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

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.

Combining Experimental Sorption Parameters with QSAR to Predict Neonicotinoid and Transformation Product Sorption to Carbon Nanotubes and Granular Activated Carbon

We recently discovered that transformation of the neonicotinoid insecticidal pharmacophore alters sorption propensity to activated carbon, with products adsorbing less than parent compounds. To assess the environmental fate of novel transformation products that lack commercially available standards, researchers must rely on predictive approaches. In this study, we combined computationally derived quantitative structure-activity relationship (QSAR) parameters for neonicotinoids and neonicotinoid transformation products with experimentally determined Freundlich partition constants (log K F for sorption to carbon nanotubes [CNTs] and granular activated carbon [GAC]) to model neonicotinoid and transformation product sorption. QSAR models based on neonicotinoid sorption to functionalized/nonfunctionalized CNTs (used to generalize/simplify neonicotinoid-GAC interactions) were iteratively generated to obtain a multiple linear regression that could accurately predict neonicotinoid sorption to CNTs using internal and external validation (within 0.5 log units of the experimentally determined value). The log K F,CNT values were subsequently related to log K F,GAC where neonicotinoid sorption to GAC was predicted within 0.3 log-units of experimentally determined values. We applied our neonicotinoid-specific model to predict log K F,GAC for a suite of novel neonicotinoid transformation products (i.e., formed via hydrolysis, biotransformation, and chlorination) that do not have commercially available standards. We present this modeling approach as an innovative yet relatively simple technique to predict fate of highly specialized/unique polar emerging contaminants and/or transformation products that cannot be accurately predicted via traditional methods (e.g., pp-LFER), and highlights molecular properties that drive interactions of emerging contaminants.

Potential application of machine learning for exploring adsorption mechanisms of pharmaceuticals onto biochars

The increasing accumulation of pharmaceuticals in aquatic ecosystems could impair freshwater quality and threaten human health. Despite the adsorption of pharmaceuticals on biochars is one of the most cost-effective and eco-friendly removal methods, the wide variation of experimental designs and research aims among previous studies pose significant challenge in selecting biochar for optimal removal. In this work, literature data of 1033 sets with 21 variables collected from 267 papers over ten years (2010-2020) covering 19 pharmaceuticals onto 88 biochars were assessed by different machine learning (ML) algorithms i.e., Linear regression model (LM), Feed-forward neural networks (NNET), Deep neutral networks (DNN), Cubist, K-nearest neighbor (KNN), and Random forest (RF), to predict equilibrium adsorption capacity (Q_e) and explore adsorption mechanisms. LM showed the best performance on ranking importance of input variables. Except for initial concentration of pharmaceuticals, Q_e was strongly governed by biochars' properties including specific surface area (BET), pore volume (PV), and pore structure (PS) rather than pharmaceuticals' properties and experimental conditions. The most accurate model for estimating Q_e was achieved by Cubist, followed by KNN, RF, KNN, NNET and LM. The generalization ability was observed by the tuned Cubist with 26 rules for the prediction of the unseen data. This study not only provides an insightful evidence for data-based adsorption mechanisms of pharmaceuticals on biochars, but also offers a potential method to accurately predict the biochar adsorption performance without conducting any experiments, which will be of high interests in practice in terms of water/wastewater treatment using biochars.

Direct Prediction of Bioaccumulation of Organic Contaminants in Plant Roots from Soils with Machine Learning Models Based on Molecular Structures

Root concentration factor (RCF) is an important characterization parameter to describe accumulation of organic contaminants in plants from soils in life cycle impact assessment (LCIA) and phytoremediation potential assessment. However, building robust predictive models remains challenging due to the complex interactions among chemical-soil-plant root systems. Here we developed end-to-end machine learning models to devolve the complex molecular structure relationship with RCF by training on a unified RCF data set with 341 data points covering 72 chemicals. We demonstrate the efficacy of the proposed gradient boosting regression tree (GBRT) model based on the extended connectivity fingerprints (ECFP) by predicting RCF values and achieved prediction performance with R-squared of 0.77 and mean absolute error (MAE) of 0.22 using 5-fold cross validation. In addition, our results reveal nonlinear relationships among properties of chemical, soil, and plant. Further in-depth analyses identify the key chemical topological substructures (e.g., -O, -Cl, aromatic rings and large conjugated π systems) related to RCF. Stemming from its simplicity and universality, the GBRT-ECFP model provides a valuable tool for LCIA and other environmental assessments to better characterize chemical risks to human health and ecosystems.

