Predicting Aqueous Adsorption of Organic Compounds onto Biochars, Carbon Nanotubes, Granular Activated Carbons, and Resins with Machine Learning

Prediction of g-C3N4-based photocatalysts in tetracycline degradation based on machine learning

Investigating the effects of g-C3N4-based photocatalysts on experimental parameters during tetracycline (TC) degradation can be helpful in discovering the optimal parameter combinations to improve the degradation efficiencies in general. Machine learning methods can avoid the problems of high cost, time-consuming and possible instrumental errors in experimental methods, which have been proven to be an effective alternative for evaluating the entire experimental process. Eight typical machine learning models were explored for their effectiveness in predicting the TC degradation efficiencies of g-C3N4 based photocatalysts. XGBoost (XGB) was the most reliable model with R2, RMSE, and MAE values of 0.985, 4.167, and 2.900, respectively. In addition, XGB's feature importance and SHAP method were used to rank the importance of features to provide interpretability to the results. This study provided a new idea for developing g-C3N4-based photocatalysts for TC degradation and intelligent algorithms for predicting the photocatalytic activity of g-C3N4-based photocatalysts.

Predicting Cd(II) adsorption capacity of biochar materials using typical machine learning models for effective remediation of aquatic environments

The screening and design of "green" biochar materials with high adsorption capacity play a pivotal role in promoting the sustainable treatment of Cd(II)-containing wastewater. In this study, six typical machine learning (ML) models, namely Linear Regression, Random Forest, Gradient Boosting Decision Tree, CatBoost, K-Nearest Neighbors, and Backpropagation Neural Network, were employed to accurately predict the adsorption capacity of Cd(II) onto biochars. A large dataset with 1051 data points was generated using 21 input variables obtained from batch adsorption experiments, including preparation conditions for biochar (2 features), physical properties of biochar (4 features), chemical composition of biochar (9 features), and adsorption experiment conditions (6 features). The rigorous evaluation and comparison of the ML models revealed that the CatBoost model exhibited the highest test R2 value (0.971) and the lowest RMSE (20.54 mg/g), significantly outperforming all other models. The feature importance analysis using Shapley Additive Explanations (SHAP) indicated that biochar chemical compositions had the greatest impact on model predictions of adsorption capacity (42.2 %), followed by adsorption conditions (37.57 %), biochar physical characteristics (12.38 %), and preparation conditions (7.85 %). The optimal experimental conditions optimized by partial dependence plots (PDP) are as follows: as high Cd(II) concentration as possible, C(%) of 33 %, N(%) of 0.3 %, adsorption time of 600 min, pyrolysis time of 50 min, biochar dosage of less than 2 g/L, O(%) of 42 %, biochar pH value of 11.2, and DBE of 1.15. This study unveils novel insights into the adsorption of Cd(II) and provides a comprehensive reference for the sustainable engineering of biochars in Cd(II) wastewater treatment.

Prediction of Cr(VI) and As(V) adsorption on goethite using hybrid surface complexation-machine learning model

This study aimed to develop surface complexation modeling-machine learning (SCM-ML) hybrid model for chromate and arsenate adsorption on goethite. The feasibility of two SCM-ML hybrid modeling approaches was investigated. Firstly, we attempted to utilize ML algorithms and establish the parameter model, to link factors influencing the adsorption amount of oxyanions with optimized surface complexation constants. However, the results revealed the optimized chromate or arsenate surface complexation constants might fall into local extrema, making it unable to establish a reasonable mapping relationship between adsorption conditions and surface complexation constants by ML algorithms. In contrast, species-informed models were successfully obtained, by incorporating the surface species information calculated from the unoptimized SCM with the adsorption condition as input features. Compared with the optimized SCM, the species-informed model could make more accurate predictions on pH edges, isotherms, and kinetic data for various input conditions (for chromate: root mean square error (RMSE) on test set = 5.90 %; for arsenate: RMSE on test set = 4.84 %). Furthermore, the utilization of the interpretable formula based on Local Interpretable Model-Agnostic Explanations (LIME) enabled the species-informed model to provide surface species information like SCM. The species-informed SCM-ML hybrid modeling method proposed in this study has great practicality and application potential, and is expected to become a new paradigm in surface adsorption model.

Coupling machine learning and theoretical models to compare key properties of biochar in adsorption kinetics rate and maximum adsorption capacity for emerging contaminants

Insights into key properties of biochar with a fast adsorption rate and high adsorption capacity are urgent to design biochar as an adsorbent in pollution emergency treatment. Machine learning (ML) incorporating classical theoretical adsorption models was applied to build prediction models for adsorption kinetics rate (i.e., K) and maximum adsorption capacity (i.e., Qm) of emerging contaminants (ECs) on biochar. Results demonstrated that the prediction performance of adaptive boosting algorithm significantly improved after data preprocessing (i.e., log-transformation) in the small unbalanced datasets with R2 of 0.865 and 0.874 for K and Qm, respectively. The surface chemistry, primarily led by ash content of biochar significantly influenced the K, while surface porous structure of biochar showed a dominant role in predicting Qm. An interactive platform was deployed for relevant scientists to predict K and Qm of new biochar for ECs. The research provided practical references for future engineered biochar design for ECs removal.

Exploring optical descriptors for rapid estimation of coastal sediment organic carbon and nearby land-use classifications via machine learning models

This study utilizes ultraviolet and fluorescence spectroscopic indices of dissolved organic matter (DOM) from sediments, combined with machine learning (ML) models, to develop an optimized predictive model for estimating sediment total organic carbon (TOC) and identifying adjacent land-use types in coastal sediments from the Yellow and Bohai Seas. Our results indicate that ML models surpass traditional regression techniques in estimating TOC and classifying land-use types. Penalized Least Squares Regression (PLR) and Cubist models show exceptional TOC estimation capabilities, with PLR exhibiting the lowest training error and Cubist achieving a correlation coefficient 0.79. In land-use classification, Support Vector Machines achieved 85.6 % accuracy in training and 92.2 % in testing. Maximum fluorescence intensity and ultraviolet absorbance at 254 nm were crucial factors influencing TOC variations in coastal sediments. This study underscores the efficacy of ML models utilizing DOM optical indices for near real-time estimation of marine sediment TOC and land-use classification.

Active Learning-Based Guided Synthesis of Engineered Biochar for CO2 Capture

Biomass waste-derived engineered biochar for CO2 capture presents a viable route for climate change mitigation and sustainable waste management. However, optimally synthesizing them for enhanced performance is time- and labor-intensive. To address these issues, we devise an active learning strategy to guide and expedite their synthesis with improved CO2 adsorption capacities. Our framework learns from experimental data and recommends optimal synthesis parameters, aiming to maximize the narrow micropore volume of engineered biochar, which exhibits a linear correlation with its CO2 adsorption capacity. We experimentally validate the active learning predictions, and these data are iteratively leveraged for subsequent model training and revalidation, thereby establishing a closed loop. Over three active learning cycles, we synthesized 16 property-specific engineered biochar samples such that the CO2 uptake nearly doubled by the final round. We demonstrate a data-driven workflow to accelerate the development of high-performance engineered biochar with enhanced CO2 uptake and broader applications as a functional material.

