"Add benzophenone for UV protection" — that word covers two different things, and the one you would buy has a smell

About 5 minutes.

When I wrote Should I add BHT, someone in the thread asked the follow-up: fine, that covers oxidation — but what about UV? Should I add benzophenone?

I said at the time that I had not looked into it. I have now, and the answer is messier than I expected.

That word covers two different things

Searching by name and CAS, the string benzophenone maps to a whole family of records:

Name	CAS	Database category	Suppliers
benzophenone	119-61-9	fragrance agents	41
oxybenzone (= benzophenone-3)	131-57-7	cosmetic UV absorber	12
benzophenone-4	4065-45-6	cosmetic UV absorber	9
benzophenone-1	131-56-6	cosmetic UV absorber	8
benzophenone-12	1843-05-6	cosmetic UV absorber	7
benzophenone-2	131-55-5	cosmetic UV absorber	6

The ones on a sunscreen label are the numbered derivatives, each with its own CAS. The unnumbered one — the one you get if you type benzophenone into a supplier's search box — is a fragrance material.

It is also the one with the most suppliers in the family: 41, more than all the others combined.

The one you would buy has a smell

The fields for CAS 119-61-9:

Field	Value
Odour type	balsamic
Odour	balsamic, rose, metallic, powdery, geranium, herbal
Strength	medium (smell in a 10% solution or less), rank 2
Substantivity	394 hours
Melting point	47–51 °C (a white crystalline powder at room temperature)
Assay	98.00 to 100.00
FEMA / JECFA / FLAVIS / CoE	2134 / 831 / 07.032 / 166

394 hours, medium strength, rose and geranium. This is not something you drop in as a protective additive on the grounds that it has no smell. It is a fragrance material, and a long-lasting one. It has a FEMA number, so it is used as a flavour material too.

Supplier descriptions include "faint, sweet rose, light herbal nuance like geranium" and "very faint powdery-rosy geranium-like odor with a slightly metallic note". Faint is not none, and 394 hours of faint is faint that stays.

For contrast, I pulled every material whose uses field contains "UV absorber". There are 45 of them, and none has an odour type. The materials actually used as UV absorbers do not even have their smell fields filled in.

The photochemistry is the bigger problem

The database's notes field says this outright:

"Benzophenone is a common photosensitizer in photochemistry. It crosses from the S1 state into the triplet state with nearly 100% yield. The resulting diradical will abstract a hydrogen atom from a suitable hydrogen donor to form a ketyl radical."

That sentence is worth stopping on.

A UV absorber and a photosensitizer do opposite jobs.

• Absorber: takes in the UV energy and dissipates it as heat. The energy stops there.
• Sensitizer: takes in the energy, crosses into a long-lived triplet state, then hands the energy to a neighbouring molecule — or goes and pulls a hydrogen off something. The energy carries on from there.

Benzophenone is the textbook case of the second. When Marazzi and colleagues recalculated its triplet kinetics in J Phys Chem Lett in 2016, they opened by calling it a paradigmatic molecule for photosensitization. Prentice and colleagues used it in 2023 precisely because it is a cheap and effective photosensitizer for photocatalytic synthesis.

It is widely used because it is good at passing light energy on.

Cuquerella and colleagues wrote a whole review on this in Accounts of Chemical Research in 2013, titled "Benzophenone photosensitized DNA damage". Their reason for choosing it is stated plainly — it is a classical and paradigmatic chromophore. And their opening explains why this matters:

Photosensitizers can extend the "active" fraction of the solar spectrum to the UVA region and beyond, which means that photosensitizers increase the probability of developing skin cancer upon exposure to sunlight.

Note the direction of that sentence. It does not say benzophenone protects you from UV. It says it makes otherwise harmless wavelengths active.

(To be clear: that paper is about DNA damage mechanisms, not about cosmetic or perfume risk assessment. I am not saying your perfume causes cancer. I am saying the mechanism runs the opposite way from what "add it to block UV" assumes.)

The GHS field

Benzophenone's GHS classification field is not empty:

H351 — Suspected of causing cancer H400 — Very toxic to aquatic life H410 — Very toxic to aquatic life with long lasting effects

I counted the background first. Of the 42,225 records in the database, only 359 (0.9%) have a GHS entry at all, rising to 337 of 3,156 (10.7%) among materials with ten or more suppliers. So having a GHS entry is itself uncommon.

Among the records that do have one, four carry H351: acetaldehyde, benzophenone, musk xylol, and (Z)-aconitic acid.

Four. Acetaldehyde has its own article in this series; musk xylol does not yet, appearing only in The four families of synthetic musk.

Read that number carefully. It means "four among the records that have a GHS entry", not "only four fragrance materials in the world are suspected carcinogens". Field coverage is 0.9%, and the missing 99.1% does not mean safe, it means no data. That trap is in Four kinds of nothing found.

And its own storage field says "protected from light"

The storage field on the same record:

"Store in cool, dry place in tightly sealed containers, protected from heat and light."

You do not add a material that needs to be kept away from light as your light protection.

So what do you do about UV

The question itself is sound — light really does change a perfume's colour and smell. But the fix is in the bottle, not the formula.

• Use opaque or dark glass. This is the only step that actually stops UV on the outside.
• Keep it cool and dark. Not the bathroom windowsill, not the car.
• Citrus materials react to light on their own account, which is a separate story: Why citrus oils are phototoxic.
• Oxidation and light exposure often happen together but are not the same thing — separated in The bottle you opened is no longer the same material.
• If you do want to use benzophenone, use it as a fragrance material — rose, geranium, powdery, balsamic, 394 hours, evaluated at 10% or less. That is a formula decision, not a protective measure.

What this doesn't establish

• I measured nothing. The whole article is database fields, database-wide counts and three papers.
• I did not check current EU cosmetic regulation. The database's regulatory field lists only JECFA, FEMA GRAS, CoE and FLAVIS, with no IFRA entry — which does not mean there is no restriction, only that this field does not carry one. Check the current rules yourself before using it.
• The Cuquerella review is about DNA damage mechanisms, not perfume risk assessment. What I take from it is that sensitizers and absorbers run in opposite directions, not any health conclusion at any dose.
• "None of the 45 UV absorbers has an odour type" describes how the database is filled in, not that they are genuinely odourless. An empty field and a zero are different things.
• The four H351 records are limited by that 0.9% coverage and must not be read as a complete list.
• I did not compare the UV absorption spectra of benzophenone and benzophenone-3. What I have is classification and notes, not spectra.
• I do not know which one the original commenter meant. They may well have meant the sunscreen family — that the word is not precise enough is the point of this article.