See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/259298244

General laws of reflection and refraction for subwavelength phase grating

Article · December 2013

Source: arXiv

CITATIONS
READS
5
5,393

4 authors, including:

Zhaona Wang
Da he Liu
Beijing Normal University
Beijing Normal University
80 PUBLICATIONS 2,005 CITATIONS
217 PUBLICATIONS 4,170 CITATIONS

Some of the authors of this publication are also working on these related projects:

photodetector View project

Nanoparticles with abundant of nanogaps for high sensitivity SERS detecting and high performance random lasers View project

All content following this page was uploaded by Da he Liu on 06 June 2014.

The user has requested enhancement of the downloaded file.

 General laws of reflection and refraction for subwavelength phase
grating
Zhaona Wang 1*, Yanyan Sun 1, Lu Han 1, Dahe Liu 1,2 *
1 Applied Optics Beijing Area Major Laboratory, Department of
Physics, Beijing Normal University, Beijing 100875
2 Key Laboratory of Nondestructive Test (Ministry of Education), Nanchang Hangkong University,
Nanchang 330063, China.
* Corresponding authors: zhnwang@bnu.edu.cn for Zhaona Wang and dhliu@bnu.edu.cn for Dahe Liu

ABSTRACT

The general reflection and refraction laws at the metasurface with the abrupt phase

shift were derived by two different methods of Fermat's principle and the boundary

conditions respectively. It is found that one or two critical angles for total internal reflection

exist when a light hits on the optical sparse material from the optical denser material, and

one such a critical angle exists when light spreads from the optical sparse material to the

optical denser material. Anomalous reflection and refraction, such as, negative reflection

and negative refraction may occur when a light passes through the metasurface, and the

conditions of their occurrence were given. Finally, a kind of metasurface based on one-

dimensional phase mask was designed to control the light propagation.

OCIS Codes: Metasurface; refraction law; reflection law; phase discontinuity

1. Introduction

The laws of reflection and refraction are the basis of geometrical optics. The behavior of a light beam at

the interface is completely determined by the optical properties of the two materials at different side of the

interface. In classical optics, it is assumed that the physical interface has an ideal boundary which does not

change any information of light waves.

With the recent development of nanofabrication technology, metamaterials [1-3] were used to control the

electromagnetic field, with which some properties that may not be found in nature can be obtained, such as,

negative refraction index [4-6], ultra-magnetic property [7], low-loss property [8] and so on. Metasurface, as an

important branch of metamaterials, has been used to achieve arbitrary control of light waves due to its special

property, for example, to control the polarization state [9], to achieve completely transmission [10], to improve

the antenna radiation performance [11-13], to obtain super absorption [14].

1
 Different from the ideal classical boundary, metasurface can change the wavefront, the phase and the

polarization state of a light wave, and the propagation of a light wave at the interface may violate the

conventional reflection and refraction laws [15,16]. In order to further understand the behavior of light at the

metasurface and to further design the devices with metasurface, it is necessary to derivate the general laws of

reflection and refraction at the metasurface which can generate the abrupt phase shift for a light wave. In the

present work, the general laws of reflection and refraction were derived by the two methods based on Fermat's

principle and the continuous boundary conditions, respectively. The relevant factors influencing light

propagation were discussed. The full transmission critical condition, the total reflection critical condition, the

conditions of abnormal reflection and abnormal refraction were also demonstrated. Finally, a metasurface made

by a one-dimensional phase mask was designed to control the propagation of light.

2. Derivation from Fermat's principle

Fermat's principle is the basic principle of optics. It should be obeyed in transformation optics [17,18]

although one needs to explore unusual geometries of space-time in the early universe [19]. Fermat’s principle

states that the ray trajectory of light between two points A and B is that of the extreme value of optical path, i.e.

B r
δ  n(r )dr  0 , n(r ) is the local refractive index. For a light wave, changing the optical path means
A
where

actually changing the phase of the light wave. Therefore, Fermat’s principle can also be stated as
B r
(r ) B
r
A
 0 , the principle of stationary phase [20–22], that is to say, the total phase (r) ) accumulated
δ  A
dϕ

r
along the actual light path from point A to point B will be a constant. If an abrupt phase shift (r )
ϕ s depending

on the coordinate rs along the interface over the scale of the wavelength is introduced, then, the total phase

shift (rurs )  B rk  dr r will be stationary, (r

r s ) respects the phase mutation caused by the metasurface at the
A

 B r r
point rs ,  A
k  d r is the phase change of light wave accumulated by gradual phase changes along the optical

path, k is the wave vector of the propagating light.