The application of machine learning methods for prediction of heavy metal by activated carbons, biochars, and carbon nanotubes

Carbonaceous materials are commonly used as adsorbents for heavy metals. The determination of the adsorption capacity needs time and energy, and the key factors affecting the adsorption capacity have not been determined. Therefore, a new and efficient method is needed to predict the adsorption capacity and explore the decisive factors in the adsorption process. In this study, three tree-based machine learning models (i.e., random forest, gradient boosting decision tree, and extreme gradient boosting) were developed to predict the adsorption capacity of eight heavy metals (i.e., As, Cd, Cr, Cu, Hg, Ni, Pb, and Zn) on activated carbons, biochars, and carbon nanotubes using 3674 data points extracted from 151 journal articles. After a comprehensive comparison, the gradient boosting decision tree had the best performance for a combined model based on all data (R2 = 0.9707, RMSE = 0.1420). Moreover, independent models were developed for three datasets classified by the adsorbent and eight datasets classified by the heavy metals. In addition, a graphical user interface was built to predict the adsorption capacity of heavy metals. This study provides a novel strategy and convenient tool for the removal of heavy metals and can help to improve the removal efficiency of heavy metals to build a healthier world.

Bioretention Design Modifications Increase the Simulated Capture of Hydrophobic and Hydrophilic Trace Organic Compounds

Stormwater rapidly moves trace organic contaminants (TrOCs) from the built environment to the aquatic environment. Bioretention cells reduce loadings of some TrOCs, but they struggle with hydrophilic compounds. Herein, we assessed the potential to enhance TrOC removal via changes in bioretention system design by simulating the fate of seven high-priority stormwater TrOCs (e.g., PFOA, 6PPD-quinone, PAHs) with log KOC values between -1.5 and 6.74 in a bioretention cell. We evaluated eight design and management interventions for three illustrative use cases representing a highway, a residential area, and an airport. We suggest two metrics of performance: mass advected to the sewer network, which poses an acute risk to aquatic ecosystems, and total mass advected from the system, which poses a longer-term risk for persistent compounds. The optimized designs for each use case reduced effluent loadings of all but the most polar compound (PFOA) to <5% of influent mass. Our results suggest that having the largest possible system area allowed bioretention systems to provide benefits during larger events, which improved performance for all compounds. To improve performance for the most hydrophilic TrOCs, an amendment like biochar was necessary; field-scale research is needed to confirm this result. Our results showed that changing the design of bioretention systems can allow them to effectively capture TrOCs with a wide range of physicochemical properties, protecting human health and aquatic species from chemical impacts.

Prediction of antibiotic sorption in soil with machine learning and analysis of global antibiotic resistance risk

Although the sorption of antibiotics in soil has been extensively studied, their spatial distribution patterns and sorption mechanisms still need to be clarified, which hinders the assessment of antibiotic resistance risk. In this study, machine learning was employed to develop the models for predicting the soil sorption behavior of three classes of antibiotics (sulfonamides, tetracyclines, and fluoroquinolones) in 255 soils with 2203 data points. The optimal independent models obtained an accurate predictive performance with R2 of 0.942 to 0.977 and RMSE of 0.051 to 0.210 on test sets compared to combined models. Besides, a global map of the antibiotic sorption capacity of soil predicted with the optimal models revealed that the sorption potential of fluoroquinolones was the highest, followed by tetracyclines and sulfonamides. Additionally, 14.3% of regions had higher antibiotic sorption potential, mainly in East and South Asia, Central Siberia, Western Europe, South America, and Central North America. Moreover, a risk index calculated with the antibiotic sorption capacity of soil and population density indicated that about 3.6% of soils worldwide have a high risk of resistance, especially in South and East Asia with high population densities. This work has significant implications for assessing the antibiotic contamination potential and resistance risk.

Using machine learning to explore oxyanion adsorption ability of goethite with different specific surface area

In this study, we developed prediction models for the adsorption of divalent and trivalent oxyanions on goethite based on machine learning algorithms. After verifying the reliability of the models, the importance of goethite specific surface area (SSA) and the average oxyanion adsorption capacities of goethite with different SSAs were calculated by shapley additive explanations (SHAP) importance analysis and partial dependence (PD) analysis. Despite there were differences in the feature importance of divalent and trivalent oxyanions, the contribution of goethite's SSA to the adsorption amount ranked the fourth based on SHAP importance, indicating SSA played the important role in oxyanion adsorption. Meanwhile, the PD values of SSA and the optimized complexation constants from surface complexation modeling (SCM) both indicated a non-monotonic relationship between the goethite with different SSA and its oxyanions binding capacity. When the total site concentration and crystal face composition were used as the machine learning model input features, the SHAP importance values of crystal faces and the PD decomposition results indicated that the (001) face showed the crucial influence on oxyanions adsorption amount. These findings demonstrated the important role of crystal face composition in goethite's adsorption ability, and provided a theoretical explanation for the variations of oxyanions adsorption amount on different SSA goethite.

Predicting apparent adsorption capacity of sediment-amended activated carbon for hydrophobic organic contaminants using machine learning

In-situ stabilization of hydrophobic organic compounds (HOCs) using activated carbon (AC) is a promising sediment remediation approach. However, predicting HOC adsorption capacity of sediment-amended AC remains a challenge because a prediction model is currently unavailable. Thus, the objective of this study was to develop machine learning models that could predict the apparent adsorption capacity of sediment-amended AC (KAC,apparent) for HOCs. These models were trained using 186 sets of experimental data obtained from the literature. The best-performing model among those employing various model frameworks, machine learning algorithms, and combination of candidate input features excellently predicted logKAC,apparent with a coefficient of determination of 0.94 on the test dataset. Its prediction results and experimental data for KAC,apparent agreed within 0.5 log units with few exceptions. Analysis of feature importance for the machine learning model revealed that KAC,apparent was strongly correlated with the hydrophobicity of HOCs and the particle size of AC, which agreed well with the current knowledge obtained from experimental and mechanistic assessments. On the other hand, correlation of KAC,apparent to sediment characteristics, duration of AC-sediment contact, and AC dose identified in the model disagreed with relevant arguments made in the literature, calling for further assessment in this subject. This study highlights the promising capability of machine learning in predicting adsorption capacity of AC in complex systems. It offers unique insights into the influence of model parameters on KAC,apparent.

Multitask Deep Learning Enabling a Synergy for Cadmium and Methane Mitigation with Biochar Amendments in Paddy Soils

Biochar has demonstrated significant promise in addressing heavy metal contamination and methane (CH4) emissions in paddy soils; however, achieving a synergy between these two goals is challenging due to various variables, including the characteristics of biochar and soil properties that influence biochar's performance. Here, we successfully developed an interpretable multitask deep learning (MTDL) model by employing a tensor tracking paradigm to facilitate parameter sharing between two separate data sets, enabling a synergy between Cd and CH4 mitigation with biochar amendments. The characteristics of biochar contribute similar weightings of 67.9% and 62.5% to Cd and CH4 mitigation, respectively, but their relative importance in determining biochar's performance varies significantly. Notably, this MTDL model excels in custom-tailoring biochar to synergistically mitigate Cd and CH4 in paddy soils across a wide geographic range, surpassing traditional machine learning models. Our findings deepen our understanding of the interactive effects of Cd and CH4 mitigation with biochar amendments in paddy soils, and they also potentially extend the application of artificial intelligence in sustainable environmental remediation, especially when dealing with multiple objectives.

Machine learning based prediction and experimental validation of arsenite and arsenate sorption on biochars

Arsenic (As) contamination in water is a significant environmental concern with profound implications for human health. Accurate prediction of the adsorption capacity of arsenite [As(III)] and arsenate [As(V)] on biochar is vital for the reclamation and recycling of polluted water resources. However, comprehending the intricate mechanisms that govern arsenic accumulation on biochar remains a formidable challenge. Data from the literature on As adsorption to biochar was compiled and fed into machine learning (ML) based modelling algorithms, including AdaBoost, LGBoost, and XGBoost, in order to build models to predict the adsorption efficiency of As(III) and As(V) to biochar, based on the compositional and structural properties. The XGBoost model showed superior accuracy and performance for prediction of As adsorption efficiency (for As(III): coefficient of determination (R2) = 0.93 and root mean square error (RMSE) = 1.29; for As(V), R2 = 0.99, RMSE = 0.62). The initial concentrations of As(III) and As(V) as well as the dosage of the adsorbent were the most significant factors influencing adsorption, explaining 48 % and 66 % of the variability for As(III) and As(V), respectively. The structural properties and composition of the biochar explained 12 % and 40 %, respectively, of the variability of As(III) adsorption, and 13 % and 21 % of that of As(V). The XGBoost models were validated using experimental data. R2 values were 0.9 and 0.84, and RMSE values 6.5 and 8.90 for As(III) and As(V), respectively. The ML approach can be a valuable tool for improving the treatment of inorganic As in aqueous environments as it can help estimate the optimal adsorption conditions of As in biochar-amended water, and serve as an early warning for As-contaminated water.

