International
Tables for Crystallography Volume A Space-group symmetry Edited by Th. Hahn © International Union of Crystallography 2006 |
International Tables for Crystallography (2006). Vol. A. ch. 13.1, pp. 836-842
https://doi.org/10.1107/97809553602060000528 Chapter 13.1. Isomorphic subgroupsIn this chapter, the (infinitely many) isomorphic subgroups of a space group are defined. The matrices relating a plane group or space group and its isomorphic subgroup are tabulated for each crystal system. Conditions on the coefficients of the matrices are also given, as are examples of the use of the matrices. Keywords: isomorphic subgroups; space groups; plane groups; symmetry operations; crystal systems. |
A subgroup of a space group is an isomorphic subgroup if is of the same or the enantiomorphic space-group type as . Thus, isomorphic space groups are a special subset of klassengleiche subgroups. The maximal isomorphic subgroups of lowest index are listed under IIc in the space-group tables of this volume (Part 7 ) (cf. Section 2.2.15 ). Isomorphic subgroups can easily be recognized because the standard space-group symbols of and are the same [isosymbolic subgroups (Billiet, 1973)] or the symbol of is enantiomorphic to that of . Every space group has an infinite number of maximal isomorphic subgroups, whereas the number of maximal non-isomorphic subgroups is finite (cf. Section 8.3.3 ). For this reason, isomorphic subgroups are discussed in more detail in the present section.
If a, b, c are the basis vectors defining the conventional unit cell of and the basis vectors corresponding to the relation holds, where (a, b, c) and are row matrices and S is a matrix. The coefficients of S are integers.1
The index of in is equal to 1, which is the ratio of the volumes [] and [abc] of the two cells. is positive if the bases of the two cells have the same handedness and negative if they have opposite handedness.
If O and O′ are the origins of the coordinate systems (O, a, b, c) and , used for the description of and , the column matrix of the coordinates of O′ referred to the system (O, a, b, c) will be denoted by s. Thus, the coordinate system will be specified completely by the square matrix S and the column matrix s, symbolized by .
An example of the application of equation (13.1.1.1) is given at the end of this chapter.
Let be the operator of a given symmetry operation of referred to (O, a, b, c) and the operator of the same operation referred to . Then the following relation applies (cf. Bertaut & Billiet, 1979). The latter expression is more conventional, the former is easier to manipulate. Identifying the rotational (matrix) and translational (column) parts of , one obtains the following two conditions: or
Here we have split w into a fractional part (smaller than any lattice translation) and which describes a lattice translation in .
The general expression of the matrix S is This general form, without any restrictions on the coefficients, applies only to the triclinic space groups P1 and ; P1 has only isomorphic subgroups (cf. Billiet, 1979; Billiet & Rolley Le Coz, 1980). For other space groups, restrictions have to be imposed on the coefficients .
For convenience, we consider first those crystal systems that possess a unique direction (the privileged axis being taken parallel to c). We also include here the monoclinic system (unique axis either c or b).
If W is the matrix corresponding to a rotation about the c axis, holds if the positive direction is the same for c and c′.2 In consequence, W must commute with S [cf. equation (13.1.1.2a)]. This condition imposes relations on the coefficients of the matrix so that S and take the following forms:
If mirror or glide planes parallel to and/or twofold rotation or screw axes perpendicular to the principal rotation axis exist, further conditions are imposed upon the coefficients and these are indicated below (cf. Bertaut & Billiet, 1979).
|
As stated above, P1 has only isomorphic subgroups and the general nature of the matrix S [equation (13.1.1.3)] requires the use of special techniques (cf. Chapter 13.2 , Derivative lattices); they apply also to .
In equation (13.1.1.2b), there occurs the choice of the origin by means of s, the nature of the lattice by means of and the nature of the symmetry operations by means of the column matrices w and w′. The three factors, origin, lattice type, screw and glide components, impose parity conditions on the coefficients of the matrix S. Only a few examples are given here.
When (W, w) and are operators of lattice translations of , say and , equation (13.1.1.2b) reduces to
Example
In a tetragonal I lattice, is either an integral or a fractional translation. If is and the matrix S is replaced by , one obtains with integers.
