Search

Limited information on the potential toxicity of ionic liquids (ILs) becomes the bottleneck that creates a barrier in their large-scale application. In this work, two quantitative structure-activity relationships (QSAR) models were used to evaluate the toxicity of ILs toward the acetylcholinesterase enzyme using multiple linear regression (MLR) and extreme learning machine (ELM) algorithms. The structures of 57 cations and 21 anions were optimized using quantum chemistry calculations. The electrostatic potential surface area (SEP) and the screening charge density distribution area (Sσ) descriptors were calculated and used for prediction of IL toxicity. Performance and predictive aptitude between MLR and ELM models were analyzed. Highest squared correlation coefficient (R2), and also lowest average absolute relative deviation (AARD%) and root-mean-square error (RMSE) were observed for training set, test set, and total set for the ELM model. These findings validated the superior performance of ELM over the MLR toxicity prediction model.

Refractive index is an important factor that determines the purity and concentration of a substance, but large sample sizes and the low efficiencies of traditional measurement methods and algorithms have led to a lack of refractive index data. To identify and control the performance of materials in chemistry and engineering, detailed refractive-index information for all substances is vital. In this study, the extreme learning machine (ELM) intelligence algorithm and a multiple linear regression (MLR) approach were used to predict the refractive index of ionic liquids (ILs) from molecular descriptors calculated by quantum chemistry. We collected 1194 refractive index data points for 115 ILs at different temperatures from 112 studies, where the average absolute relative deviation (AARD%) for the entire set was 0.855% for MLR and 0.295% for ELM. Such reliable and accurate results highlight the potential of the ELM and MLR models for predicting the refractive index and other related physical parameters of a material. [ABSTRACT FROM AUTHOR]

The partition coefficient of n -octanol/water (K ow) is generally considered a valuable characteristic in many natural sciences fields. Predictive quantitative structure-property and relationship (QSPR) models for K ow are encouraged since the experimental data are not always available. Extreme learning machine (ELM) method with the advantages of fast learning speed and good generalization ability has been applied to various domains. This work introduced COSMO descriptors (S σ -profile) of molecules as descriptors and the ELM algorithm to construct estimation models for K ow of organics. Four ELM models were developed and their outcomes were compared with the MLR models produced by the same descriptors. Results revealed that the proposed ELM models were reliable and applicable to predicting the log K ow values of compounds, especially the ELM-4 model with the best correlation coefficient R 2 (0.949) and root-mean-square error RMSE (0.358). Therefore, the proposed approaches are effective and feasible, and can be extensively generalized to forecast other important properties of chemicals. [Display omitted] • ELM models were developed for predicting the n -octanol/water partition coefficient. • The σ-profiles of chemicals were calculated as input descriptors via two methods. • Different strategies and AD analysis were employed for the verification of models. • The ELM method and σ-profiles can be used for the property prediction of compounds. [ABSTRACT FROM AUTHOR]

• The electrostatic potential surface areas (S EP) of 142 ILs were calculated. • Five quantum chemical S EP descriptors were screened for modeling. • A novel ELM model was developed to predict the ILs toxicity on Vibrio fischeri. • The R2 and RMSE of the proposed ELM model are 0.9272 and 0.3262, respectively. • ELM algorithm linking S EP can be successfully used to predict the toxicity of ILs. Ionic liquids (ILs) have attracted increasing attention both in the scientific community and the industry in the past two decades. Their risk of being inevitable released to ecosystem lights up the urgent research on their toxicity to the environment. To reduce the time and capital consumption on testing tremendous ILs ecotoxicity experimentally, it is essential to construct predictive models for estimating their toxicity. The objective of this study is to provide a new approach for evaluating the ecotoxicity of ILs. A comprehensive ecotoxicity dataset for Vibrio fischeri involving 142 ILs, was collected and investigated. The electrostatic potential surface areas (S EP) of separate cations and anions of ILs were firstly applied to develop predictive models for ecotoxicity on Vibrio fischeri. In addition, an intelligent algorithm named extreme learning machine (ELM) was employed to establish the predictive model. The squared correlation coefficients (R2), the average absolute error (AAE%) and the root-mean-square error (RMSE) of the developed model are 0.9272, 0.2101 and 0.3262 for the entire set, respectively. The proposed approach based on the high R2 and low deviation has remarkable potential for predicting ILs ecotoxicity on Vibrio fischeri. [ABSTRACT FROM AUTHOR]

The large amounts of ammonia emissions generated from industrial production have caused serious environmental pollution problems, such as soil acidification, eutrophication, the formation of fine particles and changes in the global greenhouse balance, and also greatly endanger human health. At present, effectively reducing ammonia emissions or recovering ammonia is still a huge challenge. Ionic liquids (ILs) as a new class of green solvent have been introduced for ammonia absorption with great potential, but a huge number on combination systems of ILs lead to the difficulty of measuring the ammonia solubility in all ILs by experiments (e.g., danger and cost). Hereby, this study proposed a novel approach for estimating the ammonia solubility in different ILs. A predictive model was developed based on the novel Algorithm - extreme learning machine (ELM) and the molecular descriptors of electrostatic potential surface areas (S EP) as input parameters. Besides, 502 data points of ammonia solubility in 17 ILs were gathered with a wide range of pressure and temperature. For the total set, the determination coefficient (R 2) and the average absolute relative deviation (AARD) of the developed model were 0.9937 and 2.95%, respectively. The regression plots revealed good consistency between predictive and experimental data points. Results show the good performance and reliability of the developed model, indicating that the proposed approach can be potentially applied for screening reasonable ILs to absorb ammonia from chemical industry processes. Image 1 • 502 data points of ammonia solubility in 17 ILs were collected. • Quantum chemical S EP parameters of ILs were calculated and screened for modeling. • An ELM model was developed with well-performed R 2 (0.9937) and AARD (2.95%). • The proposed model is reliable to predict the ammonia solubility. • The study is potentially helpful to screen ILs for ammonia absorption. [ABSTRACT FROM AUTHOR]