검색 전체 메뉴
PDF
맨 위로
OA 학술지
Simulation and Optimization of Nonperiodic Plasmonic Nano-Particles
  • 비영리 CC BY-NC
  • 비영리 CC BY-NC
ABSTRACT
Simulation and Optimization of Nonperiodic Plasmonic Nano-Particles
KEYWORD
Plasmonic nano particles , Optimization algorithm , Binary PSO
  • I. INTRODUCTION

    Metal nanoparticles have broad applications in electronics, photonics, chemical sensing, and imaging [1-3]. Recently there has been a growing interest in plasmonic nanoparticles. They have strong effects on light in the visible and invisible regions of the photonic spectrum for applications such as Raman scattering [4], radiative rate enhancement [5], solar cells [6], and optical biosensors [7]. Since the optical properties of these particles depend on their sizes and shapes, one of the desired goals is to control the shapes of metal and semiconductor nanoparticles. Shape control has been successfully demonstrated for gold nanoparticles using nonionic surfactants, silver underpotential deposition, and nanoreplica molding [8-11]. Nanoreplica molding has been demonstrated as a low-cost method for manufacturing a variety of devices comprised of nano structured surfaces. Recently a plasmonic nanodome array was fabricated by the nanoreplica molding process [10]. In addition to the recent interest in shape control of nanoparticles, the optical properties of noble metal particles, with their intense colors, have fascinated scientists since the turn of this century. For example, spherical gold nanoparticles show a strong absorption band in the visible region of the electromagnetic spectrum at about 520 nm [12]. In addition to the shape, size, and material of nanoparticles, the properties of transmitted light strongly depend on the localized positions and gaps between nanoparticles. Plasmonic nanoparticles with periodic structure have been reported [13, 14]. One of the most promising plasmonics nanoparticle platforms is studying the effect of deterministic aperiodic structure of nanoparticles on the properties of transmitted light [15]. Such structure, which is intermediate between a disordered system and a periodic one, enables unique control and manipulation of spatially localized plasmonic states over broadband frequency and angular spectra [16].

    In structures with periodic plasmonic nanoparticles the characteristics of transmitted light can be calculated efficiently using the coupled-dipole approximation (CDA), originally developed by Purcell and PennyPacker [17] and improved by Draine [18]. In this paper, an alternative design method called the binary coupled dipole approximation (BCDA) is described. Utilizing this method, the effect of a nonperiodic structure of nanoparticles on light properties could be studied. A big advantage of this method is its application to optimization problems by utilizing binary evolutionary computational techniques, such as binary genetic algorithms and binary particle swarm optimization. Moreover, higher absorption, scattering, and extinction coefficients are obtained using BPSO and BCDA methods. These improvements can be helpful in optical applications such as solar cells, optical bio sensors, and plasmonic nanoantennae.

    II. METHODS

       2.1. Binary Couple Dipole Approximation

    A schematic diagram of a three-dimensional array of arbitrarily shaped objects periodically arranged in the x y z plane is shown in Fig. 1(a). In the first time step the x, y, and z positions of each particle are calculated. To assign the position of each particle, it is assumed that there is a cube filled with nanoparticles equally distant from each other. The Matlab code for determining the positions of N particles is presented in Fig. 1(b). In this code, r is a particle’s position vector, N1, N2, and N3 represent the odd number of particles in the i, j, and k directions, and d represents the distance between the particles. B(i, j, k) is a three-dimensional array of binary numbers. The values 1 and 0 in the B matrix refer respectively to the presence and absence of particles in the desired locations. Next, the incident electric field is simulated. The scattered light may be illuminated by any form of incident electric field, the simplest of which is a plane wave. To calculate the E-field of each dipole at rj, the time harmonic component -iωt of the E-field is left out. The Local field arising from incident light with polar (θ) and azimuthal (φ) angles at each particle is:

    image

    Where:

    image

    For an incident field with p-polarization, the following can be written:

    image

    and for an incident field with s-polarization:

    image

    When the applied field is parallel to one of the principle axes, the polarizability α is [19]:

    image

    where V is particle volume, εm is the dielectric function of the surrounding medium, L is the shape factor, and . One advantage of the BCDA method is studying the effect of an ellipsoidal particle on light. For oblate and prolate spheroids (b=c), the following analytical expression can be given for L1 as a function of eccentricity e [19]:

    image

    where a, b, and c are seminal excess of an ellipsoid (a>b> c) and, for a sphere, L1=1/3.

