SciELO - Scientific Electronic Library Online

 
vol.68 número3The Influence of Additives Upon the Energetic Parameters and Physicochemical Properties of Environmentally Friendly Biomass PelletsExperimental Studies to Optimize Process Parameters for the Removal of Cationic and Anionic Dyes by Natural Cypress Leaves índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados

Revista

Articulo

Indicadores

Links relacionados

  • No hay artículos similaresSimilares en SciELO

Compartir


Journal of the Mexican Chemical Society

versión impresa ISSN 1870-249X

J. Mex. Chem. Soc vol.68 no.3 Ciudad de México jul./sep. 2024  Epub 04-Mar-2025

https://doi.org/10.29356/jmcs.v68i3.2073 

Articles

A DFT Study of the Hydrogen Storage Potentials and Properties of Ca, Fe, and Ti Deposited NaSi20 Fullerenes

Huixi Yang1 

Bin Liu1  * 

Hongjiang Ren2  * 
http://orcid.org/0000-0002-1954-672X

1Xi’an Key Laboratory of Textile Chemical Engineering Auxiliaries, School of Environmental and Chemical Engineering, Xi’an Polytechnic University, Xi’an 710600, PR China.

2School of Chemical Engineering, Xi’an University, Xi’an 710065, PR China.


Abstract.

In this work, the hydrogen storage materials of Ca, Fe, and Ti deposited NaSi20 clusters were investigated utilizing DFT methods (B3LYP and M06-2X) combined with the 6-311++G(d, p) basis set. The results show that Ca, Fe, and Ti atoms tend to bind with two adjacent Si atoms. The Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 can adsorb up to three, four, and six hydrogen molecules, respectively. The adsorption energy (E ads) per hydrogen molecule meets the United States Department of Energy (DOE) target for hydrogen storage materials for nH2-Ti@NaSi20 (n = 2-6) and nH2-Fe@NaSi20 (n = 1- 4), implying that NaSi20 fullerene could be a potentially suitable material for hydrogen storage.

Keywords: NaSi20 fullerenes; DFT methods; hydrogen storage

Resumen.

Utilizando métodos de la DFT (B3LYP y M06-2X) combinados con las bases 6-311++G(d, p), en este trabajo se investigaron materiales para el almacenamiento de hidrógeno a base de Ca, Fe, y Ti depositados en cúmulos de NaSi20. Los resultados muestran que los átomos de Ca, Fe, y Ti tienden a unirse a dos átomos adyacentes de Si. Los cúmulos Ca@NaSi20, Fe@NaSi20, y Ti@NaSi20 pueden adsorber hasta tres, cuatro y seis moléculas de hidrógeno, respectivamente. Las energías de adsorción por molécula de hidrógeno (E ads) de nH2-Ti@NaSi20 (n = 2-6) y nH2-Fe@NaSi20 (n = 1- 4) cumplen con el objetivo del Departamento de Energía de los Estados Unidos (DOE) lo que implica que el fullereno NaSi20 puede ser un material potencialmente adecuado para el almacenamiento de hidrógeno.

Palabras clave: NaSi20 fullerenos; métodos de la DFT; almacenamiento de hidrógeno

Introduction

With the growth of the world population and rapid economic development, as well as the shortage of fossil fuels and global environmental problems that threaten the peaceful existence of mankind and stimulate people's determination to find new energy sources [1,2]. Hydrogen as a renewable and clean energy source has several ideal characteristics, for example, it contains the maximum energy per unit weight, is abundant in the natural environment, and its combustion products do not pollute the environment like sulfur oxidation and aromatic hydrocarbon oxides [3,4]. The U. S. Department of Energy (DOE) system has identified goals for hydrogen storage materials, the average adsorption energy per H2 (average adsorption energy per H2, E ad/H2) in an ideal hydrogen storage material should be between chemisorption and physical adsorption (0.1 - 0.8 eV) to meet the hydrogen storage reversibility of the material [5].

Currently, there are many types of hydrogen storage materials. Carbon-based and chemically active hydrogen storage materials include fullerene, graphene, carbon nanotubes, etc [6,7]. Nano-structured magnesium and magnesium-based hydrides can also be used to store hydrogen through adsorption [8]. In recent years, the semiconductor cluster has become one of the research objects in cluster science and the silicon cluster is the most widely used technology in the semiconductor cluster [9-13].

Silicon is not only rich in content but is also one of the most important materials in the modern industry [11-13]. It has a wide range of applications in various areas of life, such as metallurgy, electronic manufacturing, the military industry, medical, etc [10-12]. But at the same time, the disadvantages of silicon materials are also very obvious. For example, the hollow structure of silicon and the lack of unsaturated bonds in silicon valence electrons with sp 2 hybridization deliver unstable fullerene cages (silicon nanotubes and silicon fullerenes) [10]. However, studies over the years have shown that their structures can be improved by various chemical modifications, such as encapsulating metal atoms in clusters [11] or using H adsorption to enhance the sp 3 bonding between Si atoms to improve stability [12]. Meanwhile, Ryou et al. found that the binding energy of hydrogen molecules to silicon nanotubes is less than 0.1 eV even in stable structures through DFT calculations, and then pure silicon nanotubes are not suitable for hydrogen storage [13]. Kumar et al. found that metal doping enhanced the stability and size selectivity of silicon clusters [14]. Sporea et al. stabilized silicon clusters by encapsulating alkali metal atoms Na, K, and Li inside Si20, which can provide regular shapes for Si20 clusters [15]. Moreover, Ammar et al. investigated the hydrogen storage properties of KSi20 and Ti@KSi20 using the DFT-based B3YLP and M06-2X methods. The adsorption energy values ( ads) per hydrogen molecule satisfy the U.S. DOE targets for hydrogen storage materials [16].

