Qinglong
Fang
ab,
Min
Li
a,
Xumei
Zhao
a,
Lin
Yuan
a,
Boyu
Wang
a,
Caijuan
Xia
*a and
Fei
Ma
*b
aSchool of Science, Xi’an Polytechnic University, Xi’an, 710048, Shaanxi, China. E-mail: caijuanxia@xpu.edu.cn
bState Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an, 710049, Shaanxi, China. E-mail: mafei@mail.xjtu.edu.cn
First published on 13th November 2021
First principles calculations are performed to study the effects of the interlayer distance and biaxial strain on the electronic properties and contact properties of graphene/MoS2 heterostructures. The interlayer interaction is weakened and the charge transfer from the graphene layer to the MoS2 layer is reduced with increasing interlayer distance in graphene/MoS2 heterostructures, resulting in a shift of the Fermi level to a high energy state. The n-type Schottky barrier is formed with ΦSB,N values of 0.647 eV, 0.568 eV, 0.509 eV, and 0.418 eV when the interlayer distances are 3.209 Å, 3.346 Å, 3.482 Å, and 3.755 Å, respectively. The interlayer distance and charge density difference change slightly, but the electronic structure of the graphene/MoS2 heterostructure changes obviously by applying the biaxial strain. For the biaxial strain from −4% to +6%, the ΦSB,P gradually increases for the graphene/MoS2 heterostructure, while the ΦSB,N increases initially and then decreases. Moreover, the ΦSB,N is only 0.080 eV under a biaxial strain of +6%, indicating that the Ohmic contact is nearly formed. The results demonstrate the significant effects of a biaxial strain on the physical properties of 2D heterostructures.
An important milestone is the creation of heterostructures based on graphene and other two dimensional (2D) materials, which can be assembled into three dimensional stacks with atomic layer precision.7–9 Such layered structures have already demonstrated a range of fascinating physical phenomena. Wang et al. found that for the MoS2/ZnO heterostructure the strong optical absorption in the visible region indicates that it has potential for application in photovoltaic and photocatalytic devices.10 The type-II band alignment occurs at MX2/graphene-like zinc oxide interfaces, together with the large built-in electric field across the interface.11,12 Due to the inherent weak absorption characteristics and small built-in potential of 2D material photodetectors, their external quantum efficiency is severely limited to the range of ∼0.1–1%.13,14 Duan et al.15 modulated the amplitude and polarity of photocurrent in the gated vertical garphene/MoS2 heterostructures via the electric field of an external gate. The maximum external quantum efficiency and internal quantum efficiency are estimated to be 55% and 85%, respectively. Moreover, graphene/MoS2 heterostructures display highly sensitive photodetection and gate tunable persistent photoconductivity.16 The responsivity of the heterostructures is nearly 1 × 1010 A W−1 at 130 K and 5 × 108 A W−1 at room temperature. The heterostructures could also function as a rewritable optoelectronic switch or a memory device when irradiated by the time-dependent photoillumination, where the persistent state shows almost no relaxation or decay within experimental timescales, indicating near-perfect charge retention.
Compared with metal/2D material heterostructures, the use of 2D materials to construct van der Waals vertically stacked heterostructures is one of the effective and feasible ways to reduce the effect of Fermi level pinning. Graphene has potential applications in 2D transistors due to its excellent electrical conductivity, for example, in high-performance devices and circuits based on graphene/MoS2 heterostructures, where MoS2 is used as the transistor channel and graphene as contact electrodes.17 As we all know, there are two mechanisms by which charge can be injected into a semiconductor: thermionic emission over the semiconductor and field emission across the semiconductor. The thermionic emission–diffusion theory describes the current–voltage characteristics of a metal/semiconductor heterostructure as a function of the Schottky barrier height. Carrier recombination can also be a current-limiting process if an inversion layer is present near the contact. Contrary to the bulk case, where the diffusion region extends both laterally and vertically into the semiconductor, in a metal/2D semiconductor heterostructure with no hybridization, the position of the bands vary only laterally, so that charge carriers injected far from the contact edge first encounter the flat-band region before the diffusion region. In this case, the relative contributions from thermionic emission and tunnelling become difficult to predict. On the basis of a thermionic field emission model, the barrier height at the graphene/MoS2 heterostructure was determined to be 0.23 eV and the tunability of graphene work function with electrostatic doping significantly improves the Ohmic contact.18 Wei-Qing Huang et al.19 found that the transformation from an n-type Schottky barrier to a p-type Schottky barrier can be realized in MoSxSe(2−x)/graphene heterostructures when the Se concentration is greater than 25%. Interestingly, the Schottky type, Schottky barrier height, and contact types at the interface can be tuned by an external electric field.19,20 On the other hand, because of their excellent Young's modulus, graphene and MoS2 hold promise for applications in flexible electronic devices. A highly flexible transistor was developed based on an exfoliated MoS2 channel and CVD-grown graphene source/drain electrodes, and a low Schottky barrier (∼22 meV) forms.21–24 The graphene/MoS2 heterostructures exhibit a Young's modulus that is about three times that of monolayer MoS2, while correspondingly exhibiting a yield strain that is about 30–40% smaller than that of monolayer MoS2 due to lateral buckling of the outer graphene layers.25 Zhou et al.26–28 found that the tensile strain can enhance the optical absorption in the visible range and increase the solar energy conversion efficiency for 2D heterostructures.26–28 Moreover, with an appropriate compressive strain of 2% and 3%, the WTe2-As heterostructures show transition from the type-I to type-II band alignment, which could slow down electron–hole pair recombination.29 However, to date the effects of the strain on the structural and electronic properties of the graphene/MoS2 heterostructures have not yet been studied systematically.
