Table of Contents
Fetching ...

Influence of sulphur vacancies on ultrafast charge separation in WS$_2$-graphene heterostructures

Johannes Gradl, Niklas Hofmann, Leonard Weigl, Stiven Forti, Neeraj Mishra, Camilla Coletti, Raul Perea-Causin, Ermin Malic, Isabella Gierz

Abstract

Understanding how defects influence charge separation in WS$_2$-graphene heterostructures is crucial for future applications in light harvesting and detection. Previous studies have reported widely varying lifetimes for the charge-separated state, all supposedly linked to electron trapping at sulphur vacancies. The exact impact of these defects, however, has remained unclear. Here, we deliberately introduce sulphur vacancies by annealing the heterostructures at high temperatures in ultrahigh vacuum. Angle-resolved photoemission spectroscopy (ARPES) reveals that these vacancies modify both the band alignment and doping level of the heterostructure. Time-resolved ARPES (trARPES) further shows that increasing the sulphur vacancy concentration prolongs the lifetime of electrons in the WS$_2$ conduction band but shortens the lifetime of the charge-separated state. Guided by model calculations, we attribute this behaviour to shifts in the energy alignment between sulphur vacancy states and graphene's Dirac point, combined with a reduced excitonic absorption. The model also yields a transfer time for electrons tunneling from sulphur vacancies into graphene's Dirac cone of $\sim$4ps, consistent with our trARPES measurements. Our study clarifies the role of sulphur vacancies in WS$_2$-graphene heterostructures, further improving our microscopic understanding of charge dynamics for future optoelectronic applications.

Influence of sulphur vacancies on ultrafast charge separation in WS$_2$-graphene heterostructures

Abstract

Understanding how defects influence charge separation in WS-graphene heterostructures is crucial for future applications in light harvesting and detection. Previous studies have reported widely varying lifetimes for the charge-separated state, all supposedly linked to electron trapping at sulphur vacancies. The exact impact of these defects, however, has remained unclear. Here, we deliberately introduce sulphur vacancies by annealing the heterostructures at high temperatures in ultrahigh vacuum. Angle-resolved photoemission spectroscopy (ARPES) reveals that these vacancies modify both the band alignment and doping level of the heterostructure. Time-resolved ARPES (trARPES) further shows that increasing the sulphur vacancy concentration prolongs the lifetime of electrons in the WS conduction band but shortens the lifetime of the charge-separated state. Guided by model calculations, we attribute this behaviour to shifts in the energy alignment between sulphur vacancy states and graphene's Dirac point, combined with a reduced excitonic absorption. The model also yields a transfer time for electrons tunneling from sulphur vacancies into graphene's Dirac cone of 4ps, consistent with our trARPES measurements. Our study clarifies the role of sulphur vacancies in WS-graphene heterostructures, further improving our microscopic understanding of charge dynamics for future optoelectronic applications.
Paper Structure (8 sections, 5 figures, 1 table)

This paper contains 8 sections, 5 figures, 1 table.

Figures (5)

  • Figure 1: Influence of annealing on equilibrium band structure of WS$_2$-graphene heterostructure. ARPES snapshot of WS$_2$-graphene heterostructure after degassing the sample at 600$^\circ$C in UHV (a) and after four annealing steps at 600$^\circ$C and three annealing steps at 650$^\circ$C (b). Dashed blue lines in (a) and (b) are guides to the eye indicating the equilibrium band structure Zeng2013Wallace1947. (c) Distance between WS$_2$ VB maximum and Dirac point for different annealing steps. (d) Position of the Dirac point for different annealing steps. The blue- and red-shaded areas in (c) and (d) indicate annealing at 600$^\circ$C and at 650$^\circ$C, respectively.
  • Figure 2: Influence of sulphur vacancies on non-equilibrium carrier dynamics. Panel a) shows the pump-induced changes of the photocurrent depicted in Fig. 1a for a pump-probe delay of $\sim$100 fs after photoexcitation at $\hbar\omega_{pump} = 2$ eV with a fluence of 4 mJ/cm$^2$. b) Exponential lifetime of the photoexcited electrons in the WS$_2$ CB (yellow box in a) for each annealing step. c) Exponential lifetime of the photocurrent integrated over the area marked by the black box in (a) for each annealing step. The black box in (a) is consistently placed 25 meV above the equilibrium position of the WS$_2$ VB for all annealing steps. d) Exponential lifetime of the charging shift of the WS$_2$ VB for all annealing steps. Thick lines in (b-d) are guides to the eye.
  • Figure 3: Sketch of band structure of WS$_2$-graphene heterostructure along the $\mathit{\Gamma K}$ direction based on DFT calculations from Hofmann_2DMater2023 and Hernangomez_PRB2023 to illustrate the microscopic model for charge transfer in WS$_2$-graphene developed in Krause_PRL2021. Graphene (WS$_2$) bands are drawn in black (grey). Horizontal yellow lines indicate two mid-gap states originating from sulphur vacancies. Those areas in the Brillouin zone where the states hybridize and charge transfer can occur are highlighted in red. $\Delta E_{CB}$ and $\Delta E_{VB}$ indicate the barriers that electrons and holes need to overcome before they can tunnel from the WS$_2$ CB and VB into the $\pi$-bands of graphene. $\Delta E$ is the energy offset between the VB maximum and the Dirac point from Fig. \ref{['figure1']}c. $E_0$ is the distance between the lower sulphur vacancy level and the Dirac point that determines the charge transfer time from a sulphur vacancy to graphene.
  • Figure 4: Pump-probe traces extracted by integrating trARPES snapshots from Fig. \ref{['figure2']}a over the areas marked by the yellow (a) and black boxes (b) together with exponential fits for all seven annealing steps.
  • Figure 5: Band structure dynamics for all seven annealing steps as indicated in the individual panels. a) Transient position of the WS$_2$ CB at K. b) Transient position of the WS$_2$ VB at $k=1.1$Å$^{-1}$. c) Transient band gap at $K$. d) Charging shift of the WS$_2$ VB under the assumption of a $2.0$ eV equilibrium band gap together with exponential fit (continuous line). e) Same as d) but assuming an equilibrium gap size of 1.9 eV.