The microscopy field spent thirty years relying on the wrong test for a class of particles

The "non-blinking" label on upconverting nanoparticles was a property the field treated as a law of nature. It wasn't. It was a property of one compositional regime, generalized to a class. MIT News's coverage of Sam Peng's Nature Nanotechnology paper documents what changes when the assumption is wrong: a single particle, about ten nanometers across, yields tens of thousands of localizations under continuous near-infrared light, with no imaging buffer, no oxygen scavenger, no exotic optics. The technique the Peng lab calls U-STORM inherits STORM's localization logic and drops the chemistry that kept STORM out of routine labs.

The deeper pattern is about how instrumentation gets locked in. A field that runs on a working imaging chemistry tends to write off "the particles that don't behave like our dyes." The Peng lab's reframing turns the question around: which other classes of materials were excluded not because they failed, but because the field asked them to do the wrong test? A class of particles the imaging community wrote off for three decades becomes a tool, and — the paper argues, reframing a decades-old assumption — the prior ceiling on localization precision moves by three orders of magnitude. The bottleneck, the paper suggests, was not the physics. It was a generalization that had outlived its evidence.