© A.R. De Carvalho et al., Published by EDP Sciences 2025

1 Introduction

The rapid deployment of photovoltaic energy is essential for global decarbonization, necessitating innovative solutions such as solar floating photovoltaic (FPV) systems (Almeida et al., 2022; Sahu et al., 2016). FPV systems, installed on water bodies, offer significant advantages including improved efficiency due to cooling effects, reduced water evaporation and land use optimization (Dörenkämper et al., 2021; Redón Santafé et al., 2014). Covering relatively small fractions of global reservoirs with FPV would be sufficient to help countries achieve 2050 solar energy targets (Almeida et al., 2022). It was recently identified that 643 FPV systems are currently installed on inland water bodies across the globe, most of them located in Asia (84.7%), followed by Europe (10.9%). An FPV system covers on average 34%, and up to 90%, of a lake surface (Nobre et al., 2024). Shading large portions of reservoirs could alter their biodiversity and functioning, affecting photosynthetic organisms, reducing oxygen levels, and potentially increasing methane production (Ray et al., 2024). Similarly, it might dramatically alter the air-water interface, with subsequent implications for air-water fluxes and physical, chemical, and biological properties of the recipient water body (Almeida et al., 2022; Armstrong et al., 2020).

In whole-lake experiments, the presence of FPV reduced water temperature by up to 3 °C, which could potentially decrease dissolved oxygen and alter the carbon cycle, including greenhouse gas emissions (Nobre et al., 2025). Substantial ecological changes under FPV installations, such as the massive development of benthic cyanobacteria (blue-green algae), colonization of floaters by Dreissena mussels, and an increase in the biovolume of planktonic cyanobacteria have also been reported (Sandrini et al., 2025). These impacts vary depending on the components of FPV system, its surface coverage, water body characteristics, and local climatic conditions. Although recent research have led to a better understanding of the influence of FPV on water quality and aquatic ecology (Nobre et al., 2025; Sandrini et al., 2025; Wei et al., 2025; Yang et al., 2024), the potential cascading effects on freshwater biodiversity and ecosystem functioning have yet to be studied. Moreover, far fewer studies have investigated the environmental impacts caused by the potential leaching of contaminants from FPV systems under real-world conditions (Akomea-Ampeh et al., 2025).

This paper first examines the composition of FPV materials and the degradation pathways that may lead to the release of contaminants, including metals, polymer-derived compounds, and transformation products. It then explores analytical techniques for detecting and characterizing these substances, highlighting key challenges and the need for multifaceted approaches. In this context, this paper discusses available analytical tools for broadening the chemical space, aiming to maximize the detection of diverse compounds, with a focus on the role of non-target approaches. Additionally, specialized techniques for detecting micro- and nanoplastics, as well as metals, are considered. Finally, this paper underscores the need for robust analytical strategies, carefully experimental designs, and field studies to monitor water quality effectively, ensuring that the sustainable deployment of FPV technology is aligned with environmental protection.

2 Structure and chemical composition of floating photovoltaic (FPV) systems

FPV systems combine a photovoltaic (PV) technology with a floating structure, consisting of plastic floats and a mooring system that hold them on the water body (Sahu et al., 2016) (Fig. 1A). PV modules (Fig. 1B) are based on electronic semiconductors, particularly crystalline silicon (c-Si), representing up to 80% of the market share, or thin-film semiconductor (cadmium telluride (CdTe), copper indium gallium selenide (CIGS)). These modules are integrated into a laminate structure composed of a frame, module packaging (glass front cover, encapsulant, backsheet), internal circuit (electrodes, interconnects), bypass diodes and junction boxes, cables and connectors (Aghaei et al., 2022; Al-Ezzi and Ansari, 2022). Typical c-Si PV panels contain, by weight, about 76% glass (panel surface), 10% polymer (encapsulant and backsheet foil), 8% aluminum (mostly the frame), 5% silicon (solar cells), 1% copper (interconnectors), and less than 0.1% silver (contact lines) and other metals (mostly tin and lead) (Weckend et al., 2016).

