Quick answer: Semi-solid pharmaceutical scale-up from lab to production: specific energy, tip speed, media size, residence time and heat control.
Semi-solid pharmaceutical products — ointments, creams, gels and suspended systems — are usually developed in a laboratory mixer or a small bead mill and then transferred to a production line. The transfer is where many projects lose time: the lab batch looks perfect, the first production batch does not, and the difference is attributed to "scale-up" without a clear explanation. In most cases the cause is not mysterious. It is a change in one of the variables that control dispersion and heat transfer, made without recognising that the variable mattered.
This guide explains what changes between lab, pilot and production scales for semi-solid products, which parameters should be held constant, and how to structure a scale-up trial so the production batch behaves like the lab batch.
Why semi-solid scale-up is different
In a low-viscosity suspension, mixing and pumping are relatively easy, and the main scale-up variable is the energy delivered to the particle bed. In a semi-solid, viscosity is high and often non-Newtonian, heat transfer is slower, and the product may be sensitive to both temperature and shear history. The same formulation can behave differently in a 2 L lab vessel and a 2,000 L production vessel because the surface-to-volume ratio, the wall film, the mixing pattern and the residence time distribution all change.
The practical consequence is that a semi-solid scale-up cannot rely on geometric similarity alone. A production vessel that is a scaled-up copy of the lab vessel will not necessarily produce the same dispersion, because the local shear and the heat history are different. Process similarity — the same specific energy, the same tip speed range and the same temperature profile — matters more than matching the drawing.
What changes between lab, pilot and production
At the lab scale, a batch is typically 0.5–5 L, the mixer or mill is small, and the jacket can remove heat quickly because the surface-to-volume ratio is high. At the pilot scale, 50–200 L, the batch is large enough to reveal mixing problems but small enough to adjust, and the equipment is usually a scaled version of the production design. At production scale, 500–5,000 L, heat removal becomes a limiting factor, and the residence time in a recirculation loop can be long enough to create a temperature gradient.
Each scale has a different dominant risk. The lab risk is that the formulation is not representative because the mixer is too small to reproduce the production shear. The pilot risk is that the equipment is not geometrically representative, so the result does not extrapolate. The production risk is that heat and residence time are underestimated, so the product is over-heated or over-sheared even though the energy per kilogram appears correct.
Parameters to hold constant
Specific energy is the first parameter. It is the energy delivered per kilogram of product, and it controls the amount of size reduction for a given formulation and media size. If the lab batch reaches the target at 0.2 kWh/kg, the production batch should start at the same specific energy, not at the same motor power or the same milling time.
Tip speed is the second. Tip speed controls the intensity of the shear at the media surface and the rate of wear. A production mill with a larger rotor can reach the same tip speed at a lower shaft speed, but if the tip speed is increased to shorten the cycle, the particle size and the wear rate change. Keep tip speed in the same range and adjust the number of passes instead.
Residence time is the third. In a recirculation loop, the product passes through the mill many times, and the residence time per pass is the chamber volume divided by the flow rate. If the production chamber is larger but the pump rate is scaled proportionally, the residence time per pass stays the same; if the pump rate is not scaled, the product may be over- or under-milled.
Media size and fill should also be held constant. Changing media size to compensate for a larger chamber changes the number of contacts and the wear rate, and it changes the shape of the distribution. If a larger media size is unavoidable, the lab trial should be repeated at that size rather than extrapolated.
Equipment configuration by scale
At the lab scale, a small horizontal bead mill with a ceramic chamber and a jacketed vessel is usually sufficient. The key requirement is that the mill can run at a representative tip speed and that the temperature can be controlled, because a lab mill that overheats will not predict the production result.
At the pilot scale, the equipment should be a scaled version of the production design: same chamber geometry, same separator type, same media size range and same instrumentation. This is the stage where the scale-up parameters are confirmed, and where the cleaning procedure is first tested. A pilot mill that is mechanically different from the production mill is a poor predictor, even if it is the same brand.
