Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Effective negative refractive index in ferromagnet-semiconductor superlattices

Open Access Open Access

Abstract

Problem of anomalous refraction of electromagnetic waves is analyzed in a superlattice which consists of alternating layers of ferromagnetic insulator and nonmagnetic semiconductor. Effective permittivity and permeability tensors are derived in the presence of an external magnetic field parallel to the plane of the layers. It is shown that in the case of the Voigt configuration, the structure behaves as a left-handed medium with respect to TE-type polarized wave, in the low-frequency region of propagation. The relative orientation of the Poynting vector and the refractive wave vector is examined in different frequency ranges. It is shown that the frequency region of existence for the backward mode can be changed using external magnetic field as tuning parameter.

©2006 Optical Society of America

1. Introduction

Recently, there have been many studies of a novel class of artificial materials which exhibit negative refraction of electromagnetic waves (EMW) in certain frequency ranges (see, for example, Refs. [14] and citations therein). These materials, termed as “left-handed”, are characterized by a negative permittivity and simultaneously a negative permeability, and lead to a variety of unusual optical phenomena [5]. In addition, in such materials the subwavelength focusing is possible which can be used to fabricate a perfect lens [6]. More recently, the predictions of Veselago and Pendry stimulated strong theoretical and experimental efforts to study negative refraction in various systems such as two-dimensional photonic-gap crystals [7], multilayer structures involving anisotropic materials [8], semiconductor superlattices in the presence of Bloch oscillations [9]. Experimental success has been demonstrated in ferromagnet-superconductor superlattices for millimeter waves [10].

In this paper, the problem of anomalous refraction is studied in a periodic ferromagnetsemiconductor multilayer composite, and the differences between the phase and group refracted indices are discussed. We show that in certain frequency range, left-handed behavior in such a structure is possible. Analytical expressions for both phase and group refractive indices are obtained. It is shown that the group index can be both positive and negative depending on the wave frequency and material parameters while the phase index is always positive.In section 2, we describe the physical picture of the problem and find the effective permittivity and permeability tensors in the presence of an external magnetic field. In section 3, the regions of propagation and dispersion curves of TE-wave are studied and the conditions of existence for the backward mode are obtained. In section 4, the pecularities of the phase and group refractive indices are examined and the problem of reflection at negative refraction is considered.Section 5 contains a summary of the principal results obtained in this paper.

2. Effective permittivity and permeability tensors

Consider a superlaffice (SL) or a periodic multilayer composite which consists of alternating layers of a ferromagnetic insulator (for example, ZnMnO) and nonmagnetic semiconductor (ZnO), in the presence of an external magnetic field B 0 parallel to the plane of the layers. Let’s choose a Cartesian coordinate system with z-axis along B 0 and y-axis along the optical axis of SL, so as the layers are parallel to the xz-plane; the space region y<0 is taken to be a vacuum (see Fig. 1).

 figure: Fig. 1.

Fig. 1. Geometry of the problem. A semiinfinite SL-medium is in the space region y>0; the space y<0 is taken to be a vacuum.

Download Full Size | PDF

An individual semiconductor layer of thickness l 1 is described by dielectric permittivity tensor [11]

ε̂=(εxxεxy0εxyεxx000εzz),

where

εxx=ε(1ωp2ω2ωc2),
εxy=iεωcωp2ω(ω2ωc2),εzz=ε(1ωp2ω2),

ωc=eB 0/m is the cyclotron frequency, ωp=(Ne 2/ 0 ε )1/2 is the plasma frequency, e,m and N are the charge, effective mass and density of the free charge carriers, ε is the high-frequency dielectric constant and ε 0 is the permittivity of the free space. We assume that semiconductor layers contain only one type of free charge carriers, and that the angular frequency ω of EMW is much larger compared to the frequency of the carrier collisions and is far from the excitonic frequencies so as the damping of the waves can be ignored.

In the same coordinate system, the magnetic permeability tensor of an individual ferromagnetic layer of thickness l 2 is given by [12]

μ̂=(μxxμxy0μxyμxx000μzz),

where

μxx=1ω0ωsω2ω02,μxy=iωωsω2ω02,μzz=const,

ωs=γM 0, ω 0=γH 0 is the ferromagnetic resonance frequency, γ is the magnetomechanical ratio and M 0 is the saturation magnetization. For simplicity, we assume that dielectric constant of the ferromagnetic layers is equal to ε .

