A Systematic Review, Narrative Synthesis and Quantitative Comparison on Skin Penetration and Experimental Evidence of Metal and Metal oxide Nanoparticles in Cosmetic Related Applications

Abstract

Background. Metal and Metal oxide nanoparticles (NPs) (ZnO, TiO₂, Ag, Au, CeO₂, SiO₂, Al₂O₃) are used in sunscreens, anti ageing creams, antimicrobial cosmetics, oral care products. The evidence on whether they cross the skin barrier, how their safety and efficacy differ is scattered.
Objective. To find, summarise, and compare the experimental, clinical, and case study evidence on these NPs in cosmetic and dermal exposure contexts along with focusing the direct comparisons between the types of NPs.
Methods. From the PubMed and Scopus databases searched (finalised September 2026) using the nanoparticle specific and cosmetic/skin terms, following PRISMA 2020. The Records were screened in two stages. The eligible studies were primary research, clinical trials, or case studies. The Reviews and conference abstracts were excluded. Data for the quantitative synthesis were taken from the full texts where available. The Comparisons that included head to head experiments, dose/structure response within studies, cross study numeric comparisons, and evidence density mapping.
Results. From 953 records (150 - PubMed, 803 - Scopus), 39 studies were eligible (ZnO - 25, TiO₂ - 8, Ag - 5, Au - 4, SiO₂ - 1, CeO₂ - 1, Al₂O₃ - 0) in that 8 numeric comparisons were found. ZnO NPs showed stronger antibacterial activity than TiO₂ NPs, while TiO₂ NPs had slightly higher antioxidant activity. ZnO NPs penetration increased 60–65 fold in barrier impaired skin, while TiO₂ NPs deposition was similar in intact and damaged skin. SiO₂ NPs coating reduced ZnO NPs cytotoxicity across multiple endpoints. Ag NPs and ZnO NPs were positive for skin sensitisation, with size dependent potency for Ag NPs. TiO₂ NPs was used more often than Au NPs in marketed cosmetics products. The Co exposure of ZnO NPs and TiO₂ NPs produced antagonistic toxicity. CeO₂ NPs reduced the organophosphate penetration 3–4 fold in a barrier cream.
Conclusions. Evidence is concentrated on ZnO NPs and TiO₂ NPs, mostly In vitro and animal models. ZnO shows stronger barrier status dependence than TiO₂ NPs. Al₂O₃ NPs, SiO₂ NPs, and CeO₂ NPs are under studied. Direct experimental comparisons give the most reliable cross nanoparticle insights. A pooled meta analysis was not possible due to limited numeric data, and this is stated clearly.