From this it follows that either and have the same parity and is even, or else and have opposite parities and is odd.
If (W, w) and represent the same operation in the two P lattices of and , with , for , equation (13.1.1.2b) reduces to
There are similar conditions for glide and other screw operations.
The location part s, i.e. the relative positions of the origins O and O′ of group and subgroup, is another important problem. This problem can be approached in two ways. First, only standard settings and origins (as defined in the space-group tables of this volume) of the group and its subgroup are considered. In this case, the origin relation between and must be indicated by the appropriate column s. For instance, in the matrices , , are only possible for , say . Second, one can describe based on the same origin as , i.e. O and O′ coincide. In this case, nonstandard descriptions of frequently result and one has to indicate the location of the symmetry elements of with respect to the origin of .
Unfortunately, it was not possible to incorporate in the present tables the implications that the choice of the origin has on the coefficients (cf. Bertaut, 1956; Billiet, 1978; Bertaut & Billiet, 1979).
The explicit forms of the matrices S for each space group and each plane group are given in Tables 13.1.2.1 and 13.1.2.2 without origin indications.
|
|
Example
Consider space groups and and envisage the matrix which has the determinant . Its lowest value is 2 with . The matrix then implies the axis transformation
Envisage also the matrix which has the determinant . Its lowest value for a proper subgroup is 2 with and . The axis transformation is here
This transformation is permitted for but not for where the parity rules require . Thus, the lowest value of for a subgroup of space group is 3 and the axis transformation is
There is no difficulty in reducing the preceding considerations to plane groups, where S is a matrix and s a matrix.
The matrices and the restrictions on the coefficients are listed for the plane groups in Table 13.1.2.1 and for the space groups in Table 13.1.2.2.
For the triclinic and monoclinic systems, there is an infinite choice of matrices for each index, owing to the infinite number of equivalent unit cells. For the other space groups, several different (but finitely many) choices of matrix occur. In all cases, we have restricted this choice to one matrix for each group–subgroup relation so that each subgroup is listed exactly once (apart from origin choice).
Example
No. 178, , has the matrix for all isomorphic and isosymbolic subgroups , having the lattices , , , whilst is used for all isomorphic and isosymbolic subgroups with the lattices , , . These two kinds of subgroups are obviously different, having different translation lattices. The same group, , has the matrix for the isomorphic and enantiomorphic subgroups of lattices , , whilst is used for the isomorphic and enantiomorphic subgroups of lattices , , .
In the tables, each system is preceded by the appropriate general form of the matrix, which is also given in this chapter, followed by the more specialized matrices such as . Under Conditions, we have listed the nonredundant inequalities and parity conditions3 that ensure the uniqueness of the matrix for each subgroup. Also, we have used rules in order to avoid repetition of equivalent unit cells. For instance, for trigonal and hexagonal groups (rhombohedral groups excepted), we have restricted to lie between a and a + b excluding this last vector from the sector of 60° because there is a repetition after a 60° rotation of the unit cell.
References
Bertaut, E. F. (1956). Structure de FeS stoechiométrique. Bull. Soc. Fr. Minéral. Cristallogr. 79, 276–292.Google ScholarBertaut, E. F. & Billiet, Y. (1979). On equivalent subgroups and supergroups of the space groups. Acta Cryst. A35, 733–745.Google Scholar
Billiet, Y. (1973). Les sous-groupes isosymboliques des groupes spatiaux. Bull. Soc. Fr. Minéral. Cristallogr. 96, 327–334.Google Scholar
Billiet, Y. (1978). Some remarks on the `family tree' of Bärnighausen. Acta Cryst. A34, 1023–1025.Google Scholar
Billiet, Y. (1979). Le groupe P1 et ses sous-groupes. I. Outillage mathématique: automorphisme et factorisation matricielle. Acta Cryst. A35, 485–496.Google Scholar
Billiet, Y. & Rolley Le Coz, M. (1980). Le groupe P1 et ses sous-groupes. II. Tables de sous-groupes. Acta Cryst. A36, 242–248.Google Scholar