The Coco Question

I. The Bar in the Hand

A shampoo bar sits on a kraft-paper wrapper on a shelf labelled natural, plastic-free, zero-waste. Fifty-five grams. A glassine band runs across the centre. The printed ingredients line on the back of the wrapper begins with one of four chemical names: Sodium Coco-Sulfate, Sodium Cocoyl Isethionate, Decyl Glucoside, or Sodium Lauryl Sulfate.

The bar replaces a bottle. UK shampoo-bottle recycling in mixed kerbside collection runs in the high single digits, so in most postal codes the displaced 250 ml HDPE bottle would have ended its life as litter, landfill, or downcycled flake. The bar shipped in paper from a UK or EU contract manufacturer. The shampoo-bar segment in 2025 was valued at roughly USD 11 billion globally, projected toward USD 19 billion by 2034 at a compound growth rate of 5.9%; the zero-waste sub-segment is the fastest-growing region of the parent personal-care category.¹ The format claim, on its own evidence, is true. A reader who switched to the bar for plastic-elimination has participated in a measurable material change. Nothing in this report is a complaint about that.

The other claim on the wrapper is what this report is about. The fourteen-syllable ingredient name. The "from coconut" suggestion. The implicit chemistry comparison to Sodium Lauryl Sulfate — SLS — that no published criterion on the wrapper actually performs.

The wrapper is governed by Article 19 of EU Cosmetics Regulation (EC) 1223/2009 — retained in UK law after 2020.² Article 19(1)(g) requires the manufacturer to disclose the cosmetic ingredients, in descending order of concentration above 1%, using the International Nomenclature of Cosmetic Ingredients: the INCI name. The names come from a list. The list was first compiled in 1973 by the Cosmetic, Toiletry and Fragrance Association — an American trade association now renamed the Personal Care Products Council. It was a vocabulary tool: a way for regulators and chemists across jurisdictions to be sure that one company's lauryl sulfate and another's sodium dodecyl sulfate meant the same molecule.³ The 1976 EEC Cosmetics Directive 76/768/EEC adopted it.⁴ EC 1223/2009 inherited it from the 1976 Directive. The list standardises the name. The chemistry, the pathway, the impurity profile, the pH, and the equivalence the consumer's eye reads into the wrapper sit outside it.

So the consumer outsources the chemistry adjudication to four parallel adjudicators: the brand, the certifier, the retailer, and the third-party reviewer. They do not agree.

• Whole Foods Market's Premium Body Care Quality Standards exclude Sodium Lauryl Sulfate and Sodium Laureth Sulfate by name; they retain Sodium Cocoyl Isethionate.⁵
• Sephora Clean excludes Sodium Lauryl Sulfate; it permits Sodium Cocoyl Isethionate.
• Target Clean excludes Sodium Laureth Sulfate but does not categorically exclude Sodium Lauryl Sulfate.⁶
• Boots Botanics excludes Sodium Lauryl Sulfate without citing a regulator.⁷
• Cosmébio, the French organic-cosmetic certifier, prohibits Sodium Cocoyl Isethionate on the grounds that its synthesis pathway runs through ethoxylation — the same pathway Sodium Laureth Sulfate is restricted for, and which Sodium Lauryl Sulfate does not have.⁸
• COSMOS-Organic permits Sodium Coco-Sulfate. Cosmébio publicly criticises the COSMOS position on the same surfactant.

Five adjudicators, three different exclusion lists, and at least one direct disagreement between them on which molecule to forbid. The consumer offloading the decision to a label is offloading it to a system whose labels do not agree on the answer. Article 19 is the regulation that would have to carry the parameters by which they could agree, and Article 19 standardises names. The chemistry it does not contain is what this report walks through.

II. Four Molecules, One Shopping Aesthetic

Pick up four shampoo bars from four different shelves in a UK or EU retailer. Read the first ingredient on each.

Each is sold under the same overarching shopping aesthetic — natural, plastic-free, often coconut-derived. Each contains a different molecular class. The most-recognised name in the segment, Lush, lists Sodium Lauryl Sulfate first on the bar that anchors its shampoo-bar range — the molecule the segment as a whole is positioned against.⁹

These are four different chemistries, not four spellings of one.

Sodium Lauryl Sulfate (SLS). A single twelve-carbon hydrocarbon chain ending in a sulfate ester anion (-O-SO₃⁻ Na⁺), manufactured by reacting sulfur trioxide gas with lauryl alcohol in a continuous SO₃ sulfation process. By the published purity threshold, no less than 80% of the lauryl chains in commercial SLS are exactly twelve carbons long; the rest are minor C10, C14, or C16 contributions. SLS is the dermatology reference irritant — used at 0.25% in aqueous solution as the standard control patch on the international baseline allergy test series, precisely because its irritation profile is well-characterised and predictable.¹⁰ Its synthesis route does not involve ethylene oxide; it does not carry 1,4-dioxane as a manufacturing residual. No major regulator — the US EPA, the European Scientific Committee on Consumer Safety (SCCS), the FDA, the International Agency for Research on Cancer, the National Toxicology Program, the US Cosmetic Ingredient Review (CIR) panel, or California's Proposition 65 list — has classified SLS as a carcinogen.¹¹

Sodium Coco-Sulfate (SCS). Not a single molecule. The same alkyl-sulfate chemistry as SLS, manufactured by the same SO₃ sulfation process on a different starting material — broad-cut coconut fatty alcohol rather than the single-cut lauryl alcohol that SLS uses. The product is a mixture of alkyl sulfates of varying chain length: typically 51–57% C12 (lauryl) plus 41–47% C14 (myristyl) plus minor C8, C10, C16, and C18 fractions, sulfated together as one paste.¹² By mass, roughly half of the molecule labelled "Sodium Coco-Sulfate" is the molecule labelled "Sodium Lauryl Sulfate". The other half is the same chemistry on a slightly longer chain. The ECHA-published REACH registration for SCS sits under EC number 306-683-4. The CIR safety review covers the alkyl-sulfate group as a class, with a verdict of safe in rinse-off. Pathway: identical to SLS. Composition: broader. Name: different.

