Blog / Process development
From lab route to plant: what changes when an API process scales up
Heat, mixing, time and solids all behave differently in a reactor than in a flask. Here is what a process-development team checks at each step.
A route that gives clean product in a one-litre flask can fail in a two-thousand-litre reactor without a single change to the chemistry. The reaction is the same. The physics is not.
Heat stops being free
As a vessel grows, its volume rises faster than the surface available to remove heat. An exotherm that a lab ice bath absorbs in seconds can take a plant jacket much longer. That is why additions that are instant on the bench become controlled feeds in the plant, and why reaction calorimetry is done before scale-up, not after a surprise.
Mixing changes the impurity profile
Where a reagent meets the batch matters. Poor dispersion creates local excess, and local excess creates by-products that never appeared in the lab. Agitator type, feed position and addition rate become process parameters in their own right.
Everything takes longer
Heating, cooling, charging, distilling and filtering all stretch from minutes to hours. An intermediate that was stable for the length of a lab work-up may need to hold overnight. Hold-time studies answer whether it can.
Solids are the hard part
Crystallisation, filtration and drying rarely scale in a straight line. Crystal size and form affect how fast a cake filters, how well it washes and how long it takes to dry. A slow filtration is often a crystallisation problem in disguise.
What the ladder is for
Moving through lab, pilot and small commercial batches before bulk manufacture is how these questions get answered at a cost you can afford. Each rung tests the process in equipment closer to the final plant, and the data from one rung sets the limits for the next.
A short checklist before you scale
- Is the heat of reaction measured, and can the plant remove it?
- Are critical additions defined by rate, not just quantity?
- Are hold times proven for every intermediate stage?
- Is the crystallisation controlled, including seeding where needed?
- Is the impurity profile tracked from lab batch to pilot batch?
Nivika runs processes from milligram scale in the lab to tonnes a month in the plant. To discuss a molecule, start a development brief.