Formaldehyde Testing

Formaldehyde is prohibited as a cosmetic ingredient in the EU and UK. It is also present, at low concentration, in a substantial proportion of cosmetic products on the market, because a group of permitted preservatives works precisely by releasing it slowly into the product. The law permits those preservatives, requires the product to carry a warning once the formaldehyde they release exceeds a threshold, and in 2022 lowered that threshold in the EU fifty-fold. A brand that uses a formaldehyde releaser, or a raw material that carries formaldehyde as an impurity, needs to know how much formaldehyde is in the finished product. The formulation cannot tell them; only analysis can.

Oxford Biosciences determines free and released formaldehyde in finished products and raw materials in its own analytical laboratory. The result is assessed against the threshold that applies in each market the product is sold in and goes to the assessor preparing the CPSR.

Regulatory basis

In the EU and UK, formaldehyde is classified under Regulation (EC) No 1272/2008 (CLP) as a carcinogen of category 1B, a mutagen of category 2 and a skin sensitiser of category 1. Under Article 15 of Regulation (EC) No 1223/2009 a category 1B carcinogen is prohibited in cosmetics, and formaldehyde was accordingly added to Annex II at entry 1577 by Commission Regulation (EU) 2019/831, which at the same time deleted the entries that had permitted it as a preservative at 0.2 percent under Annex V and as a nail hardener at 5 percent under Annex III.

The preservatives that release formaldehyde remain permitted under Annex V, each with its own maximum concentration. Point 2 of the preamble to Annex V requires that all finished products containing such substances be labelled with a warning where the total concentration of formaldehyde released in the finished product exceeds a threshold, irrespective of whether the product contains one or more releasers.

In the EU and Northern Ireland, Commission Regulation (EU) 2022/1181 lowered that threshold from 0.05 percent (500 ppm) to 0.001 percent (10 ppm) and changed the wording of the warning from “contains formaldehyde” to “releases formaldehyde”. It applied to products placed on the market from 31 July 2024, and products placed on the market before that date could be made available until 31 July 2026. Both dates have passed. Every product containing a formaldehyde releaser that is on the EU market today must either carry the warning or be able to show that its released formaldehyde is below 10 ppm. The amendment followed SCCS Scientific Advice SCCS/1632/21, which concluded that the 0.05 percent threshold did not protect consumers already sensitised to formaldehyde and that a fifty-fold reduction would protect the great majority of them.

In Great Britain, the retained text of Annex V has not, at the date of this page, been amended to mirror Regulation (EU) 2022/1181. The threshold remains 0.05 percent and the warning remains “contains formaldehyde”. A product sold in both Great Britain and the EU is labelled to the EU threshold, which satisfies both; a product sold only in Great Britain has more room, but the sensitisation reasoning behind the SCCS advice applies to British consumers equally, and the CPSR addresses it.

Elsewhere, Japan prohibits formaldehyde in cosmetics entirely and applies the prohibition to releasers that produce it. China’s Safety and Technical Standards for Cosmetics permit formaldehyde as a preservative up to 0.2 percent, with a prohibition in aerosols and a labelling requirement above 0.05 percent. In the United States, formaldehyde is not restricted in cosmetics at federal level, it is listed under California Proposition 65, and the FDA proposed in 2023 a rule to prohibit formaldehyde and formaldehyde-releasing chemicals in hair smoothing and straightening products; the status of that proposal is checked at the time of assessment.

The formaldehyde releasers

The following preservatives, all permitted under Annex V in the EU and UK, release formaldehyde by hydrolysis in the presence of water. The rate and extent of release depend on pH, temperature, water content, the presence of other ingredients that react with formaldehyde, and time.

Preservative (INCI)	Annex V maximum	Release characteristics
DMDM hydantoin	0.6 percent	Rapid, substantial release; the releaser most likely to exceed 10 ppm at typical use levels
Imidazolidinyl urea	0.6 percent	Moderate release, increases with temperature and at neutral to alkaline pH
Diazolidinyl urea	0.5 percent	Higher release than imidazolidinyl urea at equivalent concentration
Sodium hydroxymethylglycinate	0.5 percent	Substantial release, pH-dependent
2-Bromo-2-nitropropane-1,3-diol (bronopol)	0.1 percent	Release increases at alkaline pH and with heat; also a nitrosating agent, which is the more important safety issue where amines are present
5-Bromo-5-nitro-1,3-dioxane	0.1 percent (rinse-off only)	Similar to bronopol
Benzylhemiformal	0.15 percent (rinse-off only)	Releases formaldehyde and benzyl alcohol
Methenamine	0.15 percent	Slow release, accelerated at acidic pH
Tris(hydroxymethyl)nitromethane	0.2 percent	Moderate release

Quaternium-15, formerly the most widely used releaser, was prohibited in the EU by Commission Regulation (EU) 2021/1902 with effect from 1 March 2022 following its classification as a carcinogen of category 1B, and is listed in Annex II.

Formaldehyde also appears in products that contain no releaser, as an impurity in ethoxylated raw materials such as polysorbates, PEG esters and laureth sulphates, where it is a by-product of the ethoxylation process, and in some natural extracts. These sources are usually below 10 ppm in the finished product but not always, and a product with a high load of ethoxylated surfactants can approach the threshold without any preservative contribution at all.

Hair smoothing and straightening products of the “Brazilian keratin” type contain formaldehyde or its hydrate, methylene glycol, as the active cross-linking agent, at concentrations from 0.2 percent to several percent. In the EU and UK these products are illegal unless formaldehyde-free, and analysis of such products is the means of establishing whether a “formaldehyde-free” claim on an imported product is true.