According to our knowledge, NaSi20 should have similar hydrogen storage properties to KSi20. However, the hydrogen storage performance of NaSi20 has not been reported. And more importantly, as previously reported by researchers, Ca, Fe, and Ti metal atoms were often considered and performed well in the previous studies of modified carbon-based and boron-based hydrogen storage materials [17-19]. Meanwhile, Ca is one of the most active atoms in alkaline earth metals, Fe is the transition metal atom with more outer electrons, and Ti is the transition metal atom with fewer outer electrons. To the best of our knowledge, NaSi20 deposited with Ca, Fe, and Ti atoms has not been previously investigated as hydrogen storage materials. In this work, the storage characteristics of H2 on Na-encapsulated Si20 (NaSi20) clusters deposited in Ca, Fe, and Ti were investigated by performing DFT calculations at the theoretical level of B3LYP and M06-2X combined with the 6-311++G(d, p) basis set. The effect of electron properties on NaSi20 fullerenes deposited in Ti, Ca, and Fe was studied. Then the adsorption of n hydrogen molecules (n = 1-6) on Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 is discussed.

Computational methods

To investigate the storage properties of H2 molecules on Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 fullerenes, DFT calculations [20] were performed at the B3LYP/6-311++G(d, p) and M06-2X/6-311++G(d, p) levels of theory for the non-metallic atoms (H, Si) and an effective for the effective pseudopotential basis set LANL2DZ of Na, Ca, Fe and Ti atoms, respectively [20]. Full geometric optimizations were carried out at two levels for free H2 molecules, Si20, NaSi20, Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 substrates as well as nH2-Ca@NaSi20 (n = 1-3), nH2-Fe@NaSi20 (n = 1-4), and nH2-Ti@NaSi20 (n = 1-6) complexes. Because silicon is the most readily available semiconductor in nature, hydrogen can be adsorbed on silicon in both crystalline and amorphous forms and can be desorbed again. Several experimental and theoretical studies have been conducted by Mao WL and Mao HK to investigate whether porous silicon is capable of conducting hydrogen storage [21]. All the calculations were calculated using the Gaussian 09 program [22].

The ionization potential (IP), electron affinity (EA), and chemical hardness (η) can be expressed in equations 1-3, where E(N-1), E(N), and E(N+1) are total energies when the system has N-1, N and N+1 electrons, respectively.

IP=E(N1) E(N) (1)

EA=E(N) E(N+1) (2)

η12(IP-EA) (3)

The average binding energy (E¯b) of per atom for Si20, NaSi20, Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 is calculated by equation (4),

E¯b=Ecluster-Eatomn (4)

where Ecluster is the energy of the optimized cluster, ∑E atom is the sum of the energies of the free atoms of the cluster and n is total number of the cluster atoms. The binding energies (E bind) of Ca, Fe, and Ti atoms on NaSi20 substrate clusters are calculated by equation (5). nH2 molecule (n = 1-6) adsorption energies (E ads) and the average adsorption energy of per hydrogen molecule (E¯ads) on M@NaSi20 fullerenes (M = Ca, Fe and Ti) are calculated by equations (6) and (7), respectively.

Ebind=EM@NaSi20-(E+NaSi20EM) (5)

Eads=EnH2-M@NaSi20-(nEH2+EM@NaSi20) (6)

E¯ads=1n[EnH2-M@NaSi20-(nEH2+EM@NaSi20)] (7)

where EM@NaSi20 is the energy of the optimized Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 complexes, EM@NaSi20 is the energy of the corresponding optimized NaSi20 substrate clusters, E M is the atomic energy of the free M (M = Ca, Fe, and Ti) atoms, and EnH2-M@NaSi20 is the total energy of the optimized nH2-M@NaSi20 (M = Ca, Fe, and Ti).

The enthalpy difference (ΔH Ɵ) and free energy difference (ΔG Ɵ) of nH2-M@NaSi20 (M = Ca, Fe, and Ti) complexes are calculated by equations (8) and (9), respectively [23].

ΔHΘ=[HΘnH2-M@NaSi20-(nHΘH2+HΘM@NaSi20)]/n (8)

where, HΘnH2-M@NaSi20, HΘM@NaSi20, and HΘH2 are the enthalpies for nH2-M@NaSi20, M@NaSi20 and H2 molecules, respectively.

ΔGΘ=[GΘnH2-M@NaSi20-(nGΘH2+GΘM@NaSi20)]/n (9)

where GΘnH2-M@NaSi20, GΘM@NaSi20 and GΘH2 are the free energies of nH2-M@NaSi20, M@NaSi20, and H2 molecules, respectively.

For testing the stability of the wave function of our investigated systems, the keywords of UB3LYP, stable and nosymm are used in the examples for both 4H2/Ti@NaSi20 and 4H2/Fe@NaSi20 complexes and the results show that the wave functions are stable under the perturbations considered. The corresponding figure is given in Supplement 1.

