# pymatgen.core.surface module¶

class ReconstructionGenerator(initial_structure, min_slab_size, min_vacuum_size, reconstruction_name, termination=0)[source]

Bases: object

This class takes in a pre-defined dictionary specifying the parameters need to build a reconstructed slab such as the SlabGenerator parameters, transformation matrix, sites to remove/add and slab/vacuum size. It will then use the formatted instructions provided by the dictionary to build the desired reconstructed slab from the initial structure.

slabgen_params

Parameters for the SlabGenerator

TODO: - Right now there is no way to specify what atom is being

added. In the future, use basis sets?
Generates reconstructed slabs from a set of instructions
specified by a dictionary or json file.
build_slab()[source]
Builds the reconstructed slab by:
1. Obtaining the unreconstructed slab using the specified parameters for the SlabGenerator.
2. Applying the appropriate lattice transformation in the a and b lattice vectors.
3. Remove any specified sites from both surfaces.
4. Add any specified sites to both surfaces.
Returns: The reconstructed slab. (Slab)
get_d()[source]

Determine the distance of space between each layer of atoms along c

get_unreconstructed_slab()[source]

Generates the unreconstructed or pristine super slab.

symmetrically_add_atom(slab, point)[source]
Class method for adding a site at a specified point in a slab.
Will add the corresponding site on the other side of the slab to maintain equivalent surfaces.
Arg:

slab (Slab): The slab to modify point (frac coord): The fractional coordinate of the site

Returns: The modified slab (Slab)
symmetrically_remove_atom(slab, point)[source]
Class method for removing a site at a specified point in a slab.
Will remove the corresponding site on the other side of the slab to maintain equivalent surfaces.
Arg:

slab (Slab): The slab to modify point (frac coord): The fractional coordinate of the site

in the slab to remove.
Returns: The modified slab (Slab)
class Slab(lattice, species, coords, miller_index, oriented_unit_cell, shift, scale_factor, reorient_lattice=True, validate_proximity=False, to_unit_cell=False, reconstruction=None, coords_are_cartesian=False, site_properties=None, energy=None)[source]

Subclass of Structure representing a Slab. Implements additional attributes pertaining to slabs, but the init method does not actually implement any algorithm that creates a slab. This is a DUMMY class who’s init method only holds information about the slab. Also has additional methods that returns other information about a slab such as the surface area, normal, and atom adsorption.

Note that all Slabs have the surface normal oriented in the c-direction. This means the lattice vectors a and b are in the surface plane and the c vector is out of the surface plane (though not necessary perpendicular to the surface.)

miller_index

Miller index of plane parallel to surface.

scale_factor

Final computed scale factor that brings the parent cell to the surface cell.

shift

The shift value in Angstrom that indicates how much this slab has been shifted.

Makes a Slab structure, a structure object with additional information and methods pertaining to slabs.

Parameters: lattice (Lattice/3x3 array) – The lattice, either as a pymatgen.core.lattice.Lattice or simply as any 2D array. Each row should correspond to a lattice vector. E.g., [[10,0,0], [20,10,0], [0,0,30]] specifies a lattice with lattice vectors [10,0,0], [20,10,0] and [0,0,30]. species ([Specie]) – Sequence of species on each site. Can take in flexible input, including: A sequence of element / specie specified either as string symbols, e.g. [“Li”, “Fe2+”, “P”, …] or atomic numbers, e.g., (3, 56, …) or actual Element or Specie objects. List of dict of elements/species and occupancies, e.g., [{“Fe” : 0.5, “Mn”:0.5}, …]. This allows the setup of disordered structures. coords (Nx3 array) – list of fractional/cartesian coordinates of each species. miller_index ([h, k, l]) – Miller index of plane parallel to surface. Note that this is referenced to the input structure. If you need this to be based on the conventional cell, you should supply the conventional structure. oriented_unit_cell (Structure) – The oriented_unit_cell from which this Slab is created (by scaling in the c-direction). shift (float) – The shift in the c-direction applied to get the termination. scale_factor (array) – scale_factor Final computed scale factor that brings the parent cell to the surface cell. reorient_lattice (bool) – reorients the lattice parameters such that the c direction is the third vector of the lattice matrix validate_proximity (bool) – Whether to check if there are sites that are less than 0.01 Ang apart. Defaults to False. reconstruction (str) – Type of reconstruction. Defaultst to None if the slab is not reconstructed. coords_are_cartesian (bool) – Set to True if you are providing coordinates in cartesian coordinates. Defaults to False. site_properties (dict) – Properties associated with the sites as a dict of sequences, e.g., {“magmom”:[5,5,5,5]}. The sequences have to be the same length as the atomic species and fractional_coords. Defaults to None for no properties. energy (float) – A value for the energy.
add_adsorbate_atom(indices, specie, distance)[source]

