(426g) Coupling Reaction Kinetics, Calorimetry and Mixing Studies for Successful Scale-up of Exothermic Reactions

In this work we showcase a procedure to troubleshoot a reaction that had very good outcome in the lab (an over-reaction key impurity was maintained at ≤1.2% area, which allows to meet the dry solid specification of 0.15% in area) while in the first large scale batch (30 kg batch size), the level of this impurity reached 30% area at reaction end-point, which did not allow its purge during downstream, failing to meet the impurity specification in the dry solid. The outline course of action was to build a kinetic model for the formation of product and the key impurity and use historical data to follow as close as possible the Bourne protocol (Bourne 2003) to understand the impact of mixing on the reaction outcome.

According to the reaction mechanism (A + B --> P, P + B --> Imp), the only way for very low levels of Imp impurity is for its formation reaction rate to be much lower than the product formation reaction rate, since they are competing-consecutive reactions. Even though this seems obvious, we wanted to confirm this hypothesis with a kinetic model.

A set of experiments was carried out to better understand the formation of the impurity, namely adding to a solution of product P, a solution of 1 eq. of reagent B, at two temperatures; Additionally, four other experiments were carried out namely adding a solution of B (1eq, 3 vol) to a 3 vol solution of starting material A at the same two temperatures and two feeding durations (2h and 1h, respectively), and the reverse addition (adding a solution of A to a solution of B, with the same temperature and feeding durations described previously). Furthermore, we also added data from a DoE (which explored reaction temperature, agitation speed and feeding duration) to give more robustness to the kinetic model. Data obtained from an RC1 experiment was used as well to assess the reaction exothermicity and, thus, determine the ΔHr, which is a critical aspect for scaling-up this reaction.

To assess the impact of the agitation on the reaction, we used some of the experiments from the previously described DoE, to apply the Bourne protocol (Bourne 2003), and thus help provide the needed insights to understand which of the micro-, meso- or macro-mixing mechanism would be the controlling mechanism, and thus have a science based approach to scale-up the mixing conditions of this reaction.

It was proven that by resorting to a mechanistic modelling approach and a reduced number of experiments we were able to successfully investigate the root cause of the failed batch and scale-up the reaction to large scale conditions and avoid further failed batches in the future.

References

Bourne, John R. 2003. “Mixing and the Selectivity of Chemical Reactions.” Organic Process Research and Development 7 (4): 471–508. https://doi.org/10.1021/op020074q.