Elettra-Sincrotrone Trieste S.C.p.A. website uses session cookies which are required for users to navigate appropriately and safely. Session cookies created by the Elettra-Sincrotrone Trieste S.C.p.A. website navigation do not affect users' privacy during their browsing experience on our website, as they do not entail processing their personal identification data. Session cookies are not permanently stored and indeed are cancelled when the connection to the Elettra-Sincrotrone Trieste S.C.p.A. website is terminated.
More info
OK

Visualizing electrostatic gating effects in two-dimensional heterostructures

Electronic and optoelectronic devices utilise electric fields to manipulate material properties, controlling band structures and band alignments across heterostructures that combine metals, semiconductors and insulators. With two-dimensional materials, 2D heterostructures (2DHS) can be fabricated with atomic precision by simply stacking layers. In these, applied out-of-plane electric fields are a powerful tool that can be used to degenerately dope semiconductors, modify electronic structure through the Stark effect, and alter band-alignments between layers. As a result, out-of-plane electric fields have been used to engineer functional architectures such as high-efficiency light-emitting diodes and tunnelling transistors, and to probe many-body phenomena. 
Despite the fundamental importance of electric-field control over band structure, direct experimental measurements are challenging and have been limited. Whilst gate electrodes are routinely applied for electrical transport investigations, and many studies have reported electric-field dependent light-emission from 2DHS, these depend upon but do not directly reveal the single-particle electronic structure. Angle resolved photoemission spectroscopy (ARPES) has proven to be a powerful tool for probing the momentum-resolved valence band structure of 2D materials such as graphene and semiconducting transition metal dichalcogenides (MX2). But it is challenging to apply conventional ARPES, which typically averages over lengthscales > 100 µm, to 2DHS which are usually only a few µm across. Using the high spatial resolution and flux of the Spectromicroscopy beamline at Elettra, we have shown that submicrometre spatially resolved ARPES (µARPES) can determine band parameters and band alignments across 2DHS of mechanically exfoliated flakes. These heterostructures are similar to those used for optical spectroscopy and transport measurements, opening the way to study operating devices. 
In this work we realised a long held goal: direct momentum-resolved electronic structure measurements of in-operando microelectronic devices. To test the technique, a simple 2DHS of back-gated graphene was constructed from mechanically exfoliated flakes, as shown schematically in Figure 1a with optical images in 1b and 1c. The 2DHS was located by scanning photoemission microscopy (SPEM), using the high spatial resolution of the Spectromicroscopy beamline at Elettra, Figure 1d. The addition of electrical contacts within the analysis chamber allowed the dispersion around the graphene Dirac cone to be measured at varying gate voltage VG, Figure 1e, showing the transition from hole-doped at V< 0 to electron-doped at V> 0. The change in Dirac point with Vis quantitatively as expected, as shown by the agreement between the fit to the data in Figure 1f, validating µARPES with in situ gating.
 

Figure 1.     Visualizing electrostatic gating of monolayer graphene.(a) Schematic of a 2D heterostructure device with a grounded graphene top-contact, encapsulated by hBN, and a graphite back-gate electrode to which the gate voltage VGis applied. (b) Photograph of a device mounted in a standard dual in-line package. (c) Optical microscope image of the dotted box in (b) showing the heterostructure regions, and (d) scanning photoemission microscopy (SPEM) image of the same area (scale bar, 50 µm). (e) Energy-momentum slices near the graphene K-point, at the labelled gate voltages, from -5 V to + 5V. (f) Gate dependence of ED, the solid line is a fit assuming a linear dispersion of graphene.



We next turned to semiconducting MX2, using the gate electrode to electron dope the layers and hence directly measure the band gap by µARPES. A schematic of a WSe2device is shown in Figure 2a, with corresponding optical image Figure 2b, and SPEM map 2c. In monolayer WSe2the valence band maximum (VBM) is a spin-polarised band at the zone corner K, as shown in the schematic in Figure 2d. Energy momentum slices along Γ- K from the 1 layer, 2 layer, and 3 layer regions are shown in Figure 2e-g respectively, with the upper panels showing the dispersion at gate voltage VG= 0, and the lower ones at VG= +3.35 V. With the gate voltage applied, the chemical potential moves into the conduction band and the conduction band edge is revealed. For 1 layer (1L), the conduction band minimum (CBM) is found at K, making it a direct gap semiconductor. For 2L, the VBM is still at but the CBM moves to Q, whilst for 3 or more layers the VBM is at Γ and the CBM at Q, all indirect gaps. Extending this work further, we showed that the band gap of 1L WSeis strongly dependent on doping, showing a renormalisation (reduction) of ~ 400 meV as the carrier concentration increases to 1.5×1013cm-2. Direct comparison to gate-dependent photoluminescence from the same sample revealed a strong decrease in exciton binding energy with carrier concentration. These results demonstrate that in-operando ARPES gives access to the momentum-resolved, layer-resolved, and doping-dependent electronic structure across 2DHS. As such, it will be an important new tool to aid the understanding and development of 2D electronics, spintronics, and optoelectronics.
 

Figure 2.     Layer-number dependent conduction band edge (CBE) in WSe2. (a) Schematic of a 2DHS device for back-gating a WSe2flake. The gate voltage, VG, is applied to the graphite back gate electrode, with a grounded graphene top contact overlapping the WSe2 flake. (b) Optical and (c) SPEM images with 1L, 2L and 3L WSe2regions identified (scale bars, 5 µm). (d) Schematics of the 1L WSe2 Brillouin zone and symmetry points and of its conduction and valence band structure. (e)-(g) Energy-momentum slices along Γ- K for 1L, 2L, and 3L regions respectively: upper panels at VG= 0, and lower ones at VG= +3.35 V. The intensity in the dashed boxes is multiplied by 20. Scale bars, 0.3 Å-1. The data have been reflected about  Γ to aid comparison with electronic structure calculations (GW approximation) (red dashed lines). 

 


This research was conducted by the following research team:

Paul V. Nguyen1,Nathan P. Wilson1, Joshua Kahn1, Xiaodong Xu1,David H. Cobden1, Gabriel C. Constantinescu2, Viktor Kandyba3, Alessio Giampietri3, Alexei Barinov3,Neil R. Wilson4, Natalie C Teutsch4, Abigail J Graham4, Xue Xia4, Nicholas D.M. Hine4

 

Department of Physics, University of Washington, Seattle, Washington, USA
TCM Group, Cavendish Laboratory, University of Cambridge, Cambridge, UK
Elettra - Sincrotrone Trieste, S.C.p.A., Basovizza, Trieste  Italy
Department of Physics, University of Warwick, Coventry, UK

Contact persons:

Neil R. Wilson,  email: neil.wilson@warwick.ac.uk


 

Reference

Paul V Nguyen, Natalie C Teutsch, Nathan P. Wilson, Joshua Kahn, Xue Xia, Abigail J Graham, Viktor Kandyba,Alessio Giampietri, Alexei Barinov, Gabriel C Constantinescu, Nelson Yeung, Nicholas D M Hine, Xiaodong Xu, David H Cobden, Neil R Wilson, “Visualizing electrostatic gating effects in two-dimensional heterostructures” Nature (2019), DOI:10.1038/s41586-019-1402-1

 
Last Updated on Wednesday, 24 July 2019 11:34