Rapid detection of microplastics and nanoplastics in seconds by mass spectrometry

Abstract

Microplastics (MPs) and nanoplastics (NPs) are pervasive pollutants and their analyses by traditional mass spectrometric methods require time-intensive sample preparation (e.g., extraction, digestion, and separation). This study presents a rapid and novel method for detecting MPs and NPs using flame ionization mass spectrometry (FI-MS) in which a dried sample (e.g., powder, soil and tissue) is directly burnt or heated with a flame in front of the MS inlet. FI-MS enables decomposition and ionization of various plastics such as polyethylene terephthalate (PET) and polystyrene (PS), allowing for analysis to be completed as fast as 10 seconds per sample. As a demonstration of application of this technique, PET contaminants in 1 L of bottled water or in 0.65 L of apple juice contained in plastic bottles were quickly detected from a filter paper after sample filtration and brief drying. A 0.89 mg soil sample spiked with 6000 ppm PET microplastics was measured to contain 4.98 μg of PET (5595 ppm, quantitation error: 6.8 %). Strikingly, PS nanoplastics (200 nm size) in mouse placentas were successfully identified and quantified, highlighting the method’s ability to analyze biological tissue without tedious sample preparation. Overall, this study demonstrates the high potential of FI-MS for real-world sample analysis of MPs and NPs in environmental, biological, or consumer product samples.

Graphical Abstract

1. Introduction

Microplastics (MPs, <5 mm in size) and nanoplastics (NPs, <1 μm in size) are tiny plastic particles that have emerged as a significant environmental concern over the past few decades. There are many different sources of MP/NP contamination, including the breakdown of large plastic debris [1], the release of microbeads from cosmetic products [2], and synthetic fibers shed from textiles [3]. MPs/NPs are persistent in marine and terrestrial environments, representing potential threats to ecosystems and human health. Evidence suggests that MPs/NPs are present in the food chain, drinking water, urine, blood and even the air we breathe. Humans can be exposed to MPs/NPs through ingestion, inhalation, and dermal contact [4]. Once inside the body, MPs/NPs can accumulate in tissues, potentially leading to inflammatory responses, oxidative stress, and other adverse health and fertility effects [5,6]. Furthermore, research revealed that maternal exposure to polystyrene MPs during gestation can not only translocate to the fetal tissues [7], but also cause metabolic disorders in offspring, highlighting the risks associated with MPs/NPs [8,9]. Although the long-term effects of MP/NP exposure on human health are not yet fully understood, there is growing evidence linking MPs/NPs to various health issues, including respiratory and gastrointestinal disorders, as well as potential risks related to cardiovascular disease [10], endocrine disruption and carcinogenicity [11]. As MPs and NPs have raised tremendous public concerns, there is an urgent need to develop methods to detect and quantify the MPs/NPs in different sample matrices.

Advanced techniques, such as Fourier-transform infrared spectroscopy (FTIR) [12], Raman spectroscopy, micro-computed tomography (micro-CT) [13], and hyperspectral dark field microscopy [14–16], etc. have been developed to enhance accuracy and efficiency for detecting MPs/NPs. For example, polyethylene terephthalate (PET) contains repeating units of ester groups (-COO-) and benzene rings which show aromatic C-H bond stretching vibrations between 3100 and 2800 cm⁻¹ and ester carbonyl bond stretching at 1715 cm⁻¹ in FTIR spectra. A recent study [17] reported the detection of MPs/NPs in bottled water using stimulated Raman scattering, providing evidence that contamination of NP particles is exponentially higher than MPs. Micro-CT can distinguish MPs from sediments based on material density, which allows researchers to generate high resolution 3D images of MPs/NPs by directly scanning environmental samples. Dark field microscopy also identifies MPs/NPs utilizing a known spectral fingerprint. This technique is excellent for detecting and visualizing smaller NPs within laboratory-spiked samples and biological tissues [14–16]. However, each technique mentioned above has drawbacks. For instance, micro-CT cannot effectively distinguish different types of plastics when their densities are similar. FTIR spectroscopy is highly effective for detecting MPs larger than 20 μm, but it loses sensitivity when analyzing NPs or complex plastic mixtures due to overlapping absorption. Raman spectroscopy suffers from issues including background fluorescence interference, weak scattering signal from NPs, and long scanning time for large area imaging [18]. Dark field microscopy requires the development of a known spectral target before scanning the sample, rendering it ineffective for mixtures due to overlapping spectra.

