1:
2: \documentstyle[12pt]{article}
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
4: %TCIDATA{OutputFilter=LATEX.DLL}
5: %TCIDATA{LastRevised=Tue Jul 31 21:28:17 2001}
6: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
7: %TCIDATA{Language=American English}
8: %TCIDATA{CSTFile=article.cst}
9:
10: \textwidth=15.5cm
11: \textheight=22cm
12: \topmargin=-1.5cm
13: \def\baselinestretch{1.5}
14: %\input{tcilatex}
15:
16: \begin{document}
17:
18:
19: \begin{center}
20: \vspace*{1cm}{\Large {\bf \ Theoretical investigation on the possibility of
21: preparing left-handed materials in metallic magnetic granular composites }}
22: \medskip \\[0pt]
23: \smallskip S.T.Chui and Liangbin Hu \\[0pt]
24: \smallskip
25:
26: \vspace*{0.2cm} {\it Bartol Research Institute, University of Delaware,
27: Newark, Delaware, U.S.A} \vspace*{0.2cm}
28: \end{center}
29:
30: \centerline{\bf\large Abstract} \vspace*{1cm} We
31: investigate the possibility of preparing left-handed materials in metallic
32: magnetic granular composites. \ Based on the effective medium
33: approximation, \ we show
34: that by incorporating metallic magnetic nanoparticles into an appropriate
35: insulating matrix and controlling the directions of magnetization of
36: metallic magnetic components and their volume fraction, \ it may be possible
37: to prepare a composite medium of low eddy current loss which is
38: left-handed for electromagnetic waves propagating in some special direction
39: and polarization in a frequency region near the ferromagnetic resonance
40: frequency.
41: This composite may be easier to make on an industrial scale. In addition, its
42: physical properties may be easily tuned by rotating the magnetization locally.
43: \medskip
44:
45: PACS numbers: 73.20.Mf, \ 41.20.Jb, 42.70.Qs \vspace*{1cm}
46:
47: \newpage
48:
49: In classical electrodynamics, the response (typically frequency-dependent) of
50: a material to electric and magnetic fields is characterized by two
51: fundamental quantities, \ the permittivity $\epsilon $ and the permeability $%
52: \mu $. \ The permittivity relates the electric displacement field $\vec{D}$
53: to the electric field $\vec{E}$ through $\vec{D}=\epsilon \vec{E}$, \ and
54: the permeability $\mu $ relates the magnetic field $\vec{B}$ and $\vec{H}$ \
55: by $\ \vec{B}=\mu \vec{H}$. \ If we do not take losses into account and
56: treat $\epsilon $ and $\mu $ as real numbers, \ according to Maxwell$%
57: ^{\prime }$s equations, \ electromagnetic waves can propagate through a
58: material only if the index of refraction $n$, \ given by $(\epsilon \mu
59: )^{1/2}$, \ is real. \ \ (Dissipation will add imaginary components to $%
60: \epsilon $ and $\mu $ and cause losses, \ but for a qualitative picture, \
61: one can ignore losses and treat $\epsilon $ and $\mu $ as real numbers. \
62: Also, strictly speaking, $\ \epsilon $ and $\mu $ are second-rank tensors, \
63: but they reduce to scalars for isotropic materials). \ In a medium with$%
64: \epsilon $ and $\mu $ both positive, the index of refraction is real and
65: electromagnetic waves can propagate. \ All our everyday transparent
66: materials are such kind of media. \ In a medium where one of $\epsilon $ and $%
67: \mu $ is negative but the other is positive, \ the index of refraction is
68: imaginary and electromagnetic waves cannot propagate. \ Metals and Earth$%
69: ^{\prime }$s ionosphere are such kind of media. Metals and the ionosphere
70: have free electrons that have a natural frequency-the plasma frequency-which
71: is on the order of 10MHz in the ionosphere and falls at or above visible
72: frequencies for most metals. \ At frequencies above the plasma frequency, $\
73: \epsilon $ is positive and electromagnetic waves are transmitted. \ For
74: lower frequencies, $\ \epsilon $ becomes negative and the index of
75: refraction is imaginary and consequently electromagnetic waves cannot
76: propagate through. \ In fact, the electromagnetic response of metals in the
77: visible and near ultraviolet regions is dominated by the negative epsilon
78: concept[1]-[4].
