Non-targeted screening of per- and polyfluoroalkyl substances: Advanced methods, challenges, and environmental health

Abstract

Per- and polyfluoroalkyl substances (PFAS), a class of synthetic fluorine-containing organic compounds, pose a serious threat to the ecological environment and human health due to their persistence, bioaccumulation, and extensive toxicity. Non-targeted screening (NTS) is a key method for identifying and determining unknown PFAS, which is crucial to the understanding of their exposure pathway and health risks. Hence, this review focuses on NTS techniques for PFAS in the environment. Firstly, high-resolution mass spectrometry and ion mobility innovations enabling novel PFAS/isomer identification are evaluated. Afterwards, pretreatment optimization (e.g., solid-phase extraction and emerging adsorbents) is summarized by analyzing the advantages of each method and the challenges posed by the limited datasets, while also outlining their applicable scenarios. Analytical challenges from structural complexity (short-chain substitutes and ether-based fluorine-modulated polymers) and matrix effects are discussed. Lastly, practical implications for environmental health and the future development potential of NTS technologies for PFAS are presented. Overall, this review proposes a science-based framework for monitoring and regulatory prioritization, with the expectation of supporting PFAS management and mitigation.

Graphical abstract

Highlights

• •Achieve the full-chain integration of pretreatment and non-targeted screening.
• •Emphasize the screening of short-chain PFAS and the existence of unknown precursors.
• •Form a dynamic closed loop of regulations and risk assessment.
• •Summarize the advanced NTS method of deep fusion with the actual screening scene.

1. Introduction

With the acceleration of industrialization and modernization, the issue of emerging pollutants has gradually become a global environmental challenge. Emerging pollutants refer to those that have been recently discovered or have drawn attention, pose risks to the ecological environment or human health, and have not yet been included in management or whose existing management measures are insufficient to effectively prevent such risks [[1], [2], [3]]. Per- and polyfluoroalkyl substances (PFAS), one of the emerging pollutants that widely exist in the environment, have attracted considerable scientific and regulatory attention in recent decades. They are defined as synthetic fluorinated substances containing at least one perfluorinated methyl (−CF₃) or methylene (−CF₂−) carbon atom [4]. Among these, perfluoroalkyl substances have all hydrogen atoms in their carbon chain replaced by fluorine, while polyfluoroalkyl substances retain some C−H bonds [[5], [6], [7]]. As Fluorine (F) is the most electronegative element, and the bound fluorine is one of the most stable elements. F attracts electrons to itself by chemical bonds, giving polarity and strength to the C−F bond (about 490 kJ/mol) [6,8,9]. The strong and stable C−F bond contributes to the chemical and thermal stability of PFAS, leading to bioaccumulation in organisms and long-term environmental retention [10]. Three representative PFAS compound, perfluorooctanesulfonate (PFOS), perfluorooctanoic acid (PFOA), and perfluorohexanesulfonate (PFHxS), along with their salts and related compounds, were listed in the Stockholm Convention in 2009, 2019, and 2022, respectively. While PFOA, PFOS, and their salts are regulated under international conventions as controlled persistent organic pollutants, their widespread presence in the environment and the emergence of new substitutes, including short-chain, ether-based, and chlorinated alternatives, underscore that PFAS contamination remains a significant concern. These ongoing challenges highlight the need for continuous monitoring and assessment of PFAS as a class of emerging contaminants. Despite global restrictions, numerous unregulated alternatives continue to be released, such as short-chain perfluorocarboxylic acid (PFCA), perfluoroalkyl sulfonic acid (PFSA), and chlorinated polyfluorinated ether sulfonic acid (Cl-PFESA). Notably, perfluoroalkyl ether carboxylic acid (PFECA) and perfluoroalkyl ether sulfonic acid (PFESA) are now used in fluoropolymer manufacturing and metal plating applications, respectively, with 6:2 Cl-PFESA increasingly adopted by the electroplating industry as a primary substitute for PFOS [11].

Owing to the hydrophobic, oleophobic properties and extensive sources, PFAS have been extensively incorporated into commercial products, including textiles, surfactants, food packaging, nonstick coatings, and aqueous film-forming foam [12,13]. The widespread application has resulted in PFAS pervasive environmental release through multiple pathways, with primary emission sources spanning industrial discharges, agricultural runoff, landfill leachates, and domestic wastewater (Fig. 1a) [14,15]. PFAS circulate through the environment via six primary pathways: aquatic dispersion, marine-estuarine cycles, atmospheric deposition, soil-sludge cycles, biomagnification and food chains, and indoor microenvironments. This enables long-range transport and biomagnification, resulting in persistent exposure risks for background populations and ecosystems far from source areas. The environmental partitioning of PFAS exhibits distinct chain-length dependence. Long-chain PFAS predominantly adsorb to surface soil particles due to their hydrophobic characteristics, demonstrating limited vertical mobility. In contrast, short-chain alternatives (e.g., PFBA, PFBS) exhibit higher water solubility, facilitating deeper soil penetration and ultimately entering groundwater and seawater through the hydrological (water) cycle [16]. Moreover, PFAS concentration in the atmosphere varies seasonally, which is influenced by regional meteorological conditions, local emission sources, and long-distance transport air masses [14,17]. Such atmospheric mobility, combined with their extreme persistence, enables PFAS to undergo continuous environmental cycling between air-water-soil compartments, ultimately achieving global dispersion (Fig. 1b–f) [14]. Additionally, environmental conditions (e.g., temperature), soil characteristics, water body characteristics, characteristics of particulate matter, and molecular structure will further influence the migration of PFAS in the environment (Fig. 1a) [18]. Thus, conducting PFAS screening and investigating their presence and distribution in the environment can enhance understanding of PFAS pollution sources and pathways, while assessing potential risks to humans and wildlife. The non-targeted Screening (NTS) strategies for PFAS based on homolog and fragment can conduct in-depth research on the isomer spectrum of PFAS, which is helpful for the source analysis of PFAS and further understanding of its ecological risks [19,20].

Fig. 1.

The (a) source of PFAS and the spatial distribution of PFAS detected at various concentrations in the (b) African region, (c) Asia-Pacific region, (d) surface waters across the South American and Caribbean region, (e) water resources across the European region, and (f) North American region. Adapted from Lyu et al. [15] and Kurwadkar et al. [14] with permission from John Wiley & Sons and Elsevier, respectively.