Results and discussion

Geometric structure and electronic properties of Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters

Fig. 1 shows the optimized structures of Si20 and NaSi20 fullerenes, as well as Ca, Fe, and Ti external doping NaSi20 fullerenes, named Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters. The geometrical and electronic properties of the investigated clusters are listed in Table 1. As can be seen from Fig. 1(a), the Si20 fullerene cluster is a distorted cage with the shortest Si-Si bond length (d Si-Si) of 2.279 Å and dipole moments of 1.198 Debye at the B3LYP level, and the shortest Si-Si bond length (d Si-Si) of 2.291 Å and dipole moments of 1.215 Debye at the M06-2X level. This is similar to the value of Si-Si bond length (d Si-Si) calculated by Ammar et al., which is 2.335 Å for Si-Si bond length (d Si-Si) calculated by B3LYP method and 2.381 Å for Si-Si bond length (d Si-Si) calculated by M06-2X method. After a Na atom is embedded in the Si20 cluster, the structure of the cluster has changed into a regular shape, as shown in Fig. 1(b). Compared with pure silicon clusters without internally doped metal atoms, the shape of NaSi20 clusters is more uniform. Under the B3LYP method, the Si-Si bond length (d Si-Si) is reduced to 2.777 Å, which is similar to the results previously reported by Borshch et al. [24]. The HOMO-LUMO gap (E g) decreased from 2.163 and 3.486 eV to 1.329 and 2.512 eV, respectively. Chemical hardness (η) decreased by 38.63 % and 27.94 %, respectively, while the average binding energy between atoms (E¯b) increased by 4.5 % and 5.2 %.

Table 1 Ionization potential (IP, eV), electron affinity (EA, eV), hardness (η, eV), HOMO and LUMO energy gap (E g, eV), binding energy per atom (E b, eV), dipole moment (D, Debye), bond length (d, Å), binding energy (E bind, eV) of Si20, NaSi20, Ca@NaSi20, Fe@NaSi20 and Ti@NaSi20 as well as Mulliken charges (QM, e) of Ca, Fe, Ti atoms in the clusters. 

Si20 NaSi20 Ca@NaSi20 Fe@NaSi20 Ti@NaSi20
B3LYP M06-2X B3LYP M06-2X B3LYP M06-2X B3LYP M06-2X B3LYP M06-2X
IP 5.780 6.581 5.255 5.922 4.761 5.485 5.372 5.656 5.209 5.391
EA 3.616 3.095 3.926 3.410 3.776 3.303 3.911 3.831 3.795 3.432
η 1.082 1.743 0.664 1.256 0.493 1.091 0.730 0.913 0.707 0.979
E g 2.163 3.486 1.329 2.512 0.986 2.182 1.460 1.825 1.414 1.959
E b -4.285 -4.540 -4.091 -4.306 -3.955 -4.155 -4.137 -4.286 -4.103 -4.250
D 1.198 1.215 0.163 0.259 11.843 11.489 2.966 1.597 8.502 9.983
dSi-Si/dM-Si 2.279 - 2.777 - 3.013 - 2.367 - 2.502 -
Ebind -1.091 -0.992 -5.109 -3.877 -4.348 -3.069
QM 1.266 1.042 -0.184 -0.123 0.398 0.466

We studied the external doping of Ca, Fe, and Ti atoms at different positions of NaSi20 fullerene. The structure with the most stable energy after optimization is shown in Fig. 1. The results show that Ca, Fe, and Ti are bound to the NaSi20 cluster by bonding with two adjacent silicon atoms, which is the same as the external doping mode of Ti atoms to the KSi20 cluster in the cluster studied by Ammar et al. [14].

According to Baei's theory, the size of the energy gap (E g)value is closely related to the activity and stability of chemical reactions [25]. Under the density functional B3LYP method, the energy gap (E g) of the NaSi20 cluster modified by Fe and Ti atoms is larger than that of the NaSi20 cluster modified by Ca atoms, while the energy gap (E g) of the NaSi20 cluster modified by Ca atoms is smaller than that of the NaSi20 cluster. Similarly, with the B3LYP method, the chemical hardness (η) of the Fe@NaSi20 cluster and Ti@NaSi20 cluster increases by 9.93 % and 6.48 %, respectively, while the chemical hardness (η) of the Ca@NaSi20 cluster decreases by 25.75 %. With the M06-2X method, the Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters have smaller energy gaps than the NaSi20 clusters. The Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters also have smaller η values than the NaSi20 clusters. Under the M06-2X method, η values of Ca@NaSi20, Fe@NaSi20 and Ti@NaSi20 are reduced by 22.05 %, 27.31 % and 13.14%, respectively. Since the density functional M06-2X method considers the weak interaction, it can well describe the weak interaction between metals and clusters. At the same time, the M06-2X method is effective in calculating the energy of the reactants, the isomerization process and the reaction energy barrier. The results showed that the chemical reactivity of Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters was better than that of NaSi20 clusters. Compared with NaSi20 clusters, the average binding energy (E¯b) of NaSi20 clusters modified by Ca, Fe, and Ti atoms increases by about 3.51 %, 0.47 %, and 1.31 %, respectively, under M06-2X method. However, the average binding energy (E¯b) of NaSi20 clusters modified by Fe and Ti atoms calculated by the B3LYP method decreases by 1.12 % and 1.20 %, respectively, and for Ca@NaSi20, the mean binding energy (E¯b) increases by about 3.61 %. This is consistent with the changing trend of the binding energy of C20 and KSi20 clusters modified by Ti atoms reported by Ammar et al. [12]. It can be seen that the external doping of Ca, Fe and Ti atoms on the NaSi20 cluster reduces the stability of NaSi20 and increases the activity of the reaction.