Sodium Cocoyl Isethionate (SCI). A different chemistry class. The hydrophobic chain — a coconut-derived fatty acid — is attached not through a sulfate ester but through a carboxylic ester to a small molecule called sodium isethionate (HOCH₂CH₂SO₃Na), which carries the sulfonate head group. The chemistry, on its parent molecule, is mild. The CIR safety review of isethionate salts itemises permitted impurity ceilings: arsenic ≤3 ppm, lead ≤20 ppm, sodium chloride ≤0.8%, free fatty matter ≤10%, free fatty acid ≤18%, sodium soap ≤3%.¹³ The table is precise about what is in it. It also does not itemise 1,4-dioxane. Sodium isethionate — the upstream synthon — is itself manufactured by reacting ethylene oxide with sodium bisulfite. Ethylene oxide is the same upstream reactant that produces 1,4-dioxane as a manufacturing residual in Sodium Laureth Sulfate (SLES) — the ethoxylated cousin of SLS that the wider clean-cosmetic conversation has restricted on dioxane grounds. The CIR ceiling is silent on the residual. The Cosmetics Regulation does not require the pathway to be disclosed on the wrapper.

The US Food and Drug Administration explicitly identifies 1,4-dioxane as a manufacturing by-product of ethoxylation in cosmetic ingredients.¹⁴ New York State, since 31 December 2023, sets a 1 ppm cap on 1,4-dioxane in personal-care products sold in the state — a state law operating on the manufacturing pathway the federal regulator publicly identifies.¹⁵ Cosmébio prohibits SCI in its organic standard on this exact basis.⁸

The supplier-side mitigation is real. Innospec — one of the major commercial SCI manufacturers — publishes a brochure titled Leader in 1,4-Dioxane Free Surfactants; the cosmetic-ingredient supplier MakingCosmetics publishes a finished-grade SCI specification stating <1 ppm 1,4-dioxane and <1 ppm ethylene oxide.¹⁶ The published spec, where it is published, sits below the New York limit. None of this — pathway, grade, spec, vacuum-stripping protocol — is on the wrapper.

Decyl Glucoside. A genuinely distinct chemistry. Non-ionic — no charge on the head group. The hydrophobic chain is a ten-carbon decyl alcohol; the hydrophilic head is a glucose ring attached through an acetal bond (a sugar–oxygen–carbon linkage), produced by reacting decyl alcohol with glucose under acid catalysis. The pathway is called Fischer glycosylation; it does not involve ethylene oxide, and it does not produce sulfate esters. Decyl glucoside is the only one of the four molecules above to receive simultaneous CIR rinse-off clearance and full COSMOS-Organic certification.¹⁷

Cocamidopropyl Betaine (CAPB) — the constant co-surfactant. Below the four primary names, on most shampoo-bar wrappers, sits a fifth: cocamidopropyl betaine. It is amphoteric — carrying both positive and negative charges on the same head group — and is co-formulated with all four primary surfactants in commercial bars to boost foam and balance the irritation profile. CAPB at the parent-molecule level is mild. The American Contact Dermatitis Society named it Allergen of the Year 2004 anyway. The reason is not the parent molecule. It is the synthesis impurities. CAPB is made from coconut fatty acid plus 3-dimethylaminopropylamine (DMAPA) through an intermediate called cocamidopropyl dimethylamine — the amidoamine. Each step leaves a residual. The CIR-permitted impurity range in finished CAPB is DMAPA 0.00025% to 0.01% and amidoamine 0.05% to 5.0%.¹⁸ The labelled allergen — cocamidopropyl betaine — is not the molecule that causes the allergy. The actual sensitiser is an unlabelled impurity living inside the labelled name.

So: five molecules, five chemistry classes, one shopping aesthetic. The ingredient list discloses what to call them. What links any of them to the chemistry the consumer thinks they are buying is somewhere else.

III. The Vocabulary the Wrapper Inherits

The name on the wrapper is not chosen by the manufacturer. It is chosen by a list.

The first edition of the CTFA Cosmetic Ingredient Dictionary was published in 1973 by the Cosmetic, Toiletry and Fragrance Association — an American industry trade body, founded as the Manufacturing Perfumers' Association in 1894, renamed CTFA in 1971, and renamed the Personal Care Products Council in 2007.³ The 1973 Dictionary was a vocabulary tool. Its purpose, on the trade association's own description, was to ensure that across companies, formulations, and jurisdictions, the same molecule would be referred to by the same name. The names it assigned were the CTFA Adopted Names. Soon after publication, the US Food and Drug Administration began citing the Dictionary as the primary source of ingredient names for cosmetic labelling under the Federal Food, Drug, and Cosmetic Act.

The 1976 EEC Cosmetics Directive — Council Directive 76/768/EEC — required the cosmetic ingredient list to appear on the consumer-facing label, in descending concentration order, by a standardised name, and adopted the CTFA list as its inherited vocabulary. Subsequent Commission Decisions in 1996 and 2006 formalised the inventory.⁴ In 1993, the CTFA renamed the system International Nomenclature of Cosmetic Ingredients — INCI — to acknowledge that the EU, Japan, and others were now using it. EC 1223/2009, which superseded 76/768/EEC and is the current EU governing regulation for cosmetics, inherited the INCI vocabulary by direct reference at Article 19 and at Article 33 — the latter establishing the operative INCI glossary, published as Commission Decision (EU) 2019/701.²

Nobody in the 1973–2009 chain stopped to ask whether the list was designed to be read by a consumer as a chemistry comparator. It was a regulator's vocabulary, repurposed quietly into consumer-information infrastructure as the EU Directive landed on it. The architecture did not change. The names that ended up on consumer-facing wrappers in 2009, 2017, and 2026 are the names a US trade association compiled in 1973 for an audience that was not the consumer at all.

