UV Filter Chemistry for Accurate Dose–Response Relationships

The workshop’s first day focused on the challenges of analyzing and quantifying the structural changes UV filters undergo as they disperse into the environment, along with opportunities to improve understanding of the environmental fate, or final product or effect of these chemicals in an environment. Two speakers set the stage with perspectives on current scientific findings and methods; these speakers were then joined by three additional panelists for a focused discussion examining areas where progress has been made and opportunities for further improvement. Workshop participants then explored these issues further in small breakout groups.

THE ENVIRONMENTAL FATE OF UV FILTERS

Silvia Díaz-Cruz (Institute of Environmental Assessment and Water Research, Spanish National Research Council) reviewed current knowledge about how UV filters are released and their environmental fate. She said that many studies have confirmed the direct release of organic and inorganic UV filters from surface-water contact activities, with increased concentrations correlating with increased activity (such as more people swimming in the summer). UV filters can also be released indirectly from stormwater runoff, industrial runoff, or wastewater treatment plants. The data are extremely limited for all of these mechanisms, however, and there is even less known about potential contributions resulting from illegal dumping or discharge. An added complexity is that UV filters can occur naturally in water and sediment, and are found in many consumer products besides sunscreens, making it challenging to definitively determine sunscreen’s contribution to water contamination.

The environmental fates of UV filters vary depending on the physico-chemical properties of the particular UV filter involved, such as its water solubility, volatility, dissociation, hydrophobia, and adsorption, as well as the hydrography and water dynamics of the area in which it is released. Organic UV filters are generally hydrophobic; instead of dissolving in water they partition to particles and sediments. Inorganic UV filters are more likely to aggregate in water and land in sediments, a process that happens faster as water salinity increases.

There is limited information on how UV filters’ fates affect the environment. Impacts likely vary depending on environmental parameters such as air temperature, water temperature, salinity, and pH level, as well as how the UV filter interacts with other molecules, its particle size, and its potential for aggregation (see Figure 1).¹ UV filters can disrupt a variety of physical and chemical processes in an environment, such as through direct or indirect phototransformation, or produce potentially toxic reactive oxygen species such as oxygen peroxide. UV filters can also affect biological processes through bioaccumulation in prey, biomagnification in predators, or biotransformations in aquatic microorganisms.^2,3

___________________

¹ Schwarzenbach, R. P., Escher, B. I., Fenner, K., Hofstetter, T. B., Johnson, C. A., von Gunten, U., & Wehrli, B. (2006). The challenge of micropollutants in aquatic systems. Science (New York, N.Y.), 313(5790), 1072–1077. https://doi.org/10.1126/science.1127291.

² Bar-On, Y. M., Phillips, R., & Milo, R. (2018). The biomass distribution on Earth. Proceedings of the National Academy of Sciences of the United States of America, 115(25), 6506–6511. https://doi.org/10.1073/pnas.1711842115.

³ National Academies of Sciences, Engineering, and Medicine. 2022. Review of Fate, Exposure, and Effects of Sunscreens in Aquatic Environments and Implications for Sunscreen Usage and Human Health. Washington, DC: The National Academies Press. https://doi.org/10.17226/26381.

Given these unknowns, Díaz-Cruz emphasized the need for further research. In particular, she said it is critical to study the potential toxicity of these chemicals, including their metabolites and other transformation products; their biodegradability; the potential for long-chain biomagnification and bioaccumulation; and their impacts on aquatic plants and on critical body burdens to understand long-term biota stress. She also pointed to a particular need for studies to address the impact of inorganic UV filters on soil and aquatic organisms and to elucidate how UV filter coatings, such as aluminum, silica, and polydimethylsiloxane aggregate or dissolve.

Noting that laboratories cannot accurately recreate environmental conditions, Díaz-Cruz said it will be important to conduct field studies to address these questions, including consideration of the specific water dynamics of particular areas. To advance this work, she said there is also a need for researchers to develop and share standardized methods, field blanks, quality controls, and reliable laboratory-based or experimental bioconcentration factors and bioaccumulation factors. In addition, she suggested that monitoring programs should include repeated, replicable measurements over space and time for accurate occurrence data.