The backsheet is a critical component of solar panels, consisting of a multilayer structure that provides electrical insulation and protection against moisture and UV radiation. Fluoropolymer-based backsheets, such as those made from polyvinyl fluoride (PVF) or polyvinylidene fluoride (PVDF), offer exceptional durability and chemical resistance, representing at least 70% of the market share (Chunduri and Schmela, 2023). The core layer is most often polyethylene terephthalate (PET) and the inner layer (cell-side) is usually an ethyl-vinyl-acetate (EVA) film (Aghaei et al., 2022). Moreover, fluoropolymer coatings can be applied to various components of solar panels to enhance their resistance to weathering, abrasion, and chemical attack, thereby extending the lifespan of the panels. Fluoropolymers represent a subset of PFAS (per- and polyfluoroalkyl substances). Due to their hydrophobic properties, PFAS are widely used in water- and soil-repellent applications, including in electronic products (Glüge et al., 2020). A recent literature review highlighted several potential applications of PFAS in PV modules, including self-cleaning and anti-reflective coating, frontsheet and backsheet materials, insulating cables and wires, and as binding agent. The review also highlighted a significant lack of information regarding the specific types of PFAS used, their quantities, layer thickness, and the characteristics of the components in which their are incorporated (Nain and Anctil, 2025). Fluoropolymer-free backsheets, composed of materials like polyamides, modified PET, and polypropylene (PP), are gaining attention while becoming largely available (Chunduri and Schmela, 2023).

The floats are typically made of HDPE (high density polyethylene), known for its tensile strength, maintenance free, UV and corrosion resistance. However, other plastic materials can be used for construction of floating platforms, such as glass fiber reinforced plastic (GRP) (Sahu et al., 2016).

Fig. 1 (A) Multiple plastic floats support the PV cells. (B) An example of a common configuration of PV modules for crystalline silicon (c-Si). A glass top layer is followed by encapsulant layers that protect the solar cell and their internal circuitry. A backsheet provides insulation and protection at the bottom, while a frame encases the entire structure for mechanical stability. The junction box, located on the back of the panel, serves as the electrical connection point. Adapted from Aghaei et al. (2022) under a Creative Commons CC BY 4.0 license.

3 Degradation of FPV systems

After their deployment, FPV systems are submitted to various stress factors, including environmental stress, such as incident solar radiation, temperature, moisture, mechanical load, and soiling. These stress factors accelerate degradation mechanisms of the various components of PV modules, leading to corrosion of metallic elements, degradation of coatings, and photo-oxidation (Aghaei et al., 2022).

Solar panels contain both valuable and carcinogenic metals, such as cadmium, chromium, lead, silver, selenium, and tellurium (Sica et al., 2018). Leaching of these metals is reduced by glass-laminate encapsulation, but weathering and degradation can result in encapsulant deterioration and, consequently, metal release (Fig. 2). This concern was initially raised in the context of end-of-life of PV modules and their disposal in landfills (Sharma et al., 2021; Sica et al., 2018; Weckend et al., 2016). In leaching experiments using simulated rainwater, metal leaching from Si-based PV shown to be particularly important after one year, where silver, lead and chromium were released up to 27% of their initial content (Nain and Kumar, 2020). The calculated probability of exceeding standard surface water limits was up to 92% across several PV technologies tested, particularly for aluminum in multi-Si and mono-Si (Nain and Kumar, 2020). In another study of metal leaching under simulated disposal conditions, leaching rates were shown to increase with the age of the PV module. Metals such as lead, aluminum, zinc, and copper reached concentration up to three times higher in aged Si-based PV systems (i.e., 30-year lifespan), compared to new ones (Sharma et al., 2021). The presence of glass laminate encapsulation reduced lead leaching, whereas in its absence, lead concentration exceeded the limits for drinking water standards in all tested conditions (10 or 15 μg/L) (Sharma et al., 2021). This finding is of high concern given implications on health, including damages to the nervous system, the brain, and the kidney, which can escalate to life-threatening conditions (Wani et al., 2015).

In the case of FPV, leaching experiments of materials in direct contact with water (i.e., the PE tubing, PE caps and sealant material of the caps used in the flotation mechanism) were also performed (Mathijssen et al., 2020). The PE tubing and caps were found to release small amounts of aluminum and zinc. The sealant material leached aluminum, copper, manganese, and zinc, although most of these metals (excluding aluminum) were only detected after two weeks of exposure, and not after 24 h, suggesting slower leaching kinetics. The measured concentrations were several times lower than the internal standards for produced drinking water, however, the study was limited to a two-week weathering experiment. The authors acknowledged that long-term leaching behavior remains uncertain, that seasonal effects may influence leaching patterns, and that prolonged exposure could potentially lead to the release of additional compounds (Mathijssen et al., 2020).

