How do I prove a treatment is working, layer by layer?

Four layers, in this order: surface properties, dispersion state, formula performance, then stability. Skipping to the last one is why treatments get blamed for problems they did not cause.

Surface first, because it is cheap and direct. Contact angle of a water drop on a pressed disc should rise clearly on a hydrophobic treatment. Oil absorption should fall. FTIR shows the agent's characteristic peaks if you need proof of identity rather than effect.

Then dispersion: put the powder into the target oil phase, read fineness of grind, look for agglomerates under a microscope, and watch sediment and redispersibility on standing.

Then the real formula: viscosity, yield value, sensory panel, and the functional claim — SPF, oil control, wear. Finally stability: 40°C, freeze-thaw, centrifuge, checking for layering, coarsening and oil separation.

What makes all of it meaningful is the control. Every test needs an untreated sample of the same base powder alongside it, otherwise you cannot separate the effect of the treatment from the effect of the formula. Fix a short list of these tests and run the same list for every supplier and every batch.