Cylinder light concentrator and absorber: theoretical description

We present a detailed theoretical description of a broadband omnidirectional light concentrator and absorber with cylinder geometry. The proposed optical “trap” captures nearly all the incident light within its geometric cross-section, leading to a broad range of possible applications – from solar energy harvesting to thermal light emitters and optoelectronic components. We have demonstrated that an approximate lamellar blackhole with a moderate number of homogeneous layers, while giving the desired ray-optical performance, can provide absorption efficiencies comparable to those of ideal devices with a smooth gradient in index. ©2010 Optical Society of America OCIS codes: (260.0260) Physical Optics; (070.7345) Wave propagation References and links 1. V. M. Shalaev, W. S. Cai, U. K. Chettiar, H. K. Yuan, A. K. Sarychev, V. P. Drachev, and A. V. Kildishev, “Negative index of refraction in optical metamaterials,” Opt. Lett. 30(24), 3356–3358 (2005). 2. J. B. Pendry, “Negative refraction makes a perfect lens,” Phys. Rev. Lett. 85(18), 3966–3969 (2000). 3. Z. Jacob, L. V. Alekseyev, and E. Narimanov, “Optical Hyperlens: Far-field imaging beyond the diffraction limit,” Opt. Express 14(18), 8247–8256 (2006). 4. A. Salandrino and N. Engheta, “Far-Field Subdiffraction Optical Microscopy Using Metamaterial Crystals: Theory and Simulations,” Phys. Rev. B 74(7), 075103 (2006). 5. N. M. Litchinitser, and V. M. Shalaev, “Metamaterials: transforming theory into reality,” J. Opt. Soc. Am. B 26(12), B161–B169 (2009). 6. N. I. Landy, S. Sajuyigbe, J. J. Mock, D. R. Smith, and W. J. Padilla, “Perfect metamaterial absorber,” Phys. Rev. Lett. 100(20), 207402 (2008). 7. T. V. Teperik, F. J. García de Abajo, A. G. Borisov, M. Abdelsalam, P. N. Bartlett, Y. Sugawara, and J. J. Baumberg, “Omnidirectional absorption in nanostructured metal surfaces,” Nat. Photonics 2(5), 299–301 (2008). 8. E. E. Narimanov, and A. V. Kildishev, “Optical black hole: broadband omnidirectional light absorber,” Appl. Phys. Lett. 95(4), 041106 (2009). 9. L. D. Landau, and E. M. Lifshitz, Mechanics, Pergamon Press, Oxford, (1976). 10. R. K. Luneburg, Mathematical Theory of Optics, University of California Press, Berkeley, 1964, p. 12. 11. E. J. Post, Formal Structure of Electromagnetics: General Covariance and Electromagnetics, 1962, p. 152. 12. S. Gradshteyn, and I. M. Ryzhik, Tables of integrals, series and products, Academic Press, New York, CD-ROM Edition, 1994, Eq. (8).511.4. 13. M. Born, and E. Wolf, Principles of Optics, Electromagnetic Theory of Propagation, Interference and Diffraction of Light, 7th ed. (Cambridge University Press, New York, 1999). 14. G. P. Agrawal, and D. N. Pattanayak, “Gaussian beam propagation beyond the paraxial approximation,” J. Opt. Soc. Am. 69(4), 575 (1979). 15. M. Abramowitz, and I. A. Stegun, eds., Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, New York, Dover, (1972). 16. V. Kildishev, U. K. Chettiar, Z. Jacob, V. M. Shalaev, and E. E. Narimanov, “Materializing a binary hyperlens design,” Appl. Phys. Lett. 94(7), 071102 (2009). 17. S. D. Conte, and C. W. Boor, “Elementary Numerical Analysis: An Algorithmic Approach, 3rd edition,” 1980. 18. H. Hulst, Light scattering by small particles, Dover, 1981, p. 309.


