Sodium Laureth Sulfate: What It Does, and the Dioxane Question

Two questions bring people to that name: what it actually does, and whether it is interfering with their hormones. Both have clean answers, and the second one is no — but reaching it means pulling apart three things the internet keeps welding together: SLESSodium laureth sulfate, also written sodium lauryl ether sulfate. An ethoxylated anionic surfactant. CAS 9004-82-4., SLS, and a contaminant that rides along with one of them but not the other. That contaminant is the part worth your attention. It is not an endocrine disruptor either.

SLES is an anionic surfactantA molecule with a negatively charged water-loving head and a hydrocarbon oil-loving tail. The negative charge is what makes it foam heavily and what makes it interact with proteins on the skin surface. — a molecule built in two incompatible halves. One end is a twelve-carbon hydrocarbon tail that will not dissolve in water but will happily bury itself in oil. The other is a sulfate group carrying a negative charge, which water is glad to surround. Drop enough of them into water and they self-organise into micellesSpherical clusters of surfactant molecules, tails inward and charged heads outward. The oily interior is where grease, sebum and residue get trapped and held until the rinse carries the whole cluster away. — tiny spheres, tails tucked inside, heads facing out. Sebum and the day's grime end up in the oily interior, suspended in water that would otherwise refuse to carry them, and the rinse takes the lot down the drain.

It has two other jobs and only one of them is real work. It foams, in a way consumer testing has shown for decades makes people rate a product as cleaning better even when it is not. And it thickens: SLES solutions turn viscous in response to ordinary salt, which is how shampoo gets its pouring behaviour without a dedicated thickener. Cheap, effective, foamy, easy to formulate around. That is why it is in almost everything that lathers.

The name is the useful part. Lauryl alcohol — a twelve-carbon fatty alcohol, usually derived from palm kernel or coconut oil — gets treated with ethylene oxideA small, highly strained three-membered ring of two carbons and one oxygen. That strain makes it extremely reactive: it pops open and bolts onto whatever it meets, chaining up one unit at a time. before the sulfate group is attached. Each ethylene oxide molecule that reacts adds a short water-loving link to the chain. The eth in "laureth" is short for ether, and it is a marker: it tells you that step happened. Commercial SLES carries an average of one to three of those links — SLES-2 is the common grade. SLS carries zero. That is the entire chemical difference between the two names, and it drives everything else in this article.

Those ranges come from ordinary formulation practice and the concentration-of-use data the CIRCosmetic Ingredient Review — the US expert panel that assesses cosmetic ingredient safety gathered for its assessment of SLES and related ethoxylated alkyl sulfates — data the panel itself flagged as having real gaps Robinson et al. 2010. No brand prints its percentage. The order on the ingredient list is the only clue you get, and it stops being informative below 1%.

Yes, and it is one of the few claims in this corner of the ingredient list with a direct head-to-head measurement behind it. A group at the University of Marburg patch-tested SLS, SLES and an alkyl polyglucosideA non-ionic surfactant class built from glucose and fatty alcohols — decyl glucoside and coco-glucoside are the cosmetic examples. Milder than either sulfate, and foams considerably less. on volunteers at five concentrations from 0.125% to 2%, over exposures of 6, 12 and 24 hours, then measured the skin with TEWLTransepidermal water loss — how fast water escapes through the skin per unit area. The standard non-invasive measure of barrier damage in dermatology. and laser Doppler flowmetry at 24 hours, 7 days and 10 days after the patches came off. Their finding, in their words: a pronounced reaction to SLS, and a far milder one to SLES Löffler & Happle 2003.

The recovery curve is the more revealing half. Skin exposed to SLS still showed raised water loss at day ten, at every concentration tested. SLES-exposed skin was measurably irritated only out to about day seven. The alkyl polyglucoside was hard to detect at all even at the highest concentration, and had nothing significant left by day three. One experiment, three surfactants, a clean ordering — and the sugar-derived one is a long way behind both sulfates. Hold on to that; it comes back at the end.

The mechanism is not mysterious. The bare sulfate head on SLS sits close to the tail and is aggressive towards surface proteins — it denatures keratinThe structural protein that makes up the outer layer of skin, plus hair and nails. Surfactants that unfold keratin loosen the barrier that holds moisture in. and disorders the lipid layer that holds water in. Insert two or three ethylene oxide links and the charged head is held further out, wrapped in its own shell of water, bulkier and less able to wedge into a protein. The molecule still cleans. It does less collateral damage on the way past.

That is the trade. Switching an SLS formula to SLES buys real, measured skin tolerance, and it buys it with an extra chemical reaction — which is where the actual problem in this article comes from. Nobody made a bad decision. The consequence still has to be managed.

