Material guide: guaiacol — the flavour field literally says bacon, and it appears in five corpus formulas

Short version: a material strong enough that one drop ruins a formula, and one that barely appears on any perfumer's working list. Its real stage is food and drink, and the way it reaches you is usually that something has been burnt or roasted.

The fields

Field	Value
Database name	ortho-guaiacol (o-methoxyphenol)
CAS	90-05-1
Formula	C7H8O2, MW 124.14
Melting point	28–32 °C
Boiling point	205–206 °C @ 760 mmHg
Appearance	Colorless to pale yellow clear oily liquid to solid (est)
Odour type	Phenolic
Odour	Phenolic, smoky, spicy, vanilla, woody, medicinal, savory, meaty
Flavour	Woody, phenolic, bacon, savory, smoky, medicinal
Strength	High; evaluation recommended at 1% solution or less (strength rank 3)
Substantivity	280 hours at 100%
logP	1.30 (est)
Water solubility	18,700 mg/L @ 15 °C (exp)
Suppliers	61
Regulatory	JECFA 713, FEMA GRAS 2532, CoE 173, FLAVIS 04.005
GHS	H302 harmful if swallowed, H315 skin irritation, H319 serious eye irritation, H402/H412 aquatic
Storage	Cool, dry, sealed, protected from light, under nitrogen

The joke was right

Someone answered the question of how to add sportiness to a perfume with a run of jokes:

"It depends on the sport. For skydiving add floralozone. For ice skating or hockey add menthol. For chess don't add anything because it is not a real sport (but is great for the brain). For ballet add rose. For parkour add broken bones. For hunting add guaiacol (it smells like bacon)." (Thanospapa12345)

Checked against the database, that is the most accurate line in the thread.

Its flavour field — it is also a flavour material, FEMA GRAS 2532 — reads, in so many words:

woody; phenolic; bacon; savory; smoky; medicinal

The taste description field cites Mosciano in Perfumer & Flavorist 1997, at 2.00 ppm: "Woody, phenolic, bacon, savory, smoky and medicinal."

This is not a coincidental metaphor. A large part of the taste of bacon comes from smoke, and a large part of the taste of smoke comes from volatile phenols of this kind. The joke landed on the actual chemistry.

Melting at 28–32 °C: another material that changes state

The last beginner article in this series was about not being able to look up whether a material is a solid. Guaiacol is the textbook case, and its own database record sits on both sides of the fence.

The appearance field reads: "colorless to pale yellow clear oily liquid to solid (est)" — liquid and solid in the same field.

The melting point field gives the reason: 28–32 °C.

A summer room can sit above 30 °C and a winter one below 20 °C. The same bottle of guaiacol is an oil in July and a lump in January. That is not spoilage.

The previous round's list of materials that change state with room temperature had diphenyl oxide (27 °C) and dimethyl benzyl carbinyl acetate (30 °C). Guaiacol is the third, and its range straddles the most awkward part of the scale.

The corpus barely contains it

To be clear before the numbers: the sample is too small, so what follows is an observation and not a conclusion.

Across 954 published formulas, the guaiacol family appears in 5 (0.5%):

Spelling	Formulas
guaiacol	4
4-methyl guaiacol	2
4-ethyl guaiacol	1

(Some formulas use more than one.)

Across those 5, the median share is 4.10%, interquartile 2.99%–4.36%, maximum 11.61%.

A material marked "high strength, evaluate at 1% or less" sits at a median of 4% in the formulas that use it. The contrast is striking, and I am not going to derive anything from five formulas. Several explanations fit: these may be flavour formulas rather than perfumes; a dilution's strength may have been recorded as if it were the neat material; or those particular formulas may simply run heavy. Five cannot separate them.

One thing can be said: it is not on perfumers' working lists. Sixty-one suppliers, a FEMA number, a JECFA number, and almost no presence in published formulas. Its market is somewhere else.

Its market is food, and it is usually locked up

The natural occurrence field is unusual. An extract:

asparagus shoot; beer; boronia absolute @ 0.02%; butter; cassia bark; celery leaf; celery oil; celery seed; parmesan cheese; cinnamon; cocoa; coffee bean; roasted coffee; corn seed; fig leaf; roasted filbert; guaiacwood; ketaki flower oil india @ 0.30%; milk; white mustard; passion fruit; roasted peanut; bell pepper; peppermint leaf; popcorn; rice…

There is a pattern here: roasting, baking, fermenting, smoking. Coffee, cocoa, popcorn, roasted peanut, roasted filbert, beer, cheese. Guaiacol is one of the thermal decomposition products of lignin, the phenolic polymer in plant cell walls, so it can appear in any plant material that has been burnt or roasted.