79:
80: Although all our everyday transparent materials have both positive $\epsilon
81: $ and positive $\mu $, \ from the
82: theoretical point of view, \ in a medium with $%
83: \epsilon $ and $\mu $ both negative, \ the index of refraction is also
84: positive and electromagnetic waves can also propagate through, moreover, \
85: if such media exist, \ the propagation of waves through them should give
86: rise to several peculiar properties. \ This was first pointed out by
87: Veselago over 30 thirty years ago-when no material with simultaneously
88: negative $\epsilon $ and $\mu $ was known[5]. \ For example, \ the cross
89: product of $\vec{E}$ and $\vec{H}$ for a plane wave in regular media gives
90: the direction of both propagation and energy flow, \ and the electric field $%
91: \vec{E}$, \ the magnetic field $\vec{H}$, \ and the wave vector
92: $\vec{k}$ form a
93: right-handed triplet of vectors. \ In contrast, \ in a medium with $\epsilon
94: $ and $\mu $ both negative, \ $\vec{E}\times \vec{H}$ for a plane wave still
95: gives the direction of energy flow, \ but the wave itself (that is, the phase
96: velocity) propagates in the opposite direction, \ i.e., \ wave vector $\vec{k%
97: }$ lies in the opposite direction of $\vec{E}\times \vec{H}$ for propagating
98: waves. \ In this case, \ electric field $\vec{E}$, \ magnetic field $\vec{H}$%
99: , \ and wave vector $\vec{k}$ \ form a left-handed triplet of vectors.
100: Such a medium is therefore termed left-handed medium[5]. \ In addition to
101: this $^{^{\prime }}$left-handed$^{\prime }$ characteristic, \ there are a
102: number of other dramatically different propagation characteristics stemming
103: from a simultaneous change of the signs of $\epsilon $ and $\mu $, \
104: including reversal of both the Doppler shift and the Cerenkov radiation, \ anomalous
105: refraction, \ and even reversal of radiation pressure to radiation tension.
106: \ However, \ although these counterintuitive properties follow directly from
107: Maxwell$^{\prime }$s equations-which still hold in these unusual materials,
108: \ such type of left-handed materials have never been found in nature and
109: these peculiar propagation properties have never been demonstrated
110: experimentally. \ If such media can be prepared artificially, \ they will
111: offer exciting opportunities to explore new physics and potential
112: applications in the area of radiation-material interactions. \ Recently, \
113: interesting progress has been achieved in preparing a $^{^{\prime }}$%
114: left-handed$^{^{\prime }}$ material artificially. \
115: Following the suggestion of Pendry[1],
116: Smith and coworkers reported that a medium made up of an array of
117: conducting nonmagnetic split ring resonators and continuous thin wires can
118: have both an effective negative permittivity $\epsilon $ and negative
119: permeability $\mu $ \ for electromagnetic waves propagating in some special
120: direction and special polarization at microwave frequencies[6]. \ This is
121: the first experimental realization of an artificial preparation
122: of a left-handed material. \ Motivated by this progress, \ in this rapid communication, \
123: we propose to investigate the possibility of preparing left-handed materials
124: in another type of systems-metallic magnetic granular composites. \ The idea
125: is that, \ by incorporating metallic ferromagnetic nanoparticles into an
126: appropriate insulating matrix and controlling the directions of
127: magnetization of \ metallic magnetic particles and their volume fraction, \
128: it may be possible to achieve a composite medium that has simultaneously
129: negative $\epsilon $ and negative $\mu $ and low eddy current loss. \ This
130: idea was based on the fact that on the one hand, \ the permittivity of metallic
131: particles is automatically negative at frequencies less than the plasma
132: frequency, \ and on the other hand, \ the effective permeability of \
133: ferromagnetic materials for electromagnetic waves propagating in some
134: particular direction and polarization can be negative at frequency in the
135: vicinity of the ferromagnetic resonance frequency $\omega _{0},$ which is
136: usually in the frequency region of microwaves. \ So, \ if we can prepare a
137: composite medium in which one component is both metallic and ferromagnetic
138: and other component insulating, \ and we can control the directions of
139: magnetization of \ metallic magnetic particles and their volume fraction, \
140: it may be possible to achieve a left-handed composite medium of low eddy
141: current losses for electromagnetic waves propagating in some special
142: direction and polarization. \
143: This composite may be easier to make on an industrial scale. In addition, its
144: physical properties may be easily tuned by rotating the magnetization locally.