Introduction
Recent advances in the development of optical metamaterials, artificial structures with the desired local electromagnetic response determined by artificial subwavelength patterning of the media, have opened a wide range of new concepts and applications, from super-resolution imaging to optoelectronics and photovoltaics [1][2][3][4][5][6][7][8]. In particular, the recently proposed [8] "optical black hole", an omnidirectional light concentrator and absorber, could significantly enhance the performance of solar energy harvesting systems and lead to novel nonlinearoptical devices. In this work, we present a detailed theoretical description of this device, and develop appropriate numerical methods, which -in addition to the accurate analysis of the optical black hole, -could be applied to a broad range of metamaterial-based optical systems with centrally symmetric gradient index. We note that in the context of this paper, the term "black hole" simply refers to the highly efficient "capture" of the light incident on the device, and does not imply any profound analogy to General Relativity.
The remaining part of the paper is organized as follows. In Section 2, we introduce the initial assumptions, state the geometry of the cylindrical optical system with an axially symmetric graded index, and show basic examples of ray-tracing inside such devices, including the fundamental optical black-hole [8]. Section 3 deals with the ray optical description of the system followed by its full-wave description in Section 4. The analytical and numerical apparatus for approximating the ideal fundamental black-hole with concentric lamellar systems of uniform isotropic layers is then discussed in Section 5. Section 5 also covers derivations of the scattering and absorption cross-sections, which are then compared with their semiclassical approximations in Section 6. Finally, we summarize our work in Section 7. show light ray trajectories for p equal to −1, 1, 2, and 3 respectively, where core radius rc much smaller than shell radius rs (rc = rs).

The system
In its original proposal [8], the optical "black hole" (see Fig. 1(a)) consists of the "payload" core with the radius c r , supporting the desired functionality (such as e.g. a solar cell for concentrated photovoltaics applications), surrounded by the concentrating shell of radius s r with an axially symmetric dielectric function, ( ) r ε , changing in the radial direction as and 0 p > is a constant. In order to match the real parts of relative permittivity at the interface c r r = the radius of the core is defined by the ratio of the permittivity of the surrounding medium ( s ε ) to that of the core ( c ε ), Note that we assume that the metamaterial forming the shell, is essentially non-magnetic, so that the refractive index is n ε = .

Ray-optical description
In the semi-classical limit, the radial variation of the refractive index of the shell (where n ε = ) corresponds to the effective potential which for the dielectric permittivity defined by (2) with 2 p ≥ traps all the incident light trajectories -see Figs. 1(b)-1(e). For these ray trajectories, a straightforward integration of the Hamiltonian equations with the effective potential defined by (4) yields in the polar coordinates ( ) where m is the (conserved) angular momentum, ( ) where p φ is yet another constant defined by ( ) Case 1 p = corresponds to the gravitational potential (or electrostatic potential) and emulates the motion of a non-accelerated particle in central gravitational field; so that the parabolic trajectories (Fig. 1c) are expected. Case 2 p = gives the potential of fatal attraction [9]. In fact for all 2 p ≥ (6) corresponds to the ray trajectories "falling" into the core of the "black hole" -see Fig. 1 show the traces of light rays for p equal to −1, 1, 2, and 3 respectively; the trivial case of the beam propagating in free-space ( 0 p = ) is omitted.
This concept can be realized in both the spherical and cylindrical geometries, with the latter suitable for optoelectronic devices where the light propagation is generally realized in a waveguide geometry. In the present paper we will focus on this cylindrical realization of the optical black hole, leaving its spherical counterpart to a forthcoming work.

The wave optical description
In the present section of the paper, we develop the wave-optical description of the "black hole" light concentrator. We define a given monochromatic electromagnetic vector field by the vectors of electric field is a smooth spatiallydependent dielectric function, the free-space parameters 0 k , c , 0 ε , and 0 µ respectively denote the wavenumber, the velocity of light, the permittivity, and the permeability.
Here we assume only transverse electric (TE, ˆ( , ) e x y = E z ) and transverse magnetic (TM, ˆ( , ) h x y = H z ) waves within a nonmagnetic medium ( 1 µ = ), so we reduce the above vector equations and use in-plane scalars e and h , which are further on collectively denoted as f .

Scalar wave equations and separation of variables
Hence, both TE and TM wave Eqs. (9) and (10) reduce to a common scalar wave equation 2 2 0.
where the prime denotes the radial derivative, and ( ) 0 q r = or ( ) 2 ln q r r ε ′ = − for TE or TM modes respectively.

