Perspectives on the Presence of Environmentally Persistent Free Radicals (EPFRs) in Ambient Particulate Matters and Their Potential Implications for Health Risk

Abstract

Environmental persistent free radicals (EPFRs) represent a class of long-lived, redox-active species with half lives spanning minutes to months. Emerging as critical environmental pollutants, EPFRs pose significant risks due to their persistence, potential for bioaccumulation, and adverse effects on ecosystems and human health. This review critically synthesizes recent advancements in understanding EPFR formation mechanisms, analytical detection methodologies, environmental distribution patterns, and toxicological impacts. While progress has been made in characterization techniques, challenges persist—particularly in overcoming limitations of electron paramagnetic resonance (EPR) spectroscopy and spin-trapping methods in complex environmental matrices. Key knowledge gaps remain, including molecular-level dynamics of EPFR formation, long-term environmental fate under varying geochemical conditions, and quantitative relationships between chronic EPFR exposure and health outcomes. Future research priorities could focus on: (1) atomic-scale mechanistic investigations using advanced computational modeling to resolve formation pathways; (2) development of next-generation detection tools to improve sensitivity and spatial resolution; and (3) integration of EPFR data into region-specific air-quality indices to enhance risk assessment and inform mitigation strategies. Addressing these gaps will advance our capacity to mitigate EPFR persistence and safeguard environmental and public health.

1. Introduction

Free radicals are atoms or molecules that contain one or more unpaired electrons, a unique electronic configuration that imparts distinctive chemical reactivity [1,2]. These species include simple entities such as hydrogen atoms (with a single unpaired electron) and transition metal ions, such as iron, copper, and manganese [3,4]. The charge states of free radicals exhibit diversity, appearing as anionic radicals (negatively charged), cationic radicals (positively charged), or neutral radicals (charge-neutral) [5]. These distinct charge states dictate their differential reactivity and behavior in chemical processes [6]. For example, anionic radicals possess a greater electron-accepting capacity, while cationic radicals are more prone to donate electrons [6,7]. Neutral radicals exhibit dual redox properties, functioning as either oxidizing or reducing agents depending on the reaction conditions [4,8]. Historically, the scientific community has acknowledged that the unique electronic structure of free radicals underlies their strong bonding tendencies and high chemical reactivity, which results in their transient existence—designation as “ephemeral radicals” [9,10]. This transient nature presented significant challenges for isolation and characterization until the advent of advanced spectroscopic techniques [11]. During the 1950s, researchers discovered that positive radicals generated from emissions of coke and coal combustion exhibited prolonged stability [12], leading to their designation as Persistent Free Radicals (PFRs). While Harmon’s paradigm-shifting hypothesis initially lacked molecular validation, subsequent studies identified hydroxyl radicals (•OH) as key mediators of radiation-induced DNA strand breaks (e.g., via Fenton chemistry) and oxidative modifications to mitochondrial lipids as drivers of aging-related metabolic decline. Later investigations into tobacco smoke and carcinogenesis identified EPFRs as key constituents of cigarette smoke [13], while structurally analogous radicals were later detected in combustion-derived organic matrices, including charred biomass and diesel soot [14]. In recent years, studies have shown that significant quantities of EPFRs, including semiquinones, phenoxyl radicals, and cyclopentadienyl radicals, are substantially generated during the cooling phases of combustion or other thermal processes [15,16]. Adsorption of free radicals onto particulate matter surfaces significantly enhances their stability (from days to years), enabling quasi-infinite environmental persistence [16,17,18]. Their ubiquity in environmental matrices impacts ecosystems and human health (Figure 1) [19,20,21]. The number of papers indexed by Web of Science with the keyword “environmentally persistent free radicals” has increased greatly since 1998. Considering the importance of the pollutant, a review is thus urgently needed to summarize the current understanding and knowledge on EPFRs and, more importantly, provide guidance for future research.

This review will focus on the formation of mechanisms, analytical methodologies, atmospheric distribution patterns, and potential health impacts of EPFRs and provide insights into their environmental and toxicological implications.

2. Formation Mechanisms of EPFRs

EPFRs exhibit extraordinary stability, with lifespans ranging from hours to years in environmental matrices (e.g., PM_2.5, soil, combustion soot) [22,23]. This variability arises from differences in radical species (e.g., semiquinones vs. phenols), formation pathways (e.g., adsorption-driven stabilization vs. gas-phase radical coupling), and matrix properties such as chemical composition, temperature, and humidity [24,25,26]. Research showed that EPFRs could remain detectable via EPR spectroscopy even after prolonged exposure (more than 1 day) [27,28,29].

