Material guide: allyl hexanoate — the database's most flavour-tagged material (67 of them), and its GHS says toxic if swallowed, toxic in contact with skin

About a 5 minute read.

The short version

• Name: allyl hexanoate, also allyl caproate, CAS 123-68-2, FEMA 2032
• Odour type: fruity; odour: sweet, fruity, pineapple, tropical, ethereal, rummy, arrack, fatty, cognac, estery, juicy
• Strength: high (evaluate at 10% or less), rank 3; substantivity < 1 hour at 100%
• Suppliers: 82; Cramer Class II
• The most flavour-tagged material in the database: 67 FL tags
• GHS: H301 toxic if swallowed, H311 toxic in contact with skin

Sixty-seven tags

Last time I measured the two vocabularies in the uses field. The material carrying the most flavour tags is this one — 67 of them.

taste_descriptions states the job:

Characteristically used in pineapple flavor, but also can be seen in fruit punch, tropical, apple, orange juice, peach and rum flavors.

This is pineapple. And pineapple gets subdivided finely in the flavour industry — George and colleagues quantified 26 key aroma volatiles across five named Australian pineapple cultivars ('Aus Carnival', 'Aus Festival', 'Aus Jubilee', 'Aus Smooth' and others) by stable isotope dilution with HS-SPME-GCMS in Food Chemistry in 2024.

Those 67 tags are the same granularity in a different form.

And it is on the fragrance side too

The notes field joins the two worlds:

Allyl caproate is employed principally in the formulation of pineapple flavors but it can also be used for peach and apricot essences and for apple blossom, peach blossom, and wisteria perfume compositions. Allyl caproate is an ingredient of some lipstick perfumes. Adds a sweet juicy note to citrus flavors.

Apple blossom.

The divide I measured last time runs straight through this material — the flavour side makes pineapple with it, the fragrance side makes apple blossom. (odor_descriptions agrees: "A good fruity topnote for citrus complexes, particularly orange.")

But the GHS is unpleasant reading

ghs_hazard_statements:

H227 — combustible liquid H301 — toxic if swallowed H311 — toxic in contact with skin H401 / H411 — toxic to aquatic life, with long lasting effects

And the toxicity fields:

	Value
Oral (rat)	LD50 218 mg/kg (1964)
Oral (guinea pig)	280 mg/kg
Dermal (rabbit)	LD50 300 mg/kg (1973)

An oral LD50 in the low hundreds is low for a FEMA GRAS flavouring. (For contrast, most aroma materials' rabbit dermal LD50 reads "> 5000 mg/kg"; I have written about isovaleric acid converting to 287 and allyl isothiocyanate at 88. This one sits in that low band.)

Both 1964 entries carry an effect note: "behavioral: somnolence (general depressed activity)."

Why

Because allyl esters hydrolyse to allyl alcohol.

And allyl alcohol's hepatotoxicity has a clear mechanism. What matters in Karas and Chakrabarti's 2001 rat study, in the Journal of Environmental Pathology, Toxicology and Oncology, is the middle step:

Treatment of rats with 4-methylpyrazole (an inhibitor of liver alcohol dehydrogenase) for 30 minutes, followed by similar cotreatment with allyl alcohol and caffeine, completely prevented the elevation of plasma ALT and histological damage.

Block the enzyme and the liver injury does not happen.

So allyl alcohol is not itself the poison — it is the precursor. Alcohol dehydrogenase oxidises it to acrolein, and acrolein is what damages the periportal hepatocytes.

(That paper's actual subject is that caffeine potentiates allyl alcohol's hepatotoxicity, and phenobarbital pretreatment potentiates it further. I cite it for the enzyme-dependence, not the caffeine part.)

FEMA GRAS with an LD50 of 218 mg/kg is not a contradiction. As I wrote in the Cramer classification article, this system manages exposure. Flavour use levels sit on a ppm scale, orders of magnitude away from 218 mg/kg.

Substantivity under one hour

odor_substantivity: < 1 hour at 100.00%

Among the shortest in this series. (Compare: birch tar 400 hours, buchu mercaptan over 120.)

It is a pure top note. A boiling point of 185 °C is not low, but it leaves fast — which is exactly what a fruity opening wants. Pineapple on uncapping, then it gets out of the way.

Note that it reads "< 1 hour", not "1 hour" — that is a ceiling, running the opposite direction from the 400-hour floors I wrote about.

The occurrence field lists mushroom and baked potato

occurrence: mushroom, pineapple fruit, baked potato.

Three sources, two of them nothing to do with pineapple. This field sweeps up everything anything has been detected in — the same point I made about isovaleric acid (dolphin fat) and methyl mercaptan (orange juice next to wild strawberry).

How to use it

• Evaluate at 10% or less; it is gone within the hour. This is the opening, not the frame.
• It leads pineapple, and reaches toward peach, apricot, rum and citrus tops.
• In perfume it builds apple blossom and peach blossom. The notes field says so directly.
• Mind H311: toxic in contact with skin. Gloves; do not weigh it bare-handed.
• Toxic to aquatic life with lasting effects (H411) — the sink is not a disposal route.
• 12 months, cool and dry.

What this doesn't establish

• I have not smelled it. This is fields, whole-database statistics, and two papers.
• Karas 2001 measured allyl alcohol, not allyl hexanoate. "The ester hydrolyses to allyl alcohol" is the standard account for this class; I found no study measuring this ester's metabolism directly.
• That paper's subject is a caffeine interaction, and I cite only the enzyme inhibition step.
• George 2024 did not measure this material. I cite it to show that cultivar-level pineapple analysis has measured backing.
• I did not go to the sources for the 1964 and 1973 toxicity figures.
• "FEMA GRAS and a low LD50 are not in conflict" rests on the general logic of the TTC procedure, not on this material's specific evaluation documents.
• I did not check the IFRA text.