Two-Dimensional Hybrid Composites of SnS2 with ... - ACS Publications

Jan 10, 2019 - Pyongyang, Democratic People's Republic of Korea. •S Supporting Information. ABSTRACT: Among the recent achievements of sodium-ion...
1 downloads 0 Views 6MB Size
Download PDF
Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/IC

Two-Dimensional Hybrid Composites of SnS2 with Graphene and Graphene Oxide for Improving Sodium Storage: A First-Principles Study Kum-Chol Ri, Chol-Jun Yu,* Jin-Song Kim, and Song-Hyok Choe

Inorg. Chem. Downloaded from pubs.acs.org by UNITED ARAB EMIRATES UNIV on 01/14/19. For personal use only.

Computational Materials Design, Faculty of Materials Science, Kim Il Sung University, Ryongnam-Dong, Taesong District, Pyongyang, Democratic People’s Republic of Korea S Supporting Information *

ABSTRACT: Among the recent achievements of sodium-ion battery (SIB) electrode materials, hybridization of twodimensional (2D) materials is one of the most interesting appointments. In this work, we propose to use the 2D hybrid composites of SnS2 with graphene or graphene oxide (GO) layers as the SIB anode, based on the first-principles calculations of their atomic structures, sodium intercalation energetics, and electronic properties. The calculations reveal that a graphene or GO film can effectively support not only the stable formation of a heterointerface with the SnS2 layer but also the easy intercalation of a sodium atom with low migration energy and acceptable low volume change. The electronic charge-density differences and the local density of states indicate that the electrons are transferred from the graphene or GO layer to the SnS2 layer, facilitating the formation of a heterointerface and improving the electronic conductance of the semiconducting SnS2 layer. These 2D hybrid composites of SnS2/G or GO are concluded to be more promising candidates for SIB anodes compared with the individual monolayers.

■

LIBs, graphite has a theoretical capacity of 372 mA·h·g−1, which is not high enough to satisfy the growing demand of battery capacity. For SIBs, moreover, it is almost impossible to intercalate Na itself into graphite because of a big mismatch between the ionic radius of the Na+ cation and the interlayer distance of graphite, and thus cointercalation with organic molecules has been devised.19−21 Alternatively, graphene and its derivatives, such as GO or rGO, which can be obtained by exfoliation from graphite, have been studied extensively, confirming the doubled capacity compared with graphite, very high surface area (2680 m 2 ·g −1 ), high electron conductivity, and excellent chemical stability.14−16,22,23 However, both Li and Na bind very weakly to pristine graphene or GO, resulting in Li/Na clustering and dendrite growth. On the other hand, tin disulfide (SnS2) as the typical TMD is recognized as a promising anode material because of its high initial capacity with a unique mechanism of alkali-cation storage reaction.24−26 The bulk SnS2 has a PbI2-type layered structure and exhibits a high lithium storage capacity (average capacity of 583 mA·h·g−1 after 30 cycles) and a good cycling stability (cycle life performance of 85% after 30 cycles),25 which is a much better performance than that of conventional graphite electrodes. Lithium or sodium storage in SnS2 is performed through a two-step reaction: (1) conversion of SnS2 into metallic Sn and (2) alloying with Li or Na, forming a Li−

INTRODUCTION Recently, inorganic two-dimensional (2D) materials have attracted remarkable attention for various energy applications, including energy-harvesting and energy-storage devices, because of their unique structural and electrochemical properties.1−4 Graphene and chemically modified graphene such as graphene oxide (GO) or reduced graphene oxide (rGO) are regarded as the representative 2D materials, being characterized by layer-stacked structures coupled by weak interlayer van der Waals (vdW) and strong in-plane covalent bonding interactions.5−9 Such graphene-like 2D structures can be found in elemental analogues of graphene (silicene, germanene, phosphorene, and borophene), transition-metal oxides (TMOs), 1 0 transition-metal dichalcogenides (TMDs), 2,11,12 and transition-metal carbides/nitrides (MXenes).13 These 2D materials can be obtained by exfoliation or synthesis and can provide large surface area with abundant anchoring sites and low volume change during the de/intercalation processes of atoms or molecules. Therefore, 2D materials can be used as potential electrode materials for alkali-ion batteries with remarkably large capacity, high rate capability, good cycle performance, and low price. Many of the 2D materials, such as rGO,14−16 layered vanadium pentoxide (V2O5),17 and molybdenum disulfide (MoS2),18 have been verified to be used as efficient electrode materials for lithium-ion batteries (LIBs) and sodium-ion batteries (SIBs) by experimental and theoretical studies. In fact, as the most widely used and standard anode material for © XXXX American Chemical Society

