Quantifying UV-Driven Aging of Sub-10 µm Airborne Microplastics with High-Resolution µFTIR-ATR Imaging

Abstract

Airborne microplastics (AMPs) undergo ultraviolet (UV)-driven physicochemical aging during atmospheric transport, influencing cloud processes, greenhouse-gas release, and potential respiratory health impacts. Quantifying this transformation is particularly challenging for particles smaller than 10 µm and for polymers such as polyethylene terephthalate (PET), whose intrinsic ester carbonyl band obscures newly formed acid carbonyls in conventional infrared analyses. Here, we develop a µFTIR attenuated total reflection (µFTIR-ATR) imaging method combined with a fourth-derivative oxidation index (carbonyl ratio at 1701/1716 cm⁻¹) that resolves these overlapping bands and enables sensitive, quantitative evaluation of PET surface oxidation. The approach automates detection, identification, and oxidation analysis of particles down to ~2 µm. Laboratory UV irradiation experiments show a systematic increase in this derivative-based oxidation index with exposure dose. Application to ambient PET collected from Mt. Fuji, Tokyo, Osaka (Japan), and Siem Reap (Cambodia) reveals clear regional differences corresponding to local UV-A environments: PET from Siem Reap exhibited the highest oxidation, whereas particles from the Japanese sites showed moderate but variable aging. These results demonstrate that derivative-based µFTIR-ATR imaging provides a practical and highly sensitive tool for quantifying photo-oxidative degradation in fine AMPs and highlight the value of chemical-aging metrics for interpreting atmospheric processing and transport pathways.

1. Introduction

Airborne microplastics (AMPs) are widely distributed from urban regions to remote environments and exhibit clear long-range atmospheric transport, forming an atmospheric branch of the global plastic cycle [1,2]. During transport, AMPs undergo ultraviolet (UV) photo-oxidation, hydrolysis, and heterogeneous reactions with atmospheric constituents, leading to surface roughening, enhanced hydrophilicity, and the accumulation of water-soluble salts and organic matter [3]. In particular, polyethylene terephthalate (PET) is known to undergo photochemical degradation through photoexcitation of ester chromophores, followed by chain scission and formation of oxygen-containing functional groups such as carboxylic acids and hydroxyl groups [4,5,6]. These photochemical modifications can influence cloud microphysical processes, with recent studies reporting both suppression and enhancement of cloud condensation nuclei and ice-nucleating particle activities depending on polymer type and degradation state [7,8,9,10,11]. UV-driven polymer degradation can also release greenhouse gases such as methane and carbon dioxide [12]. In addition, recent studies indicate that chemical aging strongly influences the health relevance of AMPs [13], while toxicological experiments demonstrate that UV-aged PET microplastics can enhance airway inflammation via degradation-mediated terephthalic acid release [14]. Furthermore, the detection of AMPs in human lung tissue underscores their ability to penetrate deep into the respiratory system and persist after inhalation [15,16,17]. Together, these atmospheric and health implications highlight the need for quantitative indicators of the chemical aging of AMPs [11].

Despite this growing recognition, source apportionment of AMPs has relied primarily on polymer type, morphology, and abundance, with limited integration of chemical-aging information [3]. Incorporating UV-induced degradation as an additional dimension could help distinguish freshly emitted particles from atmospherically processed ones and provide insight into their transport pathways. However, quantitatively evaluating UV-induced chemical aging in sub-10 µm AMPs remains technically challenging. Conventional infrared microscopy typically detects particles larger than ~30 µm and overlapping absorption bands—particularly in polymers containing intrinsic carbonyl groups—complicate quantitative oxidation assessments. Raman microspectroscopy offers submicron resolution [18], but fluorescence interference and low throughput restrict its applicability to large environmental samples [19,20]. These limitations highlight the need for an analytical method capable of resolving subtle oxidation features in sub-10 µm AMPs while processing large numbers of AMPs and samples at a practical speed suitable for environmental monitoring.