145:
146: To illustrate the above idea more clearly, \ in the following we present
147: results of calculations
148: based on the effective medium theory. \ Let us consider an
149: idealized metallic magnetic granular composite consisting of two types of
150: spherical particles, \ in which \ one type of particles are metallic
151: ferromagnetic grains of radius $R_{1}$, \ the other type are non-magnetic
152: dielectric (insulating) grains of radius $R_{2}$. \ Each grain is assumed to
153: be homogeneous. \ The directions of magnetization of all metallic magnetic
154: grains are assumed to be in the same direction. In
155: length scales much larger than the grain sizes, \ the composite can be
156: considered as a homogeneous magnetic system. \ The permittivity and
157: permeability of non-magnetic dielectric grains are both scalars, and will be
158: denoted as $\epsilon _{1}$ and $\mu _{1}$.\ \ The permittivity of metallic
159: magnetic grains will be denoted as $\epsilon _{2}$ and will be taken to have
160: a Drude form: $\epsilon _{2}=1-\omega _{p}^{2}/\omega (\omega +i/\tau )$, \
161: \ where $\omega _{p}$ is the plasma frequency of the metal and $\tau $ is a
162: relaxation time. \ Such a form of $\epsilon $ is representative of a variety
163: of metal composites[8]-[9]. \ The permeability of metallic magnetic grains
164: are second-rank tensors and will be denoted as $\hat{\mu}_{2},$ \ which can
165: be derived from the Landau-Lifschitz equations[7]. \ Assuming that the
166: directions of magnetization of all magnetic grains are in the direction of \
167: the $z$-axis, \ $\hat{\mu}_{2}$ will have the following form[7]:
168: \begin{equation}
169: \hat{\mu _{2}}=\left[
170: \begin{array}{cccc}
171: \mu _{a} & -i\mu ^{\prime } & 0 & \\
172: i\mu ^{\prime } & \mu _{a} & 0 & \\
173: 0 & 0 & 1 &
174: \end{array}
175: \right]
176: \end{equation}
177: where
178: \begin{eqnarray}
179: \mu _{a} &=&1+\frac{\omega _{m}(\omega _{0}+i\alpha \omega )}{(\omega
180: _{0}+i\alpha \omega )^{2}-\omega ^{2}}, \\
181: \mu ^{\prime } &=&-\frac{\omega _{m}\omega }{(\omega _{0}+i\alpha \omega
182: )^{2}-\omega ^{2}},
183: \end{eqnarray}
184: $\omega _{0}=\gamma \vec{H_{0}}$ is the ferromagnetic resonance
185: frequency, $\ H_{0}$ is the effective magnetic field in magnetic particles
186: and may be a sum of the external magnetic field, \ the effective anisotropy
187: field and the demagnetization field; $\ \omega _{m}=\gamma \vec{M_{0}}$, $\ $%
188: where $\gamma $ is the gyromagnetic ratio, $\ M_{0}$ is the saturation
189: magnetization of magnetic particles; \ $\alpha $ is the magnetic damping
190: coefficient; $\ \omega $\ is the frequency of incident electromagnetic
191: waves. \ We shall only consider incident electromagnetic waves propagating in
192: the direction of the magnetization. \ This is the most interesting case in the
193: study of magneto-optical effects in ferromagnetic materials. \ We also
194: assume that the grain sizes are much smaller compared with the characteristic
195: wavelength $\lambda ,$ and consequently, electromagnetic waves in the
196: composite can be treated as propagating in a homogeneous magnetic system. \
197: According to Maxwell$^{^{\prime }}$s equations, \ electromagnetic waves
198: propagating in the direction of magnetization in a homogeneous magnetic
199: material is either positive or negative transverse circularly polarized. \
200: If \ the composite can truly be treated as a homogeneous magnetic system in
