G-matrix method

In Chapter 10, it is shown that by using symmetry considerations alone we may predict the number of vibrational fundamentals, their activities in the infrared and Raman spectra, and the way in which the various bonds and interbond angles contribute to them for any molecule possessing some symmetry. The actual magnitudes of the frequencies depend on the interatomic forces in the molecule, and these cannot be predicted from symmetry properties. However, the technique of using symmetry restrictions to set up the equations required in calculations in their most amenable form (the F-G matrix method) is presented in detail. 

In order to use the F-G matrix method to maximum advantage, it is first necessary to set up linear combinations of internal coordinates which provide the proper number of functions, transforming according to each of the irreducible representations spanned by the 3N - 6 normal modes of genuine... 

This type of molecule will be used to illustrate all of the procedures and techniques discussed earlier in this chapter, including setting up the vibrational secular equation by the -G matrix method. 

Because chemists seem to have become increasingly interested in employing vibration spectra quantitatively—or at least semiquantitatively—to obtain information on bond strengths, it seemed mandatory to augment the previous treatment of molecular vibrations with a description of the efficient F and G matrix method for conducting vibrational analyses. The fact that the convenient projection operator method for setting up symmetry coordinates has also been introduced makes inclusion of this material particularly feasible and desirable. 

The generalization of a force constant, k, and reduced mass, fi, from a one dimensional harmonic oscillator to the normal mode oscillators of a polyatomic molecule is accomplished by the F, G matrix methods of Wilson, et al., (1955). For the present discussion it is sufficient to know that a force constant and a reduced mass may be uniquely defined for each of 3N — 6 linearly independent sets of internal coordinate displacements in a polyatomic molecule (see also Section 9.4.12).] The harmonic oscillator Hamiltonian... 

The Wilson F-G matrix method for calculating force constants is discussed in most texts on theoretical spectroscopy. 

From the time function F t) and the calculation of [IT], the values of G may be found. One way to calculate the G matrix is by a fast Fourier technique called the Cooley-Tukey method. It is based on an expression of the matrix as a product of q square matrices, where q is again related to N by = 2 . For large N, the number of matrix operations is greatly reduced by this procedure. In recent years, more advanced high-speed processors have been developed to carry out the fast Fourier transform. The calculation method is basically the same for both the discrete Fourier transform and the fast Fourier transform. The difference in the two methods lies in the use of certain relationships to minimize calculation time prior to performing a discrete Fourier transform. 

With the availabihty of computers, the transfer matrix method [14] emerged as an alternative and powerful technique for the study of cooperative phenomena of adsorbates resulting from interactions [15-17]. Quantities are calculated exactly on a semi-infinite lattice. Coupled with finite-size scaling towards the infinite lattice, the technique has proved popular for the determination of phase diagrams and critical-point properties of adsorbates [18-23] and magnetic spin systems [24—26], and further references therein. Application to other aspects of adsorbates, e.g., the calculation of desorption rates and heats of adsorption, has been more recent [27-30]. Sufficient accuracy can usually be obtained for the latter without scaling and essentially exact results are possible. In the following, we summarize the elementary but important aspects of the method to emphasize the ease of application. Further details can be found in the above references. 

Among the methods discussed in this book, FEP is the most commonly used to carry out alchemical transformations described in Sect. 2.8 of Chap. 2. Probability distribution and TI methods, in conjunction with MD, are favored if there is an order parameter in the system, defined as a dynamical variable. Among these methods, ABF, derived in Chap. 4, appears to be nearly optimal. Its accuracy, however, has not been tested critically for systems that relax slowly along the degrees of freedom perpendicular to the order parameter. Adaptive histogram approaches, primarily used in Monte Carlo simulations - e.g., multicanonical, WL and, in particular, the transition matrix method - yield superior results in applications to phase transitions,... 

Fig. 14. Automatic correction of the thermograms by the state-function theory and the time-domain matrix methods the input f(t) the thermogram g(t) the corrected thermogram fc(t) T is the sampling period (20 sec). Reprinted from (62) with permission.

GF method calculations are simplified by the systematic behavior of the G matrix elements (Decius 1948). MUBFF calculations, however, are somewhat complicated by the force constants representing interactions between non-bonded atoms— these can be tedious to express in terms of internal coordinates. Computer programs have been written to partially automate calculations, thereby reducing the necessary effort and minimizing opportunities for errors (e.g., Schachtschneider 1964 Gale and Rohl 2003). 

G. Gidofalvi and D. A. Mazziotti, Spin and symmetry adaptation of the variational two-electron reduced-density-matrix method. Phys. Rev. A 72, 052505 (2005). 

Note that there is no simple method for converting the g-density from the spatial representation to the orbital representation and back. The g-density in the spatial representation depends on off-diagonal elements of the g-matrix in the orbital representation and the g-density in the orbital representation depends on off-diagonal elements of the g-matrix in the spatial representation. 

It is fair to say that neither of these two approaches works especially well N-representability conditions in the spatial representation are virtually unknown and the orbital-resolved computational methods are promising, but untested. It is interesting to note that one of the most common computational algorithms (cf. Eq. (96)) can be viewed as a density-matrix optimization, although most authors consider only a weak A -representability constraint on the occupation numbers of the g-matrix [1, 4, 69]. Additional A-representability constraints could, of course, be added, but it seems unlikely that the resulting g-density functional theory approach would be more efficient than direct methods based on semidefinite programming [33, 35-37]. 

---