Essential oils in a face mask, and what air does to them

There is a version of the ingredient list where the perfume is the problem and the plant oil is the solution. It is on a great many packs, and EU law does not recognise it. A distilled plant oil that makes a product smell of something is a fragrance material, and it answers to the same annex as the synthetic compound folded inside Parfum.

36 of our 215 masks carry an essential oil, and 19 different oils appear across the labels. 7 of those 19 now have to be named individually on the pack once they pass 0.001% in a leave-on product or 0.01% in a rinse-off one, which is the same pair of thresholds that has applied to the classic twenty-six fragrance substances since 2005. Most of them arrived on that list recently, through Regulation (EU) 2023/1545: lavender oil is Annex III entry 360, eucalyptus 355, lemon 353, bergamot 352. That change, and the two dates it runs on, is the subject of a separate piece. This one is about the oils themselves.

masks carry an essential oil

36masks carry an essential oilof 215

of these oils are established human contact allergens

8of these oils are established human contact allergensSCCS/1459/11, Table 7-5

of them read as fragrance-free on our own flag

10of them read as fragrance-free on our own flagno Parfum, no declared allergen

The last of those three numbers is the one worth holding on to. 10 masks here pass the name test for fragrance-free and still contain a scented oil, because the test looks for the word and the oil is not the word. What that phrase is actually required to mean is its own question.

In 2012 the EU’s Scientific Committee on Consumer Safety worked through the published evidence on fragrance substances and sorted the natural extracts into three groups: those with enough human evidence to call them established contact allergens, those with positive human data that falls short of that bar, and those nobody has published human data on at all. The distinction matters more than any single verdict, because the third group is not a clean bill of health. It is an absence of looking.

13 of the 19 oils on these labels appear somewhere in that sort. 8 are in the first group.

The oils not in that table are not absolved by their absence. Cypress, myrrh, chamomile and the rest are there because no clinical series has been published on them, which is what you would expect of materials used in small volumes. The opinion’s own indicative list exists for exactly that gap: it collects extracts with positive animal data or a known allergen inside them and no human evidence either way.

Here is the mechanism, and it is the part that almost never survives the trip from a safety opinion into a product page. Several of the most common fragrance terpenes are weak sensitisers as pure substances. Left in contact with air they oxidise, and the products of that reaction are the potent allergens. The committee’s term for a substance in that position is a prehapten: not yet the thing that sensitises you, one chemical step away from it.

The clinical difference is not subtle. The opinion tabulates patch-test results for the oxidised and the unoxidised forms of the same substances, in consecutive dermatitis patients at dermatology clinics across Europe:

The opinion states it plainly: “The non-oxidised compounds rarely cause allergic reactions.” So the entry on the back of the box, which is the thing this whole site is built on reading, cannot tell you the one fact that matters most about it. Nothing on a label records how much air the formula has met.

How common fragrance allergy actually is

Worth calibrating against, because the numbers above are clinic populations, not the street. The same opinion puts it this way: around 16% of eczema patients in the European population are sensitised to fragrance ingredients, and from studies on sectors of the population the frequency of fragrance contact allergy in the general European population can be estimated at 1-3%. Once acquired it does not go away, which is why the label is worth reading before rather than after.

Of the 54 individual chemicals the 2012 opinion classed as established human contact allergens, it singled out twelve as being of special concern because more than a hundred cases had been reported for each. Two of those twelve are Limonene and Linalool — and those two are exactly the entries that a plant oil brings onto a label with it.

They are also, after the blanket word itself, the two most common fragrance entries in this entire catalogue: Linalool on 44 masks and Limonene on 38, ranked 2 and 3 of the 14 fragrance entries we can resolve. 53 masks here name at least one of them. Citrus oil carries limonene; lavender carries linalool and linalyl acetate. When a label prints the oil and then prints those names after it, it is not naming two ingredients. It is naming a material and the parts of it the EU decided you have a right to see.

One oil here has been assessed on its own, twice, and it is the one most likely to be on the front of the box as a selling point. Melaleuca Alternifolia (Tea Tree) Leaf Oil is on 8 masks in this catalogue. It appears in neither Annex II nor Annex III, so there is no legal maximum for it and no requirement to declare it as an allergen — it is not in the 2023 amendment at all.

That is not the same as no assessment. The committee’s current opinion, adopted on 30 October 2025, considers tea tree oil safe as an anti-seborrheic and anti-microbial agent in four defended product types: up to 2.0% in shampoo, 1.0% in shower gel, 1.0% in face wash and 0.1% in face cream. It also says, in terms, “TTO is a moderate skin sensitiser.”

A face mask is not one of those four product types, so the reader has to do the mapping the opinion did not do. The honest version of it is that a mask you rinse away sits nearest the face-wash figure and a mask you leave on sits nearest the face-cream one, and those two numbers are a factor of ten apart. Of the 8 tea tree masks here, 3 are rinsed off and 5 are left on.

The 2025 opinion is also unusually direct about the thing this piece is about. The submission it assessed contained no data on the stability of the oil under the conditions of storage and use, and since the oil’s composition changes on exposure to light, heat, air or moisture, the committee said stability has to be maintained in the finished product so the components stay inside the ISO 4730 specification. The 2008 opinion, now superseded on safety but still the fullest published account of the chemistry, describes what that drift looks like: p-cymene rising up to tenfold, and in oil stored nine months under sunlight, the endoperoxide ascaridole forming.

Put the mechanism and the stability finding together and you get a conclusion that is not on any label: for this group of ingredients, how the product is packaged and how long it has been open is doing some of the work that the concentration usually does.

This is not our inference to make on its own, and it happens to be the committee’s. The 2025 opinion recommends including antioxidants in the formulation to prevent oxidation of the oil in the finished cosmetic, or using specific packaging to minimise exposure to light, in order to keep oxidation products as low as possible. The 2012 opinion makes the general version of the same point: prehapten activation, unlike activation inside the skin, can be prevented to a certain extent by avoiding air exposure during handling and storage.

Masks are unusually well split on exactly this axis. A sheet or hydrogel mask arrives in a sealed single-dose sachet, is opened once, and is thrown away — 21 of the 36 oil-bearing masks here are that kind. The other 15 are jars, tubes and pots: opened, dipped into, warmed in a bathroom and closed again, for as long as the pack lasts.

1. Read a plant oil as a fragrance, because that is what it is. If you are avoiding perfume, the oils belong on the same list. Our own fragrance filter on the catalogue is a name test and 10 masks slip through it while carrying an oil.
2. Treat “natural” as a sourcing fact, not a safety one. 8of the oils here are established human contact allergens on the EU’s own evidence review, and the ones that are not mostly have no published human data either way.
3. Prefer the sealed single dose if you react to fragrance. Not because a sheet mask is gentler — it is held against the skin for longer, under the stricter labelling threshold — but because the oil inside it has met less air.
4. Do not keep an opened oil-heavy jar for a year.The period-after-opening symbol on the back is the manufacturer’s answer to this question, and for these ingredients it is the answer to a real one.
5. If patch testing has named your allergen, the individual entries are the point. Oxidised limonene and oxidised linalool have been available as patch-test preparations since 2012, and the names appear on 53 of these labels.