How to calculate thermodynamic properties with CASTEP?
Article High-pressure thermo-elastic properties of beryl (Al4Be6Si12...
Kervin...you are asking about a plane wave DFT code. Any such code will compute an electronic (0K) energy. It is up to you as to what properties (e.g. thermodynamic, kinetc, vibrational, mechanical) you wish to compute.First and foremost, you have to make sure that your energy is fully converged. Have a look at the following publications for some examples on how to apply DFT to compute these various properties. Note that there are many other people who do these calculations and much has been done on various systems ranging in complexity. My suggestions are just to get you going.
S. Shi, P. Lu, Z. Liu, Y. Qi, Yue, L.G. Hector, Jr., S. Harris, ”Direct Calculation of Li-ion Transport in the Solid Electrolyte Interphase,” J. Amer. Chem. Soc. 134 2012) 15476-15487.
G.P.M. Leyson, W.A. Curtin, L.G. Hector, Jr., C. Woodward, “Quantitative Prediction of Solute Strengthening of Aluminum,” Nature Materials 9 (2010) 750-755.
S. Ganeshan, L.G. Hector, Jr., Z.-K. Liu,” First-principles Calculations of Impurity Diffusion Coefficients in Dilute Mg Alloys using the 8-Frequency Model,” Acta Materialia, 59 (2011) 3214-3228.
L.G. Hector, Jr., J.F. Herbst, W. Wolf, P. Saxe, G. Kresse, “Ab Initio Thermodynamic and Elastic Properties of Alkaline Earth Metals and Their Hydrides,” Phys. Rev. B 76 (2007) 014121,1-18.
L. Cui, Y. Tan, H. Zhang, L.G. Hector, Jr., S. Shi, C. Ouyang, H. Li, L. Chen, "First-principles Investigation of Atomic and Electronic Structure of M (M=Cu, Ag, Au) Adsorption on CeO2(110)," Phys. Chem. Chem. Phys. 14 (2012) 1923-1933.
As has been pointed out by Louis Hector, this tool is used for performing electronic structure calculations on periodic solids. To be able to apply statistical theories for thermodynamic quantities, one must do molecular dynamics. It is possible to use ab initio calculations to replace approximate force fields in MD. The programs that have this ability include Quantum Espresso. I do not know whether CASTEP has been incorporated into an MD code in this fashion.
To respond to Savita, CASTEP can indeed perform ab initio MD simulations.
It is not strictly necessary to run MD to compute thermodynamic properties. If you request a phonon calculation then CASTEP will compute the vibrational contributions to the partion function and get H and G as function of temperature.
Do you think the phonon calculations at various temperatures will give more accurate values than MD? What about the computation time relative to an MD run?
The thing about phonon calculations is that when they are done I have the "exact" vibrational modes within the limit of the harmonic approximation, so it is a well-defined and reproducible mode. With MD, you never know exactly how many steps you'll need and whether you've sampled the entire phase space. I do not have first hand experience, but I believe the phonons will be faster.
In CASTEP, after you have optimised the unit cell, you must calculate the phonons and, after that, you can obtain your desired thermodynamic properties.
I used to run phonon in CASTEP, they are run very long time, how will I save the time?
Phonon calculation with CASTEP is really a time-consuming work!! But the thermodynamic (especially for heat capacity as I known) from Phonon within the limit of the harmonic approximation is relative reliable.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Thermodynamics