Ethylene oxide is reactive because it is strained — a three-membered ring holding bond angles it does not want. Put it under pressure with lauryl alcohol and a catalyst and it snaps open and attaches, then another does the same to the end of the growing chain, and so on. The intended reaction is that chain. The side reaction is two ethylene oxide molecules finding each other instead and closing into a stable six-membered ring: 1,4-dioxaneA colourless, volatile cyclic ether. Not an ingredient — a byproduct formed when ethylene oxide reacts with itself during ethoxylation. Classified by IARC as Group 2B, possibly carcinogenic to humans..

This is the point that catches people out: 1,4-dioxane is not in the product because anyone put it there. It is a contaminant of the raw material, and cosmetic labelling rules cover ingredients that were deliberately added. A byproduct riding in on a surfactant is not an ingredient, so it does not appear in the list — not because anyone is hiding it, but because the list was never designed to show it. Two bottles with identical ingredient panels can differ by an order of magnitude in what they actually contain. There is no label-reading technique that solves this. That is the whole difficulty.

The hazard case rests on animal carcinogenicity, and everyone who has looked has said a version of the same thing. IARCInternational Agency for Research on Cancer evaluated 1,4-dioxane in Monograph Volume 71 in 1999 and recorded inadequate evidence in humans, sufficient evidence in experimental animals, and an overall classification of Group 2B — possibly carcinogenic to humans. The US NTPNational Toxicology Program Report on Carcinogens lists it as reasonably anticipated to be a human carcinogen "based on sufficient evidence of carcinogenicity from studies in experimental animals", while noting the human epidemiological data are not adequate to evaluate. The EPAUnited States Environmental Protection Agency calls it likely to be carcinogenic to humans on the same split. Three bodies, one honest structure: the animal data are convincing and the human data barely exist.

Then there is how much of it actually reaches you, which the FDA has measured rather than modelled. Its own absorption work found that 1,4-dioxane does penetrate animal and human skin from some preparations — lotions in particular — and then found that it evaporates readily enough to further diminish "the already small amount available for skin absorption, even in products that remain on the skin for hours." A rinse-off product gets a fraction of the contact a lotion does. The reason regulators went after the contamination anyway was not that one shower is dangerous. It is that the whole thing is avoidable.

And it has largely been avoided, which is the part of this story that rarely travels. The FDA ran ten surveys of finished cosmetics between 1981 and 1997. The 1981 average was 50 ppm average 1,4-dioxane in finished cosmetic products in FDA's 1981 survey, across a range of 2 to 279 ppm across a range of 2 to 279 ppm; by 1997 the average was 19 ppm average in FDA's 1997 survey, across a range of 6 to 34 ppm — the last survey before the detection methods changed across a range of 6 to 34 Black et al. 2001. An eleventh survey in 2008, using a method sensitive to 1 ppm, detected none at all in 80% of products and topped out at 11.6 ppm. A 2018 survey of 82 children's products found two above 10 ppm. The number went down by roughly an order of magnitude over two decades, because manufacturers were told how to fix it and mostly did.

The fix is vacuum strippingA post-reaction step in which the surfactant is held under vacuum at controlled temperature so volatile contaminants boil off. FDA has recommended it since the 1980s. Routine for suppliers selling into regulated markets; not universal. — hold the finished surfactant under vacuum and the dioxane, far more volatile than the surfactant around it, leaves. The FDA has been recommending it since the 1980s. Suppliers who do it can sell into New York, Vermont and the EU. Suppliers who skip it cannot. The bottle on the shelf looks identical either way, which is precisely why this got settled by legislation rather than by shoppers.

New York's regime — Chapter 613 of the Laws of 2019, which added Section 37-0115 to the state's Environmental Conservation Law — is the strictest in the United States. Vermont's is the newest: Act 131 of 2024 bars knowingly selling any cosmetic or menstrual product containing 1,4-dioxane at or above 10 ppm, in force since 1 January 2026. Neither does its most useful work inside its own borders. Large manufacturers do not run a separate line for one state, so meeting 1 ppm in New York generally means meeting it in Ohio and in Manchester too. A state limit on a national supply chain is a global limit with extra steps.

No — and it is worth being precise about what that "no" is resting on, because the honest version is more useful than a flat reassurance. There is no published report of SLES activating or blocking the oestrogen receptorThe protein that oestrogen normally switches on. Endocrine disruptors work by fitting into this or a related receptor and either activating it or blocking it. Fitting requires a specific molecular shape. or the androgen receptor, no animal work showing altered hormone levels from dermal exposure, and no regulator anywhere that has proposed it as a suspected endocrine disruptor. When the CIRCosmetic Ingredient Review expert panel reviewed SLES and its related sulfated ethoxylated alcohols, it reported that sodium and ammonium laureth sulfate "have not evoked adverse responses in any toxicological testing", that SLES is "a dermal and ocular irritant but not a sensitizer", and — flagging genuine gaps in the use-and-concentration data — that these ingredients "should be used only when they can be formulated to be nonirritating" Robinson et al. 2010. An irritant, an eye irritant, not an allergen. Hormones do not appear in the record at all.

