To effectively investigate MPs in soil, a thorough methodology must be employed. This ranges from soil sample collection to particles isolation in the soil matrix and other adhering substances whilst preventing any damage or artificial fragmentation to the MP sample. However, contrary to the aquatic systems, only a few studies have been conducted on this field for soil and thus research is still at an early stage than it ought to be, given the importance and complexity of MPs in soil. Most of the methodologies for detection of MPs in soil are derived from procedures on MPs in aquatic environments. This causes lack of standardization and harmonization which could lead to discrepancies and create a bottleneck for future studies. Therefore, development of an accurate methodology with identification of the type of plastic and pollution source is vital for the management and environmental protection frameworks.
Sample collection
Soil is easily influenced by constant human intervention, history and usage of the land and accumulation zones where MPs could be deposited (from surface runoff or windborne particulate matter). Therefore, it is important to identify the accurate sampling area and depth of soil that is to be collected. (Rillig et al.
2017). There are two primary methods of soil sample collection, composite and single site sampling depending on the degree of scattered debris of MPs in soil (Moller et al.
2020). Composite sampling consists of collecting similar size samples from several discrete locations within an area which is then combined and homogenized to a single sample (Scheurer and Bigalke
2018). This method is recommended since the concentration of MPs in soil is never uniform due to various interferences (Moller et al.
2020). Single site sampling is generally used in areas where limited human activities are present (Yang et al.
2021). To determine the number of sampling points, Moller et al. (
2020) have suggested to employ statistical power analysis. However, the exact ideal replicate and volume of sample are not yet conclusive as different area of sampling units were used in studies. This has significantly restricted the sample processing and created a bottleneck for subsequent analytical methods of MPs in soil.
It is imperative to define the sampling depth since the deposition of MPs is highly dependent on activities performed on the soil. Yang et al., (
2021) indicated that the surface of soil should be selected if the study is conducted on undisturbed soil (0–30 cm soil sample). It is also suggested that if concentration of MPs at varying depths of the soil is to be found, then stratified sampling should be utilized. However, the downward transportation of MPs in undisturbed soil is yet to be investigated and thus some of the collected samples might be unreliable (Rillig et al.
2017). Moller et al. (
2020) also recommended collecting a large volume of samples than the required amount for quantification, as it may be needed for subsequent purposes such as sample backup, determining moisture content and sample recovery analysis.
As a measure of quality control, guidelines introduced by the US Environmental Protection Agency have suggested taking control samples that are of a similar soil type as the main MP sample. These should be collected from a nearby vicinity unaffected by contaminants of concern (EPA
2020). This could assist in monitoring potential contamination that originated during sampling, quantifying MP’s background levels and even provide a more comprehensive understanding on soil matrices which, otherwise, could have been unclear (Thomas et al.
2020).
Drying, sieving and purification
It has been observed that the collected samples have undergone natural air drying to reduce soil humidity for better analysis in the subsequent stages although few reviewed publications have oven dried the samples to accelerate the drying process (Yang et al.
2021). This is usually conducted at a temperature range of 40–70 °C, well below the thermal deformation temperature of most plastics. Yet, the prevailing drying methods and conditions analysis are contrasting. For instance, Berg et al. (
2020) have dried the soil at 40 °C for 72 h whereas Liu et al., (
2020) have opted at 70 °C for 24 h. However, the drying temperature of MPs samples comprising of PE and PA is kept below 50 °C as high temperatures would cause impairment (Hurley et al.
2018). It must be noted that temperatures above 40 °C may affect the polymers’ physical and structural properties by glass transition, melting, or degradation (Hurley et al.
2018). Therefore, freeze drying was developed as another alternative drying option (Thomas et al.
2020). Freeze drying could effectively break soil aggregates and cell walls thereby aiding further sample preparation. Nevertheless, polymer brittleness would rise if the operating temperatures were below its glass transition temperature. Also, frost wedging may fragment the sample and cellular organic matter may be released. Furthermore, freeze drying is a slower process than air or oven drying process and often limited by the size of the freeze dryer (Thomas et al.
2020).
