Fluid Dynamic Sampling Site Characterization Improves Process Correlation During Continuous Online Sampling

Purpose Implementation of Process Analytical Technologies (PAT) is an important consideration to increase process knowledge during pharmaceutical manufacturing. Currently used batch sampling methods, however, are limited in terms of temporal resolution when compared with continuous process sampling. Here, we show the importance of fluid dynamic characterization when repurposing available batch sampling sites comprising needle insertion mechanisms to generate continuous sample streams. Methods Rheological flow profiles in a needle-based sampling site were characterized by computational fluid dynamics. Observed characteristics were confirmed with a NovaSeptum sample system in reenacted sampling situations, and correlation of sample and system flow was achieved by simultaneous measurement of an acetone gradient. Results As a result of the sampling mechanism causing distinct flow patterns around the needle body, velocity fields are shown to generate orientation-dependent sample flow. Due to the inherent relationship between flow detachment and turbulent flows, extracted sample flow depends on system flow profiles, affecting sample correlation. A combination of active sample extraction, check valve addition, and geometrical changes in sampling site design is shown to reduce sample efflux variability, allowing for system parameter-independent sample generation. The resulting sample flow has been correlated to a simulated process flow, enabling a continuous representation of at-scale systems. Conclusions Repurposing existing sampling sites requires fluid dynamic characterization in order to generate representative sample streams, which allows the exchangeable inclusion of process analytics otherwise not compatible with at-scale systems. Resulting time-dependent, discrete samples can then be utilized for on- and off-line analysis to provide additional process knowledge according to the PAT initiative.