AOCS Annual Meeting & Expo 2026

Automated Discovery of Acrylamide Formation from the Maillard reaction with the Nanoreactor

Speaker:
Croix Laconsay, Senior Scientist at Schrödinger

Abstract:
The Maillard reaction is a fundamental chemical reaction in food chemistry and occurs naturally between amino acids and reducing sugars during thermal processing. Understanding the molecular mechanisms underlying the formation of toxic byproducts from the Maillard reaction is essential for reducing exposure. Molecular modeling provides a means of understanding the energetically-accessible chemical pathways by which these toxins are created. One important example is the formation of acrylamide, a toxic compound, in foods heated above 120 °C. Experimental studies have shown that acrylamide can form from asparagine under Maillard reaction conditions.1 Examples of computational studies on this particular mechanism are rare,2 and, to the best of our knowledge, theoretical studies of acrylamide formation from asparagine have not been reported. Computational studies of chemical degradation usually involve the manual investigation of complex multistep reaction networks. These networks are explored in laborious manual efforts that involve a mix of chemical intuition and density functional theory (DFT) calculations. Various automated methods have been proposed to discover the chemically relevant elementary reaction steps.3
Inspired by the work of Jensen,4 we introduce a fully automated approach for reaction network exploration called Elementary Reaction Network. This process uses Nanoreactor, a method which utilizes a metadynamics-based simulation within a confined reaction sphere to efficiently sample elementary reaction steps. Subsequent AutoTS computations, Schrödinger’s automated DFT-based transition-state search workflow, further refine the results by locating and optimizing the transition states that connect reactants and products discovered by Nanoreactor. Our Nanoreactor-AutoTS workflow accelerates the chemical network exploration of molecular degradation mechanisms and allows unbiased exploration of the potential energy surface. In this talk, I will demonstrate the utility of this workflow in exploring degradation mechanisms relevant to the Maillard reaction.