Assume that the phase shift gradient along the interface d / is a nonzero constant, the general
dx

reflection law derived from Fermat’s principle had been given in [15]:
λ d
n sin(θ )  n sin(θ )  0
. (1)
t t i i
2π dx

where θi
is the incident angle, θt is the refractive angle; ni and nt are the refractive indices of the incident
2
medium and the output medium, respectively.
λ0 is the wavelength of light in vacuum. From Eq. (1), the

3
 refracted beam can be in arbitrary direction by introducing a suitable phase discontinuity d / dx along the

interface. Because of the nonzero phase shift gradient, the two incident angles θi will lead to different

refractive angle. That is to say, the symmetry for the incident angle will be destroyed. The critical angle for

total internal reflection will be:

nt λ0 d
θ  arcsin(  ) (2)
tc
n 2π n dx
i i

From formula (2), it can be found that the critical angle for total internal reflection is not only related to

the refractive indices of the two materials and the wavelength of light in vacuum, but also to the phase shift

gradient at the interface. It means that, when the light wave is incident on the metasurface from the optical

denser material to the optical sparse material (i.e. nt  ni ), there will be two critical angles for total internal
d d 2π
reflection with the condition of  nt λ0  1 ,  (n  n ) ) and one critical angles for

(i.e.
n 2π n dx dx t i
λ
i i 0
d 2π
total reflection with the condition of  (n  n ) . It is different from the prediction in Ref. [15] in
dx t i
λ
0

which there always exist two critical angles. Meanwhile, quite different from the classical refraction law, the

general refraction law also allows the critical angle for total internal reflection when light beam is incident from

the optical sparse material to the optical denser material (i.e. nt  ni ) with the conditions of
d 2π
 (n  n ) .
dx t i
λ
0

Fig. 1. Schematics used to derive the generalized Snell’s law of reflection. The interface between the two media is artificially
structured in order to introduce an abrupt phase shift in the light path, which is a function of the position along the interface.  and
  d are the phase shifts where the two paths (blue and red) cross the boundary.
Now, we derived the general reflection law. Considering a plane wave with a incident angle of θ ,
i

assuming that the two paths ACE and ADE are infinitesimally close to the actual path between the points A and

E (See Fig. 1), then the phase difference between them should be zero.

k0ni sin(θi )dx  (  d)  k0ni sin(θr )dx    (3)

0

4
where θr
is the reflected angle, respectively,    d are the phase discontinuities at the locations
and

5
where the two paths just cross the interface, dx is the distance between the two cross points, and the wave

number k0  2π / λ0 . Assuming the phase shift gradient along the interface d / dx is chosen to be a nonzero

constant, the Eq. (3) will leads to the general reflection law
λ0 d
sin θ   sin θ   . (4)
r 2π ni dx

Equation (4) predicts that the reflected angle changes nonlinearly with the change of the incident angle. It is

obviously different from the result of the classical reflection law. It should be addressed that there is a special

incident angle, when the incident angle is larger than it the reflected light will disappear. This behavior is

similar to the total internal reflection induced by refraction law. This critical angle can be defined as the critical

angle for full transmission

λ0 d
θ  arcsin(1  ) . (5)
r 2π ni dx

From formula (5), it can be seen that the critical angle for full transmission always exists at the

metasurface with an abrupt phase shift regardless of the both case that the light wave hits on the optical sparse

material from the optical denser material (i.e. nt  ni ), and vice versa (i.e. nt  ni ). The critical angle for full

transmission is determined by the refractive index ni of incident material, the wavelength of incident light

λ0 and the phase shift gradient

d / dx at the interface.

3. Derivation from the boundary conditions continuity

Although the general reflection and refraction laws have been derived, the interaction mechanism between

light waves and the metasurface is still an open question. Whether the boundary continuity conditions are

satisfied for the metasurface? Now, we make discussions.

z
ki
θ kr
i θr

O x
θt kt

Fig. 2. Schematics used to derive the generalized laws of reflection and refraction. There exists the interface between the two media to
introduce an abrupt phase (r to the light wave, which is a function of the position along the interface.
shift )
Considering the case, an incident plane wave is in the plane of xoz (See Fig. 2)，the x axis is just at the

interface of the two substances, the incident wave vector is

ki , the incident angle is θi , and the wave vectors

of the reflected wave and refracted wave are k and k respectively, the reflected angle and the refracted
r t