To address these challenges, we developed a micro-Fourier transform infrared attenuated total reflection (µFTIR-ATR) imaging approach [11] that enables automated detection and chemical identification of particles smaller than 10 µm and allows quantitative evaluation of subtle UV-driven surface oxidation. A key innovation of this method is a fourth-derivative oxidation index (1701/1716 cm⁻¹), which resolves overlapping ester and carboxylic acid carbonyl bands and therefore provides a sensitive aging metric even for polymers with strong intrinsic carbonyl signals such as PET. PET was selected as the target polymer because it is one of the most frequently detected AMPs in atmospheric environments [3,11], poses potential health risks due to the release of terephthalic acid during aging, and presents a rigorous analytical challenge. Finally, we apply this approach to ambient PET particles collected from Mt. Fuji, Tokyo, Osaka (Japan), and Siem Reap (Cambodia) to investigate regional differences in photo-oxidative aging and evaluate the usefulness of degradation-based indicators for interpreting atmospheric transformation processes and potential source attribution of AMPs.

2. Materials and Methods

2.1. Sampling of Airborne Aerosols

Atmospheric aerosols were collected in August at four locations: the summit of Mt. Fuji, Japan, Tokyo and Osaka, Japan, and Siem Reap, Cambodia (see

Table 1. Geographical information and mean surface UVA irradiance (July–August 2021) for each sampling site.

Site	Latitude (°N)	Longitude (°E)	Elevation (m a.s.l.)	Shortwave Radiation ** (kWh m−2 Day−1)	UVA (315–400 nm) ** (kWh m−2 Day−1)	UVA/Shortwave
Tokyo	35.69	139.72	93	4.82	0.304	0.063
Osaka	34.58	135.49	40	4.75	0.300	0.063
Mt. Fuji	35.36	138.73	3776	4.82 *	0.304	0.063
Siem Reap	13.36	103.85	17	5.34	0.338	0.063

* Mt. Fuji values represent conservative estimates based on grid-averaged surface irradiance from NASA POWER; the actual UV exposure at the summit is likely higher due to elevation effects. ** Shortwave radiation refers to ALLSKY_SFC_SW_DWN, and UVA to ALLSKY_SFC_UVA in the NASA POWER dataset.

for geographical details). At all sites, size-segregated aerosol samples were collected into >10 µm, 10–2.5 µm, and <2.5 µm aerodynamic diameter fractions. In Tokyo, Osaka, and Siem Reap, sampling was performed using a multi-nozzle cascade impactor (Tokyo Dylec Co., Ltd., Tokyo, Japan) operated at a constant volumetric flow rate of 20 L min⁻¹. Particles were collected onto PTFE-bonded glass fiber filters (TX40HI20-WW, Pallflex, Putnam, CT, USA). Samples at the summit of Mt. Fuji were collected using a high-volume air sampler (Shibata Scientific Technology Ltd., Soka, Japan) equipped with a PM_2.5 size-selective inlet, operated at a flow rate of 480 L min⁻¹. After sampling, all filters were sealed, stored in the dark at room temperature, and subjected to pretreatment prior to analysis.

2.2. Sample Pretreatment

Pretreatment followed a water-based extraction and oxidative cleanup workflow with minor modifications. Filter sections were sonicated in ultrapure water; filtrates were passed through hydrophilic PTFE membranes to remove soluble inorganics. Residues were oxidized with 30% H₂O₂ for two days to reduce organic matrix. Density separation with NaI (1.5 g cm⁻³) enriched plastic particles, and supernatants were filtered onto alumina Anodisc membranes (0.2 µm pore, 4 mm active area, Whatman, Marlborough, MA, USA). Filters were dried in a desiccator before analysis [11,21,22].

2.3. UV Irradiation Experiments

Figure 1 shows the schematic diagram of the UV irradiation experimental setup. Polymer test piece of PET (10 × 10 × 2 mm³, Standard Test Pieces Co., Ltd., Hiratsuka, Japan) was irradiated in a flow-type Pyrex chamber (Ø100 mm, height 200 mm, 0.157 L) sealed with a synthetic quartz window and silicone gasket. A xenon short-arc lamp (SX-UID 501 XAMQ, Ushio, Tokyo, Japan) with an IR-cut filter (SC1201, Asahi Spectra, Tokyo, Japan) provided UV-visible illumination while limiting thermal load. Dry N₂ (>99.99%) flowed at 0.5 L min⁻¹. To maintain the PET surface at 20 °C during UV irradiation, the cooling circulator (LTC-450A, AS ONE, Osaka, Japan) was set to 14.0 °C to offset for the temperature increase caused by UV exposure in the sealed chamber. Irradiance at the sample surface was ~10 mW cm⁻² (ISA-3151, T&D, Matsumoto, Japan). Irradiation was conducted for periods ranging from 1 day (24 h) to 7 days (168 h). To account for the spatial non-uniformity of irradiance on the sample surface, ATR imaging measurements were performed at the same positions where irradiance was measured on the polymer test pieces.