Compelling evidence of PFAS ubiquity emerges from their detection in remote ecosystems far from industrial activities, such as glacial meltwaters on the Tibetan Plateau and deep oceanic zones of the Pacific-Atlantic interface. Even in regions with stringent emission controls, PFAS persist in environmental matrices due to their resistance to both abiotic degradation (hydrolysis/photolysis) and biotic transformation [8,[21], [22], [23], [24]]. PFAS released into the environment can directly expose humans via dermal contact or inhalation; nevertheless, the predominant exposure route remains the ingestion of contaminated water and food [25]. Long-term exposure to PFAS can cause their accumulation in the body; even trace amounts can disrupt metabolism, immunity, and hormones, increasing the risk of diabetes and cancer [26,27], leading to chronic diseases in specific organs, like liver damage, thyroid disease, reduced fertility, high cholesterol, obesity, and hormone suppression. These effects can be particularly severe in infants, pregnant women, and people with compromised immune systems [28,29]. These health implications underscore the critical need for precise identification and comprehensive monitoring of PFAS in complex environmental systems. Therefore, early screening is essential to sever the exposure chain to these “forever chemicals” before disease develops. An evaluation of peer-reviewed literature in the Web of Science Core Collection yielded 12,079 articles using the search terms: (“PFAS” OR “Per- and polyfluoroalkyl substances”) AND (“screening” OR “non-target” OR “HRMS” OR “analysis” OR “detection”).

Given the pervasive application and documented hazards of PFAS, comprehensive screening constitutes a critical component of environmental risk management. However, there are some challenges in screening processes: (1) the environmental and biological sample matrices are complex and prone to interact with PFAS, interfering with extraction and screening [30]; (2) PFAS in environmental and biological samples occur at trace concentrations (typically ng/L or pg/L for liquids; ng/g or pg/g for solids), which places extremely high demands on the limits of detection (LOD), precision, and accuracy of analytical instruments [31]; (3) the complex structure of PFAS encompasses numerous types, such as perfluorooctanoic acid and perfluorosulfonic acid, with most categories containing a large number of isomers. Taking PFOA as an example, linear and branched isomers exhibit significant differences in sources, migration, transformation, and biological toxicity. Neglecting these isomers may lead to an underestimation of total PFAS levels and associated risks. However, not all PFAS categories exhibit isomerism [32]; (4) screening for PFAS homologues and precursors faces challenges, including the absence of standards, difficulty in separation, and susceptibility to transformation during the screening process. This leads to issues such as omissions, inaccurate quantification, or ion suppression phenomena during screening [33]; (5) many PFAS do not have commercial reference standards, which makes it impossible to accurately evaluate the performance indicators such as the accuracy, precision and sensitivity of analytical methods, and also makes it difficult to accurately control the quality of the test results [34]; (6) the mass spectrometry (MS) databases and structural databases related to PFAS are not complete enough, which makes it difficult to accurately identify and attribute the structures of the unknown PFAS detected during NTS, limiting the comprehensive understanding of the types and distribution of PFAS in the environment [35]. Therefore, conducting PFAS screening plays a crucial role in revealing their distribution patterns in various environmental media, clarifying pollution sources and migration paths, and tracking the emergence of emerging pollutants. It can also provide a scientific basis for the formulation of pollution prevention and control measures, the revision of environmental quality standards, and the improvement of relevant regulatory policies. Thereby, PFAS screening can effectively address the challenges of PFAS pollution and ensure environmental safety and human health [[36], [37], [38]]. Before conducting systematic screenings of unknown samples, sample pretreatment is indispensable, which can enhance screening sensitivity, accuracy, and reliability, as well as protect analytical instruments. It can also provide reliable guarantees for the screening of PFAS and lay a foundation for in-depth research on the environmental behavior, pollution sources, and ecological risk assessment of PFAS [39].

At present, in addition to traditional targeted screening, the commonly used PFAS screening methods include suspect screening and NTS based on HRMS [40]. Suspect screening and NTS are analytical techniques appropriate for identifying compounds and molecules when authentic standards and surrogates are not available [41]. Suspect screening relies on databases and previously recorded data, while NTS relies instead on feature prioritization [42]. In PFAS screening, the full scan methods (MS1) and the methods based on fragmentation (MS2) are two key MS analysis techniques. The MS1 provides molecular weight information by detecting the mass-to-charge ratio (m/z) and intensity of all ions, making it suitable for preliminary screening and qualitative analysis. In contrast, the MS2 conducts fragmentation analysis on specific precursor ions to generate m/z and intensity information of the sub-ions, which helps to improve the signal-to-noise ratio (S/N, the ratio of signal strength to noise strength in an electronic device or system), thereby providing detailed structural information to accurately identify and quantitatively analyze PFAS. Moreover, the combined use of the two can screen PFAS in environmental samples more comprehensively [42,43]. Together, these approaches can partially close the gap in detecting the growing list of PFAS. Closed-loop monitoring systems for mass suspect screening include advanced HRMS technology, coordinated analytical procedures using big data processing tools, and information sharing through chemical databases [35].

Despite growing scientific attention on PFAS, critical knowledge gaps persist regarding advanced NTS methodologies. Matrix-specific pretreatment optimization, HRMS-based suspect/NTS integration, and isomer risk assessment remain under active investigation for efficient NTS. A comprehensive review focusing on addressing these challenges is urgently needed. This review (1) analyzes the existing pretreatment technologies of PFAS and related advantages, (2) summarizes NTS methods based on HRMS and some creative combinations of several screening methods used to meet different screening needs, (3) explores the exposure-risk quantification relationships by discussing chain-length dependent environmental partitioning, isomer-specific environmental impact, and exposure pathway, (4) proposes potential solutions to address analytical standardization needs, monitoring principles and emerging PFAS prioritization. Overall, this review aims to offer research directions for PFAS screening, monitoring, and risk assessment.

2. Pretreatment technology

2.1. The necessity of pretreatment

The composition of environmental samples is complex, and various substances may interact with PFAS and cause interference, such as dissolved organic matter and suspended particles in water, humus and minerals in soil, and biological macromolecules such as proteins and lipids in biological samples. Kebarle and Tang proposed the matrix effect in 1993 during their research on electrospray ionization (ESI) mechanisms. This concept describes the inhibitory or enhancing effects of coexisting components in the sample matrix on the ionization efficiency of target analytes [44]. Through a systematic pretreatment process, the interfering components in the matrix can be removed, thereby significantly enhancing the sensitivity and quantitative accuracy of the screening method [37,39]. Currently, over 7 million PFAS-related structures are cataloged in the PubChem database (https://pubchem.ncbi.nlm.nih.gov) [45]. The NORMAN database system (https://www.norman-network.com/nds/) contains all supporting information and software tools necessary for hosting and processing data obtained through extensive target and suspect screening of PFAS in the environment [46]. PubChem provides physicochemical parameters for PFAS and natural interferents to guide adsorbent selection and internal standard design, while NORMAN offers validated natural ion “exclusion lists” and toxicity labels. These two can work together to weaken the matrix effect and reduce false positives before pretreatment. Meanwhile, the concentration of PFAS in the environment is usually at trace levels, and it is necessary to enrich PFAS through pretreatment to improve the S/N. Additionally, the instrument’s LOD is also significant. LOD refers to the minimum concentration of a target substance that an analytical method can accurately detect under specified conditions. Although pretreatment cannot alter the theoretical LOD of the instrument itself, it can elevate target analytes below the instrument’s LOD to a stable detectable signal range by increasing target concentration, reducing background interference, and enhancing signal response, which achieves a lower effective LOD in practical testing. Therefore, pretreatment is an indispensable step for enhancing the sensitivity of trace analysis, particularly in fields such as environmental monitoring, food safety, and bioanalysis [47]. In addition, PFAS comprise many homologues and isomers with similar properties but significant differences in toxicity and environmental behavior. Pretreatment can preliminarily separate and enrich PFAS, facilitating subsequent analytical methods for separation and identification, and enabling accurate analysis of each component in complex mixtures.