It is well known that if the period of SL d=l 1+l 2 is much smaller than the wavelength of EMW, the method of effective anisotropic homogeneous medium can be used to obtain the permittivity and permeability of a periodic layered composite [1316]. Using this method, after simple calculations we obtain the constitutive relations for SL, in the form

D=ε0ε̂efE,B=μ0μ̂efH,

where the relative effective permittivity tensor is given by

ε̂ef=(ε1iεa0iεaε2000ε3),
ε1=ε[1ωp12(ω2ωp22)ω2(ω2ωc2ωp22)],εa=εωcωp12ω(ω2ωc2ωp22),
ε2=ε(1ωp12ω2ωc2ωp22),ε3=ε(1ωp12ω2),
ωp12=l1dωp2,ωp22=l2dωp2,

and the relative permeability tensor has a form

μ̂ef=(μ1iμα0iμαμ2000μ3),

where

μ1=ω2ωσ2ω2ω0(ω0+ωs1),μa=ωωs2ω2ω0(ω0+ωs1),
μ2=ω2ω0(ω0+ωs)ω2ω0(ω0+ωs1),μ3=1d(l1+l2μzz),
ωs1=l1dωs,ωs2=l2dωs.ωσ2=ωs1ωs2+ω0(ω0+ωs).

In the following, we shall consider the case of the Voigt configuration, when EMW propagates in the xy-plane perpendicular to the external magnetic field. Then, from the Maxwell equations

rotH=Dt,rotE=Bt,

for the plane waves E,H~exp[i(kxx+kyy-ωt)] we obtain two solutions corresponding to the transverse electric (TE) and transverse magnetic (TM) waves. Dispersion relation for TMwave is given by

ε1kx2+ε2ky2=μ3(ε1ε2εa2)k02,k0ωc,

and practically depends only on the parameters of the effective permittivity tensor. In this wave, the magnetic field H||B 0 and does not produce any interesting effects. That is why, in the following we shall restrict ourselves to the consideration of TE-wave for which the electric field is parallel to B 0 : E={0,0,E}, and the magnetic field is parallel to the xy-plane:

H={iμakx+μ2ky,μ1kx+iμaky,0}(ωμ0μ2μv)1E,

where

μvμ1μa2μ2=ω2ωs32ω2ω0(ω0+ωs)

is the Voigt permeability,

ωs3=γ(H0+M0)(H0+M0l2d1)

corresponds to zero of µv(ω).

3. Regions of propagation and dispersion curves of TE-wave. Conditions of existence for the backward mode

The dispersion relation of TE-wave is given by

μ1kx2+μμky2=ε3μ2μνk02,

αnd describes the coupled magnon-plasmon polaritons with two different frequency regions of propagation:

min{ωr,ωp1}<ω<max{ωr,ωp1},ω>ωs3,

if the effective plasma frequency ωp 1<ωs 3, and

ωr<ω<ωs3,ω>ωp1,

if ωp 1>ωs 3 Here

ωr=(ω02+ω0ωs+ωs1ωs2sin2β)12

is the resonance frequency at which the wave number k →∞,β is the angle between the wave vector k and the+y-axis; ω=ωp 1 and ω=ωs 3 are the cutoff frequencies at which k=0.

The corresponding dispersion curves are shown schematically in Fig. 2, for three different values of the effective plasma frequency: ωp 1<ωr [Fig. 2(a)], ωr<ωp 1<ωs 3 [Fig. (2b)] and ωp 1>ωs 3 [Fig. (2c)].

 figure: Fig. 2.

Fig. 2. Dispersion curves of TE-wave for the cases: a. ωp 1<ωr, b. ωr<ωp 1<ωs 3, c. <ωp 1>ωs 3.

Download Full Size | PDF

One can see that the dispersion of the high-frequency branch is always normal: /dk >0 and the angle φ between the phase and group velocities is acute while the behavior of the low-frequency branch depends on the ratio δ=ωp 1/ωr. For δ<1, the dispersion of the branch is normal while for δ>1, it is anomal: /dk<0 and φ>π/2. To be convinced of such a behavior of the wave, let us examine the sign of the scalar product kS, where S is the time averaged Poynting vector. For TE-wave under consideration, vector S has the following form:

S={μ1μ2kx,ky,0}E22ωμ0μv.

Using Eqs. (10) and (13), we obtain

kS=(μ1kx2+μ2ky2)E22ωμ0μ2μν=12ωε0ε3E2,

according to which the wave is only forward (kS>0, φ<π/2) for ε 3>0, that is, in the frequency region ω>ωp 1. Thus, in the high-frequency regions of the wave propagation in Figs. 2(a,b,c), the SL behaves as a regular right-handed medium.

