Material guide: Lilial — a top-fifteen material in the corpus, banned in EU cosmetics, and my database never mentions it

Start with the corpus. With solvents removed, the most frequent materials across 954 public formulas rank like this:

Rank	Material	Formulas	Median dose
1	Linalool	338	5.93%
2	Benzyl acetate	317	5.00%
3	Phenethyl alcohol	274	10.00%
9	Hydroxycitronellal	161	8.00%
13	Lilial	151	6.00%

It appears in 15.8% of the formulas, ranking 13th.

And since 1 March 2022 it has been banned in cosmetics in the European Union.

Those 151 formulas cannot be made as written in the EU.

My database never mentions this

That is the half to state first.

Looking up Lilial's record (#22425, filed under the name lilyall), the regulatory field reads:

Field	Content
Regulatory listings	(blank)
Shelf life	24 months or longer
Suppliers	37

Blank. No JECFA, no FEMA GRAS, no CoE, and not one word about the EU ban.

And thirty-seven suppliers still sell it.

Reading only that record, you see a good material — 236 hours' longevity, medium strength, a muguet floral — and you put it in a formula. The record will not stop you.

This is the reason Is this material still legal exists, and Lilial is that article's best example: a database is a source of physical properties, not a source of regulatory status. Those two always get checked separately.

The short version

• 151 formulas (15.8%) in the corpus, median dose 6.00%. By occurrence rate it ranks 13th in these public documents.
• Banned in EU cosmetics from 1 March 2022, under Cosmetics Regulation 1223/2009 Annex II entry 1666 as amended by 2021/1902 — that is the cosmetics regulation, not IFRA.
• The reason is H361: suspected of damaging fertility or the unborn child. That is a reproductive toxicity classification, not sensitization.
• My database's regulatory field is blank, and 37 suppliers remain. The record carries no warning.
• A 2024 Saudi analysis found banned Lilial or Lyral in 97 of 108 marketed products. Banned is not the same as gone.
• It is not found in nature, and its odour field includes cumin — a facet that comes out at higher concentrations.

The basics

Database name	lilyall
Also known as	Lilial, butylphenyl methylpropional (BMHCA)
CAS	80-54-6
Chemical name	2-(4-tert-butylbenzyl)propionaldehyde
Formula	C₁₄H₂₀O
Molecular weight	204.31
Odour	Floral, muguet, watery, green, powdery, cumin
Odour family	floral
Strength	Medium
Longevity	236 hours at 100% (a neat value)
Appearance	Colourless to pale yellow clear oily liquid
Boiling point	250–251 °C @ 750 mmHg
Flash point	98.9 °C
Vapour pressure	0.005 mmHg @ 25 °C (est)
logP	3.60 (est)
Natural occurrence	Not found in nature
Storage	24 months or longer
Regulatory listings	(blank)
Suppliers	37
Corpus position	151 formulas (15.8%), median dose 6.00%

It was banned for a different reason than you would guess

Fragrance materials usually get retired over sensitization — Lyral went that way, on two decades of accumulated patch test data.

Lilial did not.

Look at the six GHS statements:

H302	Harmful if swallowed
H315	Causes skin irritation
H317	May cause an allergic skin reaction
H361	Suspected of damaging fertility or the unborn child
H401	Toxic to aquatic life
H411	Toxic to aquatic life with long lasting effects

H361 is the one that matters. It falls under the CMR classifications (carcinogenic, mutagenic, reprotoxic), and EU cosmetics regulation treats CMR substances on a different track from sensitizers — CMR substances are prohibited in cosmetics as a matter of principle rather than given a concentration limit.

So it did not travel through IFRA's 263 standards; it went into Annex II of the cosmetics regulation, the prohibited list. Which is also why searching IFRA will not find it.

Acute toxicity runs higher too: oral rat LD50 1,390 mg/kg with somnolence and dyspnea observed, intraperitoneal mouse LD50 700 mg/kg. Against cyclamen aldehyde's 3,810 and omega-pentadecalactone's above 5,000, that is noticeably lower.

Banned is not the same as gone

This is the part most worth knowing.

Al-Mussallam and colleagues, in Regulatory Toxicology and Pharmacology in 2024, analysed the Saudi market. Using a validated GC-MS method they quantified the 26 fragrance allergens in 108 marketed products, and ran a quantitative risk assessment to determine the risk of inducing allergies.

Two findings:

• Most allergens were present at acceptable concentrations, but 19 products carried a risk of inducing allergies;
• and Lilial and Lyral, two prohibited fragrances, were detected in 97 products.

Ninety-seven of one hundred and eight.

The authors note this is the first study in Saudi Arabia evaluating the safety of these 26 well-known fragrance allergens, and suggest it could serve as a regional standard for future research.