    The dipole moment induced in a single particle by a local electric field is given by:

    image

    Here, is the induced dipole moment, αi is the polarizability of the particle centered at , Eloc is the local electric field, and ε0 is the permittivity of free space. The local field arises from two sources: the first term is incident light, , and the second term is superposition of the retarded fields from each of the other N-1 radiating dipoles in the array. Combining these terms leads to the local field as follows:

    image
    image

    Here, , k is wave vector, and Aij are 3x3 matrices representing the interaction of two particles i and j. Now, 3N unknown dipole moments Pj should be solved for in the following exactly determined system of 3N linear equations:

    image

    where A is a 3N×3N matrix containing N_N of Aij 3x3 tensors, where N is the number of dipoles and both p and Einc are 3N vectors (i.e. each of N particles is represented by a 3-vector). These matrices are shown in Fig. 3. When this set of 3N complex linear equations is solved ( A. P = Einc), the P array of self-consistent dipole moments is obtained. The optical properties may be then calculated from this dipole array.

       2.2. Far-Field Properties of Light

    As explained in the previous section, optical absorption, scattering, and extinction can be directly calculated from the dipole array. The extinction and scattering coefficients are [12]:

    image
    image

    and the absorption coefficient is Qabs=Qext-Qsca.

       2.3. Binary Particle Swarm Optimization Algorithm

    The PSO algorithm is an optimization procedure inspired by a colony such as birds which can improve its behaviors [20]. Any element of this colony is called a particle and moves in an n-dimensional space, correcting its trajectory based on the previous actions of itself and its neighboring particles. For each particle, velocity and displacement are updated based on the following relations [21]:

    image
    image

    where k is the number of the current iteration, n is the number of particles, w is the inertia weight, c1 and c2 are acceleration parameters, and finally r1 and r2 are random parameters between 0 and 1. The best position for the i-th particle which has been stored so far is represented as [22]:

    image

    All the (p_best) are evaluated by a fitness function. The best particle among all p_best is represented as (Gbest_best_value), which in a minimization problem is the smallest member of the vector while for a maximization problem it is the largest member of . PSO was designed for continuous problems, but cannot deal with discrete problems. A new version of PSO, called Binary PSO (BPSO), was introduced by Kennedy and Eberhart in 1997 and applied to discrete binary variables. After that, many optimization problems in various areas were solved by this method. The position in BPSO is represented by a binary vector and the velocity is still a floating-point vector; however, velocity is used to determine whether the probability changes from 0 to 1 or from 1 to 0 when the positions of particles are being updated. The equation for updating the positions is then replaced with:

    image
    image

    In order to achieve maximum absorption, extinction, and scattering coefficient spectra, BPSO has been used to design a nonperiodic array of plasmonic nanospheres. The presence or absence of nanospheres could be controlled by a matrix with binary entries. As seen, BPSO is an algorithm that minimizes a cost function. To use the BPSO algorithm for maximizing the extinction coefficient spectrum, the algorithm should minimize the following function:

    image

    and for maximizing the absorption and scattering coefficient spectra:

    image
    image

    where i is the wavelength number and Qext, Qabs, and Qsca are the extinction, absorption, and scattering coefficients respectively.

