Controlling of the electronic properties of WS₂ and graphene oxide heterostructures from first-principles calculations

Abstract

Two-dimensional nanomaterials have attracted extensive interest and their heterostructures possess excellent electronic and optical properties suitable for various applications. Based on first-principles calculations, we investigated the structural stability and electronic properties of WS₂ and graphene oxide (GO) heterostructures. We considered three types of GO, including epoxy only, hydroxyl only and both epoxy and hydroxyl on the GO surface. Our results show that the interlayer binding energy per WS₂ unit increases from 0.117 eV to 0.214 eV as the surface oxygen concentration of GO increases. The band gap of the WS₂/GO heterostructures can be efficiently tuned in a wide range from 0.13 eV to 1.91 eV by changing the oxygen functionalities and the concentration. The spatial separation of the conduction band minimum and valence band maximum is observed, which are distributed in different layers. In addition, the work function of WS₂ can also be modulated by GO in the range of 4.09 eV to 6.34 eV, which potentially increases the carrier concentration and broadens the applications of WS₂ and other transition metal dichalcogenide materials in optoelectronic devices.

---

Introduction

Graphene, as the first successfully synthesized two-dimensional (2D) material, has received much attention in the past decade. Besides graphene, other kinds of 2D nanomaterials have been encouraged for exploration and investigation given the advances in nanotechnology in recent years. Among them, 2D transition metal dichalcogenides (TMDs) have attracted great interest, since they exhibit many appealing properties^1,2 for a wide range of applications.^3–7 TMDs are a kind of material (MX₂) with a hexagonal lattice like graphene, in which M represents a transition metal element from group four to six, and X presents a chalcogen (S, Se and Te) element. Pristine TMD materials mainly possess two typical configurations, 2H or 1T, and only a few compounds are stable for both 2H and 1T structures. Their electronic structures are either semiconducting or metallic. Previous studies have shown that their physical and chemical properties can be further tuned by controlling the composition, functionalization, and applying strain or external fields,^8–12 which are proposed to be used in next-generation electronic devices and electrocatalysts.^13,14 Tungsten disulfide (WS₂) has attracted increasing attention due to its unusual physical and chemical properties. Bulk WS₂ is an indirect semiconductor with a band gap of 1.35 eV,^15,16 but it becomes a direct semiconductor when changed into a monolayer, and the band gap increases to about 2.10 eV within the visible light region.^17,18

Inspired by the proposition of van der Waals heterostructures,¹⁹ more efforts have been made to synthesize heterostructures with two or more layers of various 2D materials.^20–22 To take advantage of TMDs, scientists have tried to combine them with the other 2D materials like graphene or graphene oxide (GO) both in experiments and in theory.^23–34 The formed heterostructures exhibit extraordinary properties superior to pristine TMD materials. Recent experiments have demonstrated that TMD and reduced GO (rGO) hybrid materials exhibit fantastic electrocatalytic activity in the hydrogen evolution reaction (HER).^23,24 Monolayer WS₂ on rGO shows high performance with a low Tafel slope (52 mV dec⁻¹) and remarkable durability which are attributed to the formation of an interconnected conducting bridge for rapid electron transport between WS₂ and rGO.²⁵ It has also been reported that TMD/graphene hybrids have been applied in lithium and sodium ion batteries to improve their ion storage, cyclability and rate capability.^26–28 For example, when MoS₂ is supported on graphene, the reversible capacity is significantly increased with the value of 1400 mA h g⁻¹ and the diffusion energy barrier of Li ions is reduced compared to pristine MoS₂. In addition, the WS₂/GO hybrid is proposed to be used as supercapacitor electrodes due to its capacitance value increasing by 5 and 2.5 times higher than isolated GO or WS₂ sheets.²⁹ Besides, GO is also suggested to play an important role in facilitating the growth of WS₂ nanosheets.³⁰ On the theoretical side, more efforts have been made to investigate the geometric and electronic properties of monolayer TMDs combined with other 2D substrates.^31–33 Previous results show that the substrates (such as graphene, graphene oxide, silicene, etc.) exhibited great effects on the properties of TMD materials. A recent theoretical study showed that GO can be a hole injection layer for MoS₂-based devices.³⁴ They demonstrated that the Schottky barrier heights of the formed p-type field-effect transistors (FETs) can be decreased significantly by increasing the oxygen concentration on the surface of GO. Although there are extensive studies on TMD hybrids, some key problems still need to be solved and clarified urgently. For instance, how to stabilize heterostructure combinations of TMDs and other 2D materials, and how to modulate and precisely control the properties of these heterostructures. These problems are critical to their future applications.

