Fig. 7: LLM-based agents facilitated reaction scale-up and product purification.
From: An automatic end-to-end chemical synthesis development platform powered by large language models
a Workflow for reaction scale-up and product purification copiloted by Experiment Designer and Separation Instructor agents. b The interaction between human chemists with Experiment Designer for reaction scale-up and Separation Instructor for finding the optimal eluent composition (see detailed interaction dialogs in Supplementary Table 41-45). The asterisk denotes that the numerical results were directly provided by the large language model (LLM). c Radar chart for comparing three high-yield reaction conditions obtained during self-driven reaction optimization (experiments 26, 32, and 35). d Proton nuclear magnetic resonance (1H NMR) spectrum of the purified target product (12p) in DMSO-d6 (See complete spectral information in Supplementary Information Section 6.7).
