Centimeter-scale fully suspended metal and metal oxide thin films by one-step transfer-free liquid metal capillary forming

Scientific Impact Assessment

Core Contribution

This paper introduces a genuinely novel fabrication paradigm: forming centimeter-scale, fully suspended metal, solid-metal, and metal-oxide thin films through "liquid metal capillary forming," directly analogous to soap film formation. The central insight is that the spontaneously formed, few-nanometer native surface oxide of liquid metals acts as a "surfactant-like" viscoelastic bilayer that laminates the high-surface-tension liquid metal into stable micrometer-thick films. Even more striking is the demonstration that this oxide bilayer can survive dewetting-induced liquid-metal drainage and "zip" together into a few-nanometer-thick suspended 2D oxide film with a lateral size-to-thickness ratio on the order of 10⁷. The method is one-step, transfer-free, and substrate-free — directly addressing the long-standing dependence of suspended-film fabrication on multi-step grow-and-transfer strategies that limit cleanness and lateral dimensions.

Methodological Rigor

The characterization suite is comprehensive and convincing: cross-sectional SEM, HAADF-TEM, EDS mapping, high-resolution TEM with electron diffraction, AFM, low-wavenumber Raman, optical transmittance, and micro-CT tomography. The authors ground the phenomenon in established thin-film physics — Reynolds lubrication approximation for thickness evolution and Frankel's law for the Frankel-zone thickness — and show experimental thickness profiles fitting theory across multiple metals (Ga, GaIn, InSn, In, Sn, Bi, BiInSn, Al). Phase diagrams (T_B–D) map formation regimes and are reproduced across two distinct alloy systems, strengthening generality. Control experiments in a nitrogen glovebox (no film formation without oxidation) cleanly establish the causal role of the surface oxide. The acoustic-detection demonstration is quantified with explicit sensitivity figures (1.94×10⁻⁴ m/Pa, 2 mPa pressure limit). Minor gaps: some claims (e.g., "first identification of boson peaks in liquid metal-derived 2D oxides") are peripheral, and the long-term stability/scalability of oxide films beyond lab-scale frames is not fully explored.

Potential Impact

The work could establish a new material platform for 2D-material and thin-film research. Suspended metal oxides from *non-layered* crystals (as opposed to graphene/hBN/TMDs from layered crystals) are rare, and this method makes wafer-scale versions accessible in seconds. Applications shown — rapid prototyping of thin-walled minimal-surface 3D metallic structures (catenoid, helicoid, buckyball) and ultra-sensitive acoustic diaphragms — hint at breadth across sensing, MEMS, optics, and fundamental 2D-material physics. The comparison chart (Fig. 3C) claiming ≥1 order of magnitude improvement over prior suspended 2D films by area or area-to-thickness ratio, if robust, is a meaningful advance.

Timeliness & Relevance

Liquid metals and their native oxides are an actively expanding research frontier (printed electronics, high-κ gallium oxides for 2D transistors, atomically thin oxide synthesis). The paper connects this momentum to the suspended-film/2D-materials community, addressing a recognized bottleneck (substrate-involving transfer steps limiting cleanness and lateral size). It is highly timely.

Strengths & Limitations

Strengths: Conceptually elegant cross-domain analogy (soap films ↔ liquid metal films) with strong mechanistic backing; remarkable size-to-thickness ratios; demonstrated generality across ~8 metals/alloys; thorough multimodal characterization; concrete device demonstration. The counterintuitive framing (liquid metals' enormous surface tension should *prevent* film formation, yet the oxide enables it) is well-supported.

Limitations: The oxide films are amorphous, limiting some electronic applications versus crystalline suspended oxides. Practical device integration and long-term ambient durability are only partially demonstrated. The method is constrained to metals with accessible melting points and native oxide formation. Reproducibility depends on facility-grade characterization; the fabrication itself is simple but frame geometry/temperature control parameters are only partly tabulated, and no code/detailed protocol repository is mentioned. The acoustic-sensing claim of "several orders of magnitude" higher sensitivity than other 2D films would benefit from more direct head-to-head benchmarking.

Additional Observations

This reads as a top-tier journal submission (Nature/Science format). The refutation value is notable but subtle: it overturns the intuitive assumption — explicitly stated as expected in the literature (refs 22–24) — that high-surface-tension, low-viscosity liquid metals cannot form stable thin films. The foundational potential is real: the method could become a reusable fabrication primitive for suspended-film research. Interdisciplinarity spans materials science, capillary/fluid physics, 2D-materials physics, and MEMS/acoustic sensing.