Material guide: phenylacetaldehyde — a three-month shelf life, and it caught moths in a field trap on its own

CAS 122-78-1. Thirty-one suppliers, 57 of 954 formulas. Its shelf life is the shortest this series has met: three months, where most materials carry twenty-four. It spans nine formula genres, but florals take 66.1% of its appearances, against the corpus baseline of 35.4% — one of the most concentrated specialists in the set. A 2005 study identified 44 and 45 volatiles in two Gymnadenia orchids and found only 7 and 3 physiologically active; field trapping with phenylacetaldehyde alone caught moths carrying that orchid's own pollinia.

One number first, because it is an outlier in this series.

Shelf life: 3 months.

Most materials covered here start at 24 months — phenethyl alcohol, veramoss and isoeugenol all do. The shorter ones, hexyl cinnamal and dimethyl sulfide, carry 12.

This one carries three.

The short version

A 3-month shelf life, the shortest this series has met. It is an aldehyde, it oxidises, and it also polymerises on its own.
Thirty-one suppliers, 57 of 954 formulas, median dose 1.00%.
It is one of the most concentrated specialists in the corpus. Nine genres, but florals take 66.1% of its appearances — against a corpus floral baseline of 35.4%, that is 1.87 times the baseline.
It is the oxidation state of phenethyl alcohol, and the step before it on a rose's biosynthetic route. In the corpus its co-occurrence lift with phenethyl alcohol is 2.57, second highest on that table.
A 2005 orchid study put it alone into field traps and caught moths carrying that orchid's own pollinia. The same study found the orchid significantly increases its emission at night.
The reason it is everywhere in food is chemical: it is the Strecker aldehyde of phenylalanine.

About 8 minutes.

What three months means

Aldehydes oxidise to acids, which this series has covered. Phenylacetaldehyde has a second route as well: it polymerises.

Neat phenylacetaldehyde thickens over time into resinous oligomers, and that happens at room temperature. Which is why suppliers often sell its acetal or a dilution, and why the database gives it three months.

In practice:

Buy small. You will not use 100 g in three months at a median dose of 1.00%.
Decant on arrival, blanket with inert gas or at least fill to the neck. Everything in the storage and oxidation piece applies here twice over.
Re-smell an old bottle before using it. Six-month-old phenylacetaldehyde is not the material you bought.

Flash point 87 °C, far gentler than dimethyl sulfide's −36.7 °C, and still below most of this series.

It is a specialist, for a good reason

We just measured genre breadth across the 954 formulas: sorting formulas into fourteen genres by title and counting how many each material reaches.

	phenylacetaldehyde	for comparison: Hedione
Formulas	57	165
Genres	9/14	14/14
Largest genre	66.1% floral	35.4% floral
Corpus baseline	35.4%	35.4%

Its floral share is 1.87 times the baseline.

And the concentration has a specific cause: it is hyacinth.

The database gives its odour as green, sweet, floral, hyacinth, clover, honey, cocoa, rose, powdery, fermented, chocolate, earthy. One of the collected descriptions puts it more bluntly: "very strong, green, floral, sweet, hyacinth type."

A material that is the core of one flower will naturally only appear in that flower's formulas. That is not a defect — general-purpose materials decide how usable your palette is, and specialists decide what you can make.

One step from phenethyl alcohol

Writing up phenethyl alcohol turned up something: across the 274 formulas using it, the second-highest co-occurrence lift belonged to phenylacetaldehyde (2.57, 42 shared formulas).

Because they are one oxidation state apart.

Phenylacetaldehyde (C₈H₈O, aldehyde) → reduction → phenethyl alcohol (C₈H₁₀O, alcohol)

And that step is the last one on a rose's route to phenethyl alcohol. The pathway Hirata and colleagues traced with deuterium labels in 2016 has this as its intermediate.

So perfumers put two molecules adjacent on a biosynthetic pathway into the same formula, and mostly not for that reason. They put them there because it smells right.

Their doses differ sharply: phenethyl alcohol at a median of 10.00%, phenylacetaldehyde at 1.00%. Tenfold. An aldehyde is far stronger than its corresponding alcohol, a pattern that keeps recurring in this series.

That orchid experiment

There is a fine study of what this material does in nature.

Huber and colleagues compared two closely related orchids in Oecologia in 2005 — Gymnadenia conopsea and G. odoratissima — across four populations, by day and by night, for both scent composition and pollinators.

Their method is thorough: collecting actual pollinators with hand nets, sampling floral odour by headspace sorption, identifying compounds by GC-MS, using electroantennographic detection to find which compounds are physiologically active in the pollinators, and then taking those active compounds into the field for trapping experiments.

The first result is worth keeping:

Orchid	Volatiles identified	Physiologically active
G. conopsea	45	3 (benzyl acetate, eugenol, benzyl benzoate)
G. odoratissima	44	7 (benzaldehyde, phenylacetaldehyde, benzyl acetate, phenylethyl acetate, eugenol and others)

Out of forty-odd volatiles, only three to seven produce a response in the pollinator's antenna.