In this paper, we systematically performed first-principles calculations to investigate the structural stability and electronic properties of WS₂ and GO hybrids. GO contains a range of reactive oxygen functional groups, and it is generally accepted that the oxygen groups present on GO are mainly in the form of epoxy (oxygen at bridge site, –O–) and hydroxyl (–OH) groups. We will explore the stable structures of WS₂ and GO heterostructures, and further examine the interaction and electronic effects of GO with different oxygen coverage on monolayer WS₂. Our results demonstrated that, by tuning the oxygen functionalities and concentration on the GO surface, we can efficiently control the binding strength between WS₂ and GO, together with the band gaps and work functions of the formed WS₂/GO heterostructures, which pave an effective way to create WS₂-based materials for applications in flexible electronic and optoelectronic devices.

Computational methods

All calculations are carried out by means of the projector augmented wave (PAW) method within first-principles calculations as implemented in the Vienna ab initio simulation package (VASP).^35,36 The Perdew–Burke–Ernzerhof (PBE) functional within the generalized-gradient approximation (GGA) was chosen to treat the exchange–correlation interaction of electrons.³⁷ Due to the weak interaction between layers, the van der Waals correction is taken into account by using the vdW-DF method throughout all the calculations.³⁸ The lattice constant of the graphene unit cell is 2.46 Å, while the lattice of the WS₂ unit cell with 2H configuration is 3.13 Å. In order to construct an appropriate supercell to study the WS₂/GO heterostructure, we consider building a supercell with a 5 × 5 layer of GO and a 4 × 4 layer of WS₂. In this arrangement, the lattice parameters of GO are expanded by 1.79% to match those of WS₂. To optimize the structures and calculate the electronic properties, a Morkhost–Pack K mesh of 4 × 4 × 1 is used. An energy cutoff of 500 eV is used for the plane-wave expansion of the electronic wave function. Geometric structures are fully relaxed until the force on each atom is less than 0.02 eV Å⁻¹ and the convergence criteria for energy is 1 × 10⁻⁵ eV. The vacuum thickness between adjacent layers is set to 18 Å to avoid the adjacent interaction.

Results and discussion

For the stacking patterns of WS₂ and graphene, we considered three possible configurations. As previous studies have mentioned,^39,40 a carbon (C) atom from graphene on top of a tungsten (W) atom from WS₂ is named the TM configuration; a C atom on top of a sulfur (S) atom is named the TS configuration; a C atom locating at the center of the WS₂ hexagonal ring is recognized as the H configuration. Under the structural optimizations, graphene and WS₂ still keep planar structures. Our results show that the energies are not sensitive to different stacking patterns (the optimized structures and energies are shown in Fig. S1 in the ESI†), which is consistent with MoS₂ on graphene.^31,34 Thus, we chose the H configuration for the following calculations. The distance between graphene and the top S layer of WS₂ is about 3.33 Å, which is a typical van der Waals interaction distance. Due to the interaction between WS₂ and graphene, the bond length of W–S and the bond angle of S–W–S in WS₂ decrease by 0.4% and 0.1%, respectively. Next, we will discuss the energetics and electronic properties of WS₂ and graphene with different extent of oxidation.

We define the binding energy (E_b) per WS₂ unit to identify the interactions between WS₂ and GO as follows:

E_b = (E_WS2 + E_GO − E_WS2/GO)/n

where E_WS2, E_GO and E_WS2/GO represent the total energy of monolayer WS₂, GO and WS₂/GO heterostructures, respectively. n is the number of WS₂ units in the supercell. We have considered three conditions for the formation of GO: only epoxy (e-GO), only hydroxyl (h₂-GO) and both epoxy and hydroxyl (he-GO, h₂e-GO and h₄e-GO) on the graphene surface. We take three structural motifs that have been mentioned in previous studies^41,42 to construct the structures of GO with various oxygen concentration, which are shown in Fig. 1a–c. Previous studies have reported that the oxygen functionals prefer to aggregate along the armchair direction, then the GO models with different oxygen concentration are designated in Fig. 1d–f.

