Documente Academic
Documente Profesional
Documente Cultură
DOI: 10.1109/JPHOT.2011.2128863
1943-0655/$26.00 ©2011 IEEE
IEEE Photonics Journal Generation of Surface Plasmon Polariton
Electronics and Photonics, Institute of High Performance Computing, A STAR, Singapore 138632
DOI: 10.1109/JPHOT.2011.2128863
1943-0655/$26.00 Ó2011 IEEE
Manuscript received December 21, 2010; revised February 25, 2011; accepted March 7, 2011.
Date of publication March 17, 2011; date of current version May 6, 2011. Corresponding author:
O. Kurniawan (e-mail: oka_kurniawan@sp.edu.sg).
Abstract: We study the generation of surface plasmon polariton and its extraction efficiency
from a microdisk-based plasmonic source to a metal–insulator–metal (MIM) waveguide. The
field distribution is no longer a whispering gallery mode commonly found in microdisk lasers
but hybridized with surface plasmon polariton. We observe the steady-state intensity of the
surface plasmon polariton mode inside the plasmonic microdisk structure at 1.47-m
wavelength. This field intensity is extracted efficiently to a MIM plasmonic waveguide
through a nanometer gap. The extraction efficiency reaches 60%.
1. Introduction
Nanoscale devices which work at the speed of light are exciting research areas. Conventional
photonic devices, however, have large dimensions due to the diffraction limit. Recently, it has been
proved theoretically and demonstrated experimentally that the diffraction limit can be overcome by
concentrating and channeling light using subwavelength structures, which are often called
plasmonic devices [1]. In plasmonic devices, the electromagnetic waves propagate at the interface
of a dielectric and a metal. An important research opportunity in this nanophotonics field is a surface
plasmon source.
The lack of surface plasmon source is normally resolved by using light source from free-space
which is then coupled to plasmonic devices. However, it is known that the surface plasmon mode
momentum hkS is greater than that of the free-space photon momentum at the same frequency
hk0 [2]. Therefore, a change of momentum is required to couple light from free-space to surface
plasmons at metal–dielectric interface. Three main techniques are commonly used to achieve this
coupling: using prisms [3], [4], scatterings [5], and gratings [6]. Laser is a commonly used light source
for these experiments. It is beneficial to have a surface plasmon source similar to lasers that can be
coupled directly to waveguides.
The search for a surface plasmon source has shown some progress. In 2003, Bergman and
Stockman showed that a stimulated amplification of plasmon mode can be achieved [7]–[9]. They
calculated the gain of the stimulated amplification emission when a metal structure was em-
bedded inside a gain medium. In this way, the local field was excited using the surface plasmon
amplification by stimulated emission of radiation (spaser), and the term spaser was introduced
for the first time [7]. Later on, it has been shown that spaser is able to create lasing effects
[10]–[13].
Fig. 1. Dark regions indicate metals and light regions indicate dielectric or semiconductor. (a)
Conventional microdisk laser. (b) Plasmonic microdisk with metal on top and bottom of the gain
medium. (c) Side view of the proposed structure, with a plasmonic microdisk coupled to a metal–
insulator–metal plasmonic waveguide. (d) Bird eye view of the proposed structure.
Recently, Akimov et al. showed that the light from a quantum dot can be coupled to the surface
plasmon of the metallic nanowires [14]. Koller et al. then successfully demonstrated a surface
plasmon emitter based on organic light emitting diodes and is useful for organic devices [15]. A year
later, Walters et al. proposed a silicon-based electrical source for surface plasmon polariton [16]. In
this structure, the surface plasmon was excited by using impact ionization process due to the high
electric field of the metal layers. They detected the surface plasmon by scattering out the wave
using nanoslits.
In this paper, we study an active device that emits surface plasmon polariton modes which can be
coupled directly to plasmonic devices, such as a waveguide. Such a device eliminates the
complexity of further coupling light from free space to plasmonic devices. The active device for the
surface plasmon source that we study in this work is based on a microdisk laser [17] and is shown in
Fig. 1. The modes of the microdisk resonator allow only certain frequencies to survive. The circular
shape acts as a resonator where light is reflected internally as it travels. It is well known that light
waves in a microdisk travel in a whispering gallery mode (WGM) fashion [18]. We will show later
that plasmonic microdisk behaves differently than conventional microdisk laser.
