STEM beam effect
Electron beam、Black phosphorus…

Scanning Transmission Electron Microscope (STEM)

Manipulating and engineering atom dynamics with single atom precision has long been the ultimate goal in nanoscience and nanotechnology [1]. Direct atom manipulation or beam induced structural evolution by scanning transmission electron microscopy (STEM) has several advantages, i.e. much faster speed of atom manipulation (a few tens of seconds) and manipulation at room temperature [1]. STEM has been applied in various research, such as, structural defects [2], phase transitions [3], electron beam lithography [4], assembly of atoms [5] and single-atom migration [6].

Schematic of scanning Transmission Electron Microscope (STEM) System

When electrons pass through atomically thin 2D materials, energy can be either elastically or inelastically transferred to the targeted atom. During an elastic collision, impinging electrons are scattered to high angles by the recoiling nucleus and energy and momentum are both conserved. During the inelastic scattering process, the incident electron interacts with the electronic system of the target atom resulting in electronic excitation and electrostatic charging [1]

Phase Patterning in 2D Materials

Fig.1 a) The scheme of 2D ReS2 phase transition under STEM. a,b and a + b are the three low index directions of ReS2. e– beam exposure creates a new T phase embedded in the pristine T′ phase. b–d) Atomic structures and electronic structures of T’’ (tetramerization in two directions) phase, T’ (dimerization in one direction) phase, and T (no dimerization) phase from DFT calculation, respectively.

Advanced science

Sub-Nanometer Electron Beam Phase Patterning in 2D Materials

Fig.2 STEM HAADF images of atomic-scale phase transition from pristine T’’ phase into 1D T’ or T phases, via 1D e– beam exposure direction along a, b and a+b crystal directions (scheme on the right), respectively. False color is applied to STEM images. e– beam scanning areas are marked by green and red boxes. Scale bars =1 nm.

Fig.3 c) Energy-Surface Area (E-S) relations of T, T’ and T’’ phase under the uniaxial strain along a crystal direction. Different phases are shown by different symbols: T phase, green squares; T’ phase, orange dots; T’’ phase, blue triangles. d) E-S relations of T, T’, and T’’ phase under biaxial strain. Tangent lines are presented by the gray dotted lines.

Our DFT calculations reveal the energy-surface (E-S) relations of the three phases in 1L ReS2 under strain (Fig.3 c,d). In terms of the uniaxial case, the stability superiority of the T’’ phase reduces upon a compressive strain along lattice direction a and a crossover of the total energies of T’’and T’ phases is found. Yet the T phase remains very unstable under uniaxial compressive strain, and it becomes the most stable phase when a biaxial strain is applied. The transition lattice constants are comparable with the experimentally derived lattice constants measured. Formation energies of S vacancies (single and bi- vacancies) and their associated displacement threshold energies (Td) of 1L ReS2 were revealed by DFT calculations. It indicates S-3 vacancy is the easiest one to be created, which is consistent with the experimental observation.

This work demonstrates that down to atomic precision, the focused e– beam patterning technique is capable of engineering the metallic T or T’ phase from 1D line to 2D surface at both grain domains and boundaries on the semiconducting T’’ phased ReS2 and ReSe2 monolayers. It provides an ideal patterning precision up to the sub-Å scale after aberration correction and results in phase patterning areas from several to ≈100 nm2, which is orders of magnitude greater than any conventional lithography techniques.

Selective linear etching

Chinese Physics B

Selective linear etching of monolayer black phosphorus using electron beams

Fig.4 (a) Top and side views of atomic structure of monolayer BP (V0P). The names of the zigzag-like chains and two tested atoms are marked. The upper (colored in plum) and lower (colored in light coral) chains are named nT (n is the order number of the chain) and nD, respectively. The P atoms in the upper and lower sublayers are named PnTm (m is the order number of the atom) and PnDm, respectively. (b) and (c) Trajectories of two tested P atoms in pristine monolayer BP under an FHEEB. (d) Top and side views of the atomic structure of a single-atom vacancy BP (V1P) and all five tested P atoms. (e) Calculated cross-sections for the tested atoms in pristine monolayer BP (V0P) and single-atom vacancy monolayer BP (V1P).

Fig.5 Electrical properties of predicted zigzag chain vacancy in monolayer BP. (a) Band structure and density of states of the chain vacancy. (b) PCD at bands MB1 and MB2, DCD, and atomic structure of the zigzag edge chain. (c) Band structure of double-periodic chain vacancies with and without up-and-down distortion. (c) PCD at bands MB1 and MB2, DCD, and atomic structure of the zigzag edge chain. (d) Top view (left) and side view (right) of the atomic structure of the chain vacancy with distortion.

A special zigzag chain vacancy (Fig 5d) in monolayer BP was predicted by using high-energy electron beams (FHEEBs) and knocking away P atoms one by one along a zigzag chain in the lower sublayers. The calculated electronic properties of the chain vacancy showed that there was quasi-bonding between the two edges of the vacancy (Fig 5b), and a CDW was also formed along the vacancy. Our findings help improve understanding of quasi-bonding in which covalent-like states can also be half-occupied. The chain vacancy was a dynamically stable but thermodynamically metastable state according to our comparison of the stabilities of five typical edges in monolayer BP. It was inspiring that the electron beam could create a dynamically mostly stable but thermodynamically metastable vacancy, which is difficult to obtain using conventional chemical synthesis methods but easier to achieve using an electron beam. This characteristic proves that an FHEEB can create a special environment for defect development.

This simulation was implemented using a self-developed tool aBEST (https://gitee.com/jigroupruc/aBEST). This work is expected to inspire further works that will implement more exciton modeling methods into the simulation protocol and thus provide detailed theoretical guidance for future experiments in the field of 2D material etching by FHEEBs.

REFERENCES

1. Zhao, X.; Loh, K. P.; Pennycook, S. J. Electron Beam Triggered Single-Atom Dynamics in Two-Dimensional Materials. J. Phys.: Condens. Matter 2020, 33 (6), 063001. https://doi.org/10.1088/1361-648X/abbdb9.

2. Molecular Beam Epitaxy of Highly Crystalline MoSe2 on Hexagonal Boron Nitride | ACS Nano, https://pubs.acs.org/doi/full/10.1021/acsnano.8b04037.

3. X. Zhao, Y. Ji, J. Chen, W. Fu, J. Dan, Y. Liu, S. J. Pennycook, W. Zhou, and K. P. Loh, Healing of Planar Defects in 2D Materials via Grain Boundary Sliding, Advanced Materials 31, 1900237 (2019).

4. Atomic Structure and Formation Mechanism of Sub-Nanometer Pores in 2D Monolayer MoS 2 – Nanoscale (RSC Publishing) DOI:10.1039/C7NR01127J, https://pubs.rsc.org/en/content/articlehtml/2017/nr/c7nr01127j.

5. O. Dyck, S. Kim, E. Jimenez-Izal, A. N. Alexandrova, S. V. Kalinin, and S. Jesse, Building Structures Atom by Atom via Electron Beam Manipulation, Small 14, 1801771 (2018).

6. Electron-Beam Manipulation of Silicon Dopants in Graphene | Nano Letters, https://pubs.acs.org/doi/full/10.1021/acs.nanolett.8b02406.

Ji Group@Renmin University