Gets the structure of single atom adsorption. slab structure from the Slab class(in [0, 0, 1])

Parameters: indices ([int]) – Indices of sites on which to put the absorbate. Absorbed atom will be displaced relative to the center of these sites. specie (Specie/Element/str) – adsorbed atom species distance (float) – between centers of the adsorbed atom and the given site in Angstroms.
as_dict()[source]

Dict representation of Structure.

Parameters: verbosity (int) – Verbosity level. Default of 1 includes both direct and cartesian coordinates for all sites, lattice parameters, etc. Useful for reading and for insertion into a database. Set to 0 for an extremely lightweight version that only includes sufficient information to reconstruct the object. fmt (str) – Specifies a format for the dict. Defaults to None, which is the default format used in pymatgen. Other options include “abivars”. **kwargs – Allow passing of other kwargs needed for certain e.g., "abivars". (formats,) – JSON serializable dict representation.
center_of_mass

Calculates the center of mass of the slab

copy(site_properties=None, sanitize=False)[source]

Convenience method to get a copy of the structure, with options to add site properties.

Parameters: site_properties (dict) – Properties to add or override. The properties are specified in the same way as the constructor, i.e., as a dict of the form {property: [values]}. The properties should be in the order of the original structure if you are performing sanitization. sanitize (bool) – If True, this method will return a sanitized structure. Sanitization performs a few things: (i) The sites are sorted by electronegativity, (ii) a LLL lattice reduction is carried out to obtain a relatively orthogonalized cell, (iii) all fractional coords for sites are mapped into the unit cell. A copy of the Structure, with optionally new site_properties and optionally sanitized.
dipole

Calculates the dipole of the Slab in the direction of the surface normal. Note that the Slab must be oxidation state-decorated for this to work properly. Otherwise, the Slab will always have a dipole of 0.

classmethod from_dict(d)[source]

Reconstitute a Structure object from a dict representation of Structure created using as_dict().

Parameters: d (dict) – Dict representation of structure. Structure object
get_orthogonal_c_slab()[source]

This method returns a Slab where the normal (c lattice vector) is “forced” to be exactly orthogonal to the surface a and b lattice vectors. Note that this breaks inherent symmetries in the slab. It should be pointed out that orthogonality is not required to get good surface energies, but it can be useful in cases where the slabs are subsequently used for postprocessing of some kind, e.g. generating GBs or interfaces.

get_sorted_structure(key=None, reverse=False)[source]

Get a sorted copy of the structure. The parameters have the same meaning as in list.sort. By default, sites are sorted by the electronegativity of the species. Note that Slab has to override this because of the different __init__ args.

Parameters: key – Specifies a function of one argument that is used to extract a comparison key from each list element: key=str.lower. The default value is None (compare the elements directly). reverse (bool) – If set to True, then the list elements are sorted as if each comparison were reversed.
get_surface_sites(tag=False)[source]

Returns the surface sites and their indices in a dictionary. The oriented unit cell of the slab will determine the coordination number of a typical site. We use VoronoiNN to determine the coordination number of bulk sites and slab sites. Due to the pathological error resulting from some surface sites in the VoronoiNN, we assume any site that has this error is a surface site as well. This will work for elemental systems only for now. Useful for analysis involving broken bonds and for finding adsorption sites.