High-resolution mass spectrometry (MS) is a well-established and powerful technique for chemical analysis. In contrast to traditional spectroscopic methods, it is highly sensitive and provides precise molecular weight information, enabling accurate chemical identification [19–23]. However, due to the strong hydrophobic nature and high molecule weight, synthetic plastic compounds are insoluble in water and most organic solvents, which increases the difficulty of their detection by MS. Therefore, to be able to ionize and analyze synthetic polymers, one solution is to use pyrolysis or thermal desorption methods to dissociate polymers into characteristic volatile organic compounds (VOCs) or semi-volatile organic compounds (SVOCs) that are subsequently subject to gas chromatography (GC) separation followed by MS detection, as demonstrated in the technique of pyrolysis gas chromatography/mass spectrometry (Pyr-GC/MS) [24–27] and the emerging new technique of thermal desorption proton transfer reaction mass spectrometry (TD-PTR MS) [28]. For instance, PET can be identified, based on the detection of its pyrolysis degradation products such as terephthalic acid and 4-((vinyloxy)carbonyl)benzoic acid (Scheme 1 illustrates one suggested pyrolysis decomposition pathway of PET) [25]. However, although they are powerful for analyzing MPs and NPs, both GC/MS and TD-PTR methods require a complex pyrolysis chamber and tedious sample preparation process prior to analysis (Table 1) [6,25, 27–30]. In some cases analyzing plant samples requires nitric acid digestion, density separation of plastics from sample matrix and multiple steps of organic extraction before sample introduction to the pyrolysis chamber [31,32]. Thereby, the sample pretreatment for analysis of MPs/NPs takes from hours to days (Table 1; in the case of using Pyr-GC/MS, the GC/MS analysis step takes an additional 30 min or more). In addition, MP/NPs might be lost during the lengthy sample preparation steps which causes problems for quantitative analysis. Therefore, it is critical and necessary to develop a direct, fast, accurate, and sensitive MS-based method to analyze MPs/NPs.

Scheme 1.

Proposed thermal decomposition pathway of PET.

Table 1.

Comparison of FI-MS with other MS-based methods for MP/NP analysis.

Analysis method	Sample type	Sample pretreatment time	Analysis time	Sample amount	Polymer	LOD	Ref.
FI-MS (our method)	MP powder	No pretreatment	10 s	1–2 mg	PET	–	–
	Bottled-water/apple juice	Filtration (5 min); dry (2 min)	30 s	0.65–1 L	PET	–	–
	Soil	Dry (5 min)	50 s	1 mg	PET	0.25 μga	–
	Tissue	Dry (20 min)	30 s	< 4 mg	PS	0.7b–1.3c μg	–
Pyrolysis GC/MS	Environmental and portable water	Digestion (2 days); filtration (2 h); dry (1 day)	30 min	2 L	PET	0.04 μg/L	[29]
	Plants	Dry (24 h), digestion, extraction (30 min)	30 min	3–5 g	PET	0.02 μg	[31]
	Biosolid	Sample drying (1 day), pressurized liquid extraction (5 min), solvent evaporation (30 min)	30–60 min	20–50 mg	PET	0.03 mg/g	[32]
	Human placental specimens	Digestion (3 days), ultracentrifugation (4 h), dry (24 h)	45 min	100–400 mg	PET	1.46 ppm	[27]
	Coastal sediments	Digestion and extraction (36 h), dry (72 h)	90 min	5 g	PET	0.5 μg	[33]

^aUsing 300 μm standard PS measured with FI-MS setup IV.

^bUsing 100 nm standard PS measured with FI-MS setup IV.

^cUsing 200 nm fluorescently labeled PS measured with FI-MS setup IV.

In this study, we developed a fast, direct, sensitive, and convenient method for analyzing MPs or NPs using flame ionization mass spectrometry (FI-MS). By directly burning a plastic sample (e.g., a small piece of plastic sheet, MPs/NPs powder adsorbed on a metal rod or a piece of filter paper) in a flame placed near the mass spectrometer inlet, we successfully detected characteristic degradation products from the flame thermal decomposition of the sample. Such an FI-MS analysis only takes as short as 10 seconds per sample, requiring no or minimal sample preparation (Table 1). The reason underlying the ionization of various plastic samples by flame is likely that the high temperature of the flame can cause simultaneous decomposition and ionization of resulting smaller fragments from plastic polymers. In previous studies [34,35], organic samples such as propyphenazone were ionized by flame to generate ions for detection by MS, but no analysis of plastics by FI-MS was reported before. Our results show that PET microplastic/nanoplastics from commercial bottled water and juice can be detected by FI-MS. FI-MS also enables detection and quantitation of PET from soil samples directly without extraction or isolation of plastics. Strikingly, PS nanoplastics from a mouse placenta can also be quickly identified and quantified without digestion, showing the strength of our FI-MS for analyzing MPs/NPs.

