An evaluation of methods for estimating the solubility of solids in liquids

Rights statement
Awarding institution
  • University of Strathclyde
Date of award
  • 2023
Thesis identifier
  • T16810
Person Identifier (Local)
  • 201953504
Qualification Level
Qualification Name
Department, School or Faculty
  • Crystallisation is an example of a separation and purification technique used ubiquitously in chemical manufacturing processes. For the design, development and optimisation of efficient crystallisation processes, accurate solid-liquid equilibria (i.e., solubility) is necessary. Solubility is directly measurable; however, despite years of development, available techniques have drawbacks related to the time required to complete, the amount of material required, economic cost, as well as technical challenges under certain conditions (i.e., solid form, temperature, pressure and composition). In this thesis, new model-based methods rooted in classical equilibrium thermodynamics are developed that allow the solubility of a solid in a liquid to be estimated as a function of temperature. Crucially, these methods are developed in a manner that reduce reliance on direct solubility measurements to a single temperature, and make it possible to predict solubility through measurement and mathematical modelling of thermodynamic quantities other than solubility. First, an approximate version of the van’t Hoff equation is developed and used to predict the aqueous solubility of glycine polymorphs using thermodynamic data reported in the literature at 298.15 K, only. For α-glycine, predictions in excellent agreement with direct solubility data from 260–340 K were made and additional predictions for β and γ-glycine, where less certain solubility data are available in the literature, are also provided. Then, an analytical solubility model is proposed by interpreting a criteria for thermodynamic equilibrium (i.e., equal chemical potential) in terms of a second-order Taylor series expansion. Again, this approach makes it possible to predict solubility using only a single solubility measurement and a collection of thermodynamic quantities evaluated at the chosen reference temperature. Encouraging results are presented for the aqueous solubility of α-glycine, sucrose and urea—all of which have a meaningful amount of directly determined solubility, as well as thermodynamic quantities measured at 298.15 K. In the final part of this thesis, we illustrate how to parameterise Gibbs free energy models in such a manner that solubility can be estimated precisely as a function of temperature by combining only a single measure of solubilty and a broad range of thermodynamic quantities. This retains the key advantages of the approximate approaches mentioned already, but makes it possible to reliably estimate solubility across an unbounded range of temperature, in principle. As an illustrative example, we use thermodynamic data reported in the literature for the sucrose-water system to make predictions, noting it is possible to generate estimations in excellent agreement with directly measured solubility data from the freezing point to the assumed solute-solution-vapor triple point. In summary, this thesis helps to fill a gap left by existing approaches in the ever growing suite of methods at the disposal of practitioners looking to estimate the solubility of solids in liquids.
Advisor / supervisor
  • Sefcik, Jan
  • Lue, Leo
Resource Type