The development of rechargeable batteries, particularly rechargeable lithium-ion battery (LIB), greatly promotes the application of mobile electronic devices and environmentally friendly vehicles^[1-5], which is subverting the human's life style and the way of energy application, initiating a new chapter of the new energy area. In order to further improve the performance of LIB, it is urgent to develop advanced electrode materials that provide rapid charging/ discharging rate, cyclic stability, satisfactory capacity, and safety^[6-12].

In the past several decades, graphite has been used as traditional anode material, but its energy capacity (372mAh·g^-1) and rate of Li are still far away from what are expected to meet the requirements of many applications^[13-14]. Recent years, more and more attentions were focused on two-dimensional (2D) layered materials with the characteristics of high specific surface area, large interlayer spacing, and other peculiar properties^[15-17]. For example, the well-studied graphene exhibits the unique capacity (774mAh·g^-1) because of its high charge carrier mobility, large surface area, and a broad electrochemical window^[18-22]. Up to date, graphene has been utilized as both cathode and anode materials with great success^[23-27]. MoS₂ has been explored to exhibit a high reversible lithium storage capacity and superior rate capability^[28-33]. Black phosphorus was shown to have a series of good characteristics as Li-battery electrode^[34-36]. With the decreasing of dimensionality, the 2D phosphorene has been validated as ideal anode material with a low energy barrier (0.08eV) of Li and higher average charging voltage (2.9V) than other 2D materials^[37-39]. In addition, the group Ⅳ-Ⅵ compounds MX (M=Ge, Sn; X=S or Se), which display puckered structures similar to black phosphorus and form buckled honeycomb lattices, has received much attention as electrode materials of LIB^[40-44]. Based on the interlayer spacing from 3.1 Å to 3.26 Å, their structural anisotropy directly leads to anisotropy of the migration direction of lithium ion. Sn-based materials, like SnS₂ and SnO₂ composites, provide ideal space for Li atom intercalation and exhibit high reversible capability and good cycling performances when they are used as anode materials^[45-48]. However, as far as we know, the properties and performance of SnSe₂ as electrode material have not been studied, while SnSe₂ presents prominent characteristic of large interlayer spacing up to 4.175 and hence provides more possibility for the adsorption and diffusion of the Li ions. In this work, the Li-ion adsorption properties and migration mechanism in monolayer/bilayer/bulk SnSe₂ materials have been systematically investigated.

Our study shows that the binding energy is significantly higher than that on graphene, phosphorene, MoS₂, and some other two-dimensional (2D) layered materials, indicating a strong interaction between the Li atom and SnSe₂ substrate. Bader charge analysis reveals that Li exists in cationic state with its 2s electron being completely transferred to SnSe₂. The extremely low energy barrier (0.197eV) on the monolayer SnSe₂ guarantees rapid diffusion of the Li atom. Due to the unique anionic framework structure of bulk SnSe₂ and formation of the O_h-T_h-O_h migration channel for the Li ion, the energy barrier arises to 0.483eV. Moreover, a remarkably large average voltage of 3.05V is predicted, which is even lerger than on phosphorene (2.9V). The intercalation of Li leads to a transition from semiconductor to metallic state, which gives rise to a good electrical conductivity. These findings provide insights into the Li-ion adsorption properties and migration mechanism in layered transition-metal dichalcogenides (TMDS).

1 Computational methods

All the calculations were performed using Vienna ab initio simulation package (VASP) based on density functional theory^[49]. The interactions between ion cores and valence electrons were described using the projector augmented wave (PAW) potentials^[50]. The Perdew-Burke-Ernzerhof (PBE) generalized gradient approximation (GGA) was carried out for the electron exchange-correlation interactions^[51-52]. The cutoff of plane-wave kinetic energy was set to be 550eV for all the calculations. Atomic relaxation was performed with the convergence of total energy less than 10^-5eV and all the forces on each atom smaller than 0.0 1eV/ Å. The climbing image nudged elastic band method (CI-NEB) was used to search the minimum energy path (MEP) of Li migration^[53]. For the Li adsorption and diffusion studies, we selected supercell containing 4×4 primitive cells for eliminating the interaction between Li atoms. For the monolayer/bilayer SnSe₂, we set a vacuum space of 20 Å between adjacent layers to avoid mirror interaction. K-point meshes of 5×5×1 for monolayer/bilayer SnSe₂ and 5×5×5 for bulk SnSe₂ were employed in all calculations, and denser K-point meshes of 12×12×1 and 12×12×10 were adopted for monolayer and bulk band structures. Bader charge analysis^[54] method was used to estimate the charge transfer between Li and substrate SnSe₂. In addition, van der Waals correction was taken into account using the D2 method of Grimme^[55] in the bilayer and bulk calculations.