6
angle are θr
and θt , respectively. Assuming that the tangential components of the electric field is continuous,

7
then we have

Eix  Erx  Etx (6)

iωit ki (lixmi yniz ) iωrt kr (lrxmr y nrz )( x ) iωttkt (lt xmt y nt z )( x
where Eix  Aix e Erx  Arx e )
respond the x
Etx  Atx

components of the electric fields of the incident ， reflected and refracted waves ， respectively. And

lα , mα , nα (α  i, r, t)
is three directional cosine of the wave vectors ki , kr and kt , respectively.

At the interface ( z  0 ), the incident wave vector can be denoted

as
ki  ki (sin θi ,0,cosθi ) . The boundary

conditions should be satisfied at any time t and at any point (x, y, z) , so we

have mr  mt  0 . Then, the

following relation could be obtained:

kili x  kr lr x  ( x)  ktlt x  ( (7)

x)
Substituted
lr  sin θ r , lt  sin θt into Eq. (7), we have
ω ω ω
n sin(θ )x  n sin(θ )x  ( x)  n sin(θ )x  ( x)
. (8)
i i i r t t
c c c

At the interface, Eq. (8) can be expressed as

ω ω d(x) ω d(x)
n sin(θ )  n sin(θ )   n sin(θ ) 
. (9)
c
i i
c
i r dx c t t
dx

Therefore, the general reflection law and the general refraction law identical to Eqs. (1) and (4) can be

obtained.

 λ0
sinθ  sinθ  d( x reflection law
)

r i
2π ni dx （10）
n sinθ  n sinθ
λ0 d(x) refraction law


t t i i 2π dx

It can be seen that the general reflection and refraction laws derived from Fermat's law and boundary

conditions continuity for the interface with an abrupt phase shift are exactly the same. It means that the

boundary conditions continuity is still valid in the interaction between light waves and metasurface.

4. Analysis and discussion satisfies

It should be pointed out that when the phase shift gradient

8
d/dx=0 , Eq. (10) turn into

θi  θr

n reflection law
 （11）
θ  arcsin( sinθ )
i
refraction law
 t i
 nt

It is the classical reflection and refraction laws. That is to say, the classical reflection and refraction laws are the

special cases of the general reflection and refraction laws.

9
 (a) (b)

(c) (d)
Fig. 3 The angles of reflection((a),and (c)) and refraction ( (b),and (d)) governed by the traditional laws of reflection and refraction
(blue solid line) and those given by the generalized laws of reflection and refraction (dotted line) change with the angle of incidence,
when the light transmits from the optical sparse media into the optical denser medium ( (a),and (b)) and the light incidents from the
optical denser
medium to the optical sparse medium ( (c) and (d)), with the d / dx  π / 1.5107 m-1 ( green dotted line)，d / dx  π / 3 10 m
7 -1

condition
（cyan dotted line）， d  / dx  π / 6 10 7 m -1 （magenta dotted line）， d / dx  π / 9 107 m -1 （red dotted line）.

By introducing the abrupt phase shift at the interface of the two materials, the reflected and refracted angles

are related to the incidence angle, the refractive index of the two materials, the wavelength of incident light,

and the phase shift gradient along the interface. Figure 3 demonstrates the curves of the reflected angles (Figs.

3(a) and 3 (c)) with different incident angles, while Figs. 3(b) and 3 (d) give the refracted angles changing with

different incident angles. The blue solid line is the case determined by the classical reflection and refraction

laws. Figs. 3(a) and 3 (b) are the cases that the light propagates from the optical sparse material into the optical

denser material, and Figs. 3 (c) and 3 (d) are the cases that the light transmits from the optical denser material

to the optical sparse material. In our calculations, the wavelength in vacuum was chosen as λ0  600nm ，the

green, cyan, magenta and red dotted lines show the calculations with the phase shift gradient

d / dx  π / 1.5 107 m-1 ， d / dx  π / 3 107 m-

d / dx  π / 6 107 m- and d / dx  π / 9 107 m-1 ,
1
, 1

respectively. The refraction indices

ni  1 ， nt  were chosen in Figs. 3(a) and 3(b), and the value of the
1.5

refraction indices ni  1.5 ， nt  were chosen in Figs. 3(c) and 3(d). From the Figs. 3(a) and 3(c), it can be
1

seen that the relationship between the reflected angle and the incident angle given by the general reflection law

is no longer linear. When the incident angle is larger than the critical angle θrc , the reflected light disappears.

