/ research
Robots that make the samples, software that runs the lab, and the papers that came out of it.
/ projects
in progress · ALS beamline 7.3.3
20–30experiments to hit the target, versus a grid scan of thousands
A team-built closed loop in which a Gaussian-process optimizer drives a synchrotron beamline directly, finding the processing route to a named target morphology in 20–30 experiments instead of thousands.
first author · ACS Nano 2026
100sof thin-film samples fabricated and characterized per day
Hundreds of thin-film samples fabricated in hours, measured by synchrotron scattering, and turned into the first high-throughput map of the order–disorder transition landscape in triblock copolymer–homopolymer blends.
co-first author · Macromolecules 2026
6block copolymer chemistries in one generalizable model
A machine learning framework that predicts block copolymer thin-film morphology from processing conditions, trained on high-throughput scattering data across six chemistries.
now building · Berkeley Lab
NowALS Doctoral Fellowship, Berkeley Lab, 2026–27
Current work at Berkeley Lab — a multimodal self-driving lab that uses active learning to map polymer blend morphology and phase behavior, choosing each next sample where the map is least certain.
/ software
The software that runs the lab. It drives the robots, starts each measurement, analyzes the data as it lands, and picks the next sample, with no one carrying things between instruments. New instruments plug in through adapters, so it isn't tied to one setup.
An open database of block copolymer thin-film structure, built from the robot's data across 6 chemistries, failed samples included. It's organized so machine-learning models can train on it.
/ in print
20 papers and preprints, 2 as first or co-first author.
72 citations · h-index 5 · from Google Scholar, September 2026.
No publications match that combination.
