Abstract

In a recent article (Canc\`es, Deleurence and Lewin, Commun. Math. Phys., 281 (2008), pp. 129-177), we have rigorously derived, by means of bulk limit arguments, a new variational model to describe the electronic ground state of insulating or semiconducting crystals in the presence of local defects. In this so-called reduced Hartree-Fock model, the ground state electronic density matrix is decomposed as $\gamma = \gamma^0_{\rm per} + Q_{\nu,\epsilon_{\rm F}}$, where $\gamma^0_{\rm per}$ is the ground state density matrix of the host crystal and $Q_{\nu,\epsilon_{\rm F}}$ the modification of the electronic density matrix generated by a modification $\nu$ of the nuclear charge of the host crystal, the Fermi level $\epsilon_{\rm F}$ being kept fixed. The purpose of the present article is twofold. First, we study more in details the mathematical properties of the density matrix $Q_{\nu,\epsilon_{\rm F}}$ (which is known to be a self-adjoint Hilbert-Schmidt operator on $L^2(\R^3)$). We show in particular that if $\int_{\RR^3} \nu \neq 0$, $Q_{\nu,\epsilon_{\rm F}}$ is not trace-class. Moreover, the associated density of charge is not in $L^1(\R^3)$ if the crystal exhibits anisotropic dielectric properties. These results are obtained by analyzing, for a small defect $\nu$, the linear and nonlinear terms of the resolvent expansion of $Q_{\nu,\epsilon_{\rm F}}$. Second, we show that, after an appropriate rescaling, the potential generated by the microscopic total charge (nuclear plus electronic contributions) of the crystal in the presence of the defect, converges to a homogenized electrostatic potential solution to a Poisson equation involving the macroscopic dielectric permittivity of the crystal. This provides an alternative (and rigorous) derivation of the Adler-Wiser formula.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call