Fig. 1 The building blocks of GO with (a) epoxy only, (b) hydroxyl only and (c) both epoxy and hydroxyl; (d–f) top view and side view of the GO structures with different oxygen functionalities.

For the WS₂ and e-GO hybrid, the C–O bond length is 1.47 Å, which is 0.03 Å larger than that in GO without WS₂. The distance between them is similar to that of WS₂/graphene. We have considered two types of e-GO structures, including epoxy distributed on only one side of graphene and both sides of graphene, and the structural stability of e-GO with WS₂ to form the heterostructures are compared (the binding energies are shown in Fig. S2 of the ESI†). Thermodynamically, the heterostructures of WS₂ and e-GO with epoxy located a on single side (Fig. 1d) is energetically favorable, so we chose this type of structure for further discussion. For WS₂ and h₂-GO, the hydroxyl functionals distribute on both sides of GO and induce a larger distance of 3.69 Å between WS₂ and GO compared to that of 3.33 Å between WS₂ and graphene. In addition, we investigated WS₂ and GO with both hydroxyl and epoxy functionalities, and the ratio of epoxy and hydroxyl is 1 : 2 (as shown in Fig. 1c). In this case, the distance between WS₂ and GO is 3.67 Å, which is still larger than that between WS₂ and graphene. The interaction between WS₂ and GO with different oxygen concentrations is systematically investigated and plotted in Fig. 2a. The oxygen concentration is defined as the ratio of the number of sp³ C atoms to the total number of C atoms. Generally speaking, as the concentration of the oxygen groups increases, the binding energies between WS₂ and GO gradually increase. In the case of WS₂ and e-GO, the binding energies keep constant at about 0.19 eV per WS₂ unit and starts to greatly increase from the oxygen concentration of 80%. When the GO is fully oxidized by only epoxy groups, the binding energy between GO and WS₂ can reach 0.214 eV per WS₂. It is different from the other two cases of h₂-GO and h₂e-GO. In these two cases, the binding energies continuously increases from low oxygen concentration, especially from the oxygen coverage of 40% due to the formation of a chain-like structure. Compared with these three cases, the e-GO possess the largest binding energies with WS₂, especially at low oxygen concentration. For example, at the oxygen concentration of 16%, the binding energy per WS₂ unit of e-GO is 0.187 eV, which is 0.06 eV and 0.07 eV larger than that of the h₂e-GO and h₂-GO motifs, respectively. The binding distances between C atoms from GO and the nearest S atoms from WS₂ are 3.31 Å, 3.97 Å and 4.37 Å for e-GO, h₂e-GO and h₂-GO, respectively, at the oxygen concentration of 16%. We also calculated the variation of distances between GO and WS₂ as a function of oxygen concentrations, which is plotted in Fig. S3 of the ESI.† Due to the larger buckling of the GO surface at high oxygen coverage, the distance between WS₂ and GO is defined as the vertical distance between the nearest S atoms from WS₂ and O atoms from GO. Our results show that the distances between GO (h₂-GO and h₂e-GO) and WS₂ increase as the oxygen concentration increases.

Fig. 2 (a) Binding energies between WS₂ and GO with different oxygen concentration; (b) variations of band gaps of the WS₂/GO heterostructures as the oxygen concentration increases.