What Article 19 requires is what it requires. What it does not require is where the Coco Question lives.

The synthesis pathway by which the named molecule was made is not on the wrapper. The wrapper says Sodium Cocoyl Isethionate. Whether the upstream sodium isethionate was made via ethylene oxide is not part of the disclosure.

Chain-length distribution within an INCI name that covers a mixture is not on the wrapper. The European Chemicals Agency calls these UVCB substances — Unknown or Variable composition, Complex reaction products or Biological materials. Sodium Coco-Sulfate is one. The name covers a Gaussian distribution centred at C12; the wrapper collapses the distribution to a name.

Residual-impurity profile beyond the specific Annex II–III restricted-substance list is not on the wrapper. 1,4-dioxane is an Annex II prohibited substance, but the pathway-residual-disclosure regime sits at the manufacturer-control layer, not the labelling layer. The consumer cannot see whether the SCI in the bar carries 0.1 ppm or 0.9 ppm of dioxane.

Formulation pH at use is not on the wrapper. A bar at pH 9–10 is alkaline and demonstrably damaging to hair cuticle on repeated wash. A bar at pH 5–7 matches scalp pH (5.5) and the surface of the hair shaft.

And format chemistry class is not on the wrapper. A saponified soap bar — INCI: Sodium Cocoate, Sodium Olivate, Sodium Tallowate, Sodium Castorate — is alkaline by chemistry. A syndet (synthetic-detergent) bar built on SCS, SCI, decyl glucoside, or SLS is pH-adjustable. Both sit on the same shelf, in the same category, under the same marketing aesthetic. The INCI names reveal the difference to a chemist. The marketing does not flag it to a consumer.

Article 20 of EC 1223/2009, with implementing detail in Commission Regulation (EU) No 655/2013, prohibits misleading claims; in the UK, the Advertising Standards Authority and Competition and Markets Authority can rule against an advertised cosmetic claim found to be deceptive. That is a real enforcement pathway, and it has been used — but Article 20 governs the front of the wrapper, the headline claim. Article 19 governs the back. The chemistry comparison the headline claim implies is asked of the back of the wrapper. The back of the wrapper carries names.

The food regulator faced a structurally similar problem in 1886, with butter and margarine. Margarine, invented in 1869, was sold in the US as butterine or bogus butter with no labelling distinction. The Federal Oleomargarine Act of 1886, signed by President Cleveland, defined margarine as made in imitation or semblance of butter — the first congressional standard of identity for a food product, with labelling, packaging, and tax requirements that compelled the back of the wrapper to do what the front of the wrapper was suggesting.¹⁹ Modern US food law is built on that doctrine: 21 CFR Part 166 sets the standard of identity for margarine; 21 CFR 101.13 governs light and lite claims; the Filled Milk Act of 1923 governs imitation milk products. There is no parallel rule in 21 CFR Part 701, the cosmetic labelling regulation, and there is no parallel in EC 1223/2009 — for natural, naturally derived, from coconut, or from a plant you recognise. The food sector worked this out, partially, in 1886. The cosmetic sector, in 2026, has not.

IV. What Penetrates Is the Monomer

A lay account of how a shampoo works imagines the named surfactant — millions of identical molecules — sliding en masse across the surface of the scalp like a tide of bricks. The image is wrong by an order of magnitude in two directions.

A surfactant in solution does not sit as a neat layer of single molecules. Above a concentration called the critical micelle concentration, or CMC, surfactant molecules aggregate into ball-shaped clusters — micelles. Below the CMC, the same molecules sit in solution as individual units called monomers. The CMC is a property of the molecule and the solvent. For SLS in water at 25 °C, it is around 8 millimoles per litre, roughly 0.23% by weight.²⁰ For longer-chain alkyl sulfates — the C14 and C16 fractions of SCS — the CMC is lower. Lather-strength surfactant solutions in a shampoo bar wash run well above the CMC of every chain present. Most of the surfactant in lather sits in micelles. The lather foam the consumer sees is mostly micelles holding air.

The penetrating species is not the micelle. It is the monomer — the fraction of surfactant molecules in true solution below the CMC. This is what crosses the stratum corneum (the outer skin barrier) and binds to keratin and skin lipids. Bondi and colleagues (2015), in a peer-reviewed safety assessment in Environmental Health Insights, lay out the mechanism: SLS at use concentration is largely sequestered into micelles; the monomer fraction is what the skin sees, and that fraction is small — and time-limited, because rinse-off contact is minutes, not hours.¹⁰ This is the chemistry on which the regulators have rested SLS's safe in rinse-off designation since 1983 and reaffirmed it across CIR, ECHA, and FDA review.

The monomer-below-CMC mechanism is what makes the chain-distribution gradient inside Sodium Coco-Sulfate matter. Bujak, Niziol-Lukaszewska, and Wasilewski, in a 2019 peer-reviewed primary in Tenside Surfactants Detergents, measured SCS against SLS on two skin-interaction parameters and found an inversion.¹² On the zein test — a standardised cosmetic-irritation assay using corn protein as a stand-in for keratin denaturation — SCS produced approximately 15% less zein dissolution than pure SLS. SCS was milder on that parameter. On lipid elution from stratum corneum — the parameter that measures how aggressively a surfactant strips skin barrier lipids — SCS was worse than SLS. Not the same. Worse.