1
The value of the critical angle for full transmission decreases gradually with the increase of the value of the

phase change gradient

d / dx along the interface, reaches to zero, and then becomes negative. It means that

1
there is a range of angles larger than 90º within the incidence plane in which there is no reflected light. This

feature will be useful to design transmission components or ultra-thin film of antireflective coatings. From the

Figs. 3(b) and 3(d), it is found that, when the light wave hits on the metasuface from a optical denser material

to a optical sparse material, two different critical angle for total internal reflection exist at the condition
d π d π
 107 m-1 , and there is one critical angle for total internal reflection at the condition  107 m-1 .
dx 6 dx 6

The results are consistent with the former analysis.

From Fig. 3, it can also be seen that the curves corresponding to the general reflection and refraction laws

are all partially in the second quadrant. That is to say, the reflected light and the refracted light may locate at

the same side of the normal of the interface with the incident light. Those are negative reflection and negative

refraction, respectively. One can find some interesting phenomena at some particular incident angle. For

example, ① the incident light may be opposite reflected at a certain nonzero incident angle for a appropriate

interface which just like the case of normal incidence in classical reflection law; ② the refracted light may be

in the same direction with the incident light at a nonzero incident angle which just likes no refraction. This kind

of abnormal reflection and abnormal refraction can occur simultaneously if the wavelength of incident light, the

refractive index of the two materials, and the phase shift gradient along the metasurface satisfy certain

condition.

Fig. 4 The contour of the reflected angle with the change of changes the refractive index of the incident material and
the incident angle at the condition of λ  600nm and d / dx  π / 6 107 .
0

Figure 4 gives the contour of the reflected angle for the different refractive indices of the incident material

and the incident angle under the condition of

λ0  600 nm and d / dx  π / 6 107 m-1 . It can be seen that

the incident angle and the reflected angle are in nonlinear relation, and for a same incident angle the reflected

angle decreases with the increase the refractive index of the incident material. In the contour, the curve in dark

blue or in dark red represents case of the full transmission, in which the reflected light disappears. The critical

angle of full transmission increases with increase of the refractive index of the incident material. This relation

1
is caused by the assumption of d / dx  0 . It is consistent with the results of the Fig. 3.

(a) d / dx  0 (b) d  / dx  π / 4.5107

(c) d / dx  π / 3107 （d） d  / dx  π / 1.5107 .

Fig. 5 The contour of the refracted angle changes with the refractive index contrast of two media and the incident angle at the
condition
λ0  600nm ， ni  1 ，

Figure 5 gives the contour of the refracted angle changes with the incident angle and the refractive index

contrast of two media for the case with the phase shift gradient along the metasurface d / dx  0 (as shown in

Fig. 5(a)),
d / dx  π / 4.5 107 m- (as shown in Fig. 5(b)), d / dx  π / 3 107 m-1 (as shown in Fig. 5(c)),
1

d  / dx  π / 1.5 107 m-1 (as shown in Fig. 5(d)) with the λ0  600 nm ， ni  1 , nt [3, 3] . When the
condition

phase shift gradient along the metasurface is zero, the contours distribute symmetrically in the four quadrants.

When the sign of the refraction index of the incident material is opposite to that of the refraction material,

negative refraction will occur and the refraction angle will be spatially isotropic with regard to the incident

angle, i.e, the optical path will be reversible in plane of incidence. However, when the phase gradient in the

metasurface is not zero (as shown in Fig.5 (b) 、 Fig.5 (c) and Fig.5 (d)), the symmetry of the contours is

broken.

The optical path is no longer reversible in the plane of incidence. Also, negative refraction will not occur

always for the case in which a light wave passes through two materials with opposite sign of the refractive

indices (as shown in the third quadrant in Fig.5 (c)). While, negative refraction may occur for the cases in

which the refractive indices of the two materials have same sign (as shown in the second quadrant in Fig.5 (c)).

1
When the value of the phase gradient is d / dx  π / 3 107 m-1 , the refracted angles are also positive for the

case of nint  0 , and the refracted angles are also negative value for the cases of nint  0 . The condition of

1
these cases can be denoted as sgn(ni nt )  sgn(θt regardless of the sign of the incident angle. The dividing
)

line of positive refraction and negative refraction is the normal of the interface of the two materials regardless

of their optical properties. It should be noted that, there is a critical value Rc of the ratio n / for a
i
nt

particular phase gradient with the condition

d / dx  π / 3107 m-1 , when n / n  R , the refracted wave
i t c

disappears. The value of R just corresponds to the critical angle of total internal reflection and is determined
c

by the formula (2).