2 Results and discussion 2.1 Adsorption of Li on monolayer SnSe₂ surface and in bilayer/bulk SnSe₂

Bulk SnSe₂ possesses the typical layered TMDS structure, in which per SnSe₂ layer has the sandwich-like structure with the Sn layer sandwiched between two Se layers, as shown in Fig. 1.

We investigated the adsorption of one Li atom on the surface of monolayer SnSe₂ and in the bilayer/bulk SnSe₂. In Fig. 1(a) and 1(b) shown are the two possible adsorption sites T₁ and T₂ on SnSe₂ monolayer. The T₁ site is directly above the Sn atom that is located in the second atomic layer and the T₂ site is directly above the Se atom that is located in the third atomic layer. As for bilayer/bulk SnSe₂, there are also two possible adsorption sites: the tetrahedral (T_h) site formed by three Se atoms from lower triple layers and one Se atom from upper layer and the octahedral (O_h) site at which Li binds three Se atoms from the triple layers as shown in Fig. 1(c). The binding energy (E_b) was defined

$ {E_{\rm{b}}} = {E_{{\rm{Li - SnSe}}}}_{_2} - {E_{{\rm{SnSe}}}}_{_2} - {E_{{\rm{Li}}}}, $

(1)

where E_Li-SnSe2, E_SnSe2, and E_Li are the total energies of Li-adsorbed SnSe₂, SnSe₂, and a Li atom, respectively. A negative value of E_b suggests an exothermic reaction, and a more absolute value of E_b indicates a stronger interaction between Li and SnSe₂. The values of E_b for Li are -3.01eV (T₁) and -2.98eV (T₂) on monolayer SnSe₂, -3.52eV (T_h) and -3.56eV (O_h) in bilayer SnSe₂, and -3.70eV (T_h) and -3.91eV (O_h) in bulk SnSe₂, which are significantly higher than those on graphene, phosphorene, MoS₂, and some other two-dimensional (2D) layered materials.

The spatial distributions of the charge difference between Li and SnSe₂ for monolayer/bulk SnSe₂ are illustrated in Fig. 2. The large charge deciency at Li and the charge excess around nearby the Se atom indicate strong electron transfer from Li to SnSe₂. Bader charge analysis reveals that Li transferred almost the whole charge of 2s electron to SnSe₂ (see Table 1) and thus exists in the cationic state, which gives an explanation for the strong interaction between Li and SnSe₂.

2.2 Diffusion processes of the Li-ion on monolayer and bulk SnSe₂

The performances on charging/discharging and circuit rate in Li ion battery are highly dependent on the diffusivity of Li atom in the electrode material. As shown in Fig. 1(a), three possible diffusion pathways (P₁: Sn-Sn; P₂: Se-Se; P₃: Se-Sn) are considered to explore the optimistic migration path on monolayer SnSe₂ surface. As shown in Fig. 3(a), the Li ion migrates with a remarkably low barrier of only 0.197eV along path P₁ from one T₁ site to another T₁ site, passing through a T₂ site (Fig. 3(c) and (d)). Compared with the other 2D materials, such as MoS₂, VS₂, and graphene, the Li diffusion is essentially isotropic in the plane, and the migration barriers are 0.25eV on MoS₂^[28], 0.22eV on VS₂^[56], and 0.327eV on graphene^[23], showing that the monolayer SnSe₂ is more favorite for Li ion migration. For bulk SnSe₂, it is found that Li ions prefer the O_h-T_h-O_h migration path^[57], which connects two neighboring O_h sites by passing through a face-sharing T_h site, as shown in Fig. 3(b), 3(e), and 3(f). The corresponding energy barrier arises to 0.483eV, implying that the diffusion of the Li ions on surface of SnSe₂ are much faster than in bulk SnSe₂.