Fig. 1 The optimized structures of (a) Si20, (b) NaSi20, (c) Ca@NaSi20, (d) Fe@NaSi20 and (e) Ti@NaSi20 calculated at B3LYP method. 

Interaction of H2 on Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20

In this work, hydrogen molecules are stored on optimized Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters to characterize the properties of the selected hydrogen storage materials. Fig. 2 shows the geometry of nH2-Ca@NaSi20, nH2-Fe@NaSi20, and nH2-Ti@NaSi20 clusters and in order to explore the place of adsorption more clearly, the structure diagrams from different viewpoints for 4H2-Fe@NaSi2 and 6H2-Ti@NaSi20 as well as different direction of hydrogen attacking for Ti@NaSi20 and Fe@NaSi20 are given in Fig. 3. As can be seen from Fig. 2, the H-H bonds of all H2 molecules are not broken, and they are all adsorbed by metal atoms in the form of molecules. Parameters such as dipole moment, bond length and Mulliken charge of adsorbed clusters are shown in Tables 2-4, respectively. Equations (6) and (7) were used to calculate the adsorption energy of hydrogen storage materials and the average adsorption energy data of each hydrogen molecule, as shown in Fig. 4(a) and 4(b), respectively. Density functional methods of B3LYP and M06-2X were used to optimize the structure of the H2 molecule, and the results showed that the H-H bond lengths were 0.744 Å and 0.740 Å, respectively.

Fig. 2 The optimized structures for nH2-Ca@NaSi20, nH2-Fe@NaSi20 and nH2-Ti@NaSi20 calculated at B3LYP method. 

Fig. 3 (a) The adsorption energy (E ads) and (b) the average adsorption energy per hydrogen molecule (E¯ads) for nH2-Ca@NaSi20, nH2-Fe@NaSi20 and nH2-Ti@NaSi20 at B3LYP and M06-2X methods. 

Fig. 4 The energy gap (E g) for nH2-Ca@NaSi20, nH2-Fe@NaSi20 and nH2-Ti@NaSi20 at M06-2X method. 

Adsorption of H2 on Ca@NaSi20

This section focuses on the interaction between H2 and Ca@NaSi20 clusters. As can be seen from Table 2, under the B3LYP and M06-2X methods, the interaction between a single H2 molecule and the Ca@NaSi20 cluster results in an increase of 0.023 eV and a decrease of 0.040 eV, respectively. According to Mulliken charge analysis, it can be found that the H2 molecule (QH2) gains a negative charge of 0.017e, the positive charge (QCa) on the Ca atom decreases by 0.246e, and the charge (QNa) on Na atom remains basically unchanged under the B3LYP method. This indicates that there is a transfer of charge between the Ca atom and the H2 molecule. The average adsorption energy of nH2-Ca@NaSi20 cluster is in the range of 0.019 - 0.024 eV, and the average adsorption the energy of the third H2 molecule is not different from that of the second H2 molecule. Under the density functional method of M06-2X, the adsorption energy (E ads) increases from -0.040 eV to - 0.014 eV, and the absolute value of the adsorption energy (E ads) decreases gradually when the H2 molecules adsorbed by Ca@NaSi20 cluster increase from the first H2 molecule to the third H2 molecule. As can be seen from Table 2, under the B3LYP method, when the second H2 molecule is adsorbed by Ca@NaSi20 cluster, the Ca-Si atomic bond length (dCa-Si) is 3.009 Å, and this Ca-Si atomic bond length (dCa-Si) is the same as the third H2 adsorbed by the Ca@NaSi20 cluster. Meanwhile, as shown in Table 2, when the third H2 molecule is adsorbed by the Ca@NaSi20 cluster, the distance (dCa-H) between the H2 molecule and the Ca@NaSi20 cluster increases to 6.259 Å, and the H-H bond length (dH-H) of the adsorbed third H2 molecule is the same as that of a single H2 molecule, both being 0.744 Å, that is, Ca@NaSi20 cluster did not complete the adsorption of the third H2 molecule. From the energy range analysis, it can also be seen from Table 2 and Fig. 4 that for the nH2-Ca@NaSi20 cluster, the absolute value of adsorption energy (E ads) of the Ca@NaSi20 cluster also increases with the increase of the H2 molecule. When the Ca@NaSi20 cluster adsorbs the third H2 molecule, the absolute value of the adsorption energy (E ads) is less than that of the second H2 molecule (E ads), which means that the third H2 molecule is not adsorbed by the Ca@NaSi20 cluster, indicating that the Ca@NaSi20 cluster is saturated at the adsorption of the second H2 molecule. Therefore, although Ca doped NaSi20 clusters can effectively adsorb 2 H2 molecules, the hydrogen storage capacity is weak, which is lower than the minimum value of reversible adsorption 0.1eV, and the system is unstable after adsorption, so it is not a relatively ideal hydrogen storage material.

