ANALYTIC APPROACH TO THE INTERFACIAL POLARIZATION OF HETEROGENEOUS SYSTEMS

Citation
L. Fu et al., ANALYTIC APPROACH TO THE INTERFACIAL POLARIZATION OF HETEROGENEOUS SYSTEMS, Physical review. B, Condensed matter, 47(20), 1993, pp. 13818-13829
Citations number
34
Categorie Soggetti
Physics, Condensed Matter
ISSN journal
01631829
Volume
47
Issue
20
Year of publication
1993
Pages
13818 - 13829
Database
ISI
SICI code
0163-1829(1993)47:20<13818:AATTIP>2.0.ZU;2-8
Abstract
We have obtained an analytical solution of Maxwell's equations for a d ielectric system of spherical particles in a medium. The system is pla ced between two parallel electrode plates, subject to a low-frequency alternating potential, and the solution is obtained from the correspon ding boundary-value problem for the Green function. All the multipole moments and the electric field are expressed in terms of the applied p otential at the electrodes and a matrix which depends on the system co nfiguration. The effective dielectric function is then obtained as an average over the whole sample. For disordered systems, we solve exactl y the case of two-particle distributions with short-range correlations and find that the form of the distribution plays a crucial role. In p articular, we prove that for spherically symmetric two-particle distri butions all multipole moments except dipoles are exactly zero, and the Maxwell-Garnett result, or, equivalently, the Clausius-Mossotti relat ion for spherical particles, is valid regardless of the particle conce ntration. Within the two-particle distribution, corrections to the Max well-Garnett result can only derive from nonsphericity in the distribu tion: in such a case, higher multipole moments are generally nonzero a nd may strongly affect the effective dielectric function. We provide t he explicit expressions for all the multipole moments and the effectiv e dielectric function, which can be computed straightforwardly for any given distribution. We show that an iterative unsymmetrical procedure proposed originally by Bruggeman is inconsistent with our results.