201: the case of grain sizes much smaller than the characteristic wavelength, \
202: electric and magnetic fields in the composite should also be either positive
203: or negative circularly polarized and can be expressed as :
204: \begin{eqnarray}
205: \vec{E}(\vec{r},t) &=&\vec{E}_{0}^{(\pm )}e^{ikz-\beta z-i\omega t} \\
206: \vec{H}(\vec{r},t) &=&\vec{H}_{0}^{(\pm )}e^{ikz-\beta z-i\omega t}
207: \end{eqnarray}
208: where $\vec{E}_{0}^{(\pm )}=\hat{x}\mp i\hat{y}$, $\ \vec{H}_{0}^{(\pm )}=%
209: \hat{x}\mp i\hat{y}$, $\ k={\rm Real}[k_{eff}]$ is the effective wave
210: number, $\ \beta ={\rm Im}[k_{eff}]$ is the effective damping coefficient
211: caused by the eddy current, $\ k_{eff}=k+i\beta $ is the effective propagation
212: constant. $\ $In Eqs.(4)-(5) the signs of $k$ and $\beta $ can both be
213: positive or negative depending on the directions of the wave vector and the
214: energy
215: flow. $\ $For convenience we assume that the direction of energy flow is in
216: the positive direction of the $z$ axis, \ i.e., \ we assume $\beta >0$ in
217: Eqs.(4)-(5)$,$ but the sign of $k$ still can be positive or negative. \ In
218: this case, \ if $k>0$, \ the phase velocity and energy \ flow are in the
219: same directions, \ and \ from Maxwell$^{^{\prime }}$s equation, \ one can
220: see that the electric and magnetic field $\vec{E}$ and $\vec{H}$ and the
221: wave vector
222: $\vec{k}$ will form a right-handed triplet of vectors. \ This is the usual
223: case for right-handed materials. In contrast, \ if $k<0,$ the phase velocity
224: and energy \ flow are in opposite directions, \ and $\vec{E}$, $\ \vec{H}
225: $ and $\vec{k}$ will form a left-handed triplet of vectors. \ This is just
226: the peculiar case for left-handed materials. \ So, \ for incident waves of a
227: given frequency $\omega ,$ we can determine whether wave propagations in the
228: composite is right-handed or left-handed through the relative sign changes
229: of $k$ and $\beta $. \ In the following, \ we shall determine the effective
230: propagation constant $k_{eff}=k+i\beta $ by means of the effective
231: medium approximation. \ The details of \ various kinds of effective medium
232: approximations have been discussed in a series of references[8]-[12], \ here
233: we only list the main points. \ First, \ if the composite can
234: truly be considered as a homogeneous magnetic system in the case of grain
235: sizes much smaller than the characteristic wavelength, \ then for waves
236: (positive or negative circularly polarized) propagating through the composite
237: in the direction of \ magnetization, \ \ their propagations can be described
238: by an effective permittivity $\epsilon _{eff}$ and an effective permeability
239: $\mu _{eff}$, \ which satisfy the following relations
240: \begin{eqnarray}
241: \int \vec{D}(\vec{r},\omega )e^{ik_{eff}z}d\vec{r} &=&\epsilon _{eff}\int
242: \vec{E}(\vec{r},\omega )e^{ik_{eff}z}d\vec{r}, \\
243: \int \vec{B}(\vec{r},\omega )e^{ik_{eff}z}d\vec{r} &=&\mu _{eff}\int \vec{h}(%
244: \vec{r},\omega )e^{ik_{eff}z}d\vec{r},
245: \end{eqnarray}
246: where $k_{eff}$ and $\omega $ are related by $k_{eff}=\omega \lbrack
247: \epsilon _{eff}\mu _{eff}]^{1/2}$. \ \ Although these relations are simple
248: and in principle exact, \ it is very difficult to calculate the integrals in
249: them because the fields in the composite are spatially varying in a random
250: way. \ \ One therefore must resort to various types of approximations. \ The
