Abstract

Evolution of Kinetic Motifs Through Rate-based Experimental Design in Flow Reactors

Christopher Hone, Research Scholar, University of Leeds

Continuous processing technology is transforming the way that pharmaceuticals are manufactured. Currently there is no defined scale-up path for flow technologies within high value manufacturing environments which is a major barrier to the uptake of continuous processing. Statistical approaches, such as Design of Experiments and numerical self-optimising systems, identify a single set of optimal operating conditions that maximises yield or other reaction metric. These techniques optimise well in the equipment used for the optimisation, but do not reveal an explanation as to why a response is dependent on a particular input. The primary focus of the methodology described in this paper is to generate models which describe the rate-limiting kinetics utilising continuous flow technology. The approach can be applied with small material quantities and reflects that there is limited time and resource available for process development. Scale-up using such models significantly reduces the risk compared to directly transferring laboratory conditions to the pilot plant scale. The methodology has been implemented for the rapid generation of kinetic models using sequential experimentation of the experimental design space. To arrive at a kinetic motif which best encompasses the reaction system, the largest possible process window is explored: ranging from the mildest (e.g. dilute, low temperature) to the harshest which is feasible in the equipment (e.g. concentrated, high reagent to substrate ratios, high temperature). Small scale experiments were conducted using a small scale automated coil flow reactor to collect data. The models were used to explore the design space in silico and identify the optimal operating region. Subsequently, the reactions were successfully scaled-up to a different reactor for chemical manufacture. The approach has been applied to a series of industrially relevant reaction case studies.