Self-optimizing process parameter screening in flow

一旦一步特定的合成策略已经合格，一个主要的挑战是找到一组优化的浓度和工艺条件。为此，重要的是以有效和可重复的方式筛选不同的条件，并收集分析数据。该项目将利用流量系统的超高吞吐能力。它将包括建立一个具有定制在线和在线分析设备的柔性流动反应堆。它将受益于帝国理工学院Roar工厂提供的基础设施。捕获的数据将受到机器学习算法的影响，这些算法将在一个单独的项目中开发。我们的目标是开发一种智能系统，该系统根据其通过关联一组反应参数来预测和控制反应结果的学习，自动决定下一组条件。

Once a step specific synthetic strategy has been qualified, a major challenge is to find an optimized set of concentrations and processing conditions. To this end, it is important to screen different conditions in an efficient and reproducible way, and to collect analytical data. This project will take advantage of the ultra-high throughput capability of flow systems. It will involve setting up a flexible flow reactor with bespoke in-line and at-line analytical equipment. It will benefit from the infrastructure that is available at Imperial's ROAR facility. The data captured will be subjected to machine learning algorithms that are to be developed in a separate project. The objective is to develop an intelligent system that that automatically decides on the next set of conditions based on its learnings from correlating a set of reaction parameters to predict and control the reaction outcome.