Use of artificial neural networks to select hydrophobic solvents for the extraction of ions from aqueous feeds: boron and lithium case studies

Student thesis: Master's Thesis


Deep eutectic solvents (DESs) is a new class of 'green solvents' that can be easily synthesized and tuned according to their intended use. DESs enable the utilization of different functional extractants without the need of dilution in harmful solvents. One of the many applications of interest is the liquid-liquid extraction of lithium and boron from aqueous media. The solvent extraction of boron is conducted mainly to meet drinking water standards, protect sensitive crops, or recover boron from industrial wastewater. The need for the extraction of lithium from brines stems from its increasing demand, resulting from the fact that it is the core of the continuously growing lithium-ion batteries (LIBs) industry. To aid in the screening process, an artificial neural network (ANN) capable of predicting the boron extraction efficiency from an aqueous medium is reported for the first time. The dataset used to develop the ANN included all measurements reported in the literature up to the time of writing of this thesis. COSMO-RS-based molecular parameters and the extraction operational conditions were selected as model inputs. The ANN was then utilized to design six new hydrophobic DESs that were synthesized. This was followed by the multicriteria assessment of the new DESs, based on the extraction efficiency, density, viscosity, water solubility, and solvent leachability. It is also intended to develop an ANN model for the prediction of lithium extraction efficiencies. Due to the high demand for lithium, such a model will be highly valuable. Similar to the ANN model of boron, the database will depend upon the work provided in the literature. Moreover, a number of DESs of varying combinations were tested for the liquid-liquid extraction of lithium from an aqueous solution.
Date of AwardJul 2022
Original languageAmerican English


  • Deep eutectic solvents (DESs)
  • Boron Extraction
  • Lithium Extraction
  • Artificial Neural Networks (ANN).

Cite this