Thermomechanical Nanostraining of Two-Dimensional Materials

Abstract

Local bandgap tuning in two-dimensional (2D) materials is of significant importance for electronic and optoelectronic devices but achieving controllable and reproducible strain engineering at the nanoscale remains a challenge. Here, we report on thermomechanical nanoindentation with a scanning probe to create strain nanopatterns in 2D transition metal dichalcogenides and graphene, enabling arbitrary patterns with a modulated bandgap at a spatial resolution down to 20 nm. The 2D material is in contact via van der Waals interactions with a thin polymer layer underneath that deforms due to the heat and indentation force from the heated probe. Specifically, we demonstrate that the local bandgap of molybdenum disulfide (MoS₂) is spatially modulated up to 10% and is tunable up to 180 meV in magnitude at a linear rate of about -70 meV per percent of strain. The technique provides a versatile tool for investigating the localized strain engineering of 2D materials with nanometer-scale resolution.

Keywords: 2D materials; local bandgap; molybdenum disulfide; strain nanopattern; thermal scanning probe lithography; tip-enhanced Raman spectroscopy.

Figure 1

t-SPL based strain nanopatterning in 2D materials (2DMs). (a) Conceptual illustration of the thermomechanical nanoindentation process for strain nanopatterning in the 2DM, such as TMDCs or graphene. Drawing is not to scale. (b) Cross-section scheme showing details of the heated nanotip indenting a monolayer TMDC layer on a PPA layer. Drawing is not to scale. (c) AFM topography of the written MoS₂ ripple nanostructures. (d) Three-dimensional representation of the area marked in panel c. (e) Depth profile of the selected line in panel c. The nanoindentation depth is around 4 nm and the pileup height is around 1.5 nm.

Figure 2

Versatility of the method and application to other 2D materials. (a) AFM topography of an array of nanostripes in 1L MoS₂ with designed width in the range from 2 to 100 nm and corresponding depth profile using the voltage of 7.5 V and the temperature of 950 °C. The writing direction is starting from upper right to bottom left. (b) AFM topography of nanopatterns produced with a heater temperature from 200 to 1200 °C and corresponding depth profile under the voltage of 7.5 V. The designed width is 20 nm. The writing direction is starting from bottom left to upper right. (c) Topography of 2L MoTe₂ nanoripples. (d) Topography of 1L graphene nanowells array. (e) Topography of 1L MoSe₂ nanopattern with arbitrary strain distribution (the EPFL logo is given as an example). Logo is used with permission. The writing direction in panels c, d, and e is starting from bottom right to upper left.

Figure 3

Micro-Raman characterization of the nanopatterned 1L MoS₂. (a) Topography of the nanopatterned 1L MoS₂ created by the t-SPL. (b) Raman spectra of the strained and unstrained MoS₂. Inset: schematic atomic vibration of the in-plane E¹_2g and out-of-plane A_1g modes. (c,d) Scanning Raman spectroscopic maps of the E¹_2g and A_1g peaks of the strained MoS₂, respectively. The laser scanning step in the x-axis direction is 0.2 μm and that in the y-axis direction is 0.1 μm. The Raman shift corresponding to the maximum of each spectral peak is obtained by Lorentzian line fitting.

Figure 4

Tapping-mode AFM-TERS mapping of nanostrained 1L MoS₂. (a,b) Schematic illustration of AFM-TERS mapping of strained 2D material (2DM). Drawing is not to scale. (c) AFM topography of the strained nanoripples in 1L MoS₂. The vertical lines are artifacts caused by high-frequency noise. (d) Height map showing the selected area of the nanopatterned structure in panel c. Each pixel represents an area of 10 nm × 10 nm. The green regions represent the indented MoS₂ and the orange region represents the pileup created during the nanoindentation process (in the gray pixel the TERS spectrum was accidently not measured). (e) TERS map of the A_1g peak corresponding to the height map in d. (f) TERS A_1g peak data of nanopatterned MoS₂ in the valleys and the pileup. (g) Single TERS spectrum of unstrained MoS₂ and the average TERS spectrum of strained MoS₂ in d. The spectra were extracted, smoothed and subsequently analyzed for peak shift.

Figure 5

PL characterization and analysis of bandgap modulation. (a,b) Scanning photoluminescence (PL) spectroscopic maps plotting (a) A-exciton energy peaks and (b) B-exciton energy peaks of the strained MoS₂ sample studied in micro-Raman characterization (Figure 3). The laser scanning step in the x-axis direction is 0.2 μm and that in the y-axis direction is 0.1 μm. (c) PL spectra of the strained and unstrained MoS₂. (d) Bandgap modulation as a function of the average strain. The different strains are obtained by varying the indentation depth with patterns designed to go from 1.8 to 10 nm deep. The designed width of the ripple is 50 nm and the pattern pitch is 120 nm. The average strain is calculated using the model in Supporting Information Section 9.