Material guide: diacetyl — its inhalation toxicity field reads 'not determined', and it is the molecule famous for causing a lung disease

About a 5 minute read.

The short version

• Name: diacetyl, 2,3-butanedione, CAS 431-03-8, FEMA 2370
• Odour type: buttery; odour: buttery, sweet, creamy, pungent, caramellic
• Strength: high (evaluate at 1% or less), rank 3; substantivity 44 hours at 10% in DPG
• Suppliers: 66; Cramer Class II
• Flash point 7.22 °C — lower than ethanol's
• Inhalation toxicity field: Not determined

That field first

toxicity_inhalation: Not determined.

And diacetyl is the most famous "inhalation causes disease" molecule in the entire flavour industry.

I measured in article 100 that this field looks 100% covered and is 1.7% real once "Not determined" is stripped — 731 records out of 42,225 carry an actual number.

This record is not one of the 731.

This is not the database being uniquely negligent — that field collects acute animal LC50 figures, and diacetyl's problem is chronic occupational exposure. But the effect is the same: read only this cell and you would conclude there is nothing here.

What it did

Hubbs and colleagues open their 2019 Toxicologic Pathology review:

Flavorings-related lung disease is a potentially disabling and sometimes fatal lung disease of workers making or using flavorings. First identified almost 20 years ago in microwave popcorn workers exposed to butter-flavoring vapors … In some cases, workers develop bronchiolitis obliterans, a severe form of fixed airways disease. Affected workers have been reported in microwave popcorn, flavorings, and coffee production workplaces.

Volatile α-dicarbonyl compounds, particularly diacetyl (2,3-butanedione) and 2,3-pentanedione, are implicated in the etiology. Published studies … document their ability to cause airway epithelial necrosis, damage biological molecules, and perturb protein homeostasis. With chronic exposure in rats, they produce airway fibrosis resembling bronchiolitis obliterans.

"Coffee production" is worth noting — diacetyl forms during roasting, so the problem is not confined to products with flavouring added.

And the dose figures are low

Park and Gilbert's 2018 paper in the Journal of Occupational and Environmental Medicine is a risk assessment at one microwave popcorn plant: current employees' medical histories and lung function tests, combined with air sampling over 2.7 years.

Percent-of-predicted FEV1 declined with cumulative exposure (0.40 per ppm-yr, P < 10⁻⁵) as did percent FEV1/FVC (0.13 per ppm-yr, P = 0.0004) … Lifetime respiratory impairment prevalence of one per thousand resulted from 0.005 ppm diacetyl.

0.005 ppm is five parts per billion.

(Read carefully: that is lifetime cumulative occupational exposure, not a concentration you smell once. And it is one plant's risk model, not a general standard.)

Its flash point is 7.22 °C

flash_point: 45.00 °F. TCC ( 7.22 °C. ) — and no "(est)" marker, so it is measured.

In a separate article I ranked measured flash points for materials with 15+ suppliers; the median is 92.2 °C. Diacetyl's 7.22 °C is among the six lowest, and below ethanol's 9.4 °C.

It is a flammable liquid at room temperature. With 66 suppliers, there is a decent chance there is a bottle on your bench.

It occurs in places you would not expect

The occurrence field lists dozens of sources: bilberry, blackberry, butter, cabbage, cacao, caraway, celery, cheese at 605 ± 354 µg%, chicory root, Java citronella oil, coffee, cranberry, currant, geranium oil, grape, guava, buckwheat honey, kohlrabi, lavandin oil, lavender oil…

Lavender oil and geranium oil contain diacetyl. Not added — naturally present. There is a little of it in every lavender formula you make.

(At far lower concentrations, obviously. That cheese figure of 605 µg% is on the order of 6 ppm.)

The notes field gives a specific dilution

notes opens, unusually, with an instruction:

Dilute with two parts diacetyl to 1000 parts linalool carrier of aroma of butter, vinegar, coffee, and other foods.

Two in a thousand is 0.2%. And the evaluation field recommends 1% or less. This material works in the parts-per-thousand range.

How to use it (if you are going to)

• Evaluate at 1% or less, with ventilation.
• It is the core of the butter, cream and caramel line, at ppm to parts-per-thousand levels.
• Flash point 7.22 °C — treat it as a flammable liquid. Away from heat and sparks.
• Do not inhale the vapour and do not open the bottle in a closed space. That is the one line in this article I want to lean on.
• It is a flavour material (category reads flavoring agents); its perfumery use is limited.
• Check the storage specification — it is among the 145 records told to refrigerate.

What this doesn't establish

• I have not smelled it. This is fields, whole-database statistics, and two papers.
• Both papers are occupational medicine and toxicologic pathology, measuring long-term inhalation exposure in factories — not domestic or studio use. I have no data with which to convert those numbers to your bench.
• 0.005 ppm is the output of a risk model on one plant's data, not a regulatory limit or a general threshold.
• "The inhalation field collects acute LC50 figures" is my inference about that field's contents; the database does not say what it collects.
• I quoted a concentration only for the cheese entry; the rest carry no figures, and that field sweeps up everything anything has been detected in.
• I checked no occupational exposure limit for diacetyl in any jurisdiction.