Args:
tag (bool): Option to adds site attribute “is_surfsite” (bool) to
all sites of slab. Defaults to False
Returns:
A dictionary grouping sites on top and bottom of the slab together.
{“top”: [sites with indices], “bottom”: [sites with indices}
get_symmetric_site(point)[source]
This method uses symmetry operations to find equivalent sites on
both sides of the slab. Works mainly for slabs with Laue symmetry. This is useful for retaining the non-polar and symmetric properties of a slab when creating adsorbed structures or symmetric reconstructions.
Arg:
point: Fractional coordinate.
Returns: Fractional coordinate. A point equivalent to the parameter point, but on the other side of the slab point
get_tasker2_slabs(tol=0.01, same_species_only=True)[source]

Get a list of slabs that have been Tasker 2 corrected.

Parameters: tol (float) – Tolerance to determine if atoms are within same plane. This is a fractional tolerance, not an absolute one. same_species_only (bool) – If True, only that are of the exact same species as the atom at the outermost surface are considered for moving. Otherwise, all atoms regardless of species that is within tol are considered for moving. Default is True (usually the desired behavior). ([Slab]) List of tasker 2 corrected slabs.
have_equivalent_surfaces()[source]

Check if we have same number of equivalent sites on both surfaces. This is an alternative to checking Laue symmetry (is_symmetric()) if we want to ensure both surfaces in the slab are the same

is_polar(tol_dipole_per_unit_area=0.001)[source]

Checks whether the surface is polar by computing the dipole per unit area. Note that the Slab must be oxidation state-decorated for this to work properly. Otherwise, the Slab will always be non-polar.

Parameters: tol_dipole_per_unit_area (float) – A tolerance. If the dipole magnitude per unit area is less than this value, the Slab is considered non-polar. Defaults to 1e-3, which is usually pretty good. Normalized dipole per unit area is used as it is more reliable than using the total, which tends to be larger for slabs with larger surface areas.
is_symmetric(symprec=0.1)[source]

Checks if slab is symmetric, i.e., contains inversion symmetry.

Parameters: symprec (float) – Symmetry precision used for SpaceGroup analyzer. (bool) Whether slab contains inversion symmetry.
normal

Calculates the surface normal vector of the slab

surface_area

Calculates the surface area of the slab

class SlabGenerator(initial_structure, miller_index, min_slab_size, min_vacuum_size, lll_reduce=False, center_slab=False, in_unit_planes=False, primitive=True, max_normal_search=None, reorient_lattice=True)[source]

Bases: object

This class generates different slabs using shift values determined by where a unique termination can be found along with other criterias such as where a termination doesn’t break a polyhedral bond. The shift value then indicates where the slab layer will begin and terminate in the slab-vacuum system.

oriented_unit_cell

A unit cell of the parent structure with the miller index of plane parallel to surface

parent

Parent structure from which Slab was derived.

lll_reduce

Whether or not the slabs will be orthogonalized

center_slab

Whether or not the slabs will be centered between the vacuum layer

slab_scale_factor

Final computed scale factor that brings the parent cell to the surface cell.

miller_index

Miller index of plane parallel to surface.

min_slab_size

Minimum size in angstroms of layers containing atoms

min_vac_size

Minimize size in angstroms of layers containing vacuum

Calculates the slab scale factor and uses it to generate a unit cell of the initial structure that has been oriented by its miller index. Also stores the initial information needed later on to generate a slab.