The main advantage of this structure is that a cavity structure is utilized to enhance the surface
plasmon polariton wave intensity, and that its cross section is similar to a metal–insulator–metal
(MIM) plasmonic waveguide [see Fig. 1(c) and (d)] [19], [20]. This enables one to efficiently
couple the surface plasmon polariton from the plasmonic microdisk to the MIM plasmonic
waveguide.
The plasmonic microdisk structure is studied by using a full wave 3-D finite difference time
domain (FDTD) code which incorporates combined dynamic thermal electron quantum medium
(DTEQM) and Lorentz-Drude (LD) dispersive models [21]. A 3-D simulation is required since the
light wave would normally travel in the x –y plane through a WGM while the metal confinement is on
the z-direction. This is the first reported study of such structure in three dimension. Furthermore,
our in-house FDTD code incorporates electron dynamics based on a multilevel multielectron
quantum system governed by the Pauli exclusion principle, state filling, and dynamical Fermi–Dirac
Thermalization [22]. This model enables one to treat various solid-state or atomic media. Most
previous studies were not able to take into account the electron dynamic in the semiconductor
media and only consider the semiconductor layer as a simple dielectric [16]. The metallic part, on
the other hand, is modeled by the LD dispersive model which is widely used for plasmonic
calculations [21].
Our calculation results show that the hybridized surface plasmon mode can be extracted
efficiently to the MIM waveguide. The high extraction efficiency is mainly caused by the gap
Fig. 2. (a) Electron dynamics in 4-level model. (b) Multilevel multielectron model for finite difference time
domain simulation inside the semiconductor material.
between the plasmonic microdisk and the plasmonic waveguide (see Fig. 1). We also study the
effect of varying gap distance between the plasmonic microdisk and waveguide.
where k ¼ x , y , and z, and h represent the dephasing rate at the hth energy level after excitation.
Ak is the vector potential, jUkh j2 ¼ 3
h "0 c 3 =!3ah h , and h is the Planck’s constant. NVh
0 0
and NCh are
initial values of volume densities of the energy states at level h in valence and conduction band,
respectively, while NVh and NCh are the current values. In addition, Ndiph ðr Þ is the dipole volume
density for level h within the energy width E and is specified by the number of dipoles N0 divided
by unit volume V . The interband transition frequency is denoted as !ah , while the interband
transition time is h . A more detailed description of the model can be found in [21]. This method-
ology allows us to take into account the electron transitions for stimulated emissions in the system.
Therefore, this model is able to simulate lasing effects in a microdisk structure.
To model the metallic part of the structure, we incorporated the LD dispersive model. The LD
dispersive model describes very well the relative permittivity "r of numerous metals at near infrared
and optical frequencies which are frequency dependent
!2D "
2L
"ð!Þ ¼ "ð1Þ 2 (4)
!ð! {D Þ ð!
2L Þ þ { L!
where !D is the plasma frequency, and D is the damping coefficient for the Drude model. The last
term of the above equation is a Lorentz pole, where " is a weighting factor,
L is the resonance
frequency, and L is the spectral width of the Lorentz model. For the Lorentz model a number of poles
can be included. Though multiple Lorentz pole can be added, the computation time will increase
significantly. In our case, a single Lorentz pole is found to be sufficient to study the structure [21].
The metal used in the structure is gold and parameters considered follows [27], i.e., "1 ¼
5:9673, !D =2 ¼ 2113:6 THz, D =2 ¼ 15:92 THz,
L =2 ¼ 650:07 THz, L =2 ¼ 104:86 THz,
and " ¼ 1:09.
Fig. 3. Near field pattern of (a) a conventional microdisk and (b) plasmonic microdisk. The total electric
field is in logarithmic scale. The field distribution is taken at the center plane of the semiconductor
material. The white circle denotes the microdisk edges.
