Feedback from ECHA on phases 1 and 2 was particularly critical of identity quality; “correct and unambiguous substance identification is a frequent shortcoming in registration dossiers”. A simple review of the requirements is therefore of value.
According to the REACH legislation, substances fall into three broad categories:
- Single constituent (main component typically >80%, with the remainder being impurities)
- Multi-constituent substances (no single component at >80% but two or more between >10% and <80%, with impurities also possible)
- Undefined, Variable, Complex or Biological (UVCB, defined by the starting material and the process, no impurities possible, all contained materials are constituents)
For single and multi-constituent substances, all impurities >1% must be identified and named. This limit drops to >0.1% if it is believed that the impurity may be hazardous. The total composition of the substance in both cases must add up to 100%. Note that multi-constituent substances are not just formulations. Instead, they are cases where it is physically impossible to create the two substances individually while maintaining the same properties. For UVCBs, all components present at >10% must be identified. However, naming those under 10% is also an advantage, even if only by a structure proposal. Unknowns should be grouped as far as possible and categorised according to general structure type or key similarities. The standard recommended analytic techniques are UV-Visible and IR spectroscopy for basic identity, NMR spectroscopy and/or mass spectrometry for precise identity and GC or HPLC for purity. It is important that these techniques are employed to the maximum extent to avoid the demand for repeat analysis. For example, consider also acid and alkaline environments in the UV-Visible spectrum and ensure the NMR spectrum is run fully and not just to the point of the expected peaks.
These techniques should not be considered as exhaustive. Further tests, as appropriate to the substance nature, should be added in order to justify the identity. Examples of additional techniques are: XRD, atomic absorption, carbon v metal balance, GPC, database comparisons, CHN combustion analysis, Karl-Fischer or silver nitrate titration, BET surface area and electron microscopy Traceability is essential for high quality reporting and GLP should be employed if further toxicity tests are envisaged. A good starting point on each report would be; substance name, CAS number, batch number, manufacturing data, batch expiry date, purity pre-estimate, substance nature and appearance, laboratory name, laboratory address, analysis date, operator name and signature plus laboratory manager name and signature.
A detailed interpretation of the results should be provided explaining clearly how this indicates the declared composition. The raw data, spectra and full method description should also be given. It is important that the declared composition covers all sources of the substance as handled by the legal entity. This is particularly relevant for importers whose materials may come from diverse non-EU sources. As a result, multiple spectra may be required. The lead registrant will usually release a basic Substance Identity Profile which is usually limited to just the substance identifiers (such as CAS and EC number) and maximum level of any common impurities. Prescriptive analytical instructions are not normally given. The co-registrant therefore remains fully responsible for their private analytics, identity and composition with the lead having no responsibility for certifying these or confirming any applicability to the joint registration.
Full liability for the substance identity in a submitted registration (lead or co-registration) lies with the individual registrant. ECHA and the national authorities are unable to accept any aspect of liability so unsupported statements, based on identity belief or assumptions, will generally be met with a request to provide further proof. Finally, it is vital that analytics are carried out as early as possible. Experience has shown that surprises often emerge such as the identification of previous false identity assumptions. Addressing such issues closer to the registration deadline may not be possible and could prevent a successful registration.
