Synthesis and guided assembly of niobium trisulfide nanowires and nanowire chains by chemical vapor deposition

Technical Summary: Synthesis and Guided Assembly of Niobium Trisulfide Nanowires and Nanowire Chains

Problem Statement
Transition metal trichalcogenides (TMTs), specifically those with the formula MX3, exhibit unique anisotropic physical properties due to their one-dimensional (1D) van der Waals (vdW) bonding structure. While bulk crystals of these materials display interesting phenomena such as charge density waves (CDW) and superconductivity, obtaining controlled nanoscale structures is challenging. Existing top-down approaches, such as mechanical or chemical exfoliation of bulk crystals, lack precise control over nanostructure dimensions and interfaces with support materials. Conversely, bottom-up synthesis methods like Chemical Vapor Deposition (CVD) are less established for trichalcogenides compared to dichalcogenides, and previous attempts often require complex catalyst systems (e.g., VLS with salt assistance) that may limit scalability or control. There is a need for scalable, bottom-up synthesis methods that can control the morphology, dimensions, and interfaces of 1D TMTs to enable their integration into novel devices.

Methodology
The authors employed a salt-assisted Chemical Vapor Deposition (CVD) technique to synthesize niobium trisulfide (NbS3) nanowires.

• Precursors and Conditions: The process utilized niobium (Nb) powder and sulfur (S) powder, with sodium chloride (NaCl) acting as a growth promoter. The Nb and NaCl were ground together and placed in an alumina boat, while S was placed upstream. Growth occurred at temperatures between 650–750°C with an Argon carrier gas flow of 50–150 sccm. Growth times ranged from 1 to 45 minutes.
• Substrates: Synthesis was performed on a variety of substrates to investigate growth mechanisms and guided assembly:
  • Amorphous SiO2/Si (bulk).
  • Crystalline c-sapphire (bulk).
  • Exfoliated 2D vdW materials: Graphene, hexagonal boron nitride (h-BN), and CrSBr (a 2D material with in-plane anisotropy).
• Characterization: Morphology and structure were analyzed using Scanning Electron Microscopy (SEM), Scanning Transmission Electron Microscopy (STEM), and STEM-Energy Dispersive X-ray Spectroscopy (EDS). Cross-sectional samples were prepared via Focused Ion Beam (FIB) milling.

Key Contributions and Results
The study reports the successful synthesis of NbS3 nanowires in two distinct growth modes, dependent on precursor flux and substrate conditions:

1. Mode 1: Separated Short Nanowires

   • Location: Observed primarily at the center of the substrate where precursor flux is highest.
   • Morphology: Relatively short, separated nanowires with rectangular cross-sections.
   • Dimensions: Average width of ~0.35 µm and length of ~3.6 µm (after 45 min growth).
   • Structure: Identified as the metallic NbS3 phase-II. They exhibit well-faceted surfaces and grow along the b-direction.

2. Mode 2: "Chained Nanowires"

   • Location: Observed at the edges of the growth area where precursor flux is lower.
   • Morphology: Unusually long (up to 100 µm), segmented nanowires with a sawtooth surface structure. These are formed by the end-to-end connection of multiple segments.
   • Dimensions: Segments are similar in width to Mode 1 (~0.35 µm) but the total chain length is significantly greater (average ~70 µm).
   • Growth Dynamics: Segments grow at a slight angle (approx. 8°) relative to the substrate. New segments nucleate at the contact line between the previous segment and the substrate, leading to a "chained" appearance with small angular misalignments.

3. Growth Mechanism
   The authors propose a nucleation and diffusion-limited mechanism:

   • Mode 1: High nucleation density on the amorphous substrate leads to competition for flux, resulting in short, isolated wires.
   • Mode 2: Lower flux reduces nucleation density. Initial nuclei form with a slight misorientation to the substrate. As the first segment grows, it lifts off the surface. Subsequent nucleation occurs at the contact line (where the wire meets the substrate) rather than on the wire surface, templating new segments with slight angular deviations. This results in long, curved chains composed of straight, misaligned segments.

4. Guided Assembly and Epitaxy

   • 2D Substrates: On flat graphene and h-BN, only Mode 1 growth was observed, confirming that misorientation is required for Mode 2. However, at the edges of these flakes, nanowires grew in parallel arrays, often bridging gaps between flakes.
   • Epitaxial Growth: On CrSBr (which has in-plane anisotropy) and c-sapphire, the nanowires exhibited epitaxial growth with specific orientation relationships (e.g., [010]NbS3 parallel to specific directions of CrSBr and c-sapphire). Notably, Mode 2 chains were not observed on these crystalline substrates, reinforcing the hypothesis that a misaligned first segment on a rough or amorphous surface is necessary to initiate the chained morphology.

Significance and Claims
The paper claims to demonstrate a scalable, bottom-up synthesis method for 1D TMT nanomaterials that offers control over morphology through substrate choice and growth conditions. The discovery of the "chained nanowire" morphology represents a new structural form for 1D vdW materials, achieved without the need for metal catalysts in the final product (though salt is used as a promoter).

The authors suggest that these results open pathways for:

• Scalable Synthesis: Achieving high growth yields with axial rates up to 40 µm/min.
• Directed Assembly: Utilizing substrate edges and crystalline anisotropy to guide the placement and orientation of nanowires, potentially creating dimensionally hybrid vdW heterostructures.
• Device Integration: The ability to grow 1D nanowires directly on 2D materials creates clean, abrupt interfaces with presumed low contact resistance, providing a platform to study proximity-induced effects.
• Potential Applications: The authors note the potential for these long, self-assembled chains to serve as electrical interconnects or thermal dissipation components, and suggest that patterning substrates could allow for programmable growth modes.

The study concludes that the interplay between nucleation density, substrate-induced misorientation, and anisotropic atom incorporation dictates the final morphology, offering a framework for engineering 1D TMT nanostructures.