251: simplest approximation is the effective medium approximation. \ In this
252: approximation, \ we calculate the fields in each grain as if the grain were
253: embedded in an effective medium of dielectric constant $\epsilon _{eff}$ \
254: and magnetic permeability $\mu _{eff}$. \ Consider, \ for example, \ the $i$%
255: th grain. Under the embedding assumption, \ the electric and magnetic fields
256: incident on the grain are the form of Eqs.(4)-(5):
257: \begin{eqnarray}
258: \vec{E}_{inc} &=&\vec{E}_{0}^{(\pm )}e^{ik_{eff}z-i\omega t}, \\
259: \vec{h}_{inc} &=&\vec{h}_{0}^{(\pm )}e^{ik_{eff}z-i\omega t},
260: \end{eqnarray}
261: where $\vec{E}_{0}^{(\pm )}=\hat{x}\mp i\hat{y}$ and $\vec{h}_{0}^{(\pm )}=%
262: \hat{x}\mp i\hat{y},$ \ corresponding to the positive($+$) or negative($-$)
263: circularly polarized waves. \ If the fields inside the grain can be found, \
264: then the inside fields can be used to calculate the integral over the grain
265: volume
266: \begin{eqnarray}
267: \vec{I}_{i} &=&\int_{v_{i}}\vec{E}_{i}(\vec{r},\omega )e^{ik_{eff}z}d\vec{r},
268: \\
269: \vec{J}_{i} &=&\int_{v_{i}}\vec{h}_{i}(\vec{r},\omega )e^{ik_{eff}z}d\vec{r},
270: \end{eqnarray}
271: which is required to find the integral in Eqs.(6)-(7). \ For the positive or
272: negative circularly polarized incident waves described by Eqs.(8)-(9), \ the
273: integral $\vec{I}_{i}$ and $\vec{J}_{i}$ can be written as
274: \begin{eqnarray}
275: \vec{I}_{i} &=&(\hat{x}\mp i\hat{y})I_{i}, \\
276: \vec{J}_{i} &=&(\hat{x}\mp i\hat{y})J_{i},
277: \end{eqnarray}
278: where $I_{i}$ and $J_{i}$ are scalars. \ If $I_{i}$ and $J_{i}$ can be
279: found, \ then from Eqs.(6)-(7), \ the effective permittivity $\epsilon
280: _{eff} $ \ and effective permeability $\mu _{eff}$ \ can be calculated by
281: \begin{equation}
282: \epsilon _{eff}=\frac{f_{1}\epsilon _{1}I_{1}+f_{2}\epsilon _{2}I_{2}}{%
283: f_{1}I_{1}+f_{2}I_{2}},
284: \end{equation}
285: \begin{equation}
286: \mu _{eff}=\frac{f_{1}\mu _{1}J_{1}+f_{2}\mu _{2}^{(\pm )}J_{2}}{%
287: f_{1}J_{1}+f_{2}J_{2}},
288: \end{equation}
289: where $f_{1}$ and $f_{2}$ are the volume fractions of the two types of
290: grains, $\ \mu _{1}$ is the permeability of \ non-magnetic dielectric
291: grains, $\mu _{2}^{(+)}=\mu _{a}-\mu ^{\prime }$ and $\mu ^{(-)}=\mu
292: _{a}+\mu ^{\prime }$ (see Eqs.1-3) are the effective permeability of
293: magnetic grains for positive and negative circularly polarized waves
294: respectively. \ As for the calculation of $I_{i}$ and $J_{i},$ \ we can
295: follow the method of expanding interior and exterior fields in a multipole
296: series and matching the boundary conditions[13]. \ After the coefficients of
297: the multipole expansion of interior and exterior fields are obtained by
298: matching the boundary conditions, \ $I_{i}$ and $J_{i}$ can be found
299: and subsequently be substituted into Eqs.(14)-(15). \ Since this method is
300: standard, \ we shall not present the details.\ \ In the final results,
301: Eqs.(14)-(15) reduce to one self-consistent equation:
302: \begin{eqnarray}
303: &&\sum_{i=1,2}f_{i}\sum_{l=1}^{\infty }(2l+1)[\frac{k_{eff}\psi _{l}^{\prime
304: }(k_{i}R_{i})\psi _{l}(k_{eff}R_{i})-k_{i}\psi _{l}(k_{i}R_{i})\psi
305: _{l}^{\prime }(k_{eff}R_{i})}{k_{eff}\psi _{l}^{\prime }(k_{i}R_{i})\zeta
306: _{l}(k_{eff}R_{i})-k_{i}\psi _{l}(k_{i}R_{i})\zeta _{l}^{\prime
307: }(k_{eff}R_{i})} \nonumber \\
308: &&+\frac{k_{i}\psi _{l}^{\prime }(k_{i}R_{i})\psi
309: _{l}(k_{eff}R_{i})-k_{eff}\psi _{l}(k_{i}R_{i})\psi _{l}^{\prime