On the basis of the stable WS₂/GO structures, we further investigated the electronic properties that are induced by the different oxygen concentrations of GO. Previous theoretical studies reported that the band gaps of GO can be modulated by different surface oxidation.^43–45 Thus, we propose the electronic properties of WS₂/GO heterostructures can be efficiently controlled by changing the oxygen functionalities on the GO surface. The variation of band gaps of the WS₂/GO heterostructures with different oxygen concentration are shown in Fig. 2b. For the case of WS₂ with h₂e-GO, the band gap increases linearly with increase in the oxygen concentration, and finally reaches the maximum value of 1.91 eV with an oxygen coverage of 80%, which is close to the band gap of pristine WS₂. The band gaps of the pure h₂e-GO structures without WS₂ also go up as the oxygen concentration increases but all of them are greatly larger than that of h₂e-GO with WS₂ (see the Fig. S4 in the ESI†). For example, a higher band gap of 3.11 eV is obtained for pure h₂e-GO at the oxygen coverage of 80%. Thus, electronic coupling between WS₂ and GO is proven and influences the bond length of both WS₂ and GO. For instance, take WS₂ and h₂e-GO heterostructure with the oxygen concentration of 80% as an example: the bond lengths of W–S and O–C are elongated by about 0.4% and 1.4%, respectively, in contrast to the free-standing ones. In order to make sure that the band gap variation is the general rule, more structural models have been taken into account. We have considered the other two types of GO models with hydroxyl to epoxy ratios of 1 : 1 and 4 : 1 (the structures can be seen in Fig. S5 in the ESI†). The band gaps are calculated and plotted by the dashed line in Fig. 2b. As it is shown, the trends of band gaps in both WS₂/he-GO and WS₂/h₄e-GO heterostructures are similar to those of the WS₂/h₂e-GO structures, but only with different values. Importantly, we found that the band gaps of WS₂/GO are dependent on the hydroxyl concentrations in GO, and GO with larger hydroxyl to epoxy ratio possesses larger band gaps, such as GO with a hydroxyl to epoxy ratio of 4 : 1. For example, at the same oxygen concentration of 48%, the band gaps of he-GO, h₂e-GO and h₄e-GO with WS₂ are 0.93 eV, 1.53 eV and 1.80 eV, respectively. Different from the WS₂ on GO with both hydroxyl and epoxy, the band gaps of both WS₂ with h₂-GO and e-GO exhibit an “M” shape as the oxygen concentration increases. Two peaks in the “M” shape correspond to the formation of chain-like structures, and their structures are shown in Fig. S4 of the ESI.† For the WS₂ and e-GO, the band gaps on the two peaks are 1.69 eV and 1.18 eV at the oxygen concentration of 28% and 72%, respectively, which are slightly smaller than those of the WS₂ and h₂-GO (1.77 eV and 1.49 eV at the oxygen concentration of 32% and 72%). The peak valley of the band gaps is present with the oxygen concentration at about 50%, corresponding to incomplete-chain GO structures (see the Fig. S5 in the ESI†). We also examined the electronic properties of the pure GO structures and made a comparison, which is shown in Fig. S4 in the ESI.† We find that the tendency of the band gaps of WS₂ on h₂-GO or e-GO is similar to that of GO without WS₂ at low oxygen concentration, but is different when the oxygen concentration is beyond 70% due to the relatively stronger interaction between WS₂ and GO at high oxygen concentration. Therefore, to a certain extent, the electronic properties of WS₂ and GO hybrids are dependent on the atomic structures and oxygen concentrations of GO. Due to tungsten being a heavy element, we also considered the calculations with spin–orbit coupling (SOC) effect and compared with the results without SOC effect (see Table S1, ESI†). The band gaps of WS₂ and GO heterostructures calculated using the PBE functional with SOC effect are all smaller than those without considering SOC effect, but the difference is very small. For example, the band gap of WS₂/h₂e-GO with oxygen concentration of 40% is 1.15 eV with PBE plus SOC effect, slightly smaller than 1.29 eV without considering SOC effect.

To further understand the electronic properties of WS₂/GO heterostructures, the band structures and the partial charge density distributions for pristine WS₂ and WS₂/GO hybrids are plotted in Fig. 3. Monolayer WS₂ is a semiconductor with a direct band gap at the K point, which is consistent with previous studies.^46,47 From the partial charge density distribution, the conduction band minimum (CBM) and valence band maximum (VBM) of WS₂ are both dominated by d-orbital states of W atoms in a ligand field. For WS₂/GO heterostructures, the band structures vary for different oxygen functionalities and concentrations. They are either indirect or direct semiconductors, which depend on the energy band structures of GO. For example, as shown in Fig. 3, when the oxygen concentration is 40%, WS₂ on h₂-GO and e-GO are semiconductors with direct band gaps of 1.52 eV and 1.54 eV, respectively. We also observed that the CBM and VBM are separated and distributed on different layers (see the inset figures in Fig. 3). In the case of WS₂ and h₂e-GO with an oxygen coverage beyond 40%, the CBM is distributed on GO while the VBM is located on WS₂. But, it is different from the cases of WS₂ on both h₂-GO and e-GO surfaces. In these two cases, both CBM and VBM distribute on GO at an oxygen concentration smaller than 80%, and beyond 80%, the CBM and VBM for the WS₂/GO structures are spatially separated and distributed on WS₂ and GO, respectively.