2. Experimental section

2.1. Chemicals and materials

Bottled water and apple juice were purchased from a large local retailer. PET microplastic standard with the average size of 300 μm was purchased from Goodfellow Corporation (Boulder City, NV). Standard poly(vinyl chloride) (PVC) microspheres (105–250 μm), polystyrene (PS) latex bead solutions (product code: LB1, 10 % solid) with an average particle size of 100 nm, cellulose membrane filter paper (0.7 μm pore size, 47 mm diameter), and methanol were purchased from Millipore Sigma (Temecula, CA). Standard polyethylene (PE) microspheres (10–106 μm) were purchased from Cospheric (Santa Barbara, CA). A Bic multi-purpose butane lighter was purchased from a large local retailer (average flame temperature 600–800°C) [36]. A propane torch was purchased from Bernzomatic with a torch flame temperature of 700–1600 °C [37]. The vacuum filter setup was purchased from ChemGlass Life Science CG-1424 (Vineland, NJ). Deionized water was generated using Direct-Q 5UV (Millipore Sigma). For animal tissue study, 200 nm red fluorescently labeled polystyrene (PS) nanospheres were purchased from Fischer Scientific (Dyed Red Aqueous Fluorescent Particles, 1 % PS in solution by weight, catalog #R200).

2.2. Instrument

2.2.1. FI-MS apparatus

All experiments were performed using a high-resolution Orbitrap Q Exactive mass spectrometer (Thermo Scientific, San Jose, CA). The MS instrument was set at 140,000 resolution, positive ion mode, 3 microseconds per scan, Automatic Gain Control (AGC) as 1E6, and maximum inject time as 30 microseconds. MS/MS was performed with high-energy collisional dissociation (HCD) energy as 30–50 eV and the mass selection width was set to 0.4 m/z.

Before the experiment, the commercial ion source of the mass spectrometer was removed to accommodate the setup of flame ionization ion source for analyzing different samples (different Setups I-IV are illustrated in the insets of Figs. 1–4). Basically, depending on sample nature, it can be directly introduced into a flame for decomposition and ionization (for samples like plastic sheet, and microplastic powder) or subject to FI-MS after brief drying (for soil and tissue samples).

Fig. 1.

MS spectra showing direct analysis a) a piece of clear PET plastic bottle from local retailer and b) a piece of green-colored PET plastic bottle from local retailer by FI-MS.

Fig. 4.

a) FI-MS spectra obtained from different amounts of PET microplastics (diluted with silica powder), the absolute intensities of m/z 193.05 are labeled in the spectra; b) calibration curve of PET microplastics analyzed by FI-MS, c) FI-MS of a non-spiked soil sample, and d) FI-MS spectrum of a soil sample spiked with PET microplastics.

2.2.2. FI-MS setup for direct ionization of bulk plastic materials

Bulk plastic materials (e.g., plastic bottles for bottled water) were cut into a small piece of plastic sheet (2 cm×3 cm). The plastic sheet was held using a metal forceps and placed into a flame generated using the Bic multi-purpose butane lighter for decomposition and ionization (see the Setup I shown in Fig. 1a inset). The time for heating the sample was 10 s and the resulting ions were monitored and recorded using the mass spectrometer.

2.2.3. FI-MS setup for direct ionization of MPs in a powder form

MPs were adsorbed onto a hot stainless-steel rod and then introduced into a flame generated using the Bic multi-purpose butane lighter for decomposition and ionization (see the Setup II shown in Fig. 2a inset). The time for heating the sample was 10 s and the resulting ions were monitored and recorded using the mass spectrometer.

Fig. 2.

MS spectra showing direct flame ionization of microplastics of a) PET, b) PVC and c) PE introduced by a stainless-steel rod.

2.2.4. FI-MS setup for analysis of MPs/NPs from bottled water and apple juice

All glassware used for water filtration was thoroughly rinsed and cleaned with a 50:50 water/methanol solution and subsequently dried in an oven prior to use. 1 L of purified bottled water from a local retailer was filtered through a cellulose-based fiber filter paper (0.7 μm pore size, 47 mm diameter) using a vacuum filtration system. The filter paper containing possible MP/NP particles from the water sample was then oven-dried for 30 min (or quickly dried by a hot plate in 2 min) to remove residual moisture before analysis. The dried filter paper was positioned 1 cm away from the MS inlet using a metal forceps (Setup III, Fig. 3a). The paper was then, ignited using a Bic Multi-purpose butane lighter (the lighter was removed once the paper was ignited) and MS data was collected concurrently. While the filter paper was burning, the resulting flame was aligned with the MS inlet during the FI-MS analysis.

Fig. 3.

FI-MS spectra collected from directly burning of a) a cellulose-based filter paper as a control blank, b) a cellulose-based filter paper after filtering 1 liter of commercial bottled water.

Additionally, 650 mL of apple juice contained in a plastic bottle from a local retailer was filtered in the similar way and then dried on a hot plate for 2 min to remove moisture. The filter paper was subsequently burned and analyzed using FI-MS.