Adsorption of polar and ionic organic compounds on activated carbon: Surface chemistry matters

Persistent and mobile organic compounds (PMOCs) are often detected micropollutants in the water cycle, thereby challenging the conventional wastewater and drinking water treatment techniques. Carbon-based adsorbents are often less effective or even unable to remove this class of pollutants. Understanding of PMOC adsorption mechanisms is urgently needed for advanced treatment of PMOC-contaminated water. Here, we investigated the effect of surface modifications of activated carbon felts (ACFs) on the adsorption of six selected PMOCs carrying polar or ionic groups. Among three ACFs, defunctionalized ACF bearing net positive surface charge at neutral pH provides the most versatile sorption efficiency for all studied PMOC types representing neutral, anionic and cationic compounds. Ion exchange capacity giving quantitative information of sorbent surface charges at specified pH is recognized as a frequently underestimated key property for evaluating adsorbents aiming at PMOC adsorption. A most recently developed prediction tool for Freundlich parameters in PMOC adsorption was applied and the prediction results are compared to the experimental data. The comparison demonstrates the so far underestimated importance of the sorbent surface chemistry for PMOC adsorption affinity and capacity. PMOC adsorption mechanisms were additionally investigated by adsorption experiments at various temperatures, pH values and electrolyte concentrations. Exothermic sorption was observed for all sorbate-sorbent pairs. Adsorption is improved for ionic PMOCs on AC carrying sites of the same charge (positive or negative) at increased electrolyte concentration, while not affected for neutral PMOCs unless strong electron donor-acceptor yet weak non-Coulombic interactions exist. Our findings will allow for better design and targeted application of activated carbon-based sorbents in water treatment facilities.

Spatial heterogeneity modeling of water quality based on random forest regression and model interpretation

A systematic understanding of the spatial distribution of water quality is critical for successful watershed management; however, the limited number of physical monitoring stations has restricted the evaluation of spatial water quality distribution and the identification of features impacting the water quality. To fill this gap, we developed a modeling process that employed the random forest regression (RFR) to model the water quality distribution for the Taihu Lake basin in Zhejiang Province, China, and adopted the Shapley Additive exPlanations (SHAP) method to interpret the underlying driving forces. We first used RFR to model three water quality parameters: permanganate index (COD_Mn), total phosphorus (TP), and total nitrogen (TN), based on 16 watershed features. We then applied the built models to generate water quality distribution maps for the basin, with the COD_Mn ranging from 1.39 to 6.40 mg/L, TP from 0.02 to 0.23 mg/L, and TN from 1.43 to 4.27 mg/L. These maps showed generally consistent patterns among the COD_Mn, TN, and TP with minor differences in the spatial distribution. The SHAP analysis showed that the TN was mainly affected by agricultural non-point sources, while the COD_Mn and TP were affected by agricultural and domestic sources. Due to differences in sewage collection and treatment between urban and rural areas, the water quality in highly populated urban areas was better than that in rural areas, which led to an unexpected positive relationship between water quality and population density. Overall, with the RFR models and SHAP interpretation, we obtained a continuous distribution pattern of the water quality and identified its driving forces in the basin. These findings provided important information to assist water quality restoration projects.

