This directory contains files with parameters for PC(P)-SAFT (including gc-PC-SAFT). The files named according to the pattern NameYear.json correspond to published parameters. The corresponding publication is provided in the literature.bib file.

For most applications, we recommend using the recent parametrization by Esper et al. together with binary interaction parameters from Rehner et al. In Python reading parameters from the two JSON files is done by

from feos import Parameters

params = Parameters.from_json(
    ["acetone", "hexane"], 
    "esper2023.json", 
    "rehner2023_binary.json"
)

For mixtures that lack binary interaction parameters in Rehner et al., the predicted interaction parameters by Hemprich et al. might be a good starting point. Because of the number of mixtures (> 1 M) they are not stored as JSON file, but rather as sqlite database. The file can be downloaded from the repository accompanying the publication.

from feos import Parameters

params = Parameters.from_database(
    ["acetone", "hexane"], 
    "hemprich2026.db"
)

For some applications (e.g., aqueous systems) more specialized parametrizations are required. See below for a full list of PC(P)-SAFT pure-component parameter sets available in the repository.

Parameters can also be constructed from group-contribution methods. In Python only and if you have rdkit installed in your environment, you can generate parameters directly from a SMILES code:

GcParameters.from_json_smiles(
    ["CCC(C)=O"], 
    "sauer2014_smarts.json", 
    "sauer2014_homo.json"
)

or

GcParameters.from_json_smiles(
    [Identifier(name="2-butanone", smiles="CCC(C)=O")], 
    "sauer2014_smarts.json", 
    "sauer2014_homo.json"
)

The rules that are applied in the determination of the group counts from SMILES are defined by SMARTS. All GC models that are implemented currently are compatible with the SMARTS defined in sauer2014_smarts.json.

For a more detailed description of parameter handling in FeOs, check out the example notebook.