Material guide: trans-anethole — same molecular formula as methyl chavicol, the double bond one position over, and only one of them is genotoxic

CAS 4180-23-8, 96 suppliers. The formula C10H12O maps to four records and four CAS numbers in this database: trans-anethole at 96 suppliers, estragole at 47, an unspecified-isomer anethol at 18, and cis-anethole at zero. Same formula, same molecular weight of 148.20, and the only difference is where the double bond sits on the side chain. That one position sends them into entirely different regulatory worlds. The 1999 FEMA Expert Panel assessment took a mechanism-based approach: trans-anethole's hepatotoxicity runs through the metabolite anethole epoxide, and since neither the parent nor the epoxide shows evidence of genotoxicity, the liver tumours seen in high-dose female rats were judged secondary to hepatotoxicity rather than genotoxic. Flavour intake of 54 micrograms per kilogram per day against a female rat NOAEL of 120 milligrams per kilogram per day leaves a margin of at least ten thousand fold. Estragole, one position away, is metabolised to a genotoxic 1'-sulfoxy metabolite. Two practical notes: the melting point is 21 to 23 °C, so it crystallises in the bottle when the room is cool, and the solubility field carries the line avoid iron drums. And the clouding when you add water has its own physics, mapped by a 2024 study of the ouzo ternary system.

About a 5 minute read.

Too long, didn't read

Name: (E)-anethol (trans-anethole), CAS 4180-23-8, FEMA 2086, 96 suppliers
Odor family licorice; descriptors: sweet, anise, licorice, mimosa, medicinal. Strength high, evaluate at 10% or less
Formula C10H12O, which maps to four records and four CAS numbers in this database
Longevity 20 hours, vapour pressure 0.069 mmHg, melting point 21–23 °C (it crystallises in a cool room)
Listed in all four regulatory indexes: JECFA 217, FEMA GRAS 2086, CoE 183, FLAVIS 04.010, and FCC listed
GHS: H317 skin sensitisation Category 1, and it is not among the EU's 26 named allergens
The solubility field carries an operating note: avoid iron drums

One formula, four records, four CAS numbers

Record	CAS	Suppliers	Odor family	Longevity
(E)-anethol (trans)	4180-23-8	96	licorice	20 hours
estragole	140-67-0	47	anisic	8 hours
anethol (isomer unspecified)	104-46-1	18	licorice	(empty)
(Z)-anethol (cis)	25679-28-1	0	anisic	(empty)

All four are C10H12O with a molecular weight of 148.20. The only difference is where the double bond sits.

In trans and cis anethole the double bond lies between the first and second carbons of the side chain, conjugated with the ring, and the isomers differ in whether the substituents point the same way or opposite ways. In estragole the double bond has moved to the second and third carbons, making it a terminal allyl group.

Three things fall out of that table.

First, the cis record has zero suppliers and a completely empty odor field. Nobody sells it and nobody has written down what it smells like. The market's opinion of that isomer is in that zero.

Second, there is a record simply called anethol, CAS 104-46-1, isomer unspecified, with 18 suppliers. If a quotation says only "anethole", that may be what arrives. Ask for (E) or trans, or specify CAS 4180-23-8.

Third, estragole is the subject of the basil oil piece. Same formula, one position over, and the next section is what that position costs.

Move the double bond, and the metabolic route splits

Carry the conclusion over from the basil piece: estragole is metabolised into a genotoxic and carcinogenic 1'-sulfoxy metabolite, which is why it is regulated.

trans-Anethole is different. The safety evaluation published by the FEMA Expert Panel in 1999 (PMID 10496381) took a mechanism-based approach, tracking production of the hepatotoxic metabolite anethole epoxide (AE) and using it to interpret the pathological changes observed across species and sexes of laboratory rodents in chronic and subchronic dietary studies.

The key sentence, verbatim: neither trans-anethole nor AE show any evidence of genotoxicity.

Two things with the same molecular formula ended up in two different regulatory worlds. One position apart, and the metabolism differs.

How the FEMA assessment did the arithmetic

This assessment is worth reading further, because it is a complete demonstration that "liver tumours appeared" does not equal "carcinogenic risk".

Choosing the species and sex. Female Sprague Dawley rats metabolise more trans-anethole to AE than mice or humans do, which makes them the most conservative model for human risk. Their lower activity of the "fast" detoxication enzyme epoxide hydrolase is associated with more pronounced hepatotoxicity than in males.

