An essential oil does not stand still once it is bottled. Oxygen, acid, light and heat transform constituents that carry double bonds and aldehyde groups. The consequences are threefold: the odour drifts, the original certificate of analysis no longer holds, and, most importantly, an old oil is a stronger skin allergen than a fresh one. This part presents the mechanisms only at a conceptual level, together with their practical consequences.
4.1 Autoxidation: from non-allergen to allergen
Atmospheric oxygen attacks the position adjacent to a double bond, forming hydroperoxides (–OOH group) as primary products; the hydroperoxides go on to decompose into alcohols, epoxides, ketones and aldehydes. For linalool, a theoretical computational study (DFT) describes the mechanism of hydroperoxide formation (Bäcktorp et al. 2006, doi:10.1021/jp0603278; abstract only read); the main hydroperoxide is 7-hydroperoxy-3,7-dimethylocta-1,5-dien-3-ol.
| Compound | After oxidation | Consequence | Source |
|---|---|---|---|
| Limonene | forms carvone, limonene oxide (both strong sensitisers) and carveol | unoxidised limonene does not sensitise, even in animals already sensitised to the oxidised form | Karlberg et al. 1992, doi:10.1111/j.1600-0536.1992.tb00129.x (abstract) |
| Linalool | no stabiliser, 45 weeks in air: about 30% linalool remaining, about 15% hydroperoxides | tocopherol at 200–300 ppm reduces hydroperoxide formation more than 30-fold; ECHA RAC (2015) concluded that linalool is a Category 1B skin sensitiser | ECHA RAC 2015 |
| Linalool (real products) | commercial products stored for at least 2 years: linalool oxides on average ≤ 1.8%, hydroperoxides about 0.6% | storage conditions determine the degree of oxidation | ECHA RAC 2015 |
| Linalyl acetate | 10 weeks of oxidation: sensitising concentration (EC3) falls from 25% to 3.6% | sensitising potency increases about 7-fold | Sköld et al. 2008, doi:10.1111/j.1600-0536.2007.01262.x (abstract) |
The two figures of 15% and 0.6% hydroperoxides do not contradict each other. They reflect two different conditions: forced oxidation in the laboratory and commercial products stored for at least 2 years Inference. Clinically, patch testing with oxidised linalool in 3,418 dermatitis patients gave positive rates that rose with concentration; the authors proposed a standard test concentration of 6% Measured (Christensson et al. 2010, doi:10.1111/j.1600-0536.2009.01657.x; abstract).
In tea tree, the cyclic terpenes with two double bonds (α- and γ-terpinene) are dehydrogenated to p-cymene, a stable benzene ring. A comparison of one fresh sample with one sample stored for about 25 years showed p-cymene roughly doubled, while α-terpinene, γ-terpinene and terpinolene decreased Measured (Shimadzu GC×GC application note, n = 1 versus 1, not peer-reviewed). Consequence: a genuine tea tree oil that initially met the standard may, after many years, exceed the 8% p-cymene ceiling and fall below the 14% γ-terpinene floor of ISO 4730 Inference.
4.2 Citral in acidic media: loss of lemon odour, formation of off-odours
Schieberle et al. 1988 (doi:10.1007/BF01454320; abstract only read) held citral in citric acid solution at pH 1.8 and 40 °C for 4 days and then analysed it by gas chromatography–olfactometry Measured. Citral cyclises to a p-menthane skeleton and is then aromatised into compounds with very low thresholds:
| Product | Threshold (ng/L air) |
|---|---|
| p-Cresol | 0.3–1.0 (most potent) |
| o-Cresol | 0.7–2.7 |
| p-Methylacetophenone | 2.7–10.8 |
| α,p-Dimethylstyrene | 665–2,660 |
| p-Cymene | detected by GC, not perceived by sniffing |
Because the degradation products are far more potent than the amount of citral lost, conversion of only a small fraction of citral is enough to create a medicinal off-odour and to strip away the lemon character. At alkaline pH, citral is cleaved by a different mechanism into methyl heptenone and acetaldehyde (Guo et al. 2022). Application Inference: lemongrass, litsea and lemon oils are unsuitable for low-pH products unless stabilisation measures are taken.
4.3 Isomerisation and polymerisation
- Anethole E → Z. Because the cis form is much more toxic, the JECFA specification limits cis-anethole to ≤ 1% Standard. The rate of conversion under storage conditions has no data in the sources read (NR).
- Acid-catalysed cyclisation of linalool, geraniol and nerol to α-terpineol: the reaction is an industry description Industry; according to the threshold comparison of the two pure compounds in Table 3, α-terpineol is about 50 times weaker than linalool.
- Polymerisation. Temperature, light and oxygen availability are the three main factors that degrade essential oils through oxidation and polymerisation (Turek & Stintzing 2013, doi:10.1111/1541-4337.12006; abstract). Oils rich in polyunsaturated terpenes (citrus, tea tree) thicken and darken with long storage Industry; no quantitative data are available yet.
4.4 Ageing markers measurable by GC
| Essential-oil group | Constituents that decrease | Constituents that increase |
|---|---|---|
| Tea tree (and citrus with γ-terpinene) | α-terpinene, γ-terpinene; terpinolene (tea tree) | p-cymene |
| Citrus | limonene | carvone, limonene oxide, carveol |
| Lavender, white champaca, rosewood (linalool-rich) | linalool, linalyl acetate | linalool oxides, hydroperoxides |
| Lemongrass, litsea (citral-rich) | geranial, neral | p-cresol, p-methylacetophenone (in acidic media) |
| Star anise, fennel | trans-anethole | cis-anethole |
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.