ANALYTICAL CHALLENGES IN QUANTIFYING ORGANIC UV FILTERS

Michael Gonsior (University of Maryland) discussed considerations regarding the extraction and quantification of organic UV filters and described particular challenges related to the

tautomeric behaviors (when chemical compounds exist in various confirmations simultaneously in a mixture), instability, and solubility of these chemicals (see Figure 2). He noted that these factors impact not only toxicity but also identification; researchers seeking to study a particular compound must consider the possibility that it may have changed due to reaction or degradation, confounding attempts to understand its true fate.

Organic UV filters can transform into tautomers that have very different behaviors in different environments. In addition, UV filters are not stable and tend to degrade over time. Gonsior said that laboratory storage standards are needed to avoid degradation and account for interaction, photoreaction, pH levels, and transesterification, a process in which interacting compounds exchange parts.⁴ Quantifying the solubility of UV filters is critical to understanding their environmental fate, he said, but little solubility data are available, and solubility can also vary depending on the environment. In addition, UV filter particles can create a microlayer on the sea surface or other hydrophilic photoproducts instead of dissolving.

___________________

⁴ Holt, E. L., Krokidi, K. M., Turner, M. A. P., Mishra, P., Zwier, T. S., Rodrigues, N. d. N., & Stavros, V. G. (2020). Insights into the photoprotection mechanism of the UV filter homosalate. Physical Chemistry Chemical Physics, 22(27): 15509–15519. https://doi.org/10.1039/D0CP02610G.

Extracting UV filters from environmental samples is a key challenge, and Gonsior pointed to a need for extraction standards covering parameters such as whether samples should be filtered. In particular, the use of different isolation techniques, the process of surface adsorption, and contamination from insufficient cleaning or drying can affect results and undermine the ability to compare findings across studies. He suggested that researchers should use EPA recovery standards to report adsorption loss, which some compounds are more prone to than others. In addition, he said that adopting techniques developed to analyze per- and polyfluoroalkyl substance (PFAS) could help to minimize contamination. Gonsior also described a solid-phase extraction method his team developed and applied to measuring 12 common UV filters using chromatography with optimized electrospray ionization. He said the method minimizes adsorption and contamination, avoids false positives, and enables quantification of most organic UV filters at nanomolar detection limits, including curated isotope standards. For future work, he said the team’s next priority is to incorporate recovery standards.

PANEL DISCUSSION

Scott Belanger (Procter & Gamble, retired), brought Díaz-Cruz and Gonsior together with three additional panelists to discuss opportunities to overcome the challenges in order to improve understanding the environmental fates of UV filters and inform the chemical analyses necessary to advance ERAs for these chemicals. The additional panelists were Jon Arnot (ARC Arnot Research & Consulting), Bill Mitch (Stanford University), and Kurt Reynertson (Johnson & Johnson Consumer Health). Each panelist offered opening remarks regarding areas of progress and key challenges.

Building on her previous comments, Silvia Díaz-Cruz reiterated the need to study transformation products that react to species and affect toxicity. She said it will also be important to utilize instrumentation including high-resolution mass spectrometers, to develop compound libraries or databases to identify what is visible, to conduct non-target screenings and experiments with the environmental site water and its hydrodynamics, and to further investigate biomagnification higher up the food chain.

Michael Gonsior said that progress has been made in quantifying and analyzing UV filters. To move forward, he reiterated the need to standardize measurement approaches, incorporate EPA recovery standards for different compounds, and conduct lab-comparison studies to produce consistent, reliable, and reproducible data for evaluating toxicity and environmental risk. He also highlighted the value of viewing nontarget screenings across multiple matrices such as freshwater vs. seawater or microlayer vs. dissolved, modeling for toxicity, improving understanding of how filters dissolve or partition, establishing best practices for extraction, and developing better approaches to measure biomagnification.

Jon Arnot highlighted the need for more research into the partitioning of ionizable organic chemicals to solid phases in aquatic environments, including better models of adsorption coefficients and partitioning behavior. He also noted that a better understanding of all of the chemistry involved, such as the implications of tautomerization, is needed in order to extrapolate testing results to environmental effects. Finally, he said that it is important to further study biomagnification in coral, plankton, and other lower-trophic species, though he noted that progress has been made on understanding biotransformations, which will shed light on biomagnification in the food chain.

