Interfacial Interactions of Uranium and Arsenic with Microplastics: From Field Detection to Controlled Laboratory Tests

Quality control and quality assurance

The use of plastic materials was reduced as much as possible to avoid MP background contamination. All experimental instruments and glassware were sonicated for 30 min in an ultrasonic bath with ultra-pure water (18 MΩ) and covered with aluminum foil between sonication and use. All benchtops were carefully cleaned, and all laboratory procedures were conducted in a fume hood. Field controls were included to monitor any airborne contamination. In laboratory procedures, a control containing only ultra-high purity water during filtering and digestion was included to account for any background MP interference.

Sampling Methodology

For MP analyses, three 1-L samples were collected from six locations along Paguate River and freshwater reservoirs near the Jackpile Mine of Laguna Pueblo, NM. These sites were selected based on their proximity to the mine and the Laguna community. In addition, three locations on the Rio Grande, and three on Tingley Beach, Albuquerque, NM, were selected to compare occurrence of MPs in a rural and an urban community (Supplementary Table S1 in the Supplementary Data). Samples were taken from the first 10 cm of the water surface to avoid sediment interference. A separate set of samples was taken from 20 bodies of water in the area to confirm elevated concentrations of U and As reported in previous studies from our group (Blake et al., 2017; Blake et al., 2015). Note that less samples were taken for MP analyses due to the complexity of the extraction procedures.

Extraction of MPs

To extract MPs onto filters and ensure the quality of visual assessment and polymer identification, the procedure recommended by Koelmans et al. (2019) was followed. Details about the extraction of MP are provided in the Supplementary Data.

Analyses of MPs and metals in field samples

Filters containing extracted particles were examined and imaged using a stereomicroscope (AmScope 7X-180X Trinocular Zoom Stereo Microscope) for initial visual assessment. The filters were imaged using a fluorescence microscope (Cytation 5 Cell Imaging Multi Mode Reader; Agilent Technologies) with Gen5^® software. For quantification, each filter's overall image was run through the MPVAT 2.0 macros using ImageJ (Prata et al., 2020). Each filter was then analyzed using an Attenuated Total Reflectance—Fourier Transform Infrared (ATR-FTIR) spectrometer (micro-FTIR, Thermo Nicolet iN10 MX). For each filter, the number of particles examined with ATR was 10% of the particles identified during fluorescence microscopy quantification, or a minimum of five (whichever was greater). Further details of the fluorescence and ATR-FTIR analyses are provided in the Supplementary Data.

Reagents

Sodium arsenate dibasic heptahydrate, Na₂HAsO₄7H₂O reagent (≥98%), was purchased from Sigma-Aldrich. Uranyl nitrate hexahydrate reagent, UO₂(NO₃)₂6(H₂O) (98–102%), was purchased from IBI Labs. Three types of MPs were used in these experiments: PE, PS, and PMMA (Supplementary Table S2 in the Supplementary Data). We selected the MP types based on their predominant abundance in the environment according to previous research (Di and Wang, 2018; Li et al., 2020).

We considered particles with different densities to have a representative distribution of plastics that potentially occur in the water column (Lenaker et al., 2019). PE (0.96 g/cc, 10–63 μm) and PMMA MPs (1.2 g/cc, 1–45 μm) were purchased from Cospheric. PS is used for packaging, disposable cups, and many other uses (Andrady and Neal, 2009). PS beads (200–300 μm) were purchased from Polysciences. Glass microfiber filters (Advantec GC-50 borosilicate diameter, 47 mm; Pore Size: 0.5 μm) were purchased from Cole-Parmer. The main characteristics of these MPs are provided in Supplementary Table S2 in the Supplementary Data.

Sorption experiments

These series of experiments were performed to assess the sorption of different concentrations of As and U onto PE, PS, and PMMA commercial MPs at pH 3 and pH 7. pH adjustments were made with 0.1 M HNO₃ or NaOH. A mass of 0.1 g of PE, PS, and PMMA commercial MPs was added into a borosilicate glass beaker containing a volume of 100 mL of deionized water, resulting in a concentration of 1 g MP/L, the concentration of MP to be within range (0.04–10 g/L) with other studies that evaluated the sorption of metals onto MP (Brennecke et al., 2016; Godoy et al., 2019; Holmes et al., 2014; Zou et al., 2020). Isotherms were carried out for 48 h by separately exposing the MPs to 0.05, 0.1, and 0.2 mM of U or As.

All experimental conditions were run in triplicates in a VWR Advanced Orbital Shaker Model 15000 at 150 rpm at room temperature (25°C) for 48 h. Controls without MPs and only including either U or As were included in the experiments.

