Exploiting Machine Learning and Automated Synthesis in Continuous Flow for Process Optimization of the Organocatalyzed Ring-Opening Polymerization of l‑Lactide

The ring-opening polymerization (ROP) of lactide is a well-established route for the synthesis of polylactide (PLA), a degradable and biobased polymer with applications in biomedical materials, packaging, and additive manufacturing. However, optimizing reaction conditions for efficient PLA synthesis...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:ACS polymers Au Ročník 5; číslo 5; s. 603 - 612
Hlavní autori: Clothier, Glenn Keith Kim, Taton, Daniel, Harrisson, Simon
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: United States American Chemical Society 08.10.2025
Predmet:
Catalysis
Chemical Sciences
Computer Science
Machine Learning
Mathematics
Optimization and Control
Polymers
organocatalysis
lactide
Flow chemistry
1,8-diazabicyclo[5.4.0]undec-7ene
machine learning
ring-opening polymerization
polymerization kinetics
1-8-diazabicyclo[5.4.0]undec-7ene
ISSN:2694-2453, 2694-2453
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:The ring-opening polymerization (ROP) of lactide is a well-established route for the synthesis of polylactide (PLA), a degradable and biobased polymer with applications in biomedical materials, packaging, and additive manufacturing. However, optimizing reaction conditions for efficient PLA synthesis remains challenging, with its production currently dwarfed by traditional petrochemical derived commodity polymers. In this work, we employ a continuous flow reactor to systematically explore a reaction space defined by catalyst concentration, residence time, and initiator concentration for the organocatalyzed ROP of l-lactide performed at room temperature, using 1,8-diaza­bicyclo​[5.4.0]​undec-7-ene (DBU) as catalyst, benzyl alcohol as initiator and dichloromethane as solvent. Through high-throughput experimentation coupled with online characterization, a robust data set was generated and processed with a Kernel-Based Regularized Least Squares (KRLS) model to capture system kinetics and the dependencies between initial conditions and polymer characteristics. Multiobjective Pareto optimization was subsequently used to identify conditions that maximize polymer production rate, and these conditions were experimentally validated to confirm the model accurately predicts optimal reaction conditions. Pareto-optimized parameters yield a high conversion and well-controlled PLA with low dispersities. This study highlights the advantages of continuous flow polymerization for precise control over reaction kinetics and demonstrates the potential of machine-learning-assisted optimization for efficient and scalable PLA-based materials’ exploration.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:2694-2453
2694-2453
DOI:10.1021/acspolymersau.5c00071