Material guide · 93
Material guide: amber naphthofuran (Cetalox), same formula as Ambroxan under a different CAS
· 17 min read
CAS 3738-00-9, 31 suppliers. Its formula, molecular weight, logP and estimated water solubility all match ambroxan (6790-58-5) exactly, and the PubChem IUPAC names differ only by a stereodescriptor prefix at the front. The melting range is where they part: ambroxan melts over 74-75 C, this one over 75-85 C. FEMA and the Council of Europe give both the same numbers (3471, 10514) while CAS splits them in two. The database's own note reads 'Ambrox replacer'. Recommended evaluation strength is 10 percent or less, an order of magnitude looser than ambroxan's 1 percent, yet both carry strength rank 3.
In short: this material shares a skeleton with ambroxan, and the difference is whether the stereochemistry is specified. The chemical registry treats them as two substances; food-flavour regulation treats them as one. Which one to buy comes down to the melting range and the supplier count.
Fields
| Field | Value |
|---|---|
| CAS | 3738-00-9 |
| EINECS | 223-118-6 |
| First synonym | 1,5,5,9-tetramethyl-13-oxatricyclo(8.3.0.0.(4.9))tridecane |
| PubChem first synonym | Cetalox |
| Formula | C16H28O, MW 236.398 |
| Database category | flavor and fragrance agents |
| Appearance | colorless to pale yellow liquid to solid (est) |
| Assay | 96.00 to 100.00 |
| Melting point | 75-85 C (a 10-degree range) |
| Boiling point | 102-106 C @ 25 mmHg |
| Flash point | > 100 C |
| logP | 4.70 (est) |
| Vapour pressure | 0.009 mmHg @ 25 C (est) |
| Water solubility | 2.436 mg/L @ 25 C (est) |
| Solubility | alcohol; water, slightly; fine fragrances 0.5-5%; shampoo traces-0.5%; candle traces-1% |
| Odour type | amber |
| Odour | dry; woody; amber; ambergris; musk; sweet |
| Flavour | amber; woody old wood; tobacco; powdery; dry; ambergris; animal |
| Strength | high, recommend smelling in a 10.00% solution or less (rank 3) |
| Substantivity | > 400 hours (10% in dipropylene glycol) |
| Occurrence | found in nature; Salvia sclarea oil, Greece @ 0.10% |
| Suppliers | 31 |
| Regulatory | JECFA 1240, FEMA 3471, CoE 10514, FLAVIS 13.072 |
| Structure class | III |
| Oral toxicity | not determined |
| Database note | "Ambrox replacer." |
Why this gets its own entry
Because its fields look almost exactly like ambroxan's (CAS 6790-58-5), and it still has its own CAS number.
Side by side:
| ambroxan | this material | |
|---|---|---|
| CAS | 6790-58-5 | 3738-00-9 |
| Formula | C16H28O | C16H28O |
| MW | 236.398 | 236.398 |
| logP | 4.70 (est) | 4.70 (est) |
| Water solubility | 2.436 mg/L (est) | 2.436 mg/L (est) |
| Vapour pressure | 0.009 mmHg (est) | 0.009 mmHg (est) |
| Substantivity | > 400 hours | > 400 hours |
| Strength rank | 3 | 3 |
| FEMA | 3471 | 3471 |
| CoE | 10514 | 10514 |
Nine rows agree, decimal places of the estimates included.
The PubChem IUPAC names explain it. 6790-58-5 is (3aR,5aS,9aS,9bR)-3a,6,6,9a-tetramethyl-2,4,5,5a,7,8,9,9b-octahydro-1H-benzo[e][1]benzofuran; 3738-00-9 is the same string with the leading (3aR,5aS,9aS,9bR)- missing.
Same skeleton. One specifies all four stereocentres, one specifies none.
Two fields that really do differ
Melting point. Ambroxan is 74-75 C, one degree. This one is 75-85 C, ten degrees.
A pure substance melts sharply; mix something in and the melting point drops and spreads. A 10-degree melting range is about purity, not temperature, and it fits a registry entry with no stereochemistry specified.
