The same name is not the same material: batch variation, and where natural and synthetic actually differ

The previous article asked you to record supplier and batch number in your evaluation notes. That field looks like administrative housekeeping, and it is the one people skip most. It is also the one they go hunting for later.

Here is why.

One hillside, threefold variation within a year

In 2025 Guelifet and colleagues published a very clean piece of work in Molecules. They took wild rosemary (Rosmarinus officinalis) growing in Algeria, the same plants throughout, and varied only two things: harvest season (four seasons) and distillation method (hydrodistillation, steam distillation, microwave-assisted distillation). Then they ran GC-MS.

The same four constituents dominated throughout: camphor, α-pinene, camphene, 1,8-cineole. The amounts:

Constituent	Range across seasons and methods	Fold
Camphor	21.95 – 36.52%	1.7×
α-Pinene	7.31 – 21.72%	3.0×
Camphene	7.84 – 18.23%	2.3×
1,8-Cineole	7.65 – 14.53%	1.9×

Spring harvests extracted by microwave distillation were the most enriched in oxygenated monoterpenes, up to 59.6%.

Nothing else changed. Same species, same location, same supplier, same plants on the same hillside, and one major constituent still moved by a factor of three.

The bottle you buy will say "rosemary essential oil".

Extraction method is a second, independent axis

That study varied season and method together for a reason: both change the composition, and neither shows up in the name.

In our data the scale of this is countable. Taking materials with odour descriptions and supply information, discarding single molecules and keeping only mixtures (essential oils, absolutes, extracts, resinoids), 1,078 botanicals are identifiable by name.

Of those, 214 (20%) exist in more than one extraction form in the database. The extremes:

Botanical	Forms	Which
Clove bud	7	absolute, CO₂ extract, concrete, essence, oil, oleoresin, resinoid
Rosemary	6	absolute, CO₂ extract, concrete, essence, oil, oleoresin
Celery seed	5	absolute, CO₂ extract, concrete, oil, oleoresin
Clary sage	5	absolute, concrete, essence, oil, oleoresin
Jasmine	5	absolute, CO₂ extract, concrete, essence, oil

Under the single name "jasmine" sit five separate materials. Batch variation has not even entered the picture yet.

A note that says only "jasmine" will match none of them when you read it back in six months.

The name on the bottle does not guarantee what is in it

Both of the above count as honest variation: the material is what it claims to be, it simply moves. The next case is different.

In 2021 Allenspach and colleagues used chiral gas chromatography in Scientific Reports to measure the enantiomeric ratio of α-pinene. They compared two sets:

• 54 primary essential oil samples from five Pinus species, of known provenance
• 11 commercial oils labelled Pinus sylvestris

The difference showed up in the direction itself:

Primary oils of P. sylvestris showed a positive enantiomeric excess of (+)-α-pinene, whereas commercial oils labelled P. sylvestris showed an excess of (−)-α-pinene.

The two groups have opposite optical handedness. The authors conclude that chiral analysis reveals mislabelling, use of the wrong herbal substance, and sources of adulteration in pine oil, and propose a threshold of 17.5% (−)-α-pinene enantiomeric excess for identifying them.

This goes beyond a quality gap: the contents and the name on the label come from different botanical sources.

So where do natural and synthetic actually differ?

Most beginners draw one conclusion from all this: buy natural.

That conclusion rests on an assumption: that "natural" and "synthetic" name two different kinds of thing. Our data can test it, though first we need to be clear about what the field records.

This field does not record how the bottle you buy was produced. We checked against known materials: ambroxide is recorded as natural — it does occur in ambergris, but essentially all commercial supply is synthesised from sclareol — while "lavender oil CO₂ extract" is recorded as natural/synthetic, and a plant extract plainly has no synthetic production route.

So the field records whether the molecule occurs in nature, not how it was made. Using it to say "X% of materials are natural" would be flatly wrong — exactly the kind of source-convention trap this series keeps checking for.

Read correctly, of the 5,231 materials in the base set that carry a value:

	Count	Share
Occurs in nature	2,219	42.4%
Both (occurs naturally and can be synthesised)	2,754	52.6%
Synthetic only (not known in nature)	258	4.9%

Only 4.9% of the palette consists of molecules not known in nature. More than half are explicitly both: the same molecule, extractable from a plant or synthesised, identical either way.

Vanillin is the clearest case: it occurs in vanilla beans, and the overwhelming majority of world supply is synthetic. The molecule is the same molecule.

So at the molecular level "natural versus synthetic" is largely a false dichotomy. The real difference is not chemistry. It is variance.

• A single molecule is a specification. 99% pure linalool this month and next month is, as far as your nose can tell, the same thing.
• A natural material is a distribution: dozens to hundreds of compounds whose proportions move with season, origin, harvest year, extraction method, even how long the still ran. The threefold range above is what that looks like in practice.

Neither kind is worse. They are different kinds of object: one stable but thin, one complex but moving. Professional formulae use both, precisely because they behave differently.

A figure we computed and decided not to use

A negative result worth recording, because it demonstrates the check this series keeps running.

We wanted to show that natural mixtures smell more complex than single molecules, by comparing how many descriptors each carries. The result came out backwards: single molecules have a median of 5 descriptors (mean 5.35), mixtures only 4 (mean 4.38).

That is almost certainly a fact about documentation rather than perception. Single compounds get formal study and standard odour profiles; oils frequently get one conventional phrase. The figure measures how thoroughly the source wrote things down; smell complexity never enters into it.

So we did not use it.

What this means for your notes

Five practical consequences:

1. Treat the batch number as part of the material's identity. The heading of a note should read "jasmine absolute / supplier X / batch Y", all three parts. Without the last two, the note describes something you can no longer obtain.

2. When a new batch arrives, evaluate it beside the old one, never alone. Smelled by itself, a new batch feels "about the same". Side by side, the difference appears. The article on olfactory fatigue covered why: adaptation and context. This is also why you keep retains.

3. Keep retains. One to two millilitres of every batch, sealed and refrigerated, labelled with batch number and arrival date. I learned this the annoying way: my first bottles ran out before I thought to compare, and "I think it used to smell softer" convinces nobody, least of all me. Without a retain, your notes can only ever be compared against memory, and memory loses.

4. Natural materials need two extra fields: origin and harvest year. For a single molecule those columns stay empty. For a natural they often explain more of the difference than the supplier does.

5. Be deliberate about how much a formula leans on one natural. If a formula's success rests on one particular batch of one natural material, that formula cannot be reproduced. Keep using naturals; just avoid letting a single one hold the whole structure up. The IFRA article reached the same conclusion for a completely different reason.

---

"The same name" is a convenience of the supply chain. Record at a granularity that can tell things apart, or in six months you will find yourself comparing against an average that never existed.

References

K. Guelifet et al., Seasonal and extraction-dependent variation in the composition and bioactivity of essential oils from wild Rosmarinus officinalis L., Molecules, 30(21), 4258 (2025). PMID 41226217

M. Allenspach, C. Valder, D. Flamm & C. Steuer, Authenticity control of pine sylvestris essential oil by chiral gas chromatographic analysis of α-pinene, Scientific Reports, 11, 16923 (2021). PMID 34413399

---

Next up: quantities. Why formulae are written in percentages and what makes drops unreliable, how many decimal places your balance needs, and how to make a 1% solution that comes out the same every time.