Support CIF files from ICSD with atom type symbols like Au0+

.zenodo.json

@@ -63,6 +63,11 @@
     {
       "name":"Tina Bergh"
     },
+    {
+      "name":"Dieter Weber",
+      "orcid": "0000-0001-6635-9567",
+      "affiliation": "Forschungszentrum Jülich"
+    },
     {
       "name":"Tomas Ostasevicius"
     },

CHANGELOG.rst

@@ -7,6 +7,17 @@ The format is based on `Keep a Changelog <https://keepachangelog.com/en/1.0.0>`_
 this project tries its best to adhere to
 `Semantic Versioning <https://semver.org/spec/v2.0.0.html>`_.
 
+Unreleased
+==========
+
+Added
+-----
+
+- Support for wider range of `CIF atom type symbols
+  <https://www.iucr.org/__data/iucr/cifdic_html/1/cif_core.dic/Iatom_type_symbol.html>`_,
+  notably element symbol followed by oxidation number as found in CIF files
+  recently pulled from `ICSD <https://icsd.fiz-karlsruhe.de/>`_ (#245).
+
 2025-06-02 - version 0.7.0
 ==========================
 

diffsims/release_info.py

@@ -23,6 +23,7 @@
     "Jedrzej Morzy",
     "Endre Jacobsen",
     "Tina Bergh",
+    "Dieter Weber",
     "Tomas Ostasevicius",
     "Eirik Opheim",
 ]

diffsims/structure_factor/atomic_scattering_parameters.py

@@ -16,6 +16,8 @@
 # You should have received a copy of the GNU General Public License
 # along with diffsims.  If not, see <http://www.gnu.org/licenses/>.
 
+import re
+
 import numpy as np
 
 # List of elements Z = 1-118
@@ -140,6 +142,28 @@
 # fmt: on
 
 
+def get_element(atom_type_symbol):
+    """Extract element symbol from _atom_type_symbol CIF key
+
+    See https://www.iucr.org/__data/iucr/cifdic_html/1/cif_core.dic/Iatom_type_symbol.html
+    for specification.
+
+    Parameters
+    ----------
+    atom_type_symbol : str
+        Symbol following the specification of _atom_type_symbol CIF key
+
+    Returns
+    -------
+
+    element : str
+        Alphabetic head of atom_type_symbol, notably without oxidation state.
+        Usually an element symbol, but not guaranteed by the specification.
+    """
+    reg = "^(?P<element>[A-Za-z]*)"
+    return re.match(reg, atom_type_symbol).group("element")
+
+
 def get_atomic_scattering_parameters(element, unit=None):
     """Return the eight atomic scattering parameters a_1-4, b_1-4 for
     elements with atomic numbers Z = 1-98 from Table 12.1 in
@@ -191,20 +215,26 @@ def get_atomic_scattering_parameters(element, unit=None):
 
 def get_element_id_from_string(element_str):
     r"""Get periodic element ID for elements :math:`Z` = 1-98 from
-    one-two letter string.
+    element symbol.
 
     Parameters
     ----------
     element_str : str
-        One-two letter string.
+        String starting with an element symbol, optionally delimited from other
+        following characters by a non-alphabetic character. Only the element
+        symbol is evaluated. In particular, compatible with common atom type
+        symbols in CIF files that often start with an element symbol followed by
+        e.g. an optional oxidation number, although this is not guaranteed:
+        https://www.iucr.org/__data/iucr/cifdic_html/1/cif_core.dic/Iatom_type_symbol.html
 
     Returns
     -------
     element_id : int
         Integer ID in the periodic table of elements.
     """
+    element = get_element(element_str)
     element2periodic = dict(zip(ELEMENTS[:N_ELEMENTS], np.arange(1, N_ELEMENTS)))
-    element_id = element2periodic[element_str.lower()]
+    element_id = element2periodic[element.lower()]
     return element_id
 
 

diffsims/tests/structure_factor/test_structure_factor.py

@@ -34,7 +34,10 @@
     space_group=225,
     structure=Structure(
         lattice=Lattice(3.5236, 3.5236, 3.5236, 90, 90, 90),
-        atoms=[Atom(xyz=[0, 0, 0], atype="Ni", Uisoequiv=0.006332)],
+        # Atom specified in a way found in CIF files from ICSD
+        # See also
+        # https://www.iucr.org/__data/iucr/cifdic_html/1/cif_core.dic/Iatom_type_symbol.html
+        atoms=[Atom(xyz=[0, 0, 0], atype="Ni0+", Uisoequiv=0.006332)],
     ),
 )
 ferrite = Phase(