On the other side, the power flow in the wave is in the backward direction with respect to the phase velocity, that is, kS<0, φ>π/2, only if the following conditions are fulfilled:

ε3<0,μv<0,1+μ1μ21tan2β>0.

After a simple analysis we conclude that conditions (15) can only be fulfilled simultaneously in the frequency range

ωr<ω<min{ωs3,ωp1},

that is, for the low-frequency branchs in Figs. 2(b,c).

Note that conditions (15) are quite different from those in an isotropic case when scalar ε and µ should be negative simultaneously [5]. Indeed, in (15) only one of the components of effective permittivity tensor is to be negative, namely, the diagonal component ε 3 along the direction of the external magnetic field. Also, instead of the scalar µ, the Voigt permeability µv must be negative. Moreover, an additional condition restricting the direction of the backward mode propagation appears. For instance, if µ 1<0 and µ 2>0, the third condition (15) gives tan2 β<µ 2/|µ 1|.

4. Negative refraction and problem of reflection. Phase and group refractive indices

Consider now a TE-type polarized EMW which is incident from the vacuum on the plane surface of the medium, at the angle of incidence α. In this case, the refracted wave is TE polarized too and is described by Eq.(10). According to the causality principle, the energy flux of the wave is always directed away from the interface: Sy>0. Then, using Eq.(13) we conclude that in the region (16), the normal component of the wave vector is negative, that is, the wave vector of the refracted wave is directed to the interface. In other words, with respect to the low-frequency branch, the medium behaves as a left-handed (LH) one.

Using Eq. (14), for the angle φ between the vectors k and S we obtain

cosφ=ε3μ2μvnpngμ1,

Where

np=ckω=sinαsinβ

and

ng=sinαsinγ

are the phase and group refractive indices, β is the angle of refraction and γ is the angle between the group velocity of the refracted wave V g=∂ω/ k (or the Poynting vector S) and the+y-axis. Both np and ng depend on the angle of incidence and are given by

np=[ε3μv+(1μ1μ2)sin2α]12,
ng=[ε3μv(μ2μ1)2+(1μ2μ1)sin2α]12Sgn(μ1μv).

Note that in Eq.(20a) np is positive while after Veselago[5] it is accepted to use negative sign for it. However, a careful examination of the wave refraction problem shows (for the details see Ref.[16]) that the phase refractive index determined as the ratio of the light speed in vacuum to the modulus of the phase velocity (ck/ω) is always positive while the group refractive index can be both positive and negative. The confusion in the sign of np appears because of in an isotropic LHM, one has np=|ng|.

For the backward wave under consideration, ng changes its sign at ω=ωσ, where ωσ is given in Eq.(5c) and corresponds to zero of µ 1(ω). Thus, the frequency region of existence for the backward mode (16) must be divided in two parts:

I.ωr<ω<ωσ

and

II.ωσ<ω<min{ωp1,ωs3},

where the symbols ωσ, ωr, ωp 1 and ωs 3 are key model parameters the meaning of which is explained above [see also Eqs. (4c), (9a) and (12)].

In part I, µ 1<0, Sx>0 and consequently, ng is positive; in other words, SL behaves as a positive index LH-medium [see Fig. 3(a)]. Note that in this case, the refraction occurs at an arbitrary value of the angle of incidence.

 figure: Fig. 3.

Fig. 3. Refraction of TE-wave in the case of positive index (a) and negative index (b) LHM. k 0, k R and k are wave vectors for incident, reflected and refractive waves, respectively.

Download Full Size | PDF

(a)µ 1<0, Sx>0, ng>0. (b) µ 1>0, Sx<0, ng<0. In both cases ε 3,µv<0.

In part ??, we have µ 1>0, Sx<0 and ng<0, that is, SL behaves as negative index LH (or absolute left-handed) medium. The corresponding picture of refraction is shown in Fig. 3(b). In this case, the refraction is only possible for α<αc, where the critical angle of incidence is given by

sinαc=ε3μ2μvμ1.

We have to note again that for both cases shown in Figs. 3(a) and 3(b), the medium is left-handed: the refracted wave vector is directed to the interface, and the angle between the vectors k and S is obtuse: π/2<φ<π. These vectors can only be exactly opposite in part II, and only for α=αc. However, in this case ky=0, the wave is propagated along the interface in the direction parallel to the+x-axis. The main difference between Figs. 3(a) and 3(b) is the different orientation of the Poynting vector with respect to the normal to the interface: for the case ng<0, it is in the same side of the normal as in the incident wave while for the case ng>0 it is in the opposite side, as in a regular medium.

