roadmap

[longemen3000]

• Isotherms: every adsorption isotherm under the sun, interpolators
• Arbitrary number support: ForwardDiff.Dual, Measurement, BigFloat, etc
• Methods: IAST,reverse IAST, Real Adsorbed Solution Theory (RAST), inverse RAST, etc

[pw0908]

I do wonder if we could have an extension with cDFT since it too will be able to handle adsorption isotherms?

[longemen3000]

Yeah, as I see it there is a continuum between IAST-like methods to DFT in the axis empirical <-> theoretical

[longemen3000]

One example: we can use DFT to create adsorption isotherms in a fully theoretical way, then fit an isotherm shape and use the reduced form to calculate equilibria across a range of concentrations.

RAST also requires an activity coefficient dependent on the grand potential, and you can get such dependence by fitting the DFT isotherm with a modified activity coefficient from Clapeyron, multiplied by f(Π)

[longemen3000]

related: doi.org/10.1021/acs.langmuir.9b02378

[viniviena]

Hey all. I was checking the model types and wondering that having the K as part of the model type could cause some trouble in the future. K is typically a function of temperature T in a exponential dependency with the adsorption energy like $K = K(T_{ref})\exp(-\Delta H/RT)$. Maybe splitting K in two would be better - $K_ref$ and $\Delta H$?

The methods could also include the temperature as an argument.

[longemen3000]

Seems like a good idea

[viniviena]

Cool. I will do it and make a PR

[viniviena]

I just found this package that is general suite for adsorption thermodynamics in Python. It seems to do the same things as PyIAST but they named it as a SUITE for adsorption, which makes more sense.

https://pygaps.readthedocs.io/en/master/index.html

[longemen3000]

there is also RUPTURA:
https://github.com/iRASPA/RUPTURA