The environmental persistence of EPFRs is governed by their formation pathways. The main pathway included:

(1)

EPFRs formation mediated by transition metals

Transition metals have been demonstrated to be a key factor in the formation of EPFRs. Acid-washing of fly ash to remove metals completely eliminated EPFR generation upon phenol adsorption, while metal reintroduction restored radical signals—confirming their indispensable mediation in stabilization [30]. Furthermore, mechanistic evidence from ⁵⁷Fe⁵⁷Fe-labeled montmorillonite revealed a quantitative correlation (R² = 0.89) between ⁵⁷Fe³⁺→⁵⁷Fe²⁺ reduction and EPFR concentration after anthracene adsorption, directly linking electron transfer from organic precursors to metal centers as the radical formation pathway [31]. The EPFRs generated at redox–active metal–oxide interfaces exhibit exceptional atmospheric stability, as evidenced by half lives exceeding several weeks under ambient conditions [32]. Studies indicated that the half lives of EPFRs depend on both the specific transition metal oxide and precursor molecule involved in their formation (Table 1).

The persistence of EPFRs was modulated by interfacial interactions with transition metals (such as Fe) (Figure 2). Three key environmental variables exhibit divergent stabilizing and destabilizing effects: (i) Oxygen dynamics and molecular oxygen (O₂) played dual roles through the redox cycling of Fe(II)/Fe(III). In this process, reactive oxygen species (e.g., •OH generated via the reaction Fe²⁺ + H₂O₂ → Fe³⁺ + •OH + OH⁻ under acidic conditions) facilitated the degradation of organic compounds to produce radical intermediates. Concurrently, the reduction of Fe(III) occurs—mediated by agents such as O₂•⁻ or semiquinones—which serves to regenerate Fe(II). (ii) Hydrological interactions mediate the destabilization of EPFRs through three distinct mechanisms: (a) the formation of hydration shells that sterically hinder radical recombination; (b) competition with organic pollutants for catalytically active sites; and (c) reductive dissolution of transition metal oxides, which reduces the availability of redox-active sites; (iii) photonic influences: solar irradiation drives paradoxical effects. Light irradiation promotes the degradation of organic chemicals and thus the generation of organic byproducts, including organic free radicals, which may contribute to the additional EPFR signals. On the other hand, direct light irradiation on EPFRs may facilitate their decay as well.

Indirect pathway of the formation of EPFRs included that photooxidation of organic matrices generates secondary radicals (e.g., OH, CO₃⁻·), enhancing EPFR-like signals through covalent adduct formation, and the direct pathway could be as UV absorption by organic ligands and promote EPFR depopulation via singlet oxygen (¹O₂)-mediated oxidation.

Table 1. The half life of the interaction between typical transition metal oxides and precursor molecules in the generation of EPFRs.

Particles	Organic Adsorbate	Range of g-Values	Half Life	ΔH p-p (Gauss, G)	EPFR Concentration (Spins/g)	Reaction Conditions	Reference
Fe(III)2O3/Silica	Phenol	2.0040–2.0044	69.6–111 h	5.7–6.1	~1016–1017	150–350 °C At 400 °C: Radical decomposes to cyclopentadienyl species	[25]
	Catechol	2.0057–2.0067	111 h (max)	Not reported	~1017(max at 200 °C)	150–350 °C Maximum EPFRs yield at 200 °C
	Hydroquinone	2.0053–2.0056	69.6 h	Not reported	~1016 (low yield)	150–350 °C EPFRs undetectable at 400 °C
	2Monochlorophenol	2.0033–2.0052	24 h	Not reported	<1016	150–400 °C Adsorption via H2O elimination
	1,2Dichlorobenzene	2.0057	Unmeasurable	Not reported	<1015	150–400 °C
	Monochlorobenzene	2.0033	Unmeasurable	Not reported	<1015	150 °C
Cu(II)O/Silica	Catechol	2.0061–2.0070	~74 min	Not reported	Not reported	150–400 °C	[27]
	Hydroquinone	2.0062–2.0065	~4.6 days	Not reported	Not reported	150–400 °C Stable under vacuum
	Phenol	2.0032–2.0038	~27 min	Not reported	Not reported	150–400 °C Single radical type
	2Monochlorophenol	2.0041–2.0049	~24 h	Not reported	Not reported	150–400 °C
	1,2Dichlorobenzene	2.0063–2.0068	~1 h	Not reported	Not reported	150–400 °C
	Monochlorobenzene	2.0067–2.0070	~30 min	Not reported	Not reported	150–400 °C
Ni(II)O/SiO2	Catechol	2.0051–2.0065	~2.8 days	Not reported	Medium yield	250–300 °C	[34]
	Hydroquinone	2.0044–2.0057	~2.8 days		Low yield	Up to 400 °C
	Phenol	2.0035–2.0043	5.2 days		Lowest relative yield among all adsorbates	350 °C
	2Monochlorophenol(2-MCP)	2.0042–2.0043	3.8 days		Monotonic increase with temperature	150–400 °C
	Monochlorobenzene	2.0031–2.0038	2.4 days		Higher yield than phenol at >150 °C	300–400 °C
	1,2Dichlorobenzene	2.0034–2.0041	1.7 days		Highest yield below 350 °C	300 °C
ZnO/Silica	Phenol	2.0042	46–73	6.8	4.3 × 1018	230 °C, 5 min exposure	[35]
	Monochlorobenzene	2.0037	46–73	10.0	10.4 × 1018	230 °C, 5 min exposure
	2Monochlorophenol(2-MCP)	2.0038	46–73	7.3	13.4 × 1018	230 °C, 5 min exposure
	1,2Dichlorobenzene	2.0038	42–60	9.1	17.5 × 1018	230 °C, 5 min exposure
	Hydroquinone	2.0030	42–60	10.2	32.7 × 1018	230 °C, 5 min exposure
	Catechol	2.0036	42–60	10.7	66.4 × 1018	230 °C, 5 min exposure
CuO	2,4Dichloro1naphthol	2.0034	81	4.17	2.38 × 1022	25–300 °C,5% CuO on SiO2	[36]
Al2O3	2,4Dichloro1naphthol	2.0036–2.0063	108	3.96/4.13	5.82 × 1022	25–300 °C,5% Al2O3 on SiO2
ZnO	2,4Dichloro1naphthol	2.0029–2.0039	68	4.96/4.52	5.42 × 1022	25–300 °C,5% ZnO on SiO2
NiO	2,4Dichloro1naphthol	2.0029–2.0039	86	3.47/3.16	3.72 × 1021	25–300 °C,5% NiO on SiO2
Fe(III)	Anthracene	2.0028–2.0039	22.73	Not reported	Not reported	Room temperature (~23 °C), dark conditions, relative humidity ~7%, clay interlayers	[31]
Cu(II)	Anthracene	2.0028–2.0039	21.28			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Ni(II)	Anthracene	2.0028–2.0039	11.76			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Co(II)	Anthracene	2.0028–2.0039	18.52			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Cu(II)	Benzo[a]pyrene	2.0028–2.0039	13.70			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Ni(II)	Benzo[a]pyrene	2.0028–2.0039	43.48			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Co(II)	Benzo[a]pyrene	2.0028–2.0039	58.82			~23 °C, dark conditions, relative humidity ~7%, clay interlayers