Bill Mitch stressed the need for new approaches to measure exposure, toxicity, and aqueous concentrations of UV filters and to identify the primary exposure pathways. He said that more

research is also needed to understand phase 2 metabolites, create phase 2 standards, and standardize toxicity tests, especially with regard to lighting conditions. He noted that passive sampling has progressed to capture UV filter effects in the water column, provide integrative exposure estimates, and enable measurements of phase 1 metabolites and biomass concentration, which could create a measurable steady-state concentration in tissue.

From the perspective of a sunscreen manufacturer, Kurt Reynertson said that there has been progress in analytical chemistry and contamination identification, but he noted that formulating effective sunscreens that stick to the skin remains a key focus and challenge for product developers. The formulations are complex mixtures of ingredients that can affect UV filters’ route into the environment, their fate, and their uptake by organisms. While understanding the intrinsic qualities of discrete elements like UV filters is important, Reynertson posited that it is even more important to understand the mechanisms and potential matrix effects of these complex formulations.

Following panelists’ opening remarks, Belanger moderated a discussion addressing testing challenges, how UV filters compare to other chemicals, and opportunities to advance research.

Testing Challenges

Testing UV filters in laboratory settings has multiple challenges that could be overcome so that researchers can, in Belanger’s words, “apply the right tool for the right question at the right level of specificity and at the right level of detection.” When asked how to minimize sample loss during laboratory testing, Gonsior replied that while some UV filters are more prone to loss than others, if the loss can be identified and quantified, it can be mitigated. Reynertson agreed, noting that it is important to know which compounds are susceptible to loss. Díaz-Cruz added that using weighty standards, in addition to internal standards, during testing enables both loss descriptions and recovery experiments. Mitch noted that for toxicity tests, even test tubes can induce loss, making the critical point—what amount will enter an organism—difficult to measure. He suggested that researchers should use different equipment or passive sampling. Díaz-Cruz agreed but noted that for UV filter analysis, concentrations cannot be interpreted from passive sampling unless the sampling has been validated and at least three or four samples are taken, to account for water dynamics.

Belanger asked if formulation testing of sunscreen would be more valuable than single-chemical testing, despite the analytical challenges. Reynertson said it would, adding that it is these compounds’ combination effects that make it especially hard to pinpoint effect sources and availability to organisms. Formulation testing methods differ by company but typically start with single ingredients and then gradually increase complexity. Belanger noted that understanding toxicity is harder when formulas have multiple components and therefore multiple modes of action. In addition, he said, passive dosing methods are not adequately standardized to enable appropriate risk assessment.

How UV Filters Compare to Other Chemicals

UV filters are among many human-made chemicals that can be directly or indirectly transferred into the environment. Mitch noted that his team studies wastewater effluent to determine what compounds are present and in what quantities, but these studies lack predetermined toxicity levels for sunscreens. He and Díaz-Cruz agreed that contaminant quantities are less important than concentration and toxicity levels. Arnot and Belanger noted that while quantification and exposure-route mapping are relevant, it is important to bring the focus back to environmental effects

testing for UV filters. For example, UV filters can create a surface microlayer, where Gonsior and his team have found a dramatic concentration factor that can vary depending on wave action. They also found that it is possible for this microlayer’s hydrophobic, intertwined chemicals to directly expose coral to these chemicals.

Belanger asked how UV filter compounds are different from other ionizable compounds. Gonsior replied that all ionizable compounds are challenging to work with, and he reiterated that suppression and recovery standards are needed. For example, esters can fall apart during ionization, which occurs differently in seawater and freshwater. Arnot stressed the urgent need to expand understanding of ionization in aquatic and biological test systems to overcome the challenges of studying hydrophobic chemicals, an emerging research area that may be applicable in this space. He also suggested passive dosing, in addition to passive sampling and traditional analytical methods, to learn more about solvent effects and solubility from the samples. Belanger agreed that understanding the timeline of equilibrium and degradation would present important empirical evidence. Sascha Pawlowski (BASF) noted that in addition to passive dosing, saturation columns provide the consistent exposure levels needed to test and create single-compound regulations.

Participant Suggestions to Advance Research

Reynertson underscored the need for more research into how a formulation—not just a particular compound—enters an environment. This information is critical to designing studies of true environmental relevance and toxicity because entry mechanisms also affect UV filter partitioning and organism uptake, he said. Mitch agreed and noted that formulation concentrations can be higher than individual components. He also suggested that a new, integrative approach to measuring exposure is needed, as water concentration can vary by time and space.

