Single EUV Laser Pulses Carve Silicon With Sub-Micrometer Precision

Researchers have demonstrated that a single pulse from an extreme ultraviolet (EUV) laser can carve structures into silicon with sub-micrometer precision. That is the headline result, and it deserves a careful reading before anyone redraws a lithography roadmap around it.

The demonstration matters because EUV photons are expensive. ASML's High-NA scanners cost several hundred million dollars per unit and work by projecting EUV light through masks onto resist-coated wafers. This result takes a different route: direct ablation of silicon by individual EUV laser pulses, with no mask and no resist in the loop as described.

What does the result actually show?

The claim is precise and narrow. One pulse, aimed at silicon, removes material and leaves a feature measured at less than a micrometer. The work appears in a laboratory setting, reported by AZoM, and concerns precision rather than throughput.

That distinction is everything. A lithography scanner for leading-edge chips processes well over one hundred wafers per hour and must hold overlay accuracy across full 300-millimeter substrates, pattern after pattern, layer after layer. A single-pulse experiment proves control, not economics.

Where does this sit against industrial EUV?

Commercial EUV lithography operates at a 13.5-nanometer wavelength, generated by firing a tin plasma tens of thousands of times per second inside a light source the size of a bus. Pulsed EUV lasers of the kind used in this research remain a scientific instrument, not a production tool.

The gap between the two resembles the gap between a jeweler's chisel and a stamping press. The chisel can make one exquisite mark; the press makes a million identical ones before lunch.

Direct patterning also sidesteps the photoresist chain, which is currently a genuine bottleneck in extending EUV to higher resolution. Any technique that removes resist from the equation invites attention, however early.

Who should care, and when?

For mask shops and metrology houses, pulsed EUV sources are already familiar territory. Actinic inspection of masks uses EUV wavelengths precisely because defects visible only at that wavelength matter in production. A laser that can carve and, by extension, structure silicon at sub-micrometer scale may find its first commercial home in similar niche work rather than in high-volume patterning.

For the major device makers — TSMC, Samsung, Intel — the result is a data point, not a disruption. Their exposure tools are contracted, their roadmaps run through High-NA EUV, and no首席 executive is re-planning capital expenditure over a single-pulse laboratory demonstration.

What are the honest limits?

The source describes a precision result. It does not report throughput figures, pulse repetition rates, cost per pulse, damage thresholds across a wafer, or repeatability across millions of shots. Those are the numbers that separate an interesting paper from an industrial process, and none of them appear yet.

Sub-micrometer precision, it should be noted, is also coarser than today's leading-edge features, which sit in the low tens of nanometers. The achievement lies in the mechanism — direct carving by single EUV pulses — rather than in the absolute feature size.

The verdict

One pulse, one feature, sub-micometer accuracy. As science, the result is tidy and genuinely novel in its control of EUV energy delivery. As manufacturing, it is a candle beside a lighthouse.

The work will earn its citations, and it may seed something larger: if pulse repetition rates climb and cost per pulse falls, direct EUV structuring could carve out a role in niche fabrication. Until those numbers arrive, the scanner remains king, and the laser remains a laboratory curiosity with respectable manners.