    III. NUMERICAL RESULTS

    The BCDA method is applied and the absorption spectrum is plotted for periodic and nonperiodic arrays of gold nanospheres with 1 nm diameter and 2 nm edgeto- edge separations in 1, 2, and 3 dimensions. Figure 4(b) shows the absorption spectrum as a function of wavelength in the range of 300-600 nm for a 1D nonperiodic array. The present simulations show that the absorption spectrum strongly depends on localized positions of nanospheres. On the other hand, further enhancement in absorption coefficient is possible for nonperiodic structure compare to the periodic array (Fig. 4(a)). These results are quite consistent with the results obtained using Mie theory [12]. Figure 5(a) shows this possibility for a 2D nonperiodic array. One benefit of increased material absorption coefficients is in solar cell. Then the effect of nonperiodic arrays of gold nanospheres on the absorption spectrum with a 3D nonperiodic array is investigated. Figure 5(b) shows the absorption spectrum as a function of wavelength for nonperiodic array in 3D coordinates. Figure 6(b) shows the extinction spectrum of a 2D nonperiodic array with 3x5 particles compared to that of periodic array (Figure 6(a)). The large extinction coefficient observed here was due to the interaction of metal nanoparticles via their strong near-field coupling. In this case the plasmon resonance of a nanoparticle acts as a light resonator and couples light to other resonators in the vicinity [23]. The BCDA method is applied and extinction coefficient spectra are plotted for 1, 2, 3, and 4 layers of gold nanospheres (5×5 array) with 1 nm diameter and 2 nm edge-to-edge separations. Figure 7 shows these extinction spectra in the wavelength range of 300-600 nm. Our simulations show that the extinction coefficient spectrum strongly depends on the number of layers of nanospheres, and further enhancement of the extinction coefficient is possible with multi periodic layers.

    IV. OPTIMIZATION RESULTS

    Binary PSO and BCDA are used to optimize the farfield properties of transmitted light for optical applications. Our goal is to maximize Qext by optimizing 15 binary particles in 2D arrays (3×5). Figure 8(a) shows the extinction spectrum of the optimized array (3×5) compared to that of the periodic array (Fig. 6(a)). The maximum extinction in the entire range of 300-600 nm is 9.8, which occurs at 560 nm. The large extinction coefficient observed here is due to the interaction of metal nanoparticles via their strong near-field coupling. The convergence characteristic of the BPSO method is shown in Fig. 8(b). As seen, this algorithm had good performance in terms of convergence speed.

    Next, the BPSO algorithm is applied to maximize absorption and scattering coefficients by optimizing the selection of binary particles. Figures 9(a) and Figures 9(b) show the absorption and scattering spectra of the optimized array compared to those of the periodic array. One possible application of this method can be reduction of the scattering factor to have an optical invisibility cloak based on scattering cancellation. In addition to sensor, solar cell, and cloaking applications, BPSO can enhance the Raman scattering factor.

    V. CONCLUSION

    A numerical solution called the binary-coupled dipole approximation (BCDA) was described for simulating the effects of nonperiodic plasmonic nanoparticles on light properties. This method is useful for optimization problems to find the best nanoparticle array from all possible arrays. Finally, it was shown that using a nonperiodic structure could lead to higher extinction, absorption, and scattering coefficients. This approach can be useful in optical applications such as solar cells, bio sensor, and plasmonic nanoantennae.