Fig. 4. (a) Normalized electric field for conventional microdisk (WGM) and plasmonic microdisk (SPP).
The field in plasmonic microdisk is confined at the interface (dark vertical line at 380 nm) between the
metal and the semiconductor and decays exponentially inside the metal. (b) Intensity of the electric field
of plasmonic microdisk with the presence of metal–insulator–metal waveguide.
calculation shows that the direction of the dominant field changes from radial direction (TE mode) in a
conventional microdisk to the z (normal to the disk plane) direction in a plasmonic microdisk. In
[28] and [29], the study of a similar structure has been done but limited only to the loss and the Q
factor of the device.
To investigate this hybridization further, the electric field across the metal–semiconductor–metal
layer is measured. Fig. 4(a) shows the normalized field difference between a conventional microdisk
and plasmonic microdisk along the z-direction. We took the measurement at the maximum of the
electric field in Fig. 3 perpendicular to the microdisk plane (z-direction). In Fig. 4(a), the vertical line at
380 nm denotes the interface of the semiconductor with either air (for conventional microdisk) or metal
(for plasmonic microdisk). The region on the right hand side of 380 nm line is the semiconductor
medium, and the region on the left-hand side of 380 nm line is metal for the WGM–SPP curve and air
for the WGM curve.
In conventional microdisk the field is confined inside the dielectric, where the maximum is at the
center of the semiconductor layer. Fig. 4(a) indicates that the electromagnetic field for plasmonic
microdisk is no longer in a WGM but rather is hybridized with the surface plasmon polariton mode.
Fig. 5. (a) Normalized electric field in x –y plane. (b) Normalized electric field inside the plasmonic
waveguide in y –z plane. (c) The field intensity along half of the vertical cutline in (b). (d) Field intensity
along the horizontal cut line in (b). The full width half maximum of the intensity in (b) is about 400 nm.
The electric field is confined inside the metal–insulator–metal waveguide.
A surface plasmon mode would have a maximum intensity at the interface of dielectric and metal,
and the field decays exponential inside the metal. Fig. 4(a) shows that the hybridize mode confines
the field closer to the interface between the dielectric and the metal. It can be seen that the electric
field decays exponentially inside the metal (distance less than 380 nm). The mode observed here
seems to be a hybridization between the WGM and surface plasmon polariton. A pure surface
plasmon polariton would have the maximum field intensity at the interface. Such a mode is observed
inside the waveguide after the hybrid mode is coupled to the waveguide. This result will be presented
in Fig. 5(c).
Another observation is that attaching metal layers results in a significant enhancement of the
z component of the electric field as shown in Fig. 4(b). Fig. 4(b) shows the fast Fourier transform of
the z component electric field intensity in a logarithmic scale. The electric field is probed inside the
microdisk at the center of the semiconductor layer. At 1.47 m wavelength, the intensity enhance-
ment is about 20 000 times. Similar phenomena but for a single metal layer microdisk laser have
been observed experimentally by Perahia et al. [30]. They have shown that the WGMs start to
hybridize with the surface plasmon polariton modes as metal layer area increases. In our case, we
observed that the surface plasmon polariton wave at the top metal layers are coupled with the
surface plasmon polariton wave at the bottom metal layers.
Now, when a MIM plasmonic waveguide is placed near the microdisk, the surface plasmon
polariton mode is extracted to the waveguide through the nanometer gap. Fig. 5(a) shows the total
electric field distribution when the surface plasmon polariton wave couples to the waveguide at
50 000 time step. The field distribution is taken at the middle plane, which correspond to the center
of the semiconductor layer inside the microdisk or the center of the insulator inside the waveguide.
The field intensity inside the plasmonic microdisk is higher than the intensity inside the waveguide,
as expected. The microcavity builds up the field intensity and competes with the losses until it
reaches steady state. At steady state, we observed that the field inside the waveguide propagates
from the left end of the waveguide down to the far end.
Fig. 6. (a) Normalized electric field intensity as a function of wavelength for both the microdisk and the
metal–insulator–metal waveguide. Note that Ez is the dominant mode. (b) Extraction efficiency reaches
about 60% for the 20-nm gap distance and the efficiency decreases with an increasing gap distance.
