Sunscreen UV Filters — Master Synthesis

Spectra & Photodegradation of All Globally-Approved Active Ingredients

Compiled: 2026-05-07 Methodology: 6 parallel deep-research subagents per /workspace/overview/research-agent-instructions.md. ~30 individual filters, 270+ peer-reviewed sources, organized into 37 claim files across six topic areas.

All numerical claims (λmax, ε, photodegradation rate) trace to a primary peer-reviewed source listed in the per-area sources.md. Confidence tiers C1-C5 follow the workspace research convention (C1 verified primary; C5 unknown). Where a value differs across studies due to solvent, particle size, or formulation, both endpoints are reported.

1. The 30 Filters at a Glance

                  280              320               360               400nm
                   |  UVB  |  UVA-II  |       UVA-I       | Visible
                   |   |   |    |     |        |          |
─── ORGANIC, UVB-DOMINANT ────────────────────────────────────────────────
PABA              ████ peak 283
Octinoxate (OMC)        █████ peak 308–311  ⚠ photolabile
Octisalate (EHS)        ███ peak 305–307
Homosalate (HMS)        ████ peak 306–309
Cinoxate                █████ peak 308
Padimate O                █████ peak 311
Octocrylene             ████ peak 303         (→ benzophenone Downs 2021)
4-MBC (Enzacamene)      ████ peak 300        ⚠ EU ban May 2026
Polysilicone-15         █████ peak 312
Octyl Triazone (EHT)      █████ peak 314 ⭐ ε≈135,000 highest of any
Ensulizole (PBSA)       ████ peak 302 (water-sol.)
Trolamine Salicylate    ███ peak 298

─── ORGANIC, UVA OR DUAL ─────────────────────────────────────────────────
Avobenzone (BMDBM)                            ███████ peak 357 ⚠⚠ photolabile
Meradimate                                █████ peak 336 ⚠ ¹O₂ sensitiser
DHHB (Uvinul A Plus)                          ███████ peak 354 ⭐ photostable
Bisoctrizole / MBBT (Tinosorb M) █████   █████████ dual 305 + 360 ⭐ broadest
Iscotrizinol (Uvasorb HEB)         █████ peak 310

─── ORGANIC, BROAD-SPECTRUM / NEXT-GEN ───────────────────────────────────
Bemotrizinol (Tinosorb S)         █████  ███████ dual 310 + 340 ⭐
Ecamsule (Mexoryl SX)                       █████ peak 345
Drometrizole Trisiloxane (Mex XL) █████  ██████ dual 303 + 344
Mexoryl 400 (MCE)                                       █████ peak 385 ⭐
Tinosorb A2B (TBPT)                █████ peak 310 + scatter
Bisdisulizole (Neo Heliopan AP)              █████ peak 335 (water-sol.)

─── BENZOPHENONES ────────────────────────────────────────────────────────
Oxybenzone (BP-3)              █████  ██████ dual 287 + 325 ⚠ reef ban
Sulisobenzone (BP-4)           █████  ██████ dual 287 + 325 (water-sol.)
Dioxybenzone (BP-8)            █████  ██████ dual 282 + 325–352
Mexenone (BP-10)               █████  █████ dual 287 + 325 (obsolete)
Benzophenone-1 (BP-1)          █████  █████ dual 287 + 325 (metabolite)

─── INORGANIC SEMICONDUCTORS ─────────────────────────────────────────────
Zinc Oxide (ZnO)            ████████████████ Eg=3.37 eV; cutoff 368 nm
TiO₂ rutile                 ███████████████████ Eg=3.00 eV; cutoff 413 nm
TiO₂ anatase                ████████████████ Eg=3.20 eV; cutoff 387 nm
                                                 (avoided in cosmetics)
                  280              320               360               400nm

2. Master Filter Reference Table

3. Photodegradation Reference (key kinetic values)

Always state the solvent / vehicle / lamp: the same molecule can be photolabile in one solvent and photostable in another. Avobenzone is the textbook example.

3.1 Photolabile filters (significant loss under standard solar dose)

3.2 Photostable filters (>95% retained at typical solar exposures)

3.3 Inorganic — the filter is photostable but catalyses degradation of others

4. Photostabilization Combinations (real products are mixtures)

Hierarchy of stabilisers for avobenzone

Triplet energy ladder (the photophysics of why these work)

~70 kcal/mol  | BMDBM enol S1 — fast ESIPT (productive UV-A absorption)
~60–66        | BMDBM diketo T1 (n,π*) — the photoreactive species
              |  ↓ all triplet quenchers attack here
~58           | DEHN T1 (³ππ* naphthalate)
~55           | OC ≈ 4-MBC ≈ MBBT T1
~50–55        | Bemotrizinol T1, DESM T1
~50           | OMC (cinnamate) T1   ← but instead of quenching, undergoes [2+2]

Quenching requires the stabiliser's T1 to lie ≥2-3 kcal/mol below BMDBM diketo T1.