Recommended evaluation strength. Ambroxan's field says "recommend smelling in a 1.00% solution or less". This one says "10.00% or less". A factor of ten.
Both carry strength_rank 3.
This is a field trap worth remembering. The strength rank flattens a tenfold difference. Building dilutions off the rank puts you an order of magnitude out. The number to read is the percentage inside the text.
Whether that tenfold gap was measured, I cannot tell. It may reflect a real potency difference, or it may be two records written by different people in different years. The database does not say.
The regulators disagree with the registry
| FEMA | CoE | JECFA | FLAVIS | |
|---|---|---|---|---|
| ambroxan | 3471 | 10514 | none | none |
| this material | 3471 | 10514 | 1240 | 13.072 |
FEMA and the Council of Europe assigned the same number to both. JECFA and FLAVIS list only this one, under the name 1,5,5,9-tetramethyl-13-oxatricyclo, a form carrying no stereodescriptors.
Ask whether these two are the same substance and the chemical registry says no while food-flavour regulation says yes. Neither is wrong. They are answering different questions.
Who sells it
31 suppliers, against 61 for ambroxan.
The overlap is the interesting part. Twelve suppliers list both CAS numbers, Firmenich and IFF among them. A company that considered the two interchangeable would have no reason to carry both as separate items.
Another 35 carry only 6790-58-5, and 3 carry only 3738-00-9.
Corpus
Of 954 published formulas, 95 contain an amber ether, 10.0%.
The spelling cetalox appears 27 times (8 of those tagged as a 10% dilution); amber naphthofuran appears twice. Formula authors write trade names, and the trade name Cetalox turns up under both CAS numbers in the database, retrievable from the supplier fields of this record and of ambroxan.
Median dose 0.50%, 90th percentile 2.00%.
None of the 95 formulas uses two different amber-ether spellings together.
The occurrence field
"Salvia sclarea oil, Greece @ 0.10%."
That figure lines up with how the material is made. Eichhorn and Schroeder reviewed the production routes to (-)-Ambrox in J Agric Food Chem in 2023: the traditional route starts from sclareol, the diterpene in clary sage, and proceeds by chemical modification and cyclisation. A newer route converts fermentation-derived (E)-beta-farnesene into (E,E)-homofarnesol and cyclises it enzymatically with an engineered squalene hopene cyclase.
The enzymatic route yields (-)-Ambrox, which is 6790-58-5, and the paper names it as Ambrofix.
Enzymes select a stereochemistry; chemical synthesis does not have to. That is the practical reason two CAS numbers coexist.
When you smell it
Takase and colleagues identified OR7A17, the human receptor tuned to (-)-Ambroxide, in Communications Biology in 2025. The same paper reports that non-functional alleles of this receptor are prevalent in human populations and especially common in East Asia, and that people carrying them still detect the molecule but find it less pleasant.
If it smells faint or unremarkable to you, the lot is not the first thing to suspect.
(That paper measured pleasantness, not the ability to discriminate two stereoisomers.)
What this doesn't establish
- I have not smelled this material. Everything here comes from database fields, PubChem and the papers.
- "Wide melting range means a mixture" is a general rule, not an analysis of this material. Ten degrees is consistent with several stereoisomers and equally with other impurities.
- I did not use the word racemic in any conclusion. What I confirmed is that the CAS entry specifies no stereochemistry. I have no analytical data on what is actually in circulation.
- I do not know whether the tenfold gap in recommended evaluation strength was measured or is a records artefact.
- logP, water solubility and vapour pressure are all estimates. Their exact agreement was one reason to infer a shared structure, but agreement between estimates is not agreement between measurements.
- The oral toxicity field reads "not determined". FEMA and JECFA numbers mean an evaluation process exists; this field itself is empty.
- The 12-supplier overlap came from matching supplier name strings. Identical names do not guarantee the same legal entity.
- From Takase 2025 I only have the abstract-level population statement, not per-population allele frequencies.