Original title: Quantitative control and recording of materials-synthesis processes using an automated experimentation platform
Yusuke Hashimoto, Takaya Muramoto, Hikari Terada, Harim Song +2 · September 2026
Developing new materials with data requires huge amounts of high-quality experimental records. Fully autonomous systems have high technical hurdles and limited adoption, so scientists built a simpler automated lab setup. They linked off-the-shelf tools, including robot arms, electric pipettes, an electronic balance, and web cameras, with 3D-printed parts. To run the hardware, an artificial intelligence agent powered by a large language model generated the control code. The team tested the setup by mixing two solutions to make ZIF-8A specific metal-organic framework with a structure similar to zeolites., a type of porous material called a metal-organic framework. The platform produced a white liquid suspension confirmed as ZIF-8 by X-ray measurements. Crucially, the size distribution of the particles depended strongly on how fast the pipette dispensed the liquid. Manual lab work rarely tracks or controls dispensing speed well. The automated system repeated this result across multiple runs, showing that precisely measuring hidden manual steps ensures consistent data.
Imagine baking cookies where you add the vanilla extract by hand. You might pour it quickly or slowly, but you never write down the exact speed. Some batches turn out perfect, others flat, and you cannot figure out why. Here, the researchers built a robot baker that records exactly how fast it adds each ingredient. They found that the speed of adding one solution changed the size of the crystals in a material, a detail that would have been missed in manual baking.
Materials scientists often rely on manual synthesis, where subtle parameters like how fast you add a solution go unrecorded. This platform automates and records those details, showing they can dramatically affect the product. That means more reproducible experiments and higher-quality data for discovering new materials.
The abstract doesn't explicitly state limitations, but the platform was only demonstrated on a single reaction (ZIF-8 synthesis) and does not achieve full autonomy. Its generalizability to other systems is not shown.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Quantitative control and recording of materials-synthesis processes using an automated experimentation platform — Yusuke Hashimoto, Takaya Muramoto, Hikari Terada et al.
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.