After drying stage, the soil sample is passed through a stainless-steel sieve (1–2 mm, 5 mm sieve size) to separate microplastics using various screening methods based on whether the soil is agglomerated or contains grass and other residues (Wang et al.
2019). A sieving cascade may alleviate the amount of required work, but the technique is yet to indicate how excessive sieving increases particle fragmentation, especially when the sample is an aged or freeze-dried fractions (Thomas et al.
2020).
The soil sample may consist of a large quantity of soil organic matter (SOM) which is a complex matrix at different levels of decomposition. In several conducted studies, organic matter was reported to hinder microplastic analysis. Therefore, the removal or reduction of such components remains a necessary issue during sample preparation (Thomas et al.
2020). Thus, the MP sample would need to undergo the purification process which is a vital stage for the subsequent analysis. The most common reagents used for this stage are HNO
3, NaOH, H
2SO
4 and H
2O
2. Nuelle et al., (
2014) have claimed that although H
2O
2 is widely used, the efficiency remains inconclusive as it is a time-consuming process and H
2O
2 would bleach the organic matter rather than remove it. Hurley et al. (
2018) suggested that an alternate solution would be using Fenton’s reagent which, in the past, was used to extract microplastics from wastewater samples. This method is more effective than the latter one as it would remove any organic material from complex substrates at a shorter duration. Other methods including acidic reagents such as HNO
3 and HCL have not been deemed suitable as they may degrade and melt the microplastic (Scheurer and Bigalke
2018). Alkaline reagents such as KOH and NaOH have been considered effective, but they displayed relatively low removal efficiency as humus and alkali insoluble compounds in the soil sample (or other complex samples like sewage sludge) were still present after the reagent application (Blasing and Amelung
2018). Thus, a proper testing on such complex matrices is required to establish the type of removal for organic matter.
The final stage is extracting the MPs from sediments or inorganic compounds (supernatant) that were not broken down throughout the previous stage. Density separation is the most widely used technique using extraction media such as sodium chloride, calcium chloride and zinc chloride, sodium iodide, distilled water and sodium heteropolytungstate widely used (Nakajima et al.
2019). The principle underlying this method is the difference of specific gravity of soil and plastics. Most MPs are less dense than the salt solutions and hence they float to the top, leaving the denser inorganic sediment to settle at the bottom. Therefore, the soil-to-solution ratios implemented during experiments depends on the sample size and technical setup and is decisive for MP recovery (Thomas et al.
2020). However, the ratio of soil-to-density solution varied immensely from 1:2 to 1:25 (Chen et al.
2020a,
b,
c,
d). Hence, a more harmonized method needs to be addressed.
Amongst the extraction media, sodium chloride (NaCl, 1.2 g cm
−3) and distilled water (1.0 g cm
−3), albeit its ease of access and cost, are only able to separate low density polymers such as PE, PS and PP (Li et al.
2020). For NaCl, the Na
+ may further facilitate in the dispersion of soil aggregates resulting in higher extraction efficiency (Scheurer and Bigalke
2018). Alternate extracting media is used for synthetic polymers, such as PVC (1.1–1.6 g cm
−3) and PET (polyethylene terephthalate) (1.3–1.6 g cm
−3) which have higher densities than NaCl (Moller et al.
2020). However, the type of extraction media is solely dependent on the local demand of plastics as this will determine the composition of the soil sample. Scheurer and Bigalke (
2018) contended that since PVC and PET make a relatively small contribution to the larger portion of other microplastics in the sample, a sodium chloride (NaCl) solution could be utilized. Nevertheless, this argument is solely based on the local demand of plastics which determines the composition of the soil sample and thereby the extracting media. Calcium chloride has been proposed as an optimum choice for the separation of denser particles owing to its environmental friendliness and low cost (Scheurer and Bigalke
2018). However, the presence of organic floccules was observed after the separation process as Ca
+ ion can bridge the organic molecules’ negative charge (Scheurer and Bigalke
2018). These floccules get attached to the filter membrane which would then impede the MP’s identification process (Yang et al.
2021).
Highly dense solutions including sodium polytungstate, zinc chloride, zinc bromide, and sodium iodide are effective in separating small MP fibres despite the high cost and hazardous impacts (Nakajima et al.