Exploring Optimization of Zeolites as Adsorbents for Rare Earth Elements in Continuous Flow by Machine Learning Techniques

Unsupervised machine learning (ML) techniques are applied to the characterization of the adsorption of rare earth elements (REEs) by zeolites in continuous flow. The successful application of principal component analysis (PCA) and K-Means algorithms from ML allowed for a wide range assessment of the adsorption results. This global approach permits the evaluation of the different stages of the sorption cycles and their optimization and improvement. The results from ML are also used for the definition of a regression model to estimate other REEs' recoveries based on the known values of the tested REEs. Overall, it was possible to remove more than 70% of all REEs from aqueous solutions during the adsorption assays and to recover over 80% of the REEs entrapped on the zeolites using an optimized desorption cycle.

Prediction of arsenic adsorption onto metal organic frameworks and adsorption mechanisms interpretation by machine learning

Metal-organic frameworks (MOFs) are promising adsorbents for the removal of arsenic (As) from wastewater. The As removal efficiency is influenced by several factors, such as the textural properties of MOFs, adsorption conditions, and As species. Examining all of the relevant factors through traditional experiments is challenging. To predict the As adsorption capacities of MOFs toward organic, inorganic, and total As and reveal the adsorption mechanisms, four machine learning-based models were developed, with the adsorption conditions, MOF properties, and characteristics of different As species as inputs. The results demonstrated that the extreme gradient boosting (XGBoost) model exhibited the best predictive performance (test R2 = 0.93-0.96). The validation experiments demonstrated the high accuracy of the inorganic As-based XGBoost model. The feature importance analysis showed that the concentration of As, the surface area of MOFs, and the pH of the solution were the three key factors governing inorganic-As adsorption, while those governing organic-As adsorption were the concentration of As, the pHpzc value of MOFs, and the oxidation state of the metal clusters. The formation of coordination complexes between As and MOFs is possibly the major adsorption mechanism for both inorganic and organic As. However, electrostatic interaction may have a greater effect on organic-As adsorption than on inorganic-As adsorption. Overall, this study provides a new strategy for evaluating As adsorption on MOFs and discovering the underlying decisive factors and adsorption mechanisms, thereby facilitating the investigation of As wastewater treatment.

Developing a hybrid model for predicting the reaction kinetics between chlorine and micropollutants in water

Understanding the reactivities of chlorine towards micropollutants is crucial for assessing the fate of micropollutants in water chlorination. In this study, we integrated machine learning with kinetic modeling to predict the reaction kinetics between micropollutants and chlorine in deionized water and real surface water. We first established a framework to predict the apparent second-order rate constants for micropollutants with chlorine by combining Morgan molecular fingerprints with machine learning algorithms. The framework was tuned using Bayesian optimization and showed high prediction accuracy. It was validated through experiments and used to predict the unreported apparent second-order rate constants for 103 emerging micropollutants with chlorine. The framework also improved the understanding of the structure-dependence of micropollutants' reactivity with chlorine. We incorporated the predicted apparent second-order rate constants into the Kintecus software to establish a hybrid model to profile the time-dependent changes of micropollutant concentrations by chlorination. The hybrid model was validated by experiments conducted in real surface water in the presence of natural organic matter. The hybrid model could predict how much micropollutants were degraded by chlorination with varied chlorine contact times and/or initial chlorine dosages. This study advances fundamental understanding of the reaction kinetics between chlorine and emerging micropollutants, and also offers a valuable tool to assess the fate of micropollutants during chlorination of drinking water.

Real-Time Sensor Data Profile-Based Deep Learning Method Applied to Open Raceway Pond Microalgal Productivity Prediction

Microalgal biotechnology holds the potential for renewable biofuels, bioproducts, and carbon capture applications due to unparalleled photosynthetic efficiency and diversity. Outdoor open raceway pond (ORP) cultivation enables utilization of sunlight and atmospheric carbon dioxide to drive microalgal biomass synthesis for production of bioproducts including biofuels; however, environmental conditions are highly dynamic and fluctuate both diurnally and seasonally, making ORP productivity prediction challenging without time-intensive physical measurements and location-specific calibrations. Here, for the first time, we present an image-based deep learning method for the prediction of ORP productivity. Our method is based on parameter profile plot images of sensor parameters, including pH, dissolved oxygen, temperature, photosynthetically active radiation, and total dissolved solids. These parameters can be remotely monitored without physical interaction with ORPs. We apply the model to data we generated during the Unified Field Studies of the Algae Testbed Public-Private-Partnership (ATP3 UFS), the largest publicly available ORP data set to date, which includes millions of sensor records and 598 productivities from 32 ORPs operated in 5 states in the United States. We demonstrate that this approach significantly outperforms an average value based traditional machine learning method (R2 = 0.77 ≫ R2 = 0.39) without considering bioprocess parameters (e.g., biomass density, hydraulic retention time, and nutrient concentrations). We then evaluate the sensitivity of image and monitoring data resolutions and input parameter variations. Our results demonstrate ORP productivity can be effectively predicted from remote monitoring data, providing an inexpensive tool for microalgal production and operational forecasting.

Abiotic Reduction of Organic and Inorganic Compounds by Fe(II)-Associated Reductants: Comprehensive Data Sets and Machine Learning Modeling

Iron-associated reductants play a crucial role in providing electrons for various reductive transformations. However, developing reliable predictive tools for estimating abiotic reduction rate constants (logk) in such systems has been impeded by the intricate nature of these systems. Our recent study developed a machine learning (ML) model based on 60 organic compounds toward one soluble Fe(II)-reductant. In this study, we built a comprehensive kinetic data set covering the reactivity of 117 organic and 10 inorganic compounds toward four major types of Fe(II)-associated reductants. Separate ML models were developed for organic and inorganic compounds, and the feature importance analysis demonstrated the significance of resonance structures, reducible functional groups, reductant descriptors, and pH in logk prediction. Mechanistic interpretation validated that the models accurately learned the impact of various factors such as aromatic substituents, complexation, bond dissociation energy, reduction potential, LUMO energy, and dominant reductant species. Finally, we found that 38% of the 850,000 compounds in the Distributed Structure-Searchable Toxicity (DSSTox) database contain at least one reducible functional group, and the logk of 285,184 compounds could be reasonably predicted using our model. Overall, the study is a significant step toward reliable predictive tools for anticipating abiotic reduction rate constants in iron-associated reductant systems.