Fig. 3 Band structures of (a) monolayer WS₂ and (b–f) WS₂/GO heterostructures. The inset figure shows the partial charge density at the CBM and VBM.

Work function (WF), which represents the energy difference between the vacuum energy level and the Fermi energy level, is an important factor for materials to be used in electronic devices.^48–50 The vacuum energy level is defined as the averaged effective potential across the plane that is parallel to the heterostructure, and 12 Å away from the nearest atomic layers of the heterostructure in the supercell. Our calculated work function for monolayer WS₂ is 5.65 eV, which agrees well with the previous theoretical result of 5.89 eV.⁵¹ The calculated work functions for WS₂ and GO with different oxygen functionalities and concentrations are summarized and plotted in Fig. 4. Obviously, the WF of WS₂/GO hybrids can be efficiently modulated by the surface oxidation of GO and the WF tendencies are similar to those of the band gaps. For WS₂ and h₂e-GO hybrids, the WF gradually goes up from 4.78 eV to 5.55 eV, finally close to the WF of WS₂. In order to make sure that the work function variation is the general rule, we have also calculated the other two types of GO models with hydroxyl to epoxy ratios of 1 : 1 and 4 : 1. As is shown in Fig. 4, the work functions for WS₂/he-GO and WS₂/h₄e-GO heterostructures are almost the same as those of the WS₂/h₂e-GO structures. The WF variations of both WS₂/e-GO and WS₂/h₂-GO show the “M” shape as the band gaps. It is worth noting that the WF of WS₂/e-GO hybrids becomes larger than that of pure WS₂ when the oxygen concentration is beyond 20%, while the WF of WS₂/h₂-GO hybrids are all smaller compared to that of pure WS₂ at the whole range of oxygen concentrations. The lowest WF value of 4.09 eV occurs in WS₂/h₂-GO which fully oxidized with hydroxyl groups, and the largest WF values reach 6.34 eV at the WS₂/e-GO with the only epoxy concentration at 80%. The tunable WF values modulated by GO with different oxidations can efficiently modulate the carrier concentration of WS₂, and there is the possibility to further modulate the carrier mobility by filling in the traps in WS₂ with abundant carriers. This promotes WS₂ and other TMD materials as promising electronic devices.

Fig. 4 Variations of work function of WS₂/GO heterostructures as the oxygen concentration increases.

Conclusions

In summary, we investigated the structures and electronic properties of WS₂ and GO heterostructures using first-principles calculations, and the interlayer binding energies, band structures and work functions of WS₂ and GO with various oxygen functionalities and concentrations have been obtained. The energies are not sensitive to different WS₂/GO stacking configurations, and we chose the H configuration for the calculations. We considered three conditions for the GO structures: only epoxy (e-GO), only hydroxyl (h₂-GO) and both epoxy and hydroxyl (he-GO, h₂e-GO and h₄e-GO) on the graphene surface. As the concentration of the oxygen groups increases, the binding energies between WS₂ and GO gradually increase, and e-GO possess the largest binding energies with WS₂ compared to the other GO structures. The band gaps of the WS₂/GO hybrids can be efficiently modulated in a wide range from 0.13 eV to 1.91 eV, which are dependent on the atomic structures and oxygen concentrations of GO. Based on the analysis of the band structures and the partial charge density distributions, the separated CBM and VBM for the WS₂/GO hybrids are observed, and they are distributed on different layers at certain oxygen concentrations. In addition, we also calculate the work functions of different WS₂/GO heterostructures, which possess the similar tendency of the band gaps. The lowest WF value of 4.09 eV occurs at the WS₂ and GO fully oxidized by hydroxyl groups, while the largest WF values can reach 6.34 eV for the WS₂ and GO with the only epoxy concentration at 80%. Our theoretical results provide an effective way to modulate the electronic properties of 2D TMD nanomaterials and make them ideal for high-performance electronic devices.

Acknowledgements

We acknowledge the support from the National Natural Science Foundation of China (Grant No. 21403146, 21273158, 21673149), Natural Science Foundation of Jiangsu Province (Grant No. BK20140314), a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). This is also a project supported by the Fund for Innovative Research Teams of Jiangsu Higher Education Institutions, Jiangsu Key Laboratory for Carbon-Based Functional Materials and Devices, and the Collaborative Innovation Center of Suzhou Nano Science and Technology.