참고문헌
  • 1. Huang Y, Duan X, Wei Q, Lieber C. M 2001 “Directed assembly of one-dimensional nanostructures into functional networks,” [Science] Vol.291 P.630-633 google
  • 2. Law M, Sirbuly D. J, Johnson J. C, Goldberger J, Saykally R. J, Yang P 2004 “Nanoribbon waveguides for subwavelength photonics integration,” [Science] Vol.305 P.1269-1273 google
  • 3. E Katz, I Willner 2004 “Integrated nanoparticle-biomolecule hybrid systems: Synthesis, properties, and applications,” [Angewandte Chemie International Edition] Vol.19 P.6042-6108 google
  • 4. Kneipp K, Wang Y, Kneipp H, Perelman L. T, Itzkan I, Dasari R. R, Feld M. S 1997 “Single molecule detection using surface-enhanced Raman scattering (SERS),” [Phys. Rev. Lett.] Vol.78 P.1667-1670 google
  • 5. J. S Biteen, D Pacifici, N. S Lewis, H. A Atwater 2005 “Enhanced radiative emission rate and quantum efficiency in coupled silicon nanocrystal-nanostructured gold emitters,” [Nano Lett.] Vol.9 P.1768-1773 google
  • 6. K. R Catchpole, A Polman 2008 “Plasmonic solar cell,” [Opt. Express] Vol.16 P.21793-21800 google
  • 7. Boriskina S. V, Gopinath A, Negro L. D 2008 “Optical gap formation and localization properties of optical modes in deterministic aperiodic photonic structures,” [Opt. Express] Vol.16 P.18813-18826 google
  • 8. Hormozi-Nezhad M. R, Karami P, Robatjazi H 2013 “A simple shape-controlled synthesis of gold nanoparticles using nonionic surfactants,” [An International Journal to Further the Chemical Sciences] Vol.3 P.7726-7732 google
  • 9. Personick M. L, Langille M. R, Zhang J, Mirkin C. A 2011 “Shape control of gold nanoparticles by silver underpotential deposition,” [Nano Lett.] Vol.8 P.3394-3398 google
  • 10. Wu H, Choi C. J, Cunningham B. T 2012 “Plasmonic nanogap-enhanced Raman scattering using a resonant nanodome array,” [Proc. Conference on Lasers and Electro-Optics] google
  • 11. Kawamura G. O, Nogami M, Matsuda A 2013 “Shapecontrolled metal nanoparticles and their assemblies with optical functionalities,” [Journal of Nanomaterials] Vol.3 P.1-17 google
  • 12. Le Ru E. C, Etchegoin P. G 2008 Principles of Surface Enhanced Raman Spectroscopy and Related Plasmonic Effects google
  • 13. Negro L. D, Feng N. N, Gopinath A 2008 “Electromagnetic coupling and plasmon localization in deterministic aperiodic arrays,” [J. Opt. A, Pure Appl. Opt.] Vol.10 P.064013 google
  • 14. Negro L, Feng N 2007 “Spectral gaps and mode localization in Fibonacci chains of metal nanoparticles,” [Opt. Express] Vol.22 P.14396-14403 google
  • 15. Luck J. M 1989 “Cantor spectra and scaling of gap widths in deterministic aperiodic systems,” [Phys. Rev. B] Vol.39 P.5834-5849 google
  • 16. C Forestiere, G Miano, S. V Boriskina, L Dal Negro 2009 “The role of nanoparticle shapes and deterministic aperiodicity for the design of nanoplasmonic arrays,” [Opt. Express] Vol.17 P.9648-9661 google
  • 17. M Purcell E, Pennypacker C. R 1973 “Scattering and adsorption of light by nonspherical dielectric grains,” [Astrophys J.] Vol.186 P.705-714 google
  • 18. B. T Draine 1988 “The discrete-dipole approximation and its application to interstellar graphite grains,” [Astrophys J.] Vol.333 P.848-872 google
  • 19. Bohren C. F, Huffman D. R 1998 Absorption and Scattering of Light by Small Particles google
  • 20. Bozorgi M., Granpayeh N. 2011 “Directional emission from photonic crystal waveguide output by terminating with CROW and employing the PSO algorithm,” [Journal of the Optical Society of Korea] Vol.15 P.187-195 google