TE and TM solutions for ( ) const
r ε = The solvable cases that we are using here include the trivial instance of ( ) const r ε = for both TE and TM modes. In this case, (12) reduces to the Bessel equation for r kr = , and the general solution is a linear combination of two linearly independent Bessel functions [11].
In most of the cases it is more convenient to use the following combination since it represents the unique decomposition (usually found in scattering problems) into the incident and scattered field if the domain is infinite. On the other hand if the solution is written in the domain, which contains the origin, then the additional boundary condition should be provided to avoid singularity ( 0 m b = ). For simulations we use the normalized variant of (13), which improves the condition number of the linear system matrix (Section 5.1): However in some applications, including the calculation of cross sections, a better option is an equivalent combination of the Hankel functions which gives expressions that are easier to handle for mathematical analyses (Section 5.3). Back and forth translations between formulas (13) and (15)  q r = . For a TM mode, (12) can be therefore reduced as and for a TE mode as Both (16) and (17)

Boundary conditions and the mode matching
In previous sections the general analytical solutions were written in the domains, where the dielectric function ( ) r ε is defined by either the Bessel equation or the Euler equation. At each interface between such domains, the electromagnetic field should satisfy the standard boundary conditions for the continuity of the tangential components of the magnetic and electric fields. Furthermore, the physical requirement for the behavior at the infinity and at the origin, introduce additional constrains to the solution, as described earlier in Section 4.2 In particular for a nonmagnetic, cylindrically symmetric medium without any free charges the condition of continuity of electromagnetic fields at a given interface 0 TM:

Incident light: the Gaussian beams and plane wave
While the angular momentum formalism introduced above, represents the "natural" choice for a system with rotational symmetry, to complete the description of the scattering problem, we also need to develop the angular momentum expansion of the incident field. Here, we will consider the most common examples of the illumination field: the plane wave and the Gaussian beam. Starting from the classical generating function for the Bessel functions [12]: While plane wave illumination represents perhaps the most common setup of a scattering problem, it does not convey the case when the target is illuminated by a focused beam. The latter is generally adequately described by the Gaussian beam [13], which yields an accurate description of the beam focused by linear optical elements. However, one needs to keep in mind that the Gaussian beam in its standard form is not an exact solution of Maxwell's equations in the free space, but instead corresponds to the paraxial approximation [13]. To avoid the approximate nature of this approach and resulting inaccuracy, we consider that the field, while following the Gaussian profile at the waist cross section, is in fact the exact solution of the Helmholtz wave equation with the given direction of propagation [14]: The resulting solution of (26) in each of half-spaces 0 y ≤ , 0 y ≥ can be written as a composition of y-plane waves ( ) ( ) where the amplitudes ( )

Layered Systems
Practical realizations of concentric cylinder optical and optoelectronic devices often involve a design that includes a number of (homogeneous) layers. In particular, external layers may be incorporated as a protective cover, while an internal layer between the absorber and the shell of the optical concentrator provides additional mechanical support or works as a channel for a liquid absorber. In this Section, we develop the theoretical description of light scattering and absorption by such layered systems.

Cascading the cylindrical layers
The boundary value problem of the scattered cylindrical modes inside a set of concentric cylindrical layers leads to a system of linear equations. Utilizing a direct linear algebra method for obtaining the scattered cylindrical modes inside layers with losses or axially symmetric distribution of the dielectric function could be problematic due to typically large condition numbers of the system matrices. The algorithm that we use includes a normalization step, making it possible to improve poor conditioning. However, if the algorithm were used for getting analytical formulae the normalization would not be necessary. We consider a concentric cylindrical device having 1 l − layers,  T r with yet undefined coefficients. These coefficients can be obtained from appropriate 2l boundary conditions. All cylindrical modes are linearly independent, thus each of the boundary conditions can be separated into the individual equations for each m th mode, similar to that shown in (14) presented in a more general form, where for the simplicity of notation (and from now on in this section) we suppress the index m . To satisfy the condition (22) in the core layer we put 1 0 Furthermore, the total field in the "outer" layer can be separated into the "incident" and "scattered" parts. The Sommerfeld radiation condition ( ) For the TM polarization the only change is that Inverting the matrix at the left-hand-side of (33) and evaluating its determinant as ( ) and [ , ] W T R TR T R ′ ′ = − is the Wronskian [12], which is taken for the partial solutions of the 1 i + layer. In particular, for homogeneous layers we exploit the known value of the Wronskians for the Bessel functions [15], This approach is also known as the 'cascading layers' method [16], since the solution in each layer (34) is then a composition of transfer matrices applied to the first layer ( ) The resulting system obtained after stating all 2l boundary conditions has the form: The system (37) can be solved in two sweeps, similar to the standard Thomas algorithm [17]. In this approach, the auxiliary normalized coefficients followed by the backward sweep of the reflection and transmission coefficients (39), while using the initial condition at the external interface: where l τ is known and l l l ρ ρ τ = . The above algorithm includes the normalization of the partial solution in each layer by its value on the inner interface (the field in the core layer is an exception and is normalized by its value on the outer interface 2 r r = , the latter is achieved by substitution 1 2 r r = in Eq. (32)).
The normalization significantly improves the accuracy of the mathematical operations, performed numerically with a limitation of floating-point representations and arithmetic. Finally, the procedure yields the m th mode of the field by using (32); the solution to the entire problem in the i th layer is given as a sum of the angular momentum modes,