While such experimental studies provide valuable insights, estimating actual metal release under real-world conditions remains a significant challenge. A recent and pioneering pilot study assessed metal leaching from FPV systems after one and three years of operation under field conditions in hosting basins (Akomea-Ampeh et al., 2025). Iron and manganese were identified as the predominant metals. However high heterogeneity was observed across all ponds, including relatively high concentrations in the control site (i.e., without FPV), which prevented the authors from attributing the detected metals exclusively to materials associated with the FPV infrastructure. The study calculated the heavy metal evaluation index (HEI) − a quantitative measure comparing detected concentrations to permissible drinking water limits − which indicated low metal pollution index for the two FPV-hosting basins. The HEI values were approximately two orders of magnitude below the threshold that defines low metal pollution, and none of the metals at any site exceeded regulatory limits (Akomea-Ampeh et al., 2025). However, the study was limited to only two FPV-hosting basins, a maximum of three years of FPV operation, and a narrow selection of metals (Cd, Cr, Fe, Mn, Pb, Sn). The authors identified several methodological pitfalls and emphasized the importance of study design in future assessments of FPV-related metal contamination, which is discussed in more details in Section 5.

PFAS-based processing aids are used in the production of fluoropolymers commonly used in the backsheet of FPV. PFAS are known as the “forever chemicals” due to their high persistence and low degradation tendency under natural conditions. They bioaccumulate and bio magnify in the food chain, and are recognized for their carcinogenic effects and immunotoxicity, among others (Brunn et al., 2023). PFAS synthesis byproducts are not covalently bound to the polymers and can be released into the environment during manufacturing, processing, and disposal, as well as during the lifetime of FPV systems once installed (Fig. 2). The amount of leaching from fluoropolymers depend significantly on the polymer type, their production processes and subsequent treatments (Lohmann et al., 2020). For example, emission of low-molecular weight PFAS, such as polyfluoroalkyl carboxylic acids differing by 1,1-difluoroethene (CF₂CH₂) units, has been observed downstream from PVDF manufacturing facilities near Decatur in Georgia (Newton et al., 2017). A decades-scale degradation study using microcosms demonstrated the degradation of fluorotelomer-based polymers through abiotic (e.g., hydrolysis) and biotic (e.g., microbial degradation) processes, releasing various PFAS such as fluorotelomer alcohols (FTOH), perfluorooctanoic acid (PFOA), and shorter-chain perfluorinated carboxylic acids (Lohmann et al., 2020; Washington et al., 2015).

Floating plastic structures are also prone to degradation, mainly due to UV exposure. The chemical and physical degradation of these structures causes both mechanical and molecular breakage of plastics, resulting in the release of small plastic particles, as well as soluble and/or volatile degradation products into the environment (Fig. 2). Small plastic fragments, including microplastics (i.e., particles smaller than 5 mm) and nanoplastics (i.e., particles smaller than 1000 nm and exhibiting a colloidal behavior) (Gigault et al., 2018; Weinstein et al., 2016), may be released following morphological damages, such as pits, cracks, and surface roughening of HDPE floaters (Fig. 2). Aged (micro)plastics, which have undergone oxidation and exhibit altered surface chemistry and increased surface area, are more prone to adsorb organic contaminants, potentially acting as vectors for pollution (Bhagat et al., 2022). Due to their small size, these particles are more likely incorporated into the aquatic food web, though their effects on biota health are not yet fully understood. Although HDPE is considered more chemically stable than polystyrene and PET polymers, photo-degradation studies have shown that it can release smaller organic compounds, such as long-chain alcohols, aldehydes, ketones, carboxylic acids, and hydroxy acids (Biale et al., 2021). In addition to degradation byproducts issued from the polymeric matrix, plastic additives are also of concern, as they are commonly used to enhance material performance and usability. These additives are not covalently bound to the polymer and are therefore susceptible to leaching. They include plasticizers, which increase flexibility and softness, as well as flame retardants, stabilizers, antioxidants, and coloring agents. For example, PE plastic may contain plasticizers like phthalates (diethyl phthalate, butyl benzyl phthalate), flame retardants such as polybromodiphenyl ethers, and antioxidants like Tris (2,4-di-tert-butylphenyl) phosphite (Iftikhar et al., 2024). The leaching of these additives poses potential health risks, including endocrine disruption, reproductive toxicity, and other long-term health effects (Arp et al., 2021; Iftikhar et al., 2024).