Received: October 17, 2018

A

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

and 9 O atoms in one layer. It should be noted that the GO layer with a C/O atomic ratio of 2:1 was formed by decorating epoxy groups (−O−) on the outer side of the graphene sheet. For these supercell models, fixed cell lengths of a = b = 7.79 Å and c = 30 Å were used throughout the work, providing a vacuum layer thickness of over 15 Å (Figure S1). It is worth noting that the model with the SnS2 bilayer sandwiched between two graphene or GO layers is just a model containing both the interlayer space between the two SnS2 layers (homointerface) and that between the SnS2 and graphene or GO layers (heterointerface) with the least number of layers; for a model with the SnS2 monolayer, the homointerface is not formed, while for a model with the SnS2 trilayer, one more homointerface is created and Na intercalation is expected to be uninfluenced by additional SnS2 layers because of the weak vdW interaction between the layers in such 2D materials and moreover almost perfect screening by the layers. All of the calculations in this work were carried out using the pseudopotential plane-wave method, as implemented in the QUANTUM ESPRESSO (QE) package (version 6.2).47 We utilized the projector-augmented-wave method to describe the interaction between ions and valence electrons (we used C.pbesol-n-kjpaw_psl.0.1.UPF, O.pbesol-n-kjpaw_psl.0.1.UPF, Sn.pbesol-dn-kjpaw_psl.0.2.UPF, S.pbesol-n-kjpaw_psl.0.1.UPF, and Na.pbesol-spnkjpaw_psl.0.2.UPF, which are provided in the package). The PBEsol functional48 within the generalized gradient approximation was used for the exchange-correlation interaction between valence electrons. To take into account the vdW interaction between the layers, the vdW energy provided by the exchange-hole dipole moment (XDM) method49 was added to the DFT total energy. As major computational parameters, the plane-wave cutoff energies were set as 40 Ry for the wave function and 400 Ry for the electron density and the Monkhorst−Pack special k points were set as (4 × 4 × 1) for all of the supercell models, providing a total energy accuracy of 5 meV per atom. The self-consistent convergence threshold for the total energy was 10−9 Ry, and the convergence threshold for the atomic force in structural relaxation was 4 × 10−4 Ry·bohr−1. The Fermi− Dirac function with a Gaussian spreading factor of 0.01 Ry was applied to the Brillouin zone integration. To calculate the activation barriers for Na migration, we applied the climbing-image nudged elastic band (NEB) method.50 During the NEB runs, the supercell dimensions were fixed, while all of the atomic positions were allowed to relax until the atomic forces converged within 0.05 eV·Å−1. The number of NEB images was tuned so that the distance between neighboring NEB images was less than 1 Å. We used the image energies and their derivatives to make an interpolation of the path energy profile that passes through every image point, as implemented in the neb.x code of the QE package.47 In order to estimate the binding strength between the layers, we calculated the interlayer binding energy per atom as follows:20,21

Sn or Na−Sn alloy. By contrast, only the conversion reaction occurs during lithium or sodium storage in other TMDs such as MoS2, WS2, and VS2.12 The reversible alloying reaction was found to give a large capacity of sodium storage in the first cycle but accompanied by a large volume expansion (e.g., 420% for the case of Na−Sn alloying27,28), leading to a poor capacity after a few cycles. Designing heterolayered architectures by coupling graphene or GO with TMD or TMO 2D materials has emerged as an effective strategy for enhancing the Li/Na binding strength and reducing the volume change as well.29−32 In particular, hybrid composites formed by coupling rGO and TMDs have been reported to show a good performance as anode materials for LIBs and SIBs.33−35 In such composite systems, graphenebased 2D materials can not only serve as a matrix to anchor the TMD layers but also improve the electrochemical performance of semiconducting TMDs. As a typical example of such composites, therefore, hybrid SnS2/rGO composites, including the heterointerface, can be constructed and used as excellent anode materials by combining the metallic electrical conductivity of graphene with the high storage capability of SnS2.36−41 These synthesized SnS2/rGO hybrids have been shown to exhibit excellent electrochemical performance such as a high sodium storage capacity of 649 mA·h·g−1 at a current density of 100 mA·g−1, ultralong cycle lives of 89% and 69% after 400 and 1000 cycles, and a superior rate capability of 337 mA·h·g−1 at a higher current density of 12.8 A·g−1 (28C) in only 1.3 min.37−40 To the best of our knowledge, no theoretical works for 2D hybrid SnS2/G or GO composites can be found in the literature in spite of such successful experiments and the great importance of clarifying the operational mechanism, motivating us to investigate these heterointerfaces in the atomic scale by means of first-principles modeling and simulations. In this work, we model hybrid SnS2/G and SnS2/GO composites, together with 2D-bilayered SnS2, to conduct a comparative and systematic study by using first-principles calculations within density functional theory (DFT). Performing structural optimizations, we analyze the detailed atomic structures of these 2D-layered compounds, such as the interlayer distances and bond lengths. The activation barriers for the in-plane atomic migration of Na atoms are calculated in these composite systems of SnS2/G and SnS2/GO and in the SnS2 bilayer. The local density of states (LDOS) and electronic charge-density differences are analyzed. On the basis of the calculation data, we reveal the crucial role of a heterointerface behind the sodium storage enhancement of a SnS2/GO hybrid.

■

Eb =

1 [E(d = deq) − E(d = d∞)] N

(1)

where E(d=deq) and E(d=d∞) are the DFT total energies of the supercells with the interlayer distance at equilibrium (deq) and infinity (d∞), respectively, and N is the number of atoms included in the supercell. Because of the numerical limitation, d∞ was approximated to be ∼15 Å, over which the total energy of the supercell is scarcely changed and the interaction between the layers becomes negligible. The possibility of interface formation can be estimated by calculating the formation energy per unit area defined as

COMPUTATIONAL METHODS

Regarding atomistic modeling, three kinds of 2D systems, i.e., SnS2, SnS2/G, and SnS2/GO, were built using the corresponding supercells. The SnS2 crystal has a hexagonal structure (P3̅m1 space group) with lattice constants of a = 3.649 Å and c = 5.899 Å,42 characterized by the interlayer stacking called 2H-SnS2 polytype, as was recently studied with the first-princples method.43−46 For modeling of the SnS2 layer, we used a AA-stacked SnS2 bilayer, employing the hexagonal (2 × 2) cells in plane, which contain 4 Sn atoms and 8 S atoms in one layer. For the composite systems of the SnS2 layer with a graphene or GO layer, we built sandwiched supercells, where the SnS2 bilayer is placed between the graphene or GO layers, giving a hybrid SnS2/G or SnS2/GO composite. Again, the hexagonal (2 × 2) cells were used for modeling of the SnS2 bilayer in these composite systems, while the AA-stacked hexagonal (3 × 3) cells were employed for the graphene or GO layer that contains 18 C atoms or 18 C atoms

Ef =

1 bulk [ESnS2 /G (GO) − (NSnS2ESnS + NG (GO)EG (GO))] 2 2A

(2)

where ESnS2/G (GO) and EG (GO) are the total energies of the supercells including the hybrid SnS2/G (GO) bilayer and graphene (GO) monolayer, and Ebulk SnS2 is the total energy per formula unit of the unit cell of the crystalline bulk SnS2. NSnS2 and NG (GO) are the numbers of SnS2 formula units included in the layer and of the graphene (GO) monolayers, and A is the interfacial area in the x−y plane. Because the supercell models include two equivalent interfaces between the SnS2 B