2019). This situation was observed in the Munich Plastic Sediment Separator (MPSS) designed by Imhof et al. (
2012) with a high recovery rate of 95–100% using zinc chloride solution without further extraction or incurring contamination or losses (Imhof et al.
2012). However, this recovery rate does not apply to the old MPs and given the properties of zinc chloride, it may corrode surfaces and result in a low separation of the less dense substances. Moreover, zinc chloride may react with sediments on the MP surface to create a foam that may disrupt the extraction process (Moller et al.
2020). Recent studies have indicated that sodium bromide is very effective in the extraction process as it is economically viable as a reagent and non-corrosive and non- hazardous, although it may have its own limitations due to the local plastic demand (Yang et al.
2021).
Multi-stage separation was an alternative strategy developed to encounter the limitations posed by separation solutions and to increase separation efficiency. In some studies, the same solution has been reused many times whilst others have opted applying lower and high-density solutions consecutively (Huang et al.
2020). Frere et al. (
2017) used sodium tungstate solution (1.56 g cm
−3) to separate dense MP sample (0.8–1.4 g cm
−3) from denser particles like sand and coarse sediment grains (2.65 g cm
−3). A spiking experiment was utilised to verify the full recovery of the dense MPs. The study obtained a complete recovery of each polymer types (PET, PA, PVC) without any hindrance during the identification of the polymers by Raman Spectroscopy (Frere et al.
2017). Han et al. (
2019) mixed saturated NaCl and NaI solution with a flotation density of 1.50 g cm
−3 that obtained a 90% or more recovery of most of the tested MP particles wherein the authors believed that the solution can be reused 5 times after filtration.
Novel procedures of electrostatic separation also exists which permits a significant recovery rate of 90–100% for MPs ranging from 63 μm to 5 mm (Felsing et al.
2018). But this result was obtained by conducting an experiment for 3–4 h on a soil sample lacking moisture which could raise further doubts on the aggregate formation (Felsing et al.
2018). To prevent the soil aggregation, some techniques including centrifugation, aeration, ultrasonic treatment, and continuous flow can be implemented. Grbic et al. (
2019) have developed magnetic extraction to recover MPs by creating surface modified hydrophobic iron nanoparticles that would bind to MPs samples, thereby allowing magnetic recovery. However, this experiment was conducted using freshwater MPs samples and has not yet been tested on soil matrices. The observed MPs have been fragmented and damaged caused by the magnet’s removal (Grbic et al.
2019). However, a more promising technique to extract MP from soil was built incorporating this concept by Ramage et al. (
2022) known as High Gradient Magnetic Separation (HGMS) which was able to circumvent issues associated with the former method. The HGMS proved to be more cost effective, faster and high recoveries for both high- and low-density MPs and fibres from various soil matrices. Yet one prevailing drawback of HGMS is that prior knowledge of the soil composition needs to be understood (Ramage et al.
2022). Scheurer and Bigalke, (
2018) have applied the pressurized fluid extraction (PFE) method for its efficiency wherein the analysis is performed automatically without any human influences. However, there would be some limitations due to its small sample capacity and a high level of sensitivity that leads to an inaccurate quantification. It also further failed to capture information regarding particle size required for the toxicity and mobility studies of MPs (Scheurer and Bigalke
2018).
Techniques for microplastic quantification
Identification and quantification are performed after microplastics have undergone the extraction stage. The identification of MPs is determined initially through visual inspection either by naked eye or using a microscope (Wang and Wang
2018). The morphological characteristics such as colour, shape and surface texture are the primary factors to distinguish if the item is indeed an MPs. More fluorescent staining incorporating dyes such as Evans blue, Calcofluor white and Nile red can be utilised to distinguish MPs from the surrounding matrix (Helmberger et al.
2020). However, identification via visual inspection will be subjective as it depends on various factors and could impair accuracy because of misidentification due to degradation and false positives (Wang and Wang
2018). Blasing and Amelung (
2018) noted a rate of 20–70% of error in heterogenous soil samples whilst conducting visual identification.
