Development of Universal, Automated Sample Acquisition, Preparation, and Delivery Devices and Methods for Pharmaceutical Applications

Continued emphasis in the pharmaceutical industry toward development of hybrid and continuous processes has led to a resurgence in process analytical technologies. The need to consistently and rapidly monitor the quality of material being made from these processes is a significant challenge that requires integrated, online analytics. Often the most challenging aspect of implementing real-time monitoring is ensuring that the analytical instrumentation is receiving a representative sample from the process. The Small Molecule Design and Development organization at Eli Lilly and Company has developed a highly adaptable process-to-analytics interface that has been broadly implemented across the development portfolio. The automated sample carts use only gravity and low-pressure nitrogen to obtain representative process samples, typically on the order of 0.3–6 mL. Subsequent sample preparation operations including quenching, derivatization, and most commonly dilution (10–250× is customary, but not restrictive) in a high-performance liquid chromatography-compatible diluent. Samples are then delivered to the receiving online chromatographic instrument in a manner conducive to injection and analysis for that specific platform. These carts have demonstrated robust performance, and qualification data routinely meets less than 2.0% relative standard deviation on sample-to-sample reproducibility and less than 5% carryover between samples when optimized. Implementation within and across pharmaceutical processes have spanned 5 orders of magnitude in volumetric scale, while completing tens of thousands of on-process cycles. Automated sampling has proven invaluable to our organization, significantly increasing the speed of development/optimization for continuous unit operations by providing real-time process understanding and rapid diagnosis and troubleshooting of events. Operational criteria and performance assessment of the carts are provided. Several applications will be highlighted with integration to a variety of continuous manufacturing platforms in both pharmaceutical development laboratories and good manufacturing processes manufacturing. Cart function is thoroughly described in Supporting Information.