Mineral sunscreen doesn't work the way you think it does

The story you've been sold goes like this: chemical sunscreens absorb UV and convert it to heat, which means they're harsh, potentially hormone-disrupting, bad. Mineral sunscreens — zinc oxide, titanium dioxide — sit on top of the skin and physically deflect UV radiation like a tiny mirror shield, which means they're gentle, inert, natural. One absorbs. The other reflects. Clean vs. chemical. Safe vs. scary.

Almost none of that is accurate. And photobiologists have been saying so for decades.

The mirror myth

The claim that mineral sunscreens work by reflecting UV is wrong in a measurable, quantified way. A 2016 study by Cole, Shyr, and Ou-Yang — using an optical integrating sphere to measure what ZnO and TiO₂ actually do to UV radiation across the full spectrum — found that the average reflectance of both minerals throughout the UV range is just 4–5%. That's less than SPF 2. Less UV protection via reflection than a white cotton T-shirt.

The remaining 95% of UV protection comes from absorption — specifically, from semiconductor band-gap physics. Antony Young, Emeritus Professor of Experimental Photobiology at King's College London, is direct about it: "People claim mineral or inorganic sunscreens reflect UV radiation. That is not accurate."

This is not a fringe position or a new finding. The Cole et al. paper opens with explicit frustration that scientists have had to keep reiterating this point since the 1980s. The misconception was baked in by the FDA's original 1970s monograph, which described inorganic filters as physical reflectors and scatterers. That regulatory language — written before modern semiconductor photophysics was applied to sunscreens — gave us "physical sunscreen," a term that has no meaningful scientific basis.

How ZnO and TiO₂ actually work

Both zinc oxide and titanium dioxide are wide-bandgap semiconductors. When a UV photon hits a ZnO or TiO₂ particle, it excites electrons from the valence band into the conduction band — absorbing the photon in the process. This is band-gap absorption. It is the same fundamental principle that makes solar panels convert sunlight to electricity. It is not a mirror. It is not a physical barrier. It is electronic absorption.

The numbers: ZnO has a band gap of 3.37 eV, which corresponds to an absorption cutoff around 368 nm. TiO₂ in its anatase form has a band gap of 3.2 eV, absorbing up to roughly 387 nm. Both cover UVB and most of UVA2 through this mechanism. A 2024 review in Accounts of Materials Research confirmed: "ZnO and TiO₂ demonstrate low backscattering, yet high absorbance, in the UV range, indicating that absorption is the primary mechanism of photoprotection."

The white cast, incidentally, has nothing to do with UV protection. ZnO and TiO₂ reflect and scatter visible light — wavelengths above their band gap where absorption no longer occurs. That's what makes you look grey. The same particles that leave a white cast are absorbing 95% of the UV before the visible scattering even becomes relevant.

Why the myth refuses to die

Follow the incentives. "Physical reflector" is a better marketing story than "semiconductor absorber." It creates a simple, intuitive distinction — shield vs. sponge — and positions mineral sunscreens as categorically different from, and superior to, organic filters. The "physical vs. chemical" framing is a false dichotomy that serves the clean beauty industry, not the science. As Brian Diffey, Emeritus Professor of Photobiology at Newcastle and creator of the UVA star rating system, puts it: "Everything is a chemical."

The Cole et al. paper's conclusion is unambiguous: "The functions of insoluble 'physical' or 'mineral' UV filters are fundamentally the same as those of soluble 'chemical' UV filters." Both classes absorb UV. They differ in molecular mechanism, not in the core act of absorbing radiation to prevent it from reaching your skin.

The oxybenzone scare, by the numbers

The other pillar of mineral sunscreen marketing is what it isn't: it isn't oxybenzone. The oxybenzone fear traces largely to a 2001 rat study by Schlumpf et al., which found uterine weight increases in immature rats after exposure to several UV filters, including oxybenzone (benzophenone-3). The study was real. The interpretation applied to humans was not.

The dose that produced uterine effects: 1,525 mg/kg per day, administered orally, for four days. Wang, Burnett, and Lim calculated in 2011 what systemic exposure equivalent to that dose would require via topical application in a human.

The math: at realistic application amounts (face, neck, hands, and arms — roughly 3.75 mL per day of a 10% oxybenzone formula), with a 2% dermal absorption rate, matching the oral dose from the Schlumpf study would take approximately 277 years of continuous daily use. The 34-year figure applies to full-body application at maximum recommended amounts — still well beyond any plausible exposure scenario. Wang et al. concluded that both time periods and application regimens "are essentially unattainable."