Predicting Heavy Metal Adsorption on Soil with Machine Learning and Mapping Global Distribution of Soil Adsorption Capacities

Studying heavy metal adsorption on soil is important for understanding the fate of heavy metals and properly assessing the related environmental risks. Existing experimental methods and traditional models for quantifying adsorption, however, are time-consuming and ineffective. In this study, we developed machine learning models for the soil adsorption of six heavy metals (Cd(II), Cr(VI), Cu(II), Pb(II), Ni(II), and Zn(II)) using 4420 data points (1105 soils) extracted from 150 journal articles. After a comprehensive comparison, our results showed that the gradient boosting decision tree had the best performance for a combined model based on all the data. The Shapley additive explanation method was used to identify the feature importance and the effects of these features on the adsorption, based on which six independent models were developed for the six metals to achieve better model performance than the combined model. Using these independent models, the global distribution of heavy metal adsorption capacities on soils was predicted with known soil properties. Reversed models, including one combined model for all the six metals and six independent models, were also built using the same data sets to predict the heavy metal concentration in water when the adsorbed amount is known for a soil/sediment.

Machine learning in natural and engineered water systems

Water resources of desired quality and quantity are the foundation for human survival and sustainable development. To better protect the water environment and conserve water resources, efficient water management, purification, and transportation are of critical importance. In recent years, machine learning (ML) has exhibited its practicability, reliability, and high efficiency in numerous applications; furthermore, it has solved conventional and emerging problems in both natural and engineered water systems. For example, ML can predict various water quality indicators in situ and real-time by considering the complex interactions among water-related variables. ML approaches can also solve emerging pollution problems with proven rules or universal mechanisms summarized from the related research. Moreover, by applying image recognition technology to analyze the relationships between image information and physicochemical properties of the research object, ML can effectively identify and characterize specific contaminants. In view of the bright prospects of ML, this review comprehensively summarizes the development of ML applications in natural and engineered water systems. First, the concept and modeling steps of ML are briefly introduced, including data preparation, algorithm selection and model evaluation. In addition, comprehensive applications of ML in recent studies, including predicting water quality, mapping groundwater contaminants, classifying water resources, tracing contaminant sources, and evaluating pollutant toxicity in natural water systems, as well as modeling treatment techniques, assisting characterization analysis, purifying and distributing drinking water, and collecting and treating sewage water in engineered water systems, are summarized. Finally, the advantages and disadvantages of commonly used algorithms are analyzed according to their structures and mechanisms, and recommendations on the selection of ML algorithms for different studies, as well as prospects on the application and development of ML in water science are proposed. This review provides references for solving a wider range of water-related problems and brings further insights into the intelligent development of water science.

Machine Learning: New Ideas and Tools in Environmental Science and Engineering

The rapid increase in both the quantity and complexity of data that are being generated daily in the field of environmental science and engineering (ESE) demands accompanied advancement in data analytics. Advanced data analysis approaches, such as machine learning (ML), have become indispensable tools for revealing hidden patterns or deducing correlations for which conventional analytical methods face limitations or challenges. However, ML concepts and practices have not been widely utilized by researchers in ESE. This feature explores the potential of ML to revolutionize data analysis and modeling in the ESE field, and covers the essential knowledge needed for such applications. First, we use five examples to illustrate how ML addresses complex ESE problems. We then summarize four major types of applications of ML in ESE: making predictions; extracting feature importance; detecting anomalies; and discovering new materials or chemicals. Next, we introduce the essential knowledge required and current shortcomings in ML applications in ESE, with a focus on three important but often overlooked components when applying ML: correct model development, proper model interpretation, and sound applicability analysis. Finally, we discuss challenges and future opportunities in the application of ML tools in ESE to highlight the potential of ML in this field.

Thermodynamics-Based Model Construction for the Accurate Prediction of Molecular Properties From Partition Coefficients

Developing models for predicting molecular properties of organic compounds is imperative for drug development and environmental safety; however, development of such models that have high predictive power and are independent of the compounds used is challenging. To overcome the challenges, we used a thermodynamics-based theoretical derivation to construct models for accurately predicting molecular properties. The free energy change that determines a property equals the sum of the free energy changes (ΔG_Fs) caused by the factors affecting the property. By developing or selecting molecular descriptors that are directly proportional to ΔG_Fs, we built a general linear free energy relationship (LFER) for predicting the property with the molecular descriptors as predictive variables. The LFER can be used to construct models for predicting various specific properties from partition coefficients. Validations show that the models constructed according to the LFER have high predictive power and their performance is independent of the compounds used, including the models for the properties having little correlation with partition coefficients. The findings in this study are highly useful for applications in drug development and environmental safety.