Due to the poor stability of certain PFAS in environmental samples, photolysis, hydrolysis, or defluorination reactions may occur under ultraviolet radiation (UV), high temperatures, or extreme pH conditions, producing short-chain/fluorinated intermediates. Therefore, rapid pretreatment under specific conditions can minimize losses and ensure the authenticity of test results. Meanwhile, pretreatment can remove harmful components from complex sample matrices, not only enhancing the stability of PFAS but also protecting analytical instruments, extending their service life, reducing maintenance costs, and ensuring the smooth progress of analytical work.

2.2. Pretreatment technologies

The sample pretreatment schemes are distributed in a gradient pattern: from simple membrane filtration and solvent dilution to the extraction process that requires precise control. In practical applications, solid-phase extraction (SPE) dominates, with its core being the hydrophilic-lipophilic balance (HLB) adsorbent. If the target is negatively charged, a weak anion exchange (WAX) polymer column or a polyacrylonitrile, divinylbenzene/xylene/polydimethylsiloxane fiber should be used for selective capture. This fiber belongs to the SPME fiber category, combining both adsorption and absorption mechanisms to enable the extraction of compounds across a broader range of polarities. This fiber can also be used complementarily with WAX columns. For samples with complex matrices, liquid-liquid extraction (LLE), solid-liquid extraction (SLE), ultrasonic-enhanced extraction, dispersed solid-phase microextraction (d-SPME), dispersed liquid-liquid microextraction (DLLME), or accelerated solvent extraction (ASE) can be further combined to balance the recovery rate and purification efficiency [39,48,49]. Table 1 summarizes the most common pretreatment methods at present and their respective advantages.

Table 1.

Common pretreatment methods and their respective advantages and application scenarios.

Pretreatment methods	Advantages	Application scenarios	Ref.
Protein precipitation (PPT)	PPT is simple and rapid to operate, can effectively remove proteins, reduce matrix interference, and is suitable for liquid biological samples. The organic solvent precipitation method does not require desalination treatment and has high resolution. Salting out is low-cost, environmentally friendly, and effective at stabilizing proteins and enzymes.	Biological samples, especially high-protein matrices such as serum and plasma.	[50]
Digestion method	Digestion techniques can completely break down the three-dimensional conformation of proteins, thereby fully releasing PFAS that are covalently or non-covalently bound to them. Acid-based digestion offers advantages such as simple operational steps, low reagent consumption, and low cost.	High-fat, high-protein, or complex biological samples (such as liver, muscle, eggs).	[51]
Liquid-liquid extraction (LLE)	LLE equipment has low requirements, is easy to operate, and is easy to apply on a large scale.	Liquid sample	[52]
Solid-phase extraction (SPE)	SPE efficiently enriches and purifies target compounds, enhances detection sensitivity, reduces matrix interference, and is suitable for various sample types. The offline SPE method is mature, flexible, and customizable. The online SPE has a high degree of automation, reduces human errors, and increases the analytical throughput.	Nearly all common substrates	[53]
Solid-liquid extraction (SLE)	Due to its good compatibility with granular matrices, SLE has become the preferred strategy for the separation of PFAS in solid biological samples. With techniques such as ultrasound-assisted extraction, the extraction efficiency can be improved.	Solid/Semi-solid samples	[54]
Ion-pair extraction (IPE)	IPE can effectively extract PFAS from samples with strong ionization ability and improve the extraction efficiency.	When mixed with proteins, lipids, and humic acids, and cannot be processed using conventional LLE or SPE.	[55]
QuEChERS method	The QuEChERS method is rapid, simple, inexpensive and effective, suitable for a variety of biological samples, and can reduce solvent usage and pretreatment time.	When rapid, low-cost, high-throughput simultaneous screening of short- and long-chain PFAS is required in complex food samples, biological fluids, and sediments.	[56]

Among them, SPE is a commonly used pretreatment method for PFAS [57,58]. Its advantages lie in its ability to enrich and purify target compounds, enhance detection sensitivity, and reduce matrix interference. This method is applicable to both liquid and solid samples. SPE is divided into online and offline modes. Although offline SPE has disadvantages such as complex operation and time-consuming procedures, it has high maturity and strong flexibility, and can achieve efficient extraction of specific PFAS by choosing different adsorbents and conditions. Besides, online SPE is an integrated technology that incorporates micro-sized solid-phase extraction columns into the LC-MS flow path, automatically completing enrichment, purification, and injection within 3–5 min. This approach significantly reduces solvent and human error, establishing itself as the mainstream automated pretreatment method for high-throughput monitoring of trace PFAS [49]. However, the recovery rates and enrichment efficiencies of PFAS vary among different pretreatment methods. Among them, due to the strong polarity and hydration of short-chain PFAS, as well as the weak fluorine-fluorine interaction, the recovery rates of short-chain PFAS (17%–70%) are generally lower than those of long-chain PFAS (70%–130%) [[59], [60], [61]]. Moreover, due to the weak retention of short-chain PFAS on reversed-phase/ion-exchange stationary phases, their wide elution windows and tendency to be prematurely lost, their enrichment efficiency is lower than that of long-chain PFAS [59,62,63]. To address the issues of low recovery rate and enrichment efficiency of short-chain PFAS, strategies such as introducing ion pairs, fluorine-affinity solvents, WAX optimization, or HILIC chromatography can be adopted to improve analytical performance and meet the requirements of trace monitoring and toxicity studies [64,65].

In the future development of pretreatment technologies for biological samples, the combination of new materials (such as special nanomaterials solid-phase extraction columns) with HRMS and single-cell technology will be a significant trend, which can further enhance the ability to detect low-concentration PFAS [7,37]. For example, the single-cell technology integrates capture, lysis, and online SPE enrichment at the single-cell scale into a unified pretreatment system. It can be scaled up to achieve high recovery while removing matrix components and concentrating intracellular PFAS, enabling direct HRMS determination of single-cell loading. This provides a new dimension for revealing exposure heterogeneity. Through a multi-technology collaborative strategy, precise identification and reliable quantification of full-spectrum PFAS in highly complex biological matrices can be achieved [39,49].

