Flexible and cleaner alkaloid extraction of Sophora Flavescens and Heterosmilax Japonica with the aid of process analytical technology

Abstract

Traditional alkaloid extraction processes (AEP) usually utilize large amount of organic reagents or acidic solvents, which produces considerable toxic/acidic solid and liquid waste. In order to make AEP cleaner, we take the extraction process of Sophora Flavescens and Heterosmilax Japonica as an instance and introduce process analytical technology (PAT) to analyze and improve the process. Specifically, we use near infrared spectroscopy (NIRS) to collect the process information. By combining NIRS data with online sampling and off-line high-performance liquid chromatography (HPLC) measurements, we find the hot leaching process (HLP) of the traditional AEP is quite inefficient and the extraction is quite incomplete. Hence, we make corresponding modifications to the process, including dropping the HLP step and introducing a three-phase static leaching step. Furthermore, we build predictive models using the NIRS and HPLC data in order to offer timely predictions of critical quality attributes (CQA), which makes it possible to perform dynamic controlling actions. With the help of PAT, we finally achieve flexible AEP which is cleaner and has higher extraction rate ($\sim 65\text{\%}$ more) and lower solvent consumption ($\sim 23\text{\%}$ less) than the traditional AEP. The flexible AEP also has much shorter processing time ($\sim 60\text{\%}$ less), which can lead to better energy consumption.