The equilibrium time was selected based on kinetic experiments (Supplementary Fig. S3 in the Supplementary Data) and in our previous study where we observed rapid precipitation (<1 h) and equilibrium of soluble U after 48 h using a similar concentration range (0.005–1 mM of U) (Gonzalez-Estrella et al., 2020). On the other hand, the concentration for U and As was based on (1) being within the range of concentrations of U and As that we have found in Laguna Pueblo, NM, since 2013 (Blake et al., 2017; Blake et al., 2015) and other studies that have evaluated concentration of with U and As in U mine tailings (Donahue and Hendry, 2003; Robertson et al., 2019); and (2) ensuring we achieve saturation on the MP in case the sorption followed a typical adsorption behavior.

After 48 h, the solutions were vacuum filtered through a 0.5 μm glass microfiber filter and glass frit filter unit. The filtered water samples were transferred into centrifuge tubes and the filters were placed in a petri dish and stored at 4°C. Metal adsorption was determined by quantifying the soluble concentration of U and As with Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES) and ICP-Mass Spectroscopy (ICP-MS). Each filter paper was slowly rinsed with ultra-pure water (18 MΩ) to avoid additional compounds precipitating as the remaining water evaporated from the filter surface and preserved for spectroscopy analyses.

Interaction of MPs with filtered solutions of U

An additional set of experiments was conducted to isolate the interactions between soluble U and MPs at pH 7. In these experiments, 0.02 and 0.06 mM U were used and all U solutions were filtered before exposure to the MPs to eliminate U precipitates from the solution. PP centrifuge tubes were used instead of glass to ensure that the glass was not providing a surface for heterogenous precipitation. The isotherms were run with the same parameters and conditions explained above. Controls with no MP and only including either U or As were also included in the experiment.

Characterization of MPs

PE, PMMA, and PS MPs exposed to U and As were analyzed with various spectroscopy techniques to identify any precipitation reaction on the surface. Transmission Electron Microscopy (TEM), Scanning Electron Microscopy (SEM), and Energy Dispersive Spectroscopy (EDS) were used to examine the MP morphology and quantify heavy metals binding onto the surface. Zeta potential ζ was used to measure the surface charge of MPs at pH 3 and pH 7. Details of sample preparation and SEM and TEM analyses parameters are provided in the Supplementary Data.

Quantification and characterization of MPs

All locations contained a similar range of particle concentrations (Table 1). In Laguna Pueblo, NM, a range from 7 to 149 particles/L was detected in the water samples taken from six different locations (Supplementary Fig. S1A–F and Supplementary Table S1 in the Supplementary Data). Water samples taken from three different locations of Tingley Beach, Albuquerque, NM, showed a range of 18–127 particles/L. Finally, a range from 12 to 101 particles/L was detected in water samples taken from three different locations of Rio Grande, Albuquerque, NM.

Following quantification, ATR-FTIR analyses were performed to analyze the chemical composition of the particles. A total of 25, 23, and 25 particles were analyzed from the Laguna Pueblo, Tingley Beach, and Rio Grande samples, respectively. Figure 1 shows representative particles analyzed with ATR-FTIR, Table 1 shows the spectra match of each particle analyzed, and all data from these analyses are available in the Supplementary Data (Supplementary Tables S4–S6 in the Supplementary Data). The criteria used in this study require spectra matches >60% for a particle to be considered a confirmed MP. The 25 particles analyzed from the Laguna Pueblo samples indicated spectra matching from 21% to 72%, relative to pure polymers. Two confirmed MPs were identified at the site- one rayon (72%) and one cellophane (67%) of the 25 particles examined.

OPEN IN VIEWER

FIG. 1.

Representative images of plastic-like particles found in (A) Laguna Pueblo, New Mexico (Site L5), (B) Tingley Beach, Albuquerque, New Mexico (Site T1 and T2), and (C) the Rio Grande, Albuquerque, New Mexico (Site R1 and R3).

Analysis of 23 particles randomly selected from the Tingley Beach samples indicated their spectra matched pure polymers from 25% to 73% (Table 1). Five confirmed MPs were found in these samples, including two polyester (both 73%), two rayon (67% and 63%), and one polytetrafluoroethylene (PTFE) #4 (65%), meaning that 32% of the examined particles match with a polymer spectrum. Finally, the analyses of 25 particles selected from the Rio Grande samples indicated their spectra matched from 20% to 73%, relative to pure polymers. One polyamide (PA) 6+PA 6,6 (73%) particle and one cellophane (72%) particle were confirmed as MPs from the 25 examined particles, indicating 8% of analyzed particles matched with a polymer spectrum. This analysis indicates that Tingley beach, a stagnant freshwater body located in an urban center, contained the highest number of particles confirmed as MPs.