Scattering and absorption efficiencies
For any optical system, scattering and absorption efficiencies have the physical meaning of the relative size of the equivalent "black body" target which results in the same amount of scattered and absorbed power [18]. First, we recapitulate that the total field ˆf = E z (TE) or ˆf = H z (TM) outside the device at the external boundary of the shell,

The ideal electromagnetic black hole
We consider the ideal "fundamental" (p = 2) electromagnetic black hole, see (1), (2), (3). It is a three layered system with a gradient-index shell and absorbing core, for which the formalism of Section 5.1 can be applied to calculate the exact full-wave solution.  Furthermore the approach is also used to determine the final form of the absorption efficiency. For this we use Eqs. (47)

The lamellar electromagnetic black hole
The formalism of a layered system for example can be used to study a non-ideal lamellar "black hole" optical concentrator and absorber, which approximates an ideal device ( 2.  50), the latter, abs Q , yields 72%, 84%, 90%, and 94% respectively. Fig. 4(e) depicts the field map of the ideal black hole with smooth gradient, where 99% absorption efficiency is achieved. Fig. 4. Simulated field maps for a lamellar "black hole" optical concentrator and absorber (εs = 2.1, rs = 20 µm, and rc = 8.367 µm) vs. the total number of layers, l. The device is illuminated by a plane wave with free-space wavelength λ = 1.5µm. (a) l = 3, and 72% absorption efficiency, (b) l = 5, and 84% absorption efficiency, (c) l = 9, and 90% absorption efficiency, (d) l = 17, and 94% absorption efficiency. Panel (e) depicts the reference case of the ideal black hole with smooth gradient, and 99% absorption efficiency.

Semiclassical description
The primary objective of the present section is the derivation of the semiclassical expansion ( 1 s r ≫ ) of the scattering and absorption cross sections of the "fundamental" (p = 2) electromagnetic black hole, introduced in Section 2.
Intuitively, from the ray dynamics of Section 2, we expect the absorption cross section to be close to the "geometrical" limit, so that 1 abs Q → . The ray-optical treatment of Section 5 however gives no quantitative measure of how closely this limit is actually approached in a practical device. On the other hand, the exact description of Section 5, while giving a precise quantitative account, provides no compact lucid description of the system behavior. In contrast to this behavior, the semiclassical approach, mathematically corresponding to taking the formal limit 1 s r ≫ of the exact solution, combines the clear physical picture of the actual dynamics in terms of the underlying ray trajectories, with the quantitative description of the interference and diffraction in the system.

Summary
In summary, we have developed a theoretical description of wave propagation in cylindrically-symmetric gradient-index systems, and applied this approach to light "trapping" in the recently proposed electromagnetic black hole concentrator and absorber.
In particular we have demonstrated that an approximate lamellar black-hole with a relatively limited number of homogeneous layers, while giving the desired ray-optical performance, can provide absorption efficiencies comparable with those of ideal devices with smooth gradients in index. It is forecasted that the use of non-uniform layers could further improve the performance of the lamellar device with a given number of homogeneous layers. The numerical validation of the mathematical apparatus for designing the black-hole devices opens the path for the development of test beds that could further advance the application of this light-trapping technique into a viable commercial option.