A 2004 human study applied creams containing 10% oxybenzone to subjects for four days. Slightly reduced testosterone was observed at the four-hour mark. After four days of application, hormone differences between subjects and controls had disappeared, and researchers concluded the early variations were not caused by the sunscreen.

The broader toxicology evidence has been reviewed at scale. A 2025 paper in Critical Reviews in Toxicology — Cohen et al., covering avobenzone, ensulizole, homosalate, octinoxate, octisalate, and octocrylene — concluded that all six UV filters are "not genotoxic and show no evidence of biologically relevant carcinogenic modes of action," and that human systemic exposure levels fall well below concentrations with any biologic activity.

Nuance is warranted: the Australian TGA's July 2025 safety review found that homosalate and oxybenzone, at concentrations currently permitted in Australia, have margins of safety below their 100-unit threshold — and recommended concentration restrictions. This does not mean the ingredients are unsafe. It means regulators are appropriately applying precautionary margins to the data that exists. The EU already limits oxybenzone to 6% in face products; the TGA review aligns with this direction. "Not definitively harmful" and "unrestricted use at any concentration" are different claims, and honest sunscreen communication requires holding both.

What's actually changing

If the mineral-vs-chemical framing is mostly noise, the real developments in sun protection are happening elsewhere.

In December 2025, the FDA proposed adding bemotrizinol (Tinosorb S) to the list of approved sunscreen active ingredients — the first proposed addition in 26 years. Bemotrizinol is a broad-spectrum UVA/UVB filter that has been approved in Europe since 2000 and used in Australia and Asia for over two decades. Clinical data show plasma concentrations after application rarely exceed the FDA's 0.5 ng/mL threshold, with no evidence of accumulation. If finalized, it would become only the third GRASE (Generally Recognized as Safe and Effective) sunscreen ingredient under U.S. regulation, joining ZnO and TiO₂. The irony: the incoming addition to the safe list is an organic filter.

On the mineral side, UCLA researchers published results in early 2026 on tetrapod-shaped ZnO particles — a high-temperature flame synthesis process that produces particles with porous, branching structures that resist clumping and stay evenly distributed across skin. The result is SPF ~30 performance with a warmer, skin-toned appearance rather than grey cast, without special surface coatings. It's a genuine formulation advance that addresses the one legitimate complaint about mineral sunscreens — not their mechanism of action, which was never the problem.

Also worth noting: if you are treating hyperpigmentation or melasma, filter type matters for a different reason. ZnO and TiO₂ absorb UV but do not adequately block visible light (400–700 nm), which is a significant driver of pigmentation. Tinted mineral sunscreens with iron oxides address this gap — iron oxide is currently the only widely available ingredient that blocks visible light. That's a real reason to consider a tinted formula. "It reflects UV like a mirror" is not.

What actually matters when choosing sunscreen

The evidence consistently points to a few factors that dwarf filter type in real-world UV protection:

Application amount. Most people apply 25–50% of the amount used in SPF testing (2 mg/cm²). Halving the application converts SPF 50 into approximately SPF 7. No filter chemistry overcomes this. Apply more.

Reapplication. Both organic and mineral filters degrade or rub off with sweat, water, and time. Two hours is the general guideline; sooner after swimming or sweating. Mineral filters are not inherently more durable on skin.

Broad-spectrum coverage. Check that UVA is covered — not just UVB. In the U.S., look for the "broad spectrum" label. ZnO offers better UVA1 coverage than TiO₂ alone, which is a genuine reason to favor ZnO-containing formulas if that's your concern.

Formulation you'll actually wear. The best sunscreen is the one you use consistently. If a mineral formula feels heavy or leaves you looking ashen, and that makes you skip it — that is a clinical problem. Switch to whatever you'll apply properly and reapply routinely.

Antony Young has been studying sunscreen effectiveness for decades. His summary: "SPF is SPF. The specific ingredients are not of primary importance."

The mineral-vs-chemical debate is a marketing construct dressed up as science. The actual science — the semiconductor photophysics, the toxicology, the clinical efficacy data — consistently fails to support it. Use a broad-spectrum sunscreen with an SPF you'll actually apply enough of. The rest is noise.