The two parameters move in opposite directions because of the chain-distribution mixture. The C12 fraction of SCS (the SLS-like fraction) is partly sequestered into mixed micelles by the C14 and C16 fractions, which sit at lower CMC and pull C12 in alongside them. Less C12 monomer is free in solution; less C12 monomer means less keratin denaturation, so the zein number falls. But the longer-chain C14 and C16 monomers, when they do penetrate, are more lipophilic — they partition more aggressively into the bilayer of skin lipids and elute more lipid per gram of monomer that crosses. Net effect on the keratin parameter: favourable. Net effect on the lipid-barrier parameter: not. The "SCS is gentler than SLS" claim is a partial truth that the report's research could only locate one paper to test. Bujak 2019 is what the finding rests on. It is one paper, in a respectable surfactant trade journal. Section X names the specific replication study that would update or qualify it.

Article 19 requires the INCI name. Chain-length distribution underneath the name, the CMC of the mixture, the zein value, and the lipid-elution measurement are not in scope. The named molecule on the bar is Sodium Coco-Sulfate. The molecule that interacts with the scalp is a Gaussian mixture whose dermal effects move in opposite directions on two separate parameters, one of which the name implies and one of which the name conceals.

The mechanism applies, with adjustments, to the other named molecules. SCI is largely insoluble in cold water; it requires heat and stearic-acid binders to formulate into a bar; outside its narrow stability window of pH 6 to 8, the ester linkage hydrolyses. A shampoo-bar lather is hot water plus mechanical agitation plus uncontrolled local pH. Some fraction of the SCI in the lather plume is hydrolysing during the wash, releasing free coconut fatty acid plus free isethionate at the head — and the hydrolysis product is the molecule whose synthesis pathway carries the dioxane risk. The molecule that enters the drain is not always the molecule labelled on the wrapper.

Decyl glucoside, by contrast, hydrolyses under aqueous conditions in the skin to give back fatty alcohol and glucose — both natural metabolic substrates. Its monomer interaction with the skin barrier, on the published assays, is mild. This is the mechanism behind its CIR clearance and its COSMOS-Organic permission.

Cocamidopropyl betaine penetrates as a betaine zwitterion. The DMAPA and amidoamine impurities, however, are smaller, more reactive, and capable of haptenating skin proteins — covalently attaching to a body protein and triggering an immune response. This is why the patch-test population that reacts to a CAPB-containing product is mostly reacting to the impurity, not the parent molecule. A 2023 Japanese clinical study of intractable scalp dermatitis found that 50% of patients positive on CAPB patch testing also tested positive for DMAPA, amidoamine, or both.²¹ The cohort is selected; the rate does not generalise to the general population. What it does establish, in the population that reacts, is that the labelled allergen is not the molecule causing the reaction. The wrapper carries the name. The impurity inside the name does not have to be disclosed.

V. The 1,4-Dioxane Inversion

The SLS-vs-natural reformulation conversation has rested, on the consumer-facing side, on two background ideas. First: SLS causes cancer. Second: the natural alternatives move you off the synthetic-chemical pathway. Both have a specific historical lineage. Neither survives close inspection.

The cancer claim originated in 1998, in an anonymous chain email circulated in the early commercial-internet era and tracked by Snopes and the American Cancer Society to the multi-level natural-products marketing segment of the time.²² No major regulator — IARC, the National Toxicology Program, the US EPA, the European Scientific Committee on Consumer Safety, the FDA, the CIR Expert Panel, or California Proposition 65 — has ever classified SLS as a carcinogen. California Proposition 65, the broadest precautionary cancer-warning regime in the US, does not list it.¹¹ Bondi and colleagues in 2015 published the peer-reviewed defence in Environmental Health Insights: SLS's irritation profile is real, well-characterised, and the basis on which it is used as the reference irritant in dermatology baseline patch testing — but irritation at the use levels of consumer cosmetics is well-controlled by formulation, and the cancer claim is folk-toxicology.¹⁰ The 1998 chain email did not move the regulators. It moved the marketing.

The second idea — natural alternative = off the synthetic pathway — fails specifically on Sodium Cocoyl Isethionate.

The synthesis chemistry: coconut fatty acid is condensed with sodium isethionate (HOCH₂CH₂SO₃Na) at high temperature to form SCI plus water. Sodium isethionate, the upstream synthon, is itself made by reacting ethylene oxide with sodium bisulfite. Ethylene oxide is the manufacturing reactant the wider clean-cosmetic conversation has spent two decades restricting. It is the same family of chemistry that produces 1,4-dioxane as a residual in Sodium Laureth Sulfate (SLES), the ethoxylated cousin of SLS. The US FDA explicitly identifies 1,4-dioxane as a manufacturing by-product in cosmetic ethoxylated ingredients.¹⁴ New York State, since 31 December 2023, sets a 1 ppm cap on 1,4-dioxane in personal-care products, primarily targeting SLES.¹⁵

SLS — the molecule the segment is positioned against — is not ethoxylated. It does not generate 1,4-dioxane in its synthesis. SCI — the "natural upgrade" — is downstream of ethylene oxide. The reformulation pathway from SLES to SCI is, on the synthesis-chemistry axis, an upgrade onto the same ethylene-oxide manufacturing family the SLES restriction was designed to address. The pathway from SLS to SCI is a sideways move onto a synthesis route the original SLS does not have.

This is not a claim about manufacturer failure. The published manufacturer-side controls, where they are disclosed, are real. Innospec — one of the major commercial SCI suppliers — publishes a brochure titled Leader in 1,4-Dioxane Free Surfactants. The cosmetic-ingredient supplier MakingCosmetics publishes a finished SCI specification stating <1 ppm 1,4-dioxane and <1 ppm ethylene oxide.¹⁶ Vacuum stripping at the production stage is the standard mitigation; the published spec sits below the New York legal cap. That is a real control.