References

1. M. Kin Fai, H. Keliang, L. Changgu, L. Gwan Hyoung, H. James, T. F. Heinz and S. Jie, Nat. Mater., 2012, 12, 207–211.
2. X. Duan, C. Wang, A. Pan, R. Yu and X. Duan, Chem. Soc. Rev., 2015, 44, 8859–8876.
3. D. Jariwala, V. K. Sangwan, L. J. Lauhon, T. J. Marks and M. C. Hersam, ACS Nano, 2014, 8, 1102–1120.
4. B. Radisavljevic, M. B. Whitwick and A. Kis, ACS Nano, 2011, 5, 9934–9938.
5. J. Pu, Y. Yomogida, K. K. Liu, L. J. Li, Y. Iwasa and T. Takenobu, Nano Lett., 2012, 12, 4013–4017.
6. A. K. Singh, K. Mathew, H. L. Zhuang and R. G. Hennig, J. Phys. Chem. Lett., 2015, 6, 1087–1098.
7. Y. Jing, Z. Zhou, C. R. Cabrera and Z. Chen, J. Phys. Chem. C, 2013, 117, 25409–25413.
8. L. Yang, K. Majumdar, H. Liu, Y. Du, H. Wu, M. Hatzistergos, P. Y. Hung, R. Tieckelmann, W. Tsai, C. Hobbs and P. D. Ye, Nano Lett., 2014, 14, 6275–6280.
9. K. He, C. Poole, K. F. Mak and J. Shan, Nano Lett., 2013, 13, 2931–2936.
10. N. Lu, H. Guo, L. Li, J. Dai, L. Wang, W. N. Mei, X. Wu and X. C. Zeng, Nanoscale, 2014, 6, 2879–2886.
11. H. Pan, J. Phys. Chem. C, 2014, 118, 13248–13253.
12. H. Zhang, L.-M. Liu and W.-M. Lau, J. Mater. Chem. A, 2013, 1, 10821.
13. W. J. Yu, Z. Li, H. Zhou, Y. Chen, Y. Wang, Y. Huang and X. Duan, Nat. Mater., 2013, 12, 246–252.
14. H. L. Zhuang and R. G. Hennig, J. Phys. Chem. C, 2013, 117, 20440–20445.
15. D. Braga, I. Gutierrez Lezama, H. Berger and A. F. Morpurgo, Nano Lett., 2012, 12, 5218–5223.
16. K. K. Kam and B. A. Parkinson, J. Phys. Chem., 1982, 86, 463–467.
17. S. Jo, N. Ubrig, H. Berger, A. B. Kuzmenko and A. F. Morpurgo, Nano Lett., 2014, 14, 2019–2025.
18. X. H. Wang, J. Q. Ning, Z. C. Su, C. C. Zheng, B. R. Zhu, L. Xie, H. S. Wu and S. J. Xu, RSC Adv., 2016, 6, 27677–27681.
19. A. K. Geim and I. V. Grigorieva, Nature, 2013, 499, 419–425.
20. X. Duan, C. Wang, J. C. Shaw, R. Cheng, Y. Chen, H. Li, X. Wu, Y. Tang, Q. Zhang and A. Pan, Nat. Nanotechnol., 2014, 9, 1024–1030.
21. S. Tongay, W. Fan, J. Kang, J. Park, U. Koldemir, J. Suh, D. S. Narang, K. Liu, J. Ji, J. Li, R. Sinclair and J. Wu, Nano Lett., 2014, 14, 3185–3190.
22. Y. Gong, J. Lin, X. Wang, G. Shi, S. Lei, Z. Lin, X. Zou, G. Ye, R. Vajtai and B. I. Yakobson, Nat. Mater., 2014, 13, 1135–1142.
23. Y. Li, H. Wang, L. Xie, Y. Liang, G. Hong and H. Dai, J. Am. Chem. Soc., 2011, 133, 7296–7299.
24. H. Li, K. Yu, C. Li, Z. Tang, B. Guo, X. Lei, H. Fu and Z. Zhu, Sci. Rep., 2015, 5, 18730.
25. J. Zhang, Q. Wang, L. Wang, X. Li and W. Huang, Nanoscale, 2015, 7, 10391–10397.
26. Z. Wang, T. Chen, W. Chen, K. Chang, L. Ma, G. Huang, D. Chen and J. Y. Lee, J. Mater. Chem. A, 2013, 1, 2202–2210.
