Material guide: juniper berry oil — the catalogue itself says most of what is sold is a distillery byproduct

About a 5 minute read.

Too long, didn't read

• Name: juniperus communis fruit oil (juniper berry oil), CAS 8012-91-7, FEMA 2604, 107 suppliers
• Odor family terpenic; descriptors: fresh, balsamic, terpenic, woody, carrot seed, fir needle, peppery
• The notes field says most commercial juniper berry oil is a byproduct of juniper brandy manufacturing
• CAS 84603-69-0 covers 12 records, from berries to wood to sprouts
• Two products named with juniper, juniper lactone and juniper muscone, share a different CAS and are filed as musks
• Flash point 40 °C, H226 flammable category 3, H315 skin irritation; shelf life 12 months
• Regulatory field holds FEMA GRAS and CoE 249 only; not FCC listed

The catalogue says it, and no field can hold it

The notes field, verbatim:

Commercial Juniper Berry Oil is produced by two methods: The best oil is steam distilled from the crushed, dried or partially dried, ripe berries. Occasionally water distillation is used. The greater part of all commercial juniperberry oil, however, is derived from the fermented fruits as a byproduct of the central European juniper brandy manufacturing.

Read that plainly: most of the juniper berry oil on the market is not berries distilled to make a fragrance material. It is recovered from the fruit left over after the spirit was made. Fermented, soaked in alcohol, and there is no reason for its aroma profile to match that of fresh berries.

And no field can hold this fact. The database has a regulatory field, a GHS field, a shelf life field. It has no field for "what is the mainstream process for this material". The sentence lies in free text in the notes, where no search reaches it and no filter selects on it. Who wrote the odour field covered this class of problem; this is the expensive version. The sentence that most determines what you end up buying is the one you cannot query.

Read further down the same notes and you meet this:

minimizes hepatic reperfusion injury.

A subjectless fragment of a pharmacology abstract, wedged between the process description and the TSCA definition. Not misinformation, but a visible seam. The notes field in this catalogue was assembled from several sources and the joins were never cleaned up. When you see a sentence like that, you know to read the field as a lead rather than a definition.

One CAS, twelve records, berries to sprouts

CAS 84603-69-0 carries 12 records:

Record	Suppliers	Odor family
juniper berry oil CO2 extract	12	balsamic
juniper needle oil	10	woody
juniperus communis fruit extract	9	woody
juniperus communis wood extract	6	woody
juniperus communis fruit water	3	—
juniperus communis fruit tincture	1	herbal
juniperus communis wood oil	1	woody
5 more (sprout, seed water, branch/fruit/leaf, …)	0	—

Berry, needle, wood, sprout and seed, multiplied by CO2 extraction, tincture and hydrosol. Five plant parts times four processes, under one number.

And the berry distillate that this article is about is not under that number at all. It is 8012-91-7, which carries three records: the fruit oil at 107 suppliers, a branch oil at 8 and an absolute at 1. So the CAS alone will not tell you whether you are looking at fruit or at branches. The clove piece had the same shape with bud, leaf and stem; juniper is the more extreme case.

Two products called juniper that are not juniper

Two records in the database:

Name	CAS	Suppliers	Odor family	Descriptors	Longevity
juniper lactone	109-29-5	23	musk	sweet, musk, balsamic, amber, animal	400 hours
juniper muscone	109-29-5	9	musk	sweet, powdery, musk, animal	400 hours

They share one CAS, both are filed as musks, both record 400 hours. These are macrocyclic musks, not juniper. The word juniper is part of a trade name, not a statement about the plant it came from.

This is not rare in catalogues. To work out what a material is, read the CAS and the odor family, not the name. The -olide suffix covered names lying through their endings; this is the same trick at the front.

Carrot seed in the odor field

The fifth descriptor in juniper berry oil's odor field is carrot seed, which reads like noise the first time.

Eight materials in the database carry that descriptor:

Material	Suppliers
juniperus communis fruit oil	107
juniper berry oil terpeneless	8
douglas fir oil	5
wild carrot seed absolute	5
(E)-2-nonen-1-yl acetate	5
juniper berry oil replacer	4
juniper berry fragrance	2
wild carrot seed oleoresin	1

Four of the eight are juniper-related, counting the terpeneless version, the replacer and a fragrance version. Two are wild carrot seed itself, and one is douglas fir. This is a real shared descriptor, not a typo. Juniper, douglas fir and carrot seed meet somewhere, and even a juniper replacer copies the word across.

The terpeneless version records 5 hours of longevity. The main record leaves longevity empty.

The Bulgarian oil is more than half alpha-pinene

In 2014 Höferl and colleagues at the University of Vienna and the University of Food Technologies in Plovdiv analysed Bulgarian juniper berry oil by GC/FID and GC/MS (PMID 26784665).

The composition, as reported: α-pinene 51.4%, myrcene 8.3%, sabinene 5.8%, limonene 5.1% and β-pinene 5.0%.

Those five total 75.6%, all monoterpene hydrocarbons. Three quarters of the oil is one class of molecule, and α-pinene alone is more than half of it. The odor family label, terpenic, is being honest.

Two practical fields follow from that. The flash point is only 40 °C (monoterpene hydrocarbons are light) and the shelf life is only 12 months (monoterpene hydrocarbons oxidise readily).

"Juniper" is a genus, not an oil

In 2021 Zheljazkov and colleagues compared two other species of Juniperus (PMID 34203980): J. excelsa and J. sabina.

Their main constituents do not overlap at all. J. excelsa runs to α-cedrol, limonene and α-pinene, while J. sabina runs to sabinene, terpinen-4-ol, myrtenyl acetate and α-cadinol.

The authors also note that J. sabina contains podophyllotoxin, used as a precursor for anti-cancer drugs. Within one genus, one species gives you the smell of gin and another gives you the precursor of a cytotoxic drug.

So the words "juniper essential oil" specify nothing. Write the botanical name, and write it down to the species. Same lesson as the eucalyptus piece.

Buying and storing

107 suppliers, easy to source. Asking the right question matters more here than usual:

• Ask about the process. Distilled berries, or brandy byproduct? The notes field says the latter is the mainstream. The two differ in price and in smell, and a quotation will not volunteer it.
• Ask about species and plant part. You want Juniperus communis berries. Needle oil, wood oil and other Juniperus species all exist, and some of them are nothing like what you had in mind.
• Flash point 40 °C, H226 flammable category 3. That is low for a natural oil, the same bracket as galbanum oil and Siberian fir needle oil. Treat it as flammable for shipping classification and for storage.
• Shelf life 12 months. Oils dominated by monoterpene hydrocarbons all run short. Decant, date it, keep it dark. Oxidised terpenes are a sensitisation source, the same story as the citral group.
• Optical rotation −15 to 0, refractive index 1.472–1.484, specific gravity 0.859–0.869. Those three you can check yourself on arrival.

→ Juniper berry oil in the database: full physical data, 107 suppliers and suggested blends.

What this article does not establish

• "Mostly a brandy byproduct" is the catalogue's own note, with no independent industry statistics behind it here. It gives no year and no proportion.
• I treated the pharmacology fragment as a seam without tracing which paper it came from.
• For juniper lactone and juniper muscone I confirmed only that they share CAS 109-29-5, are filed as musks and record 400 hours. I did not establish which molecule that CAS corresponds to; the database's ω-pentadecalactone sits under a different number (106-02-5).
• 51.4% is that Bulgarian lot, not a global average. Origin and year both move it.
• Zheljazkov's paper is about J. excelsa and J. sabina, not J. communis. I use it to show that a genus is not a species, not to describe this article's subject.