Table 1. The half life of the interaction between typical transition metal oxides and precursor molecules in the generation of EPFRs.

Particles	Organic Adsorbate	Range of g-Values	Half Life	ΔH p-p (Gauss, G)	EPFR Concentration (Spins/g)	Reaction Conditions	Reference
Fe(III)2O3/Silica	Phenol	2.0040–2.0044	69.6–111 h	5.7–6.1	~1016–1017	150–350 °C At 400 °C: Radical decomposes to cyclopentadienyl species	[25]
	Catechol	2.0057–2.0067	111 h (max)	Not reported	~1017(max at 200 °C)	150–350 °C Maximum EPFRs yield at 200 °C
	Hydroquinone	2.0053–2.0056	69.6 h	Not reported	~1016 (low yield)	150–350 °C EPFRs undetectable at 400 °C
	2Monochlorophenol	2.0033–2.0052	24 h	Not reported	<1016	150–400 °C Adsorption via H2O elimination
	1,2Dichlorobenzene	2.0057	Unmeasurable	Not reported	<1015	150–400 °C
	Monochlorobenzene	2.0033	Unmeasurable	Not reported	<1015	150 °C
Cu(II)O/Silica	Catechol	2.0061–2.0070	~74 min	Not reported	Not reported	150–400 °C	[27]
	Hydroquinone	2.0062–2.0065	~4.6 days	Not reported	Not reported	150–400 °C Stable under vacuum
	Phenol	2.0032–2.0038	~27 min	Not reported	Not reported	150–400 °C Single radical type
	2Monochlorophenol	2.0041–2.0049	~24 h	Not reported	Not reported	150–400 °C
	1,2Dichlorobenzene	2.0063–2.0068	~1 h	Not reported	Not reported	150–400 °C
	Monochlorobenzene	2.0067–2.0070	~30 min	Not reported	Not reported	150–400 °C
Ni(II)O/SiO2	Catechol	2.0051–2.0065	~2.8 days	Not reported	Medium yield	250–300 °C	[34]
	Hydroquinone	2.0044–2.0057	~2.8 days		Low yield	Up to 400 °C
	Phenol	2.0035–2.0043	5.2 days		Lowest relative yield among all adsorbates	350 °C
	2Monochlorophenol(2-MCP)	2.0042–2.0043	3.8 days		Monotonic increase with temperature	150–400 °C
	Monochlorobenzene	2.0031–2.0038	2.4 days		Higher yield than phenol at >150 °C	300–400 °C
	1,2Dichlorobenzene	2.0034–2.0041	1.7 days		Highest yield below 350 °C	300 °C
ZnO/Silica	Phenol	2.0042	46–73	6.8	4.3 × 1018	230 °C, 5 min exposure	[35]
	Monochlorobenzene	2.0037	46–73	10.0	10.4 × 1018	230 °C, 5 min exposure
	2Monochlorophenol(2-MCP)	2.0038	46–73	7.3	13.4 × 1018	230 °C, 5 min exposure
	1,2Dichlorobenzene	2.0038	42–60	9.1	17.5 × 1018	230 °C, 5 min exposure
	Hydroquinone	2.0030	42–60	10.2	32.7 × 1018	230 °C, 5 min exposure
	Catechol	2.0036	42–60	10.7	66.4 × 1018	230 °C, 5 min exposure
CuO	2,4Dichloro1naphthol	2.0034	81	4.17	2.38 × 1022	25–300 °C,5% CuO on SiO2	[36]
Al2O3	2,4Dichloro1naphthol	2.0036–2.0063	108	3.96/4.13	5.82 × 1022	25–300 °C,5% Al2O3 on SiO2
ZnO	2,4Dichloro1naphthol	2.0029–2.0039	68	4.96/4.52	5.42 × 1022	25–300 °C,5% ZnO on SiO2
NiO	2,4Dichloro1naphthol	2.0029–2.0039	86	3.47/3.16	3.72 × 1021	25–300 °C,5% NiO on SiO2
Fe(III)	Anthracene	2.0028–2.0039	22.73	Not reported	Not reported	Room temperature (~23 °C), dark conditions, relative humidity ~7%, clay interlayers	[31]
Cu(II)	Anthracene	2.0028–2.0039	21.28			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Ni(II)	Anthracene	2.0028–2.0039	11.76			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Co(II)	Anthracene	2.0028–2.0039	18.52			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Cu(II)	Benzo[a]pyrene	2.0028–2.0039	13.70			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Ni(II)	Benzo[a]pyrene	2.0028–2.0039	43.48			~23 °C, dark conditions, relative humidity ~7%, clay interlayers
Co(II)	Benzo[a]pyrene	2.0028–2.0039	58.82			~23 °C, dark conditions, relative humidity ~7%, clay interlayers

