Material guide · 146
Material guide: WS-23 — the first strength rank 1 material in this series, and that is the whole point of it
· 14 min read
CAS 51115-67-4, 25 suppliers, FEMA 3804. Its strength field reads 'low', rank 1 — nearly all of the hundred and forty-odd materials I have written up are rank 3, and this one is the opposite extreme. The notes field explains why: 'physiological cooling agent used in food; its effect is similar to that of menthol but without the strong minty flavour.' It separates the cooling sensation from the smell. And that separation has a basis at the receptor level — Sherkheli and colleagues noted in 2010 that known TRPM8 agonists like menthol and icilin cross-activate TRPV3 and TRPA1, while the menthol derivatives they characterised are selective for TRPM8 and do not activate the other thermo-TRPs. Its evaluation concentration is 25%, and the database holds only 14 distinct evaluation concentrations.
About a 4 minute read.
The short version
- Name: WS-23, 2-isopropyl-N,2,3-trimethylbutyramide, CAS 51115-67-4, FEMA 3804
- Odour type: minty; odour: cooling, mentholic, minty
- Strength: low, rank 1 — rare in this series
- A white crystalline solid melting at 60–64 °C; evaluated at 25%
- Suppliers: 25; oral toxicity field: not determined
- It separates the cooling sensation from the smell
Rank 1
odor_strength holds one word: low. Strength rank 1.
I have written up a hundred and forty-odd materials, and nearly every one is rank 3 — that is this database's norm, because the materials worth writing about tend to be the powerful ones.
This one is the opposite, and that is why it exists.
The notes field says it directly:
Physiological cooling agent used in food; its effect is similar to that of menthol but without the strong minty flavour. Flavouring agent for chewing gum and candies.
Menthol gives you two things: a cooling sensation and a minty smell. When you only want the first — a citrus drink that must not smell of toothpaste, say — menthol is the problem.
WS-23 is the answer to that problem.
And the separation has a basis at the receptor
Cooling is not a smell. It is a trigeminal thermal sensation, running through the TRPM8 ion channel — a neighbour in the same family as the TRPA1 that allyl isothiocyanate works through.
Sherkheli and colleagues, in the Journal of Pharmacy and Pharmaceutical Sciences in 2010, set out exactly why that separation is hard:
Known TRPM8 agonists like menthol and icilin cross-activate other thermo-TRP channels like TRPV3 and TRPA1 and mutually inhibit TRPM8. This limits the usefulness of menthol and icilin as TRPM8 ligands. Consequently, the identification of selective and potent ligands for TRPM8 is of high relevance both in basic research and for therapeutic applications.
And the group they characterised:
In the present investigation, a group of menthol derivates was characterized. These ligands are selective and potent agonists of TRPM8. Interestingly they do not activate other thermo-TRPs.
So the WS series' selling point is not merely "smells less." Menthol hits three channels; this class hits one — the cooling it gives is cleaner, not just quieter.
(I have written about WS-3, another member of the same series.)
The taste field describes a sequence in time
taste_descriptions:
Medicinal, menthol. / Initial menthol taste that fades to lingering cooling feel.
Two things happen separately: the taste leaves and the cooling stays.
That sentence is the material's specification.
Its evaluation concentration is 25%
odor_descriptions: at 25.00% in dipropylene glycol. mild cooling mentholic minty.
I measured the database's evaluation concentrations: across 7,620 records there are only 14 distinct values, and five decade rungs (100%, 10%, 1%, 0.1%, 0.01%) cover 99.6%.
25% is not on that ladder — it belongs to the remaining 0.4%.
And the reason is the same as the rank 1. Nobody evaluates a weak material at 1%; you go up until you can smell it. That ladder exists to suppress strength, and this material does not need suppressing.
(It is a white crystal melting at 60–64 °C, so evaluation has to happen in solution anyway — a different reason for dilution.)
Where it is mostly used now
Chen and colleagues developed a GC-MS method in Frontiers in Chemistry in 2025 to quantify three cooling agents at once — menthol, WS-3 and WS-23 — and their samples were:
The test aerosols were generated from laboratory-formulated e-liquids under optimized conditions. Aerosols were obtained from an electronic vaping device …
E-cigarettes.
Their opening also lists the range of use: "coolants are widely used as functional additives in various industries including food, personal care, pharmaceuticals, and tobacco."
This material's home ground is not perfume. Its cosmetic_uses reads only "fragrance", while the flavor field and that list point at chewing gum, confectionery, oral care and aerosol products.
How to use it
- It has almost no smell, so the limit on dose is sensation rather than odour. Too much is uncomfortably cold.
- It is a crystal; use a solution. The 25% evaluation concentration is why.
- For cooling without mint, this. For mint, use menthol.
- Its place in perfume is limited. If what you want is a cold effect on skin, that is skincare thinking rather than perfumery.
- The oral toxicity field is empty (
Not determined) — one of that 87.7%.
What this doesn't establish
- I have neither smelled nor tasted it. This is fields, whole-database statistics, and two papers.
- Sherkheli 2010 characterised a group of menthol derivatives, and I did not confirm whether WS-23 was among them. I cite their conclusion that selective TRPM8 ligands do not activate the other channels, not a direct measurement of this material.
- "Cooling runs through TRPM8" is prior knowledge; I did not cite the original channel identification.
- Chen 2025 is an analytical methods paper, neither a safety nor an efficacy study. I cite it for the current distribution of use.
- I checked no regulatory limit for cooling agents in food or aerosol products.
- "Too much is uncomfortably cold" is inferred from the mechanism; I have no dose data.