Ultraviolet (UV) filters are a group of chemicals with a wide range of physicochemical properties with annual production exceeding a million tonnes worldwide. They are used as important ingredients of sunscreen agents, cosmetics, skincare and lip products and hair sprays in order to protect human skin against sunburn and cancer by absorbing or reflecting UV-A (320-400 nm) and UV-B (280-320 nm) radiation as well as to prevent UV degradation of fragrances and dyes in plastics. Inorganic UV filters block UV radiation by reflection and scattering, and therefore are also called "physical filters", while organic UV filters are aromatic molecules conjugated with carbonyl groups which absorb photo-radiation in the UV region and undergo rapid vibrational relaxation back to the ground state, also called "chemical filters". Owing to their large annual production quantities and widespread usage, organic UV filters can enter the aquatic environment (i) indirectly from wastewater treatment plants (WWTPs) after entering wastewater systems following bathing or from industrial discharge and (ii) directly from recreational activities (e.g. swimming). There is increasing concern about their potential environmental impacts because of their toxicity (e.g. estrogenic activity, negative effects on reproduction in Japanese medaka (Oryzias latipes) and growth inhibition in green algae (Desmodesmus subspicatus)), bioaccumulation and ubiquity in the environment.
Though their occurrence in various environmental matrices such as surface water, sediment and biota has been reported, these studies mainly focused on European countries and only limited classes of organic UV filters were studied. As a result, the principal objective of the present study is to develop analytical methods for the simultaneous quantification of multiple classes of organic UV filters, investigate the occurrence and distribution of these chemicals from sources (i.e. wastewater in WWTPs), surface water, sediment and biota in Hong Kong and globally (e.g. China, the United States, Japan, Thailand, and the Arctic), as well as to conduct environmental risk assessments of the potential effects of organic UV filters on aquatic organisms at different trophic levels such as algae, invertebrates, fishes and corals. The first stage of this study involved (i) developing quantitative analytical methods for simultaneous multiclass determination of 12 commonly globally consumed UV filters in wastewater, including benzophenone-1 (BP-1), benzophenone-3 (BP-3), benzophenone-4 (BP-4), benzophenone-8 (BP-8), butyl methoxydibenzoylmethane (BMDM), ethylhexyl methoxycinnamate (EHMC), ethylhexyl salicylate (EHS), homosalate (HMS), isoamyl p-methoxycinnamate (IAMC), 4-methylbenzylidene camphor (4-MBC), octocrylene (OC) and octyl
dimethyl-p-aminobenzoic acid (OD-PABA), (ii) determining their concentrations, seasonal occurrence as well as removal efficiencies in wastewater samples collected from five WWTPs in Hong Kong featuring different treatment levels including both conventional (e.g. primary sedimentation, chemically-enhanced primary treatment, and biological treatment) and advanced treatment (e.g. chlorination, UV-disinfection, sand filtration, and reverse osmosis) over the course of one year, and (iii) conducting a preliminary environmental risk assessment of these compounds in the Hong Kong environment by using the predicted environmental concentrations based on effluent concentrations and available toxicity data. UV filters were quantified by the newly developed simultaneous multiclass detection method using solid phase extraction (SPE) coupled with high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). Among the 12 target compounds, BMDM, BP-1, BP-3, BP-4 and EHMC were frequently (≥80%) detected in both influent and effluent with mean concentrations ranging from 23-1290 ng/L and 18-1018 ng/L, respectively; less than 2% of samples contained levels greater than 1000 ng/L. Higher concentrations of these frequently detected compounds were found during the wet/summer season, except for BP-4, which was the most abundant compound detected in all samples in terms of total mass. The target compounds behaved differently depending on the treatment level in WWTPs; overall, removal efficiencies were greater after secondary
treatment when compared to primary treatment with >55% and <20% of compounds showing high removal (defined as >70% removal), respectively. In general, higher removal was observed for the compounds with relatively higher octanol–water partition coefficients (log Kow) values. Reverse osmosis was found to effectively eliminate UV filters from effluent (>99% removal); however, the operational costs of large-scale reverse osmosis treatment are likely to be prohibitive, particularly in developing countries. Because of the incomplete attenuation of UV filters, a preliminary risk assessment was conducted which indicated that BP-3 and EHMC discharged from WWTPs may pose high risk to fishes in the Hong Kong environment.