Understanding and Designing a High-Performance Ultrafiltration Membrane Using Machine Learning

Ultrafiltration (UF) as one of the mainstream membrane-based technologies has been widely used in water and wastewater treatment. Increasing demand for clean and safe water requires the rational design of UF membranes with antifouling potential, while maintaining high water permeability and removal efficiency. This work employed a machine learning (ML) method to establish and understand the correlation of five membrane performance indices as well as three major performance-determining membrane properties with membrane fabrication conditions. The loading of additives, specifically nanomaterials (A_wt %), at loading amounts of >1.0 wt % was found to be the most significant feature affecting all of the membrane performance indices. The polymer content (P_wt %), molecular weight of the pore maker (M_Da), and pore maker content (M_wt %) also made considerable contributions to predicting membrane performance. Notably, M_Da was more important than M_wt % for predicting membrane performance. The feature analysis of ML models in terms of membrane properties (i.e., mean pore size, overall porosity, and contact angle) provided an unequivocal explanation of the effects of fabrication conditions on membrane performance. Our approach can provide practical aid in guiding the design of fit-for-purpose separation membranes through data-driven virtual experiments.

Recent advances in g-C3N4-based photocatalysis for water treatment: Magnetic and floating photocatalysts, and applications of machine-learning techniques

Over the past decade, there has been a substantial increase in research investigating the potential of graphitic carbon nitride (g-C3N4) for various environmental remediations. Renowned for its photocatalytic activity under visible light, g-C3N4 offers a promising solution for treating water pollutants. However, traditional g-C3N4-based photocatalysts have inherent drawbacks, creating a disparity between laboratory efficacy and real-world applications. A primary practical challenge is their fine-powdered form, which hinders separation and recycling processes. A promising approach to address these challenges involves integrating magnetic or floating materials into conventional photocatalysts, a strategy gaining traction within the g-C3N4-based photocatalyst arena. Another emerging solution to enhance practical applications entails merging experimental results with contemporary computational methods. This synergy seeks to optimize the synthesis of more efficient photocatalysts and pinpoint optimal conditions for pollutant removal. While numerous review articles discuss the laboratory-based photocatalytic applications of g-C3N4-based materials, there is a conspicuous absence of comprehensive coverage regarding state-of-the-art research on improved g-C3N4-based photocatalysts for practical applications. This review fills this void, spotlighting three pivotal domains: magnetic g-C3N4 photocatalysts, floating g-C3N4 photocatalysts, and the application of machine learning to g-C3N4 photocatalysis. Accompanied by a thorough analysis, this review also provides perspectives on future directions to enhance the efficacy of g-C3N4-based photocatalysts in water purification.

First-principles, machine learning and symbolic regression modelling for organic molecule adsorption on two-dimensional CaO surface

Data-driven methods are receiving significant attention in recent years for chemical and materials researches; however, more works should be done to leverage the new paradigm to model and analyze the adsorption of the organic molecules on low-dimensional surfaces beyond using the traditional simulation methods. In this manuscript, we employ machine learning and symbolic regression method coupled with DFT calculations to investigate the adsorption of atmospheric organic molecules on a low-dimensional metal oxide mineral system. The starting dataset consisting of the atomic structures of the organic/metal oxide interfaces are obtained via the density functional theory (DFT) calculation and different machine learning algorithms are compared, with the random forest algorithm achieving high accuracies for the target output. The feature ranking step identifies that the polarizability and bond type of the organic adsorbates are the key descriptors for the adsorption energy output. In addition, the symbolic regression coupled with genetic programming automatically identifies a series of hybrid new descriptors displaying improved relevance with the target output, suggesting the viability of symbolic regression to complement the traditional machine learning techniques for the descriptor design and fast modeling purposes. This manuscript provides a framework for effectively modeling and analyzing the adsorption of the organic molecules on low-dimensional surfaces via comprehensive data-driven approaches.

Predicting adsorption of organic compounds onto graphene and black phosphorus by molecular dynamics and machine learning

With an increase in production and application of various engineering nanomaterials (ENMs), they will inevitably be released into the environment. Adsorption of various organic chemicals onto ENMs will impact on their environmental behavior and toxicology. It is unrealistic to experimentally determine adsorption equilibrium constants (K) for the vast number of organics and ENMs due to high cost in expenditure and time. Herein, appropriate molecular dynamics (MD) methods were evaluated and selected by comparing experimental K values of seven organics adsorbed onto graphene with the MD-calculated ones. Machine learning (ML) models on K of organics adsorption onto graphene and black phosphorus nanomaterials were constructed based on a benchmark data set from the MD simulations. Lasso models based on Mordred descriptors outperformed ML models built by support vector machine, random forest, k-nearest neighbor, and gradient boosting decision tree, in terms of cross-validation coefficients (Q2 > 0.90). The Lasso models also outperformed conventional poly-parameter linear free energy relationship models for predicting logK. Compared with previous models, the Lasso models considered more compounds with different functional groups and thus have broader applicability domains. This study provides a promising way to fill the data gap in logK for chemicals adsorbed onto the ENMs.

Predicting aqueous sorption of organic pollutants on microplastics with machine learning

Microplastics (MPs) are ubiquitously distributed in freshwater systems and they can determine the environmental fate of organic pollutants (OPs) via sorption interaction. However, the diverse physicochemical properties of MPs and the wide range of OP species make a deeper understanding of sorption mechanisms challenging. Traditional isotherm-based sorption models are limited in their universality since they normally only consider the nature and characteristics of either sorbents or sorbates individually. Therefore, only specific equilibrium concentrations or specific sorption isotherms can be used to predict sorption. To systematically evaluate and predict OP sorption under the influence of both MPs and OPs properties, we collected 475 sorption data from peer-reviewed publications and developed a poly-parameter-linear-free-energy-relationship-embedded machine learning method to analyze the collected sorption datasets. Models of different algorithms were compared, and the genetic algorithm and support vector machine hybrid model displayed the best prediction performance (R2 of 0.93 and root-mean-square-error of 0.07). Finally, comparison results of three feature importance analysis tools (forward step wise method, Shapley method, and global sensitivity analysis) showed that chemical properties of MPs, excess molar refraction, and hydrogen-bonding interaction of OPs contribute the most to sorption, reflecting the dominant sorption mechanisms of hydrophobic partitioning, hydrogen bond formation, and π-π interaction, respectively. This study presents a novel sorbate-sorbent-based ML model with a wide applicability to expand our capacity in understanding the complicated process and mechanism of OP sorption on MPs.

Green and facile recycling of bauxite residue to biochar-supported iron-based composite material for hydrothermal liquefaction of municipal solid waste

Industrial and municipal wastes remain significant sources of air, soil, and water pollution, thus causing adverse climate and health impacts. EU faces challenges in developing green recycling processes and reducing GHG emissions. Innovation in green catalysis is a key driver toward the fulfilment of these goals. This study demonstrated a single-step "Green Recycling" route by which different wastes e.g., industrial and bioorganic wastes are treated to produce biochar/Fe(0) (BC-Fe(0)) material. Typically, three different biomass namely organic fraction of municipal solid waste (biopulp), wheat straw (WS), and microalgae (MA) were used as green reducing agents for reducing bauxite residue (BR). Among all biomass, the high reduction potential of amino acids present in biopulp facilitated the synthesis of BC-Fe(0). BC-Fe(0) material acted as an effective catalyst for HTL of biopulp as the results showed the highest bio-crude yield (44 wt%) at 300 °C for 30 min with 10 wt% BC-Fe(0) loading (containing 2.5 wt% Fe). Furthermore, BC-Fe(0) also assisted in-situ hydrogenation and deoxygenation of chemical compounds present in the bio-liquid product, therefore bio-crude exhibited a higher H/C ratio (1.73) and lower oxygen contents (9.78 wt%) in comparison to bio-crude obtained without catalyst. However, Raw BR and reduced BR (RED) as catalysts showed no significant effect on the yield and oxygen content of bio-crude, which confirms the high catalytic activity of Fe(0) containing BC-Fe(0). Therefore, this study demonstrates the greener path for the one-step valorization of industrial and organic wastes, as an alternative to existing chemical and high temperature-based waste recycling and catalyst synthesis technologies.

