nanocatalysis/basin-hopping

Implementation of a Basin Hopping Monte Carlo (BH) global optimization algorithm for supported metal nanoalloys. Works in conjunction with Quantum Espresso.

Basin-Hopping

Implementation of a Basin Hopping Monte Carlo (BH) global optimization algorithm for supported metal nanoalloys. BH algorithm has been implemented with Python3.4, coupled to Quantum Espresso 5.2 (DFT code as calculator).

Requirements

Quantum Espresso needs to be installed before running the BH programs. Quantum Espresso 5.2 and later are supported. To download and install, please refer to the following links:

Download Quantum Espresso

User guide for Quantum Espresso

In addition, Python 3.4 or later must be installed to run this project.

Installation

To "install" this project simply download the files and put them in a prefered directory.

It is highly recommended to preserve the project file structure as it is, since it makes things much easier.

Usage

For simplicity, the usage of this project has been simplified to adding specifications into an input file "input.bh". This file has different categories for the user to fill in with data regarding the problem to solve. These are:

1. Control: This category deals with specifying the paths to both the QE input directory and the auxiliary Python programs directory.

    &control
      input_dir = /home/path/to/input
      programs_dir = /home/path/to/programs
   

2. Basin Hopping: This category deals with specifics of the problem to solve.

    &basin_hopping
      substrate_nat = //number of atoms of the substrate (in case there is any). Eg. 40
      cluster_ntyp = //Monometallic: "[AtomicSymbol:NumberOfAtoms]". Eg. [Pt:5]
                     //Bimetallic: "[AtomicSymbol_1st:NumberOfAtoms_1st], [AtomicSymbol_2nd:NumberOfAtoms_2nd]". Eg. [Pt:5], [Au:2]
      temperature_K = //Temperature in Kelvin. Eg. 500
      step_width = //Step width tu use for the basin hopping algorithm. Eg. 0.6
      iterations = //Total number of iterations, or steps in the basin hopping execution. Eg. 100
   

3. Random range: This category deals with specifying the 3D space available for generating the initial randomized cluster.

    &random_range
      x_range = //[xMin:xMax]
      y_range = //[yMin:yMax]
      z_range = //[zMin:zMax]
      z_vacuum = //[VaccumStart:VaccumEnd]
   

The penultimate option has to do with anchoring the generated structure within a certain distance from the substrate. The other three define space constraints of the cell. Note: the value of "zMax" has to coincide with "VaccumStart" in most cases.

To run the project, execute the main program as: python 3.4 basin_hopping.py input.bh

Other than the input.bh file, the other files that need specification are the QE input file input.in and the QE run script run.sh (within the input directory). These are the usual QE run files.

Credits

Author : Andres Lopez Martinez

Advisor : Dr. Oliver Paz Borbon (IF-UNAM)

Financial Support (PAPIIT-UNAM): Project IA102716

Computational resources (Miztli-UNAM): SC15-1-IG-82

SC16-1-IG-78

Project Contact: Dr. Oliver Paz Borbon [email protected]

License

Unless otherwise noted, this work is released under: http://www.apache.org/licenses/LICENSE-2.0