To create the robust data that are needed for formal ERAs, Gonsior said that studies need to be further standardized, optimized, and made reproducible. Passive and temporal samples would create a more realistic context, and he suggested that toxicology studies should be paired with analytical chemistry to uncover more of the experience of aquatic organisms. Díaz-Cruz agreed that study standards and protocols are needed, including for sample collection, which can be combined with monitoring to identify what is being measured and determine whether its bioavailability is too low to be taken up by organisms. In addition, she noted that filtering samples will impact the exposure contribution and should be avoided.

Belanger noted that another key area for future research is how UV filters transform during collection, and Gonsior agreed, adding that it is also important to determine how quickly such transformations occur. Mitch added that identifying each compound’s mechanism of toxicity is another important research area. To guide well-defined experimental test systems and create useful, relevant information about environmental conditions, Arnot suggested seeking a mechanistic understanding of both toxicokinetics and toxicodynamics to understand discrete organic chemicals and extrapolate their exposure to low-level organisms such as coral, which are generally not well understood. Those studies can then be used to measure internal and external concentrations in the test systems.

Asked to discuss test equipment that could be useful for studies in this area, Gonsior replied that he uses CDN Isotopes for readily available deuterated standards, which offers the suppression, retention, time, and confirmation for each individual compound needed to avoid false positives and optimize chromatography. He also noted that for different solvents, nuclear magnetic resonance and mass spectrometry could be used to measure the concentrations of tautomers, although quantification is somewhat difficult due to degradation.

BREAKOUT DISCUSSIONS

Participants divided into small groups for a deeper dive into specific questions about the challenges of working with UV filters, areas of progress, and examples of key needs to help address the remaining gaps. Representatives from each group summarized the outcomes of these discussions, which are combined and summarized in the sections below.

What Are the Main Chemistry Challenges When Working with UV Filters?

To inform ERAs, several participants emphasized it may be vital to further elucidate the physical and chemical properties of UV filters and the changes they undergo once released into the environment. This includes characteristics and processes such as solubility, partitioning, interactions, transformations, and biodegradation, all within the context of water dynamics and environmental conditions such as salinity. In addition, it is important to understand exposure and delivery pathways, uptake routes, and the role of metabolites and micelles. Many factors complicate a full understanding of these processes, including the wide variability in sunscreen-use patterns by region, variation in runoff amount contributed by different delivery methods, and the challenges inherent in studying contaminants in low concentrations.

Many participants pointed to a lack of interlaboratory standardization of testing, sampling methods, or quality assurance/quality control protocols as a crucial limitation to the research community’s ability to close knowledge gaps. Some participants noted that standardization is likely important to appropriately interpret and compare results, especially given the complexities of sunscreen formulations and the fact that research is carried out in multiple laboratories. Recognizing that UV filters are a diverse group of chemicals, some participants said that optimized and verified standards for extraction, handling, and analysis may be beneficial. A few participants also suggested that standardized protocols could include best practices matrices, including for the most sensitive taxa, as microbial organisms metabolize or degrade compounds very differently.

In order to be useful for creating standards and informing policy, several participants expressed a desire for results to be comparable, defensible, and reproducible via multiple methods. However, some participants identified a number of impediments to this goal. For example, in sample testing, one challenge is that each compound has different adsorption rates into glass or plastic labware; another is that the compounds or the solvents may have different purity rates, which affects toxicity; a third is the potential for contamination. Testing acute toxicity is difficult because it is hard to extrapolate bioavailability, toxicodynamics, and toxicokinetics between lab-based and environmental monitoring. Testing chronic or episodic exposure or toxicity might be an easier route, perhaps focusing on marine species that are easiest to analyze, though several participants said that studies of both chronic and acute exposure are important.

A few participants also said that testing could include a minimum level of reporting, recoverable steps, an understanding of loss potential, pulse- or steady-state dosing, and a consideration of potential organism–compound interactions. One idea is to take a “subtraction” approach; instead of studying how these chemicals add to an environment, researchers could instead remove them and measure any response. Another option is to develop chemical tracers and standards, similar to the work done to support PFAS assessments.

Are Challenges Magnified When Testing Under Certain Conditions?

