Material guide: Tonalide (AHTN) — the most discussed polycyclic musk, and not the most used one

Short version: a musk that melts at 53 °C, carries an uncommon double aquatic hazard classification, and is known on forums for being hard to dissolve. It has a lot of environmental data, and the corpus says perfumers use it far less than galaxolide. Having read the data, I will not call it worse and I will not call it fine. Their problems are not in the same box.

The fields

Field	Value
Database name	musk tetralin
CAS	21145-77-7
Trade names	Tonalid, Tonalide, Fixolide, Ganolid, Muscofix, AHTN
Formula	C18H26O, MW 258.40
Melting point	53–54 °C
Boiling point	393 °C @ 760 mmHg
Appearance	White crystalline solid (est)
logP	5.30 (est)
Water solubility	0.2879 mg/L @ 25 °C (est)
Odour type	Musk
Odour	Sweet, amber, ambrette, fruity, musk, powdery
Strength	Medium (evaluation recommended at 10% solution or less)
Substantivity	> 400 hours (10% in DPG)
Natural occurrence	Not found in nature
Shelf life	36 months
Suppliers	32
GHS	Acute toxicity, oral (Cat 4) H302; acute aquatic toxicity (Cat 1) H400; chronic aquatic toxicity (Cat 1) H410
Oral toxicity	Rat LD50 964 mg/kg
Dermal toxicity	Rabbit LD50 7940 mg/kg

The substantivity field wants reading by this series' rule: it says > 400 hours, which is a lower bound, not a measurement. Nobody reports how long it actually lasts; only that at 400 hours it is still there.

Why it comes in chunks

The answer to "big chunks need to be crushed into powder and then stirrrrrr until you get bigger biceps then Arnold" is in the melting point field.

53–54 °C. At room temperature it is a crystalline solid, and the kind that sets into large lumps.

This round's beginner article covers the handling. The key is that what is slow is the dissolution rate, not the solubility: logP 5.30 makes this a strongly lipophilic molecule that dissolves well in ethanol. Someone waiting two days at 10% was dealing with rate, not a ceiling. Gentle heat plus time solves it; crushing and stirring do less.

The 53 °C figure also explains why a warm water bath works. You are not melting it — you are warming the solvent, and the rate climbs.

One molecule, at least five names

The database's solubility field has a supplier list crammed into it, which lays this out:

Name it is sold under	By whom
Tonalid™ / Tonalide	Berjé, Associate Allied Chemicals, BOC Sciences
Fixolide	Augustus Oils
Ganolid	Agan Aroma and Fine Chemicals
Muscofix	ACS International
AHTN	Ernesto Ventós

Five trade names for CAS 21145-77-7. And the database files it under a sixth name: musk tetralin.

This is the concrete case of what this round's beginner article measured — 41 of the 120 most-used materials cannot be looked up because formulas use trade names and the database uses chemical names. Search this site for "Tonalide" and you get nothing. The rule holds: check the CAS, not the name.

Its spellings in the corpus are scattered too: tonalid (13 formulas), fixolide (musk tetralin) (1), acetyl tetralin (1), and two written out as full chemical names. Counting its usage means merging those first.

The corpus numbers don't match the volume of discussion

Pulling every musk name out of the corpus gives 42 distinct spellings, appearing in 377 formulas (39.5%).

Merging spellings of the same molecule:

	Formulas	Median dose
Galaxolide HHCB (all spellings)	110	10.00%
AHTN (all spellings)	17	3.77%
Musk ketone	84	4.00%
Ethylene brassylate	59	9.57%

Galaxolide appears 6.5 times as often, at nearly three times the median dose. Only 3 formulas use both.

That gap is worth pausing on. In the environmental argument about polycyclic musks, AHTN and HHCB are almost always named together, and AHTN often stands in as the representative because its GHS classification is the eye-catching one. In actual formulas it is a supporting player.

Its companion list confirms it holds no territory of its own: linalool (12 formulas), hedione (10), benzyl acetate (9), Vertofix (7), dihydromyrcenol (7). That is a general-purpose skeleton, not the signature of a genre.

The environmental data: three papers, and they don't line up

This section needs care, because this material is easy to write up as the bad one, and the data is not that clean.

One: it accumulates in people (1996).

Rimkus and Wolf measured 14 human adipose tissue samples and 5 human milk samples by GC/EI/MS in SIM mode:

Compound	Residue level
HHCB (galaxolide)	16–189 µg/kg fat
AHTN	8–58 µg/kg fat

ADBI, AHDI and ATII were detected at lower levels in a few samples. The contamination route the authors discuss is dermal sorption of these lipophilic compounds from cosmetics and detergents.

Two things to note. AHTN's levels run about a third of HHCB's, which points the same way as the corpus usage ratio. And this is 1996 data; EU regulation and the market have both changed since, so it cannot be read as the current situation.

Two: it interferes with mussel efflux systems (2004).