27. Y. Liu, J. Mater. Chem. A, 2014, 2, 13109–13115.
28. X. Xie, Z. Ao, D. Su, J. Zhang and G. Wang, Adv. Funct. Mater., 2015, 25, 1393–1403.
29. S. Ratha and C. S. Rout, ACS Appl. Mater. Interfaces, 2013, 5, 11427–11433.
30. J. Yang, D. Voiry, S. J. Ahn, D. Kang, A. Y. Kim, M. Chhowalla and H. S. Shin, Angew. Chem., Int. Ed., 2013, 52, 13751–13754.
31. Y. Ma, Y. Dai, M. Guo, C. Niu and B. Huang, Nanoscale, 2011, 3, 3883–3887.
32. R. Sharma, J. Aneesh, R. K. Yadav, S. Sanda, A. R. Barik, A. K. Mishra, T. K. Maji, D. Karmakar and K. V. Adarsh, Phys. Rev. B: Condens. Matter Mater. Phys., 2016, 93, 155433.
33. S. S. Li, C. W. Zhang and W. X. Ji, Mater. Chem. Phys., 2015, 164, 150–156.
34. T. Musso, P. V. Kumar, A. S. Foster and J. C. Grossman, ACS Nano, 2014, 8, 11432–11439.
35. G. Kresse and J. Furthmüller, Phys. Rev. B: Condens. Matter Mater. Phys., 1996, 54, 11169–11186.
36. G. Kresse and J. Furthmüller, Comput. Mater. Sci., 1996, 6, 15–50.
37. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
38. M. Dion, H. Rydberg, E. Schroder, D. C. Langreth and B. I. Lundqvist, Phys. Rev. Lett., 2004, 92, 246401.
39. Y. Ma, Y. Dai, W. Wei, C. Niu, L. Yu and B. Huang, J. Phys. Chem. C, 2011, 115, 20237–20241.
40. S. S. Li and C. W. Zhang, J. Appl. Phys., 2013, 114, 183709.
41. L. Wang, Y. Y. Sun, K. Lee, D. West, Z. F. Chen, J. J. Zhao and S. B. Zhang, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 82, 161406.
42. J. Yan and M. Y. Chou, Phys. Rev. B: Condens. Matter Mater. Phys., 2015, 82, 1303–1307.
43. T.-F. Yeh, J.-M. Syu, C. Cheng, T.-H. Chang and H. Teng, Adv. Funct. Mater., 2010, 20, 2255–2262.
44. J. A. Yan, L. Xian and M. Y. Chou, Phys. Rev. Lett., 2009, 103, 086802.
45. J. Ito, J. Nakamura and A. Natori, J. Appl. Phys., 2008, 103, 113712.
46. T. Georgiou, R. Jalil, B. D. Belle, L. Britnell, R. V. Gorbachev, S. V. Morozov, Y. J. Kim, A. Gholinia, S. J. Haigh and O. Makarovsky, Nat. Nanotechnol., 2013, 8, 100–103.
47. S. Ratha, A. J. Simbeck, D. J. Late, S. K. Nayak and C. S. Rout, Appl. Phys. Lett., 2014, 105, 243502.
48. R. C. Shallcross, T. Stubhan, E. L. Ratcliff, A. Kahn, C. J. Brabec and N. R. Armstrong, J. Phys. Chem. Lett., 2015, 6, 1303–1309.
49. P. Wang, D. Tanaka, S. Ryuzaki, S. Araki, K. Okamoto and K. Tamada, Appl. Phys. Lett., 2015, 107, 151601.
50. M. Zhou, R. Q. Png, S. H. Khong, S. Sivaramakrishnan, L. H. Zhao, L. L. Chua, R. H. Friend and P. K. H. Ho, Appl. Phys. Lett., 2012, 101, 1–7.
51. N. A. Lanzillo, A. J. Simbeck and S. K. Nayak, J. Phys.: Condens. Matter, 2015, 27, 175501.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6tc04487e

---