diffsims/tests/utils/test_sim_utils.py

@@ -20,6 +20,9 @@
 import numpy as np
 import diffpy
 
+from diffpy.structure import Atom, Lattice, Structure
+from orix.crystal_map import Phase
+
 
 from diffsims.utils.sim_utils import (
     get_electron_wavelength,
@@ -348,3 +351,44 @@ def test_get_kinematical_intensities(default_structure, scattering_params, answe
         scattering_params=scattering_params,
     )
     np.testing.assert_array_almost_equal(i_hkls, ([answer]), decimal=4)
+
+
+# Copied from Nickel, with unknown element Unobtainium
+# Arbitrary strings can be found in CIF files, see
+# https://www.iucr.org/__data/iucr/cifdic_html/1/cif_core.dic/Iatom_type_symbol.html
+unobtainium = Phase(
+    space_group=225,
+    structure=Structure(
+        lattice=Lattice(3.5236, 3.5236, 3.5236, 90, 90, 90),
+        # Note that the standard explicitly shows "NiFe" and "dummy"
+        # as examples
+        atoms=[Atom(xyz=[0, 0, 0], atype="UnOb", Uisoequiv=0.006332)],
+    ),
+)
+
+
+@pytest.mark.parametrize(
+    "scattering_params, answer",
+    [
+        ("lobato", 0.0),
+        (None, 1.0),
+    ],
+)
+def test_get_kinematical_intensities_unknown(scattering_params, answer):
+    latt = unobtainium.structure.lattice
+    reciprocal_lattice = latt.reciprocal()
+    reciprocal_radius = 0.2
+    unique_hkls, multiplicites, g_hkls = get_intensities_params(
+        reciprocal_lattice, reciprocal_radius
+    )
+    g_hkls_array = np.asarray(g_hkls)
+    # Debatable if this should actually be an error and not just a warning
+    with pytest.warns(UserWarning, match="not found in scattering parameter library."):
+        i_hkls = get_kinematical_intensities(
+            unobtainium.structure,
+            g_indices=unique_hkls,
+            g_hkls_array=g_hkls_array,
+            prefactor=multiplicites,
+            scattering_params=scattering_params,
+        )
+    np.testing.assert_array_almost_equal(i_hkls, ([answer]), decimal=4)

diffsims/utils/sim_utils.py

@@ -18,13 +18,15 @@
 
 import collections
 import math
+from warnings import warn
 
 import diffpy.structure
 import numpy as np
 from scipy.constants import h, m_e, e, c, pi, mu_0
 
 from diffsims.utils.atomic_scattering_params import ATOMIC_SCATTERING_PARAMS
 from diffsims.utils.lobato_scattering_params import ATOMIC_SCATTERING_PARAMS_LOBATO
+from diffsims.structure_factor.atomic_scattering_parameters import get_element
 
 
 __all__ = [
@@ -194,22 +196,29 @@ def get_vectorized_list_for_atomic_scattering_factors(
     dwfactors : numpy.ndarray
         Debye-Waller factors for each atom in the structure.
     """
-
     if scattering_params is not None:
         scattering_params_dict = get_scattering_params_dict(scattering_params)
     else:
         scattering_params_dict = {}
 
     n_structures = len(structure)
-    coeffs = np.empty((n_structures, 5, 2))
+    coeffs = np.zeros((n_structures, 5, 2))
     fcoords = np.empty((n_structures, 3))
     occus = np.empty(n_structures)
     dwfactors = np.empty(n_structures)
-    default = np.zeros((5, 2))
 
     for i, site in enumerate(structure):
-        coeffs[i] = scattering_params_dict.get(site.element, default)
-        dwfactors[i] = debye_waller_factors.get(site.element, 0)
+        element = get_element(site.element)
+        # At least emit a warning if a key is not found.
+        # Probably better to error out since the simulation will be invalid.
+        try:
+            coeffs[i] = scattering_params_dict[element]
+        except KeyError:
+            warn(
+                f"Element {element} from atom type symbol {site.element} not "
+                "found in scattering parameter library."
+            )
+        dwfactors[i] = debye_waller_factors.get(element, 0)
         fcoords[i] = site.xyz
         occus[i] = site.occupancy
 

setup.py

@@ -27,7 +27,8 @@
 # fmt: off
 extra_feature_requirements = {
     "doc": [
-        "numpydoc",
+        # Restriction due to https://github.com/pyxem/orix/issues/570
+        "numpydoc               != 1.9.0",
         "pydata-sphinx-theme",
         "sphinx                 >= 3.0.2",
         "sphinx-design",