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. Polyhedral view of relaxed atomic structures of the SnS2 bilayer and hybrid SnS2/G and SnS2/GO composites. The hybrid composites contain the two equivalent heterointerfaces between the SnS2 and graphene or GO layer. Various interlayer distances are denoted. and graphene or GO layers, as can be seen in Figure 1, the energy difference should be divided by 2 to give the formation energy of the interface. To estimate the stability of Na-intercalated compounds, we calculated the Na intercalation energy Eint as follows: bcc Eint = Esub + Na − (Esub + E Na )

the average C−C and C−O bond lengths were determined to be 1.49 and 1.42 Å, respectively, which are comparable with the previous values of 1.51 and 1.42 Å obtained by the firstprinciples calculations.53−55 Using these monolayers of SnS2, graphene, and GO, we made the supercells for the SnS2 bilayer and hybrid SnS2/G and SnS2/GO composites the same size as the monolayer supercells and performed atomic relaxation to determine the real atomic positions, as shown in Figure 1. In these bilayers, the average Sn−S bond lengths were found to be 2.63 Å, which is a little larger than that of 2.56 Å in the bulk, while the C−C and C−O bond lengths were almost the same as those of the monolayers because of the same sizes of the supercells. In order to determine the optimal interlayer distance and the corresponding binding energy between the layers, we estimated the total energy differences of the supercells by varying the interlayer distance. In accordance with variation of the interlayer distance, the supercell length c was also varied to preserve the vacuum layer thickness as 15 Å. Two different interlayer bindings can be distinguished as those between the neighboring SnS2 layers (homointerface) and between the SnS2 and graphene or GO layers (heterointerface), while four different interlayer distances can be thought of as dSn−Sn, dS−S, dC−S, and dC−C, as depicted in Figure 1. Over the interlayer distance of ∼15 Å, the total energies of the supercells were confirmed to be almost unchangeable, indicating a noninteraction or no binding between the layers. Figure 2 shows the total energy differences of the supercells with gradually increasing interlayer distances compared to that of the supercell with an infinite interlayer distance (∼15 Å). Table 1 lists the determined optimal interlayer distances and binding energies and interface formation energies. The Sn−Sn and S−S interlayer distances were found to become longer upon the formation of hybrid interfaces from 4.97 and 2.18 Å in the SnS2 bilayer to 5.04 and 2.28 Å in SnS2/G and to 5.13 and 2.37 Å in SnS2/GO. This indicates that there is an

(3)

where Esub+Na and Esub are the total energies of the supercells for the Na-intercalated compounds and pristine substrates (SnS2, SnS2/G, and SnS2/GO) and Ebcc Na is the total energy per atom of the unit cell of the Na crystal in the body-centered-cubic (bcc) phase.

■

RESULTS AND DISCUSSION Hybrid Composites. We first optimized the crystalline structure of bulk SnS2 in the hexagonal phase (P3̅m1 space group; Figure S1) by using different XC functionals. In accordance with the previous first-principles calculations,43−46 the PBEsol functional yielded slightly overestimated lattice constants (0.05% and 2% for a and c, respectively), while adding the vdW energy in the flavor of Grimme (−0.05% and −5%) or XDM49 (−0.7% and −5%) gave an underestimation compared with the experimental values of a = 3.649 Å and c = 5.899 Å.42 Although XDM gave a somewhat severely underestimated interlayer lattice constant for bulk SnS2, we choose it for further calculations of the interface in this work based on the established fact that XDM is especially good for the modeling of surfaces51 and organic/inorganic interfaces.52 Then, we performed atomic relaxation in the supercells of the individual SnS2, graphene, and GO monolayers, of which the 2D x−y planes include (2 × 2) cells for the SnS2 layer and (3 × 3) cells for the graphene and GO layers (Figure S1). The cell parameters of these supercells were set to be identical and fixed during atomic relaxation as a = b = 7.79 Å and c = 30 Å, allowing us to model the heterointerfaces from these individual layers. In fact, we carried out the variable cell relaxation in only the 2D x−y plane of the GO supercell to determine the optimal cell size in the plane. In the graphene and GO layers, C

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

interaction between the SnS2 and graphene layers. It is worth noting that all of the calculated interlayer binding energies are smaller than the binding energy between the graphene layers in graphite (−77 meV by the PBE+vdW calculation20,21) and all of the negative values indicate certain formations of stable interface systems from the individual monolayers. The interlayer distances between the graphene layers (dC−C) in the hybrid SnS2/G and SnS2/GO composites were calculated to be 13.96 and 13.87 Å, indicating that the epoxy groups of GO lead to a contraction of the total interlayer distance. The calculated formation energies of the interfaces from bulk SnS2 and the graphene or GO monolayer turned out to be positive, suggesting an endothermic feature of the interface formations. It can also be seen from the magnitudes of their formation energies that the heterointerface in SnS2/G or SnS2/GO is easier to form than the homointerface in SnS2 and the two heterointerfaces in SnS2/GO are easier to form than that in SnS2/G. Na Intercalation and Migration. To study the Na intercalation into these SnS2 composites containing the interfaces, we considered two different scenarios: (1) Na intercalation into the middle interspace between the neighboring SnS2 layers, forming substrate-Na-mid compounds, and (2) Na insertion into the heterointerspace between the SnS2 and graphene or GO layers, forming substrate-Na-het compounds, as illustrated in Figure 3a. For the case of the SnS2 bilayer, only the middle intercalation is surely possible. In each manner of Na intercalation, we identified the energetically favorable anchoring site of the Na atom by comparing the total energies of the supercells with different Na insertion sites. In the cases of middle intercalation, two different sites might be thought of: (1) the Na atom can sit on the top of the S atom of the SnS2 layer below, while at the center of the S triangle of the upper SnS2 layer, leading to the formation of a NaS4 tetrahedron (tet site), and (2) the Na atom can be placed at the spherical center of a SnS6 octahedron (oct site). In energetics, the Na-middle-intercalated SnS2 supercell in the tet site was found to be about 0.39 eV higher than that in the oct site, possibly because of a stronger bonding of the Na atom with the surrounding S atoms (Figure S2). Henceforth, only the oct site will be considered for the middle intercalations. Furthermore, it was observed for the cases of hybrid SnS2/G (GO) composites that the SnS2 layers can be shifted in the opposite direction to each other, resulting in a change of the layer stacking from AA to AB with a lowering of the total energy as 0.21 eV (0.17 eV; Figure S2). On the other hand, we distinguished three different sites for the Na heterointercalations in the cases of SnS2/G (GO) composites: (1) the Na atom can sit on the top of the Sn atom placed at the center of the S triangle in the SnS2 layer below, while on the side of the C−C bond of the upper graphene (GO) layer (top site), (2) on the hollow site at the center of the S triangle in the layer below (hollow site), and (3) at the center of C hexagon ring of the upper layer (center site). Among these Na heterointercalation sites, the top site was found to have the lowest total energy (Figure S3). Figure 3 shows the relaxed atomic structures of Naintercalated SnS2 and SnS2/G (GO) composites in the lowest-energy configurations. It is clear that the volume expansion ratio between the Na-intercalated composite (Vint) V and substrate (V0), r = Vint × 100%, which can be readily