Fig. 5(b) shows the electric field inside the waveguide in the y –z plane. We measured the field
along two cut lines which is shown in (c) and (d), respectively. Fig. 5(c) plots the surface plasmon
mode inside the waveguide where the electric field is confined at the interface between the metal
and the insulator. The electric field then decays exponentially inside the metal layers. Note that the
plot is only half side due to its symmetry. Fig. 5(d) shows the electric field is confined horizontally
inside the MIM waveguide. The full-width at half-maximum (FWHM) of the field intensity in the MIM
waveguide is about 400 nm, which is about the width of the waveguide, i.e., 360 nm.
Fig. 6(a) shows the fast Fourier transform of the electric field for both the plasmonic microdisk and
the plasmonic waveguide in logarithmic scale. The fields are measured at the center plane of the
semiconductor inside microdisk and the center plane of the dielectric layer inside the waveguide.
The observation point inside the microdisk is located about 20 nm from the gap adjacent to the
waveguide. On the other hand, the probe for field of the waveguide is located about 20 nm from the
gap inside the waveguide. It can be observed that the z component of the field inside the waveguide
follows that of the plasmonic microdisk for the wavelengths of interest from 1.3 m to 1.6 m. We
calculate the coupling efficiency by taking the Fourier transform of the field and compute the square
of the electric field, which gives us the energy. We computed the energy at both the microdisk and
the waveguide. It was calculated that the intensity at the waveguide is about 60% of that inside the
plasmonic microdisk. This is considered an efficient field extraction. As a rough comparison with a
previous work on conventional microdisk lasers, the light extraction efficiency of an elliptical micro-
disk laser was successfully increased from 9.2% to about 30% by using an external magnetic field
[25]. In another work, light from a conventional microdisk laser is coupled to a photonic waveguide
using a nanoantenna, and they claimed a maximum extraction efficiency of about 38% [31]. The
extraction efficiency of our proposed structure is 60%, which is higher than these two reported
results.
We should also compare qualitatively with other works on surface plasmon source. Akimov et al.
showed that the emission efficiency of their structure is about 27%, eventhough it was claimed that
theoretically, it should be able to reach 60% [14]. On the other hand, the extraction efficiency of
surface plasmon from organic LED was reported to be 4% in [15]. In that same work, however, it
was reported that it is possible for the organic LED to couple light emission to surface plasmon as
high as 40%. The extraction efficiency in this work gave a higher performance.
In addition, we also studied the effect of varying the gap distance between the microdisk and the
waveguides on the extraction efficiency. The gap distance between the plasmonic microdisk and
the waveguide is varied from 20 nm to 120 nm, with an interval of 20 nm. When varying the gap
distance, the other parameters are kept the constant. The effect is shown in Fig. 6(b). The plot
indicates that the extraction efficiency deteriorates as the gap distance increases. It is a well-known
fact that the field enhancement between two metal structures increases as the gap distance
decreases, and vice versa. A similar phenomenon is observed here. As the gap distance between
the microdisk and the waveguide decreases, the field enhancement in the gap increases. As a
result, higher extraction efficiency is obtained.
4. Conclusion
In this paper, a plasmonic structure that acts as a source for surface plasmon polariton waves has
been studied. The extraction efficiency is almost 60%. The extraction efficiency drops as the gap
distance between the plasmonic microdisk and the waveguide increases. Furthermore, it was
shown that the plasmonic waveguide perturbates the field inside the plasmonic microdisk. With this
device, the complexity of light extraction to a plasmonic device is simplified, and the plasmonic
source can be easily integrated on-chip. An integrated plasmonic source is essential for plasmonic
sensors, transistors, switches, and many others. Development of plasmonic source would pave the
way to enhance the applications of nanophotonics.
References
[1] S. A. Maier, Plasmonics: Fundamentals and Applications, 1st ed. New York: Springer-Verlag, May, 2007.
[2] W. L. Barnes, A. Dereux, and T. W. Ebbesen, BSurface plasmon subwavelength optics,[ Nature, vol. 424, no. 6950,
pp. 824–830, Aug. 2003.