(2)

Formation via stabilization in organic matrices

Apart from transition metals, the organic components of non-metals also play an important role in the formation of EPFRs. For example, persistent radical generation independent of metals is demonstrated in pyrolyzed lignin, where metal-free biochar develops carbon-centered radicals (g = 2.0035) with extended half lives exceeding 100 days. These radicals derive stabilization from delocalized electrons within condensed aromatic matrices, exhibiting characteristically narrow EPR linewidths (ΔHp-p < 5 G) that distinguish them from metal-mediated EPFRs (>8 G) [37].

Previous studies [38,39] showed that EPFRs concentration was significantly and positively correlated with the elemental carbon content. Thus, EPFRs generated inside the matrices of organic moieties are highly dependent on the organic structure.

The generation of EPFRs in the organic moieties typically occurred via three stages:

Step 1: Precursors, such as polycyclic aromatic hydrocarbons and phenolic compounds, undergo oxidation through mechanisms involving •OH, ozone, or transition metals. This process generates short-lived radicals; for instance, the reaction of phenol with •OH produces the phenoxy radical and H₂O. Step 2: Matrix interaction: The radical interacts with the organic matrix, forming covalent or non-covalent bonds (e.g., hydrogen bonds, π-π stacking). For example, aromatic rings in lignin stabilize phenoxy radicals through resonance.

Under conditions of light or heat, the ether bonds and carbon–carbon bonds within lignin undergo cleavage, initiating a cascade of radical-forming reactions (Figure 3). This process generates diverse free radical intermediates—including phenoxyl radicals (PhO•), methyl radicals (CH₃•), and other short-lived species—that undergo structural reorganization and stabilization. These intermediates play a critical role in EPFR formation: For instance, the cleavage of the C-O bond in a methoxy group (-OCH₃) yields a phenoxyl radical (PhO•) and a methyl radical (CH₃•) (Equation (1), Equation (2)). The phenoxyl radical reacts with molecular oxygen (O₂) to form a peroxyl radical (PhOO•) (Equation (3)), a reactive intermediate that can either rearrange or interact with other radicals. Notably, the phenoxyl radical may undergo hydrogen transfer with an adjacent carbon atom, stabilizing into a cyclic structure that lowers its energy state (Equation (4)). When this reaction occurs in the presence of water molecules and is driven by energy inputs (e.g., light or heat), it ultimately produces phenolic hydroperoxides (PhOOH) and hydroxyl radicals (HO•) (Equation (5)). These products, along with stabilized radicals from other pathways, contribute to the pool of environmentally persistent free radicals (EPFRs), whose longevity and reactivity pose significant environmental and health risks. Step 3: During stabilization the matrix donates electrons or provides steric hindrance, preventing radical re-combination or decay. This results in a metastable EPFR matrix complex with unique redox properties.

Detailed path:

Ph-O-CH₃ → PhO•+CH₃•

(1)

PhO•+H-C → PhO-H+C•

(2)

PhO•+O₂ → PhOO•

(3)

PhO•+H-C → PhOH+ C•

(4)

$$\mathrm{PhO}\text{•}+{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O} \stackrel{Pyrolysis, UV, etc.}{\to } \mathrm{PhOOH}+\text{•}\mathrm{OH}$$

(5)

3. Detection Techniques and Methods for EPFRs

3.1. EPR Spectroscopy

EPR spectroscopy is an effective analytical technique for detecting the EPFRs directly [40,41]. Originally, this method was developed through studies on paramagnetic salts like MnCl₂ [42,43]. The EPR has become a cornerstone for probing the electronic structure, crystallographic arrangements, dipole moments, and molecular geometries of complex cationic species, such as transition metal ions, free radicals, and alkali metal atoms [44,45]. The resonance spectra obtained through EPR provide detailed information regarding spin concentration, coordination environment, relaxation properties, and dynamic states of unpaired electrons within atoms or molecules [46,47].

