Really thanks for the reply. I didn't know that "Materials Studio" can do this. I tried Gaussian03, but it was crashed during the optimization.

The optimization was done by MS, but I want to demonstrate the charge density of molecule in a colorful pattern like this:

http://www.cosmologic.de/images/engineering/mol2.png

Actually, I want to estimate the activity coefficients of ionic liquids and the only software that I know are COSMOtherm or ADF COSMO-RS.

ADF has a trial version, but it only imports it's own file format plus ".cosmo" file.

I'm trying to test the output of Materials Studio in ADF COSMO-RS and see the result.

Dear Mehdi,

With ADF's COSMO-RS implementation you can also import sigma profiles generated with other programs. You can convert ASCII files to the binary format used by ADF's COSMO-RS module:

http://www.scm.com/Doc/Doc2013/COSMO-RS/page44.html

However, the parametrization is different for every COSMO-RS implementation, so it is recommended to generate the sigma profiles with ADF itself. This is easily done with the 'COSMO-RS compound' preset in the GUI which will set up the two necessary steps in one file: 1) optimization in the gas phase with the settings used for the parametrization, 2) single point COSMO calculation with infinite dielectric constant.

For an ionic liquid one would calculate the sigma profile of the cation and anion separately and make sure you would have a charge balance in the final COSMO-RS calculation. (I think people have also set up ionic liquids with COSMO-RS where the unit is a cation-anion pair).

Feel free to discuss these kind of questions with our support team or on the ADF mailing list.

Best wishes,

Fedor

Dear Fedor,

Thanks for your reply. It was very useful. I tried to optimize the cation and anion separately. Then imported them as two compounds in COSMO-RS module to calculate the activity coefficient.

I'm not sure how I should define the system. Should it be a 2-component system and mole fraction of cation and anion are defined as 0.5?

I did this procedure, but the result was from experimental value.

Thanks.

Mehdi

Dear Mehdi,

Perhaps you could send the outputs (or coskf files) to support to have a look at?

Best,

Fedor

Hi all,

I think my question is somehow related to the conversation going on here. Actually I am trying to create the cosmo file for H2S by Turbomole software. I am using B3LYP with whatever basis set that exist in the software but the result I get in the COSMOtherm in way off. Has any one had the experience of working with H2S? I appreciate it if any of you guys could help me.

Regards,

Hi Reza,

COSMO-RS has been parametrized with the BP86 functional and a prescribed basis set. So for consistency it is advisable to use the same set up.

Do you get reasonable results when you use the set procedure for generating the sigma profile?

-Fedor

Dear Mehdi,

Just to let you know that, with the help of Prof. Zhigang Lei's research group we now have a tutorial that explains how to use the 2-component procedure for ionic liquids with COSMO-RS (and also included an ionic liquid database).

http://www.scm.com/Doc/Doc2014/GUI/GUI_tutorial/page194.html

Dear Fedor,

Thanks for updating. It is what I was looking for, especially this statement:

the calculated activity coefficients have to be scaled with 0.5.

In the activity coefficient page, I think there is a typological error in the first equation.

γibin = (γibin xitern)/xibin

should be

γibin = (γitern xitern)/xibin

Regards,

Mehdi

Thanks, yes we should change that!

I wish to predict solubility of astaxanthin in mixtures of acetone and DMSO using UNIFAC software. How can i identify the functional groups of astaxanthin.

Structure of astaxanthin can be found in following link-- https://en.wikipedia.org/wiki/Astaxanthin

Can anybody help me in optimizing a ionic liquid from structure till properties calculation in ADF-COSMO RS? As i found no resources and procedure to do that. I would be highly obliged if you guys could help

Mehdi Sattari Can you please help me in this regard? I am facing the same issue that you posted here in 2014

For optimizing a single ion for use with COSMO-RS, just input your molecule (e.g. SMILES, or build from scratch), set the charge, and use the standard COSMO-RS preset: https://www.scm.com/doc/Tutorials/COSMO-RS/COSMO_result_files.html#cosmo-result-files

Here's an example of building an anion: https://www.scm.com/doc/Tutorials/Analysis/FragmentAnalysis.html#ptcl4-h2-2

There are different COSMO-RS implementations. COSMOtherm and ADF COSMO-RS are two of them. There are also open source versions sometimes also named COSMO-SAC. Some versions might be found here:

https://github.com/TUHH-TVT/ (I am one of the developers)

https://github.com/lvpp/pysac

https://github.com/usnistgov/COSMOSAC

The implementations vary in their details, but all of them rely on using a quantum chemical calculation from a specific method. T. P M Goumans has already mentioned a few years back that a specific method needs to be applied in the quantum chemical calculation to match the implementations. But you are not completely limited in the choice of quantum chemistry package as long as you can do a geometry optimization in a dielectric continuum with infinite permittivity (i.e. ideal conductor) and can export the geometry, cavity and the screening charge density. There are several quantum chemistry packages out there that allow this and are completely free or free for academia. But remember, that the COSMO-RS implementation needs to be tuned to the specific package/method.

Specifically for ions the choice of quantum chemistry might be important as some packages implement or do not implement outlying charge correction. Which should in principle be most important for anions and in general molecules of which the charge is expected to be outside of the cavity constructed by the quantum chemistry package,

In principle, we should distinguish between (i) the geometry optimization and (ii) the final single-point calculation at the optimized geometry which produces the screening charges etc. needed for the sigma-profile. While in (ii) you should use the QM method consistent with your specific COSMO-RS implementation (e.g., COSMO solvation model + DFT/BP86/PVTZ for AMS COSMO-RS), you are not limited at all with respect to (i). You can optimize the geometry in the vacuum or solvent environment, you can employ cheaper molecular mechanics tools, or even take the geometry directly from e.g. a crystal structure.

The answer whether to use the whole ionic liquid pair or the cation and anion separately in the optimization depends on which approach within COSMO-RS you prefer. It is possible to describe the cation-anion electroneutral pair as a whole or to model the counterions separetely ("ion-based approach"). Here is our benchmark for both of these approaches within ADF and also the method how we treated the geometry optimization:Article Predicting the Thermodynamics of Ionic Liquids: What to Expe...