2.4. μFTIR ATR Imaging Analysis

2.4.1. μFTIR ATR Imaging Measurement

µFTIR-ATR imaging measurements were performed using a Spectrum 3/Spotlight 400 system (PerkinElmer, Waltham, MA, USA) equipped with a Cassegrain objective and a high-refractive-index germanium (Ge) ATR crystal (n ≈ 4.0). Total internal reflection at the Ge crystal generated an evanescent field probing less than approximately 1 µm into the sample surface, providing high sensitivity to surface functional groups. A large-area Ge crystal with a contact diameter of 750 µm was used to enable efficient imaging measurements. ATR imaging surveys were conducted over five representative fields per alumina filter, corresponding to approximately 22.4% of the effective filter area. Each field covered an area of 750 × 750 µm², and spectra were acquired with a spatial sampling of 1.56 µm per pixel, which is sufficient to detect and map particles with diameters of approximately 1.5–3 µm or larger. All measurements were performed at room temperature.

2.4.2. Polymer Identification by Spectral Matching

Polymer identification was carried out using Spectrum 10 software (PerkinElmer, Waltham, MA, USA) by spectral matching against a curated in-house ATR library consisting of approximately 140 reference spectra, including virgin and UV-aged polymers, polymer additives, and inorganic and biogenic interferents. This in-house library was integrated with approximately 10,000 commercial reference spectra. Similarity scoring was based on Pearson correlation coefficients, and spectra showing high similarity scores higher than 90% were accepted as tentative polymer identifications, following established procedures [11,21,22].

2.4.3. Spectral Preprocessing and Oxidation Index Calculation

For PET particles, fourth-order derivative spectra were calculated to enhance the resolution of overlapping carbonyl absorption bands. The number of data points used for the derivative calculation was set to five, providing an optimal balance between noise suppression and wavenumber resolution [23]. The oxidation index was defined as the ratio of the carbonyl-related peak intensities at 1701 and 1716 cm⁻¹ (I₁₇₀₁/I₁₇₁₆) obtained from the fourth-order derivative ATR-FTIR spectra.

2.5. Regional UV Irradiance Data and Statistical Analysis

To quantitatively evaluate regional UV environments during the aerosol sampling period, site-averaged surface UVA irradiance data (315–400 nm) were obtained from the NASA POWER dataset for July–August 2021. Because the laboratory UV exposure experiments employed a UVA-rich light source, UVA irradiance was selected as the primary indicator of ambient UV intensity.

Daily integrated UVA values were extracted for each sampling site (Tokyo, Osaka, Mt. Fuji, and Siem Reap), and averaged over the sampling period to represent regional UV conditions. These values were used for comparison with PET particle-level oxidation indices.

Differences among sampling sites were evaluated using the Kruskal–Wallis test, followed by pairwise Mann–Whitney U tests where appropriate. Geographical information (latitude, longitude, and elevation) for each sampling site was compiled from site metadata. The resulting site-averaged surface UVA irradiance values are summarized in Table 1.

3. Results

Table 1 summarizes the geographical characteristics of the sampling sites and the site-averaged UVA irradiance during the sampling period.

3.1. Derivative-Based Quantification of PET Degradation

In the spectra of airborne PET particles (upper panel of Figure 2), the gray line represents the non-irradiated PET standard, the blue line in Tokyo, the green line in Osaka and the red line in Siem Reap. In the laboratory PET standards (lower panel), the black line represents the non-irradiated polymer, the red line the 1-day UV-irradiated sample, and the blue line the 7-day UV-irradiated sample.

Fourth-order derivative processing clearly resolved the carbonyl-related absorption bands at 1701 and 1716 cm⁻¹ in both laboratory and environmental samples [24]. In the laboratory PET standards, the relative intensity of the 1701 cm⁻¹ band increased systematically with UV irradiation time. Airborne PET particles collected in Siem Reap exhibited a stronger 1701 cm⁻¹ band than those from Tokyo, while the spectral features of Tokyo particles were comparable to those of moderately UV-irradiated laboratory samples.

In the airborne PET particles (upper panel of Figure 2), those collected from Tokyo and Osaka exhibited spectral features comparable to moderately UV-irradiated laboratory standards, whereas particles from Siem Reap displayed a stronger 1701 cm⁻¹ band, indicating a higher degree of photo-oxidative aging.

