Rubens Maarten, Vrijsen Jeroen H, Laun Joachim, Junkers Tanja
Hasselt University, Martelarenlaan 42, 3500, Hasselt, Belgium.
Polymer Reaction Design Group, School of Chemistry, Monash University, 19 Rainforest Walk, Building 23, Clayton, Vic, 3800, Australia.
Angew Chem Int Ed Engl. 2019 Mar 4;58(10):3183-3187. doi: 10.1002/anie.201810384. Epub 2018 Nov 21.
A novel continuous flow system for automated high-throughput screening, autonomous optimization, and enhanced process control of polymerizations was developed. The computer-controlled platform comprises a flow reactor coupled to size exclusion chromatography (SEC). Molecular weight distributions are measured online and used by a machine-learning algorithm to self-optimize reactions towards a programmed molecular weight by dynamically varying reaction parameters (i.e. residence time, monomer concentration, and control agent/initiator concentration). The autonomous platform allows targeting of molecular weights in a reproducible manner with unprecedented accuracy (<2.5 % deviation from pre-selected goal) for both thermal and light-induced reactions. For the first time, polymers with predefined molecular weights can be custom made under optimal reaction conditions in an automated, high-throughput flow synthesis approach with outstanding reproducibility.
开发了一种用于聚合反应的自动化高通量筛选、自主优化和强化过程控制的新型连续流动系统。该计算机控制平台包括一个与尺寸排阻色谱(SEC)相连的流动反应器。在线测量分子量分布,并由机器学习算法利用该分布通过动态改变反应参数(即停留时间、单体浓度和控制剂/引发剂浓度)对反应进行自我优化,以达到设定的分子量。该自主平台能够以前所未有的精度(与预先选定目标的偏差<2.5%)以可重复的方式针对热引发和光引发反应确定分子量。首次能够在最佳反应条件下,通过具有出色重现性的自动化高通量流动合成方法定制具有预定义分子量的聚合物。