310: }(k_{eff}R_{i})}{k_{i}\psi _{l}^{\prime }(k_{i}R_{i})\zeta
311: _{l}(k_{eff}R_{i})-k_{eff}\psi _{l}(k_{i}R_{i})\zeta _{l}^{\prime
312: }(k_{eff}R_{i})}]=0,
313: \end{eqnarray}
314: where $R_{i}$ is the radius of the $i$th type of grains, \ and
315: \begin{eqnarray}
316: k_{1} &=&\omega \lbrack \epsilon _{1}\mu _{1}]^{1/2}, \\
317: k_{2} &=&\omega \lbrack \epsilon _{2}\mu _{2}^{(\pm )}]^{1/2}, \\
318: \psi _{l}(x) &=&xj_{l}(x), \\
319: \zeta _{l}(x) &=&xh_{l}^{(1)}(x),
320: \end{eqnarray}
321: $j_{l}(x)$ and $h_{l}(x)$ are the usual spherical Bessel and Hankel
322: functions. \ Eq.(16) determine the effective product of $(\epsilon \mu
323: )_{eff},$ or equivalently $k_{eff},$ but not a single $\epsilon _{eff}$ and $%
324: \mu _{eff}.$ \ \ It can describe the change of the phase of a plane wave
325: across a slab of the composite, \ but it does not precisely describe wave
326: propagations across a slab of the composite. \ This is due to the fact we
327: make no attempt to rigorously solve the boundary-value problem for a slab of
328: composite by matching the fields inside the slab and external fields outside
329: the slab at the boundary. \ In fact, \ it is common in various types of
330: effective medium theories that for $\omega \neq 0$ the electromagnetic
331: properties of a composite cannot in general be specified by a single $%
332: \epsilon _{eff}$ and $\mu _{eff}.$ \ Since we can determine whether wave
333: propagations through the composite is left-handed or right-handed by the
334: calculation of the effective propagation constant $k_{eff},$ \ Eq.(16) is
335: enough for the problems we are discussing.
336:
337: The numerical results for a metal volume fraction $f_2$ of 0.3
338: obtained from Eq.(16) are summarized in Fig.1-Fig.2. \
339: Fig.1(a) shows the frequency dependence of the real part of the effective
340: permeability $\mu ^{(+)}$ of magnetic grains for positive circularly
341: polarized plane waves, \ Fig.1(b) and (c) show the corresponding frequency
342: dependences of the effective wave number $k$ and the effective damping
343: coefficient $\beta $ in a composite consisting of metallic magnetic grains
344: and dielectric grains. \ From Eqs.(1)-(3), \ we can see that if the magnetic
345: damping coefficient $\alpha $ is zero, $\ {\rm Re}[\mu ^{(+)}]$ will be
346: negative in the whole frequency region of $\omega >\omega _{0}($the magnetic
347: resonance frequency). \ From Fig.1(a), \ we can see that if $\alpha $ is
348: nonzero but small enough, \ there can still be a frequency region
349: near $\omega _{0}$ in which ${\rm Re}[\mu ^{(+)}]$ is negative. \ In this
350: case, \ if the amplitude of the negative $\mu ^{(+)}$ is large enough, \ \ $%
351: k $ will be negative in this frequency region \ as was shown in
352: Fig.1(b), and hence the phase velocity and energy \ flow will be in the
353: opposite directions in this frequency region, \ and $\vec{E}$, $\
354: \vec{H}$ and $\vec{k}$ will form a left-handed triplet of vectors, \ i.e.,
355: the composite will be left-handed in this frequency region for
356: positive circularly polarized plane waves. \ But if $\alpha $ is not small
357: enough, $\ {\rm Re}[\mu ^{(+)}]$ will be positive in the whole frequency
358: region, \ or though $Re[\mu ^{(+)}]$ is negative in a frequency
359: region near $\omega _{0}$, \ the amplitude of the negative $Re[\mu ^{(+)}]$
360: is not large enough, \ in this case $k$ will be positive in the whole