Parameters: initial_structure (Structure) – Initial input structure. Note that to ensure that the miller indices correspond to usual crystallographic definitions, you should supply a conventional unit cell structure. miller_index ([h, k, l]) – Miller index of plane parallel to surface. Note that this is referenced to the input structure. If you need this to be based on the conventional cell, you should supply the conventional structure. min_slab_size (float) – In Angstroms or number of hkl planes min_vacuum_size (float) – In Angstroms or number of hkl planes lll_reduce (bool) – Whether to perform an LLL reduction on the eventual structure. center_slab (bool) – Whether to center the slab in the cell with equal vacuum spacing from the top and bottom. in_unit_planes (bool) – Whether to set min_slab_size and min_vac_size in units of hkl planes (True) or Angstrom (False/default). Setting in units of planes is useful for ensuring some slabs have a certain nlayer of atoms. e.g. for Cs (100), a 10 Ang slab will result in a slab with only 2 layer of atoms, whereas Fe (100) will have more layer of atoms. By using units of hkl planes instead, we ensure both slabs have the same number of atoms. The slab thickness will be in min_slab_size/math.ceil(self._proj_height/dhkl) multiples of oriented unit cells. primitive (bool) – Whether to reduce any generated slabs to a primitive cell (this does not mean the slab is generated from a primitive cell, it simply means that after slab generation, we attempt to find shorter lattice vectors, which lead to less surface area and smaller cells). max_normal_search (int) – If set to a positive integer, the code will conduct a search for a normal lattice vector that is as perpendicular to the surface as possible by considering multiples linear combinations of lattice vectors up to max_normal_search. This has no bearing on surface energies, but may be useful as a preliminary step to generating slabs for absorption and other sizes. It is typical that this will not be the smallest possible cell for simulation. Normality is not guaranteed, but the oriented cell will have the c vector as normal as possible (within the search range) to the surface. A value of up to the max absolute Miller index is usually sufficient. reorient_lattice (bool) – reorients the lattice parameters such that the c direction is the third vector of the lattice matrix
get_slab(shift=0, tol=0.1, energy=None)[source]

This method takes in shift value for the c lattice direction and generates a slab based on the given shift. You should rarely use this method. Instead, it is used by other generation algorithms to obtain all slabs.

Arg:
shift (float): A shift value in Angstrom that determines how much a
slab should be shifted.

tol (float): Tolerance to determine primitive cell. energy (float): An energy to assign to the slab.

Returns: (Slab) A Slab object with a particular shifted oriented unit cell.
get_slabs(bonds=None, tol=0.1, max_broken_bonds=0, symmetrize=False, repair=False)[source]

This method returns a list of slabs that are generated using the list of shift values from the method, _calculate_possible_shifts(). Before the shifts are used to create the slabs however, if the user decides to take into account whether or not a termination will break any polyhedral structure (bonds is not None), this method will filter out any shift values that do so.

Parameters: bonds ({(specie1, specie2) – max_bond_dist}: bonds are specified as a dict of tuples: float of specie1, specie2 and the max bonding distance. For example, PO4 groups may be defined as {(“P”, “O”): 3}. tol (float) – Threshold parameter in fcluster in order to check if two atoms are lying on the same plane. Default thresh set to 0.1 Angstrom in the direction of the surface normal. max_broken_bonds (int) – Maximum number of allowable broken bonds for the slab. Use this to limit # of slabs (some structures may have a lot of slabs). Defaults to zero, which means no defined bonds must be broken. symmetrize (bool) – Whether or not to ensure the surfaces of the slabs are equivalent. repair (bool) – Whether to repair terminations with broken bonds or just omit them. Set to False as repairing terminations can lead to many possible slabs as oppose to just omitting them. ([Slab]) List of all possible terminations of a particular surface. Slabs are sorted by the # of bonds broken.
move_to_other_side(init_slab, index_of_sites)[source]

This method will Move a set of sites to the other side of the slab (opposite surface).

Arg:

init_slab (structure): A structure object representing a slab. index_of_sites (list of ints): The list of indices representing

the sites we want to move to the other side.
Returns: (Slab) A Slab object with a particular shifted oriented unit cell.
nonstoichiometric_symmetrized_slab(init_slab, tol=0.001)[source]

This method checks whether or not the two surfaces of the slab are equivalent. If the point group of the slab has an inversion symmetry ( ie. belong to one of the Laue groups), then it is assumed that the surfaces should be equivalent. Otherwise, sites at the bottom of the slab will be removed until the slab is symmetric. Note the removal of sites can destroy the stoichiometry of the slab. For non-elemental structures, the chemical potential will be needed to calculate surface energy.

Arg:
init_slab (Structure): A single slab structure tol (float): Tolerance for SpaceGroupanalyzer.
Returns: A symmetrized Slab object. Slab (structure)
repair_broken_bonds(slab, bonds)[source]

This method will find undercoordinated atoms due to slab cleaving specified by the bonds parameter and move them to the other surface to make sure the bond is kept intact. In a future release of surface.py, the ghost_sites will be used to tell us how the repair bonds should look like.