However, a distinction must be made that the ATR-FTIR analysis provides insight specifically into MPs >20 μm. While this is limiting, it is important to recall that larger MPs can continue to break down in the environment, potentially releasing micro- and nanoplastics below this 20 μm threshold; thus, an analysis of larger MPs still provides valuable insight. Many of the particles on each filter, especially fibers, had at least one dimension below the 20 μm detection limit of the μ-FTIR, but above the 0.5 μm detection limit of the fluorescence microscope—meaning they could be quantified, but were not eligible for ATR-FTIR analysis. An example of fibrous particles, potentially MPs, which were below the detection limit, is shown in Supplementary Fig. S1 in the Supplementary Data.

Previous studies have also found similar polymer types, including polyester (PES), PA, rayon, or cellophane (CP), and a similar particle content in freshwater. For example, a range from 3.4 to 25.8 particles/L was found in Lake Taihu in China. The most common polymer types identified were CP, PET, PES, PP, and PA (Su et al., 2016).

Fibrous and fragmented MPs were found along the middle and lower reaches of the Yangtze River Basin with concentrations varying from 0.24 to 1.8 particles/L and 0.5 to 3.1 particles/L, respectively (Li et al., 2019; Su et al., 2018). Mainly PP, PE, and polycarbonate (PC) were found in the middle of the Yangtze River Basin (Li et al., 2019), while the most dominant polymers were PES (33%), PP (19%), and PE (9%) in the lower basin area (Su et al., 2018). Similarly, a range from 0.9 to 2.4 particles/L was identified in Suzhou River, Huangpu River, and the urban creeks of Shanghai where the dominant polymer was PES (Luo et al., 2019).

MP functional chemistry

The degradation of MPs in the environment due to ultraviolet (UV) and physical weathering has become well documented in recent years (Liu et al., 2020). In this work, the functional chemistry of weathered environmental MPs is compared with pristine and pure polymer spectral reference libraries. The ATR-FTIR spectra of selected MPs found in field samples compared to the reference spectra are shown in Fig. 2. All examined spectra, including library references, are found in from Supplementary Tables S4 to S6 in the Supplementary Data. Changes in the spectra of MPs found in the samples compared to the reference spectrum may be explained due to weathering patterns, and reactions with other elements in the environment can be observed in the rayon particles (T1-1 and T2-3) in O-H region (3,700–3,000 cm⁻¹) and C-H peaks (2,900 cm⁻¹).

OPEN IN VIEWER

FIG. 2.

Infrared spectra of confirmed MPs. Dotted lines indicate polymer reference spectra. Images of MP particles L5-1, T1-1, T2-3, R6-2, L2-1, T1-2, T1-3, T3-2, T2-1, T2-2, T3-1, and R6-1 are available in the Supplementary Data. MPs, microplastics.

Cellophane particles (T1–2 and T1-3) showed changes in the C-H bending signal (1,450–1,500 cm⁻¹). Other examples include the polyester MPs (T2-1 and T22) with modifications in the O-H (3,700–3,000 cm⁻¹) and C-H peaks (2,900 cm⁻¹) compared to the reference. Discrepancy in functional chemistry between the pristine reference spectra and weathered environmental MPs likely leads to the misidentification or underidentification of MPs by current spectral identification tools. The current challenges of spectral identification highlight the need to generate more environmentally relevant spectral libraries that contain mechanically weathered and UV aged polymers.

Uranium precipitation and reactivity with the MP surface at pH 7

The reactivity of U and MPs depended on the pH. In assays supplied with PMMA and PE carried out at pH 7, the soluble concentration of U decreased significantly (p < 0.05) from 0.05 to ∼0.003 mM after 48 h of reaction (Fig. 3A–C). Surface SEM EDS analyses revealed that U precipitated on the surface of PMMA (Fig. 4A). Further TEM analyses of both the surface of the MPs and precipitates formed in the control suggest that the solid phase formed resemble Na-compreignacite on the surface of PMMA (Fig. 4B). Chemical equilibrium analyses were conducted and indicated that the solution was supersaturated with respect to schoepite and Na-compreignacite, both uranyl oxide hydrates.

OPEN IN VIEWER

FIG. 3.

Soluble U concentration in batch experiments containing (A) PMMA, (B) PE, and (C) PS and soluble As concentration in batch experiments containing (D) PMMA, (E) PE, and (F) control without MPs at pH 3 and pH 7 at 0 and a 48-h exposure. Assays were supplied with 0.05 mM U. Error bars indicate standard deviation obtained from duplicates. Asterisks represent the significant difference of soluble U concentration. PE, polyethylene; PMMA, polymethyl(meta)acrylate; PS, polystyrene.