In the context of FPV systems, the leaching experiment conducted by Mathijssen et al. (2020) also screened for organic compounds leaching from components with the highest water contact (Mathijssen et al., 2020). In addition to heavy metals, the study investigated approximately 200 compounds potentially leaching from these materials, including poly- and monocyclic aromatic hydrocarbons, volatile halogenated hydrocarbons, monocyclic (halogenated) nitrogen compounds, phenols and halogenated phenols, none of which were detected. While this represents a valuable effort to explore potential FPV-associated contamination, the target compound list was limited, especially when considering the broad range of substances that may be released from polymeric and composite materials under environmental conditions. Additionally, the relatively short duration of the leaching experiments (i.e., two weeks) limits the ability to assess long-term leaching behavior.

Maintenance and cleaning operations are another potential source of water contamination that remains to be investigated. Different cleaning procedures, relying for example on water-based solutions, brushes and soap, are available to overcome soiling and dust accumulation. To reduce the need for cleaning, self-cleaning coatings are often applied to PV modules. Fluoropolymers such as fluorinated ethylene propylene (FEP) and polytetrafluoroethylene (PTFE), are commonly used for this purpose (Nain and Anctil, 2025). However, other hydrophilic and hydrophobic coatings, such as polydimethylsiloxane (PDMS), polymethylmethacrylate (PMMA), are also available (Syafiq et al., 2022; Zahedi et al., 2021). While these coatings help maintain module efficiency, they can become a source of environmental contamination over time, particularly as they degrade from UV exposure and weathering. As their surfaces deteriorate, these coatings not only lose efficiency and accumulate more soiling but may also release degradation products and residual chemicals, which can leach into nearby water bodies and pose a threat to aquatic ecosystems (Zahedi et al., 2021).

Any material leaching from the FPV system into the aquatic environment could be transformed into new compounds, many of which are unknown and therefore not included in conventional target screening approaches of contamination monitoring. Transformation products (TPs), for instance, are formed through degradation processes of parent compounds (Fig. 2). Depending on their environmental stability and metabolic pathways, contaminants may be present either in substantially smaller or larger amounts than their TPs, posing varying levels of environmental impact. With regard to FPV systems, it is therefore of primary importance to detect not only leaching from FPV components, but also identify, and quantify TPs to further evaluate their stability and toxicity (Escher and Fenner, 2011).

Despite the presence of hazardous and degradable materials in FPV systems, most available degradation studies focus on individual components tested under controlled conditions or exposed to single stressors, often targeting specific contaminants (Aghaei et al., 2022; Mathijssen et al., 2020; Nain and Kumar, 2020). These studies rarely address the potential for environmental contamination under real-world conditions, where multiple stressors interact over time, or evaluate the leaching behavior of the entire FPV system and its impact to their hosting basins. Understanding and monitoring leaching under such complex and variable conditions requires a comprehensive and fit-for-purpose analytical framework, which is explored in the following sections.

Fig. 2 Graphical representation of potential chemical contamination resulting from the implementation of floating photovoltaic (FPV) panels. Degradation of FPV in aquatic environment is enhanced by environmental stress such as UV radiation, heating, and weathering. Metals, polymeric compounds, and polymer additives can be released from the PV unit. Micro- and nano-plastics can be generated from the floating system. Chemicals leaching from the FPV structure can further degrade and/or react with the aquatic environment leading to additional transformation products. Abbreviations: PET, polyethylene terephthalate; PVF, polyvinyl fluoride; PFAS, polyfluoroalkyl substance; HDPE, high density polyethylene.

4 Comprehensive analytical strategy

Integrating FPV leachates and their TPs into chemical analysis presents a significant challenge due to their wide range of polarities, masses, and physicochemical properties. To comprehensively assess these substances, a combination of complementary analytical methodologies is essential, not only to detect known contaminants but also to identify unknown compounds (Zahn et al., 2024). This section first explores the use of non-target approaches, which enable the detection of compounds previously unknown. Achieving this requires careful sample pre-treatment, separation techniques, and the application of multiple ionization techniques to maximize chemical coverage (Figs. 3 and SI). In addition to non-target analysis, complementary targeted approaches are crucial for quantifying specific contaminants. In parallel, methods for microplastics detection and metal analysis, both allowing for quantitative measurements, are also discussed.