2.2.5. FI-MS setup for ionizing MPs/NPs in soil or biological tissue

One gram of soil sample was collected from a local neighborhood and dried on a 50 °C hot plate for 5 min. Then 500 mg of dried soil was spiked with 3 mg of standard PET MPs (average size: 300 μm) to prepare a soil containing 6000 ppm PET. To facilitate the measurement of MPs in the soil sample, the soil sample was placed in a small glass vial (5 mm ID x 29 mm height) which was fastened using a clamp to secure its position to be ca. 0.5 cm away and below the MS inlet (see the Setup IV in Fig. 4c). For FI-MS analysis, the bottom of the vial was heated with the Bic multi-purpose butane lighter for decomposition while a propane torch flame positioned above the glass vial was employed for flame ionization of the resulting decomposition vapor at the same time with MS data recorded continuously (time for heating the sample: 50 s). There are a couple of reasons for using such a setup in which sample decomposition and ionization are separated. First, it is not easy to directly insert the soil particles into the flame. Second, using a glass vial to hold the soil sample can fix the position of the sample in the front of MS inlet, which is necessary for obtaining a reproducible ion signal for quantitative analysis. For NP analysis of mouse tissue, the same setup (Fig. 4c) was used.

2.3. Animal study design

Young adult CD-1 mice were purchased from Charles River (Wilmington, Massachusetts) at 60 days of age. Mice were housed in a Rutgers University Newark animal facility and allowed to acclimate to the facility prior to experimentation. Temperature was maintained at 22 ± 1 °C with 12 h light-dark cycles to provide a controlled housing environment. Food and water were provided ad libitum. All experiments were performed with Rutgers Institutional Animal Care and Use Committee (IACUC) approval (PROTO202100043) and a proof/certificate of approval is available upon request. At 8 weeks of age, male and female mice were trio-bred. Mice were inspected for plugs twice per day. The day at which a plug was observed was considered as gestational day (GD) 0.5 and the following day as GD 1. After plugs were observed, female mice were divided between treatment groups to receive vehicle control (ultrapure water) or 5 mg/kg of 200 nm PS NP. Oral dosing was performed by gently pipetting the dosing solution into the mouth. The dosing solutions were stored in the fridge and were mixed by gentle shaking immediately before dosing. Pregnant mice were dosed from GD 8–15 for 7 days. On GD 15, mice were euthanized by CO₂ following IACUC-approved methods. After euthanization, placentas were weighed and cut in half. Half was fixed in 4 % paraformaldehyde overnight and placed the next day in 70 % ethanol for imaging by enhanced darkfield hyperspectral microscopy. The other half of the collected placenta was snap-frozen in liquid nitrogen and stored at −80 °C for FI-MS.

In preparation for FI-MS, frozen mouse placenta tissue samples from control and PS-exposed animals were cut into approximately 1 mg pieces. Each piece was placed in a glass vial and vacuum-dried at 60°C for 20 min to remove moisture. The dried sample was then subjected to FI-MS analysis following the same procedure described for soil analysis except the heating time for tissue sample was reduced to 30 sec.

2.4. Enhanced darkfield hyperspectral microscopy

Fixed placentas were embedded in paraffin and sectioned at 5 microns. Tissues were cover slipped without staining for imaging. A CytoViva Enhanced Darkfield Hyperspectral Microscope (CytoViva, Inc., Auburn, AL, USA) was used to visualize the PS nanoplastics in sectioned placenta as previously described [15,38,39]. In brief, indirect illumination of the sample allows collection of refracted light and the visual differentiation of materials based on this unique light scatter fingerprint. Initially, the red fluorescently labeled polystyrene nanospheres described above were placed between clean glass slides and scanned to create a unique spectral library file of the material. Separate areas of the unstained control and exposed placental slides were then imaged. The ENVI 4.8 analysis software captures and combines this wavelength data in a pixel-row-by-pixel-row line scan to form a hyperspectral image and layers a spectral map algorithm to identify pixels emitting the unique wavelength fingerprint. These pixels are highlighted and identified as the experimental material (e.g., red fluorescently labeled PS NP).

2.5. Quantitative analysis

For quantitative analysis, the setup shown in Fig. 4c was adopted. In the case of quantitation of PET microplastics, a trace amount of PET standards (average size: 300 μm) was prepared via a solid dilution method using deactivated silica powder. Specifically, 1 mg of the PET standard was thoroughly mixed with 1999 mg of silica powder using a vortex mixer. From this mixture, 1 mg was precisely weighed and transferred into a glass insert, theoretically containing 0.5 μg of the PET standard. Four other different PET standard samples containing 1 μg, 2 μg, 5 μg, and 10 μg PET microplastics (total weight: 1 mg) were prepared similarly and transferred into a glass insert. For FI-MS analysis, the bottom of the vial was heated with the Bic multi-purpose butane lighter for decomposition while a propane torch flame positioned above the glass vial was employed for flame ionization of the resulting decomposition vapor at the same time with MS data recorded continuously (time for heating the sample: 50 s). Each sample was analyzed in triplicate measurements and the calibration curve was obtained by plotting the characteristic ion signal vs. the amount of the PET microplastics used for FI-MS analysis. The limit of detection (LOD) was defined as 3 times the noise signal divided by the slope of the obtained calibration curve.