[3] E. Kretschmann and H. Raether, BRadiative decay of nonradiative surface plasmons excited by light,[ Zeitschrift fuer
Naturforschung, Teil A: Astrophysik, Physik und Physikalische Chemie, vol. 23, no. 12, pp. 2135–2136, 1968.
[4] A. Otto, BExcitation of nonradiative surface plasma waves in silver by the method of frustrated total reflection,[ Z. Phys.,
vol. 216, no. 4, pp. 398–410, Aug. 1968.
[5] B. Hecht, H. Bielefeldt, L. Novotny, Y. Inouye, and D. W. Pohl, BLocal excitation, scattering, and interference of surface
plasmons,[ Phys. Rev. Lett., vol. 77, no. 9, pp. 1889–1892, Aug. 1996.
[6] R. H. Ritchie, E. T. Arakawa, J. J. Cowan, and R. N. Hamm, BSurface-plasmon resonance effect in grating diffraction,[
Phys. Rev. Lett., vol. 21, no. 22, pp. 1530–1533, Nov. 1968.
[7] D. J. Bergman and M. I. Stockman, BSurface plasmon amplification by stimulated emission of radiation: Quantum
generation of coherent surface plasmons in nanosystems,[ Phys. Rev. Lett., vol. 90, no. 2, p. 027402, Jan. 2003.
[8] M. I. Stockman, BSpasers explained,[ Nat. Photon., vol. 2, no. 6, pp. 327–329, Jun. 2008.
[9] K. Li, X. Li, M. I. Stockman, and D. J. Bergman, BSurface plasmon amplification by stimulated emission in nanolenses,[
Phys. Rev. B, Condens. Matter, vol. 71, no. 11, p. 115409, Mar. 2005.
[10] N. I. Zheludev, S. L. Prosvirnin, N. Papasimakis, and V. A. Fedotov, BLasing spaser,[ Nat. Photon., vol. 2, no. 6,
pp. 351–354, Jun. 2008.
[11] M. T. Hill, Y.-S. Oei, B. Smalbrugge, Y. Zhu, T. de Vries, P. J. van Veldhoven, F. W. M. van Otten, T. J. Eijkemans,
J. P. Turkiewicz, H. de Waardt, E. J. Geluk, S.-H. Kwon, Y.-H. Lee, R. Notzel, and M. K. Smit, BLasing in metallic-
coated nanocavities,[ Nat. Photon., vol. 1, no. 10, pp. 589–594, Oct. 2007.
[12] M. A. Noginov, G. Zhu, A. M. Belgrave, R. Bakker, V. M. Shalaev, E. E. Narimanov, S. Stout, E. Herz, T. Suteewong,
and U. Wiesner, BDemonstration of a spaser-based nanolaser,[ Nature, vol. 460, no. 7259, pp. 1110–1112,
Aug. 2009.
[13] R. F. Oulton, V. J. Sorger, T. Zentgraf, R.-M. Ma, C. Gladden, L. Dai, G. Bartal, and X. Zhang, BPlasmon lasers at deep
subwavelength scale,[ Nature, vol. 461, no. 7264, pp. 629–632, Oct. 2009.
[14] A. V. Akimov, A. Mukherjee, C. L. Yu, D. E. Chang, A. S. Zibrov, P. R. Hemmer, H. Park, and M. D. Lukin, BGeneration
of single optical plasmons in metallic nanowires coupled to quantum dots,[ Nature, vol. 450, no. 7168, pp. 402–406,
Nov. 2007.
[15] D. Koller, A. Hohenau, H. Ditlbacher, N. Galler, F. Reil, F. Aussenegg, A. Leitner, E. List, and J. Krenn, BOrganic
plasmon-emitting diode,[ Nat. Photon., vol. 2, no. 11, pp. 684–687, Nov. 2008.
[16] R. J. Walters, R. V. A. van Loon, I. Brunets, J. Schmitz, and A. Polman, BA silicon-based electrical source of surface
plasmon polaritons,[ Nat. Mater., vol. 9, no. 1, pp. 21–25, Jan. 2010.