Applied Machine Learning for Prediction of CO2 Adsorption on Biomass Waste-Derived Porous Carbons

Biomass waste-derived porous carbons (BWDPCs) are a class of complex materials that are widely used in sustainable waste management and carbon capture. However, their diverse textural properties, the presence of various functional groups, and the varied temperatures and pressures to which they are subjected during CO₂ adsorption make it challenging to understand the underlying mechanism of CO₂ adsorption. Here, we compiled a data set including 527 data points collected from peer-reviewed publications and applied machine learning to systematically map CO₂ adsorption as a function of the textural and compositional properties of BWDPCs and adsorption parameters. Various tree-based models were devised, where the gradient boosting decision trees (GBDTs) had the best predictive performance with R² of 0.98 and 0.84 on the training and test data, respectively. Further, the BWDPCs in the compiled data set were classified into regular porous carbons (RPCs) and heteroatom-doped porous carbons (HDPCs), where again the GBDT model had R² of 0.99 and 0.98 on the training and 0.86 and 0.79 on the test data for the RPCs and HDPCs, respectively. Feature importance revealed the significance of adsorption parameters, textural properties, and compositional properties in the order of precedence for BWDPC-based CO₂ adsorption, effectively guiding the synthesis of porous carbons for CO₂ adsorption applications.

Use of a large dataset to develop new models for estimating the sorption of active pharmaceutical ingredients in soils and sediments

Information on the sorption of active pharmaceutical ingredients (APIs) in soils and sediments is needed for assessing the environmental risks of these substances yet these data are unavailable for many APIs in use. Predictive models for estimating sorption could provide a solution. The performance of existing models is, however, often poor and most models do not account for the effects of soil/sediment properties which are known to significantly affect API sorption. Therefore, here, we use a high-quality dataset on the sorption behavior of 54 APIs in 13 soils and sediments to develop new models for estimating sorption coefficients for APIs in soils and sediments using three machine learning approaches (artificial neural network, random forest and support vector machine) and linear regression. A random forest-based model, with chemical and solid descriptors as the input, was the best performing model. Evaluation of this model using an independent sorption dataset from the literature showed that the model was able to predict sorption coefficients of 90% of the test set to within a factor of 10 of the experimental values. This new model could be invaluable in assessing the sorption behavior of molecules that have yet to be tested and in landscape-level risk assessments.

Critical Review of Advances in Engineering Nanomaterial Adsorbents for Metal Removal and Recovery from Water: Mechanism Identification and Engineering Design

Nanomaterial adsorbents (NAs) have shown promise to efficiently remove toxic metals from water, yet their practical use remains challenging. Limited understanding of adsorption mechanisms and scaling up evaluation are the two main obstacles. To fully realize the practical use of NAs for metal removal, we review the advanced tools and chemical principles to identify mechanisms, highlight the importance of adsorption capacity and kinetics on engineering design, and propose a systematic engineering scenario for full-scale NA implementation. Specifically, we provide in-depth insight for using density functional theory (DFT) and/or X-ray absorption fine structure (XAFS) to elucidate adsorption mechanisms in terms of active site verification and molecular interaction configuration. Furthermore, we discuss engineering issues for designing, scaling, and operating NA systems, including adsorption modeling, reactor selection, and NA regeneration, recovery, and disposal. This review also prioritizes research needs for (i) determining NA microstructure properties using DFT, XAFS, and machine learning and (ii) recovering NAs from treated water. Our critical review is expected to guide and advance the development of highly efficient NAs for engineering applications.