It is not a labelling control. The CIR's 2013 isethionate impurity ceiling itemises arsenic, lead, sodium chloride, free fatty matter, free fatty acid, and sodium soap. It does not itemise 1,4-dioxane.¹³ The Cosmetics Regulation does not require disclosure of the synthesis pathway. The wrapper does not say whether the SCI in the bar was sourced from the Innospec grade with the published <1 ppm specification or from a less-controlled grade. If every commercial SCI grade in the world tomorrow met the Innospec specification, the labelling-regime gap would still be there. The consumer paying a premium to switch from SLS specifically to avoid an ethoxylated synthesis pathway is, on the SCI option, paying a premium to switch onto one. The wrapper does not say so.

Cosmébio has named this in its own published critique of SCI as part of what it calls greenwashing — the term appears in the certifier's own publication title, not introduced here.⁸ The COSMOS-Organic standard, parallel to Cosmébio's, prohibits SCI on the same synthesis-pathway grounds. A different certifier, COSMOS-Natural, permits SCI. The four delegated adjudicators do not agree.

The pattern has a name in the green-chemistry policy literature: regrettable substitution. A vilified compound is removed under public pressure; a structural analogue is substituted; the substitute turns out to share the relevant property — or to bring a different one — that the substitution was sold against. The bisphenol-A to bisphenol-S substitution is the canonical case: BPS is comparable in endocrine activity to BPA, sometimes more persistent, and the BPA-free claim purchased on the front of the wrapper does not always match the chemistry on the back.²³ The PFOA-to-PFOS-to-GenX substitution in fluorochemistry is the parallel in the forever chemicals domain. The SLS-to-SCI move follows the architecture: vilification of the original (chain email, not regulator), substitution to a structural analogue (different molecule, same upstream reactant family), labelling that names the new molecule but does not name the pathway it travels.

VI. One Shelf, Two Chemistries

The orthogonal axis — the one the surfactant-identity question does not capture — is the format chemistry of the bar itself.

A saponified soap bar is made by reacting fats and oils with sodium hydroxide at elevated temperature; the fatty-acid components convert into their sodium salts. The INCI panel reads as plant-derived: Sodium Cocoate, Sodium Olivate, Sodium Tallowate, Sodium Castorate, Sodium Palmate. There is no synthetic surfactant. There is also no pH adjustment. A saponified soap bar at use is alkaline — typically pH 9 to 10. Scalp pH is approximately 5.5; the surface of the hair shaft is approximately 3.67. The pH gradient between an alkaline cleanser and an acidic substrate opens the cuticle scales of the hair shaft and progressively damages cuticle integrity on repeat wash. Gavazzoni Dias and colleagues, in a 2014 peer-reviewed primary in the International Journal of Trichology, document the mechanism: the alkaline pH of soap-based cleansers correlates with measurable increases in fibre friction, cuticle breakdown, and surface-charge effects on the hair shaft.²⁴

A syndet bar — synthetic-detergent — built on SCS, SCI, decyl glucoside, or SLS as primary surfactant is pH-adjustable. The formulator can set the lather pH to 5.5 to match scalp; the pH gradient against the cuticle is small; the cuticle damage mechanism does not run.

Both bars sit on the same shelf, wrapped in the same kraft paper, sold under the same natural, plastic-free, zero-waste shopping aesthetic. The INCI panel reveals the difference to a chemist by the presence (saponified soap) or absence (syndet) of Sodium Cocoate-class names. The marketing copy does not. A consumer who switches from a bottled SLES shampoo (engineered for pH 5.5 with chelators that bind hard-water calcium and magnesium ions) to a sodium-cocoate soap bar (alkaline, pH 9–10, no chelator) inherits a hair-cuticle mechanism they did not previously have to manage. The labelled surfactant changed categorically — from a syndet to a soap — but the marketing category they shopped in did not. There is no category line on the wrapper that says syndet bar or saponified soap bar. The INCI list reveals it. The shopping aesthetic obscures it.

Add the hard-water dimension. In hard-water postal codes — most of the UK and large parts of EU and US — divalent calcium and magnesium ions in tap water cross-link with anionic surfactant head groups during the wash to form lime soap: insoluble calcium-fatty-acid salts that deposit on hair and scalp. For saponified soap bars, this is the dominant failure mode. For SCS, SCI, and SLS bars, the same chemistry happens at lower magnitude — a fraction of the labelled surfactant is consumed in situ as calcium soap before it ever reaches the hair. Decyl glucoside, being non-ionic, does not form calcium soap. Bottled shampoos in hard-water markets routinely contain a chelator — EDTA, GLDA, sodium phytate, or tetrasodium glutamate diacetate — to bind the calcium and magnesium and prevent the lime-soap reaction. Many "natural" shampoo bars omit chelators because chelators are perceived as synthetic. The wrapper does not say which.

What is on the scalp during a wash, then, is not always the molecule on the wrapper. It is the molecule on the wrapper, plus the calcium soap formed in situ in hard water, plus any free fatty acid released by partial SCI hydrolysis under the lather conditions, plus any DMAPA or amidoamine impurity in the CAPB co-surfactant, plus the free monomer below the CMC of whichever chain length distribution dominates. The ingredient list carries one name per molecule. The scalp interacts with several molecules per name. The labelling regime does not require the second list.

VII. The End-of-Life Claim

The third axis — what enters the drain after the wash — is the one the segment's environmental marketing rests on most heavily. Biodegradable. Plastic-free. Eco-friendly.

The biodegradability claim is true on a single test. Under OECD 301 Ready Biodegradability protocols — the standard battery of laboratory tests in which a microbial inoculum is given the surfactant as a sole carbon source under aerobic conditions and the rate of mineralisation to carbon dioxide is measured — every named surfactant in this report passes. SLS reaches approximately 95% biodegradation at 28 days under OECD 301B. SCS, as a class, performs comparably. SCI's primary biodegradation reaches 99.6% at 14 days under OECD 301D, per the EPA HPV test plan for the substance class.²⁵ Decyl glucoside, by class data, reaches >98% elimination at 7 days under OECD 301D. Aerobic-stage biodegradation, on the OECD 301 endpoint, is true for all of them — including the SLS the segment is positioned against.

