The structure–odour relationship asks a simple question: if one detail of a molecule changes, how does the odour change? For terpenoids, the answer shows clear tendencies but many exceptions. This part uses pairwise comparisons within the same study and the same measurement matrix to separate each factor: functional group, double-bond position, cyclisation, geometric isomerism and enantiomerism.
2.1 Functional groups: a strong tendency, not a law
A convenient dataset for comparison is that of Plotto et al. 2004 (doi:10.1002/ffj.1470): thresholds measured by a three-alternative forced-choice test (3-AFC) on 33–58 people, in deodorised orange juice, together with values in water taken from the literature Measured. In the orange juice matrix, linalool has a threshold of 113 µg/L and limonene 13,700 µg/L, so linalool is about 120 times stronger.
Figure 2 ranks seven terpenes by their threshold in orange juice.
Yet α-terpineol, also an alcohol, is weaker even than limonene: 280 versus 60 µg/L in water, 25,900 versus 13,700 µg/L in orange juice. In the other direction, some hydrocarbons are very potent: in fresh Chinese ginger quantified by isotope dilution, myrcene has an OAV of about 19,000, second only to 1,8-cineole (about 65,000) Measured (Schaller & Schieberle 2020, doi:10.1021/acs.jafc.0c06733; abstract only). Conclusion: oxygenation usually lowers the threshold by 1–2 orders of magnitude on the same C10 skeleton, but the position of the functional group and the surrounding skeleton matter more than the presence of oxygen itself Inference.
Measurement in air by gas chromatography–olfactometry (GC-O) with 8 trained panellists adds two further rules Measured (Elsharif et al. 2015, doi:10.3389/fchem.2015.00057):
- Esterification weakens the odour. Linalool has a threshold of 3.2 ng/L; linalyl acetate 110.9 ng/L, about 35 times weaker. Linalyl acetate is still important in lavender because it makes up 25–47% of the oil under ISO 3515 and shifts the odour nuance towards fruity, sweet.
- Oxidation at the chain end remote from the main functional group weakens or abolishes the odour. Adding oxygen at C-8 of linalool raises the threshold about 12-fold (aldehyde form) to about 50-fold (alcohol form); the acid form is odourless.
The same study recorded individual thresholds for the same compound ranging from 0.6 to 989 ng/L, i.e. about three orders of magnitude. Every “threshold” figure is a group mean; a given individual may be hundreds of times more or less sensitive.
2.2 Double-bond position and cyclisation
| Change | Compound A | Threshold A | Compound B | Threshold B | Result |
|---|---|---|---|---|---|
| Endocyclic → exocyclic double bond | α-pinene | 9.5 | β-pinene | 140 | exocyclic form about 15 times weaker (about 23 times in orange juice) |
| Cyclisation of a tertiary alcohol | linalool (acyclic) | 5.3 | α-terpineol (cyclic) | 280 | about 50 times weaker |
| Cyclisation of a hydrocarbon | myrcene (acyclic) | 36 | limonene (cyclic) | 60 | about 1.7 times weaker |
These pairs have practical significance: in essential oils, acid is the main catalyst of the cyclisation of linalool, geraniol and citral to α-terpineol and other p-menthane skeletons, with loss of the floral and lemony notes (Part IV). On the geraniol skeleton, adding a methyl group at different positions also shifts the odour in different directions: at C-8 towards resinous, at C-4 towards lemon, at C-2 almost unchanged Measured (Sommer et al. 2022, doi:10.3390/molecules27123827).
2.3 Geometric isomers: mainly a change in potency
Geraniol (E form) has a threshold of 14 ng/L air, nerol (Z form) 60 ng/L, i.e. about 4 times weaker. The authors’ conclusion: the E/Z configuration at C-1 mainly affects potency, whereas oxidation at C-8 mainly affects odour quality Measured (Elsharif & Buettner 2018, doi:10.1021/acs.jafc.6b04534; abstract only). In terms of nuance, geraniol is described as citrus and floral; nerol adds a resinous touch (Sommer et al. 2022).
In citral, the two isomers geranial (E) and neral (Z) interconvert towards equilibrium. At pH 7 in the presence of soy protein, within 100 minutes about 37% of pure geranial converted to neral and about 60% of neral converted to geranial Measured (Guo et al. 2022, doi:10.3389/fnut.2022.929023). The geranial/neral ratio on a certificate is therefore a snapshot at the time of measurement, not a constant of the batch. No primary measurement source has been found for the separate thresholds of geranial and neral (NR).