That negative needs a qualifier, because "no" is doing two jobs at once. SLES has never been the subject of a dedicated hormone-activity investigation the way the bisphenols and the phthalates have — it was never a suspect, so nobody built the study. What exists is a molecule with no structural feature that could plausibly engage a hormone receptor, an exposure history running to billions of person-years, and no signal from anyone who has examined it for any other reason. That is a strong practical position. It is not a purpose-built programme coming back clean. The accurate label is "no evidence of endocrine activity" — not "tested and cleared."

So where does the question come from? Four places. Three of them are real chemistry attached to the wrong molecule.

The first is a genuine case of mistaken family. There is a class of ethoxylated surfactants that is properly, repeatedly documented as oestrogenic — the alkylphenol ethoxylatesIndustrial surfactants built on a phenol ring with an alkyl side chain, then ethoxylated. Used in detergents, pesticide formulations and textile processing. Largely restricted in the EU since 2005., principally the nonylphenol and octylphenol derivatives used for decades in industrial detergents and pesticide formulations. The finding was an accident: in 1991, a Tufts laboratory watching breast-cancer cells proliferate in culture with no oestrogen present traced the activity to nonylphenol leaching out of the polystyrene tubes the experiments were running in Soto et al. 1991. Follow-up work confirmed that the alkylphenolic breakdown products activate the oestrogen receptor directly White et al. 1994, and that sewage effluent containing them was weakly oestrogenic to fish Jobling & Sumpter 1993.

Both families are ethoxylated surfactants. Only one of them has the part that matters. What makes nonylphenol oestrogenic is the phenol ringA benzene ring carrying a hydroxyl group. Its size, flatness and hydrogen-bonding geometry closely mimic one end of the oestradiol molecule — close enough to fit the receptor's binding pocket. — a flat aromatic ring with a hydroxyl on it, shaped closely enough like the corresponding end of oestradiol to slot into the receptor's pocket. Sodium laureth sulfate has no ring anywhere in it. It is a straight saturated carbon chain, a short ether tail and a sulfate. There is nothing for the receptor to grip. Ethoxylation is a process, not a hazard class. Two molecules can share a manufacturing step and share nothing else.

The second is the dioxane story arriving without its label. 1,4-dioxane's hazard is rodent liver and nasal tumours after lifetime oral dosing — a carcinogenicity question. It has no established hormonal mechanism and is not classified as an endocrine disruptor by any regulator. In circulation, "contains a possible carcinogen" flattens into "bad for you" and then re-inflates as whatever the reader was already worried about. Keeping the two apart matters, because they point at different fixes: hormone activity would be a reason to avoid the ingredient, whereas a manufacturing contaminant is a reason to prefer suppliers who strip it.

The third is inherited from SLS. A chain email circulating from the mid-1990s attributed cancer risk to sulfate surfactants on the authority of a University of Pennsylvania study that nobody has ever produced. It was never a hormone claim, and it was corrected repeatedly, but it established the background sense that sulfates are the dangerous thing in a bottle — which is the bed the hormone question now grows in. The SLS article covers what the evidence on that molecule does and does not support.

The fourth is the rest of the bottle, and it is the one with substance. A shampoo containing SLES also contains a preservative system and a fragrance. Parabens have measurable, weak oestrogenic activity in laboratory assays. Phthalates travel inside the single word "fragrance". DMDM hydantoin works by releasing formaldehyde. Somebody who switched to a sulfate-free shampoo and felt better may well have changed something real — but the sulfate is the ingredient they can name, and it is rarely the ingredient that changed. SLES is the most visible thing in the bottle and the least likely to be the hormonal one.

As an ingredient in its own right, SLES sits low on any sensible priority list: a rinse-off irritant with a dose-dependent, reversible, local effect, milder than the molecule it replaced, with no systemic or hormonal case against it. If your skin tolerates it, the effort is better spent on the preservative and the fragrance further down the same list.

The contaminant is the awkward part, because it is the one thing here you cannot verify from the pack. What you can do is read for the process rather than the molecule — and the FDA publishes the exact markers to read for.

None of that is a purity ritual. Ethoxylated ingredients are useful and, stripped properly, unremarkable — and the FDA's own survey data show the stripping mostly happened. Reading for the process is narrower than avoidance: it tells you which questions a label can answer and which it cannot, and 1,4-dioxane sits firmly in the second group.

Sodium laureth sulfate does exactly what it looks like it does: strips oil, makes foam, thickens with salt. It is milder than the molecule it largely displaced, and the reason it is milder is the same reason it needs a purification step the label will never mention.

The hormone question, asked honestly, comes back empty. Not "probably fine" and not "cleared" — empty. No receptor evidence, no animal signal, no structural reason to expect either. The ethoxylated surfactants that genuinely do interfere with oestrogen signalling are the alkylphenols, which share a factory process with SLES and nothing else. If you came here worried about what is in your shampoo, you were looking at the right bottle and the wrong line. Read the preservative. Then read the word 'fragrance.'