The g-factor, a dimensionless parameter that quantifies the relationship between electron spin angular momentum and magnetic moment, serves as a structural fingerprint for paramagnetic species [48] (Figure 4). Specifically, distinct g-factor values are associated with specific electronic environments of radical-bearing molecules, thereby facilitating preliminary identification of EPFR classes [26,27]. This parameter is essential for identifying and characterizing free radicals, transition metal ions, and other species containing unpaired electrons [49,50]. The g-value can be calculated from the resonance height function in the EPR spectrum using the following equation [51,52].

$$g={\displaystyle \frac{h\nu }{\mu BB}}$$

$h$ is the Planck constant (6.626 × 10⁻³⁴ J·s); $\nu$ is the frequency (GHz); $\mu B$ is the Bohr magneton (9.274 × 10⁻²⁴ J/T); and B is the magnetic field (Gauss) [53].

Calibration can also be performed using standard samples with known g-values, such as 1,1-diphenyl-2-picrylhydrazyl (DPPH) and Mn²⁺ ions [54,55].

The peak-to-peak linewidth (ΔHp-p) is a critical parameter in EPR spectroscopy, serving as a quantitative measure of spectral broadening in spin-trapped EPR spectra [56,57] (Figure 4). This metric reflects both intrinsic radical properties (e.g., electron delocalization, hyperfine coupling) and environmental factors such as radical concentration, molecular motion, and spin–spin interactions [58,59]. For instance, delocalization of unpaired electrons across conjugated π-systems or metal-coordinated ligands (e.g., semiquinones in soot vs. hydroxylamines in aerosols) reduces spin–spin decoherence times (T₂), leading to pronounced linewidth broadening [60,61]. Furthermore, higher radical concentrations intensify exchange broadening via dipolar interactions, whereas restricted molecular motion—such as radicals immobilized on mineral surfaces—enhances motional narrowing, thereby reducing ΔH_p-p [62,63]. Such linewidth variations enable discrimination between EPFRs with distinct chemical environments—for example, localized phenoxyl radicals (narrow lines) versus delocalized π-radicals in charred biomass (broad, featureless spectra) [64,65,66].

EPFRs are categorized into three distinct classes based on their hyperfine coupling constants (g-values), which reflect differences in radical structure and electronic environment:

Low-g EPFRs (g < 2.0030): Predominantly comprise carbon-centered radicals (e.g., semiquinones, phenoxyl radicals) with localized unpaired electrons. The EPR spectra of these radicals display simple doublet or multiplet patterns, resulting from minimal electron–electron interactions [45,53,67]. These species are often associated with stable carbonaceous matrices like charcoals or soot.

Intermediate-g EPFRs (2.0030 < g < 2.0040): Represent mixed systems containing carbon-centered radicals adjacent to oxygen-containing functional groups (e.g., carbonyl or hydroxyl groups) or oxygen-centered radicals (e.g., superoxide adducts). The proximity of electronegative oxygen atoms induces splitting and hyperfine coupling, leading to complex multiple patterns in EPR spectra [54,68,69]. These radicals are common in combustion-derived organic aerosols.

High-g EPFRs (g > 2.0040): Primarily consist of oxygen-centered radicals (e.g., hydroperoxyl, alkoxyl radicals) or delocalized π-system radicals with significant spin–orbit coupling. Their intricate multiple spectra arise from strong hyperfine interactions with neighboring nuclei and low symmetry in electron density distribution [37,45,70]. These radicals are highly reactive and dominate in gas-phase combustion processes.

EPR spectroscopy was employed to analyze particulate matter (PM) samples from diverse combustion sources (Figure 5), focusing on equivalent particle-size fractions to ensure comparability [71,72,73]. Unique EPR spectral features were observed across different combustion sources, reflecting variations in radical physicochemical properties (e.g., g-values, linewidths, and signal intensities). For example, coal combustion particles exhibited notably stronger EPFR signals compared to biomass-derived particles at similar particle sizes (Figure 5) [74,75,76]. Coal-derived EPFRs showed higher g-values (2.0046), indicative of oxygen-centered radicals, aligning with abundant oxygen-functional groups in low-rank coal [76]. Biomass EPFRs (e.g., corn, wood) displayed lower g-values (2.0039–2.0041) that are consistent with carbon-centered radicals stabilized by lignin-derived aromatic structures or humic-like substances [76]. Formation of EPFRs was primarily influenced by humic-like substances (HULIS-C) (Equation (6)):

$$\mathrm{HULIS}\text{-}\mathrm{C} \stackrel{Light}{\to } \mathrm{EPFRs}$$

(6)

HULIS-C + Fe(Ⅱ) → HULIS-Fe(Ⅱ)

(7)

HULIS-Fe(Ⅱ) + O₂ → HULIS-Fe(Ⅲ) + O₂⁻

(8)

HULIS-Fe(Ⅱ) +H₂O₂ → HULIS-Fe(Ⅲ) + •OH + OH⁻

(9)

Following this, complexation between HULIS-C and Fe(II) occurs via carboxylate and phenolic functional groups, resulting in the formation of stable HULIS-Fe(II) complexes (Equation (7)). These complexes promote superoxide generation through catalytic oxidation cycles. Additionally, HULIS-Fe(II) species facilitate Fenton chemistry (Equation (8)), leading to bursts of hydroxyl radicals upon interaction with hydrogen peroxide (Equation (9)).

