Introduction to materials informatics
Hi, that's me 🤓

Outline


  • data & formats
  • software & simulations
  • collaboration & sharing


Open the Python notebooks meanwhile.

From nanometers to meters


multiscale modeling bridge multiscale modeling lengths

Software is eating the world


Crystallographic data formats 💎


POSCAR 

https://mpds.io/entry/S251615
1.0000000000000000
    0.0000   2.0245   2.0245
    2.0245   0.0000   2.0245
    2.0245   2.0245   0.0000
  Al
   1
Direct
  0.000000  0.000000  0.000000
CIF 

data_mpds_labs
_cell_length_a    15.017668
_cell_length_b    30.035337
_cell_length_c    30.035337
_cell_angle_alpha 118.543768
_cell_angle_beta  118.543768
_cell_angle_gamma 92.541335
_symmetry_space_group_name_H-M 'P1'
_symmetry_Int_Tables_number 1
loop_
 _symmetry_equiv_pos_as_xyz
 +x,+y,+z
loop_
 _atom_site_type_symbol
 _atom_site_fract_x
 _atom_site_fract_y
 _atom_site_fract_z
 Ca   0.462 -0.342 -0.179
 Be   0.362 -0.274 -0.420
...
Optimade 

{
   "attributes":{
      "immutable_id":42,
      "species":[
         {
            "chemical_symbols":[
               "Au"
            ]
         }
      ],
      "cartesian_site_positions":[
         [0,0,0]
      ],
      "lattice_vectors":[
         [0,2,2],
         [2,0,2],
         [2,2,0]
      ]
   }
}

Python 🐍 & ASE 💎



from ase.spacegroup import crystal

crystal_obj = crystal(
    ('Al', 'Y', 'Cu', 'Cu'),
    basis=[
        ( 0.235, 0.0, 0.5 ),
        ( 0.5861, 0.0, 0.0 ),
        ( 0.333, 0.666, 0.5 ),
        ( 0.0, 0.0, 0.0 )
    ],
    spacegroup=189,
    cellpar=[7.033, 7.033, 4.023, 90, 90, 120],
    primitive_cell=True
)
Have a look at the ASE website. Then head over to the Python notebooks you have opened earlier.

MPDS


MPDS: Materials Platform for Data Science phase diagrams crystal structures physical properties published literature MPDS platform

Phase diagrams, physical properties, and crystal structures are interlinked from the published literature. Just hover and click the colored circles at the logo.

Wow, many online databases?

MPDS, Materials Project, Aflowlib, OQMD, COD, Nomad

Let's link them all together 🤝

Optimade logo

Optimade queries



filter=elements HAS ALL "Al","Mg","Sc","Mn","O"
filter=chemical_formula_reduced="NaCl"
filter=elements HAS ANY "Np","Cf"
    AND elements HAS "O"
    AND nelements=2
filter=chemical_formula_anonymous="A2B"
    AND chemical_formula_anonymous="AB2"
filter=_mp_bandgap > 5.0
filter=chemical_formula_descriptive IS KNOWN
Let's play with them at the Optimade.Science aggregator. Then head over to the Python notebooks you have opened earlier.

Our community 👫


Have your online profiles ready 🎓


Conclusions 👌


  • materials informatics is just about how to handle materials data efficiently
  • Python + ASE can do any routine tasks for you
  • POSCAR & CIF are the most common formats for crystal structures, but the Optimade is coming
  • the modern materials simulations can be prepared and even conducted fully online.

PS