At production scale, the mill is usually integrated with a vacuum disperser or an emulsifying kettle, and the batch is transferred through a recirculation loop. The loop volume, the pump rate and the jacket capacity determine the temperature profile. The production line should be designed with a temperature probe at the mill outlet and a data log, so the batch record shows what the product actually experienced.
Common scale-up mistakes
Scaling by volume only. Doubling the batch volume does not double the energy per kilogram; it doubles the volume over which the same energy density is applied. The specific energy must be recalculated.
Changing media size. A larger mill is often supplied with larger media because the separator gap is larger. If the media size changes, the distribution changes. Either keep the media size and adjust the separator, or repeat the lab trial at the new size.
Ignoring heat. A lab batch can lose heat through the vessel wall in minutes; a production batch cannot. If the lab trial did not record the product temperature at the mill outlet, the production cooling requirement is unknown.
Changing the addition order. The order in which the API, stabiliser and vehicle are added affects wetting and dispersion. A production recipe that changes the order for convenience can produce a different particle size even with the same equipment.
A practical scale-up protocol
Run the lab trial on the actual formulation, not a model system. Record the specific energy, tip speed, media size, residence time, product temperature and particle-size distribution at the end of the trial. If the formulation is sensitive to shear history, take samples at two or three energy levels rather than only at the end.
At the pilot scale, reproduce the lab parameters and compare the distribution. If the distribution is wider, the cause is usually a change in residence time distribution or media size, not a change in the formulation. Adjust the pump rate or the number of passes before changing the formulation.
At production scale, start at the pilot specific energy and adjust in small increments. Take a sample after each pass and measure the distribution with the same method used in the lab. Stop when the target is reached; do not continue "for safety", because over-milling narrows the distribution but can also increase surface area and change dissolution or texture.
Selection checklist
- Is the lab mill geometrically representative of the production mill?
- Are specific energy, tip speed, media size and residence time recorded?
- Is the product temperature measured at the mill outlet, not only in the jacket?
- Is the cleaning procedure tested at the pilot scale?
- Is the analytical method for particle size fixed across all scales?
- Is the production line able to hold the required temperature during the full batch?
- Is the documentation package sufficient for your quality system?
If these points are addressed before the first production batch, scale-up becomes a predictable engineering step rather than a trial-and-error exercise. POLYC supplies laboratory, pilot and production equipment for semi-solid pharmaceutical products, including planetary mixers, vacuum dispersers, emulsifying kettles and horizontal bead mills, and can review the scale-up route with your formulation.
Frequently asked questions
Can a lab bead mill predict production particle size?
It can, if the lab mill is geometrically representative and the specific energy, tip speed, media size and residence time are recorded. A lab mill that is mechanically different from the production mill is only a rough indicator.
Should the production batch use the same milling time as the lab batch?
No. Milling time is not a scale-independent parameter. The production batch should use the same specific energy and adjust the number of passes or the flow rate to reach the target distribution.
Why does the production batch overheat?
Because the surface-to-volume ratio is lower and the batch is in the loop for longer. The cooling circuit must be sized for the full heat load, and the product temperature should be measured at the mill outlet rather than in the jacket.
Is it acceptable to change media size at production scale?
Only if the lab trial is repeated at the new media size. Changing media size changes the number of contacts and the wear rate, so the distribution cannot be extrapolated from a different size.
What should be documented during scale-up?
Specific energy, tip speed, media size, residence time, product temperature, particle-size distribution and the analytical method. These records make the production batch reproducible and support qualification.
Next step
Share your formulation, lab batch size and target distribution with POLYC MACHINE engineers. We will review the scale-up route and propose a pilot configuration that matches your production design.
Source: POLYC MACHINE — Shanghai Polyc Technology Co., Ltd. (polyc.site), manufacturer of wet grinding, dispersion and mixing equipment for coatings, inks, adhesives, battery materials, pharmaceuticals and food since 2006.
Related technical guides
Standards referenced
Method references used by process engineers when specifying and verifying the operations described in this guide.
- ISO 13320:2020 — Particle size analysis — Laser diffraction methods
- ISO 1524:2020 — Paints, varnishes and printing inks — Determination of fineness of grind