Summary

The Advisory list for self-classification of dangerous substances

When chemical substances are classified in terms of the danger they represent, their inherent properties are assessed on the basis of the knowledge and information available /2, 7, 8/. Such assessments are often carried out on the basis of results from laboratory test because the hazard classification criteria to a large extent refer to such results. Complete identification of the properties for which hazard classification criteria exist may be very time- and resource consuming. As a result, EU-harmonised classifications only exist for approximately 7.000 substances

When harmonized data are lacking, the companies are according to the legislation, before marketing obliged to self-classify the chemicals they import or produce. Lack of test data on hazardous properties of chemicals however, makes it difficult for companies to meet these obligations. As an aid for addressing this issue, the Danish Environmental Protection Agency (DK-EPA) has previously published a substance list - "The advisory list for self classification of chemical substances" – with advisory classifications for a number of endpoints and chemical substances /3, 10, 11/. The Advisory self-classification list has been created by the use of (Q)SARs ((Quantitative) Structure-Activity Relationships) to predict the intrinsic and harmful properties of chemicals and solely based on this to make advisory self-classifications on chemicals with lack of test data.

The advisory list has been published in three previous versions /3, 10, 11/. In all versions, the recommended classifications were made according to the Dangerous Substances Directive 67/548/EEC (DSD) /2/. The present report supplement the previous reports by describing a new advisory list based on the new regulation for classification and labelling of chemicals (CLP) /5/.

The updated Advisory self-classification list contains the results of a systematic assessment of 49,292 discrete^[1] organic EINECS substances in relation to the following endpoints for which (Q)SAR model predictions were available:

Mutagenicity
Carcinogenicity
Reproductive toxicity (possible harm to the unborn child)
Acute oral toxicity
Skin irritation
Skin sensitisation
Danger to the aquatic environment

The updated advisory list contains 33.835 chemicals with advisory classifications for one or more of the selected endpoints, and is available as an Excel file for download and as an online searchable database from DK-EPA's website (http://www.mst.dk).

This report describes the revisions made in order to prepare a version of the advisory classifications that meet the classification criteria set out in the new CLP-regulation for classification and labelling of chemicals /5/. The (Q)SAR models employed in the latest earlier version of the advisory self-classification list has not been revised. For details about the theoretical work ((Q)SAR-modelling and –algorithms) reference is therefore made to the latest two previous reports /10,11/.

[1] Discrete organic substance means organic substances with an unambiguous 2D structural formula.