Most of what a flower emits does nothing to what it is trying to attract. The same class of fact as 121 compounds in a jasmine absolute: quantity and effect are different things.

Then the trapping.

In field bioassays using a mixture of the active G. odoratissima compounds and phenylacetaldehyde alone, they caught a total of 25 moths, some of which carried Gymnadenia pollinia.

One molecule, in a trap in a field, caught moths carrying that orchid's pollen.

The reverse control was run too: a blend of the active G. conopsea volatiles placed in the G. odoratissima population attracted no pollinators at all.

And one more detail: the two orchids emitted different bouquets by day and by night, with G. odoratissima showing the greater difference — phenylacetaldehyde rose significantly at night.

The orchid turns this molecule up when the moths come out.

Why it is everywhere in food

The occurrence field runs long: cooked apple, apricot, asparagus, black bean stem oil, bilberry, blackberry, white bread, burdock root, cabbage, celery leaf, cherry, cocoa…

The reason is chemical: it is the Strecker aldehyde of phenylalanine.

Strecker degradation is the reaction of an amino acid with a dicarbonyl compound (or other oxidation product), losing a carbon and producing an aldehyde. Phenylalanine down that route gives phenylacetaldehyde.

Hidalgo and colleagues measured one specific version in the Journal of Agricultural and Food Chemistry in 2005: they reacted 4-hydroxy-2-nonenal, an oxidative stress product, with phenylalanine in acetonitrile-water, to see whether 4-hydroxy-2-alkenals degrade amino acids the way 4,5-epoxy-2-alkenals do.

The result: the hydroxyalkenal also degraded phenylalanine to phenylacetaldehyde, with a reaction yield of 17%.

So wherever there is phenylalanine, lipid oxidation products and heat, this molecule appears. Bread, cocoa, cooked apple, cheese — all the same reaction in different settings.

Which also joins up with the phenethyl alcohol piece: both molecules start from phenylalanine. A rose gets there with enzymes and an oven gets there with heat.

Physical properties

Property	Value
CAS	122-78-1
Formula	C₈H₈O, MW 120.15
Appearance	colourless to pale yellow clear oily liquid (est)
Boiling point	193-195 °C
Melting point	−10 °C
Flash point	87 °C
logP	1.80 (est)
Water solubility	3,026 mg/L at 25 °C (est)
Shelf life	3 months or longer
Substantivity	400 hours at 100%
Strength	high, recommend smelling at 10% or less
Suppliers	31
Regulatory listings	JECFA, FEMA GRAS, CoE, FLAVIS
GHS	H227, H302, H315, H319, H335

Putting 400 hours of substantivity next to a 3-month shelf life is instructive. The first says how long it lasts on a blotter and the second how long it lasts in the bottle. Those are different questions, and this material opens the gap between them as wide as it goes.

What this doesn't establish

I have not smelled it. The whole piece sets database fields, the formula corpus and two papers against each other.

"It polymerises" is my general statement about this class of aldehyde, not something from the database or the cited papers. The database gives only the 3-month figure, without a reason. I have no source cited for that mechanism.

The classification method and its limits are in the workhorse piece. Genres come from keyword matching on titles, which misfires, and 21.2% of formulas get no label.

"It is hyacinth" is inferred from the odour field. That field lists twelve descriptors and hyacinth is one; I pull it out because a collected description also says "hyacinth type", and that remains an editorial judgement rather than a measurement.

Huber's study covers four Swiss orchid populations of two species. Those 25 moths are the total for that field experiment, a small sample, and they do not report how many came to phenylacetaldehyde alone versus the mixture. I cite what the paper says.

"Only three to seven active compounds" is an electroantennographic result. EAD measures an electrical response in the antenna, not behaviour. A compound without an EAD response is not thereby meaningless to that insect.

Hidalgo's paper is a model system — acetonitrile and water, 37 to 80 °C. That is neither bread nor cocoa but a controlled reaction test. The 17% yield belongs to that system.

"An oven gets there with heat" is my statement of background, not that paper's conclusion. It measured one specific oxidation product reacting with phenylalanine.

The 3-month shelf life is the database's record, and it reads "3 months or longer if stored properly". That is a floor rather than an expiry date.

References

F. K. Huber, R. Kaiser, W. Sauter, F. P. Schiestl, Floral scent emission and pollinator attraction in two species of Gymnadenia (Orchidaceae), Oecologia, 142(4), 564-575 (2005). PMID 15586296. doi:10.1007/s00442-004-1750-9

F. J. Hidalgo, E. Gallardo, R. Zamora, Strecker type degradation of phenylalanine by 4-hydroxy-2-nonenal in model systems, Journal of Agricultural and Food Chemistry, 53(26), 10254-10259 (2005). PMID 16366724. doi:10.1021/jf052240+