Material guide: cis-3-hexenyl salicylate — the worst self-contradiction in a boiling point anywhere in the database

CAS 65405-77-8, 51 suppliers. Its boiling point field carries both 145 °C @ 5 mmHg and 394 °C @ 760 mmHg. Twenty-seven materials in the database register a boiling point at both of those pressures; the median gap between them is 147 °C, with quartiles at 126 and 164. This one is 249 °C, the largest in the database, 60 °C beyond the runner-up. Working back from the median puts it near 292 °C at atmospheric pressure, and its saturated analogue hexyl salicylate is registered at exactly 290 °C. In odour it is a bridge between two families: the cis-3-hexenyl half gives green leaf, the salicylate half gives soft floral and diffusion. It appears in 73 of 954 formulas at a median of 1.79%. Its only registered natural occurrence is witch hazel leaf oil at 0.01%.

About a 5 minute read.

The short version

Name: (Z)-3-hexen-1-yl salicylate, also cis-3-hexenyl salicylate, CAS 65405-77-8
C13H16O3, molecular weight 220.27, 51 suppliers
Odour type floral; odour: floral, green, metallic, herbal, balsamic
Medium strength (rank 2), substantivity 216 hours at 100%
Appears in 73 of 954 formulas, median dose 1.79% (quartiles 0.91–3.50%)
The boiling point field contradicts itself: the 5 mmHg to 760 mmHg gap is 249 °C, the largest in the database
Only registered natural occurrence: witch hazel leaf oil at 0.01%

The boiling point first

The bp field reads:

145.00 °C. @ 5.00 mm Hg | 394.00 °C. @ 760.00 mm Hg

Two numbers, two pressures. That the reduced-pressure boiling point is lower is obvious; the question is by how much.

Twenty-seven materials in the database register a boiling point at both 5 mmHg and 760 mmHg. I computed every gap:

	Gap (°C)
Median	147
Quartiles	126 – 164
Largest	249 (this material)
Runner-up	189 (cyclamen propanal)

249 °C is the largest in the database, 60 °C beyond the runner-up and 100 °C above the median.

Working back from the median: 145 + 147 = 292 °C.

And its saturated analogue hexyl salicylate (one double bond apart, molecular weight 222.28) is registered in the same database at 290 °C at atmospheric pressure.

Two degrees apart. The estimate matches the analogue; the registered value does not.

I think the 394 is wrong, by roughly 100 °C.

This check needs no chemistry

It is the cheapest kind of check I have run: a material's two boiling points check each other. You do not need to know what the molecule is, and you do not need to look up an analogue. You ask how much the temperature should rise going from 5 mmHg to atmospheric, and compare against the database median.

The vapour pressure field, by contrast, is fine. This one registers 0.000246 mmHg at 25 °C against hexyl salicylate's 0.000491. Two-fold apart, which is reasonable for one double bond. Only the 760 mmHg boiling point is broken.

It bridges two families

Split the name in half and each half belongs to a different family of perfumery materials:

The cis-3-hexenyl half is the skeleton of leaf alcohol, giving green leaf and cut grass
The salicylate half gives soft floral and balsamic character, plus very long substantivity

The formula corpus confirms the dual identity. Across the 73 formulas it appears in:

Other salicylates present	Count	Other cis-3-hexenyl esters present	Count
benzyl salicylate	31	cis-3-hexenol	23
hexyl salicylate	11	cis-3-hexenyl acetate	19
amyl salicylate	6	cis-3-hexenyl benzoate	7
cyclohexyl salicylate	4	cis-3-hexenyl tiglate	3

Both sides are there. It is not treated as either a green material or a salicylate. It routinely appears alongside relatives from both families at once.

That double bond takes it from 4 hours to 216

Take the same leaf alcohol, attach different acids, and the substantivity moves wildly:

Material	MW	Vapour pressure @25 °C	Substantivity	Strength rank
cis-3-hexenol	100.16	1.039 mmHg	4 h	3
cis-3-hexenyl acetate	142.19	1.219 mmHg	4 h	3
cis-3-hexenyl benzoate	204.26	0.002 mmHg	8 h	2
cis-3-hexenyl salicylate	220.26	0.000246 mmHg	216 h	2

Vapour pressure moves nearly four orders of magnitude and substantivity follows, which is exactly the shape I saw when I computed that relationship (rho = -0.647).

The last row overshoots, though. From the benzoate (0.002) to the salicylate (0.000246) is about 0.9 of a decade, and at the roughly threefold-per-decade rate I measured, the substantivity should go from 8 hours to around 22, not 216.

I do not know where the 216 came from. When I put this into a formula I plan around a material that lasts tens of hours, not 216.

A green leaf smell that leaves barely make

The occurrence field carries exactly one entry:

witch hazel leaf oil @ 0.01%

One natural source in the whole database, at one part in ten thousand.

What makes it interesting is that both halves are entirely natural on their own. cis-3-hexenol is leaf alcohol, released by nearly every damaged leaf, and methyl salicylate is the main component of wintergreen. The ester joining those two is something nature essentially does not make.

This is the same thing I keep hitting in this series: the molecule you smell is usually not the one the plant put out for you. This one goes further than most, in that both of its raw pieces are ordinary natural products while the product itself lives almost entirely in factories.

Dose: a median of 1.79%

It appears in 73 of the 954 formulas:

	Dose
Median	1.79%
Quartiles	0.91% – 3.50%
Max	60.0%

A very typical position. When I measured the dose distribution across 187 materials, half landed between 1% and 5%, and this one sits right in the middle of that band. The overall median for strength rank 2 is 2.69%, so it runs a little below its own rank.

The 60% maximum comes from a formula built around it as the lead; do not read it as normal.

Safety

Oral LD50 5,000 mg/kg (rat)
Registered cosmetic use is perfuming agents
The 2007 RIFM fragrance material review (PMID 18035471) covers its toxicology
A 2025 skin absorption study (PMID 39922548) measured its permeation and distribution from topical formulations directly, alongside ethyl salicylate and pentyl salicylate

Salicylates as a class carry contact dermatitis case reports; a 2006 Dermatitis paper describes a case involving both octisalate and this material. Worth keeping in mind for leave-on products.

What this doesn't establish

I did not measure a boiling point. What I ran is an internal consistency check: one material's two boiling points do not fit the database's normal relationship, and the implied value lands on its saturated analogue. That is enough to say the 394 is suspect, not enough to say 292 is correct.
Twenty-seven is a small sample. The 147 °C median gap comes from 27 materials, and the 126–164 quartile span is already wide. This material's 249 sits well outside it, but the interval itself does not carry much confidence.
I have not refuted the 216 hours, only declined to use it. I took no measurement; I pointed out that its relationship to vapour pressure is looser than that of the other three in its family.
The notes field reads None found, which is a placeholder, not a statement that there was nothing worth recording.
The 73-formula match may miss spelling variants. The corpus uses at least the spaced form cis 3 hexenyl salicylate as well, so the true count is probably a little higher.