Researchers in South Korea say they have achieved position control of two-dimensional (2D) semiconductor materials — a result that targets one of the most stubborn problems in bringing 2D crystals from the laboratory to wafer-scale manufacturing: controlling where crystal growth begins.

The team's work, reported by BigGo Finance, centers on what researchers call the "starting point" of a semiconductor crystal. In conventional processing, thin-film crystals nucleate at random locations across a substrate. That randomness produces grain boundaries, defects and non-uniform device behavior — a tolerable nuisance in large 3D films, but a hard blocker for 2D materials, where a single atomic-scale defect can dominate a device's electrical characteristics.

The Korean claim is that this starting point can now be deliberately positioned, rather than left to chance.

Why does crystal nucleation matter for 2D semiconductors?

2D materials — atomically thin layered crystals such as transition metal dichalcogenides — are widely studied as candidates for next-generation logic and memory channels. At thicknesses of a single atomic layer, electrostatic control of the channel is nearly perfect, and short-channel effects that plague silicon below a few nanometers are expected to ease.

The manufacturing obstacle has never been the material's electrical performance in isolated, lab-fabricated devices. It is uniformity at scale. A wafer covered in 2D crystals that nucleated at uncontrolled sites is a wafer covered in misaligned grains. Grain boundaries scatter carriers, inflate device-to-device variation, and make yield projections for gigascale integration unconvincing.

Position control of nucleation attacks that problem at its root. If every crystal on the wafer can be seeded at a designed location and grown to a designed footprint, device arrays built from single-crystal domains become plausible rather than exceptional.

What could this change commercially?

Any roadmap that moves 2D channels into production depends on three capabilities: growing the material directly on target substrates (or transferring it without degradation), patterning it at alignment tolerances compatible with advanced back-end or front-end integration, and doing both across full wafers with statistical uniformity.

Nucleation position control is a prerequisite for the first of those. Without it, 2D growth remains a polycrystal-forming process; with it, growth starts to resemble the epitaxial discipline that made III-V and silicon manufacturing viable.

The announcement does not, based on the available report, include published capacity, wafer-size or yield figures — a common feature of early research-stage results in this field. Whether the Korean team's method works on 200 mm or 300 mm substrates, at temperatures compatible with CMOS back-end-of-line processing, or with the throughput of a production tool, will determine whether it registers outside the literature.

How does this fit the broader 2D race?

South Korea has a structural interest in this problem. The country's device makers and national research institutes have invested in 2D materials as one of the candidate paths beyond conventional scaling, alongside gate-all-around adoption and advanced packaging. A domestic group demonstrating nucleation control keeps that capability in-country, in a field where equipment and process know-how — not just publications — determine who ultimately ships 2D-based products.

Globally, research groups and consortia in Europe, the United States, Taiwan and China have pursued the same goal through different routes: patterned seed layers, engineered substrates, and confining growth in predefined templates. The Korean result adds to a body of evidence that deterministic crystal growth of 2D semiconductors is shifting from aspiration to engineering problem.

What remains unproven?

The report available for this article is limited to the headline claim. It does not specify:

• which 2D material system the team controlled;
• the substrate, temperature range, or growth method used;
• the spatial precision of the position control achieved;
• whether device-quality films were demonstrated on the patterned crystals.

Until those details appear — typically in a peer-reviewed paper or a conference presentation — the result should be read as a directional advance in 2D process control rather than a manufacturing-ready capability.

If the technique scales to full wafers with single-crystal uniformity, it removes one of the principal arguments against 2D semiconductors in volume production and strengthens the competitive position of any manufacturer that licenses or replicates it first.