3.2. Gas Chromatography-Mass Spectrometry (GC-MS)

Although GC-MS cannot directly detect EPFRs due to their radical nature and non-volatility, it serves as a crucial tool in EPFR research by facilitating the analysis of precursor substances, degradation products, and environmental behaviors.

Active free radicals (e.g., peroxy radicals, •OH, and other reactive species) are commonly detected using indirect methods such as fluorescence (e.g., laser-induced fluorescence), chemiluminescence, chromatography, mass spectrometry (e.g., chemical ionization mass spectrometry, GC-MS, liquid chromatography-mass spectrometry [LC-MS]), and EPR spin-trapping technology [2,21,77]. In particular, mass spectrometry technologies measure the reaction products of active free radicals and chemical reagents.

For instance, high-performance liquid chromatography-mass spectrometry (HPLC-MS) and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS) were integrated to characterize the free radical adducts. Additionally, the semiquinone radical signals were effectively separated and identified [78]. In the investigation of the catalytic mechanism underlying benzene oxidation by copper oxides, EPR was utilized to detect free radical signals, including anthrone-type and anthraquinone-type radicals, while GC-MS was employed to identify both intermediate and final products resulting from these free radical reactions. This study elucidates the role of copper compounds as effective POPs [79].

3.3. Spectroscopic Methods

Spectroscopic techniques are among the most prominent approaches for the detection of free radicals, particularly through Ultraviolet-Visible (UV-Vis) spectroscopy and Infrared (IR) spectroscopy. UV-Vis spectroscopy is fundamentally based on electronic transitions induced by molecular absorption at specific wavelengths of light. The presence of unpaired electrons in free radicals leads to characteristic absorption bands within the UV-Vis region. This method is especially effective for detecting free radicals in solutions, where variations in absorbance at different wavelengths can provide insights into radical concentration and properties. It demonstrates high sensitivity and low detection limits [80]. IR spectroscopy utilizes the absorption of specific infrared wavelengths by unpaired electrons in radicals to generate distinctive IR absorption bands, facilitating structural and property analysis [81].

3.4. Complementary Characterization Methods for EPFR Studies

Various other characterization techniques, including nuclear magnetic resonance (NMR) and X-ray diffraction (XRD), have also been employed in the detection of EPFRs.

The identification of EPFRs has been systematically validated through advanced spectroscopic characterization techniques, particularly EPR and NMR. Current evidence emphasizes the catalytic role of EPFRs in promoting the •OH, primarily facilitated by the OH⁻ or H₂O through interactions with photogenerated holes (h⁺) [82]. To elucidate the role of EPFRs in photoinduced degradation processes, attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectroscopy was employed as a crucial tool for identifying molecular structural changes during interactions with pollutants [83]. In the study of CuₓO/SiO₂ catalysts for the oxidation of benzene to generate environmentally persistent free radicals, XPS and XRD were employed to confirm the coexistence of Cu(I) and Cu(II) within the catalyst. EPR directly detected the signal of phenoxyl radicals captured on the catalyst surface (g ≈ 2.003), thereby confirming the generation of O₂⁻ and·OH in aqueous solution [84]. The sorption of phenol onto Fe(III)CaM was characterized through FT-IR and XRD spectroscopy. FT-IR analysis identified distinct absorption bands within the ranges of 1345–1595 cm⁻¹ and 694–806 cm⁻¹, which correspond to vibrational modes associated with the functional groups of phenol. Concurrently, XRD data indicated an expansion in the interlayer spacing of Fe(III)CaM, further supporting the structural modifications induced by phenol uptake [57]. The integration of EPR and surface-enhanced Raman spectroscopy (SERS) enables the direct observation of the types and concentration variations of free radical intermediates on metal surfaces. Meanwhile, density functional theory simulations elucidate the stabilization mechanisms of metal–radical complexes [15,16]. In the investigation of the formation of EPFRs in fly ash and their correlation with sulfur content, EPR spectroscopy was utilized to quantitatively assess the concentration of EPFRs. X-ray Absorption Near Edge Structure (XANES) analysis was conducted to examine the binding forms of sulfur with metal oxides, thereby confirming the impact of sulfur on the generation of EPFRs [30]. In the study of the radical-mediated thermochemical formation mechanism of 2,3,6-trichlorophenol (TCP) catalyzed by copper oxide supported on silica to generate polychlorinated dibenzo-p-dioxins and furans (PCDD/Fs), in situ electron paramagnetic resonance spectroscopy, gas chromatography/quadrupole time-of-flight mass spectrometry (GC/Q-TOF-MS), and gas chromatography/high-resolution mass spectrometry (GC/HRMS) were used to identify the reaction intermediates, and it was demonstrated that the radicals attached to the particulate matter have a longer persistence [36]. Recent advances in specialized techniques, such as low-temperature matrix isolation electron paramagnetic resonance (LTMI-EPR), have been used to identify semiquinone, phenoxyl, and hydroxycyclopentadienyl radicals [85].