2.4 Enantiomers: from a few-fold difference to odorous versus odourless
| Compound | Form A | Form B | Source |
|---|---|---|---|
| Carvone | (R)-(−): spearmint | (S)-(+): caraway, dill | Leitereg et al. 1971 (DOI verified only; odour descriptions are textbook knowledge) |
| (Z)-β-Santalol | (−): creamy, lactonic, characteristic sandalwood | (+): odourless | Birkbeck 2017 (review) |
| Nootkatone | (+): 0.6–1.0 µg/L water (grapefruit odour) | (−): 400–800 µg/L | Sommer et al. 2022, citing the literature |
This difference underpins the most important authenticity testing technique in use today. Plants usually make one enantiomer in very high proportion (for example (R)-linalool at about 94% in lavender), whereas ordinary synthetic material is a 50:50 (racemic) mixture. Adding synthetic material skews the enantiomeric ratio even when the overall percentages remain within the standard (Part V).
2.5 Sesquiterpenes and woody notes
In the sesquiterpene tier, the gap between the main constituent and the odour carrier is even greater.
| Material | Main constituent by mass | Character-impact compound | Source |
|---|---|---|---|
| Sandalwood Santalum album | (Z)-α-santalol; total α + β-santalol above 90% in East Indian oil | (−)-(Z)-β-santalol, GC-O dilution factor 1024 | Birkbeck 2017, doi:10.2533/chimia.2017.823 |
| Vetiver | khusimol 3.4–13.7% (9 origins) | (+)-2-epi-ziza-6(13)-en-3-one, 29 pg/L | Champagnat et al. 2006 (abstract); List et al. 2020 |
| Patchouli | patchoulol 31.9–68.4% (from sources cited in the review) | NR; the industry often cites norpatchoulenol, but no readable primary source has been found | Khairan et al. 2020 (review) |
| Cedarwood Juniperus | J. virginiana: α-cedrene 27.2%, thujopsene 27.6%, cedrol 15.8%; J. ashei: thujopsene 60.4%, cedrol 19.0% | NR | Adams 1987, doi:10.1007/bf02859346 |
| Agarwood Aquilaria | 2-(2-phenylethyl)chromones 59 + 60: 43.9–73.0% (wild kynam) and 72.4–95.6% (cultivated kynam) in the extract | no single compound yet; chromones are odourless at room temperature and become odorous on heating | Yang et al. 2021, doi:10.3390/molecules26020436 |
Two consequences Inference. First, ISO 3518 specifies both Z-α- and Z-β-santalol but does not distinguish the (−)/(+) enantiomers, whereas only the (−)-(Z)-β form carries the characteristic sandalwood odour; “meets ISO” therefore reflects authenticity more than odour quality. Second, the smell of agarwood when burnt or heated differs in nature from the smell of distilled agarwood oil: most of the resin mass is chromones, which become odorous only on pyrolysis and barely pass into the distilled oil (Part VI). When evaluating Vietnamese agarwood, these two contexts must be kept apart.
The sandalwood market illustrates the practical risk: of 6 commercial essential oils labelled “sandalwood” sold in the EU, none met ISO 3518. Three were in fact Amyris balsamifera (“West Indian sandalwood”, containing no santalol), one was a mixture of synthetics, one contained 74.6% benzyl benzoate; the only S. album sample had 37.1% α-santalol, below the 41% minimum Measured (Kucharska et al. 2021, doi:10.3390/molecules26082249).
2.6 Summary of the rules
- On the same C10 skeleton, an oxygen functional group usually lowers the threshold by 1–2 orders of magnitude, with clear exceptions (α-terpineol weak; myrcene, α-pinene strong).
- Esterification and oxidation remote from the main functional group weaken the odour; carboxylic acids are usually odourless or fatty.
- Cyclisation usually weakens fresh (floral, lemony) notes and pushes the nuance towards woody, resinous.
- E/Z isomers mainly change potency (about 4-fold for geraniol/nerol); enantiomers can differ from a few-fold to odorous versus odourless, and this cannot be predicted by simple rules.
- In the sesquiterpene tier, the character-impact compound is usually a trace constituent; the percentage of the main constituent is an authenticity criterion, not an odour criterion.
A monograph compiled from 718 sourced facts collected under a common protocol and cross-checked independently. Items labelled Inference are the compilers’ own judgements. Diagrams are newly drawn from published data. Scientific reference material — not legal or medical advice, and no substitute for the text of the standards.