Domain morphology plays a pivotal role not only for the synthesis of high-quality 2D transition metal dichalcogenides (TMDs) but also for the further unveiling of related physical and chemical properties, yet little has been divulged to date, especially for metallic TMDs. In addition, solid precursor as a transition metal source has been conventionally introduced for the synthesis of TMDs, which leads to an inhomogeneous distribution of local domains with the substrate position, making it difficult to obtain a reliable film. Here, we tailor the domain morphologies of metallic NbSe 2 and NbSe 2 /WSe 2 heterostructures using liquid-precursor chemical vapor deposition (CVD). We find that triangular, hexagonal, tripod-like, and herringbone-like NbSe 2 flakes are constructed through control of growth temperature and promoter and precursor concentration. Liquid-precursor CVD ensures domain morphologies that are highly reproducible over repeated growth and uniform along the gas-flow direction. A domain coverage of ∼80% is achieved at a high precursor concentration, starting with tripod-like NbSe 2 domain and evolving to the herringbone fractal. Furthermore, mixing liquid W and Nb precursors results in sea-urchin-like heterostructure domains with long-branch-shaped NbSe 2 at low temperature, whereas protruded hexagonal heterostructure domains grow at high temperature. Our liquid precursor approach provides a shortcut for tailoring the domain morphologies of metallic TMDs as well as metal/semiconductor heterostructures. KEYWORDS: domain morphology, monolayer niobium diselenide, lateral NbSe 2 −W x Nb 1−x Se 2 heterostructure, liquid precursor, chemical vapor deposition

Atomic dopants and defects play a crucial role in creating new functionalities in 2D transition metal dichalcogenides (2D TMDs). Therefore, atomic‐scale identification and their quantification warrant precise engineering that widens their application to many fields, ranging from development of optoelectronic devices to magnetic semiconductors. Scanning transmission electron microscopy with a sub‐Å probe has provided a facile way to observe local dopants and defects in 2D TMDs. However, manual data analytics of experimental images is a time‐consuming task, and often requires subjective decisions to interpret observed signals. Therefore, an approach is required to automate the detection and classification of dopants and defects. In this study, based on a deep learning algorithm, fully convolutional neural network that shows a superior ability of image segmentation, an efficient and automated method for reliable quantification of dopants and defects in TMDs is proposed with single‐atom precision. The approach demonstrates that atomic dopants and defects are precisely mapped with a detection limit of ≈1 × 1012 cm−2, and with a measurement accuracy of ≈98% for most atomic sites. Furthermore, this methodology is applicable to large volume of image data to extract atomic site‐specific information, thus providing insights into the formation mechanisms of various defects under stimuli.

The 1D wire TaS3 exhibits metallic behavior at room temperature but changes into a semiconductor below the Peierls transition temperature (Tp), near 210 K. Using the 3ω method, we measured the thermal conductivity κ of TaS3 as a function of temperature. Electrons dominate the heat conduction of a metal. The Wiedemann–Franz law states that the thermal conductivity κ of a metal is proportional to the electrical conductivity σ with a proportional coefficient of L0, known as the Lorenz number—that is, κ=σLoT. Our characterization of the thermal conductivity of metallic TaS3 reveals that, at a given temperature T, the thermal conductivity κ is much higher than the value estimated in the Wiedemann–Franz (W-F) law. The thermal conductivity of metallic TaS3 was approximately 12 times larger than predicted by W-F law, implying L=12L0. This result implies the possibility of an existing heat conduction path that the Sommerfeld theory cannot account for.