Monitoring batch-to-batch reproducibility using direct analysis in real time mass spectrometry and multivariate analysis: a case study on precipitation.

Batch-to-batch variability of traditional Chinese medicine (TCM) has already heavily limited its industrial development. By utilizing direct analysis in real time mass spectrometry (DART-MS) and multivariate statistical analysis, the present study introduced a novel and rapid methodology for TCM manufacturing process monitoring. Danshen alkaline precipitation was employed as an example operating unit to demonstrate the effectiveness of this new technique. A total of 15 batches run under normal operating conditions were used to develop a reference principal component analysis model and then enabled the establishment of multivariate control charts. Hotteling T(2) and DModX charts were applied to examine batch-to-batch reproducibility of 12 test batches. Artificial variations including starting material change and process fault were identified, which was in a good agreement with conventional analysis result. Further understanding of the deviating process behavior was achieved by investigating the contribution plot of abnormal batches. The results showed that DART-MS, in conjunction with multivariate analysis, provided valuable information on model process and gave us a new perspective for TCM manufacturing process monitoring.