Table 2 The dipole moment (D, Debye), bond length (d, Å), adsorption energy (E ads, eV) and Mulliken charges (Q, e) of Ca@NaSi20

D dCa-Si dCa-H dH-H E ads E¯ads QNa QCa QH2
B3LYP B3LYP B3LYP B3LYP B3LYP M06-2X B3LYP M06-2X B3LYP B3LYP B3LYP
H2-Ca@NaSi20 11.881 3.012 4.054 0.745 0.023 -0.040 0.023 -0.040 -0.311 1.020 -0.017
2H2-Ca@NaSi20 11.816 3.009 4.133 0.745 0.048 -0.041 0.024 -0.021 -0.323 1.029 0.005
5.552 0.745 -0.019
3H2-Ca@NaSi20 11.857 3.009 4.107 0.745 0.057 -0.042 0.019 -0.014 -0.321 1.030 0.005
5.411 0.745 -0.021
6.259 0.744 -0.001

Adsorption of H2 on Fe@NaSi20

This section concerns the interaction of the H2 molecules with the Fe@NaSi20 cluster. As can be seen from Table 3, the bond length of the H2 molecule increases from 0.744 Å to 0.802 Å for a single H2 molecule after adsorption, increasing by 0.058 Å. After the adsorption of one H2 molecule, the bond length of the Si-Fe atom (dFe-Si) in the Fe@NaSi20 cluster is 2.320 Å under the B3LYP method, which is 0.047 Å shorter than the Fe@NaSi20 cluster without adsorption of the H2 molecule. The increasing of hydrogen molecular bond length (dH-H) and the shortening of Si-Fe atom bond length (dFe-Si) indicate that the H2 molecule is successfully adsorbed by the Fe@NaSi20 cluster. In terms of energy range, when the number of adsorbed H2 molecules increases from 1 to 4, the absolute value of adsorption energy (E ads) calculated by the M06-2X method gradually increases from 0.291 eV to 0.508 eV, and the absolute value of average adsorption energy (E¯ads) gradually decreases from 0.291 eV to 0.127 eV. The average absolute adsorption energy of all hydrogen molecules adsorbed by clusters is greater than 0.1eV, which is within the reversible adsorption energy range of 0.1 - 0.8eV. Meanwhile, as can be seen from Fig. 5, the absolute value of the energy gap (E g) increases with the increase in the number of H2 molecules adsorbed by the Fe@NaSi20 cluster. When the fourth H2 molecule is adsorbed, the energy gap (E g) reaches saturation. When the fifth H2 molecule is adsorbed by the Fe@NaSi20 cluster, the absolute value of the energy gap (E g) becomes smaller and increases from -2.689 eV to -1.348 eV. It can be concluded that the maximum number of H2 molecules adsorbed by Fe@NaSi20 is 4. From the discussion of the energy gap, adsorption energy and average adsorption energy, it can be seen that the Fe@NaSi20 cluster cannot effectively adsorb the fifth H2 molecule. As can be seen from the Mulliken charge QFe of the Fe atom in Table 3, the negative charge density of Fe atom increases gradually during the adsorption of the H2 molecule, while the charge density QNa of the metal atom remains basically unchanged, and the charge density reaches saturation when the fourth H2 molecule is adsorbed. Therefore, it can be seen that the Fe@NaSi20 cluster increases the negative charge density in the outermost orbital of the Fe atom through charge transfer between the Fe atom and H2 molecule in the adsorption process of H2, so that Fe atom cannot store the fifth H2 molecule, and the main hydrogen storage role in the system is determined by Fe atom. Therefore, Fe modified NaSi20 cluster has strong hydrogen storage performance. Fe atoms in the Fe@NaSi20 cluster can store up to 4 H2 molecules, and the absolute value of average adsorption energy is within 0.1 - 0.8 eV. Fe@NaSi20 cluster can realize a reversible cycle of hydrogen storage performance. Compared with the calculation results of the Ca@NaSi20 cluster, Fe@NaSi20 has better hydrogen storage performance.

Table 3 The dipole moment (D, Debye), bond length (d, Å), adsorption energy (E ads, eV) and Mulliken charges (Q, e) of Fe@NaSi20

D dFe-Si dFe-H dH-H E ads E¯ads QNa QFe QH2
B3LYP B3LYP B3LYP B3LYP B3LYP M06-2X B3LYP M06-2X B3LYP B3LYP B3LYP
H2-Fe@NaSi20 3.359 2.320 1.771 0.802 -0.175 -0.291 -0.175 -0.291 -0.408 -0.298 0.119
2H2-Fe@NaSi20 3.397 2.260 1.629 0.851 0.156 -0.345 0.078 -0.172 -0.346 -0.434 0.296
1.635 0.845 0.265
3H2-Fe@NaSi20 3.020 2.275 1.637 0.838 0.076 -0.374 0.025 -0.125 -0.344 -0.614 0.250
1.635 0.842 0.230
1.756 0.796 0.126
4H2-Fe@NaSi20 3.083 2.289 1.636 0.843 0.067 -0.508 0.017 -0.127 -0.368 -0.630 0.257
1.637 0.839 0.255
1.769 0.793 0.005
3.387 0.744 -0.023

Fig. 5 Frontier molecular orbitals of 4H2-Fe@NaSi20 (the top row) and 4H2-Ti@NaSi20 (the bottom row). 

