| Literature DB >> 31415148 |
Makars Šiškins1, Martin Lee1, Farbod Alijani2, Mark R van Blankenstein1, Dejan Davidovikj1, Herre S J van der Zant1, Peter G Steeneken1,2.
Abstract
Two-dimensional (2D) materials with strong in-plane anisotropy are of interest for enabling orientation-dependent, frequency-tunable, optomechanical devices. However, black phosphorus (bP), the 2D material with the largest anisotropy to date, is unstable as it degrades in air. In this work we show that As2S3 is an interesting alternative, with a similar anisotropy to bP, while at the same time having a much higher chemical stability. We probe the mechanical and optical anisotropy in As2S3 by three distinct angular-resolved experimental methods: Raman spectroscopy, atomic force microscopy (AFM), and resonance frequency analysis. Using a dedicated angle-resolved AFM force-deflection method, an in-plane anisotropy factor of [Formula: see text] is found in the Young's modulus of As2S3 with Ea-axis = 79.1 ± 10.1 GPa and Ec-axis = 47.2 ± 7.9 GPa. The high mechanical anisotropy is also shown to cause up to 65% difference in the resonance frequency, depending on crystal orientation and aspect ratio of membranes.Entities:
Keywords: 2D materials; Raman spectroscopy; arsenic trisulfide (As2S3); mechanical anisotropy; multimode resonances; nanoelectromechanical systems (NEMS)
Year: 2019 PMID: 31415148 PMCID: PMC6764108 DOI: 10.1021/acsnano.9b06161
Source DB: PubMed Journal: ACS Nano ISSN: 1936-0851 Impact factor: 15.881
Figure 1Crystal structure and atomic force microscopy (AFM) images of As2S3. (a) Crystal lattice of a single-layer As2S3. Projection along the b-axis. Gray box indicates a single unit cell. (b) Projection along the a-axis. Inversion center is indicated. (c) Projection along the c-axis. (d, e) AFM images of a nanometer-thin exfoliated layer of As2S3 on SiO2/Si as fabricated and after ∼3 months in air. Scale bars: 1 μm.
Figure 2(a) Raman spectra of a bulk As2S3 crystal (thickness, t ≈ 50 nm) for a polarization of incident red laser light (λ = 632 nm) along two crystalline axes (a and c). (b) Polarization-dependent Raman intensity spectra obtained by sample rotation with normal incident light (along the b-axis of a crystal). (c) Thickness-dependent Raman spectra under λ = 632 nm excitation. (d–f) Polarization dependence for three Raman-active modes. Vibrational modes in a, b, and d–f are labeled in accordance to previous works[40−45] on As2S3.
Figure 3Detailed angle-resolved study of the Young’s modulus of As2S3. (a) Tapping mode AFM image of the sample. Scale bar: 4 μm. (b) Typical force–deflection curve obtained at different angles of a rectangular cavity with respect to crystalline axes. Top inset: Optical image of the sample. Scale bar: 8 μm. Bottom inset: Principle of the measurement. (c) Angle-resolved effective Young’s modulus of the membrane. Red line is a fit to eq . Young’s moduli for in-plane crystalline directions (E and E) are indicated. (d) Raman spectra along each axis. (e) Effective Young’s modulus along a and c crystalline axes measured for the best five samples of various thickness (from 9 to 25 nm).
Figure 4Measurement setup and measured angle-resolved frequency response of the device. (a) Schematic of the measurement setup. (b) Angle-dependent frequency response of rectangular resonators rotated by 90° with respect to the a-axis. (c) Comparison of anisotropic response observed in motion (laser interferometry, black) and statics (AFM force-indention, red).
Figure 5Change in resonance frequency of vibrational modes of an As2S3 plate. (a) Finite element model (FEM) for the first six modes of a resonator compared to the measured magnitude (M, M) of resonance peaks at 0° and 90° rotation of the rectangular membrane with respect to the c-axis of the crystal. (b) Optical image of a device with orientation and scales indicated. Dimensions are a = 5 μm and b = 10 μm. Scale bar: 5 μm. Both blue and red lasers are focused at the position indicated by the red circle in the lower panel. (c) Vibrational mode frequencies with corresponding errors compared to the FEM model.