4.1 Non-target analysis (NTA) of organic contaminants

Non-target analysis (NTA) aims at identifying chemicals without prior knowledge of their molecular formula or structure (BP4NTA, 2024). In recent years, NTA has advanced considerably, driven by improvements in analytical instrumentation and computational techniques (Hollender et al., 2017). High-resolution mass spectrometry (HRMS) methods, coupled with separation techniques like liquid or gas chromatography (LC or GC-HRMS), are commonly employed in NTA for their high sensitivity, selectivity, versatility, and ability to detect both known and unknown substances across a broad mass range. HRMS is a gas phase technique that measures the accurate mass-to-charge ratio (mass accuracy < ± 0.001 Da) of ionized compounds present in a sample (liquid, gas, or solid). The accurate mass allows for the identification of elemental compositions, isotopic patterns, and potentially elucidate molecular structures by means of tandem MS experiments supported by various chemical compound databases. Coupling HRMS with chromatography provides an additional layer of information, as chromatography helps separating complex mixtures into individual components based on their physicochemical properties, such as polarity and volatility.

The simultaneous detection of thousands of compounds, including previously unidentified ones, allows to track eventual changes in the water chemical profile across gradients of time, space, or treatment, such as following the deployment of a FPV system. Statistical analysis of the data, using principal component analysis, clustering, regression analysis, and data reduction techniques, allows to pinpoint key features responsible for these changes and to further tailor the complete characterization of the associated compound(s). Each feature corresponds to an ionized molecule, but fully elucidating all features in a dataset and identifying unknown compounds remains a laborious task at present. Therefore, prioritization strategies are necessary in any NTA investigation to focus identification efforts on the most relevant signals (Aggerbeck et al., 2024; Hollender et al., 2023). In this context, NTA has been successfully used in several freshwater studies (Carpenter et al., 2019; Ruff et al., 2015), including the investigation of temporal trends of polar micropollutants in a riverbank filtration system in the Rhine basin (Albergamo et al., 2019). In this work, more than 18,000 molecular features were detected and subsequently clustered, with prioritization based on signal intensity. Following extensive data processing, the identified clusters were interpreted in the context of historical anthropogenic emissions and restoration measures, leading to the identification of 25 compounds, which had not previously been reported in bank filtrates or natural waters. Their identification provided new evidence that diverse polar compound classes can persist and migrate through riverbank filtration systems over extended periods.

NTA has also proven effective in elucidating longitudinal pollution patterns of organic micropollutants along river systems (Beckers et al., 2020). Distinct pollution signatures were identified, corresponding to inputs from wastewater treatment plant effluents, tributary confluence, and diffuse or intermittent sources, including small point discharges and groundwater inflows. Among 38 representative compounds associated with these patterns, 25 were successfully identified, providing the usefulness of NTA in prioritizing and managing complex environmental measures.

A key specificity of NTA data using HRMS is the ability to create a “digital archive” of the analyses. This allows for the retrospective examination of data as new concerns or knowledge about specific substances emerge, providing a dynamic and invaluable resource for ongoing environmental monitoring and research. Archived effluent data from wastewater treatment plants were retrospectively analyzed to identify high-priority unknown compounds, defined as those with the highest intensities and highest temporal frequency linked to point sources across samples from previous environmental monitoring campaigns (Lai et al., 2021). The findings were directly relevant to environmental monitoring and protection, leading to 21 tentative identifications for unknown compounds, and the recommendation of several additional candidate compounds for further identification efforts.

Altogether, NTA is a powerful tool for detecting a broad range of organic compounds, but it still faces key limitations. Quantification is not yet straightforward, workflows can be time-consuming, and the approach is not routinely applied in standard monitoring. As such, targeted methods remain the gold standard for accurate concentration measurements, while being blind to compounds outside the target list. Nevertheless, NTA offers a unique advantage: it is currently the only technique that allows for a comprehensive screening of environmental samples without prior knowledge of the contaminants. This makes NTA an indispensable tool in scientists’ toolbox for identifying unknown or emerging substances.