3.5. Chemical Methods (Free Radical Trapping Assays)

In addition to the aforementioned techniques, free radical trapping assays are widely employed in experimental research. These methods utilize specific trapping agents (e.g., DMPO (5,5-dimethyl-1-pyrroline-N-oxide), PBN (α-phenyl-N-tert-butylnitrone)—that react with free radicals to form stable spin adducts [68]. These adducts can subsequently be detected and characterized by using EPR [49,86].

The EPR spectrum pattern varied depending on the spin trap used. When DMPO was employed as the spin trap, a characteristic 4-peak spectrum was observed (Figure 6), consistent with its typical use in detecting •OH radicals. The signal intensity of the •OH adduct decreased over time, indicating radical decay (Figure 5). In contrast, the spin trap TEMPONE-H produced a 3-peak spectrum and was used to quantify total radical concentrations, as it reacts broadly with multiple radical species [87].

4. Sources and Distribution of EPFRs in Atmospheric Particulate Matter

The atmospheric concentrations of EPFRs typically range from 10¹⁶ to 10¹⁸ spins/g [88,89]. These radicals are primarily composed of carbon-centered radicals, oxygen-centered radicals, or heteroatom-substituted carbon-centered radicals, including those containing nitrogen or sulfur species. EPFRs are generated through diverse anthropogenic and natural processes, including combustion activities (e.g., biomass burning, fossil fuel combustion), vehicular emissions, industrial discharges, and photochemical reactions [90,91,92,93]. Among these, combustion sources dominate atmospheric EPFR contributions, accounting for approximately 60.6% of EPFRs in urban PM_2.5 as identified by source apportionment studies employing positive matrix factorization (PMF) [94]. This predominance arises because combustion processes—particularly incomplete burning of carbonaceous materials—generate stabilized radicals on particulate matter via pyrolysis and subsequent surface adsorption [95]. For instance, EPFRs in soot and ash exhibit prolonged lifetimes due to charge delocalization and interactions with mineral matrices, enabling their transport over regional scales [93]. Elevated concentrations of EPFRs have been documented during haze events compared to clear periods, with a preferential accumulation in fine (PM_1.0) and ultrafine (PM_0.1) particulate matter. Seasonal analyses further reveal that atmospheric concentrations of EPFRs in rural areas during culturally significant periods (e.g., the Lunar New Year) surpass urban PM_2.5 levels, demonstrating robust correlations with ionic markers (K⁺, Cl⁻) and polycyclic aromatic hydrocarbons (PAHs [96]. Comparative analyses of atmospheric PAHs and EPFRs during both heating and non-heating periods have revealed a significant increase in particulate-phase and gaseous-phase PAH concentrations during heating intervals. Investigations into the formation of EPFRs during thermal treatment of naturally stable waste have revealed significant differences between pyrolysis and incineration processes. Pyrolytic biochar is primarily composed of O-centered and C-centered radicals, while incineration fly ash exhibits negligible concentrations of EPFRs. Our recent data indicate that the mean g-factor of the EPFRs was measured at 2.0042 ± 0.0002 and 8.7616 ± 3.8196 in PM_1.1; 2.0043 ± 0.0003 and 5.6444 ± 0.7327 in PM_1.1–2.0; and 2.0043 ± 0.0002 and 7.0606 ± 3.1867 in PM_2.0–3.3, respectively. These values suggest that the samples collected from the atmosphere of Xuanwei (a region in Yunnan Province, China, known for its notably high incidence of lung cancer) predominantly consisted of oxygen-centered radicals, specifically phenoxyl and semiquinone radicals. This finding is consistent with typical organic free radicals reported to exhibit g-values ranging from 2.0039 to 2.0046 in chimney soot, coal, soil, and total suspended particles (TSP) sourced from the Xuanwei area [97].

Road dust and vehicular emissions are significant sources of EPFRs in urban environments, with their contributions exhibiting notable spatiotemporal variability [94]. Seasonal analyses indicate that street dust accounts for up to 39% of EPFRs in PM_2.5 during spring [77]. Furthermore, assessments of spatial distribution reveal elevated concentrations of EPFRs in high-traffic areas compared to ecological zones, primarily driven by tailpipe emissions (approximately 90%) and secondary transformations of volatile organic compounds within road dust matrices [98]. Beyond direct emission processes, EPFRs are also generated through secondary atmospheric mechanisms, particularly photochemical reactions. Experimental evidence indicates that moderate irradiation enhances the formation of EPFRs in humic substances, where radical stability is influenced by the functional group composition and physicochemical properties of organic matrices [99]. During the atmospheric aging of organic aerosols, gas-phase EPFRs have been shown to accelerate oxidation kinetics and alter aerosol hygroscopicity and optical properties, thereby affecting climate-relevant particle behavior [100]. Secondary photochemically derived EPFRs in particulate matter exhibit distinct characteristics—shorter lifetimes, increased reactivity, and heightened health risks—compared to predominantly combustion-generated radicals. This underscores their dynamic role in air pollution chemistry [101]; EPR analyses confirm that particles derived from biomass burning contain high initial concentrations of EPFRs. However, their lifetimes are significantly reduced under visible light irradiation (400–700 nm), highlighting light-dependent attenuation dynamics [102].