Permanganate activation by glucose-derived carbonaceous materials for highly efficient degradation of phenol and p-nitrophenol: Formation of hydroxyl radicals and multiple roles of carbonaceous materials

Owing to its inertness toward refractory organic pollutants and the release of Mn²⁺, the use of permanganate was limited in soil and groundwater remediation. The present study proposed an improvement strategy based on glucose-derived carbonaceous materials, which enhanced the potential of permanganate degrading organic pollutants. The glucose-derived carbonaceous material with 1000 °C charring temperature was named C1000, which was exploited in activating KMnO₄ for the elimination of refractory organic contaminants. The addition of C1000 in the KMnO₄ system triggered the degradation of refractory p-nitrophenol and quicken phenol degradation. Unlike the detection of Mn(III) species in a solo KMnO₄ system, the presence of C1000 facilitated the formation of ^•OH in the KMnO₄ system, which was confirmed by the use of quenchers such as methanol, benzoic acid, tertiary butanol, and carbonate. Additionally, the glucose-derived carbonaceous material played multiple roles in improving the performance of permanganate, including the enrichment of organic pollutants, donation of electrons to permanganate, and acting as an electron shuttle to facilitate the oxidation of organic pollutants by permanganate. The study's novel findings have the potential to expand the use of permanganate in the remediation of organic pollutants.

A machine learning based approach for estimating site-specific partition coefficient Kd of organic compounds: Application to nonionic pesticides

The partitioning coefficient K_d for a specific compound and location is not only a key input parameter of fate and transport models, but also critical in estimating the safety environmental concentration threshold. In order to reduce the uncertainty caused by non-linear interactions among environmental factors, machine learning based models for predicting K_d were developed in this work based on literature datasets of nonionic pesticides including molecular descriptors, soil properties, and experimental settings. The equilibrium concentration (C_e) values were specifically included for the reason that a varied range of K_d corresponding to a given C_e occurred in a real environment. By transforming 466 isotherms reported in the literature, 2618 paired equilibrium concentrations of liquid-solid (C_e-Q_e) data points were obtained. Results of SHapley Additive exPlanations revealed that soil organic carbon, C_e, and cavity formation were the most important. The distance-based applicability domain analysis was conducted for the 27 most frequently used pesticides with 15952 pieces of soil information from the HWSD-China dataset by setting three C_e scenarios (i.e., 10, 100, and 1000 μg L^-1). It was revealed the groups of compounds showing log K_d < 0.06 and log K_d > 1.19 were composed mostly of those with log K_ow of -0.800 and 5.50, respectively. When log K_d varied between 0.100 and 1.00, it was impacted by interactions among soil types, molecular descriptors, and C_e comprehensively, which accounted for 55% of the total 2618 calculations. It could be concluded that site-specific models developed in this work are necessary and practicable for the environmental risk assessment and management of nonionic organic compounds.

Optimizing the Quality of Machine Learning for Identifying the Share of Biogenic and Fossil Carbon in Solid Waste

Insights into carbon sources (biogenic and fossil carbon) and contents in solid waste are vital for estimating the carbon emissions from incineration plants. However, the traditional methods are time-, labor-, and cost-intensive. Herein, high-quality data sets were established after analyzing the carbon contents and infrared spectra of substantial samples using elemental analysis and attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR), respectively. Then, five classification and eight regression machine learning (ML) models were evaluated to recognize the proportion of biogenic and fossil carbon in solid waste. Using the optimized data preprocessing approach, the random forest (RF) classifier with hyperparameter tuning ranked first in classifying the carbon group with a test accuracy of 0.969, and the carbon contents were successfully predicted by the RF regressor with R² = 0.926 considering performance-interpretability-computation time competition. The above proposed algorithms were further validated with real environmental samples, which exhibited robust performance with an accuracy of 0.898 for carbon group classification and an R² value of 0.851 for carbon content prediction. The reliable results indicate that ATR-FTIR coupled with ML algorithms is feasible for rapidly identifying both carbon groups and content, facilitating the calculation and assessment of carbon emissions from solid waste incineration.

Enhancing Biochar-Based Nonradical Persulfate Activation Using Data-Driven Techniques

Converting biomass into biochar (BC) as a functional biocatalyst to accelerate persulfate activation for water remediation has attracted much attention. However, due to the complex structure of BC and the difficulty in identifying the intrinsic active sites, it is essential to understand the link between various properties of BC and the corresponding mechanisms promoting nonradicals. Machine learning (ML) recently demonstrated significant potential for material design and property enhancement to help tackle this problem. Herein, ML techniques were applied to guide the rational design of BC for the targeted acceleration of nonradical pathways. The results showed a high specific surface area, and O% values can significantly enhance nonradical contribution. Furthermore, the two features can be regulated by simultaneously tuning the temperatures and biomass precursors for efficient directed nonradical degradation. Finally, two nonradical-enhanced BCs with different active sites were prepared based on the ML results. This work serves as a proof of concept for applying ML in the synthesis of tailored BC for persulfate activation, thereby revealing the remarkable capability of ML for accelerating bio-based catalyst development.

Improved Machine Learning Models by Data Processing for Predicting Life-Cycle Environmental Impacts of Chemicals

Machine learning (ML) provides an efficient manner for rapid prediction of the life-cycle environmental impacts of chemicals, but challenges remain due to low prediction accuracy and poor interpretability of the models. To address these issues, we focused on data processing by using a mutual information-permutation importance (MI-PI) feature selection method to filter out irrelevant molecular descriptors from the input data, which improved the model interpretability by preserving the physicochemical meanings of original molecular descriptors without generation of new variables. We also applied a weighted Euclidean distance method to mine the data most relevant to the predicted targets by quantifying the contribution of each feature, thereby the prediction accuracy was improved. On the basis of above data processing, we developed artificial neural network (ANN) models for predicting the life-cycle environmental impacts of chemicals with R² values of 0.81, 0.81, 0.84, 0.75, 0.73, and 0.86 for global warming, human health, metal depletion, freshwater ecotoxicity, particulate matter formation, and terrestrial acidification, respectively. The ML models were interpreted using the Shapley additive explanation method by quantifying the contribution of each input molecular descriptor to environmental impact categories. This work suggests that the combination of feature selection by MI-PI and source data selection based on weighted Euclidean distance has a promising potential to improve the accuracy and interpretability of the models for predicting the life-cycle environmental impacts of chemicals.

A data-driven approach for understanding the structure dependence of redox activity in humic substances

Humic substances (HS) can facilitate electron transfer during biogeochemical processes due to their redox properties, but the structure-redox activity relationships are still difficult to describe and poorly understood. Herein, the linear (Partial Least Squares regressions; PLS) and nonlinear (artificial neural network; ANN) models were applied to monitor the structure dependence of HS redox activities in terms of electron accepting (EAC), electron donating (EDC) and overall electron transfer capacities (ETC) using its physicochemical features as input variables. The PLS model exhibited a moderate ability with R² values of 0.60, 0.53 and 0.65 to evaluate EAC, EDC and ETC, respectively. The variable influence in the projection (VIP) scores of the PLS identified that the phenols, quinones and aromatic systems were particularly important for describing the redox activities of HS. Compared with the PLS model, the back-propagation ANN model achieved higher performance with R² values of 0.81, 0.65 and 0.78 for monitoring the EAC, EDC and ETC, respectively. Sensitivity analysis of the ANN separately identified that the EAC highly depended on quinones, aromatics and protein-like fluorophores, while the EDC depended on phenols, aromatics and humic-like fluorophores (or stable free radicals). Additionally, carboxylic groups were the best indicator for evaluating both the EAC and EDC. Good model performances were obtained from the selected features via the PLS and sensitivity analysis, further confirming the accuracy of describing the structure-redox activity relationships with these analyses. This study provides a potential approach for identifying the structure-activity relationships of HS and an efficient machine-learning model for predicting HS redox activities.