For the quantitative analysis of PS nanoplastics, 0.5–10 μL of a 10-fold diluted standard PS nanoplastics latex solution (NPs size: 100 nm) were carefully pipetted into the glass insert, which theoretically contained 0.005–0.1 mg of PS nanoplastics. The water was then removed using speed-vac for 5 min at 60 °C. Each sample was tested in triplicate measurements in a similar manner as described above, except that the sample was heated for only 30 seconds using the butane lighter.

3. Results and discussion

3.1. Direct analysis of plastic materials by FI-MS

To test the feasibility of ionizing plastic materials by FI-MS via flame decomposition and ionization, we placed a rectangular strip (2 cm×3 cm), cut from a clear PET plastic bottle, near the mass spectrometer (MS) inlet and applied a Bic Multi-purpose butane lighter to directly burn the plastic strip, as illustrated in Fig. 1a inset. The resulting ions were monitored and detected with MS. Upon ignition, as shown in Fig. 1a, four major PET decomposition product ions were detected including the protonated vinyl benzoate (m/z 149.06), the protonated vinyl terephthalate (m/z 193.05), and the protonated di-vinyl terephthalate (m/z 219.07) and the protonated terephthalic acid dimer (m/z 385.09). In particular, the peak of m/z 193.05 is dominant with a high intensity of 3.3E7 in Fig. 1a. These characteristic decomposition compounds were previously detected by using Pyr-GC/MS [25]. We also examined a strip cut from a green-colored PET plastic bottle, observing the same four major ions with additional background peaks (Fig. 1b), in comparison to the MS spectrum obtained from the clear plastic strip (Fig. 1a). One peak at m/z 259.09 (Fig. 1b) was identified as the fragment ion [C₁₆H₁₁N₄]⁺ (measured mass m/z 259.09718, theoretical mass m/z 259.09782, mass error 2.4 ppm) from copper phthalocyanine (CuPC), a green pigment commonly used in plastic products [40]. This result demonstrates that plastics can be directly decomposed and ionized using a flame to generate characteristic ions for MS detection which only takes 10 s, suggesting a much faster and more handy method for their detection than traditional MS methods such as Pyr-GC/MS.

3.2. Direct analysis of microplastics by FI-MS

After success of directly examining plastics sheet using FI-MS, we reasoned that our method could be applicable to ionize MPs and NPs, as flame ionization does not depend on material sizes. As shown in Fig. 2a inset, PET microplastics were introduced into a flame using a stainless-steel rod for FI-MS analysis. As shown in Fig. 2a, ionization of the PET standard MPs (300 μm) yielded peaks, m/z 149.06, m/z 193.05, m/z 219.07 and m/z 385.09, identical to those observed from the plastic sheet from a clear PET bottle (Fig. 1a). Again, the peak of m/z 193.05 is dominant in the spectrum with an intensity of 2.2E7 (Fig. 2a). This result shows the capability of our FI-MS for analyzing not only bulk PET plastic materials like plastic sheets but also MPs. The total ionization and data acquisition time for each sample was only 10 seconds, while other methods such as pyrolysis GC-MS take hours to days for analyzing MPs, due to tedious sample pretreatment as mentioned before. To confirm the assigned structures of ions generated from flame ionization of PET micoplastics, tandem mass spectrometry (MS/MS) was performed. For instance, upon collision induced dissociation (CID), the protonated vinyl terephthalate (m/z 193.05) generated from FI of PET microplastics gave rise to fragment ions m/z 149.02 and 121.03, probably due to consecutive losses of CO₂ and C₂H₄ (Figure S1).

Other MPs such as PE and PVC were also tested by our FI-MS method. For the PVC microplastics, a number of known PVC thermal decomposition products were detected, matching with previous reports [29,41, 42]. As shown in Fig. 2b, three major aromatics from the pyrolysis of PVC were detected, such as naphthalene (measured m/z 129.07018, theoretical m/z 129.06988, mass error 2.9 ppm), 1-methylnaphthalene (measured m/z 143.08591, theoretical m/z 143.08553, mass error 2.7 ppm) and phenanthrene (measured m/z 179.08592, theoretical m/z 179.08553, mass error 3.3 ppm). For FI-MS analysis of PE microplastics, as shown in Fig. 2c, three hydrocarbons with different numbers of carbons differed by 14 Da (corresponding to one moiety of CH₂) were detected by FI-MS, such as hepta-1,6-diene (measured m/z 97.10150, theoretical m/z 97.10118, mass error 3.3 ppm), octa-1,7-diene (measured m/z 111.11700, theoretical m/z 111.11683, mass error 1.5 ppm) and nona-1,8-diene (measured m/z 125.13254, the theoretical m/z 125.13248, mass error 0.4 ppm). These detected ions indicate the occurrence of the PE polymer C-C bond cleavage during flame ionization [29,43].

