Material guide: acetoin — its vapour pressure puts it in the top band, and it lasts 28 hours

About a 5 minute read.

The short version

• Name: acetoin, 3-hydroxy-2-butanone, CAS 513-86-0, FEMA 2008
• Odour type: buttery; odour: sweet, buttery, creamy, dairy, milky, fatty
• Strength: high (evaluate at 1% or less), rank 3; substantivity 28 hours at 100%
• Vapour pressure 2.69 mmHg — the top band, where the median substantivity is 4 hours
• Suppliers: 105; storage: refrigerate, 12 months
• Melting point 90–91 °C, boiling point 147–148 °C — that pairing is itself the clue

It does not follow the rule

Last time I computed three bands from vapour pressure: above 1 mmHg is top (median substantivity 4 hours), 0.01 to 1 is heart (24 hours), below 0.01 is base (212 hours).

Acetoin's vapour pressure is 2.69 mmHg, squarely in the top band.

And its substantivity is 28 hours — seven times that band's median, and past the heart band's median too.

This material is what the remaining 0.353 of that ρ = −0.647 looks like.

The clue is in the physical properties

Melting point	90.00 to 91.00 °C
Boiling point	147.00 to 148.00 °C
Appearance	colourless to pale yellow liquid to solid

A material whose melting point sits only 57 degrees below its boiling point is very unusual. Most room-temperature liquids melt below zero — diacetyl at −2.4 °C, linalool lower still.

And the appearance field says "liquid to solid", not "liquid".

The likeliest explanation is that those two numbers are measuring different things. Acetoin dimerises on standing into a crystalline solid, and that crystal melts high — the melting point is the dimer's, the boiling point is the monomer's.

If part of what you bought is dimer, it has to dissociate before it can evaporate off a blotter — which may be where the 28 hours comes from.

(This is an inference. The database does not say which form either number describes, and I have no experiment to check it. But that pairing of numbers is plainly wrong, and dimerisation is a known property of this material.)

It is diacetyl's other half

The notes field says it plainly:

a product of fermentation. It is a component of the butanediol cycle in microorganisms. In mammals it is oxidized to carbon dioxide. Constituent of beer, wine, fresh or cooked apple, leek, corn, honey, cocoa, butter, cheeses, roasted coffee and other foodstuffs.

The butanediol cycle: acetoin ⇄ diacetyl, with 2,3-butanediol at the other end.

Last article's material, diacetyl, is the other end of this line — diacetyl is the oxidised form, acetoin the reduced one, and bacteria can run it in either direction.

Menéndez and colleagues, studying the ripening of Arzúa-Ulloa cheese in Galicia in the International Journal of Food Microbiology in 2000, used this very pair as their marker:

Two control batches of cheese were made with an acid-aromatic starter containing Lactococcus lactis subsp. lactis and Lactococcus lactis subsp. lactis var. diacetylactis … On day 1 of ripening the highest diacetyl-acetoin content was noted in the control batches.

The organism is literally named diacetylactis — named for doing this. And the researchers treat the two molecules as one measurement, because in cheese they are one pool.

What this means for a formula: if you want "butter", diacetyl is the sharp end and acetoin is the round one. The evaluation fields say the same — diacetyl's description includes "pungent"; acetoin's does not. It reads sweet, buttery, creamy, dairy, milky, fatty.

And the butanediol end is a signal to plants

The same cycle reaches somewhere unexpected. Bhavya and colleagues in Microbiological Research, 2026:

Rhizobacterial volatile organic compounds are increasingly being recognized for their role in plant-microbe interactions … Based on its performance, Bacillus subtilis DLD-8 was selected for detailed analysis. GC-MS profiling detected 36 VOCs under stress and 24 under non-stress …

And the most effective molecule they identified was 2,3-butanediol — acetoin's reduced product in the butanediol cycle. Acting through a MYB transcription factor, it enhances drought tolerance in mung bean.

A volatile a bacterium releases when short of water, which a plant reads as a signal. And that molecule is a neighbour on the same metabolic road as the butter smell.

(That paper has nothing to do with fragrance; I cite it to show what this cycle's products do in nature.)

It is in the 145 refrigerate records

storage: refrigerate in tightly sealed containers. Shelf life 12 months.

I measured those 145: the refrigerate group is about suppressing reaction rate, not volatility. And acetoin is exactly the reactive kind — it dimerises, and it oxidises to diacetyl.

Here the reason for refrigeration is concrete.

How to use it

• Evaluate at 1% or less. Rank 3.
• It is the round end of the butter/dairy line. For sharpness use diacetyl; for body use this.
• 28 hours is longer than its vapour pressure suggests, so it stays around longer than you would expect.
• Refrigerate, and expect crystals. Solid at the bottom of the bottle may be dimer rather than contamination.
• Oral LD50 > 5000 mg/kg (two 1970s sources); acute toxicity is not the issue.
• It is in beer, wine, cheese and coffee — it is the signature of fermentation.

What this doesn't establish

• I have not smelled it. This is fields, whole-database statistics, and two papers.
• "The melting point is the dimer's" is an inference. The database does not label the form, and I did not consult the primary measurement.
• "The 28 hours comes from dimerisation" is more speculative still — I can say the numbers are unusual, not that this is the cause.
• Neither paper is fragrance research, and neither measured this material's behaviour in perfume. Menéndez measured cheese ripening; Bhavya measured drought tolerance in mung bean.
• Bhavya 2026 measured 2,3-butanediol, not acetoin. I cite it to explain the cycle, not because it tested this material.
• I did not check the IFRA text.