[17] S. L. McCall, A. F. J. Levi, R. E. Slusher, S. J. Pearton, and R. A. Logan, BWhispering-gallery mode microdisk lasers,[
Appl. Phys. Lett., vol. 60, no. 3, pp. 289–291, Jan. 1992.
[18] K. J. Vahala, BOptical microcavities,[ Nature, vol. 424, no. 6950, pp. 839–846, Aug. 2003.
[19] L. Liu, Z. Han, and S. He, BNovel surface plasmon waveguide for high integration,[ Opt. Express, vol. 13, no. 17,
pp. 6645–6650, Aug. 2005.
[20] H. T. Miyazaki and Y. Kurokawa, BSqueezing visible light waves into a 3-nm-thick and 55-nm-long plasmon cavity,[
Phys. Rev. Lett., vol. 96, no. 9, p. 097401, Mar. 2006.
[21] I. Ahmed, E. Khoo, O. Kurniawan, and E. Li, BModeling and simulation of active plasmonics with the FDTD method by
using solid state and Lorentz–Drude dispersive model,[ J. Opt. Soc. Amer. B, Opt. Phys., vol. 28, no. 3, pp. 352–359,
Mar. 2011.
[22] Y. Huang and S.-T. Ho, BComputational model of solid-state, molecular, or atomic media for FDTD simulation based on
a multi-level multi-electron system governed by Pauli exclusion and Fermi–Dirac thermalization with application to
semiconductor photonics,[ Opt. Express, vol. 14, no. 8, pp. 3569–3587, Apr. 2006.
[23] C. Walther, G. Scalari, M. I. Amanti, M. Beck, and J. Faist, BMicrocavity laser oscillating in a circuit-based resonator,[
Science, vol. 327, no. 5972, pp. 1495–1497, Mar. 2010.
[24] M. Kuttge, F. J. Garcia de Abajo, and A. Polman, BUltrasmall mode volume plasmonic nanodisk resonators,[ Nano Lett.,
vol. 10, no. 5, pp. 1537–1541, May 2010.
[25] E. H. Khoo, I. Ahmed, and E. P. Li, BEnhancement of light energy extraction from elliptical microcavity using external
magnetic field for switching applications,[ Appl. Phys. Lett., vol. 95, no. 12, p. 121104, Sep. 2009.
[26] E. H. Khoo, E. P. Li, I. Ahmed, Y. Huang, and S. T. Ho, BLight energy extraction from the minor surface arc of an
electrically pumped elliptical microcavity laser,[ IEEE J. Quantum Electron., vol. 46, no. 1, pp. 128–136, Jan. 2010.
[27] A. Vial, A.-S. Grimault, D. Macias, D. Barchiesi, and M. L. de la Chapelle, BImproved analytical fit of gold dispersion:
Application to the modeling of extinction spectra with a finite-difference time-domain method,[ Phys. Rev. B, Condens.
Matter, vol. 71, no. 8, p. 085 416, Feb. 2005.
[28] Y.-H. Chen and L. J. Guo, BAnalysis of surface plasmon guided sub-wavelength microdisk cavity,[ in Proc. 21st Annu.
Meeting IEEE LEOS, 2008, pp. 320–321.
[29] M. Kim, P. Ku, and J. Guo, BSurface plasmon enabled sub-wavelength nano-cavity laser,[ in Proc. 20th Annu. Meeting
LEOS, 2007, pp. 252–253.
[30] R. Perahia, T. P. M. Alegre, A. H. Safavi-Naeini, and O. Painter, BSurface-plasmon mode hybridization in
subwavelength microdisk lasers,[ Appl. Phys. Lett., vol. 95, no. 20, pp. 201114-1–201114-3, Nov. 2009.
[31] H. T. Hattori, Z. Li, D. Liu, I. D. Rukhlenko, and M. Premaratne, BCoupling of light from microdisk lasers into plasmonic
nano-antennas,[ Opt. Express, vol. 17, no. 23, pp. 20 878–20 884, Nov. 2009.