Universal machine-learning algorithm for predicting adsorption performance of organic molecules based on limited data set: Importance of feature description

Adsorption of organic molecules from aqueous solution offers a simple and effective method for their removal. Recently, there have been several attempts to apply machine learning (ML) for this problem. To this end, polyparameter linear free energy relationships (pp-LFERs) were employed, and poor prediction results were observed outside model applicability domain of pp-LFERs. In this study, we improved the applicability of ML methods by adopting a chemical-structure (CS) based approach. We used the prediction of adsorption of organic molecules on carbon-based adsorbents as an example. Our results show that this approach can fully differentiate the structural differences between any organic molecules, while providing significant information that is relevant to their interaction with the adsorbents. We compared two CS feature descriptors: 3D-coordination and simplified molecular-input line-entry system (SMILES). We then built CS-ML models based on neural networks (NN) and extreme gradient boosting (XGB). They all outperformed pp-LFERs based models and are capable to accurately predict adsorption isotherm of isomers with similar physiochemical properties such as chiral molecules, even though they are trained with achiral molecules and racemates. We found for predicting adsorption isotherm, XGB shows better performance than NN, and 3D-coordinations allow effective differentiation between organic molecules.

Machine learning-assisted exploration for carbon neutrality potential of municipal sludge recycling via hydrothermal carbonization

In the context of advocating carbon neutrality, there are new requirements for sustainable management of municipal sludge (MS). Hydrothermal carbonization (HTC) is a promising technology to deal with high-moisture MS considering its low energy consumption (without drying pretreatment) and value-added products (i.e., hydrochar). This study applied machine learning (ML) methods to conduct a holistic assessment with higher heating value (HHV) of hydrochar, carbon recovery (CR), and energy recovery (ER) as model targets, yielding accurate prediction models with R² of 0.983, 0.844 and 0.858, respectively. Furthermore, MS properties showed positive (e.g., carbon content, HHV) and negative (e.g., ash content, O/C, and N/C) influences on the hydrochar HHV. By comparison, HTC parameters play a critical role for CR (51.7%) and ER (52.5%) prediction. The primary sludge was an optimal HTC feedstock while anaerobic digestion sludge had the lowest potential. This study provided a comprehensive reference for sustainable MS treatment and industrial application.

Ecological risk assessment based on soil adsorption capacity for heavy metals in Taihu basin, China

Due to the toxicity, bioaccumulation, non-biodegradability and perseverance of heavy metals, their risk assessment is essential for soil quality management. The Hakanson potential ecological risk index (RI), which considers the effects of heavy metal concentration and toxicity, has been widely used in soil ecological risk assessment. However, RI overlooks the influence of soil properties on the mobility and availability of heavy metals in risk assessment. To fill this gap, this study sought to develop an improved ecological risk index (IRI), which incorporates soil adsorption into RI, and applied it to evaluate the ecological risk of heavy metals in the soil of the Taihu basin, China. The soil adsorption models based on the Gradient Boosting Decision Tree (GBDT) was used to predict the soil adsorption capacity of five heavy metals (i.e. cadmium, chromium, copper, lead, zinc). The soil adsorption capacity in 1446 sites in the Taihu basin was predicted by the GBDT models and was assigned as the weight of IRI. The risk assessment results of the five metals in the Taihu basin showed that 40% of the sites were at a moderate risk level and 60% of the sites were at a slight risk level based on the RI. The value of IRI in the basin ranged from 11.1 to 75.5, with a mean value of 28.1. IRI differed from RI in spatial distribution due to the influence of soil adsorption. The comparative analysis between the metal contents in sediments and surrounding soils confirmed the tremendous influence of soil adsorption on ecological risks, indicating that soil adsorption should be taken into consideration in soil risk assessment.

Machine learning in the evaluation and prediction models of biochar application: A review

This article reviews recent studies applying machine learning (ML) approaches to biochar applications. We first briefly introduce the general biochar production process. Various aspects are contained, including the biochar application in the elimination of heavy metals and/or organic compounds and the biochar application in environmental and economic scopes, for instance, food security, energy, and carbon emission. The utilization of ML methods, including ANN, RF, and NN, plays a vital role in evaluating and predicting the efficiency of biochar absorption. It has been proved that ML methods can validly predict the adsorption effectiveness of biochar for water heavy metals with higher accuracy. Moreover, the literature proposed a comprehensive data-driven model to forecast biochar yield and compositions under various biomass input feedstock and different pyrolysis criteria. They said a 12.7% improvement in prediction accuracy compared to the existing literature. However, it might need further optimization in this direction. In summary, this review concludes increasing studies that a well-trained ML method can sufficiently reduce the number of experiment trials and working times associated with higher prediction accuracy. Moreover, further studies on ML applications are needed to optimize the trade-off between biochar yield and its composition.

Machine learning-assisted evaluation of potential biochars for pharmaceutical removal from water

A popular approach to select optimal adsorbents is to perform parallel experiments on adsorbents based on an initially decided goal such as specified product purity, efficiency, or binding capacity. To screen optimal adsorbents, we focused on the max adsorption capacity of the candidates at equilibrium in this work because the adsorption capacity of each adsorbent is strongly dependent on certain conditions. A data-driven machine learning tool for predicting the max adsorption capacity (Q_m) of 19 pharmaceutical compounds on 88 biochars was developed. The range of values of Q_m (mean 48.29 mg/g) was remarkably large, with a high number of outliers and large variability. Modified biochars enhanced the Q_m and surface area values compared with the original biochar, with a statistically significant difference (Chi-square value = 7.21-18.25, P < 0.005). K- nearest neighbors (KNN) was found to be the most optimal algorithm with a root mean square error (RMSE) of 23.48 followed by random forest and Cubist with RMSE of 26.91 and 29.56, respectively, whereas linear regression and regularization were the worst algorithms. KNN model achieved R² of 0.92 and RMSE of 16.62 for the testing data. A web app was developed to facilitate the use of the KNN model, providing a reliable solution for saving time and money in unnecessary lab-scale adsorption experiments while selecting appropriate biochars for pharmaceutical adsorption.

Investigation on the evolution of hydrothermal biochar

The objective of this study was to visualize trends and current research status of hydrothermal biochar research through a bibliometric analysis by using CiteSpace software. The original article data were collected from the Web of Science core database published between 2009 and 2020. A visual analysis network of national co-authored, institutional co-authored and author co-authored articles was created, countries, institutions and authors were classified accordingly. By visualizing the cited literature and journal co-citation networks, the main subject distribution and core journals were identified respectively. By visualizing journal co-citations, the main research content was identified. Further the cluster analysis revealed the key research directions of knowledge structure. Keyword co-occurrence analysis and key occurrence analysis demonstrate current research hotspots and new research frontiers. Through the above analysis, the cooperation and contributions of hydrothermal biochar research at different levels, from researchers to institutions to countries to macro levels, were explored, the disciplinary areas of knowledge and major knowledge sources of hydrothermal biochar were discovered, and the development lineage, current status, hotspots and trends of hydrothermal biochar were clarified. The results obtained from the study can provide a reference for scholars to gain a deeper understanding of hydrothermal biochar.