The second stage of this study focused on (i) determination of the concentrations, spatial occurrence and distribution of these 12 UV filters in surface water samples, especially seawater samples, collected from countries including China (Hong Kong, n=60; Shantou, n=4 and Chaozhou, n=3), United States (New York City, n=6 and Los Angeles, n=4), Japan (Tokyo Bay, n=8), Thailand (Bangkok, n=2) and the Arctic (n=14), as well as their temporal distribution in Hong Kong over the course of a year; and (ii) conducting an ecological risk assessment for both the worst-case scenario and using a probabilistic approach. This study is the first to report the occurrence of UV filters in marine surface waters in these locations. The number of
compounds detected, Hong Kong (12) > Tokyo (9)=Bangkok (9) > New York (8)=Los Angeles (8) > Arctic (6) > Shantou (5)=Chaozhou (5), generally increased with population density. The presence of these compounds in the Arctic might be due to a combination of inadequate wastewater treatment and oceanic/atmospheric long-range transport. The median concentrations of all detectable UV filters were below 250 ng/L while BMDM, BP-3 and EHMC were widely distributed in all cities and in the Arctic with detection frequencies ≥ 30%. BP-4 did not show temporal differences in the Hong Kong samples, likely because it is an ingredient not only in sunscreens but also in personal care products (PCPs) without seasonal usage. Probabilistic assessment indicated that the risk of BP-3, EHMC and 4-MBC causing growth inhibition of green algae (Desmodesmus subspicatus) and affecting egg development of Japanese medaka (Oryzias latipes) was generally < 50%, while BP-3 and EHMC could pose high risk to hard corals (Acropora sp. and A. pulchra) in the form of coral bleaching in the worse-case scenario. The ubiquitous occurrence and global distribution of organic UV filters reflects the impacts of human use of PCPs on aquatic environments, while the potential ecological effects of these compounds as indicated by probabilistic risk assessment provides an early warning about the release or transport of these emerging contaminants, particularly to remote regions such as the Arctic. Because of their ubiquity, high lipophilicity and inadequate occurrence data in the marine environment,
it is also necessary to develop quantitative analytical methods for detection of organic UV filters in sediment and biota samples in order to understand their environmental partitioning as well as environmental fate, and to more fully assess their potential negative impacts and risk to marine ecosystems.
The third stage of the study thus focused on (i) development of quantitative analytical methods for simultaneous multiclass determination of 11 commonly globally consumed UV filters (all of the compounds measured previously except BP-4) in sediment by using pressurized liquid extraction (PLE); (ii) determination of the occurrence, spatiotemporal distribution and composition profile of the target chemicals in marine sediment samples collected throughout Hong Kong during August 2012, February and June 2013, and Tokyo Bay during July 2013; and (iii) conducting a probabilistic ecological risk assessment using the measured concentrations. In Hong Kong, seven of the 11 target UV filters were detected in the samples, with median concentrations ranging from <LOD to 21 ng/g. Generally, the detection frequencies of UV filters were higher in the wet season than in the dry season, and BMDM, BP-1, BP-8 and EHMC showed detection frequencies >50%. All compounds were detected at similar frequencies in locations associated with direct and indirect sources, except for BMDM and OD-PABA, whose levels were 40% higher at locations associated with direct sources compared with those associated with indirect sources. Composition profiles showed that BMDM, EHMC and OD-PABA were the predominant compounds in sediment samples accounting for more than 60% of the total UV filter occurrence in each season, while all of the benzophenone derivatives accounted for <35%, likely due to their relatively lower estimated soil organic carbon partition coefficients (log Koc) values (benzophenone-derivatives: 2.8-3.8; other detectable compounds: 4.4-6.8). Probabilistic ecological risk assessment showed that the likelihood of EHMC causing toxic effects on reproduction in snails was over 84% and 32% for two species, respectively, suggesting potential risks of UV filters to benthic organisms as well as possible wider effects on the marine food web.
As the first three parts of this study indicated the ubiquity of UV filters in the aquatic environment, their bioaccumulation and occurrence in marine organisms should be investigated. As a result, an analytical method was developed for quantifying 4-MBC in zebrafish embryos and larvae using liquid-liquid extraction by n-hexane, nitrogen concentration and quantification by HPLC-MS/MS; matrix spike recoveries including relative standard deviation (R.S.D.) were 92% ± 6.6. A method for simultaneous extraction and quantification of multiple classes of UV filters in biota samples (i.e. fish and bivalves) was partially optimized by using PLE, followed by SPE coupled with HPLC-MS/MS. Several PLE parameters including extraction temperature, static time, number of extraction cycles and packing material for in-cell purification were optimized, and the method was validated in field-collected green-lipped mussels (Perna viridis). The optimized method was also applied to analysis of several fish species collected from Hong Kong and Japan.
A complete study of the occurrence, distribution and ecological risk of organic UV filters from sources to environment has been presented in this thesis. The widespread occurrence and potential ecological risks of these compounds in the aquatic environment should raise further concern about (i) improving removal efficiencies from sources such as WWTPs; (ii) regular environmental monitoring and risk assessment, particularly the need for more toxicological information for marine species; (iii) maximum authorized levels in PCPs; and (iv) possible toxicity of their transformation products.
| Date of Award | 3 Oct 2014 |
|---|
| Original language | English |
|---|
| Awarding Institution | - City University of Hong Kong
|
|---|
| Supervisor | Margaret Burkhardt MURPHY (Supervisor) & Kwan Sing Paul LAM (Co-supervisor) |
|---|