Ri is a column vector containing the three laboratory-fixed coordinates Xi, Yi, Zi of nucleus i. R is a column vector containing the three laboratory-fixed coordinates of the center-of-mass of the nuclei. S is a 3 × 3 rotation matrix relating the orientation of the molecule-fixed axes x,y,z, to the orientation of the corresponding laboratory-fixed axes X,Y,Z. We fix conventions for the Eulerian angles χ θ φ by taking S to have the form
where cχ = cosχ, sχ = sinχ, etc. The quantity ai is a column vector containing the three molecule-fixed coordinates of nucleus i at equilibrium, and di is a column vector containing the three molecule-fixed components of the instantaneous displacement vector of nucleus i from its equilibrium position.
Equation (9) can be understood most easily by reading it from the right-hand side, as illustrated in Fig. 2. We consider the nuclei to be brought into their instantaneous positions in the laboratory by four steps. (a) Each nucleus is placed at its equilibrium position ai, as in Fig. 2a. (b) Each nucleus is displaced from its equilibrium position by an amount di, as in Fig. 2b. (c) The entire molecular framework is rotated using the rotation matrix S(χθφ), as in Fig. 2c. (d) The entire molecular framework is translated by the vector R, as in Fig. 2d.
For completeness, though they will not be referred to again in this article, the center-of-mass and Eckart conditions defining (explicitly) the vector R and (implicitly) the rotational angles (χθφ), are given in (eq. 11).
The qualitative ideas required to relate the point-group and permutation-inversion operations are as follows. We seek a set of transformations for the dynamical variables comprising vibration-rotation-translation phase space, such that after the di, χ, θ, φ, and R on the right-hand side of (eq. 9), have been replaced by the (di)new, χnew, θnew, φnew, and Rnew, the new right-hand side will be consistent with a left-hand side in which Ri has been replaced by (Ri)new = ±Rj, with a sign choice and subscript choice in agreement with the prescription of the permutation-inversion operation under consideration.
It is sometimes convenient, when a given symmetry operation P is to be applied to the dynamical variables of configuration space, to visualize it as the product of three separate operators 1P, 2P, 3P, which act on the vibrational, rotational, and translational variables, respectively. While we shall not use this notation explicitly in Section 4.2 and Section 4.3, we shall deal separately and in turn with vibrational, rotational, and translational transformations. When considerations involving translation of the molecule as a whole are not important, the operator 3P can be ignored, as has been done in Fig. 3 and Fig. 4 illustrating the vibrational and rotational transformations discussed below.