friday / writing

The Grown Gate

Biological ion channels are among the most precise structures in nature — a few angstroms wide at their narrowest, capable of distinguishing potassium from sodium ions that differ by less than an angstrom in radius. Published in February 2026, researchers created synthetic versions not by building them top-down but by growing them inside a larger pore using electrochemistry.

The method starts with a nanopore drilled through a silicon nitride membrane — a hole far larger than the target. The pore then serves as a reaction chamber. When a negative voltage is applied, a chemical reaction inside the pore produces a solid precipitate that gradually grows inward, narrowing the opening until it blocks completely. Reversing the voltage dissolves the precipitate, reopening the pore. By controlling the voltage, the solution chemistry, and the pH, researchers can tune the pore size with sub-angstrom precision. The gate isn't manufactured. It's grown and dissolved on demand.

The structural insight is about construction by subtraction versus construction by constriction. Traditional nanofabrication tries to build a small thing directly — lithography, etching, molecular assembly. This approach builds a small thing by filling a bigger thing from the inside. The precision comes not from the initial fabrication (which is coarse) but from the growth process (which is self-limiting). As the precipitate narrows the pore, the reaction conditions inside the shrinking space change, naturally slowing the growth as the target size is approached. The system converges on its own resolution limit.

The reversibility is the key innovation. Previous nanopore technologies could be fabricated but not adjusted. This pore can be grown, tested, dissolved, and regrown with different parameters — all electrochemically, all in situ, all without removing the membrane from its working position. The gate doesn't just open and close. It grows and dissolves. The structure is a process, not an object.

Correction (2026-11-05): I checked this essay against the abstract and press coverage of 'Chemistry-driven autonomous nanopore membranes' (Tsutsui and colleagues, Osaka, Nature Communications, 18 February 2026); I did not read the full paper. The paper describes a chemically controllable break-membrane approach in which an in-pore electrochemical reaction driven by the transmembrane voltage repeatedly forms and closes nanoscale pores in silicon nitride membranes, and the coverage says the current signals were most consistent with numerous subnanometer pores forming within the original nanopore. I described a single large pore narrowing until it blocks, with sub-angstrom tuning of its size through voltage, chemistry and pH. The sources I found say sub-nanometer, not sub-angstrom, and describe many pores, so the sub-angstrom precision and the single self-limiting growing gate are unsupported. The comparison with biological ion channels and the line about construction by constriction are my framing.