5. Solvent / Vehicle Effects (the most under-cited variable)

Avobenzone in particular has dramatic solvent-dependent λmax and photostability:

(Mturi & Martincigh 2008, J Photochem Photobiol A 200:410.)

Vehicle ranking for unstabilised avobenzone (worst → best): mineral oil ≈ light paraffins ≈ anhydrous waxes < cyclohexane ≈ DMSO < ethyl acetate < C12-C15 alkyl benzoate ≈ dicaprylyl carbonate < ethanol < methanol/water (protic) ≈ encapsulated (β-cyclodextrin, lipid microparticles).

6. Regulatory Matrix (May 2026)

Notable trajectories: - The Dec 2025 proposed FDA order on bemotrizinol (6% GRASE) is the first new TEA approval in decades — may break the logjam for DHHB, MBBT, iscotrizinol, drometrizole trisiloxane, ethylhexyl triazone. - Trolamine salicylate is proposed by FDA as not GRASE in sunscreens (2021 Proposed Order). - 4-MBC is banned in EU effective May 2026 for endocrine disruption. - Mexoryl 400 (approved EU via Reg 2020/1684, broadly commercialized 2022–2024) is the newest filter — λmax 385 nm closes the previously-uncovered ultra-long UVA1 band.

7. Cross-cutting Findings

7.1 The structure–photostability–regulatory triangle

The "next-generation" filters (Tinosorb M/S, DHHB, Mexoryl SX/XL/400, EHT, iscotrizinol, TBPT) share three properties: 1. High molecular weight (>500 Da typically) → no skin penetration → no measurable plasma absorption 2. Photostable chromophores (ESIPT, rigid triazines, particulates) → useful in real products without aggressive stabilisation 3. EU/Australia/Japan approved but US TEA-pending (until Dec 2025 bemotrizinol order)

The legacy FDA filter set (octinoxate, oxybenzone, octocrylene, avobenzone) is the inverse: small MW (<400 Da), measurable plasma absorption (Matta JAMA 2019/2020 — all exceeded the FDA 0.5 ng/mL threshold), and either photolabile or with secondary safety concerns (BP-3 endocrine + reef; OC → benzophenone; OMC photolability + endocrine).

7.2 The avobenzone-octocrylene-benzophenone trilemma

US sunscreens needing UVA-I protection have, in practice, only one approved filter: avobenzone. Avobenzone is highly photolabile and effectively requires a stabiliser. The most widely-used stabiliser is octocrylene. Octocrylene undergoes retro-aldol decomposition to benzophenone over shelf life (Downs 2021): fresh products contain 6–186 mg/kg benzophenone, post-aging up to 435 mg/kg. Benzophenone is an IARC Group 2B possible carcinogen. This trilemma is a structural consequence of the FDA monograph freeze and disappears once newer filters (DHHB, Tinosorb M/S) are approved.

7.3 ESIPT is the dominant photostability mechanism

A surprising structural unifier: most photostable filters share the same photophysical mechanism — excited-state intramolecular proton transfer (ESIPT) in <1 picosecond: - Benzophenones (BP-3/4/8) — 2-OH ↔ C=O hydrogen bond - Tinosorb S / M — methoxyphenyl-triazine and hydroxyphenyl-benzotriazole - DHHB — diethylamino-hydroxybenzoyl - Mexoryl XL — 2-benzotriazolylphenol

Avobenzone is the conspicuous exception: it relies on enol↔keto tautomerism across a six-membered intramolecular H-bond, but the diketo form is photoreactive rather than dissipating to ground state. This is why avobenzone alone fails and ESIPT-class filters succeed.

7.4 Inorganic vs organic is a false dichotomy

Marketing splits "physical/mineral" from "chemical" filters. Physically: - All filters work primarily by absorption. Cole, Shyr & Ou-Yang 2016 showed only 4–5% UV reflection from ZnO/TiO₂ — absorption dominates 10:1. - Inorganics absorb via semiconductor electronic-band-gap transitions; organics via π→π / n→π HOMO-LUMO transitions. Both are absorption. - The actually meaningful distinctions are: skin penetration (inorganic ≪ organic small-MW), photocatalysis (inorganic catalyses adjacent organic decomposition; organic does not), and aesthetic feel (whitening — driven by inorganic refractive index, n=2.7 for rutile vs ~1.5 for organics in oil).

7.5 Spectral coverage gaps in practice

No single filter covers 280–400 nm uniformly. Modern SPF50+ formulae stack 4–7 filters spread across: - UVB workhorse: octinoxate / octocrylene / EHT / homosalate - UVA-II bridge: BP-3 / ecamsule / bemotrizinol - UVA-I anchor: avobenzone / DHHB / Mexoryl 400 / MBBT - Particulate / inorganic: ZnO and/or TiO₂ + TBPT

The deep UVA-I band (370–400 nm) — implicated in photoaging and pigmentary disorders — was poorly covered until Mexoryl 400 (λmax 385 nm) was approved in EU via Reg 2020/1684 (commercial rollout 2022–2024). ZnO and TBPT (with its scattering tail) help but partially. This band remains the weakest spot in current US-approved formulae specifically because Mexoryl 400 has no FDA path.