3. Full scan methods (MS1)

In order to better understand PFAS, accurately identifying PFAS in complex environmental matrices is indispensable [12]. The classification of PFAS and their common representatives is shown in Fig. 2 [25,66]. According to the number of fluorinated carbon atoms, PFAS are informally classified as long-chain, short-chain, or ultra-short chain [15,22,67,68]. It is worth noting that long-chain PFAS may be more harmful to the human body due to their high bioaccumulation and toxicity [69]. Short-chain PFAS have high mobility and are difficult to capture. Despite their relatively low accumulation, they still pose potential toxicity to specific groups such as fetuses [66]. It is urgent to track their sources, distribution, as well as health and environmental risks [[70], [71], [72]]. Toxicity studies of ultra-short chain PFAS are limited, and their potential health effects are still unclear. In the past, due to the immaturity of technology, the high cost of detection, and the lack of strict supervision, the research on short-chain PFAS was long overshadowed by that on long-chain PFAS [73]. Thus, further research is still needed [73,74]. PFAS are everywhere and often overlooked, yet they seriously endanger health [75]. Therefore, there is an urgent need to develop a rapid sensing platform that combines high sensitivity and on-site applicability to achieve immediate identification of PFAS, thereby supporting informed decision-making for precisely reducing their environmental risks.

Fig. 2.

The classification of PFAS. Adapted from Gaillard et al. [25] with permission from Elsevier.

At present, the methods for screening PFAS existing in the environment are mainly divided into targeted screening and NTS. Traditional targeted screening methods focus on a limited number of known compounds, mainly conducting qualitative and quantitative analyses of known PFAS present in the environment. These compounds may not be able to capture the entire spectrum of PFAS present in the environment, which may lead to the neglect of many unidentified or newly emerging variants. To address this limitation, NTS technology is of crucial importance [76,77]. In order to determine the relative abundance of non-target PFAS, we conducted a semi-quantitative analysis. In short, based on structural similarity, each proposed non-target PFAS structure is matched with the quantitative target PFAS in the targeted screening. Then, we calculated the peak area ratio between non-target PFAS and the corresponding matching target PFAS in the sample, and obtained the concentration of non-target PFAS based on the product of this ratio and the concentration of target PFAS. These methods can detect and estimate the concentration of PFAS in a wider range, thereby conducting a more comprehensive assessment of environmental pollution [78,79]. In addition to HRMS-based techniques, another analytical tool to quantify the extent of unknown PFAA precursors is the total oxidizable precursor (TOP) assay [80]. This method directly oxidizes samples with alkali, completely converting all oxidizable PFAS precursors into stable PFCA (e.g., PFOA). After cooling, filtration, and pH adjustment, the generated PFCA is quantified via SPE enrichment and LC-MS/MS analysis, enabling calculation of the total original oxidizable precursor content in the sample [81]. Owing to the optimization of the PFAS sample pretreatment process, supplemented by the coordinated advancement of NTS, suspect screening, and HRMS, the characteristic peaks that were once difficult to quantify in the mass spectrometry spectra can now be analyzed one by one, significantly expanding the chemical space for detectable PFAS. The integration of the above-mentioned technologies not only provides high-confidence data for the subsequent standard methods to incorporate more PFAS but also lays a methodological foundation [39].

Current PFAS screening strategies primarily rely on three major analytical platforms: LC-MS, GC-MS, and HRMS. By integrating multidimensional separation with ultra-high-resolution/high-mass-accuracy detection, these platforms enable both targeted quantification and non-targeted discovery within a single analytical workflow. This approach provides the technical foundation for accurate qualitative identification and sensitive quantitative determination of PFAS [8,82]. Based on the depth of mass spectrometry analysis, existing methods can be further categorized into two primary levels: MS1 and MS2. These two levels complement each other in terms of principles, functions, and application scenarios: MS1 performs only first-level mass spectrometry acquisition, focusing on utilizing precise mass numbers, isotope distribution, and retention time (RT) to achieve rapid initial screening and homolog profiling analysis; while MS2 employs collision-induced dissociation (CID) on selected parent ions following MS1, utilizing high-resolution detection of fragment ions to obtain structural information and thereby enhance identification confidence. However, its effectiveness is constrained by fragmentation efficiency and database coverage [83,84]. In practical applications, MS1 and MS2 are often used in combination to improve the accuracy and reliability of the analysis. NTS for PFAS typically involves preliminary screening of HRMS data using mass defect filtering. Subsequently, homologous series are constructed based on characteristic neutral losses (e.g., CF₂, CF₂CF₂, and CF₂O) and product ion sequences, enabling family classification and structural inference of unknown fluorinated compounds. The following section will systematically outline the key implementation points for both strategies in PFAS analysis.

In the MS1 methods, charged ions enter the mass spectrometer and undergo a first stage of MS. In the MS1 stage, a mass analyzer, such as a quadrupole, time-of-flight, or orbitrap, separates all incoming ions according to their m/z. The MS detector records the signal strength of these ions to generate a first mass spectrum. This spectrum shows the relationship between the m/z of each component in the sample and its corresponding signal strength, enabling the identification and analysis of the compounds in the sample. MS1 methods have the advantages of high sensitivity, strong selectivity, and rapid analysis, and are widely used in the fields of environmental monitoring, metabolomics, drug development, and food safety.

3.1. Mass defect filtering

Mass defect filtering is an analytical technique based on HRMS data for the rapid screening of compounds with a specific range of mass defects (MD). Its core principle includes the identification of MD and the application of MD filtering. MD refers to the difference between the exact mass of a compound and its nearest integer mass [43]. MD filtering refers to the rapid screening of compounds with similar structural characteristics by setting a specific MD range. This method is particularly useful for identifying compounds with specific repeating units, such as CF₂ [83].

MD filtering is commonly used to identify homologous sequences. PFAS usually have specific repetitive units (such as CF₂) and are of various types. Therefore, MD filtering can reduce the dimensionality of the data and improve the analysis efficiency [85]. The key to PFAS screening is to calculate the MD of each ion and set the specific MD range. Screening for ions with specific MD should exclude irrelevant background ions. The selected ions are analyzed to confirm the presence of PFAS (Fig. 3a) [86]. MD filtering can perform rapid screening, quickly identifying potential PFAS from large amounts of data, thereby enhancing analytical efficiency. It is applicable to various sample types, including water samples, soil, sediment, and biological tissues. In addition, the high sensitivity of HRMS enables the screening of PFAS at extremely low concentrations. However, when dealing with complex samples, there may be multiple compounds with similar MD, leading to false positive results. However, MD filtering relies on HRMS data, which places high demands on the instrument [87].

Fig. 3.

The workflow of PFAS screening. Adapted from Bugsel et al. [83] with permission from Elsevier.

3.2. MD/C-m/C approach

Although MD filtering can meet most of the requirements for PFAS screening, NTS is based on HRMS, which can comprehensively characterize the compounds in the samples and generate a large amount of data, thus requiring effective data reduction techniques to identify potential PFAS. Therefore, new technologies are needed to prioritize the measured HRMS data in order to enhance the efficiency of PFAS screening. Kaufmann et al. proposed a method to distinguish and prioritize PFAS from other compounds by creating a two-dimensional scatter plot (MD/C-m/C plot) [83], normalizing the MD and mass (m) of the compounds to the number of carbon atoms, respectively.

