Material guide: cananga oil — it shares a CAS number with ylang's three grades, and its specific gravity range does not overlap theirs

About a 4 minute read.

The short version

• Name: cananga oil, CAS 68606-83-7, 40 suppliers
• Odour type floral; odour: sweet, floral, balsamic, green, spicy, animal, woody, waxy, leathery
• Medium strength (rank 2), substantivity 116 hours; FEMA 2232, FEMA GRAS
• Its CAS number is identical to ylang ylang grades I, II and III
• But its specific gravity (0.906–0.923) does not overlap ylang complete oil's (0.948–0.968)
• Its optical rotation is −15 to −30 against ylang III's −60 to −40, which does not overlap either
• Only 8 formulas in the corpus, at a median dose of 0.86%

One CAS number, four materials

Last article I worked through ylang's 22 records and found that the three grades (I, II, III) share CAS 68606-83-7 while the unqualified complete oil carries 8006-81-3.

This material's CAS is also 68606-83-7.

So a matching rule of "same CAS means same material" merges cananga with ylang I, II and III.

They are the same plant, Cananga odorata. What differs is the cultivated form and the origin. The notes field says:

Water distilled from the flowers of Cananga Odorata Hook cultivated in Java.

Note water distilled, not the steam distillation ylang usually gets. Different method, different thing coming out.

Two physical fields disagree

The CAS says one material; the measurements say otherwise:

	cananga oil	ylang ylang flower oil
CAS	68606-83-7	8006-81-3
Specific gravity @25 °C	0.90600 – 0.92300	0.94800 – 0.96800
Refractive index @20 °C	1.49500 – 1.50500	1.49700 – 1.51100
Substantivity	116 hours	140 hours
Suppliers	40	93

The two specific gravity ranges do not overlap at all. The upper bound of 0.923 sits below the lower bound of 0.948, with 0.025 between them.

That is practically useful: you can tell these two apart with a density bottle. The refractive index cannot (1.495–1.505 against 1.497–1.511, almost entirely overlapping).

Optical rotation separates them too: cananga at −15 to −30 against ylang grade III's −60 to −40, two ranges ten degrees apart and not touching.

Writing up nutmeg and mace I ran into the opposite situation: two tissues of one tree whose refractive indices overlapped past telling apart. This pair has at least one field that separates them.

Two words in the odour field that no ylang has

	Odour field
cananga oil	sweet, floral, balsamic, green, spicy, animal, woody, waxy, leathery
ylang complete oil	floral, petal, neroli, lily, spicy
ylang grade I	floral, petal, jasmin, neroli, lily of the valley, spicy, balsamic, earthy, woody, winey

Neither animal nor leathery appears anywhere in the ylang series.

That fits with cananga being cheap and treated as a ylang substitute. It is not a worse ylang; it is a different material leaning animal and leathery. If you want ylang's clean petal character, swapping in cananga adds a layer you did not ask for.

Eight formulas in the corpus

Formulas	8 (cananga oil 7, cananga essence 1)
Median dose	0.86%
Quartiles	0.60% – 2.50%
Max	10.0%

Against ylang's 152. This is a minor material.

And only 1 of the 8 also uses ylang. They rarely appear together in practice, which matches their odours pointing different ways: nobody puts in two materials for the same effect.

The median dose of 0.86% is low too. The overall median I measured is 2.11%, and merged ylang runs 2.50%. It is dosed three times leaner than ylang, because that animal note does not need much.

The inverse of the last article

Last round's Verdox carries a different CAS from Vertenex while their molecular formula, molecular weight, vapour pressure, substantivity and strength rank are all identical, and only their odour types differ.

This material is the reverse: the same CAS, with specific gravity, optical rotation, substantivity and odour field all different.

	Identifier	Physical fields	Odour
Verdox / Vertenex	different	same	different
cananga / ylang I–III	same	different	different

In both cases the identifier and the measurements tell different stories, and the nose was right both times.

The practical rule is simple: do not decide whether two materials are the same from one field. CAS numbers merge things that differ; physical fields split things that do not.

Safety and regulatory

• FEMA GRAS, FEMA 2232
• toxicity_oral reads Not determined, which is a placeholder
• Zhang and colleagues measured the anxiolytic effect of Cananga odorata essential oil in mice and identified its major active constituents in Phytomedicine in 2016 (PMID 27912874)

That paper says "Cananga odorata essential oil". Under this database's divisions, that could be cananga oil or it could be any of the ylang grades. The paper uses the botanical name and the database uses trade names, with no bridge between them.

What this doesn't establish

• I measured neither the specific gravity nor the optical rotation. The non-overlapping ranges are a comparison of field values, not of my measurements.
• "Water distillation and steam distillation extract different things" is a general principle. The notes field says only that cananga is water distilled; ylang's method comes from the ylang record's own notes ("by steam distillation of the flowers").
• Eight formulas will not support a robust median dose. Take 0.86% as a direction, not a figure.
• "Animal" and "leathery" are words in the odour field, not things I smelled.
• I did not look up the original CAS registry definition behind 68606-83-7 covering four materials. Under the TSCA definition it is "extractives and their physically modified derivatives, Cananga odorata", which is broad by design, so sharing it is not necessarily an error. It does mean CAS cannot be your only basis for identification.