Reading conventions
Essential oil composition in this monograph is given as GC peak area percentage unless stated otherwise. This is the most common way of reporting, but it is not a true mass percentage, because each compound has a different detector response factor. Each composition figure is accompanied by the sample origin where available, because essential oil composition varies with region, cultivar, season and distillation method.
Each statement in the main text carries an evidence-type label: Measured data measured on samples in a peer-reviewed paper; Standard a standard or regulation; Industry a description of industry experience, without measured data; Inference a conclusion drawn by the compilers from the facts. Data that could be read only from a paper’s abstract are marked as such.
Odour thresholds come from three measurement systems that cannot be compared directly with one another: thresholds in air (ng/L or ppb by volume), thresholds in water or in a food matrix (µg/L), and compilation tables that do not state the measurement medium. Every “X times stronger” comparison in this monograph uses only values from the same study and the same medium.
For ISO standards, the monograph quotes only the ranges of the main specified constituents, as reference points for reading a certificate. The full table of each standard — including every constituent and every origin column — must be consulted directly in the ISO text or the equivalent TCVN.
Part I. From biosynthesis to odour profile
The composition of an essential oil is not random. It reflects the plant’s metabolic pathways, the plant part that is distilled, and what happens to the molecules after they leave the plant. This part sets out three basic concepts used throughout the monograph: the two terpenoid “tiers”, the methylation step that determines the chemotype of the phenylpropanoids, and the difference between mass percentage and contribution to odour.
1.1 Two pathways to the isoprene unit and the two tiers of an essential oil
All terpenoids are built from a 5-carbon unit (IPP and DMAPP). Plants have two independent pathways for making this unit, located in two different compartments of the cell Measured. The mevalonate (MVA) pathway, in the cytosol and partly in the peroxisomes, supplies the 15-carbon precursor (FPP) for sesquiterpenes. The methylerythritol phosphate (MEP) pathway in the plastids supplies the 10-carbon precursor (GPP) for monoterpenes and the 20-carbon precursor (GGPP) for diterpenes (Gutensohn et al. 2021, doi:10.3389/fpls.2021.691754).
The two compartments do exchange precursors, but not extensively. In transgenic tomato, the sesquiterpene product (nerolidol, made in the cytosol) was emitted at 60 to 584 times lower levels than the monoterpene product (linalool, made in the plastids). Boosting an enzyme of the pathway in the same compartment raised nerolidol 5.7-fold, whereas boosting an enzyme of the other pathway raised it only 2.5-fold Measured.
| Tier | Pathway, cell compartment | Example constituents | Typical essential oils | Role in fragrance |
|---|---|---|---|---|
| Monoterpenes C10 | MEP, plastids | limonene, linalool, citral, menthol, 1,8-cineole | citrus, lemongrass, mint, eucalyptus, lavender | evaporate quickly: top and middle notes |
| Sesquiterpenes C15 | MVA, cytosol | santalol, khusimol, patchoulol, cedrol, agarofuran | sandalwood, vetiver, patchouli, cedarwood, agarwood | low vapour pressure: base notes, fixatives |
| Diterpenes C20 | MEP, plastids | sclareol | rare in steam-distilled oils; more abundant in resins, concretes, absolutes | low volatility; semi-synthesis feedstock (e.g. Ambrox, Part III) |
Consequence for reading a certificate of analysis Inference: the ratio of monoterpenes to sesquiterpenes depends on the plant part (leaf, fruit peel versus wood, root) and on technique. Prolonged distillation draws in more of the heavy sesquiterpene fraction; time-based fractional distillation, as with ylang-ylang (Part V), separates these two tiers into different product grades.
1.2 Phenylpropanoids: the methylation step determines the chemotype
The second group of aromatic compounds starts from the amino acid phenylalanine (the shikimate pathway) and runs via cinnamic acid to monolignol alcohols such as coniferyl alcohol. This alcohol is acetylated and then reduced to a phenylpropene by an NADPH-dependent reductase. It is the enzyme that determines the position of the double bond: eugenol (terminal double bond, allyl group) and isoeugenol (double bond conjugated with the ring, propenyl group) are made by two homologous enzymes and are not interconverted in the plant Measured (Koeduka et al. 2006, doi:10.1073/pnas.0603732103; abstract only).
The final step is usually attaching a methyl group to the phenolic OH (O-methylation): eugenol to methyl eugenol, chavicol to estragole, anol to anethole. In basil, different lines accumulate either eugenol or estragole as the dominant constituent; phenylpropanoid genes account for about 13% and genes for methyl-donor production for about 14% of the transcripts in the glandular trichomes Measured (Gang et al. 2001, doi:10.1104/pp.125.2.539; abstract only).
| Side chain | Examples | Chemical and safety characteristics |
|---|---|---|
| Allyl (–CH₂–CH=CH₂) | eugenol, estragole, methyl eugenol, safrole, myristicin | The ethers can be metabolised to 1′-hydroxy derivatives that are genotoxic (SCF 2001) |
| Propenyl (–CH=CH–CH₃) | isoeugenol, anethole, isosafrole | Have E/Z geometric isomers; cis-anethole is 4.8–20 times more toxic than trans-anethole by LD50 in rodents (JECFA 1967) |
| Conjugated aldehyde | cinnamaldehyde | Skin sensitiser |
1.3 Percentage is not odour: the odour activity value
A GC-MS certificate lists peak area percentages. The nose does not read percentages; it responds to concentration relative to each compound’s threshold. The quantity linking the two is the odour activity value (OAV): concentration divided by detection threshold, both measured in the same matrix (Grosch 2001, Chem Senses 26:533). An OAV greater than 1 means the compound is likely to contribute to the odour.
Three examples show the gap between mass and odour:
- Bulgarian rose. β-Damascenone makes up only 0.14% of the oil but carries about 70% of the total “odour units” Measured (Demole et al. 1970, via the 2015 Cotton summary page of the University of Bristol, a secondary source).
- Cold-pressed Ponkan mandarin peel. Limonene makes up about 80% and total monoterpene hydrocarbons more than 89%, but the characteristic odour is determined by two aliphatic aldehydes, octanal and decanal; limonene serves only as a background Measured (Sawamura et al. 2004, doi:10.1271/bbb.68.1690; abstract only).
- Vetiver. The main odorant is a trace ketone, (+)-2-epi-ziza-6(13)-en-3-one, with a threshold of 29 pg/L air, about 150 times stronger than khusimone, the compound once thought to be dominant. The most abundant constituents by mass are not the character odorants Measured (List et al. 2020, doi:10.1002/anie.202014609; read via a summary).
OAV has two limitations to bear in mind. First, it depends strongly on the matrix used to measure the threshold (Figure 1). Second, a high OAV does not guarantee a large contribution: in a model aroma of Grenache rosé wine comprising 24 compounds with OAV above 0.5, omitting β-damascenone (the second-highest OAV) changed the odour only slightly Measured (J Agric Food Chem 2002, doi:10.1021/jf0115645; abstract only). To find out which compounds really determine an odour, the odour must be reconstituted from reference compounds and each compound removed in turn (omission test).
A character-impact compound is one that on its own is enough to evoke the identity of the material: 1-p-menthene-8-thiol for grapefruit, methyl anthranilate for Concord grape, 2-methoxy-3-isobutylpyrazine for bell pepper, β-damascenone for rose. Parts II and III show that these compounds typically have extremely low thresholds and are present in very small amounts.
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.