Classification and Regression Machine Learning Models for Predicting Aerobic Ready and Inherent Biodegradation of Organic Chemicals in Water

Machine learning (ML) is viewed as a promising tool for the prediction of aerobic biodegradation, one of the most important elimination pathways of organic chemicals from the environment. However, available models only have small datasets (<3200 records), make binary classification predictions, evaluate ready biodegradability, and do not incorporate experimental conditions (e.g., system setup and reaction time). This study addressed all these limitations by first compiling a large database of 12,750 records, considering both ready and inherent biodegradation under different conditions, and then developing regression and classification models using different chemical representations and ML algorithms. The best regression model (R² = 0.54 and root mean square error of 0.25) and classification model (the prediction accuracy from 85.1%) achieved very good performance. The model interpretation indicated that the models correctly captured the effects of chemical substructures, following the order of C═O > O═C-O > OH > CH₃ > halogen > branching > N > 6-member ring. The consideration of chemical speciation based on pK_a and α notations did not affect the regression model performance but significantly improved the classification model performance (the accuracy increased to 87.6%). The models also showed large applicability domains and provided reasonable predictions for more than 98% of over 850,000 environmentally relevant chemicals in the Distributed Structure-Searchable Toxicity database. These robust, trustable models were finally made widely accessible through two free online predictors with graphical user interface.

Predicting pesticide dissipation half-life intervals in plants with machine learning models

Pesticide dissipation half-life in plants is an important factor to assessing environmental fate of pesticides and establishing pre-harvest intervals critical to good agriculture practices. However, empirically measured pesticide dissipation half-lives are highly variable and the accurate prediction with models is challenging. This study utilized a dataset of pesticide dissipation half-lives containing 1363 datapoints, 311 pesticides, 10 plant types, and 4 plant component classes. Novel dissipation half-life intervals were proposed and predicted to account for high variations in empirical data. Four machine learning models (i.e., gradient boosting regression tree [GBRT], random forest [RF], supporting vector classifier [SVC], and logistic regression [LR]) were developed to predict dissipation half-life intervals using extended connectivity fingerprints (ECFP), temperature, plant type, and plant component class as model inputs. GBRT-ECFP had the best model performance with F1-micro_binary score of 0.698 ± 0.010 for the binary classification compared with other machine learning models (e.g., LR-ECFP, F1-micro_binary= 0.662 ± 0.009). Feature importance analysis of molecular structures in the binary classification identified aromatic rings, carbonyl group, organophosphate, =C-H, and N-containing heterocyclic groups as important substructures related to pesticide dissipation half-lives. This study suggests the utility of machine learning models in assessing the environmental fate of pesticides in agricultural crops.

The application of machine learning methods for prediction of metal immobilization remediation by biochar amendment in soil

Biochar has been used widely in heavy metal contaminated sites as a soil remediation agent. However, due to the diversity of soils, biochars, and heavy metal contamination status, the remediation efficiency is difficult to measure, owing to a variety of parameters such as soil, biochar properties, and remediation procedure. Thus, an appropriate method to predict the remediation results and to select the appropriate biochar for the remediation is required. We initially created a database on soil remediation by biochars, which has 930 datasets with 74 biochars and 43 soils in it, based on collecting and organizing data from published literatures. Then, using data from the database, we modeled the remediation of five heavy metals and metalloids (lead, cadmium, arsenic, copper, and zinc) by biochars using machine learning (ML) methods such as artificial neural network (ANN) and random forest (RF) to predict remediation efficiency based on biochar characteristics, soil physiochemical properties, incubation conditions (e.g., water holding capacity and remediation time), and the initial state of heavy metal. The ANN and RF models outperform the lineal model in terms of accuracy and predictive performance (R² > 0.84). Meanwhile, model tolerance of the missing data and reliability of the interpolation were studied by the predicted outputs of the models. The results showed that both ANN and RF have excellent performances, with the RF model having a higher tolerance for missing data. Finally, through the interpretability of ML models, the contribution of factors used in the model were analyzed and the findings revealed that the most influential elements of remediation were the type of heavy metals, the pH value of biochar, and the dosage and remediation time. The relative importance of variables could provide the right direction for better remediation of heavy metals in soil.

Machine learning may accelerate the recognition and control of microplastic pollution: Future prospects

Microplastics (MPs, sizes <5 mm) have been found to be widely distributed in various environments, such as marine, freshwater, terrestrial and atmospheric systems. Machine learning provides a potential solution for evaluating the ecological risks of MPs based on big data. Compared with traditional models, data-driven machine learning can accelerate the realization of the control of hazardous MPs and reduce the impact of MPs at both local and global scales. However, there are some urgent issues that should be resolved. For example, lack of MP databases and incomparable literatures causing the current MP data cannot fully support big data research. Therefore, it is imperative to formulate a set of standard and universal MP collection and testing protocols. For machine learning, predictions of large-scale MP distribution and the corresponding environmental risks remain lacking. To accelerate studies of MPs in the future, the methods and theories achieved for other particle pollutants, such as nanomaterials and aerosols, can be referenced. Beyond predication alone, the improvement of causality and interpretability of machine learning deserves attention in the studies of MP risks. Overall, this perspective paper provides insights for the development of machine learning methods in research on the environmental risks of MPs.

Application of machine learning in predicting the adsorption capacity of organic compounds onto biochar and resin

Detailed prediction of the adsorption amounts of organic pollutants in water is essential to the clean development and management of water resources. In this study, Kriging and polyparameter linear free energy relationship model are coupled to predict adsorption capacity of organic pollutants by biochar and resin. It's based on 1750 adsorption experimental data sets which contains 73 organic compounds on 50 biochars and 30 polymer resins. The Kriging-LFER model shows better accuracy and predictive performance for adsorption (R² are 0.940 and 0.976) than the published NN-LFER model (R² are 0.870 and 0.880). Local sensitivity analysis method is adopted to evaluate the influence of each variable on the adsorption coefficient of resin and find out that top sensitive parameters are V and log C_e, to guide parameter optimization. Data's uncertainty analysis is presented by Monte Carlo method. It predicts that the adsorption coefficient will range from 0.062 to 0.189 under the 95% confidence interval. The Kriging-LFER model provides great significance for understanding the importance of various parameters, reducing the number of experiments, adjusting the direction of experimental improvement, and evaluating the fate of organic pollutants in the environment.

An Intelligent Deep Learning Model for CO 2 Adsorption Prediction

In this paper, we propose a supervised deep learning neural network (D-CNN) approach to predict CO2 adsorption form the textural and compositional features of biomass porous carbon waste and adsorption features. Both the textural and compositional features of biomass porous carbon waste are utilized as inputs for the D-CNN architecture. A deep learning neural network (D-CNN) is proposed to predict the adsorption rate of ${\text{CO}}_2$ on zeolites. The adsorbed amount will be classified and predicted by the D-CNN. Three tree machine learning models, namely, gradient decision model (GDM), scalable boosting tree model (SBT), and gradient variant decision tree model (GVD), were fused. A feature importance metric was proposed using feature permutation, and the effect of each feature on the target output variable was investigated. The important extracted features from the three employed model were fused and used as the fusion feature set in our proposed model: fusion matrix deep learning model (FMDL). A dataset of 1400 data items, on adsorbent type and various adsorption pressure, is used as inputs for the D-CNN model. Comparison of the proposed model is done against the three tree models, which utilizes a single training layer. The error measure of the D-CNN and the tree model architectures utilize the mean square error confirming the efficiency of 0.00003 for our model, 0.00062 for the SBT, 0.00091 for the GDM, and 0.00098 for the GVD, after 150 epochs. The produced weight matrix was able to predict the ${\text{CO}}_2$ adsorption under diverse process settings with high accuracy of 96.4%.

