented Wave (PAW) technique and the Vienna Ab initio Simulation Package (VASP5.4.4) were used to calculate density functional theory (DFT). The Perdew–Burke–Ernzerhof (PBE) function is used to characterize the exchange–correlation energy in the generalized gradient approximation (GGA) [23,24,25]. The modeling of hydrogen adsorption and desorption by PBE is reasonably accurate and in good agreement with the experimental results from ultrahigh vacuum, according to prior benchmark studies and discussions [26, 27]. We also consider spin polarization and van der Waals corrections using the DFT-D3 method [28, 29]. The dipole moment is modified along the z-direction in each calculation. A (4 × 4 × 1) Γ-centered k-point grid was used to achieve structural relaxation, and a plane wave basis set with a 500 eV energy cutoff was used to represent valence electrons. Geometries were considered relaxed whenever the sum of all atom forces fell below 0.02 eV Å−1. To calculate the effect of spin polarization on the system, we set ISPIN = 2. Since there were dipole moments in the slabe model, which can have a certain impact on the results. Therefore, to make the results more accurate, we set LDIPOL = TRUE and IDIPOL = 3.

To investigate the charge change brought on by the addition of SACs (Δρ), we calculated the charge density difference. Charge density difference, one of the most important techniques for studying electronic structures, was calculated using the formula:

$$ \Delta \rho = \rho_{AB} - \rho_{A} - \rho_{B} $$
(1)

where Δρ represents the change value in charge density, ρAB represents the charge density when two fragments of A and B are together, and ρA and ρB respectively represent the charge densities when A and B are individually present at the same position. The Local Orbital Basis Suite towards Electronic-Structure Reconstruction (LOBSTER) software is used to calculate the Crystal Orbital Hamilton Population (COHP) approach, which is used to investigate the bonding mechanism of chemical bonds in the system [30,31,32,33].

In this work, we chose the most stable 110 surface of MgH2 to study the dehydrogenation phenomenon [34]. The catalysts were loaded with transition metal atoms using three N3-doped graphene 2d materials as substrates. To successfully establish a heterojunction structure between the catalyst and MgH2, we constructed a 4 × 2 MgH2 supercell (Fig. 1a). The size of the catalyst was also adjusted to 12.33 × 12.83 Å to reduce the lattice mismatch (Fig. 1b). A 15 Å vacuum layer was set up in the z-direction to separate the images. The heterojunction structure was optimized by setting different layer spacings, and it was found that the heterojunction layer spacing was about 2.75 Å and the energy was about −834.06 eV after optimization (Fig. 2b). The final structure of this work is shown in Fig. 1c.

Fig. 1
figure 1

(a) Side and top views of MgH2(110) surface. (b) Side and top views of a graphene-doped nitrogen substrate loaded with transition metal atoms. SACs-TM (TM = Ti, V, Cr, Mn, Fe, Co, Ni, Cu, and Zn). (c) Side view of an MgH2/SACs-TM heterojunction. White and orange spheres represent H and Mg, respectively. Brown and blue spheres represent C and Ni (using Ni as an example), respectively

Full size image
Fig. 2
figure 2

(a) Illustrations of the Bond 1 and Bond 2 in MgH2. White and orange spheres represent H and Mg, respectively. (b) Distance-energy relationship between MgH2 and SACs-Ni