OPEN IN VIEWER

FIG. 4.

Spectroscopy analyses of PMMA MPs exposed to 0.06 mM of U for 48 h experiments were performed at pH 7, (A–C) TEM images of precipitates onto the commercial PMMA MP surface and (D, E) SEM/EDS analyses confirming U accumulated on the MP surface. EDS, Energy Dispersive Spectroscopy; SEM, Scanning Electron Microscopy; TEM, Transmission Electron Microscopy.

The Na-bearing U solids were the primary phases in our study because NaOH was used to adjust the pH. The decrease in U observed in the control without MPs at pH 7 was likely caused by homogenous precipitation of uranyl solids. Heterogenous precipitation took place with the presence of MPs, which provided surface sites for U solids to deposit and precipitate. These results suggest that U homogenous and heterogenous precipitation processes are relevant mechanisms that may be observed in aquatic environments supersaturated with U. Past studies have confirmed Na-compreignacite and schoepite precipitates formed at pH 7 in a similar concentration used in our study (Gorman-Lewis, Burns, et al., 2008; Gorman-Lewis, Fein, et al., 2008; Kanematsu et al., 2014).

To further investigate the surface interaction mechanism of U and MPs, PMMA MPs were exposed to three different U concentrations (0.05, 0.1, and 0.2 mM) at pH 7 for 48 h. PMMA MPs were selected for these experiments because they have the smallest particle size (1–45 μm), largest surface area (0.86 ± 0.87 m²/g), most negatively charged surface (−42.83 ± 5.17 mV) compared to PE and PS at pH 7 (Supplementary Table S2 and Supplementary Fig. S2 in the Supplementary Data), and the concentration of U decreased the most in the sorption experiments amended with PMMA MPs. The soluble U concentration significantly (p < 0.05) decreased in assays supplied with and control without PMMA MPs (Fig. 5). These findings show that homogenous and heterogenous precipitation of U onto the surface of MPs are key mechanism for U reactivity in the system studied.

OPEN IN VIEWER

FIG. 5.

Soluble U concentration in batch experiments containing (A) PMMA and (B) control (no MPs) at pH 7 at 0 and a 48-h exposure. Error bars indicate standard deviation obtained from triplicates.

Interaction of MPs with prefiltered U solutions

Additional experiments were run with U concentrations of 0.02 and 0.06 mM, which were prefiltered to ensure that U precipitates were the conduct in which U interacted with MP and not the soluble fraction. The filtered U solution exposed to the three MPs and the control at pH 7 slightly decreased (Supplementary Fig. S4 in the Supplementary Data). The use of glass vials was eliminated in this section as it can influence precipitation reactions; PP tubes was used instead. These findings imply that homogenous and heterogenous precipitation are still occurring in the system even with the substitution from glass to plastic and with the extra filtration step to remove U precipitate before MP exposure.

Although U precipitated homogenously in the control without MPs, the SEM analyses confirmed U mineral precipitated heterogeneously on the MP surface. The EDS analyses also showed U and Na compositions and did not show any silica (Si) (Fig. 4). A paired-samples t-test showed the decrease of U concentration is significantly different, t(2) = 64.27, p = 0.0002, (p < 0.05). Our results demonstrate that the precipitation process drives the interaction between the MPs and uranium, and not the soluble ions in the system. Other studies have reported that metals interacted with MPs. Most interactions reported in these studies are attributed to adsorption reactions between cations and MPs (Godoy et al., 2019; Munier and Bendell, 2018; Rochman et al., 2014; Zou et al., 2020).

Lack of reactivity of As (pH 3 and 7) and U (pH 3) with MPs

All experiments supplied with As and PE, PS, and PMMA MPs at pH 3 and pH 7 remained close to the initial concentration (0.05 mM) after the 48 h of exposure (Fig. 3D–F). Similar results were found for the assays supplied with U at pH 3. The controls showed no change in the concentration as well.

Although sorption of As (III) onto PTFE and PS MPs at pH ranging from 3 to 7 has been reported (Dong et al., 2020; Dong et al., 2019), we found no sorption of As (V) onto MPs at pH 3 and 7. Lack of sorption of As (V) may be explained by difference of charges; that is, As (V) is predominantly negative at this pH range (H₂AsO₄^- and HAsO₄^2-), whereas As (III) species are uncharged (Benjamin, 2014). Lack of sorption of U at pH 3, predominantly UO₂²⁺ at acidic pH, likely results from the lack of electrostatic attraction as well. Previous studies have stated that hydrophobic and electrostatic interactions are two predominant mechanisms for the sorption of contaminants on MPs (Tourinho et al., 2019).