5. Toxicological Studies of the EPFRs

EPFRs have been implicated in triggering respiratory irritation and exacerbating symptoms of bronchial asthma. Inhalation of exogenous radicals adsorbed onto atmospheric particulate matter disrupts endogenous redox homeostasis, leading to systemic health risks, including DNA damage, cardiovascular issues, and respiratory diseases [24,103,104]. These associations have been documented in studies linking such exposures to oxidative stress [33,103,105]. Furthermore, chronic exposure to EPFRs has been associated with accelerated skin aging [46,106,107].

Experimental research underscores the central role of the EPFRs in mediating toxicity through their interactions with transition metals (e.g., Fe³⁺, Cu²⁺) in PM. These EPFR-transition metal complexes act as potent generators of ROS, including superoxide (•O₂⁻) and hydroxyl (•OH) radicals, which drive localized oxidative stress and inflammatory cascades at target sites [108,109,110].

It has been demonstrated that cells treated with EPFRs derived from CuO-fly ash exhibit significant loss of viability and impaired antioxidant mechanisms, such as depleted glutathione levels and suppressed superoxide dismutase activity, which compromise cellular redox balance. Additionally, biomarkers indicative of lipid peroxidation (e.g., malondialdehyde) and protein oxidation (e.g., 8-oxo-2′-deoxyguanosine) are elevated in cells exposed to EPFRs. The impact of EPFRs is not only evident at the cellular level but also observable in animal models and in vitro lung models. Short-term exposure to ZnO/MCB particles containing EPFRs induces significant pulmonary toxicity both in vitro and in vivo. Notably, the toxicity associated with ZnO/MCB is greater than that observed with ZnO alone, thereby establishing a direct link between EPFRs and respiratory damage. [23,111]

Beyond the respiratory tract, EPFRs impact systemic health. Chronic EPFR exposure correlates with vascular endothelial dysfunction, characterized by reduced nitric oxide bioavailability and increased expression of pro-inflammatory adhesion molecules. Epidemiological data link this to heightened risks of hypertension and myocardial infarction [112,113]. EPFRs could cross the blood–brain barrier, inducing microglial activation and pro-inflammatory cytokine release (e.g., TNF-α, IL-6). This neuroinflammatory milieu promotes amyloid-β aggregation and neuronal apoptosis, elevating susceptibility to Alzheimer’s disease and other neurodegenerative disorders [114,115].

6. Summary and Perspectives

This review synthesizes current knowledge on EPFR formation mechanisms, detection approaches, environmental behavior, and biological impacts. Despite recent advances, critical knowledge gaps still remain, particularly in understanding molecular-level formation pathways, spatiotemporal distribution, and long-term health effects under chronic low-dose exposure scenarios. Future research must address these challenges through standardized methodologies and predictive risk assessment frameworks. Below, we outline four key research priorities for advancing EPFR research:

• Formation Mechanisms

A comprehensive understanding of EPFR generation requires integration of in situ characterization techniques (e.g., synchrotron X-ray absorption spectroscopy) with computational modeling. Particular attention should focus on secondary photochemical processes, including charge–transfer reactions in atmospheric aqueous phases that govern radical propagation and aging under variable environmental conditions (humidity, irradiation).

2.

Analytical Method Development

Current analytical challenges in complex matrices (soil, PM2.5) demand advanced techniques, such as GC-EPR-MS systems, to improve detection sensitivity and matrix discrimination. The establishment of standardized quantification protocols for trace-level EPFRs in multi-pollutant environments is essential. Parallel development of portable EPR sensors could revolutionize real-time ambient monitoring and inform pollution control strategies.

3.

Environmental Transport and Fate

A systematic investigation of interfacial processes (air-water-soil) is needed to predict EPFR mobility and bioavailability. Key research directions could include humic acid-mediated binding dynamics in aquatic systems, hydrophobicity and ionic strength effects on partitioning, and synergistic interactions with multiple pollutants (heavy metals, PAHs) that may amplify ecotoxicological impacts.

4.

Biological and Health Impacts

Priority research areas include establishing chronic exposure thresholds through dose-response studies, elucidating oxidative stress cascades leading to ecosystem disruption (e.g., microbial community shifts), and mechanistic links to human pathologies. Critically, integrating EPFR data into region-specific air quality indices—through the development of standardized EPFR metrics and predictive models—would enhance risk assessment frameworks and inform evidence-based mitigation strategies. Addressing these interdisciplinary gaps will strengthen our capacity to reduce EPFR persistence in the environment, protect vulnerable populations, and promote sustainable ecosystem health.