The claim narrows considerably under closer reading.

Municipal wastewater treatment, in the EU and most of the developed world, is a two-stage process. Aerobic treatment at the treatment plant is the first stage. Anaerobic digestion of the sludge — where the organic load is converted, in the absence of oxygen, to biogas (methane and carbon dioxide) for energy recovery — is the second. OECD 301 measures the first stage only. At the second stage, the picture for alkyl polyglucosides — the chemistry class of decyl glucoside, the segment's "greenest" surfactant — inverts.

Ríos and colleagues at the University of Granada, in a 2016 peer-reviewed primary in Environmental Science and Pollution Research, measured anaerobic biodegradation kinetics for alkyl polyglucosides under methanogenic digester conditions.²⁶ At a concentration of 100 mg of carbon per litre, alkyl polyglucosides inhibit biogas production. Linear APGs reached >70% mineralisation under methanogenic conditions; branched APGs resisted anaerobic degradation entirely. The shorter the alkyl chain, the faster the anaerobic biodegradation; the longer chains (decyl, lauryl) are slower. Aerobic-stage >98% in 7 days is real. The anaerobic-stage behaviour is the inversion: the surfactant marketed for its environmental fate inhibits the second stage of the process by which municipal wastewater treatment recovers energy from organic load.

The 100 mg/L laboratory concentration is well above typical municipal-wastewater APG levels at the digester inlet — by approximately three orders of magnitude. This is not a claim that household shampoo-bar use is poisoning municipal digesters. The finding is that the biodegradable marketing claim, asserted on a single aerobic test, does not survive contact with the second stage of the system the wastewater is actually moving through. Biodegradable is a single-test claim. The test does not say under all conditions. The wrapper does.

The other named surfactants raise their own end-of-life questions. SCI, partly hydrolysing under shampoo-bar lather conditions, releases free isethionate into the drain — a relatively small molecule whose own environmental half-life is shorter than SCI's, but which carries the synthesis pathway the chemistry started on. SCS, on the OECD 301B aerobic endpoint, behaves like the alkyl-sulfate class to which it belongs. SLS, on the same test, reaches 95%. The aquatic-toxicity ladder is mixed: SLS marine LC50 is 4.1 mg/L in published REACH data; decyl glucoside fresh-water LC50 is in the 96–115 mg/L range. The directional ordering between named natural surfactant and named synthetic surfactant depends, again, heavily on the test species, the matrix, and the test stage. A clean directional finding on aquatic fate is not available on the OECD endpoint as published.

The inversion is structural: the surfactant marketed for its environmental superiority — decyl glucoside — is the one that inhibits the anaerobic stage of the system the wastewater enters. The single-test aerobic claim and the multi-stage real-system behaviour are not the same chemistry.

VIII. The Provocation

The discipline of this report at Stage 5 was an explicit Provocation: if even one of the four named substitutes — most plausibly decyl glucoside in an APG-led syndet bar without CAPB — turns out to be a meaningfully different molecule with materially better behaviour on every parameter the report tests (irritation, lipid elution, impurity load, contact-allergy rate, aerobic biodegradation), then the labelling regime is doing its job and the report's structural finding is a literacy gap, not a regime gap.

The Provocation is the right test. The evidence does not survive it.

Decyl glucoside, on the parameters the Stage 8 packet supports, is genuinely chemically distinct from the alkyl sulfates: non-ionic, structurally different head group, no ethylene-oxide pathway, full COSMOS-Organic clearance. On irritation it is mild; the CIR's 2013 final safety assessment of decyl glucoside and other alkyl glucosides documents the dermatology data and concludes safe when formulated to be non-irritating.¹⁷ On manufacturing-impurity load (the 1,4-dioxane axis), it favours: APGs are not ethoxylated and do not generate dioxane.

Two parameters fail the meaningfully better on every parameter specification.

Contact-allergy rate. The American Contact Dermatitis Society named alkyl glucosides — decyl glucoside specifically — as the Allergen of the Year 2017.²⁷ The North American Contact Dermatitis Group standard panel has tested decyl glucoside since 2009. Soriano and colleagues in 2021 reported a UK and Ireland sensitisation rate of approximately 1.3% (14 of 1,099 patch-tested patients).²⁸ Loranger and colleagues at McGill in 2017 documented the trend rising from 1.3% to 2.2% across their cohort.²⁹ The rate does not exceed the 5% threshold the report's own Section X commits to as the inversion point. It is not above 5%. It is also not zero, not declining, and not meaningfully better than the alkyl sulfates SCS or SLS, whose patch-test rates in the same registries are comparable or lower. The contact-allergy axis is at best neutral. The ACDS Allergen of the Year designation is, on its own evidence, a recognition that the molecule is on a rising sensitisation trajectory in the dermatology registries.

Anaerobic biodegradation. Ríos et al. 2016, the peer-reviewed primary cited above, documents that APGs at 100 mg of carbon per litre inhibit biogas production under methanogenic digester conditions, with branched APGs resistant entirely.²⁶ The aerobic-stage biodegradation claim survives. The anaerobic-stage claim does not.

The Provocation as specified — meaningfully better on every parameter — is defeated. Decyl glucoside is a real chemistry difference from SLS on most parameters and a partial improvement on most environmental endpoints. It is not a clean win on the sensitisation axis or the anaerobic-stage axis. The substitution is partially meaningful. It is not the real and meaningful on every parameter finding the Provocation demanded as the literacy-gap reframing.