In the above discussions, it is assumed that the abrupt phase shift is a continuous function of the position

along the metasurface, and the phase shift gradient d / dx is a constant. Therefore, the transfer of the energy

of the incident wave cross the metasurface is determined by the general reflection and refraction laws

regardless of the polarization state of incident light. This phenomenon of abnormal refraction caused by the

discontinuity of the phase shift along the interface is quite different from the relation established in

metamaterials with the negative permittivity, negative permeability and the anisotropy of the permittivity

tensor[15].

5. Design of the metasurface

The general reflection and refraction laws can be used to design elements to control the wave front by

introducing the abrupt phase shift in the optical path. V-shaped antenna array [15,16] had been designed to

achieve the abrupt phase shift along the interface based on the surface plasma resonance, and verified

experimentally by reorganizing the plasma antenna unit. However, the V-shaped antenna can only generate the

abrupt phase shift for the cross-polarized scattered light, but still, the original polarization scattering light can

not be controlled.

If a metasurface generating the abrupt phase shift for arbitrary polarization direction is designed, the

reflected light and refracted light will be controlled to achieve arbitrary manipulation. In fact, a phase mask can

be used as the metasurface. Assuming that the interface is a one-dimensional phase grating which can change

abruptly the phase of the light wave. The thickness of the grating is about tens of nanometers, its transmission

function is t( x)  t eiφ( x ) , and the phase shift φ(x)  2π x / T along the direction of the grating vector, T
0
is

is the period of the grating. So, we have

dφ / dx  2π / T . The thickness of such a structure is much smaller

than the wavelength of light, thus, it can be treated as a kind of metasurface. Pasting the metasurface on a glass

1
surface, we have ni  1 nt  1.5 . When a light with the wavelength of passes through the metasurface at
, λ0

1
 2λ0
a incident angle of θ , the refracted angle can be written as sin(θ )  sin(θ )  . Usually, the optical phase
i t i
3T
can be changed by changing the geometry path of light and/or the refractive index of the material.

The numerical simulations were done using the Comsol software to study the propagation behavior of light

crossing the ultra-thin interface. In our simulations, the refractive index gradually increased along the x-axis

direction (See Fig. 1) with dn / dx  1 / 60 nm1 , while the thickness was kept unchanged. The computational

domain was 1.8 μm  2 μm , the thickness of the metasurface was 30 nm ,

ni  1 nt  were chosen.
and
, 1.5

In order to reflect the variation of the refractive index along the metasurface, the cross section of the whole

interface between the air and the glass was divided into sub-areas with the size of 30 nm  30 nm ,

the

difference of the refractive indices between the two adjacent sub-areas was n  0.5 , the frequency of the light

was chosen as f  10 1014 Hz ( λ  0.3μm ), and the refractive index at right side was larger than that at left
0
side. The metasurface can change the phase of the light wave with
dφ / dx  π / λ0 . Figure 6 shows the

simulating results at different incident angles. It can be seen that different refracted angle corresponds to

different incident angle, and negative refraction may occur (See Fig. 6(c)). These results satisfy the relation
1
1.5sin(θ )  sin(θ )  and it verified the previous results. It should be pointed out that the method for
t i
2
impedance matching was employed to minimize the affection of reflection so that to well demonstrate the

simulating results. Similar results were obtained for other wavelengths in our simulations.

(a) θ  0 (b) θi  π / 6
i

(c) θi  π / 12 (d) θ i  π / 6

1
 Fig. 6 The simulation results when the light waves with 14
f  10 10 at different angles
frequency
Hz

In the above simulations, it was assumed that the refractive index changes linearly along the metasurface.

1
Several techniques can be used to achieve refractive index change, such as electro-optical effect, accusto-

optical effect, in designing a variety of ultra-thin optical devices.

Acknowledgements

The authors thank the National Natural Science Foundation of China (grant Nos. 11074024 and 11104016) and

the Specialized Research Fund for the Doctoral Program of Higher Education (SRFDP) Education ministry of

China for the financial support. We also would like to thank Dr. Jun Zheng for her help with the simulations.