Where the dose turns. At low exposure, trans-anethole is efficiently detoxicated, primarily by O-demethylation in rodents and omega-oxidation in humans, with epoxidation only a minor pathway. At high doses in rats, particularly females, a metabolic shift occurs, increasing epoxidation and the formation of AE.

The observed threshold. In chronic dietary studies, hepatotoxicity was observed when the estimated daily hepatic production of AE exceeded 30 mg AE/kg body weight. In female rats, chronic hepatotoxicity and a low incidence of liver tumours were reported at a dietary intake of 550 mg trans-anethole/kg body weight/day, under which daily hepatic AE production exceeded 120 mg/kg.

How the conclusion follows. Because neither compound is genotoxic, the authors concluded that the hepatocarcinogenic effect in female rats occurs by a non-genotoxic mechanism and is secondary to hepatotoxicity from continuous exposure to high hepatocellular AE.

The margin. Flavour intake is 54 µg/kg body weight/day. The female rat NOAEL from a two-plus year study is 120 mg/kg body weight/day, which produces AE at 22 mg/kg/day, at least ten thousand times the 0.002 mg/kg/day produced by flavour use.

What to take is not only the conclusion but the shape of the argument: identify the molecule that causes the toxicity, then show that its pathway does not switch on at real human exposures. A complete worked example of the method described in how to read fragrance safety studies.

The clouding when you add water has physics behind it

trans-Anethole's water solubility is only 98.68 mg/L (est), while it dissolves completely in alcohol. Which produces something anyone who has drunk an anise spirit has seen.

In 2024 Archer and colleagues at Loughborough and Edinburgh studied that system (PMID 38748245). Their opening states the mechanism cleanly: ouzo's ingredients are water, alcohol and trans-anethole oil. The oil is insoluble in water but completely soluble in alcohol, so when water is added to the spirit, the available alcohol is depleted and the mixture exhibits spontaneous emulsification. The process is commonly known as the louche or ouzo effect.

What they added was the part of this ternary system nobody had studied: the properties of the coexisting phases. They measured the tie-lines connecting coexisting phases, determined the critical point (the plait point, in a ternary system), and measured surface tension and density across alcohol concentrations, alongside a thermodynamic theory that, with suitable interaction parameters, captures nearly all features of the experiments.

The practical reading is direct: any alcoholic formula containing anethole can go cloudy when water is added. That is not the formula breaking; it is the ternary phase diagram entering a two-phase region. Control the water content to avoid it, or accept the louche as a feature.

It crystallises at 21 °C, and then there is "avoid iron drums"

Two fields that cost people money.

Melting point 21–23 °C. The appearance field reads "white colorless crystals". In a room slightly below normal temperature, what is in the bottle is a solid. A crystallised bottle arriving in winter is normal physics and warms back out, but melt it fully and mix it before weighing, or the proportion you take will be off.

The solubility field has "avoid iron drums" wedged into it. That sentence has nothing to do with solubility; it is a storage warning that landed in the wrong column. The juniper piece covered seams in the notes field, and this is the same phenomenon in a different field: important handling information sitting in the wrong box, where no search will find it.

One more comparison worth making. Its first GHS line is H317 skin sensitisation Category 1, and yet it is not among the EU's 26 named allergens. That article measured the point: a labelling list and a hazard classification are two lists with different purposes, and disagreement between them is normal.

Buying and storing

96 suppliers. Things to settle before buying:

Write the CAS out. For the trans isomer, specify 4180-23-8. "Anethole" alone may get you the unspecified 104-46-1.
The grades field lists USP (technical) and FCC (food grade). The use decides the grade.
Do not decant into iron containers. That line in the database is in the wrong field but it is real.
A crystallised delivery is not a defect. Melting point 21–23 °C; warm it, melt it fully, mix, then weigh.
Strength high, evaluate at 10% or less, and longevity is only 20 hours. It is a loud early middle, not a base.
Shelf life 24 months, sealed and dark. Flash point 90.56 °C.

→ trans-Anethole in the database: full physical data, 96 suppliers and suggested blends.

What this article does not establish

I have not established that the cis isomer has no suppliers because of toxicity. The database shows zero suppliers and an empty odor field, with no reason given.
The FEMA assessment is from 1999, reflecting the evidence and the methods of its time. Current regulation follows the current text.
That assessment addressed dietary intake, not skin contact or inhalation. The H317 line is about something else.
Estragole's genotoxicity comes from the sources in the basil piece, not from either paper cited here. The two do not corroborate each other.
The ouzo paper measured a water–ethanol–trans-anethole ternary system. A real formula contains dozens of other materials, so the phase diagram does not transfer directly.