The chemistry does not vindicate any one of the four substitutes as the clean substitution the segment markets the category as. The substitution is heterogeneous: real for decyl glucoside on some parameters; partial for SCS on the zein-vs-lipid inversion; pathway-vulnerable for SCI on the dioxane axis; impurity-driven for CAPB on the labelled-vs-actual sensitiser. None of the four is the meaningfully better on every parameter substitute the literacy-gap framing required.

The labelling-regime gap survives. The report's spine — Article 19 standardises the name, not the chemistry — is the regime gap, not a literacy gap. And the four delegated adjudicators do not agree on which substitute to forbid. If the consumer cannot tell from the wrapper, and the four parallel outsourcing channels — brand, certifier, retailer, third-party reviewer — also cannot agree, then the report's structural finding is not the consumer needs to read harder. It is that the system the consumer is reasonably outsourcing to does not contain the parameters by which the four channels could align. Article 19 is the regulation that would have to carry those parameters. It does not.

IX. The Levers

The chemistry is done. The reader has the right to know what to do with it.

There are two tiers — what does not require buying anything, and what to choose differently when something needs replacing.

Tier 1 — what does not require buying anything.

Read the INCI for format chemistry class first. If the first ingredient is Sodium Cocoate, Sodium Olivate, Sodium Tallowate, or Sodium Castorate — and there is no SCS, SCI, decyl glucoside, or SLS as a primary surfactant — the bar is a saponified soap bar at pH 9 to 10. If the first ingredient is one of the four synthetic surfactants, the bar is a syndet bar, pH-adjustable to scalp pH. The two are not interchangeable for hair, regardless of the marketing category they share. A consumer who has experienced cuticle damage, frizz, or the great unwash transition after switching to a "natural" shampoo bar should consider whether the bar they switched to is saponified soap rather than syndet. The INCI panel reveals this. The marketing rarely does.

Check for the secondary surfactant. If Cocamidopropyl Betaine appears on the INCI panel — it appears on most of them — the unlabelled DMAPA and amidoamine impurity question is on the table even when the primary surfactant is "natural." A consumer with chronic scalp, facial, or eyelid dermatitis from a "natural" shampoo bar may be reacting to the CAPB impurity rather than to the primary surfactant they switched away from. CAPB-free formulations exist as a published consumer category for this reason. The diagnostic move is to read the INCI for Cocamidopropyl Betaine before switching primary surfactants.

Check for chelator presence. In hard-water postal codes — most of the UK; large parts of the EU and US — EDTA, GLDA, Sodium Phytate, or Tetrasodium Glutamate Diacetate on the INCI panel signals a formulation engineered for hard-water performance. Their absence in a "natural" bar in a hard-water postal code sets up the calcium-soap deposition mechanism the consumer did not previously have to manage with their bottled shampoo.

Treat "SLS-free" as a procurement-gate-compliance claim, not a dermatology claim. The four retail "clean" gates — Whole Foods Premium Body Care, Sephora Clean, Target Clean, Boots Botanics — operate published exclusion lists. None cite a regulator finding that SLS is unsafe in rinse-off, because no major regulator has issued one. The exclusion is a private retail policy. The label is the policy's output, not the chemistry's verdict. SLS-free tells the consumer a retailer's preference. It does not tell the consumer what the chemistry under the substitute is doing.

Tier 2 — what to choose differently when something needs replacing.

This report does not name brands as recommendations. The properties that close the gap are observable from the label.

The bar that maximally resolves the four hidden axes is a syndet bar at pH 5–7, with an alkyl-polyglucoside-led primary surfactant, without cocamidopropyl betaine, with a chelator if used in a hard-water postal code, and ideally with a COSMOS-Organic certification mark (which excludes SCI on the ethoxylation pathway grounds Cosmébio names). The chemistry is structurally distinct from SLS; the synthesis pathway is not ethoxylated; the labelled-vs-actual sensitiser problem is bypassed at the secondary surfactant; the hard-water deposition mechanism is mitigated. On the contact-allergy axis the rate is not zero — Soriano 2021 documents approximately 1.3% sensitisation on patch testing — and a consumer with a history of patch-test-confirmed reactions to alkyl glucosides should consider this. The decyl-glucoside-led syndet is the closest the available chemistry comes to a clean substitution from SLS on the parameters this report tests. It is not the meaningfully better on every parameter substitute the marketing category implies. It is meaningfully better on most.

These are properties, not products. A reader can read for them on any bar wrapper. The labelling regime requires the names; the names allow the properties to be inferred — partially, with the discipline above.

X. What Would Change This Analysis

Seven specific findings would update or invert this report. Each is named with the evidence type that would establish it.

1. A 2024 to 2026 final opinion from the EU Scientific Committee on Consumer Safety, the German Federal Institute for Risk Assessment (BfR), the French ANSES, or the US Cosmetic Ingredient Review Expert Panel that establishes a quantitative consumer-exposure threshold for SCI residual 1,4-dioxane, with a measurement method and a labelling-regime disclosure requirement. This would convert the dioxane inversion from a labelling-regime gap into a labelling-regime fix in progress. It would not eliminate the broader regime gap, but it would remove the SCI-specific load-bearing chemistry finding. The 2025 SCCS final opinions roster does not contain a standalone surfactant opinion. The CIR 2013 isethionate ceiling does not itemise dioxane. The finding does not currently exist.

2. A direct head-to-head clinical Repeat Insult Patch Test (RIPT) of SLS, SCS, SCI, and decyl glucoside in finished shampoo-bar matrix, at use concentration, at use pH, and at use contact time, showing decyl glucoside materially superior on irritation, lipid elution, transepidermal water loss, and contact-allergy rate. The published comparative work, including Bujak 2019, is mostly aqueous-solution single-surfactant testing. A finished-matrix head-to-head would convert the substitution-is-heterogeneous finding into a substitution-is-real-for-decyl-glucoside finding. The Tier 2 lever above would harden. The labelling-regime gap would remain — the consumer still cannot identify which formulation is APG-led without reading the INCI panel for primary-surfactant position — but the chemistry argument would shift.