Figure 2. Interlayer binding energy per atom with varying interlayer distances of (a) dSn−Sn and dS−S (inset) in the SnS2, SnS2/G, and SnS2/GO bilayers and (b) dC−S in the SnS2/G and SnS2/GO bilayers.

Table 1. Interlayer Distance (d), Interlayer Binding Energy per Atom (Eb), and Interface Formation Energy (Ef) in the SnS2, SnS2/G, and SnS2/GO Bilayers dSn−Sn (Å) dS−S (Å) Eb (SnS2−SnS2) (meV) dC−S (Å) Eb (C−SnS2) (meV) dC−C (Å) Ef (J·m−2)

SnS2

SnS2/G

SnS2/GO

4.97 2.18 −54

5.04 2.28 −22 3.19 −28 13.96 0.34

5.13 2.37 −16 3.10 −30 13.87 0.13

0.81

attractive interaction between the SnS2 layer and graphene or GO layer, and the strength of the SnS2−GO interaction can be said to be higher than that of the SnS2−G interaction because of the further lengthening of dSn−Sn or dS−S in the case of the GO hybrid. Accordingly, the corresponding interlayer binding energy per atom between the neighboring SnS2 layers, Eb(SnS2−SnS2), also decreases in magnitude from −54 meV in SnS2 to −22 meV in SnS2/G and to −16 meV in SnS2/GO. In the cases of hybrid composites of SnS2/G and SnS2/GO, the interlayer distance dC−S in the GO hybrid (3.10 Å) was determined to be shorter than that in the graphene hybrid (3.19 Å), while the binding energy of the SnS2 layer with the GO layer (−30 meV) was found to be higher in magnitude than that with the graphene layer (−28 meV). These also indicate that the oxidation of graphene can enhance the

0

D

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Table 2. Overview of Na Intercalation and Diffusion in Both Middle and Hetero Manners, Including the Interlayer Distances d, Volume Expantion Ratio r, Na Intercalation Energy Eint, and Na Migration Energy Emig SnS2 mid dSn−Sn (Å) dS−S (Å) dC−S (Å) dC−C (Å) r (%) Eint (eV) Emig (eV)

6.21 3.37

119 −0.30 0.52

SnS2/G

SnS2/GO

mid

het

mid

het

5.91 3.09 3.26 15.15 109 −1.29 0.37

4.86 2.10 3.87 14.76 106 −0.76 0.32

6.02 3.19 3.19 15.13 109 −1.12 0.48

4.93 2.17 3.65 14.56 105 −0.35 0.38

heterointercalated SnS2/G (GO) as anode materials, which need to possess an electrode potential as low as possible. We then investigated diffusion of the Na atom by calculating their activation barriers, which can give a suggestion of the rate capability and cycling stability of electrode materials in SIB applications. For the cases of middle intercalation, we considered the Na migration path from the site determined by atomic relaxation to the adjacent image site, which consists of seven NEB image configurations. It should be noted that the interval of the neighboring image points was set at around 1 Å, and the migration paths were determined to be almost straight line along the a direction on the x−y plane, as shown in Figure 4a. In Figure 4b, we show the path energy profile

Figure 3. (a) Schematic illustration of Na intercalation into the middle interspace and heterointerspace of SnS2 and graphene or GO composites. Side views of relaxed atomic structures of Na-intercalated (b) SnS2, (c) SnS2/G, and (d) SnS2/GO compounds. The left panel is for the Na middle intercalations, the central panel for the substrates, and the right panel for the Na heterointercalations. Typical interlayer distances in angstrom units are denoted for comparison.

estimated by using the outermost interlayer distances because of the same in-plane surface area, is distinctly larger in the nonhybrid SnS2 bilayer (∼119%) than in the hybrid SnS2/G (GO) composites. This indicates that the hybrid of SnS2 with graphene or GO can effectively reduce the big volume change of SnS2 itself upon a de/sodiation process. For both cases of the hybrid SnS2/G and SnS2/GO composites, meanwhile, the volume expansion ratios upon Na middle intercalation (109%) are larger than those of upon heterointercalation (105%). As presented in Table 2, interlayer distances such as dSn−Sn, dS−S, and dC−C are clearly smaller in the Na-heterointercalated compounds than in the middle-intercalated ones. From these results, Na intercalation into the heterointerface between the SnS2 and graphene or GO layers is more profitable for SIB application than that into the homointerface with respect to the volume change. The intercalation energies were calculated to be negative, indicating an exothermic reaction of Na intercalation from the bilayer substrate and Na metal. The magnitude of the intercalation energy for the cases of the heterointerfaces was smaller than that for the homointerface, giving an expectation of a good performance of Na-

