Direct Visualization of Large-Scale Intrinsic Atomic Lattice Structure and Its Collective Anisotropy in Air-Sensitive Monolayer 1T'- WTe2

Abstract

Probing large-scale intrinsic structure of air-sensitive 2D materials with atomic resolution is so far challenging due to their rapid oxidization and contamination. Here, by keeping the whole experiment including growth, transfer, and characterizations in an interconnected atmosphere-control environment, the large-scale intact lattice structure of air-sensitive monolayer 1T'-WTe₂ is directly visualized by atom-resolved scanning transmission electron microscopy. Benefit from the large-scale atomic mapping, collective lattice distortions are further unveiled due to the presence of anisotropic rippling, which propagates perpendicular to only one of the preferential lattice planes in the same WTe₂ monolayer. Such anisotropic lattice rippling modulates the intrinsic point defect (Te vacancy) distribution, in which they aggregate at the constrictive inner side of the undulating structure, presumably due to the ripple-induced asymmetric strain as elaborated by density functional theory. The results pave the way for atomic characterizations and defect engineering of air-sensitive 2D layered materials.

Keywords: Te vacancy; WTe2 monolayer; air-sensitive 2D materials; anisotropic ripple; large-scale atomic mapping.

Figure 1

Optical and STEM imaging of air‐sensitive WTe₂ monolayer. Optical images of 1T′‐WTe₂ exposed in air a) right after the growth and b) after 5 min. A clear dimming of the contrast from a) purple to b) pale purple is visible. Optical images of 1T’‐WTe₂ in the glove‐box c) right after the growth and d) after 48 h. No obvious change of contrast is visible. Large‐scale high‐angle annular dark field (HAADF) STEM images of large‐scale WTe₂ monolayer prepared e) in air and f) in the inert environment of the glove‐box, respectively. The sample surface is deteriorated and fragmented due to the oxidation and contamination in air, while the large‐scale intact atomic lattice structure is clearly visible when the entire process is isolated with air.

Figure 2

Atomic characterizations of the collective distorted lattice in continuous monolayer WTe₂. a) HAADF‐STEM image of WTe₂ monolayer. Corrugated areas are marked by orange and blue shades representing the ascending and descending sections of the ripple structure, respectively. The direction of the crystal plane (110) is marked by white dotted lines, which is also the boundary of the ripple structure. The arrows indicate the direction of the ripple. b,e) Atomic models of the marked orange and blue regions in (a), shown from top view. The deformed region and its bending direction can be clearly distinguished by the orientation of the distorted quadrilateral unit (marked by red diamonds), with b) the upward‐bending structure shaded in orange and e) the downward‐bending in blue, respectively. c,f) Corresponding STEM simulation images and d,g) atomic models from the side view. The ripple propagates dominantly perpendicular to the (110) crystal plane. h,j) HAADF‐STEM image of ripples propagating perpendicular to the (100) and (110) crystal planes. i,k) The spatial distribution mapping of distorted quadrilateral units in (h, j). The degree of distortion is reflected by the quadrilateral color changing from blue (less distortion) to red (severer distortion). The distribution of the distorted quadrilateral units is used as a key feature to distinguish the direction of ripple spread.

Figure 3

Anisotropy of continuous ripple in large‐scale WTe₂ monolayer. a,b) STEM image of ripples propagating perpendicular to the (110) crystal planes and the corresponding spatial distribution mapping of the distorted quadrilateral units. The bending directions of ripple are determined by the distorted quadrilateral units, labeled as “Bend up” and “Bend down” in (a). The two bending sections are separated by green dotted lines. The boundary of the quadrilateral units in (b) shows that all ripples propagate perpendicular to (110) direction. c) TEM selected area electron diffraction (SAED) pattern of 1T’ WTe₂ monolayer without tilting. DPs of (330) and (3−30) are marked by blue circles; (360) and (3−60) by orange circles. The tilt‐axis α and β are represented by grey dashed lines. d,e) FWHM of the intensity of four DPs in free‐standing monolayer WTe₂ as a function of tilt angle d) α and e) β ranging from 0° to 25°. Dashed lines are the linear fits yielding the average roughness. The asymmetric FWHM broadening of (330)/(360) versus (3–30)/(3–60) along α and β axis suggests an anisotropic broadening of the corresponding relrods in the reciprocal space. f) Schematic diagram of anisotropic ripples in free‐standing WTe₂ monolayer. The atomic structures are shown in insets, where different arrow size represents the fluctuation degree of ripple structure in that direction. g) The evolution of DPs with tilt angle α in monolayer WTe₂. Each angle corresponds to a black solid line. For each tilt angle, the blue solid line represents a cross‐section for relrods of (330) and (360) in the reciprocal space.

Figure 4

Anisotropic distribution of intrinsic Te vacancies modulated by ripple structure in WTe₂ monolayer. a–d) HAADF‐STEM images and the atomic structure of four distinct Te vacancy sites in WTe₂ monolayer. Red dotted circles represent the positions of the missing Te atoms. e) Atomic structure of the bending region from the side view. The bending is divided into the upper and lower regions, which are symmetric according to the change of the slope. The bending trend of W atom chain is depicted by green semicircles, including constrictive and tensile sides. f) Statistical counting of the four Te vacancy sites in the upper and lower deformation regions. Te vacancies are distributed preferentially on the constrictive parts. g) Calculated formation energy of the four distinct Te vacancy sites in the lower region as a function of the bending degree. Te vacancies located in the upper constrictive layer (site 1^up and site 2^up) maintain a lower formation energy than that in the tensile layer (site 1^down and site 2^down). h) Ripple‐induced modulated formation energy of the four Te vacancies in the supercell of ripple. Chalcogen atoms located at top and bottom layers are represented by blue and orange shades, respectively. The formation energy of Te vacancies fluctuates along with the ripple morphology. For detail in the statistics, see the Statistical Analysis section.

Figure 5

Anisotropic response of the ripple to the differently oriented grains and structure in partially oxidized case. a) STEM image of ripples near grain boundary (GB) in monolayer WTe₂. Yellow arrows indicate the possible strain direction from the GB. The crystal plane of (100), (110), and (1−10) on both sides are marked by white dashed lines. The direction of the ripple in the right grain is indicated by the gray arrow. b) Atomic‐scale STEM image of fine step‐like ripple structures within partially oxidized WTe₂ monolayer. Fine step‐like bending regions are shaded in blue. c) Schematic diagram of the transition process from a continuous smooth ripple structure to fine step‐like structure in WTe₂ monolayer induced by oxidization.