Material guide: 2-acetyl pyrazine, evaluate at 0.10% or less, and most people smell it neat first

In short: the point of this material is written in its strength field. Most materials in this library read "high"; this one reads "very high", with a recommended evaluation strength of 0.10% or less. And it is a solid, not a liquid.

Fields

Field	Value
CAS	22047-25-2
First synonym	1-pyrazin-2-ylethanone
Formula	C6H6N2O, MW 122.127
Database category	flavor and fragrance agents
Appearance	white to pale yellow crystalline powder (est)
Assay	99.00 to 100.00
Melting point	76-78 C
Boiling point	188-190 C @ 760 mmHg
Flash point	> 100 C
logP	0.20 (est)
Vapour pressure	0.188 mmHg @ 25 C
Water solubility	66,520 mg/L @ 25 C (measured)
Solubility	alcohol; dipropylene glycol
Odour type	popcorn
Odour	popcorn; nutty; corn chip; bread crust; chocolate; hazelnut; coffee; musty; roasted; potato
Flavour	roasted; nutty; bready; yeasty; popcorn; corn chip
Strength	very high, recommend smelling in a 0.10% solution or less
Substantivity	400 hours (10% in dipropylene glycol)
Shelf life	24 months
Storage	cool, dry, tightly sealed, protected from heat and light
Occurrence	roasted almond; bread; cocoa; coffee; filbert; peanut; popcorn; potato chip; sesame oil
Suppliers	80
Regulatory	JECFA 784, FEMA 3126, CoE 2286, FLAVIS 14.032
Oral toxicity	rat gavage LD50 > 3000 mg/kg
GHS	H315 skin irritation; H319 serious eye irritation; H335 may cause respiratory irritation
Note	component of roasted sesame seed aroma

That "very high"

The strength field reads, literally: very high, recommend smelling in a 0.10% solution or less.

Most strong materials in this library read "high". "Very high" is a minority wording, and it comes with the recommended strength pushed down to 0.10%.

Setting them side by side helps:

Material	Strength field	Evaluate at
2-acetyl pyrazine	very high	0.10%
2,3,5-trimethyl pyrazine	high	0.01%
2-acetyl pyridine	high	1.00%
para-cresol	high	0.10%
rose oxide	high	1.00%
amber naphthofuran	high	10.00%

Two pyrazines, one at 0.10% and one at 0.01%, a factor of ten apart. There is no family rule in this field; check each material.

Somebody on the forum who had just received this material described it as smelling like feet, like a tar playground for children, with nothing gourmand about it. They later admitted they had not calculated any dilution and had smelled it neat the day it arrived. Neat is a thousand times the recommended strength, and the odour field already carries musty and roasted, which at that multiple bury everything else.

It is a solid

The appearance field says white to pale yellow crystalline powder, melting at 76 to 78 C.

Worth knowing before buying, because it decides whether you make a solution first. Its near relative 2-acetyl pyridine is a liquid, and both share the odour type popcorn.

The good news is it dissolves easily. Water solubility of 66,520 mg/L is a measured value, logP is only 0.20, and the solubility field lists alcohol and dipropylene glycol. A solid melting at 76 C is usually hard to dissolve (the arithmetic is in the maltol article), but this one is polar enough not to be a problem.

(A logP of 0.20 flags something else: it is hydrophilic, and will not stay in an oil phase in a system containing water.)

Everything in the occurrence field has been heated

Roasted almond, bread, cocoa, coffee, filbert, peanut, popcorn, potato chip, sesame oil.

Nothing in that list is raw. This is not extracted from a plant; it is made by heating one.

Li and colleagues heated 18 amino acids at 160 C for an hour in Food Research International in 2025 and tracked volatile formation. One result is counter-intuitive: even with no reducing sugar present, amino acids with aliphatic hydroxyl side chains (serine, threonine) and with amide groups (glutamine, asparagine) generated pyrazines.

Adding a sugar mix simulating tea leaf composition shifted the volatiles toward furans, aldehydes and acids, modulating pyrazine formation.

"Sugar plus amino acid equals pyrazine" is too coarse. Which amino acid, whether sugar is present, and which sugar all change the result.

(That paper's matrix was based on tea, with applicability stated by the authors.)

Structure and odour: this family has models

Wailzer and colleagues applied a neural network to pyrazines in J Med Chem in 2001. Ninety-eight compounds, split into 32 green, 43 bell-pepper and 23 nutty, classified from structural descriptors alone at 93.7% accuracy. They also predicted thresholds, with a Pearson R of 0.926 on the bell-pepper test set.

Within the pyrazine family, structure predicts both odour class and threshold.

The composition of those 98 needs noting. Nutty is only 23% of them, with bell-pepper at 43%. In this database's 236 pyrazine records, among those with an odour type, nutty is the most common. The two samples differ in composition because Wailzer's was selected for a bell-pepper study.

Corpus

Across 954 published formulas, pyrazines appear in 6, or 0.6%, and 2-acetyl pyrazine appears twice, one of those written as 2-acetyl pyrazine 0.1%.

Across the pyrazine class, 73% of uses carry a dilution tag.

Two formulas is too few to compute a dose.

What this doesn't establish

• I have not smelled it. This is database fields, corpus and two papers.
• On "very high is uncommon", I did not count occurrences of that wording across the library. I compared against the other materials I have written up.
• The substantivity figures, 400 hours (10% in dipropylene glycol) against acetyl pyridine's 9 hours (neat), are not comparable. The same problem appears in the rose oxide article.
• Water solubility is measured and logP is estimated, so mixing them needs care.
• "A solid melting at 76 C is usually hard to dissolve" is a general rule, and this material is an exception because of its polarity. I have no measured solubility for it in ethanol.
• Wailzer 2001 works inside the pyrazine family and does not extrapolate to other ring systems.
• Li 2025's matrix was based on tea, with applicability stated by the authors.
• The occurrence field gives no content percentage for any item, so I do not know which food carries the most.
• Two corpus formulas cannot support anything about dosing.