EUV Lithography
# Carbon Nanotube EUV Pellicle Membrane
A EUV pellicle is a thin, free-standing membrane stretched a few millimeters above the mask to catch particles before they can land on the pattern itself, and the membrane material has to satisfy two requirements that normally work against each other: it must be nearly transparent to 13.5nm light, and it must survive the intense thermal load of repeated high-power EUV exposure without tearing. Any particle landing directly on an unprotected mask gets imaged onto the wafer as a defect, repeating on every single die across the entire wafer since the mask pattern is reused thousands of times. A pellicle intercepts that particle a few millimeters above the mask surface instead, far enough out of focus that the particle's own shadow doesn't print. The problem is that any material standing directly in the EUV beam path also absorbs some of that light and heats up under the beam's power — a membrane that's too absorptive either blocks too much usable light or heats enough to physically fail.
Carbon nanotube membranes are being pursued specifically because an aligned CNT mesh combines very high in-plane thermal conductivity with very low areal mass, letting the membrane spread absorbed heat laterally fast enough to avoid a damaging hot spot, while staying thin enough that EUV transmission losses stay acceptable. A membrane material that's a poor thermal conductor accumulates heat right where the beam hits, and local temperature spikes are what cause pellicle membranes to fail — tearing, sagging out of the optical path, or simply degrading transmission over repeated exposures. Carbon nanotubes conduct heat exceptionally well along their length, so a mesh of aligned or interwoven nanotubes spreads absorbed thermal energy out across the whole membrane area almost as fast as it's deposited, keeping any single point from overheating even as cumulative exposure time adds up across a scanner's operating life.
The tradeoff that keeps CNT pellicle membranes in active development rather than full production is manufacturing a mesh uniform and defect-free enough across the full pellicle area — any gap, bundle, or non-uniform region in the nanotube network creates a local weak point in exactly the heat-spreading and transmission properties the material is chosen for. A single point of poor nanotube coverage behaves like the conventional thin film's hot spot all over again, defeating the purpose of switching materials in the first place. That manufacturing-uniformity challenge, not the underlying thermal physics, is the main reason CNT pellicles remain a research-and-qualification topic rather than a universal replacement for existing pellicle materials in production scanners today.
The tin-droplet source and collector mirror determine how much EUV light and heat load exist in the first place, and the pellicle membrane is one of the surfaces that heat load lands on directly — carbon nanotube membranes are a bet that a fundamentally different thermal-conduction mechanism can survive that load better than conventional thin films, at the cost of a harder manufacturing problem.