Adsorption of H2 on Ti@NaSi20

This section discusses the interaction of H2 molecules with Ti@NaSi20 clusters. Table 4 shows the geometric parameters of the optimized nH2-Ti@NaSi20(n = 6) cluster. Under the B3LYP method, the bond length (dH-H) of the adsorbed H2 molecule increases from 0.744 Å to 0.846 Å, and as can be seen from Fig. 2, there is no break in the hydrogen bond of all H2 molecules adsorbed by nH2-Ti@NaSi20(n = 6) system. It indicates that H2 molecules are adsorbed on the cluster in molecular form. The Ti-Si atomic bond length (dTi-Si) increases with the number of H2 adsorbed. When the sixth hydrogen molecule is adsorbed, the maximum value of the Ti-Si atomic bond length (dTi-Si) is 2.656 Å. When the seventh molecule is adsorbed by the Ti@NaSi20 cluster, the Ti-Si bond is broken. That is, the Ti-Si atomic bond length (dTi-Si) reaches the maximum value when the sixth H2 molecule is adsorbed. Therefore, the Ti@NaSi20 system can adsorb up to 6 H2 molecules according to the bond length analysis. Under the B3LYP method, the dipole moment of the H2-Ti@NaSi20 system is 0.250 Debye smaller than that of Ti@NaSi20, and under the M06-2X method the dipole moment of the H2-Ti@NaSi20 system is 1.022 Debye smaller than that of Ti@NaSi20, with the decrease of dipole moment the stability of the clusters is decreased and the activity is increased. Meanwhile, Mulliken charge analysis results show that from the analysis of the charge QTi on the Ti atom, due to the charge transfer and redistribution around the metal atoms and H2 molecules, compared with the Ti@NaSi20 cluster, the positive charge of the Ti atom in H2-Ti@NaSi20 system decreases by 0.045 e, and QTi gradually decreases as the number of H2 molecules adsorbs increases, and reaches the minimum value at the adsorption of the sixth H2 molecule. In addition, Ti atoms have fewer electrons in their outermost shell and can accept more electrons than Fe atoms, which is one reason why Ti@NaSi20 can attach more H2 molecules than Fe@NaSi20 clusters. The absolute values of the adsorption energy (E ads) of the first H2 molecule adsorbed by the Ti@NaSi20 cluster were 0.185 and 0.925 eV, respectively, calculated by the B3LYP and M06-2X methods. With the number of adsorbed H2 molecules gradually increasing to 6, the absolute value of adsorption energy (E ads) first increases and then decreases, and the absolute value of average adsorption energy (E¯ads) first decreases, then increases and subsequently decreases with the increase of the adsorbed H2 molecules, presenting an irregular change, which is related to the types of doped atoms and the types of atoms covered by silicon clusters [1]. But it's within the normal range. When the Ti@NaSi20 cluster adsorbs the fourth hydrogen molecule, the absolute values of adsorption energy (E ads) and average adsorption energy (E¯ads) reach the maximum value under the calculation of B3LYP and M06-2X, and the absolute values of adsorption energy (E ads) are 0.512 eV and 2.091 eV, respectively. The absolute values of average adsorption energy (E¯ads) are 0.128eV and 0.523eV, respectively. In this case, the Ti@NaSi20 cluster has the best adsorption effect on H2 molecules, the system is the most stable and can realize reversible storage of H2 molecules. The energy gap (E g) of the system is observed to change. The energy gap (E g) and mean adsorption energy (E¯ads) of the system change in the same trend, reaching the maximum value when the fourth H2 molecule is adsorbed, and the value is basically unchanged when the sixth H2 molecule is adsorbed. In summary, the analysis of charge density, adsorption energy, mean adsorption energy, and energy gap of the Ti@NaSi20 cluster shows that the Ti@NaSi20 cluster reaches the maximum value of mean adsorption energy (E¯ads) when it adsorbs 4 H2 molecules, and the saturation state of H2 molecular adsorption when it adsorbs 6 H2 molecules. In addition, under the calculation of M06-2X, the mean adsorption energy (E¯ads) of Ti@NaSi20 clusters for each hydrogen molecule is within the range of 0.1- 0.8 eV, which can realize reversible adsorption of hydrogen. It is also proved that Ti@NaSi20 clusters have a better storage effect on H2 molecules than Ca@NaSi20 and Fe@NaSi20 clusters.

Table 4 The dipole moment (D, Debye), bond length (d, Å), adsorption energy (E ads, eV) and Mulliken charges (Q, e) of Ti@NaSi20

D dTi-Si dTi-H dH-H Eads E¯ads QNa QTi QH2
B3LYP B3LYP B3LYP B3LYP B3LYP M06-2X B3LYP M06-2X B3LYP B3LYP B3LYP
H2-Ti@NaSi20 8.252 2.465 1.957 0.791 0.185 -0.925 0.185 -0.925 -0.344 0.353 -0.155
2H2-Ti@NaSi20 7.850 2.452 1.870 0.824 0.261 -0.966 0.130 -0.483 -0.347 0.297 -0.131
1.855 0.825 -0.118
3H2-Ti@NaSi20 6.133 2.470 1.856 0.826 -0.270 -0.694 -0.090 -0.231 -0.353 0.252 -0.026
1.855 0.840 -0.007
1.866 0.826 0.095
4H2-Ti@NaSi20 9.268 2.546 2.020 0.776 -0.512 -2.091 -0.128 -0.523 -0.366 0.279 -0.036
2.120 0.768 -0.054
2.028 0.775 0.094
1.980 0.798 0.305
5H2-Ti@NaSi20 6.426 2.535 1.912 0.814 -0.413 -1.552 -0.083 -0.311 -0.404 0.220 -0.021
1.907 0.816 0.008
2.014 0.793 0.101
1.875 0.818 0.149
1.844 0.836 0.190
6H2-Ti@NaSi20 7.089 2.656 1.818 0.846 -0.279 -1.580 -0.047 -0.263 -0.390 0.240 -0.020
1.916 0.811 0.041
1.970 0.803 0.103
1.917 0.803 0.204
1.995 0.796 0.191
1.921 0.810 0.161