3.3. FI-MS detection of microplastics/nanoplastics in commercial bottled water and juice

Micro- or nano-plastic contamination in bottled water emerged as a significant public concern. A previous study analyzed 259 plastic bottles from 11 global brands, revealing that 93 % of the samples showed signs of MP contamination [44]. The research attracted widespread public attention and prompted extensive investigations into MPs, including their detection and quantification in water, food, and the environment, as well as studies on the potential health effects of MPs/NPs [4,17,29,45, 46]. In this study, we tested the possibility of analysis of MPs in bottled water by FI-MS. First, as a control, a clean filter paper was first ignited near the MS inlet, allowing the paper to self-combust to generate ions for MS detection. Across three trials, no characteristic peak of m/z 193.05 from PET was detected (Fig. 3a), confirming the absence of PET contamination on the clean filter paper. Subsequently, 1 liter of bottled water was filtered through another piece of clean filter paper (700 nm pore size) using a vacuum filtration system, with the expectation that MPs/NPs of sizes larger than 700 nm would be captured by the filter. The collected filter paper was immediately dried and then burned in front of the MS inlet. Under the same FI condition, m/z 193.05 signal was detected (mass error: 0.2 ppm) when the filter paper was ignited (shown in Fig. 3b). To confirm that the detected signal of m/z 193.05 from the water sample was truly from PET and to avoid false positive assignment, MS/MS of m/z 193.05 was conducted. However, due to the weak signal of m/z 193.05 in Fig. 3b, it was challenging to obtain a clear MS/MS spectrum. To increase the ion intensity, in our experiment, a 200 V voltage was applied on a metal plate placed in front of the filter paper to generate an electric field (shown in Figure S2–a). The electric field pushed the positive ions generated by FI into MS and increased the m/z 193.05 signal by 65 times (Figure S2–b), allowing us to perform MS/MS analysis. As illustrated in Figure S2–c, MS/MS of the ion showed two characteristic fragment ions of m/z 149.02 and m/z 121.03, which is in an agreement with the MS/MS spectrum of m/z 193.05 generated from FI of PET microplastics standard (Figure S1) and confirms the existence of PET in this bottled water sample. This experiment was repeated two more times by filtration of 1 L of bottled water followed with FI-MS analysis of the filter paper and the generated ion of m/z 193.05 was confirmed each time by MS/MS (Figures S2–d and S2–e). This result shows the PET detection reproducibility from bottled water using our method. PVC and PE microplastics/nanoplastics were not found in this bottled water since no target PVC and PE decomposed products (m/z 97.10150 and 111.11700 for PE and m/z 129.07018, 143.08591, and 179.08592 for PVC) was detected in the acquired MS spectrum. These results suggest the feasibility of detecting MPs/NPs from water samples, after simple filtration and drying. In this case, the detected PET particles from the bottled water sample could come from weathering or decomposition of the plastic bottle which was comprised of PET material, or from the water purification and manufacturing process [47].

Furthermore, 650 mL of apple juice was filtered using filter paper and analyzed with FI-MS and MS/MS using the setup shown in Figure S2–a with the applied electric field. As shown in Figure S3–a, the characteristic PET peak at m/z 193.05 was detected in the acquired FI-MS spectrum (Figure S3–a), and its characteristic fragment ions at m/z 149.02 and m/z 121.03 in MS/MS spectrum (Figure S3–b). It tells that the apple juice contained PET microplastics/nanoplastics as well. This test result highlights that the analysis of MPs using our FI-MS is not only applicable to water sample but also can be used to samples with complicated matrices such as apple juice.

3.4. Detection and quantitation of PET microplastics in soil by FI-MS

Besides identification of MPs, our results also showed the feasibility of quantitative measurement of MPs. For quantitative analysis, it is important to get reproducible ion signals for sample analysis. In the case of FI-MS, the ion signals of plastics are affected by the positions of both flame and plastic samples. As explained above, for quantitative analysis, we adopted the FI-MS Setup IV (illustrated in Fig. 4c inset) in which a MP sample was placed in a small glass vial that was fastened and placed close to the MS inlet using a metal clamp on a ring stand. The bottom of the vial was heated with the Bic multi-purpose butane lighter for decomposition while a propane torch flame was positioned above the glass vial to trigger flame ionization of the resulting decomposition vapor at the same time with MS data recorded concurrently. To achieve the preparation of various trace amounts of PET microplastics, a solid dilution method was employed [45]. Basically, standard PET microplastics (average size: 300 μm) were mixed with inert silica powder in varying weight ratios, as described earlier, allowing precise preparation of PET microplastics samples with weights ranging from 0 to 10 μg contained in silica powder (total sample weight: 1 mg). As seen in Fig. 4a, the characteristic PET peak of m/z of 193.05 had increased intensity with increased amount of PET microplastics and a linear calibration curve was obtained with R² to be 0.98 (Fig. 4b). Note that there was a small background signal of m/z of 193.05 in the blank sample with intensity of 8.92 E3, which came from the propane flame (it is likely that the propane fuel contained impurities). The LOD of PET microplastics was calculated to be 0.25 μg (see detailed data and calculations in Table S1). This result demonstrated the capability for quantitative analysis of PET microplastics.

