I am studying about adsorption of gases on some specific compounds. I want to analyze the surface chemistry, change in gibbs free energy behind the experimental results. Can i do such calculations using some software ?
I don't think that quantum chemistry software such a gaussian is a solution. Those are designed to model processes such as steps of chemical reactions, and evaluation of bonds energies. Modelling physisorption and all electrons would be VERY expensive here, and with bad accuracy for the low energies of physical adsorption.
I would go for methods which are specific for physical adsorption, such as GCMC. Perhaps you can try the software RASPA in which you can model the whole isotherm and obtain adsorption energies easily.
Are you talking about the experiments????if you need to calculate the binding energy of the adsorbed molecules (gas) on to a specific adsorbent (homogeneous or heterogeneous), you can calculate it from their adsorption isotherms. We developed series of binding site energy distribution functions. Just fit the data in Sips, Toth, Freundlich and then convert the parameters to binding site energy distribution....
if it is homogeneous surface then just fit the isotherm to langmuir isotherm..The KL value in Langmuir constant will be the binding energy of your material...
If you are talking about the theoretical modeling:
Then you have to rely on GCMC calculations where you can calcualte the isosteric heat which will be the binding energy. In GCMC you can study the adsorption of only one gas molecule on to adsorbent and then u can calcualte the adsorption isosteric heat at infinite dilution..you can use DL_POLY, DL_MONTE, MUSIC or RASPA....its your choice...
Thanx alot for your answers.Actually i was looking for some simulation software so that i could use those theoretical data to support my experimental results .I will look forward to GCMC calculations and RASPA. Since, i dont have a hand in programming and i am a beginner in this domain so its gonna take a while....
if u r new to this area, then begin with MUSIC or DL_MONTE, they are more easier than RASPA...DL_MONTE is very matured project and you will have lots of user support and technical support...
If you do not have much time to spend, then i would recommend MUSIC as its more easy (still you need couple of months to understand and possible another year to get familiarized)...
Prof. Tina Duren made a simple but complete tutorial in this page:
Currently i am writing my own tutorial for MUSIC as a way to thank Prof Randall Snurr who made these codes free to researchers like us.Soon i will upload in my uni page..if i did i will write you here to check the link....
To know the basics of forcefields(FF) use the tutorial of GULP..they explain in detail about FF so a beginner can understand...Good luck with your research.
NOTE: Also pay some attention to the materials u are using...if its amorphous materials then you can model it globally as still there are no universal model to treat the adsorbents, if it is crystalline then everything is fine....
References
==========
Below i list some of our papers where we developed site energy distribution functions, which you can use to characterize the bindig energy of the material from adsorption isotherms (these wokrs are not cited as we expected, but still they are useful and you can avoid using expensive inverse gas chromatography experiments to do this job),
Figuring out what software to use to model surface adsorption depends heavily on how you plan to model the surface. Consider the adsorption reaction:
A + S ---> A/S
Here A is the gas-phase species being adsorbed. S is the bare surface and A/S is the surface with the adsorbate bound to it. The energy change for the reaction is expressed as:
E(A/S) - E(A) - E(S)
Here E(X) is the energy of species X. There are a number of techniques that could be used to compute E(X), but it is critical that the same method be used for all three species. (In such calculations, one typically assumes DeltaE = DeltaH, which amounts to neglecting P-V work. The above also neglects entropic contributions. One approach to approximate DeltaS is by computing the vibrational DOS for each species, and treating A as an ideal gas using the Sackur-Tetrode equation.)
Referencing the above, your basic question could be restated: What code should I use to compute E(X)? One of the most important questions considerations is how the surface will be modeled. There are two basic approaches: 1) Model the surface as unit cell of an infinitely repeating slab using periodic boundary conditions. 2) Model the surface as a large cluster that exposes the surface of interest. Approach #1 is generally accepted to give a more accurate model of a true surface as there are no unphysical edge effects and the surface material will possess the band structure of the corresponding solid. If approach #2 is used one has to make sure the cluster is sufficiently large that the adsorbate feels no edge effects. In practice this is often very difficult to achieve. Even if the adsorbate can be kept adequately far from an adjacent edge, the problem remains of dangling bonds at the edges of the slab influencing the electronic structure. Once can often achieve a closed-shell system by judiciously cleaving the cluster and terminating dangling bonds with -H, but often at the cost of using the wrong stoichiometry.
To follow approach #1, one needs a code with periodic boundary conditions implemented. There are quite a few choices. VASP, CASTEP, ABINIT, Dmol, ADF, CRYSTAL... are commonly used. Some use pseudopotentals to model the core electrons and treat only the valance electrons explicitly, others are all-electron codes. The latter approach being computational more demanding. Of those that use pseudopotentials, some use a plane wave basis, others an atom centered basis. These are considerations in selecting the code to use.
Despite its shortcomings, approach #2 is very common. I suspect the reasons for this are twofold. First, many computational people come from a molecular quantum chemistry background, so they are comfortable with codes that employ molecular boundary conditions. Second, approach #2 opens up a much larger selection of codes.