These representative spectra demonstrate that the fourth-derivative ATR-FTIR approach can discriminate subtle differences in PET oxidation states among AMPs, with quantitative evaluation of particle-to-particle variability and regional differences presented in Figure 3.

3.2. Comparison of Ambient PET Samples from Different Regions

While the representative spectra in Figure 2 illustrate qualitative differences in PET oxidation states resolved by the fourth-derivative ATR-FTIR approach, quantitative assessment of particle-to-particle variability and regional differences requires analysis of a large number of particles. Figure 3 presents the distributions of the fourth-derivative oxidation index (I₁₇₀₁/I₁₇₁₆) for ambient PET particles collected from each sampling site.

Distinct regional differences are observed among the four sampling sites. Statistical analysis using the Kruskal–Wallis test confirmed significant differences in the oxidation index distributions, and post hoc pairwise Mann–Whitney U tests indicated that Siem Reap samples were significantly more oxidized than those from the Japanese sites.

PET particles from Mt. Fuji exhibit oxidation index values comparable to, and in some cases slightly higher than, those from Tokyo and Osaka. Notably, the Mt. Fuji samples show a broader distribution of oxidation indices, indicating considerable variability among individual particles. In contrast, Tokyo and Osaka samples display narrower distributions with intermediate oxidation index values.

PET particles from Siem Reap consistently show the highest oxidation index values, with distributions clearly shifted toward higher oxidation states compared with the Japanese sites.

4. Discussion

4.1. Interpretation of Photochemical Degradation Signatures

The observed increase in the acid carbonyl band at 1701 cm⁻¹, along with the broadening of the carbonyl and hydroxyl regions in both laboratory-irradiated and ambient PET particles, is consistent with characteristic pathways of UV-induced polymer degradation. PET photodegradation typically proceeds through photoexcitation of the ester chromophore, followed by chain scission, formation of peroxy radicals, and subsequent generation of oxygen-containing functional groups, including carboxylic acids and hydroxyl moieties. These transformations align with established photochemical mechanisms reported for PET [4,5,6].

Within this mechanistic framework, the fourth-derivative-based oxidation index provides a quantitative measure of surface oxidation by resolving newly formed acid carbonyls from the intrinsic ester carbonyl band. This approach enables direct comparison of photochemical degradation signatures among PET particles collected under different atmospheric conditions.

4.2. Relationship Between Oxidation Index and Regional UV Exposure

The regional differences observed in the fourth-derivative oxidation index (Figure 3) generally reflect variations in ambient UV exposure during the sampling period. As summarized in Table 1, site-averaged surface UVA irradiance during July–August 2021 was highest in Siem Reap and lower at the Japanese sampling sites. Correspondingly, PET particles from Siem Reap exhibited significantly higher oxidation indices than those from Japanese sites, as confirmed by non-parametric statistical analysis.

Among the Japanese sites, PET particles from Mt. Fuji displayed oxidation index values comparable to those from Tokyo and Osaka, but with a broader distribution. This variability likely reflects heterogeneous atmospheric transport histories and residence times prior to sampling, rather than differences in local UV intensity alone. Although high-altitude environments generally experience enhanced UV irradiance, the lower ambient temperatures at Mt. Fuji may partially suppress oxidative reaction kinetics, offsetting the effect of increased UV exposure.

These results suggest that ambient UVA intensity provides a useful first-order indicator of photo-oxidative aging of airborne PET microplastics, while additional environmental factors—including atmospheric transport, temperature, and exposure history—contribute to particle-to-particle variability. Importantly, the fourth-derivative oxidation index should be interpreted as a semi-quantitative proxy for photochemical aging, rather than a direct measure of cumulative UV dose.

4.3. Comparison with Previous Studies

Previous studies using ATR-FTIR or µFTIR imaging have primarily focused on polymer identification and qualitative assessment of environmental weathering, rather than on quantitative evaluation of photochemical degradation in microplastics [20]. Although large-area ATR systems and Ge crystal-based imaging approaches can detect sub-10 µm particles, these studies mainly emphasized particle detection performance and did not establish robust metrics for quantifying oxidation at the particle level.