These discrepancies can be rectified and verified with spectroscopic and thermoanalytical techniques such as Raman spectroscopies, Fourier Transform infrared spectroscopy (FTIR) and Pyrolysis gas chromatography–mass spectrometry particularly for smaller sample sizes. The FTIR spectroscopy measures the amount of IR radiation absorbed by MPs sample thereby providing an analysis of its molecular composition (Chen et al.
2020a,
b,
c,
d). Absorption peaks produced from the IR spectrum would correspond to the vibration frequencies of the samples’ atomic bonds and consequently present a rapid and reliable representation of the plastic structure (Chen et al.
2020a,
b,
c,
d). Amongst the research studies, Chen et al. (
2020b) demonstrated that smaller MPs were identified by the μ-FTIR with minimum particle size of 10 μm and larger ones by ATR-FTIR (Attenuated total reflectance FTIR). The ATR-FTIR technique, however, is known to cause damage to certain MPs due to the large pressure applied by the equipment’s probe. To counteract this problem, FPA-FTIR (focal plane array FTIR) can be implemented, enabling scanning of more tentative MPs types (Loder et al.
2015). However, Loder et al. (
2015) noted that μ-FTIR requires a longer duration for sample measurement, (over 20 h) causing a risk of loss and contamination of coloured or small sample plastics. Raman spectroscopy is an alternate yet promising technology identifying the MPs sample and is considered more advantageous than the FTIR. It encompasses a better spatial resolution (1 μm) and requires low amounts of the MPs sample (Araujo et al.
2018). Raman spectroscopy is also known to simultaneously perform analysis on wet samples while distinguishing pigments or fillers (Zhao et al.
2018). However, studies have shown that this method is more susceptible to fluorescence interference and sample heating due to applying laser as the light source which may lead to the degradation of MP fragments (Zhao et al.
2018). Also, even though Raman spectroscopy can provide reliable information on MP identification, the process is highly time consuming (He et al.
2018). Furthermore, Paul et al. (
2019) claimed that this technique tends to perform less efficient when non-coastal soils were considered and thus restricts its potential usage in the context of this research. A potential obstacle further lies with the sensitivity of FTIR and Raman spectroscopy to water, atmospheric carbon dioxide, SOM, and traces of clay present in the sample which would then require more extensive removal during the sample preparation stage (Xu et al.
2019a,
b). Thomas et al. (
2020) pointed out that the time needed to obtain measurements via these applications may impede screening and monitoring studies that are to be conducted.
According to a study conducted by Wang et al. (
2019), mass spectrometry techniques consisting of pyrolysis–gas chromatography–mass spectrometry (Pyr–GC–MS), thermogravimetric analysis-mass spectrometry (TGA–MS), and thermal extraction desorption-gas chromatography–mass spectrometry (TED-GC–MS) can also be utilized for MP identification. Mass Spectrometry works by separating the constituents of the ionized MP sample depending on their m/z value (Wu et al.
2023). These techniques encompass several advantages, from high sensitivity and easy operation to reliable and quick analysis that can be utilised for detection of molecules along with the identification of a material’s composition (Wu et al.
2023). These technologies were also shown to analyse samples with no further pre-treatment. However, Wang et al. (
2019) argued that the morphology of the plastic fraction and its colour could be damaged or hampered during the evaluation of the sample. Thomas et al. (
2020) indicated that there will be the issue of preparing homogenous aliquots (less than 100 mg) for these techniques if the MP sample was not extracted through using organic solvents which will then require cryomilling. Paul et al. (
2019) have tested another conventional process analytic—Near infrared (NIR) spectroscopy which was successful in achieving a high throughput analysis. The electromagnetic spectrum part of NIR ranges from 667–2500 nm (in the middle of IR and UV) and has a greater capacity to penetrate deeper, and making it very viable in larger samples.
Depolymerization method with the aid of alkali-assisted heating is a recent identification technique suggested by Wang et al. (
2017) where types of MPs such as polycarbonate (PC) and PET are identified and quantified by establishing their constituent compounds. However, samples consisting of a wide range of key components could cause measurement issues and thus the method would need further investigation.