This method visualizes the results by drawing data points on the MD/C-m/C two-dimensional graph, allowing for intuitive observation of the distribution characteristics of different compounds. Due to its unique chemical structure, such as high fluorine content, PFAS usually form specific distribution areas on this map. Introducing MD/C-m/C into NTS assists in the identification and quantification of unknown PFAS, helping to reduce data dimensions and improve analytical efficiency [88,89]. The process of screening PFAS by the MD/C-m/C method is shown in Fig. 3b. After obtaining the MS data, the MD of each ion is calculated and normalized to the carbon number to obtain MD/C and m/C. These values are then plotted respectively on two-dimensional graphs to identify PFAS through specific ranges. This method can rapidly screen potential PFAS from large datasets to enhance analytical efficiency. However, in complex samples, there may be multiple compounds with similar MD/C and m/C values, which can lead to false positive results.

3.3. Kendrick mass defect (KMD) analysis

With the development of HRMS technology, the amount of data generated has increased dramatically, and traditional analysis methods have been difficult to meet the demand. In addition to the MD/C-m/C method, Kendrick proposed a mass scale based on CH₂ = 14.0000 in 1963, along with the derived MD, which converts the IUPAC mass of a compound to a mass based on specific functional groups (such as CH₂). This was originally intended to address the difficulty in identifying isomers in petroleum samples, and was later known as the KMD. This method displays homologues as horizontal lines in the KMD plot, making them easily identifiable. It provides an efficient data processing approach for rapidly screening homologous sequences and structure-specific compounds [82]. The application of KMD in PFAS screening mainly lies in the screening of homologues, which can quickly locate and classify compounds carrying this structural unit. Besides, KMD can assist in NTS to identify unknown PFAS in complex environmental matrices [90]. When processing HRMS data, KMD can reduce the data dimension and improve the analysis efficiency [91].

The workflow of KMD analysis for PFAS screening is shown in Fig. 3c. After obtaining the mass spectrometry data, the organic mass (OM) of the compound is normalized to integer masses of repeating units, such as CF₂, to calculate KM and KMD. Then, PFAS are screened by setting a specific KMD range [85]. The advantage of KMD analysis is that it can efficiently simplify data and improve analysis efficiency. Notably, no MS2 data is required, which expands the application scope of this method. Combining KMD with other multi-dimensional parameters can significantly improve the identification accuracy of highly fluorinated PFAS. However, for partially fluorinated PFAS with a relatively high hydrogen substitution ratio, the screening sensitivity significantly decreases as the mass loss value approaches that of common organic backgrounds. In addition, when the signal is too strong or too weak, the error in isotope abundance determination increases, thereby weakening the reliability of KMD-based calculation [83].

3.4. Chemical formula assignment

With the development and application of HRMS technology, a method of chemical formula allocation has been gradually formed. The method uses precise mass information provided by HRMS, combined with parameters such as isotopic abundance and distribution, to infer possible chemical formulas. The core principle is to use a high-resolution mass spectrometer (such as Orbitrap MS or FT-ICR MS) to measure the exact mass of the ions in the sample, then analyze the isotope pattern of the ions to further confirm the molecular formula, and finally, based on the precise mass and isotope pattern, combined with chemical laws and databases, generate possible molecular formulas [92]. Besides, chemical rules, such as element composition, double bond equivalents, etc., are applied to filter and screen possible molecular formulas to reduce false positive results [43]. In the processing of complex samples, the distribution of chemical formula can reduce the data dimension and improve the analytical efficiency.

The chemical formula assignment workflow for PFAS screening is shown in Fig. 3d. This method can lock potential PFAS at high throughput in complex matrices, significantly shortening the analysis cycle. This approach also reduces the dependence on instrument configuration and experimental environment, thus having a broader prospect for laboratory promotion [83]. With the further optimization and application of this method, it is expected to play a greater role in environmental monitoring, toxicological research, and pollution control of PFAS [93].

4. Methods based on fragmentation (MS2)

The MS2 method, also known as secondary MS or tandem MS (MS/MS), is a high-precision mass analysis technique that involves coupling two or more MS analyzers for further structural analysis of the molecules in the sample. The principle is that the fragment ions generated by MS1 enter a second mass spectrometer, where they are separated and detected according to their m/z. The MS2 spectra show the m/z and relative strength of these fragment ions, and by analyzing these data, detailed structural information of the parent ions can be obtained.

The MS2 method is suitable for the screening of compounds in complex samples. Besides, MS2 can improve the screening sensitivity of specific molecules, especially for low-concentration compounds. In this method, the identity of the molecules in the sample is confirmed by matching the MS2 spectrum with the standard spectrum. The MS2 method has been widely used in bioanalysis, proteomics, metabolomics, and environmental monitoring, making it an indispensable tool in modern MS. Compared with the MS1, MS2 achieves structural confirmation through characteristic fragment ions, eliminating co-ions of the same mass and reducing false positives by over 50%. Its SRM/PRM mode removes matrix noise, lowers the LOD, and distinguishes co-eluting isomers. It also provides traceable fragmentation information for unknown PFAS, thereby comprehensively outperforming MS1 in specificity, sensitivity, and quantitative reliability [83].

4.1. Diagnostic fragments

A diagnostic fragment refers to the formation of characteristic fragment ions during the fragmentation process of certain compounds in MS. These characteristic fragment ions can be used as fingerprints to identify specific compounds. This approach involves selecting a specific parent ion (precursor ion) from MS1 and subsequently fragmenting it via CID, high-energy collisional dissociation (HCD), or other fragmentation techniques to generate smaller fragments. Then, the characteristic fragments are identified by analyzing the m/z of the generated fragment ions. Finally, the identified characteristic fragment ions are compared with the characteristic fragment ions of known compounds, so as to identify the compounds [94].

This method can improve the confidence of PFAS screening [95,96]. The workflow of diagnostic fragment screening for PFAS is shown in Fig. 4a. After obtaining the MS data, a specific parent ion is selected and crushed to create a fragment ion. Then, the characteristic fragment ions are identified. PFAS are screened by setting specific diagnostic fragment ions and fragment mass differences, and their identities are confirmed through database comparison. Finally, the selected ions are analyzed, and the structure is identified in combination with other information [97]. The advantage of this method is that the diagnostic fragments provide detailed structural information of the compounds, which can help identify PFAS in complex samples and improve the detection sensitivity of specific PFAS through the detection of characteristic fragment ions. However, the screening of diagnostic fragments relies on the MS2 data of known compounds, which has limitations. In complex samples, it is prone to false-positive results [83].

Fig. 4.

The workflow of PFAS screening. Adapted from Bugsel et al. [83] with permission from Elsevier.

4.2. Fragment differences and neutral loss

Fragment differences and neutral loss are techniques based on MS2 data used to identify characteristic fragment ions of compounds. Fragment differences refer to the formation of multiple fragment ions by the parent ion under CID or other fragmentation techniques in MS. By calculating the mass differences between these fragment ions, compounds with specific structural characteristics can be identified. Neutral loss refers to the fact that in the process of fragmentation, the parent ion may lose a neutral molecule (such as water molecules, methanol molecules, etc.) to form fragment ions [98].