Read that number carefully. Saudi Arabia does not operate under EU cosmetics regulation, so those 97 products were not necessarily illegal locally — the paper measures how much is still out there, not how much is in breach. But for someone formulating, the implication is direct: a material banned in one jurisdiction keeps circulating in global supply chains and on shelves. The 37 suppliers in my database are the same fact from the other side.

A 2026 paediatric respiratory study

Tunçerler and colleagues, in the Annals of Allergy, Asthma & Immunology in 2026, ran a prospective case-control study enrolling 80 recently diagnosed children with wheezing and 80 healthy controls (August 2023 to February 2024), recording laundry habits and product ingredients and measuring lung function by impulse oscillometry.

Softener use was higher in the wheezing group (P = .011). Chemical analysis showed significantly higher frequency of sodium lauryl sulfate, nonionic active substances, cocoglycoside, methylisothiazolinone, butylphenyl methylpropional (that is, Lilial), linalool, coumarin and oxygen-based bleaching agents in the detergents used by the wheezing group (P < .05). That group showed higher airway resistance and lower reactance (P < .05), and children using additive-free detergents had lower airway resistance (P < .05).

This is association, not causation. A case-control design compares the frequency of ingredients between two groups; the list runs to eight classes of substance, Lilial being one, and the study did not and could not attribute the effect to any single ingredient. The authors' own conclusion says "a potential link".

It is here because it demonstrates something specific: a material banned in EU cosmetics still turns up in other product categories and other markets, with researchers reading its name off a detergent label in 2026.

(If your formula will reach children, that article has the full procedure.)

About that cumin

Worth a technical note: the odour field reads floral, muguet, watery, green, powdery, cumin.

The cumin facet emerges at higher concentrations, and it is why plenty of people dislike Lilial — the clean muguet turns into a sweaty, cuminic register once you push the dose. Cyclamen aldehyde's mustiness is the same phenomenon in the same slot.

It is not found in nature, like cyclamen aldehyde, heliotropin and helional — almost the entire muguet slot consists of molecules nature does not make.

If you own a bottle

Establish your jurisdiction first. EU cosmetics: prohibited. Elsewhere, the rules differ and you have to check. This article cannot decide for you.

If you are in the EU and making something for people to use or to sell: you cannot use it. And every one of those 151 public formulas containing it needs rewriting.

For study and evaluation only: it remains a good teaching material for understanding the muguet slot, and it is one of the more obtainable reference points in that functional position. But treat it as a historical material rather than a palette member.

On replacements: the corpus's neighbours in that slot are hydroxycitronellal (161 formulas, 8.00%) and cyclamen aldehyde (64 formulas, 2.00%), plus several purpose-built substitutes on the market. Check each one against current regulations — this slot moves faster than any other family in perfumery.

What this doesn't establish

I did not re-verify the regulatory text this round. Annex II entry 1666 and the 1 March 2022 date are notes I verified previously. Check the current EU regulation yourself before relying on them — that is precisely what this article is about, and I should not commit the error I am describing.

A blank regulatory field does not mean the database is wrong. That field records flavour listings (JECFA, FEMA, CoE, FLAVIS), and Lilial is not on those lists, so blank is correct. What is wrong is reading that field as "regulatory status". That misreading is the point of this article.

The 151 formulas reflect the age of those documents. That corpus of public demo formulas is 74% patent or sourced examples, most predating 2022. So "15.8% of formulas are now non-compliant" is a statement about historical documents, not about anything currently on sale.

The Saudi study is not a compliance audit. It measured ingredient concentrations in marketed products and ran a risk assessment; 97 of 108 is a detection rate, not a violation rate. I did not check local regulation.

The paediatric study is associational. Eighty against eighty, many ingredients compared simultaneously, with no way to attribute the difference in respiratory function to Lilial. It is here to describe circulation, not to claim an effect.

Compare acute LD50 figures cautiously. Numbers from different decades, species and routes of administration sit in one table here; the direction is readable, the ratios are not.

References

A. S. Al-Mussallam et al., Quantitative risk assessments of skin sensitization for 26 allergens in different consumer products in the Saudi market, Regulatory Toxicology and Pharmacology, 153, 105714 (2024). PMID 39368536. doi:10.1016/j.yrtph.2024.105714

G. Tunçerler et al., Fresh clothes, hard breaths: Laundry washing habits, detergents, softeners, and impaired respiratory functions in children with wheezing, Annals of Allergy, Asthma & Immunology, 136(1), 85–91 (2026). PMID 40651564. doi:10.1016/j.anai.2025.07.006

→ Lilial in the database: full physical data and suppliers. Check regulatory status separately. → Is this material still legal: the procedure for checking it yourself.