Skip to content

Latest commit

 

History

History
73 lines (68 loc) · 3.46 KB

band_align.md

File metadata and controls

73 lines (68 loc) · 3.46 KB
Raw

Band alignment

Quick Start

from ectoolkits..analysis.band_align import BandAlign
inp = {
     "input_type": "cube",
     "ave_param":{
         "prefix": "./00.interface/hartree/Hematite-v_hartree-1_",
         "index": (1, 502),
         "l1": 4.8,
         "l2": 4.8,
         "ncov": 2,
         "save": True,
         "save_path":"00.interface"
     },
     "shift_param":{
         "surf1_idx": [124, 125, 126, 127, 128, 129, 130, 131],
         "surf2_idx": [24, 25, 26, 27, 28, 29, 30, 31]
     },
     "water_width_list": [8, 9, 9.5, 10, 10.5, 11, 12, 13],
     "solid_width_list": [1, 2, 3, 4]

}
ba = BandAlign(inp)
#quick view of hartree fluctuation
fig = ba.plot_hartree_per_width('water')
fig = ba.plot_hartree_per_width('solid')

# detail information is accessible in
ba.water_hartree_list
ba.solid_hartree_list

Prepare Input Data

One has to collect all hartree cube files from continuous MD simulations in one directory with same prefix and suffix of .cube. For example, in directory ./00.interface/hartree/, one should prepare cube file named Hematite-v_hartree-1_1.cube, Hematite-v_hartree-1_2.cube, ..., Hematite-v_hartree-1_501.cube.

Explanation for Parameters

density

  • input_type: cube for cp2k cube files
  • prefix: define prefix for cube files
  • index: define the index for cube files, see Prepare Input Data
  • l1: parameter for nanosmoothing. l1 equals to natural number (n) times periodic length (pl) for oxides slab. For example, the z position of a layer for $\mathrm{SnO_2}$(110) slab is defined as the ensemble-averaged z positions of $\mathrm{Sn}$ atoms in the layer. Then, the periodic length is inter-layer distance between two layers. Assuming pl is 2.4 $\mathrm{\AA}$, l1 can be 2.4, 4.8, and 7.2 $\mathrm{\AA}$.
  • l2: parameter for nanosmoothing. l2 equals to natural number (n) times periodic length (pl) for water. Since water dose not have periodicity, we recommend to set l2 same as l1.
  • ncov: parameter for nanosmoothing. ncov is number of convolution. ncov equals to 2 for interfaces and equals to 1 for surface (slab-vacuum). When ncov is 1, l2 won't be used.
  • save: whether to save post processed hartree data. We recommend to set it as True.
  • save_path: save_path for post processed hartree data.
  • surf1_idx: place holder
  • surf2_idx: place holder
  • water_width_list: Width for bulk water. Bulk Water Width is plotted in above figure. Since one has to test multiple width value, the width is input as a list.
  • solid_width_list: Width for Bulk Solid. Bulk Solid Width is plotted in above figure. Since one has to test multiple width value, the width is input as a list.

Plot Band Alignment Data

After Obtain band positions from post processing, you can plot these data using plot_band_alignment

from ectoolkits.plots.band_align import plot_band_alignment

#prepare your band positions data in dictionary format
ba_data = {
    "sys_1": {
        "vbm": 1.0,
        "cbm": -0.5
    },
    "sys_2": {
        "vbm": 1.2,
        "cbm": -0.3
    },
    ....
}
fig = plot_band_alignment(ba_dict=ba_data, show_diff=True, vac_value=False)
fig.savefig("bandalignment.png")

We assume the vbm and cbm values are in SHE scale. If the vbm and cbm values are referred to Vacuum scale, set vac_value=True. If you want to plot the difference between band position of first system and other systems, set show_diff=True. density