Adsorption of neutral organic compounds on polar and nonpolar microplastics: Prediction and insight into mechanisms based on pp-LFERs

Adsorption of 18 neutral organic compounds (OCs) on polar (polybutylene succinate (PBS) and polycaprolactone (PCL)) and nonpolar (low-density polyethylene (LDPE) and polystyrene (PS)) microplastics (MPs) were investigated. The adsorption coefficients (K_d) varied with ranges of 130-42,002, 124-27,768, 6.40-10,713, and 1.52-10,332 L kg^-1 for adsorption on PCL, PBS, LDPE, and PS MPs, respectively. The polar MPs showed greater adsorption capacities than nonpolar MPs. Non-specific interaction, i.e. hydrophobic partition played a crucial role in the adsorption of OCs on all MPs, while polar interactions also contributed significantly to the greater adsorption on polar MPs. Poly-parameter linear free energy relationships (pp-LFERs) with multiple linear regression (MLR) and feedforward network (FN) were then employed to model the adsorption of OCs on MPs so as to obtain deep insights into adsorption mechanisms. The MLR models achieved R_adj² of 0.90-0.97 and root mean square error (RMSE) of 0.13-0.38 log units, while the FN models achieved R_adj² of 0.85-0.90 and RMSE of 0.21-0.60 log units. The MLR models are more accurate under selected equilibrium concentrations while FN models are capable of making predictions under varying equilibrium concentrations. Lastly, both MLR and FN models showed good prediction on literature adsorption data on nonpolar MPs.

Remarkable adsorbent for removal of bisphenol A and S from water: Porous carbon derived from melamine/polyaniline

Recently, contamination of water resources with various organics such as bisphenols is a problem worldwide. Here, we developed nitrogen-enriched porous carbons (N-PDCs) from pyrolysis of melamine-loaded polyaniline (PANI), for the first time. The N-PDCs and PANI-derived carbons (PDCs, without using melamine) were characterized and applied in adsorptive removal of two typical bisphenols, such as bisphenol A and S (BPA and BPS, respectively), from water under a wide range of conditions. Via this research, we found that one N-PDC (N-PDC-700, obtained at 700 °C) showed very remarkable performances in adsorption of BPA (Q₀: 961 mg/g) and BPS (Q₀: 971 mg/g) under pH of 7.0. In other words, N-PDC-700 has Q₀ value for BPS around 2 times as much as that of the most effective adsorbent, MIL-101-NH₂. Moreover, the Q₀ value of N-PDC-700 for BPA is the second highest, after the sp² C dominant N-doped carbon. The plausible adsorption mechanism could be suggested based on the adsorption of BPA under a wide range of pH values. Finally, the N-PDC-700 was easily recycled for several uses, suggesting the potential application in adsorption of bisphenols from water.

Quantitative structure activity relationships (QSARs) and machine learning models for abiotic reduction of organic compounds by an aqueous Fe(II) complex

Due to the increasing diversity of organic contaminants discharged into anoxic water environments, reactivity prediction is necessary for chemical persistence evaluation for water treatment and risk assessment purposes. Almost all quantitative structure activity relationships (QSARs) that describe rates of contaminant transformation apply only to narrowly-defined, relatively homogenous families of reactants (e.g., dechlorination of alkyl halides). In this work, we develop predictive models for abiotic reduction of 60 organic compounds with diverse reducible functional groups, including nitroaromatic compounds (NACs), aliphatic nitro-compounds (ANCs), aromatic N-oxides (ANOs), isoxazoles (ISXs), polyhalogenated alkanes (PHAs), sulfoxides and sulfones (SOs), and others. Rate constants for their reduction were measured using a model reductant system, Fe(II)-tiron. Qualitatively, the rates followed the order NACs > ANOs ≈ ISXs ≈ PHAs > ANCs > SOs. To develop QSARs, both conventional chemical descriptor-based and machine learning (ML)-based approaches were investigated. Conventional univariate QSARs based on a molecular descriptor E_LUMO (energy of the lowest-unoccupied molecular orbital) gave good correlations within classes. Multivariate QSARs combining E_LUMO with Abraham descriptors for physico-chemical properties gave slightly improved correlations within classes for NCs and NACs, but little improvement in correlation within other classes or among classes. The ML model obtained covers reduction rates for all classes of compounds and all of the conditions studied with the prediction accuracy similar to those of the conventional QSARs for individual classes (r² = 0.41-0.98 for univariate QSARs, 0.71-0.94 for multivariate QSARs, and 0.83 for the ML model). Both approaches required a scheme for a priori classification of the compounds for model training. This work offers two alternative modeling approaches to comprehensive abiotic reactivity prediction for persistence evaluation of organic compounds in anoxic water environments.