It is important to note the possibility of tuneability of both regions of anomalous refraction (21a) and (21b) by means of external magnetic field variation. Using Eqs. (5c), (9a) and (12) for the characteristic frequencies ωσ,ωs 3 and ωr, it is easy to see that with increasing magnetic field the low frequency region (21a) shifts toward the high frequency side without any change of its width. The tuneability of the negative index region (21b) is more interesting and promising from experimental point of view: for a given value e of ωp 1>ωs 3, the lower limit of the region increases monotonically with increasing magnetic field while the interval of the region increases initially, reaches a maximum value at ωs 3=ωp 1, then decreases rapidly and vanishes at ωσ=ωp 1.

Finally, let us consider briefly the problem of reflection from the surface of LHM. Using the standard continuity conditions for the tangential field components at the interface, we obtain the reflection coefficient for our TE-polarization case in the form

R=(μvcosα+np2sin2α)2+(μaμ2)2sin2α(μvcosαnp2sin2α)2+(μaμ2)2sin2α,

where np is given by Eq.(18a). It is easily to see that the function R(α) has a minimum value Rm at α=αm, where

Rm=BCABC+A,
tan2αm=μ2(Cε3μvB)AB,
A=ε3μ2μvμ1,B=1ε3μ2,C=Aμv+ε3(μ1μv).

For a given value of ω in the range (16), all the parameters A, B,C are positive. The minimum exists only if the condition

μvε3>1+2μa2Aμ2

is fulfilled. Normal component of the refracted wave vector at α=αm is given by

ky=k0CB1cosαm.

Note that in different frequency ranges (21a,b), angle dependence of the function R(α) is different [see Figs. 4(a,b)].

 figure: Fig. 4.

Fig. 4. Angle dependence of the reflection coefficient from the surface of LHM with positive (a) and negative (b) group refractive index.

Download Full Size | PDF

In these figures, the coefficient of reflection at the normal incidence is given by

R0=(ε3μvε3+μv)2,

and the angle αc at which the total reflection occurs, is given by Eq.(22).

5. Conclusions

The obtained results may be summarized as follows. The dispersion, refraction and reflection of electromagnetic waves are studied analytically in ferromagnet-semiconductor multilayer in the presence of an external magnetic field. It is shown in terms of effective medium approach and Voigt configuration that such a multilayer or a superlattice can behave as a left-handed medium for a TE-polarized wave.

General necessary conditions are found for anomalous refraction [see Eq.(15)] which are quite different from those in an isotropic case considered in [5]. It is shown that a backward mode exists only if the effective plasma frequency, at which the wave has a cutoff, exeeds the resonance frequency given in Eq.(12). Unlike the isotropic case, only one of the diagonal components of the effective permittivity tensor, namely, the component along the external magnetic field direction should be negative. Also, instead of the scalar µ(ω) in an isotropic case, the Voigt permeability µv(ω) given in Eq.(9) must be negative. Finally, an additional condition restricting the direction of the refracted wave propagation must be fulfilled.

A frequency range is found in which all the necessary conditions are fulfilled simultaneously. It is shown that the range can be devided in two parts, in one of which (low frequency part) the group refractive index is positive while in the second one it is negative. These parts are separated by the characteristic frequency ωσ given in Eq.(5c). The main difference between these parts is the different orientation of the Poynting vector of the refracted wave with respect to the normal to the interface [see Figs. 3(a) and 3(b)]. In both these regions, the medium is left-handed: the refracted wave vector is directed to the interface and makes an obtuse angle φ>π/2 with the Poynting vector.

It is shown that with increasing magnetic field, the positive index region shifts toward the high-frequency side without any change of its width while the negative index region becomes wider initially, reaches a maximum value and then decreases monotonically. Application of this result is promising from experimental point of view, taking into account the possibility of tuneability of the regions by means of external magnetic field variation.