Figure 4. (a) Top view of Na migrations along the one-dimensional lines within the middle interspaces of SnS2, SnS2/G, and SnS2/GO composites and (b) the corresponding path energy profiles.

corresponding to the migration paths, from which the activation energy barriers were determined to be 0.52, 0.37, and 0.48 eV for the case of SnS2, SnS2/G, and SnS2/GO, respectively. Because the lower activation energies in the hybrid SnS2/G or GO composites than in pristine SnS2 itself, it can be seen that the hybrid of SnS2 with the graphene or GO layer can facilitate the diffusion of Na. Figure 5 presents the results of the Na migration within the heterointerface between the SnS2 and graphene or GO layer. E

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 5. (a) Top view of Na migrations within the heterointerspaces of hybrid SnS 2 /G and SnS 2 /GO composites and (b) the corresponding path energy profiles.

We used five NEB image configurations due to the shorter length of the path. As shown in Figure 5a, the migration paths are 2D on the x−y plane, passing the top, hollow, and center sites, as described in the above subsection. In Figure 5b, we show the path energy profiles in the hybrid SnS2/G and SnS2/ GO composites and the activation barriers as 0.32 and 0.38 eV, which are clearly lower than the corresponding Na migrations within the middle interface. The migration barriers are also listed in Table 2. In summary, the Na atom can migrate more smoothly in the heterointerspace than in the middle interspace, and a hybrid with graphene is slightly more beneficial than that with GO. It is worth noting that the calculated activation energies are comparable with those for Na or Li migration in graphite (0.40 eV)56 and for Na-molecule migration (0.40 eV).20 Electronic Properties. To further understand such an enhancement of Na insertion by a hybrid of SnS2 with graphene or GO and make clear the role of the heterointerface, we carried out an analysis of the electronic properties such as the density of states and electronic charge-density differences. Such electronic properties can provide us with valuable insight for interaction between the layers and charge transfer in the event of heterointerface formation, which can be useful for understanding the enhancement of the electrochemical properties in the 2D hybrid materials. To obtain intuitive insight for charge transfer between SnS2 and graphene or GO layers during the interface formation, we calculated the electronic charge-density difference (Δρ) as follows: Δρ = ρSnS

2 /G

(GO)

− (ρSnS + ρG (GO)) 2

Figure 6. Isosurface plot of the electronic charge-density differences (left panel) and integrated LDOS (right panel) in (a) SnS2 bilayer, (b) SnS2/G, and (c) SnS2/GO composites. The isosurface values are ±0.0005 e·Å−3 in (b) SnS2/G and ±0.0013 e·Å−3 in (c) SnS2/GO. Brown (green) represents electron accumulation (loss). The Fermi level is set at zero.

composites at values of ±0.0005 and ±0.0013 e·Å−3, respectively. For the case of the hybrid SnS2/G composite, it is shown that the C atoms of the graphene layer lose some electrons, as represented by green-colored charge depletion, while the Sn atoms gain the electrons, as represented by brown-colored charge accumulation (Figure 6b, left panel). In Figure 6c, for the hybrid SnS2/GO composite, we can also see the electronic charge accumulation around the O atoms as well as the Sn atoms, while charge depletion can be seen around the C atoms of the GO layer as well. In these hybrid composites,

(4)

where ρSnS2/G (GO), ρSnS2, and ρG (GO) are the charge densities of the hybrid composites, individual SnS2, and graphene or GO layers, respectively. In Figure 6, we show the isosurface plot of the electronic charge-density differences in the hybrid SnS2/G and SnS2/GO F

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

■

therefore, it is clear that the C atoms act as electron donors, whereas the Sn and O atoms act as electron acceptors, indicating a strong interaction between the SnS2 and graphene or GO layers and thus stable formation of the heterointerfaces. Because of the influence of O atoms of the epoxy groups in the GO hybrid, the extent of charge transfer from the C atoms to the O and Sn atoms seems to be enhanced compared with the graphene hybrid, leading to a slight strengthening of the interlayer interaction between the GO and SnS2 layers. When the Na atom is intercalated into the bilayered substrates, the Na and Sn atoms were found to give electrons, leading to enhanced interaction between the layers (Figure S4). We present the LDOS integrated on the x−y plane and thus along the z axis in the right panel of Figure 6, where the Fermi level is set at zero. As shown in Figure 6a, the SnS2 bilayer model was found to have a band gap of ∼1.2 eV between the valence-band maximum and the conduction-band minimum (CBM), which is smaller than that of bulk (2.28 eV),44 indicating a semiconducting feature of the SnS2 bilayer. In the right panel of Figure 6b,c for the hybrid cases, however, the Fermi level can be seen as very close to the CBM belonging to the SnS2 layer, resulting in enhancement of the electronic conductivity in a hybrid of SnS2 with graphene or GO. Moreover, we can see the occupied valence states of the graphene layer and denser LDOS of the GO layer close to the Fermi level in these hybrids, indicating a good in-plane electronic conductivity due to the pz electron states, as was already well-known before.19,21 If the Na atom intercalated into the heterointerface, the SnS2 layers’ unoccupied states were filled by the injected electrons from the Na atom, and thus the Fermi level was placed on the CBM (Figure S4). These reveal enhancement of the electrochemical properties of the hybrid SnS2/G (GO) composites compared to the pristine SnS2 bilayer by the formation of a heterointerface through electron transfer from the graphene or GO layer to the SnS2 layer.

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.8b02962.