Sorption of PFOS in 114 Well-Characterized Tropical and Temperate Soils: Application of Multivariate and Artificial Neural Network Analyses

The influence of soil properties on PFOS sorption are not fully understood, particularly for variable charge soils. PFOS batch sorption isotherms were conducted for 114 temperate and tropical soils from Australia and Fiji, that were well-characterized for their soil properties, including total organic carbon (TOC), anion exchange capacity, and surface charge. In most soils, PFOS sorption isotherms were nonlinear. PFOS sorption distribution coefficients (K_d) ranged from 5 to 229 mL/g (median: 28 mL/g), with 63% of the Fijian soils and 35% of the Australian soils showing K_d values that exceeded the observed median K_d. Multiple linear regression showed that TOC, amorphous aluminum and iron oxides contents, anion exchange capacity, pH, and silt content, jointly explained about 53% of the variance in PFOS K_d in soils. Variable charge soils with net positive surface charges, and moderate to elevated TOC content, generally displayed enhanced PFOS sorption than in temperate or tropical soils with TOC as the only sorbent phase, especially at acidic pH ranges. For the first time, two artificial neural networks were developed to predict the measured PFOS K_d (R² = 0.80) in the soils. Overall, both TOC and surface charge characteristics of soils are important for describing PFOS sorption.

Modelling of the adsorption of Pb, Cu and Ni ions from single and multi-component aqueous solutions by date seed derived biochar: Comparison of six machine learning approaches

Biochar is an effective material for the removal of heavy metals from wastewater. Operational conditions, such as metal initial concentration, temperature, contact time as well as the presence of competing ions can impact the effectiveness of the treatment process. While several models have been proposed for modelling the adsorption process, no model currently exists that accounts for the mutual interactions of key process parameters on the adsorption capacity in multi-solute systems. The aim of this study is to address this gap in knowledge by formulating a multi-input multi-output (MIMO) model, which takes into account the effect of mutual interactions of key factors while predicting heavy metals adsorption capacity of the biochar in single and multi-solute systems. In this study, we use machine learning models, specifically several ANN models, radial basis and gradient boosting algorithms to model the MIMO process. The results of our models provide highly accurate predictions (R² > 0.99). The generalized regression network provided the best match to the experimental data. This approach can allow operators to predict how the adsorption system will respond to changes in the operations and hence provide them with a tool for process optimization.

Coupling a Feedforward Network (FN) Model to Real Adsorbed Solution Theory (RAST) to Improve Prediction of Bisolute Adsorption on Resins

When predicting bisolute adsorption, the adsorbed solution theory (AST) and real adsorbed solution theory (RAST) either frequently show high prediction deviations or require bisolute adsorption data. Emerging feedforward network (FN) models can provide high prediction accuracy but lack broad applicability. To avoid those limitations, adsorption experiments were performed for a total of 12 single solutes and 55 bisolutes onto two widely used resins (MN200 and XAD-4). Different FN-based models were then built and compared with AST and RAST, based on which a new modeling strategy coupling FN to RAST and requiring only single-solute data was proposed. The root-mean-square error (RMSE) of predictions by the FN-RAST is 0.082 log units for 50 bisolute adsorption on MN200, much lower than that by AST (0.164) and slightly higher than that by RAST (0.069) or the best FN model (0.068). The FN-RAST model further provided satisfactory predictions for 5 bisolute adsorption on XAD-4 (RMSE = 0.10), which is comparable to that by RAST (0.10) and much lower than those by AST (0.26) and FN model (0.38). Therefore, the FN-RAST enjoys both satisfactory prediction accuracy and some broad applicability. The values of Abraham descriptors E and S were also founded to help assess/compare the nonideal behavior in different bisolute mixtures.