Orbital analysis

The two systems with the best adsorption performance were selected for the wave function analysis. Fig. 6 shows the distribution of the highest occupied molecular orbitals (HOMO) and the lowest unoccupied molecular orbitals (LUMO) in the Fe@NaSi20, Ti@NaSi20, 4H2-Fe@NaSi20 and nH2-Ti@NaSi20 (n=3-6) clusters, whose analysis can be used to guide the design and characterization of novel functional hydrogen storage fullerenes. The active points are the regions of high local charge density, in which the red and green colors represent the positive and negative wave functions, respectively. As can be seen from Fig. 6, in the Fe@NaSi20 cluster the electron cloud contribution of the Fe atoms in HOMO-LUMO is almost zero. But after the cluster adsorbed 4 H2 molecules, the electron cloud around Fe atom in the 4H2-Fe@NaSi20 cluster increased and HOMO distribution increased, which represents charge transfer between Fe atoms and H2 molecules. This is consistent with the previous charge analysis of QFe and QH2. For the Ti@NaSi20 cluster, HOMO and LUMO are partially distributed on Ti and partially distributed on the NaSi20 cluster when there is no H2 molecule adsorbed. With the increase of H2 molecules from 3 to 6, the distribution of frontier orbitals on Ti gradually decreases. When four H2 molecules are adsorbed, there are a few unoccupied orbitals around Ti atoms, indicating that they can still adsorb H2 molecules, which is consistent with the previous analysis trend.

Fig. 6 (a) Enthalpy difference (ΔH Ɵ) and (b) free energy difference (ΔG Ɵ) for nH2/Ca@NaSi20, nH2/Fe@NaSi20 and nH2/Ti@NaSi20 calculated at M06-2X method. 

Adsorbed H2 prediction

Theoretically, the Effective Atomic Number Rule (EAN) can be used to calculate the maximum adsorption capacity of hydrogen atoms of hydrogen storage material with surface doped metal atoms [26]. The equation is as follows,

nH=18nvensenme (10)

where nve is the number of valence electrons of the metal atom, nse is the bonding number of the metal to the substrate, and nme is the bonding number between neighboring metal atoms. Considering the 3d orbitals of transition metal atoms Ti and Fe, the valence electron numbers (nve) of Ca, Fe, and Ti atoms are 2, 8, and 4, respectively. Since M is adsorbed above the Si-Si bond, nse = 2, the n H of M@NaSi20 (M = Ca, Fe, and Ti) is calculated to be 14, 8, and 12, respectively, which means that the maximum adsorption numbers of hydrogen molecules are 7, 4, and 6. Compared with the calculated results, it can be seen that the hydrogen storage performance of Ca@NaSi20 is less than the theoretical results, so the hydrogen storage performance is poor. The maximum hydrogen storage number of M@NaSi20 (M = Fe and Ti) is in accordance with the theoretical expectation, i.e., the number of H2 molecules that can be adsorbed is in accordance with EAN. At the same time, we tried to adsorb the third, 5th, and 7th H2 molecules on Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20, respectively. None of the results calculated by the density functional method can converge, which further indicates that Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 can adsorb at most 2, 4, and 6 H2 molecules.

Thermodynamic analysis

The enthalpy difference (ΔH Ɵ) can express the strength of the interaction and the free energy difference (ΔG Ɵ) is a measure of the spontaneity of the reaction, which are the key thermodynamic parameters for the interaction between hydrogen and the sorbent materials. ΔH Ɵ and ΔG Ɵ are graphed in Fig. 7 as a function of the number of hydrogen molecules for nH2-Ca@NaSi20, nH2-Fe@NaSi20, and nH2-Ti@NaSi20 complexes. As shown in Fig. 7a, ΔH Ɵ values are negative for all the considered structures, indicating exothermic reactions. For all the complexes of the nH2-Ti@NaSi20 system only the free energy difference of 1H2-Ti@NaSi20, 2H2-Ti@NaSi20, 4H2-Ti@NaSi20, 5H2-Ti@NaSi20 is negative (ΔG Ɵ < 0), which shows that they are spontaneous reactions. In the nH2-Ti@NaSi20 system, when n = 1 the absolute value of ΔHƟ decreases as the number of H2 molecules increases, and the absolute value of ΔH Ɵ is 20.45 kcal/mol. It can be seen from Fig. 7b, in all the complexes, 3H2-Ca@NaSi20 has the largest value of free energy difference (ΔG Ɵ), which is 10.50 kcal/mol. As a negative ΔH Ɵ value indicates stronger stability, and a negative ΔG Ɵ value indicates stronger reversal ability of the reaction, with the increase of the number of hydrogen molecules, the stability decreases, the hydrogen storage capacity gradually reaches saturation, and the desorption ability of H2 molecules increases [27-29].

