The emulsion split

A layer of oil on top, or a watery pool at the bottom, or a cream that has gone grainy. The emulsion is doing what emulsions do when nothing stops them.

What you are seeing

An emulsion is metastable: energy created it and thermodynamics is slowly undoing it. Visible separation is the end of a process that began the day it was made.

So the question is never why it separated. It is why it separated this soon.

The three mechanisms

Creaming is droplets rising or sinking without merging. It looks like separation and often redisperses on shaking. It is a warning rather than a failure.

Coalescence is droplets merging because the interfacial film failed. It does not redisperse. This is the real failure.

Phase inversion is the system flipping from oil-in-water to water-in-oil or back, usually under heat or electrolyte stress. The product changes character entirely.

What actually caused it

An HLB mismatch between the emulsifier and the oil phase, which is a formulation error present from the first batch.

Electrolytes, which is the surprise. Magnesium salts, zinc oxide and high preservative loads dehydrate the ethoxylate chains on many emulsifiers, lowering the effective HLB and pushing the system toward inversion. Adding a mineral sunscreen or raising the preservative after a marginal challenge test both do this.

Process: insufficient shear, phases combined at different temperatures, or cooled too fast.

Temperature in transit or storage, which is what freeze-thaw testing exists to find.

Catching it before the customer does

Centrifuge at 3,000 to 5,000 rpm for thirty minutes as a fast screen: it accelerates phase separation and tells you in half an hour what a shelf would take months to reveal.

Then the full protocol: accelerated storage at 40 and 45 °C for four and eight weeks, and three to five freeze-thaw cycles between −10 and +25 °C. Products that survive all of that rarely split in a bathroom.

References