Contents: Legislation: CLP and the GHS; Mixtures; Product classification; How to classify mixtures: testing; How to classify mixtures: bridging principles; How to classify mixtures: threshold concentrations; Expert judgement and weight of evidence; References
You must classify chemical substances and mixtures before they are placed on the market. This guide tells you how the classification processes works for mixtures. Substance classification is described in an earlier post.
Legislation: CLP and the GHS
Chemical classification is a shorthand description of the hazards of a product, and the cornerstone of chemical hazard communication in the supply chain, as expressed in safety data sheets and on labels.
The Classification, Labelling and Packaging (CLP) Regulation  specifies the criteria for the classification of hazardous properties of chemical products in the EU.
The CLP Regulation is based on the United Nation’s (UN) Globally Harmonized System of Classification and Labelling of Chemicals (GHS) . Versions of the GHS are implemented in all industrialized nations, and provides an ‘off-the-peg’ chemical control policy for developing countries.
One might assume that all countries would fully adopt the GHS, thereby eliminating differences in chemical classification throughout the world. However, there are important reasons why this ideal situation is not realised, as discussed in a previous blog post on the classification of substances.
The requirement to classify and label according to the CLP Regulation applies to most chemical products supplied in the EU, such as coatings, detergents, and air fresheners. There are exemptions for certain products already regulated by other legislation, such as:
- Medicinal and veterinary products, and medical devices
- Cosmetic products
- Food, and food additives and animal feeds, and feed additives
Brexit has complicated the picture. Great Britain has separate legislation to the EU and Northern Ireland (which retains the EU system). The GB legislation on classification and labelling  is closely aligned with the EU CLP Regulation. The EU legislation, as it was at 31 December 2020, was transposed into UK legislation, with changes relating to the roles and responsibilities of the UK authorities.
The UK government will take account of recent and future changes in EU legislation, but new transpositions into UK law will be delayed, changed, or not implemented at all, so divergence over time is inevitable.
The duplication and divergence of EU and GB legislation places an increased regulatory burden on GB-based chemical businesses.
In the legal framework for chemical classification, products can be:
- Substances (often a single chemical component, but can be more complex, eg petroleum substances),
- Mixtures (deliberate combination of two or more substances), or
- Articles (products dependent more on shape than chemical composition).
The majority of chemical products are mixtures, eg lubricants, adhesives, cleaning liquids.
The classification of substances is described in a previous post.
Articles are not within the scope of the CLP Regulation and do not require hazard classification.
Classification is the categorization of chemical hazard (see our Compendium of chemical hazards) into easily identifiable, standard descriptions, such as flammable, toxic, irritating, or hazardous to the aquatic environment.
The details of the criteria and definitions, and references to test methods, are given Annex I of the CLP Regulation as follows:
- Part 2: Physical hazards
- Part 3: Health hazards
- Part 4: Environmental hazards
- Part 5: Supplemental EU hazards
The GHS and CLP Regulation break chemical hazard into a number of Hazard Classes; 29 in total. Not all chemicals hazards are covered (eg example dust explosivity and ecological hazards to the terrestrial environment), so that a substance ‘does not meet the criteria for classification’ may still pose a risk.
The classification comprises the Hazard Classes, their subdivisions (usually called Categories, but also called Types or Divisions), and the associated hazard statements.
For each Hazard Class, with its Category subdivisions, there is a definition, which is linked with the test method to give a set of criteria. This is described in detail on the substance classification post.
All data pertinent to the classification of the mixture should be considered for hazard classification:
- Test results
- Bridging (ie the read-across of data from a similar mixture)
- Use of threshold concentrations, above which the product become classified
- Expert judgement (eg form placed on the market, bioavailability)
Each of these methods of classifications is discussed below.
How to classify mixtures: testing
Theoretically, testing and classification of mixtures can be similar to that for substances: (1) the mixture is tested according to the prescribed test method; (2) the results are compared with the criteria for the respective Hazard Class and Categories; (3) if the criteria are fulfilled, then the substance is classified in that Hazard Class.
Direct testing of the mixture, when applicable, gives definitive results for classification, and takes precedence over other methods.
However, some thought has to be given to the appropriateness of tests when applied to mixtures.
Testing is useful for many physico-chemical properties of mixtures, such as melting point (or range), density, surface tension, and vapour pressure. It is not valid for properties where the ingredients of the mixture separate and express their individual values, for example in the partition coefficient.
For physical hazards, such as flammability, explosivity, or corrosivity to metals, it is usually recommended to test the mixture if there is concern for the hazard. There is usually no concern for a hazardous property of a mixture if none of the substances in the mixture has that property.
However, there are cases where the hazardous substances in a mixture create a new hazard. An oxidising solid and a flammable liquid can produce an explosive mixture.
In the EU and UK, animal rights are important considerations for testing chemical products. Thus, most mixtures are evaluated on the basis of their component substances and their concentration, rather than through animal tests.
EU and UK laboratories have an ethics committee, that will likely deny a request for tests on mixtures involving animals, unless there is a safety concern in not testing, such as a potential for synergistic effects from the components of a mixture.
Some tests for toxicological hazards have been developed that use animal parts (ex vivo tests) or cells (in vitro tests). They are not suitable for all mixtures, but if appropriate can be used for classification purposes. Most common of these are for skin corrosion or irritation, eye damage or irritation, skin sensitisation, and the Ames test as an indicator of mutagenic potential.