By analyzing fragment differences and neutral loss events in MS2 data from NTS, PFAS with specific structural features can be rapidly identified, thereby aiding in the discovery of unknown PFAS within complex environmental matrices. Besides, combining other data (such as precise mass, isotopic pattern, RT, etc.) can improve the confidence of PFAS screening [95]. The workflow of fragment differences and neutral losses in PFAS screening is shown in Fig. 4b. The main steps are selecting a specific parent ion and fragmenting it to create a fragment ion. Then, calculate the mass difference between the fragment ions, detect the neutral molecule loss events, and establish a list of characteristic fragment ions and neutral loss events. PFAS are screened by setting specific fragment differences and neutral loss patterns, and identified in combination with database comparison [96]. This strategy can provide detailed molecular structure information, which is helpful for accurately identifying PFAS in complex matrices. In addition, it significantly enhances sensitivity and selectivity and can meet the analysis requirements of trace compounds. However, as the MS2 data-based approach relies on known compounds, its ability to identify novel PFAS is limited. Furthermore, in complex environmental matrices, there may exist multiple compounds with similar fragment differences and neutral loss patterns, which will lead to false positive results [83].

4.3. Fragment ion flagging

Fragment ion flagging is a MS2 data-based technique used to label characteristic fragment ions to help identify specific compounds. It works by first selecting a specific parent ion from MS1. Then, fragmentation processes will break the parent ion into smaller fragment ions by CID, HCD, or other fragmentation techniques. Finally, the m/z of the generated fragment ions is analyzed, and the characteristic fragment ions are identified [58]. In addition, these characteristic fragments are flagged for subsequent analysis and screening. In NTS, by analyzing the characteristic fragment ions in MS2 data, this method can rapidly identify PFAS with specific structural features. By combining with other data, fragment ion flagging can improve the confidence level of PFAS screening [99].

The workflow of fragment ion flagging for PFAS screening is shown in Fig. 4c [100]. This method can provide rich structural information to help identify PFAS in complex samples. In addition, the sensitivity and anti-interference ability have been significantly enhanced, meeting the analysis requirements of trace compounds. However, due to its reliance on the MS2 data of known compounds, its ability to identify novel PFAS is limited [83].

4.4. Ion mobility mass spectrometry

Ion mobility mass spectrometry (IM-MS) achieves two-dimensional separation and precise identification of ions in complex matrices by successively utilizing the differences in gas-phase migration rates and m/z. Its working principle is to ionize sample molecules to form charged ions. Then, ion mobility separation is carried out. Under the drive of an external electric field, charged ions migrate directionally in inert buffer media such as nitrogen or helium. Then, ions separated by mobility enter the mass spectrometer. The mobility of ions is related to their collision cross-section (CCS). Smaller ions move faster, while larger ions move more slowly. Further separation and detection are carried out based on the m/z. Finally, by analyzing the migration time of ions and the m/z, ion mobility spectra and MS are generated for screening and quantitative analysis (Fig. 4d) [100,101].

In 2019, Ahmed et al. first applied differential mobility spectrometry (DMS) to the separation and detection of PFAS such as perfluoroalkyl compounds [102]. Yukioka et al. proposed a PFAS screening method combining data-independent acquisition (DIA) and ion mobility spectrometry in 2021 [103]. This method distinguishes the target precursor ions and fragment ions from the co-eluting ions through the drift time, improving the recognition efficiency of PFAS. Then, in 2025, Boatman et al. studied the reporting practice of PFAS screening confidence when LC, GC, and ion mobility spectrometry are combined with HRMS, and proposed a simple and unified confidence guideline. Combining the characteristics of PFAS and the CCS value of the ion mobility spectrum has improved the confidence level of PFAS screening [104]. In NTS, IM-MS can rapidly identify and characterize unknown PFAS in complex samples by analyzing migration time and the m/z of ions, but the identification ability of this method is also limited by these parameters [83].

5. Other methods

5.1. Methods combined with TOP

The total oxidizable precursors (TOP) assay is a pretreatment strategy that couples chemical oxidation with targeted analysis [105]. By quantitatively converting potential perfluoroalkyl acid precursors in the sample, this approach estimates the proportion of unknown PFAS that can be oxidized and detected as PFAAs, providing supplementary information for subsequent precise screening [106]. These precursors represent potential new sources of PFAA in the environment but are often overlooked in most studies, resulting in a serious underestimation of PFAS levels [42,107]. In order to screen out more potential PFAS in the environment, Li et al. proposed a PFAS screening method that combines NTS with TOP assistive methods, and effectively utilized this method to successfully detect PFAS that had not been detected before [108]. Moreover, Zhao et al. employed a comprehensive approach that combines screening based on HRMS with TOP analysis to identify emerging PFAS and potential TFA precursors in north China and predict their environmental behavior. In this method, three PFAS screening strategies are adopted respectively or in combination: suspect screening, NTS based on homologues, and NTS based on fragments. It successfully identifies novel PFAS in the environment and can also predict their behavior under oxidative conditions.

The workflow of this method mainly consists of data collection and preprocessing, suspect screening, NTS, and TOP analysis combined with NTS. The core part of this method lies in the fact that after NTS, the samples are oxidized with TOP to convert the precursors into known final products (such as PFCAs). The composition of PFAS before and after oxidation is compared through targeted analysis and non-targeted analysis (HRMS) to identify incompletely oxidized precursors or new final products. This integrated strategy combines TOP analysis with NTS, overcoming the limitation of traditional TOP targeting only PFCAs to comprehensively reveal multiple oxidation end-products. Employing multidimensional data mining techniques, including suspect database matching, homology screening, and fragment analysis, significantly enhances the resolution of complex PFAS mixtures and provides an effective tool for environmental monitoring and risk management. However, while the combined use of TOP and NTS expands the identification boundaries for unknown PFAS, several bottlenecks remain to be overcome before achieving robust, comparable total exposure screening within regulatory frameworks. These include insufficient standardization of oxidation conditions, lack of databases for ultra-short-chain and ether-type products, inadequate control of matrix interference, and gaps in solid-phase product resolution.

5.2. Mass spectrometry without chromatography

Given the complex and variable nature of the sample matrix, pretreatment such as purification and enrichment must be carried out before PFAS screening. The trace-level occurrence of PFAS require the method to have both high sensitivity and extremely low LOD. Therefore, after the pretreatment is completed, liquid chromatography-electrospray tandem mass spectrometry (LC-ESI-MS/MS) or LC-HRMS is generally used in studies for targeted quantification and NTS. However, the LC separation process is time-consuming, consumes excessive solvents, and generates a large amount of chemical waste, making the LC technology unsuitable for high-throughput analysis. Besides, matrix suppression of ion signals may occur when using MS. To reduce the suppression of mass spectrometry signals by the substrate, it is necessary to remove the substrate before it enters the mass spectrometry system. This step is usually accomplished through extraction or desalination, but it will consume additional time. When the sample matrix is complex or the PFAS concentration is at trace levels, achieving the required sensitivity is particularly difficult. Therefore, Hassan et al. reported the feasibility of rapid screening for PFAS in environmental water bodies using paper-based spray ionization mass spectrometry (PS-MS), confirming that the method can complete the detection within minutes. The desalination paper spray mass spectrometry (DPS-MS) method was further developed to solve the problem of ion suppression caused by salt or soil matrix [109].

