Technion's electrolysis process swaps oxygen for an industrial epoxide at 98% efficiency

For a plant manager, this paper raises two questions: what the 98% measures, and who buys the epoxide.

The first has a plain answer. According to the team, the figure means only about 2% of the electrical charge was lost to competing reactions ^[3]. It describes where the current went, for both the hydrogen and the epoxide ^[2]. On its own terms that is a strong result. Energy use and cost per kilogram are different quantities, and the Interesting Engineering report does not include either ^[3].

Green hydrogen has been held back by its cost compared with hydrogen made from fossil fuels ^[6]. Interesting Engineering's pitch is that two useful outputs give producers an extra economic incentive ^[14]. The work so far is the researchers' own experiments ^[13].

The result also depends on the whole system. According to Ruan, the performance comes from optimizing the complete electrochemical system, not a single electrode or reaction ^[11]. If that holds at scale, an operator would be maintaining a redox mediator and a solvent environment along with the electrodes ^[12].

The new system builds on Rothschild's earlier work separating the reactions that normally produce hydrogen and oxygen ^[9]. The team has filed a US patent application for it ^[8].

The second question depends on the site. Epoxides go into polymers, coatings, adhesives, pharmaceuticals and other products ^[5], so the demand exists in general. A coproduct lowers the effective cost of a plant's hydrogen only if a buyer near that plant takes the volume the electrolyzer makes.

I'd sort any candidate site on two axes. One is offtake: an epoxide buyer within reach, at volumes matched to the hydrogen you need. The other is run time: published data on sustained industrial operation, enough to price mediator and electrode replacement over years. A site with offtake and run-time data is ready for a pilot conversation. Offtake without the data puts a site on a watch list, with a named contact at the buyer. Without offtake, the epoxide is just another stream to handle, however efficiently it is made. Right now every site sits on the no-data side, because the study does not establish that the technology is ready for commercial-scale deployment, and the report says further development is needed to show how it performs under sustained industrial conditions ^[13].