Exchange-correlation matrices

Loop structure for contribution to exchange-correlation matrices from a sphere Exc - Exc + w f... 

Algorithm for construction of implicit derivatives of exchange-correlation matrices... 

With Eq. (67) in Eq. (66), the resulting Kohn-Sham exchange-correlation matrix elements (although rooted only in the exchange) yield the integral representation ... 

The hKs matrix is analogous to the Fock matrix in wave mechanics, and the one-electron and Coulomb parts are identical to the corresponding Fock matrix elements. The exchange-correlation part, however, is given in terms of the electron density, and possibly also involves derivatives of the density (or orbitals, as in the BR functional, eq. (6.25)). 

What we have not discussed so far is how the contribution of the final components of the Kohn-Sham matrix in equation (7-12), i. e., the exchange-correlation part, can be computed. What we need to solve are terms such as... 

Once the coefficients for the expansion of the exchange-correlation term have been evaluated, all matrix elements can be calculated analytically. The Obara and Saika [47] recursive scheme has been used for the evaluation of the one and the two electron integrals. The total energy is therefore expressed in terms of the fitting coefficients for the electronic density and the exchange-correlation potential. 

Spin-dependent operators are now introduced. The external potential can be an operator Vext acting on the two-component spinors. The exchange-correlation potential is defined as in Eq. [27], although Exc is now a functional Exc = Exc[pap] of the spin-density matrix. The exchange-correlation potential is then... 

This leads to an exchange-correlation potential in the form of a 2 x 2 Hermitian matrix in spin space... 

This long-range correlation effect shows up in both the first-order density matrix and the exchange-correlation hole for finite systems [19]. We concentrate here on the exchange-correlation hole. The general asymptotic form of the pair density is then... 

To study the structure of the exchange-correlation energy functional, it is useful to relate this quantity to the pair-correlation function. The pair-correlation function of a system of interacting particles is defined in terms of the diagonal two-particle density matrix (for an extensive discussion of the properties of two-particle density matrices see [30]) as... 

Z is the nuclear charge, R-r is the distance between the nucleus and the electron, P is the density matrix (equation 16) and (qv Zo) are two-electron integrals (equation 17). f is an exchange/correlation functional, which depends on the electron density and perhaps as well the gradient of the density. Minimizing E with respect to the unknown orbital coefficients yields a set of matrix equations, the Kohn-Sham equations , analogous to the Roothaan-Hall equations (equation 11). 

It is seen that the basic operator (65) and strength matrix (66) have now simple expressions via 6p f) from (62). The operator (65) has exchange-correlation and Coulomb terms. For electric multipole oscillations (dipole plasmon,. ..), the Coulomb term dominates. 

The problem is that this exchange correlation potential does not in a simple fashion depend on only the density or the diagonal part of the one-matrix. The antisymmetry property of the two-matrix, which yields the last term in Eq. (33) as well as the apperance of the coefficents rkj kj, prevent this. Only some further (ad hoc) assumptions will allow an exchange-correlation potential that is simply a function of the electron density. This can be achieved in a number of ways. For a full discussion on this subject see [11]. 

Here Ho is the kinetic energy operator of valence electrons Vps is the pseudopotential [40,41] which defines the atomic core. V = eUn(r) is the Hartree energy which satisfies the Poisson equation ArUn(r) = —4nep(r) with proper boundary conditions as discussed in the previous subsection. The last term is the exchange-correlation potential Vxc [p which is a functional of the density. Many forms of 14c exist and we use the simplest one which is the local density approximation [42] (LDA). One may also consider the generalized gradient approximation (GGA) [43,44] which can be implemented for transport calculations without too much difficulty [45]. Importantly a self-consistent solution of Eq. (2) is necessary because Hks is a functional of the charge density p. One constructs p from the KS states Ts, p(r) = (r p r) = ns Fs(r) 2, where p is the density matrix,... 

---