  • 21. Emami F, Akhlaghi M 2012 “Gain ripple decrement of S-band Raman amplifier,” [IEEE Photon. Technol. Lett.] Vol.24 P.1349-1352 google
  • 22. Akhlaghi M., Emami F. 2013 “Fuzzy adaptive modified PSO-algorithm assisted to design of photonic crystal fiber Raman amplifier,” [Journal of the Optical Society of Korea] Vol.17 P.237-241 google
  • 23. Nishijima Y, Rosa L, Juodkazis S 2013 “Long-range interaction of localized surface plasmons in periodic and random patterns of Au nanoparticles,” [J. Appl. Phys. A] Vol.113 google
OAK XML 통계
이미지 / 테이블
  • [ FIG. 1. ]  (a) 3-dimensional array of nanoparticles. (b) Matlab code for determining the position of N number of nanoparticles.
    (a) 3-dimensional array of nanoparticles. (b) Matlab code for determining the position of N number of nanoparticles.
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ FIG. 2. ]  Prolate (left, a>b=c) and oblate (right, a<b=c) ellipsoids.
    Prolate (left, a>b=c) and oblate (right, a<b=c) ellipsoids.
  • [ FIG. 3. ]  A, P, Einc matrix from equation 10.
    A, P, Einc matrix from equation 10.
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ ] 
  • [ FIG. 4. ]  (a) QAbs for a 1D periodic array with B=[1 1 1 1 1] . (b) QAbs for a 1D non periodic array with B=[1 0 1 0 0].
    (a) QAbs for a 1D periodic array with B=[1 1 1 1 1] . (b) QAbs for a 1D non periodic array with B=[1 0 1 0 0].
  • [ FIG. 5. ]  (a) Absorption coefficient for a 2D nonperiodic structure with B=[0111 0;01001;10110;01100;01110], (b) Absorption coefficient for 3D nonperiodic structure with B = [10001; 00010;00100;01011;10100], B(:,:,2) = [10101;0101 0;0 0 0 0 0; 1 0 0 1 0; 0 0 1 0 0].
    (a) Absorption coefficient for a 2D nonperiodic structure with B=[0111 0;01001;10110;01100;01110], (b) Absorption coefficient for 3D nonperiodic structure with B = [10001; 00010;00100;01011;10100], B(:,:,2) = [10101;0101 0;0 0 0 0 0; 1 0 0 1 0; 0 0 1 0 0].
  • [ FIG. 6 ]  (a) Extinction spectrum of a periodic 2D array with 3*5 particles. (b) Extinction spectrum of nonperiodic 2D array with 3*5 particles with B = [0 1 0;0 1 0;1 0 0;1 0 0;1 0 1].
    (a) Extinction spectrum of a periodic 2D array with 3*5 particles. (b) Extinction spectrum of nonperiodic 2D array with 3*5 particles with B = [0 1 0;0 1 0;1 0 0;1 0 0;1 0 1].
  • [ FIG. 7. ]  Extinction coefficient spectra for (a) a 1-layer periodic array, (b) a 2-layer periodic array. (c) a 3-layer periodic array, and (d) a 4-layer periodic array.
    Extinction coefficient spectra for (a) a 1-layer periodic array, (b) a 2-layer periodic array. (c) a 3-layer periodic array, and (d) a 4-layer periodic array.
  • [ FIG. 8. ]  (a) Extinction spectrum of the optimized array (3*5) with B = [ 0 1 0;0 1 1;0 0 1;1 0 0;1 0 0], (b) Convergence characteristic of BPSO.
    (a) Extinction spectrum of the optimized array (3*5) with B = [ 0 1 0;0 1 1;0 0 1;1 0 0;1 0 0], (b) Convergence characteristic of BPSO.
  • [ FIG. 9. ]  (a) Absorption spectrum of the optimized array (3*5) with B=[0 0 1;101;1 10;010;000]. (b) Scattering spectrum of the optimized array (3*5) with B = [0 01; 0 0 1;100;100;000].
    (a) Absorption spectrum of the optimized array (3*5) with B=[0 0 1;101;1 10;010;000]. (b) Scattering spectrum of the optimized array (3*5) with B = [0 01; 0 0 1;100;100;000].
(우)06579 서울시 서초구 반포대로 201(반포동)
Tel. 02-537-6389 | Fax. 02-590-0571 | 문의 : oak2014@korea.kr
Copyright(c) National Library of Korea. All rights reserved.