Some participants emphasized that testing for contamination by UV filters and related chemicals is a challenging and time-consuming undertaking, whether in the laboratory or in the

field. These challenges are compounded by the complexity of environmental variables, such as pH and light levels, which cause different matrix effects and confound measurements. In both laboratory and field settings, conditions can make it particularly difficult to reproducibly study solubility, partitioning, interaction, metabolization, transformation, and concentration fluctuations. In addition, measuring sediment exposure, tissue body burdens, and UV absorption rates from different light sources can further complicate studies.

Testing in the water column amplifies the difficulties of extraction, detection, and determining solubility, partitioning, and toxicity concentrations (a key data point for ERAs, along with temporal resolutions and randomized and diel studies). Testing is easier in a laboratory, some participants noted, but the results would be most useful if they were reproducible, quantifiable, representative of environmental systems and variability, and scalable. Then the results could be used to create a “reference state” applicable across different settings, conditions, and laboratories.

What Progress Is Being Made in Addressing These Challenges?

Progress is being made, and several participants suggested the research community could continue to focus on what is known in order to help advance assessments. In terms of addressing knowledge gaps, specific areas of progress discussed by participants include findings related to photostability, biodegradation, and metabolite identification and quantification. In terms of research methods, some participants pointed to the use of reference materials, the development of deuterated compounds and standards, the development of methods for testing in natural conditions, and PFAS passive sampling methods as useful steps forward. In addition, analytical chemistry is improving simultaneous assessment and quantification of compounds, which could help to overcome partitioning, concentration, and toxicity measurement challenges, a few participants noted.

Finally, many participants said that different points of view are beginning to converge and that in general the research community may have a clearer focus on identifying information needs and formulating problems to study. As guidelines and more reproducible ERA data are gradually becoming available, some participants noted that there is a growing awareness that this issue may benefit from significant cross-disciplinary collaborations to make further progress.

What Standardizations, Innovations, and/or Other Focused Efforts Are Suggested to Move Forward on Addressing These Challenges?

Some participants considered a number of opportunities to overcome fragmentation in this area of research and begin to address the remaining gaps and barriers. For example, a participant suggested focused, collaborative efforts by analytical chemistry developers and users to create interlaboratory standardized, centralized, shareable, reliable, and reproducible analytical methods and/or best practices for monitoring, sampling, transport, storage, data usage, reporting, and quality assurance/quality control procedures for studies of UV filter contamination. Other participants suggested a synthetic matrix and standard reference materials, such as for organisms, seawater, water depth level, or background concentration levels, as well as interlaboratory comparison studies and high-resolution mass spectrometry image and video tools. Identifying where biological effects happen, identifying or mitigating cross-contamination, and improving understanding of risk were also suggested as important areas of focus.

To avoid duplication of effort and errors, some participants stressed the benefits for open and transparent sharing of data and methods—for example, by including detailed information

about study methodology as supplementary information with published studies, as well as by sharing methods and results from unpublished or “failed” studies. Finally, a few participants underscored the importance of collaboration among government (especially EPA), academia, and industry, supported by appropriate policy and funding along with transparent public reporting on research investments and results.

What Are Existing Research Programs, Capabilities, and Infrastructure That Can Contribute to Addressing Gaps in Research on UV Filter Chemistry?

Participants considered roles for a variety of organizations in enabling further progress toward understanding UV filters and their potential environmental impacts. For example, the National Institute of Standards and Technology could help create needed standards, organizations such as EPA and the Organization for Economic Co-operation and Development could create and disseminate guidelines and tools, and EPA, FDA, and other federal agencies could sponsor grants, workshops, and other initiatives to incentivize and support this work.

Rather than working on individual projects in a vacuum, some participants urged the research community to focus on fostering a coordinated, cross-disciplinary scientific community. Organizations could pool their resources to leverage funding and expertise, and groups already working on exposure, risk, and benefits assessments could work to build stronger and more global collaborations, many participants suggested. Models for multidisciplinary, multisector collaborations exist, including in the area of sunscreens where the research community came together to establish standards around sun protection factors (SPFs).

Examples of particular organizations and programs that could be leveraged to help advance collaborations around UV filters identified by participants include Canada’s Experimental Lakes Area, EPA’s Safe and Sustainable Water Resources and Chemical Safety for Sustainability research programs, the Florida Department of Environmental Protection, the U.S. Geological Survey, PFAS research programs, and the International Collaboration on Cosmetic Safety. Finally, several participants suggested leveraging the resources of industrial equipment manufacturers like SCIEX and Phenomenex.