 ｬ žMoS ｭ ž ｮ NOAEL SED (39) see below. T he accepted safety limit of an MoS value should be equal or more than 100. During in vitro studies, avobenzone, octinoxate, OCR, oxybenzone, and padimate O, using mineral oil as carrier, were assessed, with respect to their skin penetration properties, and were found to cause no harm to the underlying cells of the epidermis (40). As sunscreen effi cacy may depend on vehicle formulation, some researchers investigated the vehicle effects on UV fi lter skin penetration and permeation. In vitro release and skin permeation of two widely used UV fi lters, OMC (5% w/w) and BMBM (1% w/w), form topical formulations with different features [oil-in-water (O/W) emulsions with different viscosity, water-in-oil (W/O) emulsion, and oils with different lipophilicities]. To mimic in-use conditions, experiments were carried out repeating sunscreen application on the skin surface for three consecutive days. BMBM release from all these vehicles was very low, thus leading to poor skin permeation. The vehicle composition signifi cantly affected OMC release and skin permeation, and slight increases of OMC permeation were ob- served after repeated applications. From skin permeation data, SED and MoS values of BMBM and OMC were calculated for all the investigated formulations after a single ap- plication and repeated applications. SE D was estimated, as reported by Søeborg et al. (41) and by the Scientifi c Commit- tee on Consumer Safety (2016), using the following equation: 3 2 1 2 mg mg DA 10 SSA cm day Kg body weight cm SED , day 60 kg a F  ｬ  ｬ Ng ž ž q q ｭ ｭq ž ž ｮ Ng ｭ ž ｭ ｭ ž ｭ ž wh er e DAa (μg/cm2) is the dermal absorption reported as amount/cm2, resulting from in vitro skin permeation experiments SSA is the skin surface area, expected to be treated with the formulation under investigation F is the frequency of application of the inves- tigated formulation and 60 kg is the default human body weight. The r esults showed that both vehicle composition and the pattern of application affected BMBM and OMC skin permeation. However, all formulations investigated could be con- sidered safe under in-use conditions (42). UV FI LTER SUBSTANCE CLASSIFICATION INORG ANIC UV FILTERS There are two inorganic fi lters (also known as mineral fi lters): titanium dioxide (TiO2) and zinc oxide (ZnO) both are metal oxide particles. These molecules absorb, refl ect, and refract UV photons but function in photoprotection primarily by absorbing UV radiation. The initial formulations of mineral fi lter–containing sunscreens often left a white, chalky JOURNAL OF COSMETIC SCIENCE 308

appearance on the skin, which is most noticeable in dark-skin individuals. New formulations were encouraged by decreasing the particle size, culminating in the usage of nanoparticles. Oxidative stress and cellular toxicity would be a serious concern about ZnO and TiO2 in case they could penetrate the SC, enter the dermis, and ultimately the blood supply. Fortunately, both in vivo and in vitro studies have found that systemic absorption is not possible because their permeation reaches only the SCSC. This could be due to the nanoparticles’ tendency to concentrate into larger structures (30–150 nm), which cannot penetrate into the skin (43–45). However, they might be dangerous when inhaled, and thus there use, as spray sunscreen products, containing nanoparticles, is restricted. ORGANIC UV FILTERS Organic UV fi lters belong to several organic categories that are classifi ed as UVA and UVB fi lters because of their absorption properties. A great number of organic UV fi lters contain aromatic moieties conjugated with carbonyl groups. Excited delocalized electrons of the aromatic moieties make absorption into the UV range possible. Further substitution affects the absorption strength of various UV fi lter classes. Dibenzo ylmethane (1,3-diphenyl-1,3-propanedione) Derivatives. Cosmet ic products contain the most common UVA fi lter, i.e., butyl methoxydibenzoylmethane (BMDBM), which replaced the former 4-isopropyl dibenzoylmethane (I-DBM), in accordance with the de- mands of a balanced UVA/UVB protection. The production of BMDBM was interrupted in 1993 because it was thought to be photo-allergizing. It is also implied that photodeg- radation products can also be reactive and may cause further contact allergies (46-48). However, photodegradation can be minimized, by using a UV fi lter combination or by encapsulation (16). Benzophen one Derivatives. BP-3 and BP-4 are proven to have good photostability and broad-spectrum protection against the whole UVB and UVA ranges. That is why their use is extended to sunscreens and various skincare products. However, they do not offer complete UVA protection, unless they are combined. According to various studies, BP-3 tends to permeate the human skin, leaving traces in urine and breast milk, and is consid- ered to have endocrine potential, as mentioned earlier (toxicity studies) (9,49,50). Twenty-fi ve volunteers applied a commercially available sunscreen containing 4% BP-3 morning and night for 5 d. Their urine was measured during those 5 d and further 5 d after the last application. They were divided into groups A (unirradiated) and B. Group B received UV radiation according to the skin type. BP-3 in urine was analyzed with a high- performance liquid chromatography method (HPLC). The volunteers excreted 1.2–8.7% (mean 3.7%) of the total amount of BP-3 applied. There was no signifi cant difference between the two groups (p 0.99, t-test). It has been proven that there is a positive correlation between the use of cosmetics con- taining BPs and their presence in human milk. Human milk samples were collected from mothers of three different cohorts in 2004, 2005, and 2006, who gave birth to a singleton child at the University Women’s Hospital Basel. The age of the mothers across all three study cohorts was similar, with a mean of 32.3 years. Milk samples were taken in August/ September, October, and November/December 54.72% of the study participants used sunscreens and 60.38% used other cosmetic products containing UV fi lters. UV fi lters were detected in 46 of 54 or 85.19% of breast milk samples (BP-3: 13.21% of total) (51). DISTRIBUTION OF UV FILTERS ON THE SKIN 309