Herein, first principles calculations are done to systematically study the effects of the interlayer distance and biaxial strain on the electronic structure and contract properties of the graphene/MoS2 heterostructures. Although the charge transfers from the graphene layer to MoS2 layer, the electronic band structure seems to be a simple sum of those of each constituent when graphene stacks on top of the MoS2 layer. And the contact properties are insensitive to the interlayer distance. While the biaxial strain engineering is a valuable method to modulate the electronic structure and contact properties of the graphene/MoS2 heterostructures.
The thermodynamic stability of graphene/MoS2 heterostructures is evaluated by calculating the binding energy (Eb) as shown in the following equation:
Eb = [Egraphene/MoS2 − (Egraphene + EMoS2)]/NC, | (1) |
Δρ = ρgraphene/MoS2−ρgraphene−ρMoS2, | (2) |
The equilibrium structure is yielded by applying the two-step optimization on the designed lattice. For step one, the interlayer distance is changed and the binding energy was calculated, and also the results were fitted into the well-known Buckingham potential equation:
(3) |
Next, the calculated electronic band structure of the graphene/MoS2 heterostructure is displayed in Fig. 2(c). For comparison, the electronic band structures of pristine graphene and isolated MoS2 monolayers are also shown in Fig. 2(a and b). For the graphene/MoS2 heterostructure, the electronic band structure seems to be a simple sum of those of each constituent. The linear dispersion bands of graphene appear in the large energy gap of MoS2, and the character of the electronic band structure of pristine graphene seems to be preserved, indicating the weak interaction between graphene and MoS2. In Fig. 3(a), the band alignment is schematically demonstrated to show the formation of interlayer excitons. In particular, excited electrons from graphene migrate to MoS2, while holes go oppositely owing to the difference between their Fermi levels, and the electrons and holes can be held together by strong Coulomb interactions. The charge difference of graphene/MoS2 heterostructure is calculated. As expected, the charge redistribution mainly occurs at the interface between the layers, with an accumulation of electrons on the MoS2 layer and a depletion of charges on the graphene layer [Fig. 3(b and c)]. Furthermore, it can be found that the potential of graphene (8.881 eV) in the graphene/MoS2 heterostructure is deepened as shown in Fig. 3(d). The large potential drop of the grapene/MoS2 heterostructure indicates a strong electronic field across the interface, which may affect the kinetics of photo-generated carriers.44
Fig. 2 Band structures of (a) graphene and (b) MoS2 monolayers, as well as (c) graphene/MoS2 heterostructure. The Fermi level is indicated by the green line. |
Fig. 4 displays the charge density difference of the graphene/MoS2 heterostructure as the interlayer distance changes. From Fig. 4(a), one can observe that the fluctuation of the differential charge density curve increases when the interlayer distance is reduced from 3.755 Å to 3.209 Å, indicating the more charge transfers from the graphene layer to the MoS2 layer. Moreover, according to Mulliken charge analysis, 0.120 e, 0.107 e, 0.088 e, and 0.062 e transfer from the graphene layer to MoS2 layer when the interlayer distances are 3.209 Å, 3.346 Å, 3.482 Å, and 3.755 Å, respectively. The Fermi level of the graphene shifts downwards due to its charge transfers to the MoS2 layer, while the electronic band structures of pristine graphene and MoS2 monolayers are preserved. The band gaps of the MoS2 layer in heterostructures are 1.747 eV, 1.737 eV, 1.747 eV, and 1.757 eV when the interlayer distances are 3.209 Å, 3.346 Å, 3.482 Å, and 3.755 Å, respectively (Fig. 5). All these results indicate that the interlayer distance is effective to control the charge transfer, while limited to regulate the electronic structure (Table 1).
