Nanoimprint lithography physically stamps a pattern into a resist using a rigid template, and its patent claims typically cover template fabrication, release layers, and pattern transfer mechanics. Directed self-assembly instead relies on block copolymers organising themselves into a pattern guided by underlying chemical or topographic cues, so its claims tend to cover polymer formulations, guiding structures, and defect reduction chemistry. In this dataset both routes are searched together because commercial process flows increasingly combine them, but they sit on different prior art bases and should be searched separately during clearance work.

Filing peaked at 50 families in 2017 and settled to a flat midpoint of 26 by 2022, with the most recent year showing near-zero activity. Part of that recent drop is an artefact of publication lag, since patent applications typically surface 18 months or more after filing. The multi-year plateau before the most recent year, however, suggests genuine consolidation: the foundational techniques were claimed early by a small group of filers, and later entrants have had less open claim space to work with.

Equipment and process specialists such as Canon show substantial historical filing volume, often alongside dedicated nanotechnology subsidiaries as co-assignees. Materials suppliers including Merck Patent GmbH, and academic groups tied to institutions such as the University of Massachusetts and the University of Texas System, also appear repeatedly. Several of the most active filers show zero filings in the most recent tracked year, which reflects the sector-wide slowdown rather than any single company's specific circumstances.

Relative to the dense coverage in core photolithography (G03F) and general semiconductor (H01L) classes, sub-areas such as defect-density metrology for DSA patterns, roll-to-roll throughput scaling, and overlay accuracy correction across multilayer stacks show thinner claim density. A first claim in these branches is more likely to clear existing prior art than one filed against the heavily cited foundational release-layer or template-fabrication claims. Confirming this requires a targeted novelty search against the specific sub-claim language, not just the IPC subclass count.

The most-cited record in this dataset, US5772905A, was filed decades ago and by the standard 20-year term would have expired, meaning its citation weight reflects historical influence on later filings rather than current enforceability. Newer highly cited records, such as the block copolymer infiltration synthesis patent from 2014, are more likely to remain in force and worth checking against active development work. Citation count alone should not be read as a proxy for which patents currently block a given process; expiry status and claim scope both need separate verification.