Luckenbach, Corsi and Epel tested nitro musks (musk ketone, musk xylene) and polycyclic musks (Galaxolide HHCB, Celestolide ADBI, Tetralide AHTN, Traseolide AITI) against multixenobiotic resistance (mxr) transporters in the gill tissue of the marine mussel Mytilus californianus.

The result: all tested musks inhibited transport activity in the low micromolar range, at an effective concentration range similar to quinidine and about a hundred times higher than verapamil. Musk ketone and musk xylene also inhibited efflux of dye from pre-loaded tissue.

The authors' inference is that beyond accumulating themselves, these compounds may act as chemosensitizers, letting other toxic substrates accumulate in cells.

This paper does not score AHTN separately from the other musks. Its conclusion is that all of them do it.

Three: but on zebrafish thyroid, AHTN did better (2023).

The newest and most detailed of the three. Chae's group used embryo-larval zebrafish to test three frequently used synthetic musks: musk ketone (MK), HHCB and AHTN. Experimental concentrations for HHCB and AHTN were chosen to include the maximum levels reported in ambient water.

After five days of exposure:

• Musk ketone or HHCB significantly decreased larval T4 concentration, at levels as low as 0.13 µg/L.
• AHTN exposure up-regulated crhβ, nis, ugt1ab and dio2 but did not alter T4 level, which the paper reads as "suggesting its lesser thyroid disrupting potential".
• All three caused hypoactivity in the larval fish. Genes related to neurogenesis or development such as mbp and syn2a were down-regulated, though the transcriptional patterns differed between the three.

The authors conclude that it requires attention that HHCB and AHTN could influence thyroid hormone or behaviour at levels close to those observed in the environment.

Taken together the picture is this: on human accumulation AHTN sits below HHCB; on mussel efflux inhibition both are named together; on zebrafish thyroid AHTN comes out better than HHCB; on behaviour both have effects.

This series has previously written about someone looking for an environmentally friendly, non-endocrine-disrupting musk. The finding there was that of 36 musk records with ten or more suppliers, 23 have no GHS field at all. AHTN's problem is not that its data looks bad. It is that it has data. Macrocyclic musks look clean partly because nobody has measured them.

So what about those two GHS lines

H400 and H410 (acute and chronic aquatic toxicity, Category 1) are genuinely uncommon for a fragrance material. Combined with logP 5.30 and a water solubility of 0.2879 mg/L, they all say the same thing: it does not dissolve in water, it partitions into fat and sediment, and it does not break down easily.

What that means for someone blending at home, stated precisely:

• This is not a skin safety warning. Dermal toxicity is rabbit LD50 > 7940 mg/kg, which is low. The H302 (harmful if swallowed) is not a contact route either.
• It is a discharge warning. The GHS precautionary statements include P273 "avoid release to the environment" and P391 "collect spillage". Rinsing a beaker into the sink is that route.
• But one person's quantity is negligible, and the scale of this problem is downstream. Polycyclic musks reach the environment mainly through heavy use of detergents and personal care products discharged via wastewater treatment, not through your 5 g.

I am not going to write personal use up as the crux of an environmental responsibility; that would not be honest. What is worth knowing here is that if you are making something to sell, or perfuming a cleaning product, those two classifications follow your product.

Whether to buy it

• If you want a musk that is cheap, diffusive and lasts past 400 hours, this delivers, and 32 suppliers make it easy to get.
• But the corpus says most perfumers reach for galaxolide (6.5×). If you want to connect with published formulas, that is the common one.
• You will be handling a solid. Warm water bath and time, not crushing and stirring.
• If you are making something to sell, H400/H410 affect labelling and disposal, so check first.
• If you are replacing it for environmental reasons, switching to galaxolide is not an improvement on either the 1996 human residue data or the 2023 thyroid results. This series' position is unchanged: macrocyclic musks look cleaner partly because there is less data on them.

What this doesn't establish

• Rimkus 1996 has a small sample (14 adipose, 5 milk) and is thirty years old. EU regulation and the market have changed since; it cannot be read as current.
• Luckenbach 2004 does not score AHTN separately. Its conclusion is that all tested musks inhibited transport, from which nothing follows about AHTN being better or worse than the rest.
• Chae 2023's "lesser thyroid disrupting potential" is relative to the musk ketone and HHCB tested alongside it, not an absence of harm. The same paper measured AHTN causing hypoactivity in larval fish.
• None of the three is an exposure assessment. They measure effects, not how much a person wearing perfume is exposed to. Any leap from here to personal risk is one I have not made.
• Seventeen formulas is a very small sample. The 3.77% median dose rests on those 17 and would carry a wide interval.
• The name merging in the corpus was done by hand. Combining tonalid, fixolide (musk tetralin), acetyl tetralin and two full chemical names into one material is a CAS-based judgement, but the corpus itself has no CAS field, so other spellings may have been missed.
• The > 400 hours substantivity is a lower bound. The actual value is unknown.