3. Publicly available datasheets from the major commercial SCI manufacturers — Stepan, Innospec, Clariant, Galaxy Surfactants, Croda — for finished SCI grades, showing 1,4-dioxane residuals below 1 ppm at the production specification, validated by a third-party test laboratory and not solely by manufacturer self-declaration. Innospec and MakingCosmetics already publish <1 ppm specifications. A broader, third-party-validated industry-wide spec would shift the framing: manufacturer-side mitigation is real; the labelling regime still does not disclose the pathway or the controls; the consumer relies on supplier reputation invisibly.

4. A 2024 to 2026 NACDG, IVDK, ESSCA, or BfR contact-dermatitis registry showing decyl glucoside sensitisation rate above 5% in shampoo-bar-using populations. This would invert the Tier 2 lever entirely. The decyl-glucoside-led syndet recommendation would not hold. The available data — Soriano 2021 (UK/Ireland, 1.3%), Loranger 2017 (McGill, 1.3% to 2.2%) — is below the threshold. The ACDS Allergen of the Year 2017 designation is qualitative; the underlying registry data is quantitative.

5. A retailer — Whole Foods, Sephora, Target, Boots — publishing the regulatory or peer-reviewed evidence on which their SLS exclusion list is based, showing the basis is dermatology-evidence-driven rather than retailer-marketing-driven. The published exclusion lists do not currently cite a regulator. A published evidentiary basis would weaken the adjudicator-incoherence finding that this report rests on; the chemistry findings would not change.

6. The European Commission, the US Modernization of Cosmetics Regulation Act framework, or the UK Office for Product Safety and Standards publishing a 2024 to 2026 cosmetic standard of identity rule for "natural," "naturally derived," or "from [plant]" claims, with disclosure requirements for the synthesis pathway when the labelled molecule is wholly synthetic. This would close the regulatory gap the report maps. The report would become historical — useful as a this was the gap; this is how it was closed piece, not as a current consumer-information report.

7. A failure to replicate the Bujak 2019 finding — the lipid-elution-greater-than-SLS result on SCS — in an independent peer-reviewed study at the same use concentration and methodology. The chain-distribution-gradient finding on Section IV would soften. The report would still hold at Sections III and V to VIII (the labelling-regime gap is robust to this); the most specific irritation-versus-barrier-elution finding would be dropped.

The methodology is capable of producing each of these outcomes. The labelling-regime gap is robust to most of them. Only finding 6 — the cosmetic standard of identity rule — would close the regime gap directly. The chemistry findings are supporting material; the report's strength is independent of which of the four named substitutes the chemistry vindicates on which parameter.

XI. The Wrapper Was Built for Someone Else

The 1973 CTFA Cosmetic Ingredient Dictionary compiled the names so that one chemist's lauryl sulfate would be another regulator's sodium dodecyl sulfate in a continent-wide market. The 1976 EEC Cosmetics Directive adopted the list. EC 1223/2009 inherited it. Article 19 standardises the name. The chemistry the consumer reads into the name fifty-three years later is not what the architecture was specified to carry, because the consumer was not in the room when it was specified.

What the wrapper does not say, in 2026, is what the chemistry does between the lather and the drain. It does not say which synthesis pathway the molecule travelled, which chain-length distribution sits inside the coco root of the name, which production-grade impurity profile was used, what pH the lather will run at, whether the bar is saponified soap or syndet, whether the formulation contains a chelator for hard water, or what the surfactant does at the second stage of the wastewater treatment plant. It says the name. The name is a vocabulary, fifty-three years old, and the consumer is asked to perform chemistry adjudication on it.

The food sector solved a structurally similar problem in 1886 — the Federal Oleomargarine Act, one page of statute, defining margarine as made in imitation or semblance of butter, with labelling, packaging, and tax requirements that compelled the back of the wrapper to do what the front of the wrapper was suggesting. Modern US food law extends the doctrine: standards of identity for margarine, light and lite claims, filled milk. The cosmetic sector has no parallel rule for natural, naturally derived, or from coconut. The gap between what a cosmetic name implies and what the chemistry behind the name does is a question the regulator has not yet asked.

The Magic Wand candidate for this report is not a brand, a certification, or a product. It is a structural artefact: a cosmetic standard of identity — the analogue of the 1886 Oleomargarine Act, drafted into EC 1223/2009 (and into UK retained law, and into the US Modernization of Cosmetics Regulation Act) — defining natural, naturally derived, and from [plant] with disclosure of the synthesis pathway when the labelled molecule is wholly synthetic; with chain-length distribution disclosure for UVCB substances; with an ethoxylation: yes/no flag; with a formulation pH at 1% solution line; with a format chemistry class — saponified soap or syndet line; and with a labelling-side flag for retailers and third-party reviewers operating without a regulator-grade evidentiary base. Disclosure of these parameters is not technically onerous. The absence reflects the regulation's 1973 inheritance, not the chemistry's 2026 understanding.

The bar in the hand was chosen for a real reason: the bottle it replaces has a low recycling rate, the supply chain is shorter, the format reduces plastic at the point of use. That is a measurable material change, and nothing in this report argues against it. The choice between saponified soap and syndet, between with CAPB and without, between APG-led and SCS-led, between Cosmébio-prohibited SCI and COSMOS-permitted decyl glucoside is now visible on the back of the wrapper, in the names this report has walked through.

The wrapper still cannot say which substitution is real and which is naming. The reader can now read it for what it does say, and read past it for what it does not.