Two buyers order what looks like the identical raw material - same INCI, same CAS, same EO number - and one drum arrives with 1 ppm of 1,4-dioxane while the next arrives near 30 ppm. The certificate difference is not an accident. 1,4-dioxane is a by-product of the ethoxylation reaction itself, and whether it leaves the plant depends on whether the reactor is followed by a stripping stage and what grade the customer asked for. If your purchase order does not state a ppm limit, the supplier's default grade decides.
Where the residue comes from
Ethoxylated surfactants are made by reacting a starter (fatty alcohol, sorbitan ester, castor oil) with ethylene oxide. Along the way, ethylene oxide can dimerise into 1,4-dioxane. The amount depends on reaction conditions: temperature, catalyst, reaction time, and in the sulfation of alcohol ethoxylates (AES/SLES), on the SO3-to-feed ratio - push it slightly above optimum and dioxane multiplies several-fold. For AES made under well-controlled modern conditions, literature reports roughly 20-30 ppm as shipped, about ten times lower than typical levels in the 1980s.
What makes the residue stubborn is its physical profile: 1,4-dioxane boils at 101 °C, mixes with water in all proportions, and is chemically stable. It survives blending, dilution and repackaging. Nothing downstream removes it - only a vacuum steam stripping or nitrogen purge step at the plant does. Vacuum steam stripping can cut the level by roughly a factor of eight, down to 1-5 ppm; suppliers sell these as tiers - a standard grade (often <20 ppm), a low-dioxane grade (<10 ppm) and, at a price, an ultra-low grade (<5 ppm).
The limits, by market
1,4-dioxane is classified IARC Group 2B (possibly carcinogenic) and residual ethylene oxide is Group 1. No market allows it as an intentionally added substance; the differences are in the trace-level tolerance for finished products:
| Market / body | Limit for finished product | Basis |
|---|---|---|
| New York State, USA | 10 ppm cosmetics; household cleansing and personal care 2 ppm from Dec 31, 2022, tightened to 1 ppm from Dec 31, 2023 | ECL Articles 35/37, enforced by NYS DEC |
| EU | Prohibited as intentionally added; SCCS/1570/15 accepts 10 ppm as an acceptable trace level | Annex II of Regulation (EC) 1223/2009; SCCS opinion |
| ASEAN | Prohibited as intentionally added; not more than 10 ppm as trace impurity | ASEAN Cosmetic Directive Annex II; Guidelines on Limits of Contaminants |
| China | Not more than 30 mg/kg in the finished cosmetic product | Safety and Technical Standards for Cosmetics |
| California, USA | Proposition 65 warning threshold of 30 micrograms/day exposure | Exposure-based, no fixed product ppm |
Read the table twice and notice the asymmetry: China is the loosest at 30 ppm, New York personal care the strictest at 1 ppm. A global SKU must therefore be built around the strictest end-user rule, not the average.
The dilution arithmetic (and why raw-material ppm decides compliance)
Every limit above applies to the finished product, but you buy raw materials. The conversion is simple: finished ppm = sum of (ingredient weight % x ingredient ppm) / 100.
- A 70% active surfactant shipped at 10 ppm, dosed into a shampoo at 10% of the formula, contributes about 1 ppm.
- Ethoxylated ingredients add up: 6% ceteareth-25 at 15 ppm plus 2% polysorbate-20 at 10 ppm plus 1% PEG-40 hydrogenated castor oil at 10 ppm gives 0.9 + 0.2 + 0.1 = 1.2 ppm in the finished bottle - already over New York's 1 ppm personal-care line before any other contribution.
That is why a 1 ppm finished-product limit is really a raw-material specification of <5 ppm for every ethoxylate in the formula, purchased as the low-dioxane tier. If your formula carries five or six ethoxylated components, ask for each COA line; a generic "compliant" statement without a number is not verifiable.
What to write into the purchase order
- A per-lot numeric value for 1,4-dioxane (and ideally residual ethylene oxide), not a compliance statement. Head-space GC-MS is the standard method with detection limits of 0.5-1 ppm; ask which method was used and the detection limit.
- Batch-to-batch consistency: if a supplier declares <10 ppm, ask for the range of recent lots. A responsible manufacturer keeps the statistics.
- For pharmacopoeial-grade polysorbates, note the compendial expectation is already tight: ethylene oxide not more than 1 ppm and 1,4-dioxane not more than 10 ppm. Technical grades may exceed both and are not interchangeable.
- Remember the CAS trap: the CAS number is a family identifier, not a grade. Ceteareth-2, -20, -21 and -25 all share CAS 68439-49-6; beheneth-10/20/25 share 26636-40-8. The dioxane spec - and the EO number - are what actually distinguish two drums.
Where our range sits
We supply the classic ethoxylated emulsifier families with low-residue documentation on request: Ceteareth-2 (CAS 68439-49-6), Ceteareth-20, Ceteareth-21, Ceteareth-25, Beheneth-10 (CAS 26636-40-8), Beheneth-20, Beheneth-25, Steareth-21 (CAS 9005-00-9), Polysorbate 20 (CAS 9005-64-5), Polysorbate 60 (CAS 9005-67-8), Polysorbate 80 (CAS 9005-65-6), Polysorbate 85 (CAS 9005-70-3) and PEG-40 Hydrogenated Castor Oil (CAS 61788-85-0). For how the EO number shapes performance across these families, see our fatty alcohol ethoxylate guide and the PEG ester ester-count guide; for the saturated-basis raw material behind the castor-oil ethoxylates, see tomorrow's companion piece on iodine value and unsaturation.