A BRIEF DISCUSSION ON THE DIFFERENT FORMULATIONS OF THE COUPLED MODE METHOD IN CHIRAL MEDIA: APPLICATION TO THE PARALLEL-PLATE CHIROWAVEGUIDE A. Go ´ mez, A. Vegas, and M. A. Solano University of Cantabria Dpto. De Ingenierı ´a de Comunicaciones Avda. de los Castros s/n 39005 Santander, Spain Received 10 January 2004 ABSTRACT: The application of the coupled-mode method (CMM) for calculating the propagation constants and the electromagnetic-field distribu- tion in chirowaveguides is well known. In this paper, two different formula- tions of the method are revised and compared, and are then extended to partially filled parallel-plate waveguides. © 2004 Wiley Periodicals, Inc. Microwave Opt Technol Lett 42: 181–185, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mop. 20245 Key words: chiral media; parallel-plate waveguides; coupled-mode method 1. INTRODUCTION The application of the coupled-mode method (CMM) to the study of electromagnetic-field propagation inside a waveguide that is fully or partially loaded with isotropic, anisotropic, and bi-isotro- pic materials is not new [1, 2]. This method consists in expanding a certain set of the electromagnetic-field components in terms of basis functions that correspond to the empty waveguide-field com- ponents. Two different formulations of the method have been proposed for chirowaveguides. In the first one, the components of the electric field E and the magnetic field H are expanded as a lineal combination of the electric and magnetic fields of the modes of the same waveguide, but without the chiral media, that is, the empty waveguide. We will call this EH formulation. In this formulation, it is essential to express the longitudinal amplitudes of the elec- tromagnetic field as a function of the amplitudes of the transverse components. Two different strategies can be used to do this. In the first one, which is normally used in the literature for bi-isotropic media, it is not necessary to solve any systems of equations; instead, the amplitudes are obtained directly (direct EH formula- tion) [3]. In the other strategy, E z and H z are not expressed directly as functions of the transverse components. This implies that it is necessary to solve a set of linear equations and, therefore, a matrix inversion must be performed (indirect EH formulation) [4]. In both EH formulations the normal components of H vanish at the perfect electric walls because they are expanded in terms of functions ( H -field for an empty waveguide), which are zero at these walls. However, if a chiral media is in contact with a perfect electric conductor wall, the normal components of H are not zero on the wall. Instead, the normal components of the magnetic induction B are zero. This drawback of the EH formulation can be remedied if the components of the magnetic induction B are expanded [5], instead of the components of H . This is called the EB formulation. A good discussion of the EB and EH formulations can be found in [6]. The goal of this paper is to extend the EB formulation to parallel-plate waveguides partially filled with chiral media, which has only been developed for totally filled waveguides. Also, if discontinuities between these kinds of structures need to be char- acterized, it is necessary to find the magnetic field. Then, we also propose a way of finding H (which is not directly obtained as in the EH formulation) from the EB formulation. Finally, we discuss the advantages and disadvantages and point out a common error found in the literature with regard to EH formulations. 2. THEORY 2.1. The EH-Formulations Let us consider the problem (illustrated in Fig. 1) where the planes y = 0 and y = b are perfect electric walls, thereby constituting a parallel-plate waveguide. Inside this waveguide there is a chiral material characterized by the following constitutive relations [7]: D =E - j 0 0 H , (1a) B = H + j 0 0 E , (1b) where is the dielectric permittivity, is the magnetic permeabil- ity, and is the Pasteur parameter. The way to obtain the algebraic system of equations for EH formulations, which is based on the different form of relating E z , D z , H z , and B z by means of Eq. (1), can be found in [2, 3] (the direct strategy) and in [4] (the indirect strategy). It must be noted that in [3] the development for the H z component [Eq. (1)] is not correct, because a term independent of the transverse coordinates is not included. For magnetic media, it has been proved [8] that the absence of this term produces incor- rect results, both in the propagation constants and in the electro- magnetic field. The same occurs for bi-isotropic media [4]. On the other hand, the direct strategy produces, in general, poorer con- vergence results than the indirect strategy for the propagation constants (that is, for the same number of basis functions, the value of the propagation constant of any mode is more accurate using indirect rather than direct strategy) and also poorer results in the profile of the electromagnetic field. This apparent and surprising behavior (because one might think that better results would be obtained with direct strategy, where no matrix inversion must be performed than with indirect strategy, where matrix inversion must be performed) is due to another remarkable situation based on indirect strategy, which ensures a much better numerical treatment of the discontinuous nature of the longitudinal components of the electromagnetic field [4, 7]. The complete expressions for these two formulations can be seen in detail in [4]. 2.2. The EB Formulation In the EB formulation, the magnetic field H is not developed directly using T-functions. In this case, H is substituted by the magnetic induction B . The expansions are given by Figure 1 Parallel-plate waveguide partially filled with a slab of chiral medium (PEC: perfect electric conductor) MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 42, No. 3, August 5 2004 181