MP contamination in soil is an emerging problem for the environment, ecology, and agriculture. Many different methods have been used to characterize and quantify the MPs/NPs in soil, such as hyperspectral imaging technology and Pyr-GC/MS [48–50]. A previous study reported that PET microplastics in industrial areas was measured ranging from 300 to 67,500 mg/kg (300–67,500 ppm). This method used a complicated high pressure fluid extraction to separate MPs from other compounds in the soil such as fats and oil, prior to analysis [49]. In our study, we also tested MPs in soil samples by FI-MS. In our experiment, one soil sample for this study was collected from a residential neighborhood distant from industrial activity and dried to remove moisture. First, we took 1 mg of the soil and used the FI-MS Setup IV (Fig. 4c) to analyze the PET in the collected soil. From the acquired MS spectrum (Fig. 4c) of this soil sample, the decomposed PET product ion of 193.05 was detected with weak intensity. To distinguish whether the m/z 193.05 signal was produced by PET contamination or by background, MS/MS of the m/z 193.05 was conducted but its characteristic fragment ion at m/z 149.02 was not detected (Figure S4–b), suggesting that the m/z 193.05 signal from the soil (Fig. 4c) was not from PET and the soil we collected was free from PET.

Second, to further test the feasibility of identifying and quantifying PET in the soil sample, the soil sample was spiked with standard PET microplastics (average size: 300 μm), creating a 6000 ppm PET microplastics-containing soil sample. A 0.89 mg of the spiked sample was tested using the Setup IV (Fig. 4c). The signal intensity of the PET-specific peak at m/z of 193.05 (Fig. 4d) was 6.65E4 for this PET-spiked soil sample. To avoid false positive detection, MS/MS analysis of m/z 193.05 from the PET-spiked soil sample was conducted and its characteristic fragment ion of m/z 149.02 was detected (Figure S4–a), confirming that the signal m/z 193.05 was from PET indeed. Using the calibration curve in Fig. 4b, the PET mass detected in this PET-spiked soil was quantified as 4.98 μg (5595 ppm). Compared with the theoretical value of 6000 ppm, the quantification error was 6.8 % (Table S2). This result emphasizes that our FI-MS can not only identify PET microplastics from the soil but can also accurately quantify it.

3.5. Detection and quantitation of PS nanoplastics in tissue by FI-MS

In our study, we further examined the detection of NPs by FI-MS. First, FI-MS analysis of standard PS nanoplastics (size: 100 nm) was successfully conducted, as shown in Fig. 5a. Ions of PS decomposition products such as 1,3-diphenylpropane (measured m/z 195.11692, theoretical m/z 195.11683, mass error 0.5 ppm) and but-3-ene-1,3-diphenylpropene (measured m/z 207.11690, theoretical m/z 207.11683, mass error 0.4 ppm) were detected. Upon MS/MS fragmentation of m/z 195.12 (Fig. 5b), fragment ions m/z 167.09 and m/z 117.07 were seen, likely due to losses of C₂H₄ and C₆H₆, respectively. For the MS/MS fragmentation of m/z 207.12 (Fig. 5c), the peak at m/z 129.07 indicates a loss of C₆H₆ via from β-elimination and the peak at m/z 91.05 indicates a loss of C₉H₈. These observed decomposition product ions from PS matched the previously reported results from pyrolysis GC/MS studies [51,52], indicating that PS nanoplastics were thermally degraded via C–C bond cleavage to form diaromatic products. This result shows that FI-MS can be used to measure NPs. Furthermore, a calibration curve with a good linearity (R² = 0.99, Fig. 5d) was obtained by measuring different amounts of standard 100 nm PS nanoplastics and the LOD for detecting PS nanoplastics (100 nm size) was calculated to be 0.7 μg (Table S3). We further tested the fluorescently labeled PS nanoplastics with size of 200 nm using FI-MS in the same way and a calibration curve was obtained (R² = 0.99). The LOD was calculated to be 1.3 μg (Table S4). This result suggests that FI-MS for NPs analysis can be quantitative and sensitive.

Fig. 5.

a) FI-MS spectrum of standard PS nanoplastics; MS² spectra of b) m/z 195.12 and c) m/z 207.12; and d) calibration curve for FI-MS analysis of 100 nm standard PS nanoplastics.