Why “free” and “released” formaldehyde are different measurements

Formaldehyde in a product containing a releaser is in equilibrium: some has been released and is present as free formaldehyde and its hydrate, and the rest remains bound in the releaser molecule. The equilibrium moves with time, temperature and pH. A product tested a week after manufacture may contain 5 ppm of free formaldehyde and 60 ppm at the end of its shelf life. The labelling threshold refers to the total concentration of formaldehyde released in the finished product, and the SCCS advice was explicit that this means the formaldehyde available to the consumer, not the amount present on the day of a single measurement.

The analysis therefore addresses both. Free formaldehyde is the concentration present at the time of testing, determined by a method that does not itself decompose the releaser. Released formaldehyde is the total available formaldehyde, determined after the releaser has been driven to complete hydrolysis under controlled conditions, and it is the figure the labelling decision rests on for a product that will be stored and used for months. Where a stability study is running, free formaldehyde is also determined on the aged samples at the end of the study, which gives the actual figure at the end of life and is the strongest evidence for or against the warning.

How the analysis is done

1. Sample preparation

A weighed portion of the product, typically 1 g, is dispersed in water, with a co-solvent where the product is anhydrous or strongly hydrophobic, and buffered. For free formaldehyde, the extraction is performed at ambient temperature and the sample is taken to derivatisation immediately, so that no further release occurs during preparation. For released formaldehyde, the dispersion is heated under conditions that hydrolyse the releaser fully, then cooled and derivatised. A spiked sample, prepared by adding a known amount of formaldehyde to a second portion of the product, is carried through the same steps to confirm that the matrix does not suppress recovery.

2. Derivatisation

Formaldehyde is a small, highly reactive, volatile molecule with no chromophore, which makes it difficult to measure directly. It is converted to a stable, strongly UV-absorbing derivative by reaction with 2,4-dinitrophenylhydrazine (DNPH) in acidic solution, forming formaldehyde-2,4-dinitrophenylhydrazone. The reaction is quantitative and the derivative is stable for days. Because every aldehyde and ketone in the sample forms a hydrazone, the derivatisation is not selective, and the separation that follows is what makes the method specific.

3. Chromatography and detection

The derivatised extract is injected onto a reversed-phase C18 high-performance liquid chromatography column and eluted with an acetonitrile-water gradient. The formaldehyde-DNPH hydrazone elutes at a characteristic retention time, separated from the hydrazones of acetaldehyde, acetone, benzaldehyde and the other carbonyl compounds that fragrance materials and raw materials contribute, and from excess DNPH reagent. Detection is by UV absorbance at 360 nm, where the hydrazone absorbs strongly. Identity is confirmed by retention time against a certified formaldehyde-DNPH standard and, where any doubt remains, by diode-array spectrum or by mass spectrometry.

Colorimetric methods, in which formaldehyde reacts with acetylacetone and ammonium acetate to form a yellow dihydropyridine measured at 412 nm, or with chromotropic acid to form a violet product, are faster but are subject to interference from other aldehydes and from coloured products. They are appropriate as a screen and were the basis of the original EU official methods for free formaldehyde in Directives 82/434/EEC and 90/207/EEC, but the HPLC-DNPH method is the one used for a result that has to bear a labelling decision at the 10 ppm level.

4. Calibration and quality control

The method is calibrated with a series of formaldehyde standards, prepared from a certified formalin solution whose concentration has been verified by iodometric titration, derivatised alongside the samples. Each batch includes a reagent blank, a calibration check, a duplicate and the spiked recovery sample. The limit of quantification in the product is approximately 1 mg/kg, a tenth of the EU labelling threshold, so that a result reported as below the limit is a meaningful negative.

5. Reporting

Free and, where determined, released formaldehyde are each reported in mg/kg and as a percentage by mass of the product. The report states the sample’s age and storage history, since the free formaldehyde figure depends on them, and compares each figure with the labelling threshold in each market the product is sold in.

Interpreting the result

A released formaldehyde result above 10 ppm in a product for the EU market leaves the brand with two options: carry the warning “releases formaldehyde” on the label, or reformulate to a non-releasing preservative system. Most brands reformulate, because the warning has become a commercial liability and because the sensitisation reasoning behind it is sound. The report identifies which ingredient is contributing and by how much, so that reformulation can target the right one; a product with a releaser at low concentration and a high load of ethoxylates may be over the threshold without the releaser being the main source. Where a product is reformulated, a repeat analysis on the new formulation is the evidence that the warning is no longer required, and it is filed with the CPSR.

What the report contains

• Sample identification, form, batch, date of manufacture where known and date of receipt
• Method summary: extraction conditions for free and released formaldehyde, derivatisation, chromatographic and detection conditions
• Free formaldehyde and released formaldehyde in mg/kg and percent by mass, with the limit of quantification
• Quality control summary: blank, calibration check, duplicate and spiked recovery
• Comparison with the labelling threshold in each relevant market
• Identification of the likely source where the result requires it
• Signature of the analyst and the Head of Laboratory Sciences

What to send

20 g or mL of product from a single batch, in its final packaging where possible, with the date of manufacture. The full formulation with percentages is required so that releasers and ethoxylated ingredients can be identified before analysis. Where aged samples from a stability study are available, send them alongside the fresh sample.

How it fits with the CPSR

The formaldehyde result supports the impurities and traces section of Annex I, Part A, Section 2, the labelling decision under Annex V, and the sensitisation assessment in Part B. Where we prepare the CPSR, the analysis is ordered by the assessor whenever the formulation contains a releaser or a substantial ethoxylate load, and the labelling conclusion is written from the data. Where a client brings a supplier’s declaration that a raw material is “formaldehyde-free”, we check what was measured, by what method, at what limit and on what date before relying on it.