SDS Section 11: Toxicological information

The section that has to show its working

Section 2 asserts a classification. Section 11 is where the assertion is justified — the health hazard classes one at a time, the data behind each of them, and what exposure does to a person by route, by dose and by duration.

Both the EU and UN texts open by naming the reader: medical professionals, occupational health and safety professionals and toxicologists. That is unusual on a safety data sheet, and it sets the standard of the writing. Section 4 is written for whoever is standing next to the casualty; section 11 is written for the clinician the casualty is taken to, and for the assessor deciding whether the classification holds.

The test is blunt. If section 2 says Skin Sens. 1, section 11 has to say what that rests on. If section 2 says nothing about carcinogenicity, section 11 still has to say something — the class does not disappear because it was not triggered.

11.1 — one answer per hazard class, including the negative ones

Start with the heading, because most sheets get it wrong. Regulation (EU) 2020/878 renamed the subsections of section 11. The current EU headings are 11.1 “Information on hazard classes as defined in Regulation (EC) No 1272/2008” and 11.2 “Information on other hazards”. A sheet still printing the pre-2020 subsection titles is, on that evidence alone, being generated from a superseded template.

The ten classes are a fixed list, not a menu

REACH Annex II and UN GHS Annex 4 give the same ten health hazard classes: acute toxicity; skin corrosion/irritation; serious eye damage/irritation; respiratory or skin sensitisation; germ cell mutagenicity; carcinogenicity; reproductive toxicity; STOT-single exposure; STOT-repeated exposure; aspiration hazard. Both texts then add the same sentence: these hazards shall always be listed on the safety data sheet.

“Always” is doing real work. The list is not the classes the product happens to be classified for — it is the set of questions the section answers, ten times, whatever the answers turn out to be.

”Not classified” is four different statements

Subsection 11.1.1 is the requirement most often flattened into a blank. Where the sheet states that a substance or mixture is not classified for a hazard class, it must say which of four things is true: there is a lack of data; obtaining the data was technically impossible; the data were inconclusive; or the data were conclusive but insufficient for classification — and in that last case the sheet must specify “based on available data, the classification criteria are not met”. Subsection 11.1.4 supplies the other half: where the criteria are not met, the information supporting that conclusion has to be given.

Those four states are not interchangeable. “No study exists” and “a study exists and came back below the criteria” are opposite pieces of evidence, and only one of them tells the reader the hazard was looked for.

The data have to describe the product as sold

Subsection 11.1.2 fixes what the section is about: the data apply to the substance or mixture as placed on the market. For a mixture, the properties described are those of the mixture as a whole, except where Article 6(3) of CLP applies; where they exist, the relevant properties of the hazardous substances in it are given as well.

Where the mixture has not been tested as a whole for a given health effect, 11.1.10 sends the author back to the substances listed in section 3. That is the structural join between the two sections, and it is why a vague concentration range in section 3 quietly weakens section 11 too.

Where the data run out

Two subsections handle absence, and they pull in the same direction. 11.1.9 permits data on similar substances or mixtures — provided the similar substance or mixture is identified, and provided the sheet states clearly that specific data were not used or not available. 11.1.11.2(c) closes the other gap: where information on interactions between substances in a mixture is not available, assumptions shall not be made, and the health effects of each substance are listed separately instead.

Read-across is allowed; invention is not; and the reader has to be able to tell which one is in front of them.

Numerical measures of toxicity, and what they are for

Subsection 11.1.2 in the EU text and item (d) in the US text name the same family of values — LD50, LC50 and acute toxicity estimates. They sit in section 11 because they are the only part of it a downstream user can compute with.

An acute toxicity estimate is route-specific: oral, dermal, inhalation. It is what a classifier feeds into the additivity calculation when a mixture has not been tested as a whole. The inputs are the substances declared in section 3 and the values published in section 11 of their own sheets; the output is the mixture’s ATE by route, and from that its acute toxicity category. The ATE mixture calculator runs that chain.

Two consequences follow. A section 11 that offers only a qualitative description leaves the formulator downstream unable to classify anything built from the product. And a section 11 whose values cannot be reconciled with the category shown in section 2 is telling the reader that one of the two was not rechecked at the last revision.

11.2 — the hazards CLP has no class for

Subsection 11.2 is new in the 2020/878 layout, and it exists because the classification system does not cover everything a health professional needs.

11.2.1 Endocrine disrupting properties. Where a substance has been identified as having endocrine disrupting properties in subsection 2.3, section 11 must carry the human-health side of that finding: brief summaries of the information derived from applying the assessment criteria laid down in Regulations (EC) No 1907/2006, (EU) 2017/2100 and (EU) 2018/605. Endocrine disruption is not a CLP hazard class, which is exactly why it needed its own place in the sheet.

11.2.2 Other information. Other relevant information on adverse health effects must be included even when the classification criteria do not require it. It is the provision that stops “not classifiable” being read as “not worth mentioning”.

Where the US sheet asks a different question

OSHA’s Appendix D covers much of the same ground as a flat lettered list — routes of exposure, symptoms, delayed, immediate and chronic effects, numerical measures, interactive effects — and then asks something the EU text never asks: whether the chemical is listed in the NTP Report on Carcinogens (latest edition), or has been found to be a potential carcinogen in the IARC Monographs (latest edition), or by OSHA.

That is a listing question, not a classification question, and the two can diverge. A substance can carry no harmonised carcinogenicity classification in the EU and still appear in an IARC monograph; a US sheet has to disclose that either way. The EU handles the same ground through classification: for substances subject to registration, section 11 must carry the result of comparing the available data with the CLP criteria for CMR categories 1A and 1B, following point 1.3.1 of Annex I to REACH.

Appendix D item (g) is the other US-only requirement. Where specific chemical data are not available, the preparer must indicate that alternative information is used and the method used to derive it — the text’s own example is drawing on a class of chemicals and applying SAR. The EU asks the author to name the analogue; the US asks the author to name the method.

How section 11 connects to the rest of the sheet

• Section 2 is the claim and section 11 is the evidence — the EU text requires section 11 to be consistent with the classification and with the registration or chemical safety report.
• Section 3 supplies the substances 11.1.10 falls back on for any health effect the mixture has not been tested for.
• Section 4 is the operational consequence of 11.1.5 to 11.1.7: the symptoms described there are the ones first aid has to answer.
• Section 8 selects controls against the routes of exposure 11.1.5 identifies — a skin notation on an exposure limit should already be explained by what section 11 says about dermal uptake.
• Section 12 is the same exercise for organisms other than people. The two get confused; human health data belong here, not there.
• Section 16 carries the revision history, the only place a reader can see that section 11 moved when the classification did.

What a good section 11 looks like

1. Are all ten classes present? Count the headings. Ten questions, ten answers, no silent omissions — and where the answer is that data are missing, the sheet says which kind of missing.
2. Can you trace each statement to a study type? Human, animal or in vitro, with the species or the cell type identified. Subsection 11.1.7 requires that distinction, and it is the first thing a toxicologist looks for.
3. Does it agree with sections 2, 3 and 8? Section 11 is where those three are justified. Where the toxicology and the classification disagree, the sheet has not been reviewed as a document.