In PS-MS analysis, the triangular test strips are ultrasonically cleaned sequentially with acetone, pure methanol, and a methanol-water solution before the sample solution is applied. After drying, they are placed before the mass spectrometer inlet, methanol is dropped, and high-voltage ionization is applied (Fig. 5a). In DPS-MS analysis, the solution is applied to the tip of the triangular paper and then contacts the gold-plated fabric, where it is desalinated in situ by capillary action. The paper is fixed before the inlet, soaked in methanol, and then ionized with high voltage (Fig. 5b). This method enables rapid screening of multiple PFAS at low ppt levels and direct analysis of high-salinity soil within 3 min, with no pretreatment is required. DPS-MS combines speed, portability, and low-ppt sensitivity for rapid analysis without chromatographic equipment; in situ desalting of gold-plated fabric ensures 70%−120% recovery in complex matrices, with the same paper substrate applicable to water, soil, and packaging extracts. However, the absence of chromatographic separation leads to overlapping signals from branched/straight-chain isomers, and high organic matter content may still cause signal drift. With limited species coverage and reliance on portable mass spectrometry, it lacks ISO/EPA certification, making it more suitable for field screening rather than as an enforcement-grade quantitative tool.

Fig. 5.

The workflow of (a) PS-MS (b) DPS-MS and the high-throughput SPME- Microfluidic Open Interface (MOI) -MS system (c) schematic diagram of the high-throughput extraction column and (d) the MOI-MS interface. Adapted from Hassan et al. [109] and Zhou et al. [110] with permission from Elsevier and American Chemical Society, respectively.

Direct injection is prone to ion suppression and a sudden drop in sensitivity due to the matrix or high salt content. Moreover, trace amounts of PFAS are more difficult to detect when they are not enriched. SPME (solid-phase microextraction) integrates sampling, extraction, purification and enrichment in a single step, achieving simple and efficient sample pretreatment. Therefore, Zhou et al. customized a high-throughput SPME-microfluidic open interface (MOI)-MS system for PFAS analysis (Fig. 5c and d). This compatible SPME system, when used in conjunction with direct MS, can rapidly screen PFAS in drinking water and complex food matrices. In addition, this system is capable of simultaneously processing up to 48 samples using SPME and directly coupling with MS through an automatic microfluidic open interface (MOI) without the need for chromatographic separation [110]. This method enables detection of 18 PFAS compounds at the nanogram level within 3 min without chromatography, offering high throughput and broad matrix adaptability. However, it does not provide isomer separation, exhibits insufficient robustness and coverage, and lacks standard certification. It is suitable only for rapid batch screening rather than quantitative enforcement testing.

5.3. Optical methods

Due to the trace concentrations of PFAS, most analyses are performed in specialized laboratories using chromatographic separation methods coupled with mass spectrometry (HPLC-MS/MS). These analytical protocols may involve preconcentration steps, which prolong analysis time and reduce method reproducibility and sensitivity, thereby limiting rapid on-site screening [111]. Considering the limitations of current analytical methods, developing a rapid, simple, and portable screening method is essential. Colorimetric assays stand out as a technique characterized by minimal operation, rapid response, and low-cost equipment. Based on their principles, colorimetric assays can be categorized into nanoparticle (NP)-aggregation-based and redox dye-based types. NP aggregation-based methods achieve screening by inducing surface changes in AuNPs/AgNPs through PFAS exposure, leading to alterations in absorbance or color. Redox dye-based methods primarily involve the formation of ion pairs between anionic PFAS and cationic dyes (e.g., MB, MG), causing shifts in UV-visible absorption peaks or changes in intensity [34]. Colorimetric assays require no chromatography, enabling rapid testing at low cost, making them suitable for on-site preliminary screening. However, they are prone to high false-positive rates due to interferences like humic acids. Without chromatographic separation, errors may arise from the presence of isomers. These methods are only suitable for semi-quantitative analysis and not for legally binding quantitative testing.

Fluorescence sensing methods achieve highly sensitive detection by monitoring changes in fluorescence intensity (on or off) following binding between fluorescent probes and PFAS. In addition, multi-parameter detection (such as the combination of fluorescence, absorption, and scattering) further enhances its screening ability and can be used for the preliminary analysis of water and soil. However, this method is vulnerable to background interference and requires complex sample pretreatment, which limits its application. Due to the problems of insufficient sensitivity and poor selectivity of traditional optical methods in complex environmental matrices, Zhang et al. constructed a lanthanide-based MOF fluorescence array that leverages the electrostatic/hydrophobic differences arising from the fluorine tail and head-end charge disparity of PFAS. This approach achieved 100% accurate classification of seven typical perfluorinated compounds and their mixtures without requiring pretreatment, providing a simple and accurate tool for high-throughput screening of trace PFAS (Fig. 6a) [112]. Additionally, Chen et al. developed a fluorescent sensor array composed of three topologically distinct porphyrin-zirconium MOFs, successfully distinguishing six PFAS types through differential static quenching patterns induced by PFAS adsorption [113]. Zha et al. achieved high-precision rapid differentiation of multiple PFAS in water by leveraging the unique fluorescent response generated through competitive binding between β-cyclodextrin polymers and dyes, combined with machine learning [114]. In addition, Wang et al. immobilized metal nanoclusters on a ZIF-on-MIL MOF framework to construct a machine learning-assisted proportional fluorescence sensor array, enabling rapid and specific identification and differentiation of multiple PFAS (Fig. 6b) [115]. These fluorescence array methods do not require pretreatment, can complete readings quickly, and have a low LOD. At the same time, by integrating machine learning, the classification accuracy rate for 6 to 7 typical PFAS is over 95%, and the chip cost is suitable for on-site high-throughput screening. However, the absence of chromatographic separation leads to errors in the quantification of short-chain PFAS and branched isomers. Moreover, fluorescence suppression occurs when humic acid is present. A single array covers fewer than 10 targets and has a low response to ethers. It lacks ISO/EPA certification and is currently only suitable for primary screening rather than law enforcement quantification.

Fig. 6.

Schematic Illustration of (a) the Ln-MOF fluorescent sensor array and (b) the synthesis of M@ZIF-on-MIL and ratio fluorescence sensor array construction. Adapted from Zhang et al. [112] and Wang et al. [115] with permission from Elsevier and American Chemical Society, respectively.

5.4. Machine learning methods

NTS for PFAS has revealed previously unknown pollutant categories. The current problem is that for many sites contaminated by PFAS, individual PFAS are only partially known. To solve this problem, Zweigle et al. coupled KMD homologous sequences with their self-developed FindPFΔS mass difference algorithm to construct a multidimensional non-target system. This approach enables the simultaneous identification of 17 PFAS compounds in composite agricultural soils and validates the mass difference method’s capability to detect non-homologous structures [90]. Moreover, there are many types of PFAS in nature that have not been discovered. To capture more unknown PFAS in environmental samples, Heuckeroth et al. developed an ICP-MS/MS-ESI-MS/MS dual-channel platform capable of simultaneously detecting fluorine elements and molecular ions without relying on the KMD series, enabling universal discovery of low-fluorinated unknown organic fluorine compounds [116].