And two papers describe its second property in nature: it is often not free.

One: smoke taint in wine (2014)

Mayr's group at the Australian Wine Research Institute studied smoke taint — wine made from vineyards exposed to bushfire smoke. They note that guaiacol, 4-methylguaiacol, syringol, 4-methylsyringol, and o-, m- and p-cresol, as well as their glycoconjugates, are present at elevated concentrations in smoke-tainted wine.

Using sensory descriptive analysis, they added free volatile phenols together with their glycosidically bound forms to red wines to mimic smoke taint. The result: adding the volatile phenols together with the glycoconjugates gave the strongest off-flavour.

They then chased the mechanism with in vitro and in vivo experiments, and concluded that enzymes present in human saliva can release the volatile aglycones from their glycoconjugates, even under low pH and elevated ethanol conditions.

So a glass can smell barely smoky and then be unlocked enzymatically in the mouth after you drink it. The authors suggest this mechanism may play a more general role in the flavour and aftertaste of wine.

Two: smoky aroma in tomato (2013)

The same phenomenon is worked out more completely in tomato. Tikunov's group studied the key tomato fruit aroma attribute termed "smoky", which phenylpropanoid volatiles are responsible for.

Their central finding: release of these volatiles from their glycosylated precursors, rather than their biosynthesis, is the major determinant of smoky aroma in cultivated tomato.

Using a combinatorial omics approach they identified the non-smoky glycosyltransferase1 (NSGT1) gene. Its expression is induced during fruit ripening, and the enzyme converts the cleavable diglycosides of the smoky-related volatiles into noncleavable triglycosides, preventing their deglycosylation and release when the tissue is disrupted — that is, when you bite it.

In an nsgt1/nsgt1 background that step does not happen, the diglycosides can be cleaved, and the volatiles are released. Reverse genetics showed NSGT1-mediated glycosylation to be the molecular mechanism behind the major quantitative trait locus for smoky aroma, and sensory trials on transgenic fruit in which functional NSGT1 was restored showed a significant and perceivable reduction in smoky aroma.

Taken together: molecules like guaiacol frequently exist in plants as glycosides, where they cannot be smelled. An enzyme — the plant's, a microbe's, or your own saliva's — has to remove the sugar before they appear.

There is one direct implication for perfumery. What you use is neat guaiacol, the unlocked form. Its abundance in a natural material does not mean you can smell that much of it; conversely, the amount you put in a formula is fully active.

Whether to use it

• Strength rank 3, evaluation recommended at 1% or less, which puts it among the highest of the 83 materials this series has covered. Its class is indole and dimethyl sulfide: get the amount wrong and the formula is gone.
• Substantivity 280 hours. It does not leave early. A mistake stays.
• It carries GHS skin irritation (H315) and serious eye irritation (H319). Not unusual for a fragrance material, but combined with "high strength, handle at 1% or less", gloves are a reasonable call.
• The oral toxicity field carries a human datum: oral-human LDLo 43 mg/kg, sourced to Deichmann's Toxicology of Drugs and Chemicals, 1969. Stated carefully: LDLo is the lowest known lethal dose, which is a case report rather than an experiment, and perfume is not ingested. It is listed here because it is in the record, not because it bears on topical use.
• Store under nitrogen. The storage field adds that specifically; phenols oxidise.
• If you want smoke, leather, whisky or roasted directions, this is the core molecule. Start at the 0.01% level.

What this doesn't establish

• The corpus gives five formulas, and I draw nothing from the 4.10% median. They may be flavour formulas, they may record a dilution as if neat, or those formulas may simply run heavy. Five cannot distinguish.
• Neither paper is perfumery research. One studies smoke taint in wine, the other tomato breeding. What is cited is the mechanism they establish — glycoside-bound forms and enzymatic release — not any conclusion about perfume.
• Mayr 2014's release mechanism is in-mouth. It explains flavour after swallowing and does not transfer to perfume on skin.
• Tikunov 2013 measured tomato. NSGT1 is a tomato gene, and the mechanism need not be the same in other plants.
• "Bacon" is the flavour field's description, under taste conditions at 2 ppm. The odour field reads phenolic, smoky, meaty, which is not identical. The joke landed, but what it landed on was a taste description.
• The oral-human LDLo of 43 mg/kg is 1969 case data from a book; I did not trace it to its original source and would not derive any usage conclusion from it.
• The 28–32 °C melting point is the database's value. That field mixes measured and estimated entries, and I did not verify this one separately.