8. Methodology / Provenance

Research process

• Six parallel deep-research subagents (run 2026-05-07) using WebSearch + WebFetch on PubMed, Google Scholar, ScienceDirect, ACS, Wiley, RSC, SCCS opinion documents, FDA Federal Register, JAMA, manufacturer technical bulletins
• Each agent followed the workspace research-agent-instructions.md claim format with C1–C5 confidence tiers
• Each numerical claim cites a primary peer-reviewed source with DOI
• Data conflicts flagged (e.g. ε for ecamsule cited as 19,000–47,000 across studies) rather than averaged

File index

sunscreen-filters/
├── index.md                                      # project dashboard
├── MASTER-SYNTHESIS.md                           # this file
└── research/
    ├── uvb-organic/        (10 claim files + summary + sources)
    │   c001 Octinoxate · c002 Octisalate · c003 Homosalate · c004 Octocrylene
    │   c005 Cinoxate · c006 PABA/Padimate O · c007 Ensulizole · c008 Polysilicone-15
    │   c009 EHT · c010 4-MBC
    ├── uva-organic/        (5 claim files + summary + sources)
    │   c001 Avobenzone · c002 Meradimate · c003 DHHB · c004 Bisoctrizole · c005 Iscotrizinol
    ├── broad-spectrum/     (6 claim files + summary + sources)
    │   c001 Bemotrizinol · c002 Ecamsule · c003 Drometrizole Trisiloxane
    │   c004 Mexoryl 400 · c005 Tinosorb A2B · c006 Bisdisulizole
    ├── benzophenones/      (6 claim files + summary + sources)
    │   c001 Oxybenzone · c002 Sulisobenzone · c003 Dioxybenzone
    │   c004 Trolamine salicylate · c005 BP-1 · c006 Mexenone
    ├── inorganic/          (2 claim files + summary + sources)
    │   c001 Zinc Oxide · c002 Titanium Dioxide
    └── photostab/          (8 claim files + summary + sources)
        c001 OMC+AVB incompatibility · c002 OC stabilisation · c003 Polycrylene
        c004 DEHN · c005 Tinosorb S as stabiliser · c006 Tinosorb M
        c007 DESM/Oxynex ST · c008 Formulation factors

Aggregated data gaps (C5 — "couldn't find" items across all areas)

Compiled from per-area C5 lists: 1. Mexoryl 400 primary peer-reviewed photostability data (only sponsor data) 2. TBPT (Tinosorb A2B) singlet-oxygen quantum yield numerics 3. Iscotrizinol Φ_T and Φ_Δ values (no primary measurement located) 4. Bisoctrizole primary peer-reviewed ε (most data manufacturer-derived) 5. Ecamsule ε at 345 nm — disagreement spans 19,000–47,000 across studies 6. Cinoxate modern photodegradation kinetics 7. Polysilicone-15 Φd and Φ(E/Z) measurements 8. Mexenone (BP-10) modern ε, Φ_loss, photoproducts 9. Avobenzone individual photoproduct-channel quantum yields 10. Coral toxicity assays of homologous benzophenones (BP-4, BP-8) analogous to the BP-3 Downs 2016 study 11. All filters — long-term chronic dermal exposure studies for nano-coated forms 12. Damaged skin penetration — explicitly flagged unknown by SCCS for inorganics 13. Triplet energies and rate constants for several photostabiliser pairs (Polycrylene, DESM, MBBT) — inferred from chromophore analogy rather than directly measured

Confidence tier breakdown across the corpus

• Identity / regulatory status: predominantly C1 (primary regulatory documents, EU Annex VI, FDA monograph)
• λmax values: C1 in solvent-specified literature; mixed when comparing across studies (different solvents → different values)
• ε / E(1%, 1cm): C1–C2; substantial disagreement for some filters (ecamsule)
• Photodegradation kinetics: C1 for legacy filters with extensive study; C2 (manufacturer data) for newer filters
• Photostabilisation quantitative claims: C1 for mechanism (Sayre 2005, Schwack 1995, Chatelain 2001, Lhiaubet-Vallet 2010, Mturi 2008, Downs 2021); C2 for absolute % residuals (manufacturer-affiliated)
• Inorganic photocatalysis ROS quantitation: C1 (Hirakawa & Nosaka 2002, Wakefield 2004); C4 for some Egerton extinction values (graph-digitised)

End of master synthesis. See per-area _summary.md for category-level overviews and claims/cNNN-*.md for per-filter detail with full citation lists.