■

Unit cell of SnS2 bulk, supercells of monolayer SnS2, graphene, and GO, supercells of Na-intercalated compounds optimized in a homointercalation manner (SnS2-Na-mid, SnS2/G-Na-mid, and SnS2/GO-Na-mid) with different Na atom sites and in a heterointercalation manner (SnS2/G-Na-het and SnS2/GO-Na-het) and the energy diagrams according to the different models, and additional electronic charge-density differences and LDOS integrated in the x−y plane and along the c direction for SnS2-Na-mid, SnS2/G-Na-mid, SnS2/GONa-mid, SnS2/G-Na-het, and SnS2/GO-Na-het (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (C.-J.Y.). ORCID

Chol-Jun Yu: 0000-0001-9523-4325 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is supported as part of the fundamental research project “Design of Innovative Functional Materials for Energy and Environmental Application” (Project 2016-20) funded by the State Committee of Science and Technology, Republic of Korea. Computation has done on the HP Blade System C7000 (HP BL460c), which is owned by the Faculty of Materials Science, Kim Il Sung University.

■

■

CONCLUSIONS In this work, we provide a valuable understanding of the material properties of 2D hybrid composites composed of SnS2 and graphene or GO layers that can be used as effective anode materials of SIB. Various supercell models of the pristine SnS2 bilayer and the hybrid SnS2/G or GO composites have been suggested, and their atomic structures and energetics were investigated using first-principles calculations. Through analysis of the calculated interlayer distances and binding and formation energies, it was revealed that a heterointerface between the SnS2 and graphene or GO layers can be formed, while oxidation of graphene can enhance the interaction between the layers. Furthermore, we have devised two different models for Na intercalation, such as middle intercalation into the interspace between the SnS2 layers and heterointercalation into the interspace between the SnS2 and graphene or GO layers, and it was found that heterointercalation is more probable because of the higher intercalation energy, lower migration energy, and smaller volume change. To gain a deep understanding of the material properties of these 2D composites, the electron-rich charge-density differences and integrated LDOS were investigated, demonstrating electron transfer from the graphene layer to the SnS2 layer and the improvement of electronic conduction by the formation of hybrid composites. In conclusion, the 2D hybrid SnS2/G or GO composite, including the heterointerface, can be a promising candidate for anode materials of SIB.

REFERENCES

(1) Shi, L.; Zhao, T. Recent Advances in Inorganic 2D Materials and Their Applications in Lithium and Sodium Batteries. J. Mater. Chem. A 2017, 5, 3735−3758. (2) Kang, W.; Wang, Y.; Xu, J. Recent Progress in Layered Metal Dichalcogenide Nanostructures as Electrodes for High-Performance Sodium-Ion Batteries. J. Mater. Chem. A 2017, 5, 7667−7690. (3) Georgakilas, V.; Tiwari, J. N.; Kemp, K. C.; Perman, J. A.; Bourlinos, A. B.; Kim, K. S.; Zbo-ril, R. Noncovalent Functionalization of Graphene and Graphene Oxide for Energy Materials, Biosensing, Catalytic, and Biomedical Applications. Chem. Rev. 2016, 116, 5464−5519. (4) Quesnel, E.; Roux, F.; Emieux, F.; Faucherand, P.; Kymakis, E.; Volonakis, G.; Giustino, F.; Martin-Garcia, B.; Moreels, I.; Gursel, S. A.; et al. Graphene-Based Technologies for Energy Ap- plications, Challenges and Perspectives. 2D Mater. 2015, 2, No. 030204. (5) Yu, X.; Cheng, H.; Zhang, M.; Zhao, Y.; Qu, L.; Shi, G. Graphene-Based Smart Materials. Nat. Rev. Mater. 2017, 2, 17046. (6) Kaplan, A.; Yuan, Z.; Benck, J. D.; Govind Rajan, A.; Chu, X. S.; Wang, Q. H.; Strano, M. S. Current and Future Directions in Electron Transfer Chemistry of Graphene. Chem. Soc. Rev. 2017, 46, 4530− 4571. (7) Ambrosi, A.; Chua, C. K.; Latiff, N. M.; Loo, A. H.; Wong, C. H. A.; Eng, A. Y. S.; Bo- nanni, A.; Pumera, M. Graphene and Its Electrochemistry − An Update. Chem. Soc. Rev. 2016, 45, 2458−2493. (8) Zhu, P.; Sumpter, B. G.; Meunier, V. Electronic, Thermal, and Structural Properties of Graphene Oxide Frameworks. J. Phys. Chem. C 2013, 117, 8276−8281.