References

1. S. Zouhdi, A. Sihvola, A. P. Vinogradov (2008-12). Metamaterials and plasmonics: fundamentals,

modelling, applications. (Springer, New York, 2008).

2. W. Cai, V. Shalaev, Optical metamaterials: fundamentals and applications (Springer, New York, 2009).

3. N. Engheta, R. W. Ziolkowski, Metamaterials: physics and engineering explorations (Wiley-IEEE Press,

New York, 2006).

4. D. R. Smith, J. B. Pendry and M. C. K. Wiltshire. “Metamaterials and negative refractive index”. Science

305(5685): 788-792(2004).

5. J. Valentine1, S. Zhang, T. Zentgraf, E. Ulin-Avila1, D. A. Genov1, G. Bartal1 and X. Zhang, “Three-

dimensional optical metamaterial with a negative refractive index”. Nature 455, 376-379 (2008).

6. C. García-Meca, J. Hurtado, J. Martí, A. Martínez, W. Dickson and A. V. Zayats. “Low-loss multilayered

metamaterial exhibiting a negative index of refraction at visible wavelengths”. Phys. Rev. Lett. 106,

067402 (2011).

7. C. Enkrich, M. Wegener, S. Linden, S. Burger, L. Zshiedrich, F. Schmidt, J. F. Zhou, Th. Koschny, and C.

M. Soukoulis, “Magnetic metamaterials at telecommunication and visible frequencies”, Phys. Rev. Lett.,

95, 203901 (2005).

8. A. Boltasseva and H. A. Atwater. “Low-loss plasmonic metamaterials”. Science 331(6015): 290(2011).

9. Y. Zhao and A. Alù. “Manipulating light polarization with ultrathin plasmonic metasurfaces”. Phys. Rev.

B 84(20): 205428(2011).

10. J. W. Lee, M. A. Seo, J. Y. Sohn, Y. H. Ahn, D. S. Kim, S. C. Jeoung, Ch. Lienau, and Q-Han Park.

“Invisible plasmonic meta-materials through impedance matching to vacuum”. Optics

Express .13(26):10681-10687 (2005).

11. E. Saenz, I. Ederra, P. Ikonen, S. Tretyakov and R. Gonzalo, “Power transmission enhancement by means

1
 of planar meta-surfaces”. J. Opt. A: Pure Appl. Opt. 9: S308–S314 (2007).

12. E. Saenz, I. Ederra, R. Gonzalo, S. Pivnenko, O. Breinbjerg and P. de Maagt, “Coupling reduction

between dipole antenna elements by using a planar meta-surface”. IEEE Transactions on Antennas and

Propagation. 57(2):383-394 (2009).

13. E. Saenz, I. Ederra, P. de Maagt and R. Gonzalo, “Highly efficient dipole antenna with planar meta-

surface”, Electron. Lett. 43(16): 850-851(2007).

14. B. K. Alici, A. B. Turhan, C. M. Soukoulis, E. Ozbay, “Optically thin composite resonant absorber at the

near-infrared band: a polarization independent and spectrally broadband configuration”. Optics Express 19

(15):14260-14267(2011).

15. N. Yu, P. Genevet, M. A. Kats, F. Aieta1, J. P. Tetienne, F. Capasso, and Z. Gaburro, “Light propagation

with phase discontinuities: generalized laws of reflection and refraction”, Science 334 (333):333-

337(2011).

16. X. Ni, N. K. Emani, A. V. Kildishev, A. Boltasseva, and V. M. Shalaev. “Broadband light bending with

plasmonic nanoantennas”. Science Express. 335(6067):427-430(2011).

17. J. B. Pendry, D. Schurig, D. R. Smith, “Controlling electromagnetic fields”, Science

312(5781) :1780–1782 (2006).

18. U. Leonhardt, “Optical conformal mapping”, Science 312(5781):1777–1780(2006).

19. I. I. Smolyaninov, E. E. Narimanov, “Metric signature transitions in optical metamaterials”. Phys. Rev.

Lett. 105, 067402 (2010).

20. S. D. Brorson, H. A. Haus, “Diffraction gratings and geometric optics”. J. Opt. Soc. Am. B 5, 247 (1988).

21. R. P. Feynman, A. R. Hibbs, Quantum mechanics and path integrals (McGraw-Hill, New York, 1965).

22. Max Born, and Emil Wolf, Principles of Optics, 7th ed. Cambridge University Press, 1999, Cambridge.

12
View publication stats