Machine Learning to Predict the Adsorption Capacity of Microplastics

Abstract

Nowadays, there is an extensive production and use of plastic materials for different industrial activities. These plastics, either from their primary production sources or through their own degradation processes, can contaminate ecosystems with micro- and nanoplastics. Once in the aquatic environment, these microplastics can be the basis for the adsorption of chemical pollutants, favoring that these chemical pollutants disperse more quickly in the environment and can affect living beings. Due to the lack of information on adsorption, three machine learning models (random forest, support vector machine, and artificial neural network) were developed to predict different microplastic/water partition coefficients (log Kd) using two different approximations (based on the number of input variables). The best-selected machine learning models present, in general, correlation coefficients above 0.92 in the query phase, which indicates that these types of models could be used for the rapid estimation of the absorption of organic contaminants on microplastics.

Photo-oxidation of Micro- and Nanoplastics: Physical, Chemical, and Biological Effects in Environments

“…Generally, the photo-oxidation of MNPs may enhance the adsorption of hydrophobic OPs and reduce the adsorption of hydrophilic OPs due to enhanced O-containing functional groups. , However, in some cases, it is complex and depends on many factors including polymer types of MNPs, oxidation degree, and physicochemical properties of OPs. ,, There is a lack of knowledge to evaluate the effects of the photo-oxidation degree and the physicochemical properties of OPs on their adsorption processes on MNPs. Integrating experimental data with advanced analytical methods, such as molecular dynamic simulations and machine learning, , offers a comprehensive approach for understanding the adsorption mechanism of MNPs after photo-oxidation, and establish the multidimensional relationship between the adsorption capacity of MNPs, the photo-oxidation degree (e.g., the carbonyl index and hydroxyl index, and the physicochemical properties of pollutants (e.g., molecular weight, log K OW , zeta potential, type, and number of functional groups).…”

“…69,73,80 There is a lack of knowledge to evaluate the effects of the photo-oxidation degree and the physicochemical properties of OPs on their adsorption processes on MNPs. Integrating experimental data with advanced analytical methods, such as molecular dynamic simulations 229 and machine learning, 230,231 offers a comprehensive approach for understanding the adsorption mechanism of MNPs after photo-oxidation, and establish the multidimensional relationship between the adsorption capacity of MNPs, the photo-oxidation degree (e.g., the carbonyl index and hydroxyl index, and the physicochemical properties of pollutants (e.g., molecular weight, log K OW , zeta potential, type, and number of functional groups). coexisting natural colloids.…”

Traceability of Microplastic Fragments from Waste Plastic Express Packages Using Near-Infrared Spectroscopy Combined with Chemometrics

“…Microplastic pollution may be one of the most widespread and persistent anthropogenic changes to Earth’s surface [ 22 ]. MPs are emerging pollutants that may play roles as mediators in the environment [ 23 , 24 ]. MPs can be carriers of other toxic pollutants, including both organic and inorganic pollutants [ 25 ], bacteria, and deadly viruses [ 23 ], and can quickly disperse toxic substances into the environment, thus affecting organisms.…”

“…Astray et al used random forests, support vector machines, and artificial neural networks to study the adsorption capacities of different MPs to organic pollutants, and the correlation coefficient of the best selected machine learning model was greater than 0.92, indicating that these types of models could be used to quickly estimate the adsorption of organic pollutants by MPs [ 26 ]. Cid-Samamed et al believed that the addition of mono- and bivalent salt surfactants via polymer modification could assist in trapping microplastics and achieve the aggregation of microplastics in aquatic systems [ 24 ]. At present, the methods used for the analysis and identification of MPs can be divided into the three following categories: (1) physical morphology characterization analysis methods (i.e., microscopic imaging analysis), including scanning electron microscopy (SEM), scanning electron microscopy–energy dispersive spectrometry (SEM-EDS), atomic force microscopy (AFM), and fluorescence microscopic imaging (FMI); (2) chemical component spectral analysis, including near-infrared (NIR) spectrometry, Fourier transform infrared (FTIR) spectrometry, and Raman spectroscopy; and (3) thermal analysis techniques, such as differential scanning calorimetry (DSC), pyrolytic gas chromatography–mass spectrometry (Pyr-GC/MS), and thermogravimetric analysis (TGA).…”

Machine Learning Prediction of Adsorption Behavior of Xenobiotics on Microplastics under Different Environmental Conditions

“…Measuring the adsorption and evaluating the magnitude of these impacts for every combination of compound, polymer composition, and environmental condition are daunting tasks. With the advent of machine learning (ML), the adsorption capacities of microplastics for unstudied chemicals can be predicted with a high degree of accuracy. , In this study, we obtained the adsorption capacity and affinity of 16 polymer compositions for over 40 xenobiotics under environmental conditions. Using these data, we trained two ML models, Random Forest (RF) and Artificial Neural Network (ANN), which were used to predict Linear, Freundlich, and Langmuir isotherms for a set of xenobiotics and polymer compositions.…”