Supplementary Information:

Photo-tunable transfer characteristics in MoTe2-MoS2 vertical hetero-structures

Arup Kumar Paul, ManabendraKuiri, DipankarSaha, BiswanathChakraborty, Santanu Mahapatra, A.K Soodand Anindya Das

A.Device Characterization details

Here we have presented additional characteristics of the Devices D1, D3 and D4, and details of the DFT calculation. The IDS –VDS characteristics of individual flakes of device D1 is presented in Figure S1. Fig.S1 (a) shows the IDS –VDS response of the MoS2 flake. Though, there exist a slight asymmetry, the response is highly linear. The asymmetry might be because of non-equality of gold contacts with MoS2. The linearity suggests there is no significant Schottky barrier at the contacts. Fig.S1 (b) shows the IDS –VDS response of the few layered MoTe2 flake. As can be seen the response is linear, so good Ohmic contact has been established on MoTe2.

In Figure S2 we show the detail characteristics of the device D3 on which all the opto-electronic measurements were performed. Fig.S2 (a) - (b) are IDS – VDS characteristics of MoS2 and MoTe2 flakes respectively. As can be seen MoS2 characteristics are similar to that of D1 and MoTe2 response is linear. Fig.S2 (c) shows the gate response of the flakes. In this case MoTe2 flake shows ambipolar behavior and MoS2 shows its usual n type behavior. In Fig.S2 (d) we show theIDS- VDS characteristics of the junction in D3. Though this junction also shows rectification behavior, but this is not perfect rectification as D1.



The IDS –VDS response of the device D4 is shown in Figure S3. This device does not show any rectification behavior.

The comparison between electrical measurements with and without light, for individual flakes in D3 is shown in Figure S4.For MoS2, the gate response and IDS-VDS responses are shown in Fig.S4 (a) and (b), respectively. Similarly, MoTe2 responses are plotted in Fig.S4 (c) and (d). It is evident from the figures that for both MoTe2 and MoS2, the responses do not change significantly in presence of light.

In Figure S5 we compare dark and light IDS –VDS response of junction of device D3, for different VBG. It can be seen clearly, that the response under light is dependent on the applied bias voltage and also on the gate-voltage. The dependence of IDS on VBG under illumination shows the photo-doping effect.

B. Computational details:

The software package Atomistix Tool Kit (ATK) has been used to carry out the first-principles based density functional theory (DFT) calculations [1]. We employ the generalized gradient approximation (GGA) exchange correlation along with the Perdew-Burke-Ernzerhof (PBE) functional to derive all the electronic and structural properties [2]. Besides, for the norm-conserving pseudo-potentials, we utilize the OPENMX (Open source package for Material eXplorer) code [3], [4]. The basis sets used for “Mo”, “Te”, and “S” are “s3p2d1”, “s2p2d2f1”, and “s2p2d1” respectively. For the purpose of geometry optimization of the unit cells, we have used the LBFGS (Limited memory Broyden Fletcher Goldfarb Shanno) algorithm (force-tolerance ~ 0.01 eV/Å, maximum stress ~ 0.001 eV/ Å3). Asdescribed in Figure S6 (a)-(b), the optimized in-plane lattice constants for the MoS2 and MoTe2 unit cells are a = b = 3.197 Å and a = b = 3.608 Å respectively. Next, we perform the binding energy calculation and obtain an equilibrium distance of ~ 3.45 Å where the MoTe2 surface binds with the MoS2 monolayer. Nonetheless, the strain mismatch between the MoTe2 and MoS2 monolayers is restricted to a significantly low value (mean absolute strain ~0.16%). The vdW heterostructure of MoS2 and MoTe2 (with a = b = 9.57 Å) has been shown in Fig.S6 (c).


Considering the electronic properties of the fully relaxed structures, we find a direct bandgap of 1.7 eV (which is consistent with the experimental observation of [5], and the other theoretical calculation [6]) for

the MoS2 unit cell (Fig.S7 (a)). However, for the MoTe2, bandgap is comparatively much lower (~ 0.97 eV, as shown in Fig.S7 (b)). The bandstructure of the MoS2-MoTe2 vdW supercell is illustrated in Fig.S7 (c). The composite vdW heterostructure has an effective bandgap of 0.86 eV. For the calculation of electronic properties, we opt for a Monkhorst-Pack grid of 3×3×1 (in X-Y-Z directions). Besides, the density-mesh-cutoff (in the numerical accuracy settings) is taken as 90 Hartree. Apart from that, in order to consider the vdW interactions, we have further included the Grimme’s dispersion correction (DFTD2)
[7].

For electrical transport calculations using the DFT-NEGF (non-equilibrium Green’s function) combination, we set the periodic boundary conditions, in the X-Y directions, whereas in Z direction we take Dirichlet boundary condition (Monkhorst-Pack grid 1×3×99) in order to solve the Poisson’s equation. Instead of multilayer, here we opt for the monolayer MoTe2 to reduce the enormous computational budget. However, the phenomena related to interlayer carrier separation and charge coupling (as described in the main text), will remain valid for the multilayer sample too.

References:

  1. AtomistixToolKit (ATK) with Virtual NanoLab, Version 2015.1, QuantumWise A/S, Available at
  1. J. P. Perdew, K. Burke, and M. Ernzerhof, “Generalized Gradient ApproximationMade Simple,” Phys. Rev. Letters, vol. 77, no. 18, 3865,1996.
  1. T. Ozaki, “Variationally optimized atomic orbitals for large-scale electronic structures,” Phys. Rev. B, 67, 155108, 2003.
  1. T. Ozaki and H. Kino, “Numerical atomic basis orbitals from H to Kr,”Phys. Rev. B, 69, 195113, 2004.
  1. K. F. Mak et al., “Atomically Thin MoS2: A New Direct-Gap Semiconductor,” Phys. Rev. Lett., 105, 136805, 2010.
  1. D. Saha and S. Mahapatra, “Atomistic Modeling of metallic-to-semiconducting phase boundaries in monolayer MoS2,” Appl. Phys. Lett., 108, 253106, 2016.
  1. S. Grimme, “Semiempirical GGA-type density functional constructed with a long-range dispersion correction,” J. Comput. Chem., vol. 27, no. 15, pp. 1787-1799, 2006.