Conclusions

In this work, the storage characteristics of H2 molecules on NaSi20 fullerenes deposited by Ca, Fe, and Ti were investigated at the theoretical level of the DFT-based B3LYP and M06-2X methods. The results show that the encapsulated Na atom into the Si20 cluster delivers the regular shape to the NaSi20. The deposition of the Ca, Fe and Ti atoms on NaSi20 clusters decreases the ionization potential (IP), HOMO-LUMO energy gap (E g), and hardness (η), and increases the dipole moment (D), which confirms that Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters are less stable and therefore more reactive than NaSi20 clusters.

Additionally, the Ca@NaSi20, Fe@NaSi20, and Ti@NaSi20 clusters are saturated by two, four, and six H2 molecules, respectively. The adsorption energy values (E ads) per hydrogen molecule meet the U. S. DOE target for hydrogen storage materials for nH2-Ti@NaSi20 (n = 2 - 6) and nH2-Fe@NaSi20 (n = 1-4), which implies that NaSi20 fullerenes may be a potentially suitable material for hydrogen storage. The calculated enthalpy differences emphasize that some hydrogen molecules are physisorbed on the studied clusters with an effect greater than one.

Acknowledgements

This study was financially supported by the Natural Science Basic Research Plan in Shaanxi Province of China (Program No. 2022JM-075) and the Three-year action plan project of Xi'an University (Program No. 21XJZZ0001-11) as well as the Youth Innovation Team of Shaanxi Universities (Environmental Pollution Monitoring and Control Innovation Team, 51).

References

1. Hoel, M.; Kverndok, K. Resour Energy Econ. 1996, 18, 115-136. [ Links ]

2. Kale, P. Int J Hydrogen Energy. 2012, 37, 3741-3747. [ Links ]

3. Bailleux, C. Int J Hydrogen Energy. 1981, 6, 461. [ Links ]

4. Mazloomi, K. Renew Sust Energ Rev. 2012, 16, 3024-3033. [ Links ]

5. Anafcheh, M. Int J Hydrogen Energy. 2018, 43, 12271-12277. [ Links ]

6. Xu, W. Prog Chem. 2006, 18, 200-201. [ Links ]

7. Gong, J. M. Natural Gas Chem. Indus. 2010, 5, 71. [ Links ]

8. Sahaym, U. J Mater Sci. 2008, 43, 5395. [ Links ]

9. Wang, Q.; Sun, Q.; Jena, P.; Kawazoe, Y. J Chem Theory Comput. 2009, 5, 374-379. [ Links ]

10. Zhang, D. J.; Ma, C.; Liu, C. B. J Phys Chem C. 2007, 111, 17099. [ Links ]

11. Khanna, S. N.; Rao, B. K.; Jena, P. Phys Rev Lett. 2002, 89, 016803. [ Links ]

12. Kumar, V.; Kawazoe, Y. Phys Rev Lett. 2003, 90, 055502. [ Links ]

13. Ryou, J.; Hong, S.; Kim, G. Solid State Commun. 2008, 148, 469-471. [ Links ]

14. Kumar, V.; Kawazoe, Y. Phys Rev B. 2007, 75, 155425. [ Links ]

15. Sporea, C.; Rabilloud, F. J Chem Phys. 2007, 127, 164306. [ Links ]

16. Ammar, H. Y.; Badran, H. M. Int J Hydrogen Energy. 2021, 46, 14565-14580. [ Links ]

17. Williamson, A. J.; Reboredo, F. A.; Galli, G. Appl Phys Lett. 2004, 85, 2917-2919. [ Links ]

18. Yoon, M.; Yang, S.Y.; Hicke, C.; Wang, E. Phys Rev Lett. 2008, 100, 206806. [ Links ]

19. Wang, C. J.; Tang, C. M.; Zhang, Y. J.; Gao, F. Z. Chem Res Chin Univ. 2014, 35, 2131-2137. [ Links ]

20. Kohn, W.; Sham, L. J. Phys Rev. 1965, 140, 1133-1138. [ Links ]

21. Mao, W. L.; Mao, H. K. Proc Natl Acad Sci USA. 2004, 101, 708-710. [ Links ]

22. Frisch, M. J. Gaussian 09, revision D.01, Wallingford, CT: Gaussian, Inc, 2009. [ Links ]

23. Guo, C.; Wang, C. Int J Hydrogen Energy. 2019, 44, 10763-10769. [ Links ]

24. Borshch, N. A.; Pereslavtseva, N. S.; Kurganskiĭ, S. I. Semiconductors. 2006, 40, 1457-1462. [ Links ]

25. Baei, M. T.; Koohi, M.; Shariati, M. J Mol Struct. 2018, 1159, 118-134. [ Links ]

26. Kubas, G. J. Kluwer Academic/Plenum Publishing. 2001. [ Links ]

27. Padhee, S. P.; Roy, A.; Pati, S. Int J Hydrogen Energy. 2021, 46, 906-921. [ Links ]

28. Liu, Z. C.; Ruan, Y. M.; Li, F.; Zhang, G. F.; Zhao, M. Int J Hydrogen Energy. 2021, 46, 4201-4210. [ Links ]

29. Pluengphon, P.; Tsuppayakorn-aek, P.; Inceesungvorn, B. Int J Hydrogen Energy. 2020, 45, 25065-25074. [ Links ]

Received: May 30, 2023; Accepted: December 22, 2023; Published: April 26, 2024

*Corresponding author: Bin Liu; Hongjiang Ren, email: hjren@xawl.edu.cn

Creative Commons License This is an open-access article distributed under the terms of the Creative Commons Attribution License