In brief

QuestionWhat changes when an electrochemical experiment can choose its own next conditions?
ResultThe platform was demonstrated on three distinct reaction campaigns, including C–N coupling, anodic amino-acid modification, and decarboxylative alkylation.
Why it mattersThis moves AI from post-experiment analysis into the decision loop that selects the next electrochemical experiment.
Electrochemical self-optimization for the synthesis of densely functionalized molecules
Eduardo Rial-Rodríguez et al. · Chem · Published 19 March 2026
DOI: 10.1016/j.chempr.2026.102952

The electrochemical problem

Electrosynthesis offers unusual reaction pathways and can replace stoichiometric redox reagents with electrons, but optimization becomes difficult when voltage or current, residence time, concentration, reagent identity, and other continuous and categorical variables all interact.

For densely functionalized medicinal-chemistry targets, every experiment can also be relatively expensive. This creates a natural use case for a system that learns from each result and chooses the next experiment rather than scanning a fixed grid.

What the researchers did

The authors built a modular flow-electrochemical platform that integrates reaction preparation, electrolysis, online UHPLC-DAD analysis, and Bayesian optimization in a closed loop. The platform turns analytical results into the next set of experimental conditions without routine human intervention during the optimization cycle.

The system was tested in three chemically different campaigns: a nickel-catalyzed C–N cross-coupling, anodic modification of an amino acid, and decarboxylative alkylation of a natural product.

What they found

The platform could optimize single or multiple objectives while handling both continuous and categorical variables. Engineering features, including a gas-release module, were incorporated so that anodic reactions could run within the automated workflow.

Once favorable slug-flow conditions had been identified, the authors translated them to continuous-flow operation for product isolation. That step is important because it connects algorithmic optimization with a practically usable synthesis condition.

Why this matters

This is a clear example of AI becoming part of the experimental control loop. The algorithm does not merely interpret finished data; it allocates the next experimental measurement.

Electrochemistry is well suited to this idea because many reaction variables are digitally controllable. The difficult part is not only the optimization algorithm but the integration of pumps, electrochemical hardware, sampling, analysis, and chemical safety into a reliable autonomous workflow.

One important caution

Autonomy does not remove chemical constraints. The search space, objectives, analytical method, reactor design, and safety envelope are still specified by researchers, and Bayesian optimization can only search within that engineered framework.

Likewise, success on three reaction campaigns demonstrates versatility but not universal transfer to every electrosynthetic transformation.

Primary source

Eduardo Rial-Rodríguez et al. Electrochemical self-optimization for the synthesis of densely functionalized molecules. Chem (2026). DOI: 10.1016/j.chempr.2026.102952.

Research-note facts and quantitative values were checked against the publisher record linked above. The “Why this matters” and “One important caution” sections are EchemAI editorial interpretation.