G

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (9) Dabhi, S. D.; Gupta, S. D.; Jha, P. K. Structural, Electronic, Mechanical, and Dynamical Properties of Graphene Oxides: A First Principles Study. J. Appl. Phys. 2014, 115, 203517. (10) Kubota, K.; Yabuuchi, N.; Yoshida, H.; Dahbi, M.; Komaba, S. Layered Oxides as Positive Electrode Materials for Na-Ion Batteries. MRS Bull. 2014, 39, 416−422. (11) Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L. J.; Loh, K. P.; Zhang, H. The Chemistry of Two-Dimensional Layered Transition Metal Dichalcogenide Nanosheets. Nat. Chem. 2013, 5, 263−275. (12) Pumera, M.; Sofer, Z.; Ambrosi, A. Layered Transition Metal Dichalcogenides for Electrochemical Energy Generation and Storage. J. Mater. Chem. A 2014, 2, 8981−8987. (13) Naguib, M.; Mochalin, V. N.; Barsoum, M. W.; Gogotsi, Y. 25th Anniversary Article: MXenes: A New Family of Two-Dimensional Materials. Adv. Mater. 2014, 26, 992−2014. (14) Ali, G.; Mehmood, A.; Ha, H. Y.; Kim, J.; Chung, K. Y. Reduced Graphene Oxide as A Stable and High-Capacity Cathode Material for Na-Ion Batteries. Sci. Rep. 2017, 7, 40910. (15) Liu, T.; Kavian, R.; Kim, I.; Lee, S. W. Self-Assembled, RedoxActive Graphene Electrodes for High-Performance Energy Storage Devices. J. Phys. Chem. Lett. 2014, 5, 4324−4330. (16) Wang, Y.-X.; Chou, S.-L.; Liu, H.-K.; Dou, S.-X. Reduced Graphene Oxide with Superior Cycling Stability and Rate Capability for Sodium Storage. Carbon 2013, 57, 202−208. (17) Feng, J.; Sun, X.; Wu, C.; Peng, L.; Lin, C.; Hu, S.; Yang, J.; Xie, Y. Metallic Few-Layered VS2 Ultrathin Nanosheets: High TwoDimensional Conductivity for In-Plane Supercapacitors. J. Am. Chem. Soc. 2011, 133, 17832−17838. (18) Mortazavi, M.; Wang, C.; Deng, J.; Shenoy, V. B.; Medhekar, N. V. Ab Initio Characterization of Layered MoS2 as Anode for SodiumIon Batteries. J. Power Sources 2014, 268, 279−286. (19) Ri, G.-C.; Yu, C.-J.; Kim, J.-S.; Hong, S.-N.; Jong, U.-G.; Ri, M.H. First-Principles Study of Ternary Graphite Compounds Cointercalated with Alkali Atoms (Li, Na, and K) and Alkylamines towards Alkali Ion Battery Applications. J. Power Sources 2016, 324, 758−765. (20) Yu, C.-J.; Ri, S.-B.; Choe, S.-H.; Ri, G.-C.; Kye, Y.-H.; Kim, S.C. Ab Initio Study of Sodium Cointercalation with Diglyme Molecule into Graphite. Electrochim. Acta 2017, 253, 589−598. (21) Yu, C.-J.; Ri, U.-S.; Ri, G.-C.; Kim, J.-S. Revealing the Formation and Electrochemical Properties of Bis(trifluoromethanesulfonyl) Imide Intercalated Graphite with FirstPrinciples Calculations. Phys. Chem. Chem. Phys. 2018, 20, 14124− 14132. (22) Lin, D.; Liu, Y.; Liang, Z.; Lee, H. W.; Sun, J.; Wang, M.; Yan, K.; Xie, J.; Cui, Y. Layered Reduced Graphene Oxide with Nanoscale Interlayer Gaps as A Stable Host for Lithium Metal Anodes. Nat. Nanotechnol. 2016, 11, 626−632. (23) Bonaccorso, F.; Colombo, L.; Yu, G.; Stoller, M.; Tozzini, V.; Ferrari, A. C.; Ruoff, D. S.; Pellegrini, V. Graphene, Related TwoDimensional Crystals, and Hybrid Systems for Energy Conversion and Storage. Science 2015, 347, 1246501. (24) Kim, T.-J.; Kim, C.; Son, D.; Choi, M.; Park, B. Novel SnS2Nanosheet Anodes for Lithium-Ion Batteries. J. Power Sources 2007, 167, 529. (25) Seo, J.-W.; Jang, J.-T.; Park, S.-W.; Kim, C.; Park, B.; Cheon, J. Two-Dimensional SnS2 Nanoplates with Extraordinary High Discharge Capacity for Lithium Ion Batteries. Adv. Mater. 2008, 20, 4269−4273. (26) Zhong, H.; Yang, G.; Song, H.; Liao, Q.; Cui, H.; Shen, P.; Wang, C.-X. Vertically Aligned Graphene-Like SnS2 Ultrathin Nanosheet Arrays: Excellent Energy Storage, Catalysis, PhotoConduction, and Field-Emitting Performances. J. Phys. Chem. C 2012, 116, 9319. (27) Wang, J. W.; Liu, X. H.; Mao, S. X.; Huang, J. Y. Microstructural Evolution of Tin Nanoparticles during In Situ Sodium Insertion and Extraction. Nano Lett. 2012, 12, 5897−5902.