Fig. 4 (a) Plane averaged charge density difference and (b) three dimensional isosurface of the charge density difference of the graphene/MoS2 heterostructure at various interlayer distances. |
Fig. 5 Band structures of the graphene/MoS2 heterostructure at various interlayer distances: (a) 3.209 Å, (b) 3.346 Å, (c) 3.482 Å, and (d) 3.755. |
Strain (%) | d (Å) | L C–C (Å) | L Mo–S (Å) | E b (meV C−1) |
---|---|---|---|---|
−4 | 3.246 | 1.402 | 2.381 | −10 |
−2 | 3.386 | 1.430 | 2.393 | −55 |
0 | 3.414 | 1.460 | 2.406 | −39 |
+2 | 3.423 | 1.489 | 2.421 | 47 |
+4 | 3.437 | 1.518 | 2.436 | 195 |
+6 | 3.453 | 1.547 | 2.453 | 397 |
ε = (a−a0)/a, | (4) |
Fig. 6 shows the charge density difference of the graphene/MoS2 heterostructure under the biaxial strain. It is found that the fluctuation of the differential charge density curve decreases when the biaxial strain changes from −4% to +6%, indicating the less charge transfers from the graphene layer to MoS2 layer [Fig. 6(a)]. The same phenomenon can observe through the three dimensional isosurface of the charge density difference of the graphene/MoS2 heterostructure as shown in Fig. 6(b). According to Mulliken charge analysis, 0.062 e, 0.059 e, 0.055 e, and 0.051 e are transferred from the graphene layer to the MoS2 layer under a biaxial strain of −4%, −2%, +2%, and +6%, respectively. Thus, by applying biaxial strain, the interlayer distance and charge density difference change slightly. However, there is an obvious change in the electronic structure of the graphene/MoS2 heterostructure by applying the biaxial strain. Although the electronic band structures of pristine graphene is still preserved, the band structure of MoS2 substantial changes under the biaxial strain. In particular, the band gaps of the MoS2 layer in the heterostructure are 1.669 eV, 1.806 eV, 1.352 eV, and 0.618 eV under a biaxial strain of −4%, −2%, +2%, and +6%, respectively (Fig. 7).
Fig. 6 (a) Plane averaged charge density differences and (b) three dimensional isosurface of the charge density difference of the graphene/MoS2 heterostructure under a biaxial strain. |
Fig. 7 Band structures of the graphene/MoS2 heterostructure under a biaxial strain: (a) −4%, (b) −2%, (c) +2%, and (d) +6%. |
It is obvious that the graphene/MoS2 heterostructure is characterized by the metal/semiconductor heterostructure. In the graphene/MoS2 heterostructure, the n-type Schottky barrier height (ΦSB,N) is ΦSB,N = ECBM − EF, whereas the p-type Schottky barrier height (ΦSB,P) is ΦSB,P = EF − EVBM. ECBM, EVBM, and EF represent the conduction band minimum, valence band maximum, and Fermi level, respectively. Fig. 8 displays the evolution of the n- and p-type Schottky barrier height (ΦSB,N and ΦSB,P) of the graphene/MoS2 heterostructure at the various interlayer distance and under the biaxial strain. When the interlayer distance is reduced, the ΦSB,N gradually increases, but ΦSB,P gradually decreases. Moreover, the value of ΦSB,P is always larger than that of ΦSB,N, indicating a n-type Schottky contact. The ΦSB,N values are 0.647 eV, 0.568 eV, 0.509 eV, and 0.418 eV when the interlayer distances are 3.209 Å, 3.346 Å, 3.482 Å, and 3.755 Å, respectively [Fig. 8(a)]. For applying the biaxial strain from −4% to +6%, the ΦSB,P gradually increases for the graphene/MoS2 heterostructure, while the ΦSB,N increases initially and then decreases [Fig. 8(b)]. Furthermore, the ΦSB,N is only 0.080 eV under a biaxial strain of +6%, indicating that the Ohmic contact is nearly formed.
Fig. 8 The evolution of the n- and p-type Schottky barrier height (ΦSB,N and ΦSB,P) of the graphene/MoS2 heterostructure: (a) at various interlayer distances and (b) under a biaxial strain. |
This journal is © The Royal Society of Chemistry 2022 |