4.2 Metal analysis

Accurate detection and quantification of metals, such as lead, mercury, arsenic, and cadmium, require analytical techniques like inductively coupled plasma mass spectrometry (ICP-MS) and atomic absorption spectroscopy (AAS). In ICP-MS, the sample is introduced into a high-temperature plasma, where it is atomized, ionized, and further evaluated by MS, allowing a multi-element analysis. In AAS, the sample is atomized typically in a flame or graphite furnace. As the atoms absorb light at specific wavelengths, the decrease in light intensity is measured and related to the concentration of the metal in the sample (Wilschefski and Baxter, 2019). These methods offer high sensitivity and specificity, enabling the identification and quantification of trace metal concentrations in water samples (Fig. 3). Sample preparation typically involves steps such as filtration to remove particulate matter, acid digestion to break down complex matrices.

A non-target approach for metals analysis was recently demonstrated by Tirez et al. (2024), who applied a quantitative workflow using ICP-MS/MS and a Discrete Analyzer to surface water samples. The study emphasized the value of integrating non-target metal quantification with geochemical modeling to assess the concentration of ecotoxicologically relevant free metal ions. These data can support chemical fingerprinting and strengthen interpretations of inorganic pollution in complex environmental matrices. The approach also complements organic NTA in advancing a comprehensive chemical assessment (Tirez et al., 2024).

Fig. 3 Sample analysis workflow for assessing a complex array of contaminants potentially issued from FPV systems. The workflow starts with sample collection and preparation. The samples can then be analyzed based on a combination of analytical techniques, ranging from, but not limited to, pyrolysis-GC-MS and ICP-MS for identification of plastics, additives, and metals, to HRMS-based techniques for organic compounds. HRMS encompasses different approaches, including target analysis, suspect screening, and non-target analysis. Common layout of these analytical strategies usually consists of a first chromatographic separation to reduce matrix complexity, followed by an ionization step that allows to charge and transfer the analytes to the mass spectrometer. Different complementary ionization methods exist, differing in their working principle, allowing to cover a large range compound polarity. The generated ions, resolved by retention time, can optionally be separated based on their size and shape using ion mobility. The compound identification is then based on fragmentation and mass detection of the precursor ion (*) and its corresponding fragments. The data treatment steps involve different levels of data processing to monitor and/or identify the chemicals, the reporting, and the storage of the data for eventual retrospective analysis.

4.3 Micro and nano plastic analysis

The identification of microplastic in environmental matrices is commonly performed by spectroscopy methods, such as infrared or Raman spectroscopy, and by thermo-analytical techniques, such as pyrolysis-GC-MS. A major limitation of spectroscopic methods is their ability to identify and screen only single particles with a minimum size of typically 25 μm. Additionally, their selectivity is insufficient for identifying specific monomers, mixtures, additives, and degradation products. The emerging use of pyrolysis-GC-MS for the detection of micro- and nanoplastics involves pyrolyzing a sample under an inert atmosphere and analyzing its fragmentation profile using GC-MS (La Nasa et al., 2020). This technique allows the identification of polymers, such as PE, PMMA or PTFE, based on the molecular profile of the products produced during thermal decomposition (Tsuge et al., 2011; Yakovenko et al., 2020). The pyrolysis fingerprint can also provide information on degradation products, such as those resulting from photo-oxidation process, and plastic additives (Akoueson et al., 2021) (Fig. 3).

Following polymer identification, polymer concentration can be determined using external calibration curves, ideally created in a matrix similar to the samples under analysis and using analytical standards, i.e., virgin plastic particles. A specific marker of the polymer pyrolysis products is chosen, and its concentration in a sample is estimated against the same marker from the calibration curve. This process assumes a similar fragmentation pattern and pyrolysis yield between the polymer in the sample and in the calibration curve.

5 Chemical mapping and experimental design

To broaden the chemical coverage, the integration of diverse ionization, separation, and detection techniques is required. The use of orthogonal analytical techniques provide complementary information on the chemical space, allowing more effective identification and reducing potential interferences (Detailed information in SI). Upstream of the chosen analytical technique, sample preparation is a crucial step as it serves to remove matrix interferences, concentrate analytes, and ensure that the sample is compatible with the chosen analytical technique. By eliminating unwanted components, such as proteins, lipids, or salts, it enhances the accuracy and sensitivity of the analysis (Fig. 3). However, every step in the sample preparation process carries a risk of compound loss, which may result in incomplete or biased outcomes (Mechelke et al., 2019), underscoring the importance of understanding the limitations and selectivity of the chosen sample preparation approach.

