Getting started

1.High-throughput adsorption modeling

(1) Prepare the input files

config.yaml should be prepared in your working folder.

Config.yaml file Contains two parts:

• StuctInfo parts

• Species parts (optional)

• Model parts

The format of the whole config.yaml is as follows:

StrucInfo:
  file: POSCARfile
  struct:
    element: Au
    lattice_type: fcc
    latttice_constant: 4.16
    facet: ['111','100'] #This is comment
    supercell: [3,3]     #Default: [3,3], can be undefined
    layers: 4            #Default: 4, can be undefined
    dope:
      Cu: [3]

# This part alias for Species
Species:
  file:
    'NH3+': NH3+.xyz #struct_file
    'NH3-': NH3-.xyz #struct_file
  sml:
    'O2': O=O #smiles

Model:
  ads:
    - ['NH3+',1]
    - ['NO', 2]
    - [s: 'C(=O)O',1]
    - [f: 'NH3+',1]
    - [s: 'C[Al](C)C',1Al]
  coads:
    - ['NH3','O',1,1]

StuctInfo parts contains the information about substrate, the keywords should be chosen as following two types: file or struct

• file: Path , this keywords will read substrate from POSCAR file

• struct: this keywords will generate substrate according to following parameter:

  • element: bulk phase element

  • lattice_type: lattice type

  • latticd_constant: lattice parameter

  • facet: crystal plane, should be a list, start with [, separate with , and end with ], facet index should be a str start with ' end with ', like '100', '111'

  • supercell: supercell of the substrate in the xy-plane

  • layers: layers of the substrate in z axis

  • dope: dope of the substrate, the formate is dope element : [dope type1, dope type2] before : is the doped element, after the : is the dope type, the dope type can be chosen as follows: 0 corresponds to no doping, 1, 2 and 3 correspond to surface layers doped with 1, 2 and 3 atoms, respectively. 1L and b1 represent surface layer substitution and bulk equivalent proportional substitution.

Species part is an optional part, which used to alias the species name, including:

• sml: using Smiles toe represent the adsorption species

• file: adsorption species reading from structure file, such as xyz format and vasp format

Model part contains the adsorption modeling parameters, including:

• ads: using - [ adsorbate formular , adsorption sites type] to represent one adsorption status, where adsorption formular should be a str start with ' end with '. Different adsorption status should start with new line, adsorption sites type can be chosen from 1 or 2.

• coads: using - [ads1, ads2, ads1 sites, ads2 sites], ads1 and ads2 is two adsorbate species formular, ads1 sites and ads2 sites is adsorption sites type, can be chosen from 1 and 2.

• Each adsorbate species formular has three representation:

  • str format such as 'ads1', this format will automatically generate species according to chemical formular

  • smile format such as s: 'ads1' or sml: 'ads1', this format will generate spcecies according to the smile formular

  • file format such as f: 'ads1' or file: 'ads1', this format will read structure from file ‘ads1’, now only ‘xyz’ and ‘vasp’ structure file format is supported

• New in version 1.0.5: Adding the symbol of the element contained in the adsorbate followed by the type of adsorption site generates a configuration in which all atoms corresponding to that element are adsorbed atoms, and automatically adjusts the orientation so that the adsorbed atom is at the bottom. Such as - [s: 'C[Al](C)C',1Al] will generatte configurations with Al atoms as adsorbed stoms

To avoid ambiguity, it is recommended that SMILES be used when declaring complex species. Users can modify the corresponding parameters to achieve customized modeling according to their research needs.

(2) Run script

Running command htmat ads can automate the enumeration and construction of all possible configurations, and ultimately output structure files in the VASP format, like Au_Cu_111_1_NH_0.vasp, Furthermore, for the adsorption models on doped surfaces, we use the get_binding_adatom function to determine the types of surface atoms that the adsorbate binds to in the preliminary configurations. We only select configurations where the adsorbate is bound to a surface atom that contains a dopant atom, in order to reduce the number of final output structures.

(3) Completion of the input file

Run the command htmat complete, it will add the default parameters according to the current config.yaml and output the complete_config.json file.

2.Automated construction of reaction transition state calculation process

3.Automated extraction of calculation results

4.Automated extraction of descriptors