References and Links

1. R.A. Shelby, D.R. Smith, and S. Schultz., “Experimental Verification of a Negative Index of Refraction,” Science , 292, 77, 2001. [CrossRef]   [PubMed]  

2. P. Markos and C.M. Soukoulis, “Transmission studies of left-handed materials,” Phys. Rev. B 65, 033401 (2002). [CrossRef]  

3. D.R. Smith, S. Schultz, P. Markos, and C.M. Soukoulis, “Determination of effective permittivity and permeability of metamaterilas from reflection and transmission coefficients,” Phys. Rev. B 65, 195104 (2002). [CrossRef]  

4. S.G. Parazzoli, R.B. Gregor, K. Li, B.E. Koltenbah, and M. Tanielian, “Experimental verification and simulation of Negative Index of refraction using Snell’s law,” Phys. Rev. Lett. 90, 107401 (2003). [CrossRef]   [PubMed]  

5. V.G. Veselago, “The electrodynamics of substances with simultaneously negative ε and µ,” Sov. Phys. Usp. 10, 509 (1968) [CrossRef]  

6. J.B. Pendry, “Negative refractions makes a perfect lens,” Phys. Rev. Lett. 85, 3966 (2000). [CrossRef]   [PubMed]  

7. S. Foteinopoulou and C.M. Soukoulis, “Electromagnetic wave propagation in 2D photonic crystals:A study of anomalous refractive effects,” Phys. Rev. B72, 165112 (2005).

8. A. Mandatory, C. Sibilia, M. Bertolotti, S. Zhukovsky, J.W. Haus, and M. Scalora, “Anomalous phase on onedimensional, multilayer, structures with birefringent materials,” Phys. Rev. B70, 165107 (2004).

9. A.S. Raspopin, A.A. Zharov, and H.L. Cui, “Spectrum of electromagnetic excitations in a dc-biased semiconductor superlattice,” J. Appl. Phys. 98, 103517 (2005). [CrossRef]  

10. A. Pimenov, A. Loidl, and P. Przyslupski, “Negative refraction in Ferromagnet-Fuperconductor Superlattices,” Phys. Rev. Lett. 95, 247009 (2005). [CrossRef]   [PubMed]  

11. R.H. Tarkhanyan, “On the Theory of Surface Waves in a Uniaxial Semiconductor Slab,” Phys. Status Sol.(b) 72, 111 (1975). [CrossRef]  

12. D. Polder, “Theory of electromagnetic Resonance,” Philos. Mag. 40, N 300, 99 (1949).

13. V.M. Agranovich and V.E. Kravtsov, “Notes on crystal optics of superlattices,” Sol. St. Commun. 55, 85 (1985). [CrossRef]  

14. R.H. Tarkhanyan and A.G. Nassiopoulou, “Electromagnetic instability of surface waves in semiconductor superlattices,” J. Nanosci. Nanotech.3, 549 (2003), and “Influence of magnetic field on electromagnetic instabilities in semiconductor superlattices,” J. Nanosci. Nanotech4, 1085 (2004). [CrossRef]  

15. D.J. Bergman, X. Li, and Y.M. Strelniker, “Macroscopic conductivity tensor of a three-dimensional composite with a one-or two-dimensional microstructure,”Phys. Rev. B71, 035120 (2005).

16. R.H. Tarkhanyan and D.G. Niarchos, “Wave refraction and backward magnon-polaritons in left-handed antiferomagnetic/semiconductor superlattices,” JMMM (to be published).

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (4)

Fig. 1.
Fig. 1. Geometry of the problem. A semiinfinite SL-medium is in the space region y>0; the space y<0 is taken to be a vacuum.
Fig. 2.
Fig. 2. Dispersion curves of TE-wave for the cases: a. ωp 1<ωr , b. ωr <ωp 1<ωs 3, c. <ωp 1>ωs 3.
Fig. 3.
Fig. 3. Refraction of TE-wave in the case of positive index (a) and negative index (b) LHM. k 0, k R and k are wave vectors for incident, reflected and refractive waves, respectively.
Fig. 4.
Fig. 4. Angle dependence of the reflection coefficient from the surface of LHM with positive (a) and negative (b) group refractive index.

Equations (42)

Equations on this page are rendered with MathJax. Learn more.