Quantitative assessment of polymer oxidation has often relied on the conventional carbonyl index (CI), which has been widely applied to polymers such as polyethylene and polypropylene [25,26]. However, this approach is unsuitable for PET, because the intrinsic ester carbonyl band of PET overlaps with newly formed carboxylic acid carbonyls generated during photo-oxidation. As a result, the CI underestimates the extent of oxidation in PET and lacks sensitivity to early-stage degradation.

In contrast, the fourth-derivative ATR-FTIR approach employed in this study enables spectral separation of newly formed acid carbonyl bands from the intrinsic ester carbonyl band of PET. This allows surface oxidation to be quantified at the level of individual airborne particles, even when the oxidation is subtle. As demonstrated by the regional comparisons presented here, this approach provides enhanced sensitivity for detecting systematic differences in photochemical aging that would be difficult to resolve using conventional FTIR-based metrics. By extending quantitative oxidation analysis to airborne PET microplastics, which experience distinct atmospheric transport pathways and UV exposure histories, the present work complements previous studies that have largely focused on marine or terrestrial microplastics and provides new insight into atmospheric aging processes from a particle-level perspective.

4.4. Methodological Considerations and Uncertainties

Several methodological considerations should be taken into account when interpreting the oxidation index results presented in this study. First, the oxidation index from fourth-order derivative ATR-FTIR spectra reflects chemical modifications at the particle surface rather than bulk polymer degradation, owing to the shallow penetration depth of the ATR evanescent field. While this surface sensitivity is advantageous for probing early-stage photo-oxidation, it also implies that subsurface degradation processes are not captured.

Second, the interpretation of oxidation indices in relation to ambient UV exposure is influenced by environmental factors beyond UV intensity alone. At high-altitude sites such as Mt. Fuji, ambient UV irradiance may be higher than at urban lowland sites; however, lower ambient temperatures can reduce oxidative reactions kinetics and partially offset the effect of enhanced UV exposure [5,6]. Regarding sample pretreatment, a substantial fraction of ambient PET particles exhibited very low oxidation index values, comparable to those of non-irradiated reference materials. This indicates that the pretreatment procedure did not induce uniform or artificial oxidation, and any effect on PET surface chemistry was limited under the conditions applied.

In addition, particles collected at high-altitude sites are likely affected by heterogeneous atmospheric transport pathways and residence times prior to sampling. Beyond these considerations, several limitations and opportunities for future work should be noted. This study focused on PET as a representative polymer, and extension of the fourth-derivative ATR-FTIR approach to other polymers containing carbonyl functionalities, such as poly(methyl methacrylate) (PMMA) and ethylene–vinyl acetate (EVA), would help assess the broader applicability of degradation-based indicators. Moreover, the use of site-averaged UV irradiance data provides a regional context but cannot fully resolve particle-specific exposure histories. Combining chemical oxidation metrics with atmospheric transport analyses would further improve interpretation of photochemical aging in AMPs.

5. Conclusions

This study established a derivative-based µFTIR-ATR imaging approach for quantifying the photo-oxidative aging of airborne polyethylene terephthalate (PET) microplastics smaller than 10 µm. The fourth-derivative oxidation index (I₁₇₀₁/I₁₇₁₆) newly formed acid carbonyl groups could be spectrally separated from the intrinsic ester carbonyl band of PET, enabling sensitive evaluation of surface oxidation at the level of individual particles. Laboratory UV irradiation experiments confirmed that this derivative-based index increases systematically with irradiation time, demonstrating its suitability for detecting early-stage photo-oxidative degradation.

Application of this method to ambient PET particles collected from Mt. Fuji, Tokyo, Osaka, and Siem Reap revealed clear regional differences in oxidation index distributions. PET particles from Siem Reap exhibited significantly higher oxidation indices than those from the Japanese sites, whereas particles from Mt. Fuji, Tokyo, and Osaka showed intermediate but variable oxidation states. These regional patterns were broadly consistent with differences in ambient UV exposure. In contrast, the wide particle-to-particle variability observed at the high-altitude Mt. Fuji site highlights the influence of heterogenous atmospheric transport histories, in addition to local UV conditions.

Overall, the fourth-derivative µFTIR-ATR framework presented here provides a practical and semi-quantitative tool for assessing the photochemical aging of AMPs. While the oxidation index should not be interpreted as a direct measure of cumulative UV dose, it serves as a sensitive chemical indicator of surface-level aging that complements conventional particle characterization metrics. Extension of this approach to other polymer types and integration with atmospheric transport analyses will further advance understanding of the atmospheric life cycle and environmental implications of AMPs.