Loading [MathJax]/extensions/TeX/AMSsymbols.js
SIRIUS 7.5.0
Electronic structure library and applications
All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Pages
Symmetry

Definition of symmetry operation

SIRIUS uses Spglib to find the spacial symmetry operations. Spglib defines symmetry operation in fractional coordinates:

\[ {\bf x'} = \{ {\bf R} | {\bf t} \} {\bf x} \equiv {\bf R}{\bf x} + {\bf t} \]

where R is the proper or improper rotation matrix with elements equal to -1,0,1 and determinant of 1 (pure rotation) or -1 (rotoreflection) and t is the fractional translation, associated with the symmetry operation. The inverse of the symmetry operation is:

\[ {\bf x} = \{ {\bf R} | {\bf t} \}^{-1} {\bf x'} = {\bf R}^{-1} ({\bf x'} - {\bf t}) = {\bf R}^{-1} {\bf x'} - {\bf R}^{-1} {\bf t} \]

We will always use an active transformation (transformation of vectors or functions) and never a passive transformation (transformation of coordinate system). However one should remember definition of the function transformation:

\[ \hat {\bf P} f({\bf r}) \equiv f(\hat {\bf P}^{-1} {\bf r}) \]

It is straightforward to get the rotation matrix in Cartesian coordinates. We know how the vector in Cartesian coordinates is obtained from the vector in fractional coordinates:

\[ {\bf v} = {\bf L} {\bf x} \]

where L is the 3x3 matrix which clomuns are three lattice vectors. The backward transformation is simply

\[ {\bf x} = {\bf L}^{-1} {\bf v} \]

Now we write rotation operation in fractional coordinates and apply the backward transformation to Cartesian coordinates:

\[ {\bf x'} = {\bf R}{\bf x} \rightarrow {\bf L}^{-1} {\bf v'} = {\bf R} {\bf L}^{-1} {\bf v} \]

from which we derive the rotation operation in Cartesian coordinates:

\[ {\bf v'} = {\bf L} {\bf R} {\bf L}^{-1} {\bf v} \]