All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

• Added henry_coefficients and ideal_gas_enthalpy_of_adosrption to PoreProfile. #263

• Updated quantity dependency to 0.10. #262
• Updated num-dual dependency to 0.11. #262
• Updated numpy and PyO3 dependencies to 0.23. #262
• Updated gauss-quad dependency to 0.2. #261

• Updated interfaces according to removal of HelmholtzEnergyDual and HelmholtzEnergy in feos-core. #226
• Changed return values of HelmholtzEnergyFunctional::contributions from dyn FunctionalContribution to HelmholtzEnergyFunctional::Contribution. #226

• Removed FunctionalContributionDual trait. #226

• Updated quantity dependency to 0.8. #238
• Updated num-dual dependency to 0.9. #238
• Updated numpy and PyO3 dependencies to 0.21. #238

• Updated feos-core dependency to 0.6.

• Implemented HelmholtzEnergyFunctional for EquationOfState to be able to use functionals as equations of state. #158
• Added the possibility to specify the angles of non-orthorombic unit cells. Currently, the external potential must be specified if non-orthorombic unit cells are calculated. #176

• HelmholtzEnergyFunctional: added Components trait as trait bound and removed ideal_gas method. #158
• DFT<F> now implements Residual and furthermore IdealGas if F implements IdealGas. #158
• What properties (and contributions) of DFTProfile are available now depends on whether an ideal gas model is provided or not. #158
• Pore volumes and other integrated properties (loadings, energies, ...) are now always reported in units of volume. For translationally symmetric pores this is achieved by multiplying with the reference length (1 Å) in every direction with symmetries. #181

• Removed DefaultIdealGasContribution #158
• Removed getters for chemical_potential (for profiles) and molar_gibbs_energy (for Adsorption1D and Adsorption3D) from Python interface. #158

• Added an error, if the unit cells in 3D DFT are too small and violate the minimum image convention. #176

• Updated quantity dependency to 0.7.
• Updated num-dual dependency to 0.8. #137
• Updated numpy and PyO3 dependencies to 0.20.

• Added new methods drho_dmu, drho_dp and drho_dt that calculate partial derivatives of density profiles to every DFT profile. Also includes direct access to the integrated derivatives dn_dmu, dn_dp and dn_dt. #134
• Added enthalpy_of_adsoprtion and partial_molar_enthalpy_of_adsorption getters to pores and adsorption isotherms. #135

• PlanarInterface now has methods for the calculation of the 90-10 interface thickness (PlanarInterface::interfacial_thickness), interfacial enrichtment (PlanarInterface::interfacial_enrichment), and relative adsorption (PlanarInterface::relative_adsorption). #64
• Added a matrix-free Newton DFT solver that uses GMRES for the linear subproblem. #75
• Solver metrics like execution time and the residuals are automatically logged during every computation. The results can be accessed via the solver_log field of DFTProfile. #75

• Added Send and Sync as supertraits to HelmholtzEnergyFunctional and all related traits. #57
• Renamed the 3d_dft feature to rayon in accordance to the other feos crates. #62
• internally rewrote the implementation of the Euler-Lagrange equation to use a bulk density instead of the chemical potential as variable. #60
• Renamed the parameter beta of the Picard iteration and Anderson mixing solvers to damping_coefficient. #75
• Removed generics for units in all structs and traits in favor of static SI units. #115

• Fixed the sign of vector weighted densities in Cartesian and cylindrical geometries. The wrong sign had no effects on the Helmholtz energy and thus on density profiles. #120

• Updated pyo3 and numpy dependencies to 0.18. #119
• Updated quantity dependency to 0.6. #119
• Updated num-dual dependency to 0.6. #119

• The 3D DFT functionalities (3D pores, solvation free energy, free-energy-averaged potentials) are hidden behind the new 3d_dft feature. For the Python package, the feature is always enabled. #51

• Adsorption::equilibrium_isotherm, Adsorption::adsorption_isotherm, and Adsorption::desorption_isotherm only accept SIArray1s as input for the pressure discretization. #50
• For the calculation of desorption isotherms and the filled branches of equilibrium isotherms, the density profiles are initialized using a liquid bulk density. #50

• Added getters for the fields of Pore1D in Python. #30
• Added Steele potential with custom combining rules. #29

• Made FMT functional more flexible w.r.t. the shape of the weight functions. #31
• Changed interface of PairCorrelationFunction to facilitate the calculation of pair correlation functions in mixtures. #29

• Moved the implementation of fundamental measure theory to the feos crate. #27

• Added grand_potential_density getter for DFT profiles in Python. #22

• Renamed AxisGeometry to Geometry. #19
• Removed PyGeometry and PyFMTVersion in favor of a simpler implementation using PyO3's new #[pyclass] for fieldless enums feature. #19
• DFTSolver now uses Verbosity instead of a bool to control its output. #19
• SurfaceTensionDiagram now uses the new StateVec struct to access properties of the bulk phases. #19
• Pore1D::initialize and Pore3D::initialize now accept initial values for the density profiles as optional arguments. #24
• Internally restructured the DFT structure to avoid redundant data. #24
• Removed the m function in FluidParameters, it is instead inferred from HelmholtzEnergyFunctional which is now a supertrait of FluidParameters. #24
• Added optional field cutoff_radius to ExternalPotential::FreeEnergyAveraged. #25

• Updated pyo3 and numpy dependencies to 0.16.
• Updated quantity dependency to 0.5.
• Updated num-dual dependency to 0.5.
• Updated feos-core dependency to 0.2.
• Updated ang dependency to 0.6.
• Removed log dependency.

• The pore volume for Pore3D is now also accesible from Python. #16

• The pore volume using Helium at 298 K as reference is now directly accesible from Pore1D and Pore3D. #13

• Removed the unsendable tag from python classes wherever possible. #14

• HelmholtzEnergyFunctionals can now overwrite the ideal_gas method to provide a non-default ideal gas contribution that is accounted for in the calculation of the entropy, the internal energy and other properties. #10

• Removed the functional field in Pore1D and Pore3D. #9

• Fixed the units of default values for adsorption isotherms. #8

• Updated rustdct dependency to 0.7.

• Initial release