361: frequency region, \ as was shown in Fig.1(b). \ In this case, \ the
362: composite is right-handed for\ positive circularly polarized waves in the
363: whole frequency region. \ The calculations also show that if the radius of
364: metallic grains are small enough and the volume fraction of metal components
365: is smaller than the threshold value of the insulator-metal transition, \
366: which is approximately $1/3$ \ in our model, \ the losses caused by eddy
367: current are very small and the composite is essentially an insulator. This
368: can be seen from Fig.1(c), \ in which the damping coefficient $\beta $ is
369: very small compared with the amplitude of the wave number $k,$ \ i.e., \ the
370: eddy current losses are very small \ in the cases shown in Fig.1. If the
371: volume fraction of metal components is larger than the threshold value, \
372: the composite will be essentially a metal, \ and the damping coefficient $%
373: \beta $ will be much larger than the amplitude of wave number $k$(not shown
374: in the figure). \ In Fig.2(a) we show the frequency dependence of the real
375: part of the effective permeability $\mu ^{(-)}$ of magnetic grains for
376: negative circularly polarized waves, \ and in Fig.2(b) we show the
377: corresponding frequency dependence of the effective wave number $k$ in a
378: composite consisting of the metallic magnetic grains and dielectric grains.
379: We can see that for negative circularly polarized waves, $\ {\rm Re}[\mu
380: ^{(-)}]$ is positive in the whole frequency region no matter how small $%
381: \alpha $ is, \ and correspondingly, \ $k$ is positive in the whole frequency
382: region, \ i.e., the composite is right-handed in the whole frequency region
383: for negative circularly polarized waves no matter how small $\alpha $ is.
384:
385: In conclusion, \ we have discussed the possibility of preparing a
386: left-handed material in metallic magnetic granular composites based on the
387: effective medium approximation.
388: Our model analysis shows that, \ by incorporating
389: metallic magnetic nanoparticles into an appropriated insulating matrix and
390: controlling the directions of magnetization of metallic magnetic components
391: and their volume fraction, \ it may be possible to prepare a composite
392: medium of low eddy current losses which is left-handed for electromagnetic
393: waves propagating in some special direction and polarization in a
394: frequency region near the magnetic resonance frequency.
395: %\ Though this simple
396: %model analysis shows that there exists the possibility of preparing a
397: %left-handed medium in metallic magnetic granular composites, \ from the
398: %experimental point of view, \ there are a lot of doubts on this possibility
399: %because the conditions assumed in the theoretical calculations may be rather
400: %difficult to be satisfied in real experiments. \ Of course, \ the validity
401: %of the simple effective medium approximation applied in this paper may also
402: %be a question.
403: \ Further theoretical investigations by
404: %improved effective medium type of \ theories and
405: other approximations such as first-principle numerical calculations may be
406: needed to further confirm the possibility shown in this paper.
407:
408: S. T. Chui was supported in part by the Office of Naval Research and by the
409: Army Research Laboratory through the Center of Composite Materials at the
410: University of Delaware. We thank John Xiao for helpful discussions.