ε ̂ = ( ε x x ε x y 0 ε x y ε x x 0 0 0 ε z z ) ,
ε x x = ε ( 1 ω p 2 ω 2 ω c 2 ) ,
ε x y = i ε ω c ω p 2 ω ( ω 2 ω c 2 ) , ε z z = ε ( 1 ω p 2 ω 2 ) ,
μ ̂ = ( μ x x μ x y 0 μ x y μ x x 0 0 0 μ z z ) ,
μ x x = 1 ω 0 ω s ω 2 ω 0 2 , μ x y = i ω ω s ω 2 ω 0 2 , μ z z = const ,
D = ε 0 ε ̂ e f E , B = μ 0 μ ̂ e f H ,
ε ̂ e f = ( ε 1 i ε a 0 i ε a ε 2 0 0 0 ε 3 ) ,
ε 1 = ε [ 1 ω p 1 2 ( ω 2 ω p 2 2 ) ω 2 ( ω 2 ω c 2 ω p 2 2 ) ] , ε a = ε ω c ω p 1 2 ω ( ω 2 ω c 2 ω p 2 2 ) ,
ε 2 = ε ( 1 ω p 1 2 ω 2 ω c 2 ω p 2 2 ) , ε 3 = ε ( 1 ω p 1 2 ω 2 ) ,
ω p 1 2 = l 1 d ω p 2 , ω p 2 2 = l 2 d ω p 2 ,
μ ̂ e f = ( μ 1 i μ α 0 i μ α μ 2 0 0 0 μ 3 ) ,
μ 1 = ω 2 ω σ 2 ω 2 ω 0 ( ω 0 + ω s 1 ) , μ a = ω ω s 2 ω 2 ω 0 ( ω 0 + ω s 1 ) ,
μ 2 = ω 2 ω 0 ( ω 0 + ω s ) ω 2 ω 0 ( ω 0 + ω s 1 ) , μ 3 = 1 d ( l 1 + l 2 μ z z ) ,
ω s 1 = l 1 d ω s , ω s 2 = l 2 d ω s . ω σ 2 = ω s 1 ω s 2 + ω 0 ( ω 0 + ω s ) .
rot H = D t , rot E = B t ,
ε 1 k x 2 + ε 2 k y 2 = μ 3 ( ε 1 ε 2 ε a 2 ) k 0 2 , k 0 ω c ,
H = { i μ a k x + μ 2 k y , μ 1 k x + i μ a k y , 0 } ( ω μ 0 μ 2 μ v ) 1 E ,
μ v μ 1 μ a 2 μ 2 = ω 2 ω s 3 2 ω 2 ω 0 ( ω 0 + ω s )
ω s 3 = γ ( H 0 + M 0 ) ( H 0 + M 0 l 2 d 1 )
μ 1 k x 2 + μ μ k y 2 = ε 3 μ 2 μ ν k 0 2 ,
min { ω r , ω p 1 } < ω < max { ω r , ω p 1 } , ω > ω s 3 ,
ω r < ω < ω s 3 , ω > ω p 1 ,
ω r = ( ω 0 2 + ω 0 ω s + ω s 1 ω s 2 sin 2 β ) 1 2
S = { μ 1 μ 2 k x , k y , 0 } E 2 2 ω μ 0 μ v .
k S = ( μ 1 k x 2 + μ 2 k y 2 ) E 2 2 ω μ 0 μ 2 μ ν = 1 2 ω ε 0 ε 3 E 2 ,
ε 3 < 0 , μ v < 0 , 1 + μ 1 μ 2 1 tan 2 β > 0 .
ω r < ω < min { ω s 3 , ω p 1 } ,
cos φ = ε 3 μ 2 μ v n p n g μ 1 ,
n p = c k ω = sin α sin β
n g = sin α sin γ
n p = [ ε 3 μ v + ( 1 μ 1 μ 2 ) sin 2 α ] 1 2 ,
n g = [ ε 3 μ v ( μ 2 μ 1 ) 2 + ( 1 μ 2 μ 1 ) sin 2 α ] 1 2 Sgn ( μ 1 μ v ) .
I . ω r < ω < ω σ
II . ω σ < ω < min { ω p 1 , ω s 3 } ,
sin α c = ε 3 μ 2 μ v μ 1 .
R = ( μ v cos α + n p 2 sin 2 α ) 2 + ( μ a μ 2 ) 2 sin 2 α ( μ v cos α n p 2 sin 2 α ) 2 + ( μ a μ 2 ) 2 sin 2 α ,
R m = BC A BC + A ,
tan 2 α m = μ 2 ( C ε 3 μ v B ) AB ,
A = ε 3 μ 2 μ v μ 1 , B = 1 ε 3 μ 2 , C = A μ v + ε 3 ( μ 1 μ v ) .
μ v ε 3 > 1 + 2 μ a 2 A μ 2
k y = k 0 C B 1 cos α m .
R 0 = ( ε 3 μ v ε 3 + μ v ) 2 ,
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.