Complications with sample analysis
Insufficient research conducted in the case of analytical methods will cause limited development of such technologies in future studies. The situation is even further exacerbated when observing that most of the available methodologies derived from MP analysis are in the aquatic environment. This is a critical factor as very limited quality control and external proficiency tests have been conducted in this area (Moller et al.
2020). For small sample size, the results may not be reliable and if larger samples are selected the duration of the analytical process will be too long, and expensive. Therefore, the size and quantity of sample is significantly essential for the accurate results. As strategies are varied in sample measurements, some studies indicated the volume as a basis for sample measurement and other used mass (Yang et al.
2021). The geographical location is also another essential factor since collecting and regulating the quantity of samples is related to the local plastic demand and has a direct proportional relationship with the type of MPs found in soil. Moreover, the definition of MP concentration needs to be illustrated as some studies have applied the weight of MPs per kg soil instead of the number of MPs per kg soil (Zhou et al.
2020). Zhou et al. (
2020) also noted the necessity of obtaining the analytical procedure of collected samples when comparisons of MPs concentration from different regions need to be reported.
There are possibilities of losses of MPs from the soil sample especially during the sieving process. Some operations have a sieving size of 2 mm (taken as standard) but others would apply a sieving size of 1 mm which can cause that MPs greater than 1 mm to be discarded (Blasing and Amelung
2018). This would then require more consideration during identification and quantification stage as MP abundance in the sample is based on the sieve size. Furthermore, Yang et al. (
2021) have pointed out that the MP size required for the FTIR spectroscopy depends on how easily it can be handled by tweezers for transferal compared to the Raman spectroscopy technique which even identifies smaller MPs sizes. Hence, it highlights the importance of all stages in the analytical methodology for sample identification. Additionally assessing pristine and aged MPs is an important factor as it plays a vital role in the MPs identification process. This is due to the fact that weathering of such aged particles may cause great losses if an improper identification method is selected (Chen et al.
2020a,
b,
c,
d).
Key differences in analytical methods for aquatic systems
While the analytical methods for microplastics in various environmental media share common procedures, there is notable variability across different matrices. The main difference lies mostly in sampling. As discussed previously, in aquatic systems, sampling techniques involve considering a water column with depths that vary based on the research objectives. For instance, surface water sampling, a prevalent approach, aims to unravel the occurrence of microplastics (MPs) at the surface (Yu et al.
2016). Neuston nets are commonly employed for surface collection at depths ranging from 0 to 0.5 m (Anderson et al.
2017), while bongo nets are utilized for water columns of medium depth, and benthic nets are deployed for seabed sampling. Research in this field has reported a significant depth range, extending from 50 to 60 μm (surface microlayer) to 212 m (Stock et al.
2019).
Similar to soil sampling, the presence of MPs in aquatic environments depends on factors such as the collected water column, net opening area, and mesh size. Studies indicate a wide range of mesh sizes from 0.053 to 3 mm, with the net aperture playing a crucial role in determining the efficiency of microplastic capture. A flow meter is mounted at the entrance of the net to determine the filtered water volume which can thereby be used to find the MP concentration (Wagner and Lambert
2017). Alternatives such as 3D hydrodynamic—numerical modelling and acoustic doppler current profilers are other means to find the water volumes, the latter of which can be implemented for sampling different depths at one location (Stock et al.
2019).
As methodologies for assessing MPs in aquatic systems continue to evolve, the diverse array of sampling techniques reflects the complexity of studying MPs across different environmental compartments. In marine research samples obtained from below the water surface have no mention of specific sampling depths and vary according to the goal of the research conducted. For sediments near water bodies, which is regarded as a long-term sink for MPs, sampling will differ with location which can be categorized as the tideline, intertidal and supralittoral environments (Mai et al.
2018). In beaches, most studies have conducted sampling within a depth of 5 cm with equipment such as trowels, shovels, spoons used for sample extraction. Wang et al. (2018) notes that although some studies have recommended using tweezers for extracting, it may overlook smaller MPs thus underestimating the abundance of MPs in that location (Hidalgo-Ruz et al.
2012). For sublittoral zones Wang et al. (2018) suggests the usage of grabbers or box corers for superficial sediments. However, in such cases, it is important to first identify and comprehend the various sediment layers since the usage of these tools causes disturbances.