NTS often performs poorly in complex matrices, has difficulty identifying atypical fragment PFAS, and has a high false positive rate and an uncertain structure. Furthermore, existing research is limited by time and space and does not cover a wide range of long sequences, making it difficult to systematically assess potential ecological and health risks. To expand the chemical space for detectable PFAS, Li et al. established a parallel workflow integrating Compound Discoverer and FluoroMatch 2.0 based on online SPE-LC-HRMS, overcoming the limitations of the MD to achieve extended identification of low-fluorinated/partially fluorinated PFAS in drinking water and their degradation products (Fig. 7) [117]. Moreover, Jiao et al. proposed a two-layer homologous network framework that clusters similar compounds and eliminates >90% of false-positive features, successfully identifying 94 PFAS (including 36 novel structures) and significantly accelerating the discovery and confirmation of PFAS in complex matrices [118]. Machine learning multi-dimensional NTS simultaneously captures multiple low-fluorine/non-homologous PFAS, reducing false positives while achieving simultaneous element-molecule quantification and comprehensive coverage of degradation products. However, inconsistencies in parameters introduce errors, preventing differentiation between branched/linear isomers. Additionally, the lack of databases for short-chain and ether PFAS, along with insufficient toxicity data for new structures, limits isomer toxicity assessment and inclusion in regulatory control lists.

Fig. 7.

The workflow of online SPE combined with LC-HRMS and Q-Exactive Orbitrap system. Adapted from Li et al. [117].

6. Conclusions

Driven by technological advances, the expanding PFAS database reveals unique environmental persistence and bioaccumulation hazards distinct from conventional pollutants. Comprehensive screening remains fundamental for risk assessment and regulatory intervention of PFAS. Pretreatment of samples to achieve the purpose of enriching PFAS and removing impurities can help to better carry out PFAS screening. Based on the selected reference criteria, targeted screening can quickly identify PFAS. This strategy, with its high specificity, can achieve precise qualitative and quantitative analysis of known and potential emerging pollutants, providing a reliable basis for pollution level assessment. However, due to the limitations of the selected standards, those unknown PFAS cannot be identified. Thereby, NTS has been developed to expand the screening scope of PFAS, especially short-chain PFSAs and ether-based fluorotelomers. Various integrated techniques have been developed for efficient PFAS NTS. For example, methods that combine TOP with mass spectrometry can significantly enhance the quantitative analytical ability of unknown peaks in the spectrum. Overall, PFAS are closely related to daily human activities and exposure pathways. Strengthening the application prospects of PFAS screening in environmental monitoring and pollution control is of great significance for environmental health. This review aims to systematically organize existing commonly used NTS methodologies, identify standardization gaps under short-chain, ether-based polymers and matrix effects, and advocate for a globally shared database and an exposure-structure-toxicity prioritization framework to achieve a closed-loop process from PFAS discovery to regulatory action.

7. Perspectives

The value of advanced PFAS screening lies not only in expanding chemical data but also in translating it into actionable health protection knowledge. To this end, tools such as miniature samplers and artificial intelligence are being developed to clarify the links between environmental exposure and health risks through environmental health analysis. Consequently, the future of PFAS assessment hinges on this synergistic integration, ensuring technological advancements directly translate into enhanced early warning capabilities and targeted risk management strategies for population and ecosystem safety.

7.1. Technological development

The heterogeneity of PFAS exposure pathways warrants further investigation. Global multi-regional surveys have revealed that PFAS are widely present in various environmental media, and their distribution pattern has no significant correlation with regional industrial or economic levels, suggesting that daily consumer goods and household products may be important exposure sources for such substances in developing countries. Source-apportionment models linking products to environmental loads of PFAS could be developed [14,119]. Developed regions should establish comprehensive, high-precision closed-loop demonstration systems, while underdeveloped areas should adopt rapid screening and regional data-sharing for low-cost pathways. These complementary approaches will provide replicable PFAS risk management templates for diverse economic gradients worldwide. The current PFAS screening technology is evolving towards high sensitivity, integrated methods, intelligent decision-making, miniaturization of equipment, and on-site portability. Future efforts must prioritize breakthroughs in ultra-sensitive field sensors, simultaneous monitoring of short-chain compounds and precursors, and regional collaborative data platforms. This will enable a shift from reactive responses to proactive prevention in addressing emerging pollutants. Future PFAS research should place greater emphasis on gender-specific susceptibility. It is reported that PFAS may be more harmful to women than men, but current toxicological conclusions have not yet fully clarified the specific mechanisms by which PFAS cause related diseases in women [10]. Moreover, interdisciplinary cooperation (such as analytical chemistry, environmental science, and toxicology) can be carried out to promote the construction of a global monitoring network and data sharing platform. Additionally, NTS can combine with multi-omics and big data to study the synergistic effects of PFAS with other pollutants, as well as the application of artificial intelligence in NTS. The established NTS technical roadmap, data formats, and prioritization model can be directly applied to emerging pollutants such as brominated flame retardants (BFRs), chlorinated paraffins (CPs), and microplastics (MPs). This approach provides a universal methodology and regulatory template for future NTS development, application, and risk management of emerging pollutants.

7.2. Environmental health analysis

The long-term persistence of PFAS in the environment, their toxicity and harmfulness to organisms, as well as their bioaccumulation, have drawn extensive attention. Under current pollution levels, drinking water and food intake are the primary pathways driving health risks, with children already exhibiting non-carcinogenic risks (HQ > 1). Although skin contact and inhalation contribute minimally, they can accumulate through consumer products, leading to long-term internal exposure. Prioritizing control of PFAS concentrations in drinking water and food can bring overall risks back within acceptable ranges. Thus, conducting a series of studies on PFAS is of utmost urgency. Comprehensive PFAS screening and analysis enable rapid hazard identification and reduced assessment timeframes, while providing regulatory authorities with precise, comparable, and legally valid data. This directly drives core management processes such as limit value establishment, pollution source reduction, product substitution, and polluter-pays principles. By transforming unknown contaminants into known substances and converting unknown exposures into quantifiable risks, PFAS screening provides a comprehensive exposure profile and quantifiable risk benchmarks for environmental health assessments. Simultaneously, it generates end-to-end evidence for health policy development, including priority lists, scientifically grounded thresholds, source access controls, responsibility allocation, and effectiveness evaluations. Additionally, it facilitates targeted prevention and control alongside multi-departmental collaborative governance.

CRediT authorship contribution statement

Shuan Yang: Writing – review & editing, Writing – original draft, Validation, Supervision, Investigation. Huan Yi: Writing – review & editing, Visualization, Validation, Supervision. Danling Ma: Writing – review & editing, Validation, Supervision. Lixi Zeng: Writing – review & editing, Visualization, Validation, Supervision.

Declaration of competing interest

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