Thickness-directed oxide-templated conversion of sputtered tungsten films into WS2 nanotubes.

Controlling the topology of curved transition-metal dichalcogenides remains challenging because shell formation, precursor conversion and hollowing occur simultaneously during growth. Here, we show that the thickness of sputtered tungsten (W) films provides a direct handle for selecting WS2 nanotube topology during charcoal-assisted oxide-template sulfurization. Varying the initial W thickness from 1 to 100 nm drives a transition from laterally extended WS2-like domains to hollow nanotubes and, finally, to retained-core nanotubes exhibiting shell-front mismatch and spiral-like shell wrapping. Sulfur-free controls, Raman spectroscopy, grazing-incidence X-ray diffraction, transmission electron microscopy and elemental mapping reveal that this transition is governed by the continuity, accessibility, and recession of WO x -derived precursor structures. As an initial test of functionality, multiphoton microscopy further shows that the oxide-core nanotube assemblies are optically active nonlinear emitters, generating second-harmonic, sum-frequency and four-wave-mixing signals that are readily detected under conditions where a WS2 monolayer reference is near the detection limit. These results establish film thickness as a synthetic handle for curved transition-metal dichalcogenide architectures and identify these self-formed nanostructures as a promising platform for nonlinear nanophotonics.