Full size image

3 Results and Discussion

3.1 Performance of H-Mg Bond Length of Pure MgH2 and MgH2/SACs-TM

To understand what happens after the addition of SACs-TM, we performed structural optimization calculations for all heterojunction models. After convergence of the structural optimization, we found that the bond length of H-Mg in pure MgH2 is 1.85 Å (including Bond 1 and Bond 2). However, a significant increase in the bond length of H-Mg bond in MgH2 occurs regardless of which single atom catalyst is added to MgH2. The largest increase in the bond length of the H-Mg bond was observed for the addition of SACs-Ni, with the bond length reaching about 2.03 Å (Bond 2). In addition, the SACs-Co, V and Cu also increase the bond length of the H-Mg bond to about 2.01 Å (the maximum of Bond 1 and Bond 2), which is slightly less effective than Ni. SACs-Fe had a similar effect on Bond 1 and Bond 2. However, the other SACs-TM have slightly different effects on Bond 1 and Bond 2. Among them, SACs-TM (TM = Ti, V, Cr, Zn) has a stronger effect on Bond 1 and a slightly less effect on Bond 2. On the contrary, SACs-TM (TM = Mn, Co, Ni, Cu) have a stronger effect on Bond 2. We attribute the different effects of different single-atom catalysts on Bond 1 and Bond 2 to the fact that the structure of the heterojunction system will be somewhat different after optimization. It is easily found that in the presence of single-atom catalysts, the bond energy decreases as the bond length of H-Mg becomes longer, thus reducing the energy required to break the H-Mg bond. The addition of these single-atom catalysts weakens the H-Mg interactions and thus facilitates the dehydrogenation process (Fig. 3).

Fig. 3
figure 3

Bond 1 and Bond 2 (H-Mg) bond length changes before and after the addition of SACs-TM (TM = Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn)

Full size image

3.2 Electron Localization Function of Pure MgH2 and MgH2/SACs-TM

To understand the mechanism of H-Mg bond weakening, we analyzed the catalytic mechanism of single-atom catalysts on the MgH2 surface from another perspective. ELF is one of the means to study the electronic structure, where the strength of the H-Mg bond can be analyzed [35]. First, we performed electron localization function (ELF) analysis on pure MgH2 to understand the charge distribution of H atoms on the 110 surface. Then, we further performed ELF analysis on the heterojunction system with the addition of different single-atom catalysts to study the change of H atom charge distribution after the addition of SACs. The red region is used to indicate the higher electron density, while the blue region indicates the lower electron density. We can clearly see that there is a very full red region around the H-Mg bond on the MgH2 surface without the addition of a single-atom catalyst. This indicates that there are more electrons distributed in this part, indicating that the H-Mg bond (bond 1 and bond 2 are very stable). In addition, the red region around the H-Mg bond in the second layer is slightly reduced, but the H-Mg bond here is still more stable. However, the red region around the H-Mg bond on the MgH2 surface is greatly reduced after the addition of the monatomic catalyst. In particular, the red area around Bond 1 and Bond 2 shrinks to a greater extent, because this area is most affected by the catalyst. On the contrary, the red area around the H-Mg bond in the second layer has only a small change. Therefore, we believe that the single-atom catalyst mainly has a greater effect on the charge distribution on the MgH2 surface. By the electron cloud distribution analysis, we can confirm that the H-Mg bonds (Bond 1 and Bond 2) on the surface of MgH2(110) are weakened by the action of the monatomic catalyst. Among all the catalysts, the H-Mg bonds in the presence of SACs-Mn, Co, Cu and Ni are weakened more than the other transition metal atoms and there is almost no electron distribution above the bonds. This is consistent with the previous results on the bond length variation. Some studies [36] have pointed out that the interaction between the unsaturated d-layer electrons of metals and the valence electron of H will weaken the H-Mg bond, thus promoting the dehydrogenation process of MgH2, which is consistent with the research results in this paper (Fig. 4).

Fig. 4
figure 4

ELFs of pure MgH2 and MgH2/SACs-TM (TM = Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn). The section is the 010 plane of the H atom nearest to the metal atom. In the color bars, 1 represents a fully localized state, 0.5 represents a uniform electronic gas state, and 0 represents a fully non-localized state

Full size image

3.3 Charge Density Difference of MgH2/SACs-TM

To understand the mechanism of H-Mg bond weakening, we analyze the heterojunction system from the electronic structure level. Analysis of charge transfer is an essential method to determine whether the bond strength becomes weaker or not. Therefore, we performed charge density difference calculations for all heterojunction systems with doped catalysts. From the calculation results, it can be seen that charge transfer occurs in all the heterojunction systems after the addition of catalysts. The amount of charge transfer is more compared to the previous MgH2/SACs system in the N4 system [20]. Therefore, it is reasonable to believe that the dehydrogenation energy barrier is lower in this system. In detail, different charge transfers occur in different heterojunctions. Among them, both the catalyst and MgH2 in the MgH2/SACs-Ti system lose their charges, and the charges accumulate at the space layer between the heterojunction heterojunctions. However, most of the catalyst in the other heterojunction systems flows to MgH2, while a small amount of charge accumulates at the space layer between the heterojunctions. This charge transfer may be one of the important reasons for the weakening of the H-Mg bond (Fig. 5).