If test results are available, then they override the results from bridging or use of threshold concentrations for classification purposes.
Some toxicological tests are not suited for mixtures; carcinogenic, mutagenic, and reproductive toxicity hazards of a mixture are evaluated only on the basis of the components and their concentration.
Ecotoxicity testing on chemical mixtures is rare.
From an animal welfare perspective, testing on fish is not allowed in the EU and UK.
The two other levels of the food-chain used for hazard testing, Daphnia (water-flea) and algae are allowed for testing, but technical issues may make testing expensive, or even impossible for mixtures.
Biodegradability, and important factor in hazard classification, is not a meaningful concept for mixtures. Imagine a sand and sugar mixture: the sugar is biodegradable, but the sand is not, and it is not helpful to report a value for the mixture.
How to classify mixtures: bridging principles
The use of bridging principles in the classification of mixtures is useful in only limited circumstances, particularly when some test results on a similar mixture are already available.
If a formulation with test data for a particular hazard is tweaked, for example with small changes in concentration or by dilution, then the assessor may be able to draw conclusions on the classification of the new formulation, using bridging principles in the CLP Regulation, Annex I.
This may be useful for consumer product manufacturers with frequent changes to formulations, or those marketing concentrates and diluted forms.
How to classify mixtures: threshold concentrations
The usual way of classifying mixtures for health and environmental effects is based on the classification of the ingredients and their concentration.
In general, physical hazards (eg flammability, explosivity) do not have threshold concentrations, and testing is recommended for correct classification.
For accurate classification, the assessor has to know the identity (eg name, EC or CAS number) and percentage of all hazardous ingredients.
For formulators, ingredient classification and percentage should be available on the safety data sheet (SDS) provided with your raw materials, particularly if your SDS complies with EU or UK Regulations. Composition information can be more difficult to obtain from non-EU/UK suppliers, but the importer has a legal obligation to ensure correct hazard classification of their products.
Searching for substance hazard data is therefore an integral part of the process for classifying mixtures. ECHA’s ‘Search for Chemicals’ is a valuable resource. The formulator of mixtures should be able to demonstrate due diligence in assessing the hazards of their products.
The procedure for mixture classification has the following steps:
- Identification of relevant ingredients (ie hazardous ingredients above a minimum cut-off value)
- Identification of concentration limits for each Hazard Class of the ingredients, which would cause classification of the product. Generic concentration limits are given in the CLP Regulation, Annex 1, but some substances have specific concentration limits (SCLs) in their harmonised classification (CLP Regulation, Annex VI) or REACH registration dossier (see substance classification article).
- Allocation of the hazard class and sub-division of the mixture (or as ‘not meeting the critieria for classification’).
While this process is potentially very complicated, it is manageable for most mixtures with a limited number of ingredients. In situations where the assessor has many mixtures with complicated formulations, specialist software can aid the calculation of mixture classifications.
For some hazard classes, the concentration of each relevant ingredient is additive. For example, two ingredients classified as Skin Irritant, Category 2 and each present at 3% in a mixture, will have the same effect on the mixture classification as either ingredient at 6%.
A special additivity formula is available for acute toxicity (CLP Regulation, Annex I, 126.96.36.199.).
Aquatic toxicity is also considered additive.
Some hazards are not considered additive because the mechanism of action is specific to the particular chemical. This is the case for skin sensitisation, specific target organ toxicity (STOT), carcinogenicity, and mutagenicity. So if you have a mixture containing several ingredients with a classification of STOT RE 2, H373 (may cause damage to organs through prolonged or repeated exposure), but all are present below 10%, then the mixture would not meet the critieria for classification for this Hazard Class.
In some cases, irritation/corrosion is not considered additive; for strong acids and bases the pH is considered a more reliable classification criterion. There are lower concentration limits for surfactants, and other ingredients considered to act with a non-additive mechanism (aldehydes, phenols and inorganic salts are mentioned in the Regulation).
Expert judgement and weight of evidence
Expert judgement and weight of evidence in hazard classification usually relates to using non-standard information for the purposes of classification. Some examples where expert judgement is required are:
- Test data is not from standard methods, or has deficiencies, so judgement is required to evaluate their relevance, eg intravenous route of exposure, impure test substances, unusual test species (including epidemiology in humans)
- Available data is inconclusive or conflicting. In these circumstances, often the worst-case classification is appropriate.
- Weight of evidence can be used when multiple data are inconclusive on their own but provide similar results, such as poorly reported literature values combined with structure–activity relationships or calculated results.
Further, expert judgement can to reduce the severity of the classification determined from generic concentration limits.
There are some specific circumstances where this may be helpful:
- If the form or physical state of the mixture is different from that of the tested component substance, eg inhalation classification for granular solids or solutions of solids where exposure via inhalation route can be precluded
- Where the hazardous component of a mixture is not biologically available, eg by encapsulation in a solid, insoluble matrix
Use of these derogations is controversial, and supporting evidence is recommended.
[back to Legislation: CLP and the GHS]
 Regulation (EC) No 1272/2008 of the European Parliament and of the Council of 16 December 2008 on classification, labelling and packaging of substances and mixtures (as amended).
 Globally Harmonized System of Classification and Labelling (GHS); Eighth Revised Edition; United Nations; 2019.
 The Chemicals (Health and Safety) and Genetically Modified Organisms (Contained Use) (Amendment etc.) (EU Exit) Regulations 2019; UK Statutory Instrument 2019 No. 720.