270997 Integrated Continuous-Flow System for Automated Online Optimization of Chemical Synthesis

Thursday, November 1, 2012: 2:15 PM
323 (Convention Center )
Nikolay Zaborenko1, Marc Champagne2, Wei-Ming Sun3, Michael T. Miller4, Todd D. Maloney4 and Edward Sheldon5, (1)Small Molecule Design and Development, Eli Lilly and Company, Indianapolis, IN, (2)Analytical R&D, Eli Lilly, Indianapolis, IN, (3)Chemical Product Research and Development, Eli Lilly &Co., Inc., Indianapolis, IN, (4)Analytical Technologies Design and Development, Eli Lilly & Co., Indianapolis, IN, (5)Analytical Technologies Design & Development , Eli Lilly & Co., Indianapolis, IN

A great deal of experimental effort and cost is expended in chemical research and development towards mapping reaction response surface areas and determining optimal operating conditions.  Frequently, such experimentation is performed by synthetic chemists in flask (batch) reactors, leading to overall large material consumption and time expenditure.   Automated integration of continuous-flow micro-scale reactors with fluid delivery, reactor temperature control, and analytical equipment, combined with a decision-making algorithm, has been previously demonstrated in an academic setting to be a very powerful tool for real-time optimization with minimal material usage or operator involvement.[1]  Herein, we present the application of this concept to a pharmaceutical R&D laboratory system.  DeltaV was used as the communication framework to integrate and communicate with a network of robust high-pressure syringe pumps, a recirculating heater/chiller, and an on-line PAT HPLC, allowing widespread remote user interfacing.  MATLAB was used as the system engine, coordinating all of the devices and applying the process algorithm.   The system was used to apply Dantzig’s simplex algorithm to a reaction step in an actual investigational pharmaceutical route.  The reaction was performed in a micro-scale tube reactor, with the algorithm automatically optimizing product yield and selectivity by varying residence time, temperature, and equivalents of two components.  The system design is fully modular, allowing for a wide variety of optimization or model investigation algorithms, reactor configurations, and target fitness functions.

[1] McMullen, J. P. 2010. Automated microreactor system for reaction development and online optimization of chemical processes.  Ph. D. thesis, Massachusetts Institute of Technology, Cambridge, MA.

Extended Abstract: File Not Uploaded
See more of this Session: Optimization and Predictive Control II
See more of this Group/Topical: Computing and Systems Technology Division