Database for liquid phase diffusion coefficients at infinite dilution at 298 K and matrix completion methods for their prediction

Abstract

Experimental data on diffusion in binary liquid mixtures at 298±1 K from the literature were systematically consolidated and used to determine diffusion coefficients D∞ij of solutes i at infinite dilution in solvents j...

Rational method for defining and quantifying pseudo-components based on NMR spectroscopy

“…We recommend selecting the reference component in a way that D ∞ ref can be adopted from the literature. As an alternative, it can also be determined experimentally, which is, however, usually tedious, 60 or estimated using a prediction method. 20,21,61–63…”

“…From D ∞ Ũ , in turn, the molar mass M Ũ of component Ũ can be calculated using basically any predictive model for self-diffusion coefficients at infinite dilution. We have used the SEGWE model 20,21 in the present work, which is a semi-empirical extension of the Stokes–Einstein equation 64 and was found to be the best available semi-empirical model for predicting self-diffusion coefficients in a recent study: 60 where D ∞ Ũ is the self-diffusion coefficient of pseudo-component Ũ at infinite dilution, M Ũ is the molar mass of Ũ, k B is the Boltzmann constant, η S and M S are the dynamic viscosity and molar mass of the solvent, respectively, T is the temperature, and ρ eff is a lumped parameter of the SEGWE model, called effective density, whose default value 21 ρ eff = 627 kg m −3 was used here. Calculating the molar mass M Ũ from eqn (5) requires solving a cubic equation and choosing the appropriate solution.…”

“…We recommend selecting the reference component in a way that D N ref can be adopted from the literature. As an alternative, it can also be determined experimentally, which is, however, usually tedious, 60 or estimated using a prediction method. 20,21,[61][62][63] From D N U ˜, in turn, the molar mass M U ˜of component U ˜can be calculated using basically any predictive model for selfdiffusion coefficients at infinite dilution.…”

AI in Process Industries – Current Status and Future Prospects

“…For basic and detail engineering, the KEEN project explored various ways, in which artificial intelligence can support engineering and process development and help to minimize errors in the early engineering phases. An example is process design based on machine learning and process flow diagrams (PFD) together with consistency checks in piping and instrumentation diagrams (P&ID) using deep learning [22]. The combination of the model with optimization and its connection to a process simulator showed enormous potential in a feasibility study.…”

Predicting Temperature‐Dependent Activity Coefficients at Infinite Dilution Using Tensor Completion

“…These matrices are usually only sparsely filled with experimental data. Since their first application in this field in 2020 [15], MCMs have been used very successfully to predict different thermodynamic mixture properties, including activity coefficients [15][16][17][18], Henry's law constants [19], and self-diffusion coefficients [20]. These MCMs have outperformed physical baseline methods for predicting the corresponding properties.…”