411:
412: \begin{thebibliography}{10}
413: \bibitem[1]{1} J.B.Pendry, A.J.Holden, W.J.Stewart, and I.Youngs, {\it %
414: Phys. Rev. Lett} {\bf 76}, 4773(1996)
415:
416: \bibitem[2]{2} R.H.Ritchie and A.Howie, {\it Phil. Mag. A } {\bf 44},
417: 931(1981)
418:
419: \bibitem[3]{3} T.L.Ferrell and P.M.Echenique, {\it Phys. Rev. Lett} {\bf 55},
420: 1526(1985)
421:
422: \bibitem[4]{4} A.W.Howie and C.A.walsh, {\it %
423: Microsc.Microanal.Microstructure.{\bf 2}}, 171(1991)
424:
425: \bibitem[5]{5} V.G.Veselago, {\it Sov.Phys.Usp. {\bf 10}}, 509 (1968)
426:
427: \bibitem[6]{6} D.R.Smith, W.J.Padilla, D.C.Vier, S.C.Nemet-Nasser,
428: S.Schultz, {\it Phys. Rev. Lett.} {\bf 67}, 3578(2000)
429:
430: \bibitem[7]{7} C.P.Slichter, {\it Principle of Magnetic Resonance},
431: Springer-Verlag Berlin, Heidelberg(1978)
432:
433: \bibitem[8]{8} R.Burridge, S.Childress, and G.papanicolaou(edition), {\it %
434: Macroscopic Properties of Disordered Media}, Springer-Verlag Berlin,
435: Heidelberg(1982)
436:
437: \bibitem[9]{9} J.C.Garland and D.B.Tanner(edition), {\it Electrical
438: Transport and Optical Properties of Inhomogeneous Media}, American Institute
439: of Physics, New York(1978)
440:
441: \bibitem[10]{10} D.Stroud and F.P.Pan, \ {\it Phys. Rev.B} {\bf 20}, 455(1979)
442:
443: \bibitem[11]{11} P.Sheng, {\it Phys. Rev. Lett} {\bf 45} , 60(1980)
444:
445: \bibitem[12]{12} W.Lamb, D.M.Wood, and N.W.Ashcroft, {\it Phys. Rev.B} {\bf 21%
446: } , 2248(1980)
447:
448: \bibitem[13]{13} H.C.Van de Hulst, {\it Light scattering by Small particles}%
449: (Chapter 9), \ Dover Publications Inc., \ New York(1981); \ \ J.D.Jackson, \
450: {\it Classical electrodynamics, \ }2nd ed., Wiley, \ New York(1975)
451: \end{thebibliography}
452:
453: {\Large \newpage }
454:
455: \begin{center}
456: {\normalsize Caption }
457: \end{center}
458:
459: {\normalsize \bigskip }
460:
461: {\normalsize \noindent Fig.1. (a) The frequency dependencies of the
462: effective permeability }$\mu ^{(+)}${\normalsize \ of magnetic grains and
463: the corresponding frequency dependencies of (b) the effective wave number }$%
464: k ${\normalsize \ and (c) the effective damping coefficient }$\beta $%
465: {\normalsize \ of the composite for positive circularly polarized waves
466: propagating in the direction of magnetization. \ (The plasma frequency }$%
467: \omega _{p}${\normalsize \ is usually in the visible or ultraviolet
468: frequency region and the ferromagnetic resonance frequency }$\omega _{0}$%
469: {\normalsize \ is usually in the microwave frequency region. For simplicity,
470: \ hereafter we will set }$\omega _{0}/\omega _{p}=10^{-5}${\normalsize . \
471: The other parameters are: \ }$\omega _{m}/\omega _{0}=4.0${\normalsize , \ }$%
472: \omega _{p}R/c=0.2${\normalsize , \ }$f_{2}=0.3${\normalsize , \ \ }$\alpha $%
473: {\normalsize \ is shown in the figures). \newline
474: \newline
475: Fig.2. (a) The frequency dependence of the effective permeability }$\mu
476: ^{(-)}${\normalsize \ of magnetic grains and the corresponding frequency
477: dependencies of (b) the effective wave number }$k${\normalsize \ and (c) the
478: effective damping coefficient }$\beta ${\normalsize \ of the composite for
479: negative circularly polarized waves propagating in the direction of
480: magnetization. (The parameters are: }$\omega _{m}/\omega _{0}=4.0$%
481: {\normalsize , }$\omega _{p}R/c=0.2${\normalsize , \ }$f_{2}=0.3$%
482: {\normalsize , \ }$\alpha ${\normalsize \ is shown in the figures.) \newline
483: \newline
484: }
485:
486: \end{document}
487: