When a cream separates in transit, the first technical question is not "which emulsifier is bad" but "is this an O/W that inverted, or a W/O that was never really an O/W". The HLB value printed on a TDS is measured at 25 °C and says nothing about what happens at 80 °C on the production line or 55 °C inside a summer sea container. Ethoxylated emulsifiers quietly lose hydrophile character as they warm up, and the emulsion type itself can flip. This article covers the three accepted ways to identify emulsion type, the phase inversion temperature (PIT) behind them, and the procurement numbers that keep an O/W an O/W.
Three tests, one standard — and why two of three must agree
China's national standard GB/T 35827-2018 (General methods on determination of cosmetics — Identification of the emulsion type) prescribes three qualitative tests for creams and lotions:
| Test | Principle | Procedure | Judgement |
|---|---|---|---|
| Conductivity | The continuous phase conducts: water does, oil does not | Calibrated meter, range ≥10 µS/cm; immerse the electrode, watch for a reading | Any reading = O/W; no reading = W/O |
| Dilution | An emulsion dilutes into its own continuous phase | One drop into distilled water, stir 30 s (viscous samples may be warmed to 40±5 °C) | Disperses evenly = O/W; oil droplets or layering = W/O |
| Dye | The external phase takes up the dye | 1 g sample on a watch glass, two 1 mL drops of water-soluble dye (1 part dye + 9 parts water, e.g. brilliant blue), wait 2 min | Surface stains = O/W; stays unstained = W/O |
The standard requires at least two tests to agree before you call the type, and that rule exists for good reasons. A high-viscosity O/W may disperse slowly in cold water (hence the 40±5 °C allowance), an air bubble on the electrode can mask a reading, and highly structured gel networks can confuse any single observation. Run all three; they take minutes.
Why conductivity is the decisive one
Water conducts; oil essentially does not. An O/W tracks its water phase and reads readily; a W/O reads near the oil phase — effectively dead on the meter. That is why the standard sets the operational threshold as "any reading on a range of at least 10 µS/cm". It is the fastest single test, and in most disputes it settles the question. Just remember it answers "which phase is continuous", not "which phase is more abundant": a W/O can be 60% water and still read non-conductive.
PIT: the temperature HLB never told you about
Ethoxylated nonionic emulsifiers dissolve in water because their polyoxyethylene chains hydrogen-bond with water. Heat those chains and the hydrogen bonding weakens: the polymer dehydrates, the molecule becomes more lipophilic, and its effective HLB falls with temperature. At one specific temperature the affinities for oil and water balance exactly — the phase inversion temperature (PIT, also called the HLB temperature). There the interfacial tension hits a minimum and both emulsion types are unstable; above it a W/O is the stable form, below it an O/W.
Three practical consequences:
- Emulsifying near the PIT gives the finest droplets because interfacial tension is lowest there — this is Shinoda's PIT emulsification method. The droplet size is decided by how fast you cool through the inversion window; slow cooling means coalescence no matter how good the homogeniser is.
- Storage temperature should sit well below the PIT. A nano-emulsion study (n-decane/water with a C12 laureth ether) found the optimum storage temperature about 20 °C below the PIT, giving stable 35–54 nm droplets; departures from that window drove Ostwald ripening.
- The PIT is a property of the system, not of the surfactant: it moves with the oil phase, the phase ratio and the EO distribution of a commercial blend. A PIT quoted in a paper for another oil is a reference point, not your number.
Cloud point: the field proxy you can actually put in a spec
Measuring the PIT of a full formula is specialist work; the 1% aqueous cloud point is on most TDS and tracks the same dehydration mechanism (the same behaviour that drives PPG polyether cloud points in defoamers). Typical published values:
| Emulsifier | CAS | Cloud point (1% aq) | Margin vs 55 °C container |
|---|---|---|---|
| Polysorbate-20 | 9005-64-5 | ≈76 °C (literature 76–77 °C) | ≈21 °C |
| PEG-40 Hydrogenated Castor Oil | 61788-85-0 | 65–75 °C | 10–20 °C |
| Polysorbate-80 | 9005-65-6 | ≈65 °C (literature 56–65 °C) | 1–10 °C |
Container interiors are conventionally assessed at 55 °C. The arithmetic above is not a pass/fail verdict — the formula's PIT differs from the neat surfactant's cloud point — but the ranking is robust: more EO units, higher cloud point, wider margin. It is exactly why a higher-EO Ceteareth-25 outranks a Ceteareth-2 (oil-soluble, deliberately lipophilic) as the hydrophilic anchor of a hot-climate O/W, and why a Polysorbate-80-only system sits closest to its inversion window in a summer container. Note that the cloud point only applies once the emulsifier is in water: during the melt at 75–85 °C the emulsifier sits in the oil phase and nothing inverts. The dangerous moment is the slow cool-down through the inversion window, and the slow coast at 45–50 °C during hot filling.
Ionic emulsifiers don't invert with heat — their weakness is salt
Soap- and sulfate-based self-emulsifying bases (glyceryl stearate SE, cetearyl alcohol & sodium lauryl sulfate) have no EO chain to dehydrate, so they show no thermal inversion and ride through hot processing. Their fragility is electrolyte: hard water, calcium/magnesium salts and pH drift precipitate the soap and collapse the system. Ethoxylated nonionics are the mirror image — tolerant of salt, bounded in heat. Modern liquid stabilisers such as acrylamide/ammonium acrylate copolymer (and) mineral oil (and) laureth-7 and alkyl glucoside pairs (cetearyl alcohol & cetearyl glucoside) add a third route with its own electrolyte and pH profile. The design question is which stress your product actually meets in its lifetime — and as our earlier piece on self-emulsifying bases showed, One card can serve three industries only if you know which failure mode you are buying into — see our self-emulsifying bases guide.
The temperature ladder every O/W should clear
| Stage | Typical temperature | What can go wrong | Design answer |
|---|---|---|---|
| Melt & emulsify | 70–85 °C | Emulsifier is in the oil phase — no inversion yet; the risk is cooling too slowly through the inversion window | Rapid cooling, controlled cool-down path |
| Hot filling | 40–50 °C | Approaching cloud points of Polysorbate-80-class anchors | Check the margin arithmetic before the season |
| Sea transport | up to 55 °C | Lowest margin on low-EO or C18:1 anchors | Prefer Polysorbate-20 / high-EO ceteareths, or add an ionic co-emulsifier |
| Shelf storage | 15–30 °C | — | The window the formula was designed for |
What to lock in the PO
- INCI + EO number + CAS, triple lock. The CAS is a family number: Ceteareth-2 and Ceteareth-25 share 68439-49-6, and Oleth-2 and Oleth-20 share 9004-98-2 — only the EO number on the order separates them.
- Cloud point with the method (1% aqueous, named standard). A cloud point without a method is not a spec — the same number measured differently moves by degrees.
- For polysorbates, add peroxide value ≤5 meq/kg and 1,4-dioxane ≤10 ppm to the document set; both are residues of the ethoxylation step (see our 1,4-dioxane residue guide) and both age the product in ways a cloud point never shows.
- For W/O systems invert the logic: sorbitan oleate, sorbitan trioleate and PEG-30 dipolyhydroxystearate (and) sorbitan sesquioleate are chosen for water-phase tolerance and re-emulsification behaviour, not for a cloud point. If fragrance is going into the same system, re-check the solubilisation ratio first — the micelle rules from our fragrance solubilisation guide apply unchanged.
We supply the full emulsifier ladder referenced above, all with lot-level COA: Ceteareth-2 (68439-49-6), Ceteareth-25 (68439-49-6), Ceteareth-20 (and) Cetearyl Alcohol (68439-49-6; 67762-27-0), Oleth-20 (9004-98-2), Laureth-4 (5274-68-0), Polysorbate-20 (9005-64-5), Polysorbate-80 (9005-65-6), PEG-40 Hydrogenated Castor Oil (61788-85-0), Glyceryl Stearate SE (11099-07-3), Cetearyl Alcohol (and) Sodium Lauryl Sulfate (67762-27-0; 151-21-3), PEG-100 Stearate (9004-99-3), Sorbitan Oleate (1338-43-8), Sorbitan Trioleate (26266-58-4), PEG-30 Dipolyhydroxystearate (and) Sorbitan Sesquioleate (827596-80-5; 8007-43-0), Cetearyl Alcohol (and) Cetearyl Glucoside (67762-27-0; 246159-33-1) and Acrylamide/Ammonium Acrylate Copolymer (and) Mineral Oil (and) Laureth-7 (26100-47-0; 8042-47-5; 9002-92-0). Send us your target process and shipping temperatures — we will help you pick the anchor whose margin clears the ladder, and quote the cloud point with the method line attached.