(28) Ellis, L. D.; Hatchard, T. D.; Obrovac, M. N. Reversible Insertion of Sodium in Tin. J. Electrochem. Soc. 2012, 159, A1801− A1805. (29) Pomerantseva, E.; Gogotsi, Y. Two-Dimensional Heterostructures for Energy Storage. Nature Energy 2017, 2, 17089. (30) Chauhan, H.; Soni, K.; Kumar, M.; Deka, S. Tandem Photocatalysis of Graphene-Stacked SnS2 Nanodiscs and Nanosheets with Efficient Carrier Separation. ACS Omega 2016, 1, 127−137. (31) Pan, Q.; Xie, J.; Zhu, T.; Cao, G.; Zhao, X.; Zhang, S. Reduced Graphene Oxide-Induced Recrystallization of NiS Nanorods to Nanosheets and the Improved Na-Storage Properties. Inorg. Chem. 2014, 53, 3511−3518. (32) Zhou, L.; Kong, X.; Gao, M.; Lian, F.; Li, B.; Zhou, Z.; Cao, H. Hydrothermal Fabrication of MnCO3@rGO Composite as An Anode Material for High-Performance Lithium Ion Batteries. Inorg. Chem. 2014, 53, 9228−9234. (33) Geim, A. K.; Grigorieva, I. V. Van der Waals Heterostructures. Nature 2013, 499, 419−425. (34) David, L.; Bhandavat, R.; Singh, G. MoS2/Graphene Composite Paper for Sodium-Ion Battery Electrodes. ACS Nano 2014, 8, 1759−1770. (35) Xie, X.; Ao, Z.; Su, D.; Zhang, J.; Wang, G. MoS2/Graphene Composite Anodes with Enhanced Performance for Sodium-Ion Batteries: The Role of the Two Dimensional Heterointerface. Adv. Funct. Mater. 2015, 25, 1393−1403. (36) Chang, K.; Wang, Z.; Huang, G.; Li, H.; Chen, W.; Lee, J. Y. Few-Layer SnS2/Graphene Hybrid with Exceptional Electrochemical Performance as Lithium-Ion Battery Anode. J. Power Sources 2012, 201, 259−266. (37) Ma, C.; Xu, J.; Alvarado, J.; Qu, B.; Somerville, J.; Lee, J. Y.; Meng, Y. S. Investigating the Energy Storage Mechanism of SnS2rGO Composite Anode for Advanced Na-Ion Batteries. Chem. Mater. 2015, 27, 5633−5640. (38) Qu, B.; Ma, C.; Ji, G.; Xu, C.; Xu, J.; Meng, Y. S.; Wang, T.; Lee, J. Y. Layered SnS2- Reduced Graphene Oxide Composite − A High-Capacity, High-Rate, and Long-Cycle Life Sodium-Ion Battery Anode Material. Adv. Mater. 2014, 26, 3854−3859. (39) Zhang, Y.; Zhu, P.; Huang, L.; Xie, J.; Zhang, S.; Cao, G.; Zhao, X. Few-Layered SnS2 on Few-Layered Reduced Graphene Oxide as Na-Ion Battery Anode with Ultralong Cycle Life and Superior Rate Capability. Adv. Funct. Mater. 2015, 25, 481−489. (40) Xu, W.; Zhao, K.; Zhang, L.; Xie, Z.; Cai, Z.; Wang, Y. SnS2@ Graphene Nanosheet Arrays Grown on Carbon Cloth as Freestanding Binder-Free Flexible Anodes for Advanced Sodium Batteries. J. Alloys Compd. 2016, 654, 357−362. (41) Luo, B.; Fang, Y.; Wang, B.; Zhou, J.; Song, H.; Zhi, L. Two Dimensional Graphene-SnS2 Hybrids with Superior Rate Capability for Lithium Ion Storage. Energy Environ. Sci. 2012, 5, 5226. (42) Burton, L. A.; Whittles, T. J.; Hesp, D.; Linhart, W. M.; Skelton, J. M.; Hou, B.; Web- ster, R. F.; O’Dowd, G.; Reece, C.; Cherns, D.; Fermin, D. J.; Veal, T. D.; Dhanak, V. R.; Walsh, A. Electronic and Optical Properties of Single Crystal SnS2: An Earth-Abundant Disulfide Photocatalyst. J. Mater. Chem. A 2016, 4, 1312. (43) Kumagai, Y.; Burton, L. A.; Walsh, A.; Oba, F. Electronic Structure and Defect Physics of Tin Sulfides: SnS, Sn2S3, and SnS2. Phys. Rev. Appl. 2016, 6, No. 014009. (44) Whittles, T. J.; Burton, L. A.; Skelton, J. M.; Walsh, A.; Veal, T. D.; Dhanak, V. R. Band Alignments, Valence Bands, and Core Levels in the Tin Sulfides SnS, SnS2, and Sn2S3: Ex- periment and Theory. Chem. Mater. 2016, 28, 3718−3726. (45) Skelton, J. M.; Burton, L. A.; Oba, F.; Walsh, A. Chemical and Lattice Stability of the Tin Sulfides. J. Phys. Chem. C 2017, 121, 6446−6454. (46) Skelton, J. M.; Burton, L. A.; Jackson, A. J.; Oba, F.; Parker, S. C.; Walsh, A. Lattice Dynam- ics of the Tin Sulphides SnS2, SnS and Sn2S3: Vibrational Spectra and Thermal Transport. Phys. Chem. Chem. Phys. 2017, 19, 12452−12465. (47) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; et al. QUANTUM ESPRESSO: A Modular and Open-Source H

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Software Project for Quantum Simulations of Materials. J. Phys.: Condens. Matter 2009, 21, 395502. (48) Perdew, J. P.; Ruzsinszky, A.; Csonka, G. I.; Vydrov, O. A.; Scuseria, G. E.; Constantin, L. A.; Zhou, X.; Burke, K. Restoring the Density Gradient Expansion for Exchange in Solids and Surfaces. Phys. Rev. Lett. 2008, 100, 136406. (49) Otero de la Roza, A.; Johnson, E. R. Many-Body Dispersion Interaction from the Exchange-Hole Dipole Moment Model. J. Chem. Phys. 2013, 138, No. 054103. (50) Henkelman, G.; Uberuaga, B. P.; Jónsson, H. A Climbing Image Nudged Elastic Band Method for Finding Saddle Points and Minimum Energy Paths. J. Chem. Phys. 2000, 113, 9901−9904. (51) Hong, S.-N.; Kye, Y.-H.; Yu, C.-J.; Jong, U.-G.; Ri, G.-C.; Choe, C.-S.; Kim, K.-H.; Han, J.- M. Ab Initio Thermodynamic Study of the SnO2 (110) Surface in An O2 and NO Environ- ment: A Fundamental Understanding of the Gas Sensing Mechanism for NO and NO2. Phys. Chem. Chem. Phys. 2016, 18, 31566−31578. (52) Johnson, E. R. Non-covalent Interactions in Quantum Chemistry and Physics; Elsevier, 2017; pp 215−248. (53) Šljivančanin, Z.; Milošević, A. S.; Popovic, Z. S.; Vukajlović, F. R. Binding of Atomic Oxy- gen on Graphene from Small Epoxy Clusters to A Fully Oxidized Surface. Carbon 2013, 54, 482−488. (54) Yan, J.-A.; Chou, M. Y. Oxidation Functional Groups on Graphene: Structural and Electronic Properties. Phys. Rev. B: Condens. Matter Mater. Phys. 2010, 82, 125403. (55) Wang, L.; Sun, Y. Y.; Lee, K.; West, D.; Chen, Z. F.; Zhao, J. J.; Zhang, S. B. Stabil- ity of Graphene Oxide Phases from FirstPrinciples Calculations. Phys. Rev. B: Condens. Matter Mater. Phys. 2010, 82, 161406. (56) Nobuhara, K.; Nakayama, H.; Nose, M.; Nakanishi, S.; Iba, H. First-Principles Study of Alkali Metal-Graphite Intercalation Compounds. J. Power Sources 2013, 243, 585−587.

I

DOI: 10.1021/acs.inorgchem.8b02962 Inorg. Chem. XXXX, XXX, XXX−XXX