Next, we examined the possibility of using our FI-MS technique to directly analyze tissue samples. As described in the experimental section, pregnant female mice were divided between treatment groups to receive vehicle control (ultrapure water) or 200 nm fluorescently labeled polystyrene NPs, from which placenta tissues with and without PS nanoplastics contamination were obtained. A darkfield hyperspectral microscopy image of the mouse placenta is shown in Fig. 6a, which confirms the presence of PS NPs in the tissue from the mouse fed with PS NPs. Next, both a control placenta tissue sample from the mouse that was not fed with PS nanoplastics (0.72 mg) and a test sample of placenta tissue from the mouse fed with 200 nm PS nanoplastics (3.66 mg) were tested using the FI-MS Setup IV (Fig. 4c), immediately after brief drying. The signal intensity of m/z 195.12 from the control tissue (red curve, Fig. 6b) was 9.24E2, which is close to the blank signal (8.48E2 Table S4), indicating that the control tissue lacked PS nanoplastics. Indeed, MS/MS of the signal of m/z 195.12 from the control tissue did not give rise to its characteristic fragment ion at m/z 167.09 (data not shown). In contrast, the test tissue sample from PS-fed mouse showed a higher ion intensity of 2.62E3 for m/z 195.12 (black curve, Fig. 6b). Upon MS/MS analysis of m/z 195.12, m/z 167.09 (Figure S6) was detected, verifying the existence of PS nanoplastics in PS-fed mouse placentas tissue. This result is in line with the darkfield hyperspectral microscopy image shown in Fig. 6a. Based on the calibration curve (Figure S5), the quantity of PS nanoplastics contained in the test tissue sample was 2.04 μg (i.e., 557 ppm). This quantitative result appears to be in line with a previous study by Campen et al. reporting PS content ranging from 6.5 to 685 ppm across 65 human placental specimen samples [27], supporting the validity of our quantification result. In combination, dark field microscopy paired with our innovative technique rapidly identified the cellular location of the PS NPs and polymer concentration within a biological sample. Compared with other methods, by direct analysis of tissue, our approach does not require time-consuming sample preparation such as tissue digestion and MP/NP extraction [28,29,53].

Fig. 6.

a) Darkfield hyperspectral microscopy image of mouse placenta. Pixel matches from a placenta exposed to 200 nm fluorescently labeled PS nanoplastic are indicated with white arrow and b) FI-MS spectra (background subtracted) of a placenta tissue from the mouse fed with 200 nm fluorescently labeled PS nanoplastics and a control placenta tissue from the mouse that was not fed with the PS nanoplastics.

4. Conclusions

In this study, we successfully developed a rapid and sensitive FI-MS method for detecting and quantifying trace levels of MPs/NPs (e.g., PET and PS), in diverse sample matrices, including bottled water, apple juice, soil, and mouse placenta tissue. In comparison to traditional methods (Table 1) for analyzing MPs/NPs such as Pyr-GC/MS, our approach offers several striking advantages. First, it is very quick (the analysis time can be in seconds with no or minimum sample pre-treatment). Second, by using high resolution MS and MS/MS analysis, our method provides detailed molecular information for MPs and NPs with high specificity. Third, while our LOD for PET (0.25 μg) is comparable to those of traditional methods (0.02–0.5 μg) for MP or NP analysis (Table 1), we used much less samples (e.g., we only used 1–4 mg of tissue/soil samples, whereas traditional methods used 20 mg-5g samples). It is of note that MS-based techniques do not identify the sizes of MPs/NPs within the sample, and should be paired with a visual technique to confirm intact MP/NP deposition. In summary, our method could have broad potential for applications in a range of environmental and biological matrices to meet today’s challenge of severe plastic pollution issues that our world is facing.

Environmental implication

Microplastics (MPs) and nanoplastics (NPs) are omnipresent in the environment and presents a critical environmental issue. They can accumulate in tissues, potentially leading to inflammatory responses, oxidative stress, and other adverse health effects. Trace detection of these compounds from complex samples is crucial for securing ecosystems and human health. This paper reports a novel and unprecedented method, which is both fast (as fast as 10 seconds per sample analysis) and sensitive (e.g., LOD of sub-micrograms), for trace detection and quantification of MPs and NPs in various environmental samples including water, juice, soil and biological tissue.

Supplementary Material

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jhazmat.2025.138322.

HIGHLIGHTS.

• Developed a sensitive and rapid mass spectrometric method for detecting microplastics (MPs) and nanoplastics (NPs).

• PET particles in bottled water or apple juice were quickly detected.

• A soil sample spiked with PET microplastics was successfully quantified (quantitation error: 6.8%).

• PS nanoplastics (200 nm size) in mouse placentas were also quickly identified and quantified.

Data availability

Data will be made available on request.