[Adsorption Characteristics of Tetracycline by CuFeO2-modified Biochar].

CuFeO2-modified biochars were prepared through co-precipitation and hydrothermal methods, and the composites had high efficiency removal for tetracycline (TC) from water. The CuFeO2-modified biochar with a 2:1 mass ratio of CuFeO2 to BC450 (CuFeO2/BC450=2:1) demonstrated the best adsorption performance. The kinetic process of TC adsorption by CuFeO2/BC450=2:1 was well fitted with the intraparticle diffusion model, suggesting that the adsorption process was controlled by film and pore diffusion. Under the condition of neutral pH and 298 K, the maximum adsorption capacity of the Langmuir model of CuFeO2/BC450=2:1 was 82.8 mg·g-1, which was much greater than that of BC450 (13.7 mg·g-1) and CuFeO2(14.8 mg·g-1). The thermodynamic data suggested that TC sorption onto CuFeO2/BC450=2:1 was a spontaneous and endothermic process. The removal of TC by CuFeO2/BC450=2:1 increased first and then decreased with increasing pH, and the maximum adsorption occurred under the neutral condition. The strong adsorption of TC by CuFeO2/BC450=2:1 could be attributed to better porosity, larger specific surface area, and more active sites (e.g., functional groups and charged surfaces). This work provided an efficient magnetic adsorbent for removing antibiotics.

Formation of iodinated aromatic DBPs at different molar ratios of chlorine and nitrogen in iodide-containing water.

Variations in iodinated aromatic disinfection byproducts (DBPs) in the presence of I- and organic compounds as a function of reaction time in different molar ratios (MRs) of HOCl:NH3-N were investigated. Up to 17 kinds of iodinated aromatic DBPs were identified in the breakpoint chlorination of iodide (I-)/organic (phenol, bisphenol S (BPS) and p-nitrophenol (p-NP)) systems, and the possible pathways for the formation of iodinated aromatic DBPs were proposed. The reaction pathways include HOCl/HOI electrophilic substitution and oxidation, while the dominant iodinated DBPs were quantified. In the I-/phenol system (pH = 7.0), the sum of the concentrations of four iodinated aliphatic DBPs ranged from 0.32 to 1.04 µM (triiodomethane (TIM), dichloroiodomethane (DCIM), diiodochloromethane (DICM) and monoiodoacetic acid (MIAA)), while the concentration of 4-iodophenol ranged from 2.99 to 12.87 µM. The concentration of iodinated aromatic DBPs remained stable with an MR = 1:1. When the MR was 6:1, iodinated aromatic DBPs decreased with increasing reaction time, in which the main disinfectant in the system was active chlorine. This study proposed the formation mechanism of iodinated aromatic DBPs during the breakpoint chlorination of iodide-containing water. These results can be used to control the formation of hazardous iodinated aromatic DBPs in the disinfection of iodine containing water.

Recursive Random Forests Enable Better Predictive Performance and Model Interpretation than Variable Selection by LASSO.

Variable selection is of crucial significance in QSAR modeling since it increases the model predictive ability and reduces noise. The selection of the right variables is far more complicated than the development of predictive models. In this study, eight continuous and categorical data sets were employed to explore the applicability of two distinct variable selection methods random forests (RF) and least absolute shrinkage and selection operator (LASSO). Variable selection was performed: (1) by using recursive random forests to rule out a quarter of the least important descriptors at each iteration and (2) by using LASSO modeling with 10-fold inner cross-validation to tune its penalty λ for each data set. Along with regular statistical parameters of model performance, we proposed the highest pairwise correlation rate, average pairwise Pearson's correlation coefficient, and Tanimoto coefficient to evaluate the optimal by RF and LASSO in an extensive way. Results showed that variable selection could allow a tremendous reduction of noisy descriptors (at most 96% with RF method in this study) and apparently enhance model's predictive performance as well. Furthermore, random forests showed property of gathering important predictors without restricting their pairwise correlation, which is contrary to LASSO. The mutual exclusion of highly correlated variables in LASSO modeling tends to skip important variables that are highly related to response endpoints and thus undermine the model's predictive performance. The optimal variables selected by RF share low similarity with those by LASSO (e.g., the Tanimoto coefficients were smaller than 0.20 in seven out of eight data sets). We found that the differences between RF and LASSO predictive performances mainly resulted from the variables selected by different strategies rather than the learning algorithms. Our study showed that the right selection of variables is more important than the learning algorithm for modeling. We hope that a standard procedure could be developed based on these proposed statistical metrics to select the truly important variables for model interpretation, as well as for further use to facilitate drug discovery and environmental toxicity assessment.