Arg:

slab (structure): A structure object representing a slab. bonds ({(specie1, specie2): max_bond_dist}: bonds are

specified as a dict of tuples: float of specie1, specie2 and the max bonding distance. For example, PO4 groups may be defined as {(“P”, “O”): 3}.
Returns: (Slab) A Slab object with a particular shifted oriented unit cell.
generate_all_slabs(structure, max_index, min_slab_size, min_vacuum_size, bonds=None, tol=0.001, max_broken_bonds=0, lll_reduce=False, center_slab=False, primitive=True, max_normal_search=None, symmetrize=False, repair=False, include_reconstructions=False)[source]

A function that finds all different slabs up to a certain miller index. Slabs oriented under certain Miller indices that are equivalent to other slabs in other Miller indices are filtered out using symmetry operations to get rid of any repetitive slabs. For example, under symmetry operations, CsCl has equivalent slabs in the (0,0,1), (0,1,0), and (1,0,0) direction.

Parameters: structure (Structure) – Initial input structure. Note that to ensure that the miller indices correspond to usual crystallographic definitions, you should supply a conventional unit cell structure. max_index (int) – The maximum Miller index to go up to. min_slab_size (float) – In Angstroms min_vacuum_size (float) – In Angstroms bonds ({(specie1, specie2) – max_bond_dist}: bonds are specified as a dict of tuples: float of specie1, specie2 and the max bonding distance. For example, PO4 groups may be defined as {(“P”, “O”): 3}. tol (float) – Threshold parameter in fcluster in order to check if two atoms are lying on the same plane. Default thresh set to 0.1 Angstrom in the direction of the surface normal. max_broken_bonds (int) – Maximum number of allowable broken bonds for the slab. Use this to limit # of slabs (some structures may have a lot of slabs). Defaults to zero, which means no defined bonds must be broken. lll_reduce (bool) – Whether to perform an LLL reduction on the eventual structure. center_slab (bool) – Whether to center the slab in the cell with equal vacuum spacing from the top and bottom. primitive (bool) – Whether to reduce any generated slabs to a primitive cell (this does not mean the slab is generated from a primitive cell, it simply means that after slab generation, we attempt to find shorter lattice vectors, which lead to less surface area and smaller cells). max_normal_search (int) – If set to a positive integer, the code will conduct a search for a normal lattice vector that is as perpendicular to the surface as possible by considering multiples linear combinations of lattice vectors up to max_normal_search. This has no bearing on surface energies, but may be useful as a preliminary step to generating slabs for absorption and other sizes. It is typical that this will not be the smallest possible cell for simulation. Normality is not guaranteed, but the oriented cell will have the c vector as normal as possible (within the search range) to the surface. A value of up to the max absolute Miller index is usually sufficient. symmetrize (bool) – Whether or not to ensure the surfaces of the slabs are equivalent. repair (bool) – Whether to repair terminations with broken bonds or just omit them include_reconstructions (bool) – Whether to include reconstructed slabs available in the reconstructions_archive.json file.
get_integer_index(miller_index)[source]

Converts a vector of floats to whole numbers

get_recp_symmetry_operation(structure, symprec=0.01)[source]

Find the symmetric operations of the reciprocal lattice, to be used for hkl transformations :param structure: conventional unit cell :type structure: Structure :param symprec: default is 0.001

get_symmetrically_distinct_miller_indices(structure, max_index)[source]

Returns all symmetrically distinct indices below a certain max-index for a given structure. Analysis is based on the symmetry of the reciprocal lattice of the structure. :param structure: input structure. :type structure: Structure :param max_index: The maximum index. For example, a max_index of 1

means that (100), (110), and (111) are returned for the cubic structure. All other indices are equivalent to one of these.
miller_index_from_sites(supercell_matrix, coords)[source]

Get the Miller index of a plane from two vectors formed from three cartesian coordinates. If you use this module, please consider citing the following work:

Sun, W., & Ceder, G. (2018). A topological screening heuristic
for low-energy , high-index surfaces. Surface Science, 669(October
2017), 50–56. https://doi.org/10.1016/j.susc.2017.11.007

Parameters: supercell_matrix – 3x3 matrix describing the supercell or unit cell coords – List of three (numpy arrays) points as cartesian coordinates in the corresponding cell The Miller index
reduce_vector(vector)[source]