Water is the most commonly investigated environmental matrix in NTA studies, with solid-phase extraction (SPE) emerging as the predominant sample preparation technique (Hajeb et al., 2022). This approach enables sample enrichment, which is particularly important for detecting contaminants present at sub-ppb levels. For instance, enrichment factors of 200 have been reported using SPE cartridges (Albergamo et al., 2019). Enrichment strategies may be necessary for the detection of organic compounds, metals, and microplastics in FPV-hosting basin.

In the context of NTA, the detection and identification of compounds should be followed by reporting the confidence level of a compound annotation that reflects the available evidence of its identity (Schymanski et al., 2014). The development of well-defined suspect lists, tailored to specific regulatory questions and environmental contexts, is crucial for improving the accuracy of the process. Designing small-scale experiments, e.g. laboratory leaching experiments of individual components, can help generate a chemical fingerprint of materials and their degradation products, providing a suspect list for targeted screening in field samples. Similarly, controlled mesocosm studies offer a valuable opportunity to simulate real-world aquatic conditions, enabling the integration of components fingerprint with metabolic signatures from microorganisms, macroinvertebrates, and fish, creating a more comprehensive mesocosm-based chemical profile.

Extracting meaningful patterns from complex datasets, applying multivariate statistics and machine learning models, can help researchers to correlate changes in chemical signatures with environmental or study variables (Albergamo et al., 2019; Beckers et al., 2020; Carpenter et al., 2019). In their work, Akomea-Ampeh et al. (2025) highlighted three recommended key actions related to experimental design to improve definitive conclusions regarding FPV impacts on water quality and risk assessment (Akomea-Ampeh et al., 2025). First, background testing should be conducted to assess pre-existing contamination levels, ideally through water quality monitoring before FPV installation. This includes analyzing contaminants and physicochemical parameters both prior to and immediately after deployment. Second, the authors emphasize the importance of systematic, long-term monitoring that captures seasonal, spatial, and depth-related variability. That is essential to distinguish FPV-related contamination from background variations and external influences, such as rainfall runoff, industrial discharge, atmospheric deposition or seasonal variations, requires robust experimental designs. A Before-After-Control-Impact (BACI) (before-after-control-impact) design is also recommended and has been recently applied in a FPV-related whole-lake study (Nobre et al., 2025), as it incorporates temporal variations and the presence of control sites, thereby minimizing the impact of unmeasured covariables on the observed effects (Chevalier et al., 2019). Ideally, chemical profiles elucidation using non-target approaches could be included and performed at high-temporal frequency, e.g., monthly. Third, the authors call for standardized data collection and analysis practices, including clearly defined control conditions and the incorporation of complementary analytical tools, such as water isotope analysis, to better explain contaminant fluxes at FPV sites.

Various factors may influence contaminant leaching from FPV systems, including the type of FPV installation, its surface coverage, water body characteristics, and local climatic conditions (Sandrini et al., 2025). Therefore, it is essential that studies document and report relevant metadata on both lake infrastructure and the specific FPV system deployed (Akomea-Ampeh et al., 2025).

In addition, sharing suspect lists of compounds detected in FPV-related studies, whether through laboratory leaching experiments or whole-lake monitoring, is essential to build a comprehensive database for identifying potential contaminants. Finally, creating a digital archive of samples that enables retrospective screening when new compounds are identified or when specific questions are raised (Alygizakis et al., 2019; Lai et al., 2021) is a crucial step toward understanding the occurrence and distribution of FPV-associated contaminants and developing prioritization strategies for risk assessment and management.

6 Environmental and health impact

The outlined analytical roadmap proposed in this paper has been tailored to the context of assessing FPV-associated contamination. Combining complementary analytical techniques to address the specific contaminants emerging from FPV systems implementation ensures a more comprehensive evaluation of potential environmental and health risks. However, it is important to acknowledge that applying such a broad range of techniques is both time-consuming and resource-intensive, necessitating a strategic balance between analytical depth and practical feasibility.

The co-occurrence of different organic contaminants, metals, and microplastics should be considered to more accurately capture the overall chemical exposure associated with FPV systems. Including both sediment and water analysis will further support the identification of contaminant sources and their environmental fate, such as potential accumulation in sediments, infiltration into groundwater, or integration into the food web. To assess the latter, it is essential to monitor contaminant levels in biota, such as macroinvertebrates and fish, as these reflect the extent to which pollutants are taken up and transferred through trophic levels.