2.3 Average open circuit voltage for monolayer/bulk SnSe₂-based LIB

The open-circuit-voltage (OCV) is widely used to characterize the charging/discharging performance of Li ion battery. Thus, to further understand the performance of SnSe₂-based Li ion battery, the open-circuit-voltage has been derived. In theory, the open circuit voltage curve can be obtained by calculating the average voltages over different Li concentrations. The charging/discharging process of SnSe₂-based anode can be assumed as

$ {\rm{L}}{{\rm{i}}_{x1}}A + ({x_2} - {x_1}){\rm{L}}{{\rm{i}}^ + } + ({x_2} - {x_1}){{\rm{e}}^ - } \leftrightarrow {\rm{L}}{{\rm{i}}_{x2}}A. $

(2)

Therefore, the average voltage of Li_xSnSe₂ in the x₁≤ x ≤ x₂ range can be evaluated using equation

$ V \approx \frac{{{E_{{\rm{Li}}}}{{_{_{x1}}}_{{\rm{SnSe}}}}_{_2} - {E_{{\rm{Li}}}}{{_{_{x2}}}_{{\rm{SnSe}}}}_{_2} + ({x_2} - {x_1}){E_{{\rm{Li}}}}}}{{({x_2} - {x_1}){\rm{e}}}}, $

(3)

whereE_Lix1SnSe2, E_Lix2SnSe2, and E_Li are the energies of Li_x1SnSe₂, Li_x2SnSe₂, and metallic Li, respectively. In this work, a series of Li_xSnSe₂ (x=0.04, 0.083, 0.125, 0.25, 0.5) are considered by adopting the 1×1, 2×2, 2×4, 3×4, and 5×5 supercells with one Li atom absorbed on monolayer SnSe₂ surface. The voltage profile and geometrical configuration are shown in Fig. 4, and there is a slight drop from 3.2V to 2.84V. The calculated average voltage by averaging numerically the voltage profile is 3.05V in the 0≤x≤ 0.25 range, which is obviously higher than the values of 2.9V for phosphorene^[38], 1.5V for graphite and TiO₂^[25], and 0.93V for VS₂^[56], indicating that the monolayer SnSe₂ provides a higher charging voltage. Furthermore, for the intercalation of Li in bulk-SnSe₂, a series of Li concentrations were considered by using 1×1, 2×2, 2×4, 3×4, and 5×5 supercells with one Li atom intercalated in bulk-SnSe₂. The obtained average OCV is about 3.7V, which is even higher than the value for monolayer SnSe₂. Besides, the OCV does not drop a lot when the Li concentration increases from 0.04 to 0.25.

2.4 The electronic properties of Li-intercalated monolayer/bulk SnSe₂

As we know, the Li ion battery has been one of promising battery systems, used to power a great deal of devices in many applications, such as low-current cells applied in portable electronics and memory backup and high-current cells used in military applications. Thus, it is essential for us to investigate the electronic properties of electrode materials.

As revealed by Bader charge analysis, the Li atoms transfer almost the whole charge of 2s electron to substrate SnSe₂, which affects the electronic structure of the whole system distinctly, such as shift of Fermi energy level compared to the original band structures. The electronic band structures of SnSe₂ with different Li concentrations have been calculated, and the results are shown in Fig. 5. Obviously, Fermi energy levels shift up with increasing of Li concentration, and the whole system transits from original semiconducting state to metallic state during the process of Li intercalation, giving rise to a good electrical conductivity.

3 Conclusion

In conclusion, based on first-principle calculations, we performed a systematically study on the adsorption properties and migration mechanism of the Li atoms in monolayer/bilayer/bulk SnSe₂ systems. Our study shows that the Li atom is able to form stable adsorption with SnSe₂ with a higher binding energy than with some common 2D layered materials. Upon adsorption, the Li atom transfers its 2s electron to substrate SnSe₂ and exists in the cationic state. The adsorption complex becomes metallic with Li interaction, giving rise to a good electrical conductivity, which is essential for an electrode. The energy barrier is only 0.197eV for Li in monolayer SnSe₂, significantly lower than those on other 2D materials, such as MoS₂, graphene, SnS₂, VS₂, etc. Due to the unique anionic framework of bulk SnSe₂, the O_h-T_h-O_h migration channel turns out to be preferred. Furthermore, the average voltage for monolayer SnSe₂ has been estimated to be 3.05V, which is obviously suitable for high charging voltage applications. On the basis of our present findings, monolayer SnSe₂ is expected to be one of candidates for electrode materials. These findings provide insights into the Li-ion adsorption properties and migration mechanism in layered TMDS.

Appendix