Fig. 5
figure 5

Calculated charge density differences of MgH2/SACs-TM (TM = Ti, V, Cr, Zn, Cu, Mn, Co, Fe, and Ni) induced by the addition of SACs. The yellow and blue isosurfaces indicate electron accumulation and loss, respectively. All the results are plotted with an isovalue of 0.01 e·Å−3. White, orange, and brown spheres represent H, Mg, and C, respectively

Full size image

3.4 Crystal Orbital Hamilton Population of Pure MgH2 and MgH2/SACs-TM

To further reveal the mechanism of H-Mg bond weakening, this work calculates the integral crystal orbital Hamiltonian population (ICOHP) integral values of H-Mg bonds in different systems by means of the COHP method. To facilitate the comparison, the ICOHP values of the H-Mg bonds (Bond 1 and Bond 2) in different systems were statistically analyzed. From the ICOHP values, the sizes of Bond 1 and Bond 2 of pure MgH2 are almost identical at −0.49. Also, we found that the COHP analysis is almost identical and neither of them occupies the antibond orbital. Therefore, we can assume that the H-Mg bonds (Bond 1 and Bond 2) on the surface of pure MgH2 (110) are very stable. However, all ICOHP values changed after the addition of single-atom catalysts. Also, the addition of monatomic catalysts leads to the appearance of antibonding orbitals as can be seen in the COHP analysis. Therefore, we can assume that the catalyst causes a decrease in the stability of the H-Mg bonds (Bond 1 and Bond 2). Compared to the results of previous studies, the value of ICOHP in this work is more variable and the weakening of the H-Mg bonds is more pronounced. We speculate that this phenomenon occurs because the single-atom catalysts in the present work carry metal atoms closer to the H atoms. This again confirms that SACs can promote MgH2 dehydrogenation by weakening the H-Mg bond (Figs. 6, 7).

Fig. 6
figure 6

(a–j) COHP analysis of the H-Mg Bond 1 and Bond 2 in MgH2 and MgH2/SACs-TM

Full size image
Fig. 7
figure 7

Comparison of the integral values of H-Mg Bond 1 and Bond 2 between MgH2 and MgH2/SACs-TM in the system. The top half is from this work, using a N3-doped catalyst, and the bottom half is from the previous work, using a N4-doped catalyst

Full size image

After the above analysis, it has been confirmed that the single atom catalyst doped with N3 has a better weakening effect on the H-Mg bond. And further understanding of the weakening mechanism of H-Mg bonds. Next, study the effects of different coordination environments on catalysts to find a descriptor. Finally, we hope to design catalysts using the descriptor to provide theoretical guidance for the commercial application of Mg based solid hydrogen storage materials.

4 Conclusion

In summary, we optimized pure MgH2 and nine MgH2/SACs-TM heterojunction systems using DFT-D3 calculations of spin polarization, respectively. We found that the MgH2 dehydrogenation performance of all heterojunction systems was improved by the addition of SACs. This was mainly manifested by a significant increase in the outermost H-Mg bond length of MgH2(110) after the addition of the catalyst. However, the addition of different SACs-TM can have different effects on Bond 1 and Bond 2. We speculate that this is attributed to some differences in the conformation of different heterojunction systems after optimization. To in-depth analyze the binding properties of SACs and MgH2 and the weakening mechanism of the H-Mg bond, electron localization function, charge density difference and crystal orbital Hamiltonian population analyses were performed on the heterojunction. Furthermore, this work provides important guidance for the design of novel MgH2/SACs heterojunction materials and offers a solution for the slow kinetics of MgH2 dehydrogenation in hydrogen storage.