The toxicological hazard of newly identified compounds or mixtures are often unknown, therefore further toxicological assessment is essential to evaluate their potential health risks. In addition to toxicological assays, computational tools − such as those predicting toxicity endpoints based on structural similarity to previously evaluated chemicals for a toxicity parameter − are increasingly being explored (Canchola et al., 2025). Integrating NTA with effect-directed analysis (EDA), for instance, enables researchers to directly link detected compounds to their potential toxicological effects (Cha et al., 2019). While such approaches provide critical insights into biological effects, it is equally important to design realistic risk assessment studies that account for real-world exposure scenarios, co-occurrence of contaminants, environmental dynamics, and long-term ecological impacts. Where applicable, incorporating health-based thresholds, such as the heavy metal evaluation index (HEI), can support the evaluation of FPV-associated health risks (Akomea-Ampeh et al., 2025). Altogether, these strategies strengthen compounds prioritization and supports monitoring efforts, helping to track the spread and relevance of contaminants and ultimately guiding evidence-based policymaking.

7 Conclusion

Conducting field studies to evaluate the real-life effects of FPV systems implementation, such as its impact on ecosystems, ecological functions, and water quality, is crucial to guide regulations and promote the deployment of this technology. Regarding the potential chemical contamination arising from the deployment and weathering of FPV systems, monitoring studies relying on NTA, combined with the application of target analytical techniques, presents both opportunities and challenges. NTA enables the detection and identification of a wide range of unknown compounds, providing valuable insights into complex environmental samples. Despite the powerful analytical capabilities of NTA, these studies must be conducted within the constraints of time and costs, necessitating efficient workflows, strategic resource allocation, and experienced scientists. Combining orthogonal analytical approaches, employing diverse ionization, separation, and detection techniques, is essential to increase the chemical coverage and to better elucidate the fingerprint of FPV systems in water bodies and its impact on the overall ecosystem. This should be combined with well-designed field experiments to minimize the risk of false associations. Comprehensive approaches ensure the detection and identification of a wide range of compounds, facilitating the design of realistic risk assessment studies and strategies to minimize the chemical impact on the environment. The non-exhaustive analytical toolkit and experimental design strategy discussed in this manuscript aim at ensuring that the deployment of renewable energy solutions aligns with worldwide environmental protection goals.

Acknowledgments

Project Floatix, funded by OFB (Office Français de la Biodiversité, France).

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence this paper.

References

All Figures

Fig. 1 (A) Multiple plastic floats support the PV cells. (B) An example of a common configuration of PV modules for crystalline silicon (c-Si). A glass top layer is followed by encapsulant layers that protect the solar cell and their internal circuitry. A backsheet provides insulation and protection at the bottom, while a frame encases the entire structure for mechanical stability. The junction box, located on the back of the panel, serves as the electrical connection point. Adapted from Aghaei et al. (2022) under a Creative Commons CC BY 4.0 license.

In the text

Fig. 2 Graphical representation of potential chemical contamination resulting from the implementation of floating photovoltaic (FPV) panels. Degradation of FPV in aquatic environment is enhanced by environmental stress such as UV radiation, heating, and weathering. Metals, polymeric compounds, and polymer additives can be released from the PV unit. Micro- and nano-plastics can be generated from the floating system. Chemicals leaching from the FPV structure can further degrade and/or react with the aquatic environment leading to additional transformation products. Abbreviations: PET, polyethylene terephthalate; PVF, polyvinyl fluoride; PFAS, polyfluoroalkyl substance; HDPE, high density polyethylene.

In the text

Fig. 3 Sample analysis workflow for assessing a complex array of contaminants potentially issued from FPV systems. The workflow starts with sample collection and preparation. The samples can then be analyzed based on a combination of analytical techniques, ranging from, but not limited to, pyrolysis-GC-MS and ICP-MS for identification of plastics, additives, and metals, to HRMS-based techniques for organic compounds. HRMS encompasses different approaches, including target analysis, suspect screening, and non-target analysis. Common layout of these analytical strategies usually consists of a first chromatographic separation to reduce matrix complexity, followed by an ionization step that allows to charge and transfer the analytes to the mass spectrometer. Different complementary ionization methods exist, differing in their working principle, allowing to cover a large range compound polarity. The generated ions, resolved by retention time, can optionally be separated based on their size and shape using ion mobility. The compound identification is then based on fragmentation and mass detection of the precursor ion (*) and its corresponding fragments. The data treatment steps involve different levels of data processing to monitor and/or identify the chemicals, the reporting, and the storage of the data for eventual retrospective analysis.

In the text