The full name of the PASP package is Property analysis and simulation package for materials (material property analysis and simulation package), also known as Latticemodel (lattice model). It was developed by the research group of Professor Xiang Hongjun of Fudan University. Phys. 154, 114103 (2021)]. The package currently integrates many functions, such as symmetry analysis, effective Hamiltonian methods, Monte Carlo simulations, machine learning methods, global structure search methods, etc. By combining this package with first principles computing software such as VASP, we can easily simulate and study various physical properties of crystalline materials, as well as study the microscopic mechanisms of various physical properties from the perspective of interaction between local modes.


In magnetic, ferroelectric, multiferroic and other types of materials, it involves degrees of freedom such as spin, localized phonon mode, and lattice strain; in alloy materials, it involves degrees of freedom of element species; when crystalline materials contain organic molecules and may also involve degrees of freedom such as orientation and displacement of organic molecules; in addition, degrees of freedom such as charge amount and orbital order may also need to be considered. In the effective Hamiltonian method of PASP, we describe the various possible degrees of freedom mentioned above with different types of local modes, and then we can analyze the various possible unequal interactions between the different local modes; we can also use The MLMCH (full name Machine Learning Method for Constructing realistic effective Hamiltonian) method screens out the important interactions and constructs a simplified effective Hamiltonian; then the effective Hamiltonian can be used to perform Monte Carlo simulations (including parallel annealing Monte Carlo simulations) , the English abbreviation is PTMC) or molecular dynamics simulation to study the physical properties such as the ground state and phase transition temperature of the system.


In addition, neural network potential methods are also supported in PASP's effective Hamiltonian methods (including Monte Carlo simulations), which can be used to deal with more complex interaction forms; tight-binding models in the PASP package can be used to study energy Bands, magnetic interaction parameters, magnetoelectric coupling parameters and other properties; in addition, combining PASP's global structure search method with first-principles calculations, the ground-state structure of complex systems can be predicted.


References:

F. Lou, X. Y. Li, J. Y. Ji, H. Yu, J. S. Feng, X. G. Gong, and H. J. Xiang*, J. Chem. Phys. 154, 114103 (2021).
Kai Liu, Jinlian Lu, S. Picozzi, L. Bellaiche*, and H. J. Xiang*, Phys. Rev. Lett. 121, 027601 (2018)
P. S. Wang, W. Ren, L. Bellaiche, and H. J. Xiang*, Phys. Rev. Lett. 114, 147204 (2015)
H. J. Xiang*, Bing Huang, Erjun Kan, Su-Huai Wei, and X. G. Gong, Phys. Rev. Lett. 110, 118702 (2013)
H. J. Xiang*, E. J. Kan, Y. Zhang, M.-H. Whangbo*, and X. G. Gong*,Phys. Rev. Lett. 107, 157202 (2011)