Rapid assessments of light-duty gasoline vehicle emissions using on-road remote sensing and machine learning

In-time and accurate assessments of on-road vehicle emissions play a central role in urban air quality and health policymaking. However, official insight is hampered by the Inspection/Maintenance (I/M) procedure conducted in the laboratory annually. It not only has a large gap to real-world situations (e.g., meteorological conditions) but also is incapable of regular supervision. Here we build a unique dataset including 103,831 light-duty gasoline vehicles, in which on-road remote sensing (ORRS) measurements are linked to the I/M records based on the vehicle identification numbers and license plates. On this basis, we develop an ensemble model framework that integrates three machining learning algorithms, including neural network (NN), extreme gradient boosting (XGBoost), and random forest (RF). We demonstrate that this ensemble model could rapidly assess the vehicle-specific emissions (i.e., CO, HC, and NO). In particular, the model performs quite well for the passing vehicles under normal conditions (i.e., lower VSP (<18 kw/t), temperature (6-32 °C), relative humidity (<80%), and wind speed (<5 m/s)). Together with the current emission standard, we identify a large number of the 'dirty' (2.33%) or 'clean' (74.92%) vehicles in the real world. Our results show that the ORRS measurements, assisted by the machine-learning-based ensemble model developed here, can realize day-to-day supervision of on-road vehicle-specific emissions. This approach framework provides a valuable opportunity to reform the I/M procedures globally and mitigate urban air pollution deeply.

Adsorptive removal of herbicides with similar structures from water over nitrogen-enriched carbon, derived from melamine@metal-azolate framework-6

Based on the recent concern on the pollution of water bodies with herbicides, adsorptive removal of typical herbicides with similar chemical structures, e. g. clofibric acid (CLFA), methylchlorophenoxypropionic acid or mecoprop (MCPP) and 2,4-dichlorophenoxyacetic acid (2,4-D) from water was studied using a porous nitrogen-enriched carbon. To prepare the nitrogen-enriched carbon, pyrolysis of a melamine (MLM) incorporated metal-azolate framework-6 (MLM(x)@MAF6; x = 0-50 M % of the ligand 2-ethylimidazole for MAF6), that was prepared for the first time via an in situ method, was carried out. The MLM(x)@MAF6-derived carbons (MDC6M(x)s) were characterized and used in the removal of CLFA, MCPP and 2,4-D from water. We found that the MDC6M(25), obtained from MLM(25)@MAF6 with 25% MLM (as the optimum precursor composition), showed the highest maximum adsorption capacity (Q₀) of 1031 mg/g for CLFA, compared with any reported adsorbents, so far. The physicochemical properties of CLFA, as well as adsorbents and adsorptions under wide pH conditions, were employed to propose a plausible adsorption mechanism including hydrogen bonding. Remarkably, the porous carbon with enriched nitrogen, derived from MAF6 loaded MLM via in situ method, was very competitive in herbicides adsorption because of the contribution of well-dispersed nitrogen sties on the adsorbent. Finally, MDC6M(25) was suggested as a potential adsorbent for the removal of herbcides, including CLFA, MCPP and 2,4-D, from water, which is highly attractive to mitigate the environmental issue, especially, water pollution by various herbicides.

Revolutionizing Membrane Design Using Machine Learning-Bayesian Optimization

Polymeric membrane design is a multidimensional process involving selection of membrane materials and optimization of fabrication conditions from an infinite candidate space. It is impossible to explore the entire space by trial-and-error experimentation. Here, we present a membrane design strategy utilizing machine learning-based Bayesian optimization to precisely identify the optimal combinations of unexplored monomers and their fabrication conditions from an infinite space. We developed ML models to accurately predict water permeability and salt rejection from membrane monomer types (represented by the Morgan fingerprint) and fabrication conditions. We applied Bayesian optimization on the built ML model to inversely identify sets of monomer/fabrication condition combinations with the potential to break the upper bound for water/salt selectivity and permeability. We fabricated eight membranes under the identified combinations and found that they exceeded the present upper bound. Our findings demonstrate that ML-based Bayesian optimization represents a paradigm shift for next-generation separation membrane design.

Data-Driven Machine Learning in Environmental Pollution: Gains and Problems

The complexity and dynamics of the environment make it extremely difficult to directly predict and trace the temporal and spatial changes in pollution. In the past decade, the unprecedented accumulation of data, the development of high-performance computing power, and the rise of diverse machine learning (ML) methods provide new opportunities for environmental pollution research. The ML methodology has been used in satellite data processing to obtain ground-level concentrations of atmospheric pollutants, pollution source apportionment, and spatial distribution modeling of water pollutants. However, unlike the active practices of ML in chemical toxicity prediction, advanced algorithms such as deep neural networks in environmental process studies of pollutants are still deficient. In addition, over 40% of the environmental applications of ML go to air pollution, and its application range and acceptance in other aspects of environmental science remain to be increased. The use of ML methods to revolutionize environmental science and its problem-solving scenarios has its own challenges. Several issues should be taken into consideration, such as the tradeoff between model performance and interpretability, prerequisites of the machine learning model, model selection, and data sharing.

Insights into the adsorption of pharmaceuticals and personal care products (PPCPs) on biochar and activated carbon with the aid of machine learning

The science-informed design of 'green' carbonaceous materials (e.g., biochar and activated carbon) with high removal capacity of recalcitrant organic contaminants (e.g., pharmaceuticals and personal care products (PPCPs)) is indispensable for promoting sustainable wastewater treatment. In this study, machine learning (ML) incorporating PPCPs and biochar properties as well as adsorption conditions were applied to build adsorption prediction models and explore the contributions of various biochar's inherent properties to their PPCPs adsorption capacity. The results demonstrated that the models developed by detailed biochar properties (e.g., surface functionality and hierarchical porous structure) from advanced microscopic and spectroscopic techniques were more accurate (i.e., the root-mean-square error decreased by 18-24%) than those by general information such as bulk elemental composition and total pore volume. The relative importance of surface carbon functionalities ranked in the order of C-O bond > CO bond > non-polar carbon for predicting the adsorption capacity. According to the partial dependence analysis, the average pore diameters of adsorbents that were larger than the maximum diameter of PPCPs molecules by 1.5-fold to 2.5-fold favored the PPCPs adsorption. This study reveals new insights into the adsorption of PPCPs and provides a comprehensive reference for the sustainable engineering of biochar adsorbents for PPCPs wastewater treatment.

Predicting Solute Descriptors for Organic Chemicals by a Deep Neural Network (DNN) Using Basic Chemical Structures and a Surrogate Metric

Solute descriptors have been widely used to model chemical transfer processes through poly-parameter linear free energy relationships (pp-LFERs); however, there are still substantial difficulties in obtaining these descriptors accurately and quickly for new organic chemicals. In this research, models (PaDEL-DNN) that require only SMILES of chemicals were built to satisfactorily estimate pp-LFER descriptors using deep neural networks (DNN) and the PaDEL chemical representation. The PaDEL-DNN-estimated pp-LFER descriptors demonstrated good performance in modeling storage-lipid/water partitioning coefficient (log K_{storage-lipid/water}), bioconcentration factor (BCF), aqueous solubility (ESOL), and hydration free energy (freesolve). Then, assuming that the accuracy in the estimated values of widely available properties, e.g., logP (octanol-water partition coefficient), can calibrate estimates for less available but related properties, we proposed logP as a surrogate metric for evaluating the overall accuracy of the estimated pp-LFER descriptors. When using the pp-LFER descriptors to model log K_{storage-lipid/water}, BCF, ESOL, and freesolve, we achieved around 0.1 log unit lower errors for chemicals whose estimated pp-LFER descriptors were deemed "accurate" by the surrogate metric. The interpretation of the PaDEL-DNN models revealed that, for a given test chemical, having several (around 5) "similar" chemicals in the training data set was crucial for accurate estimation while the remaining less similar training chemicals provided reasonable baseline estimates. Lastly, pp-LFER descriptors for over 2800 persistent, bioaccumulative, and toxic chemicals were reasonably estimated by combining PaDEL-DNN with the surrogate metric. Overall, the PaDEL-DNN/surrogate metric and newly estimated descriptors will greatly benefit chemical transfer modeling.