Plasmonic Demultiplexer and
Chenglong Zhao and Jiasen Zhang*
lating light in two dimensions with sub-
wavelength scales,1have become a very ac-
tive and attractive research field.2,3Taking
advantage of the intrinsic two-dimensional
(2D) nature of SPPs, which provides many
opportunities for miniaturizing current opti-
cal devices to micro- and nanosize, various
elements for the 2D SPP optics4analogous
to the conventional 3D counterparts have
been developed, such as lenses,5?10
microscopy.15By combining those plas-
monic devices together, it may be possible
to develop an integrated plasmonic circuit
and provide a next-generation information
network with improved bandwidth and
Plasmonic demultiplexers, which are
length division multiplexing (WDM) sys-
tems, need to be developed, and a high
resolution is necessary to achieve a narrow
channel spacing in high capacity plasmonic
networks. Some SPP dispersing elements
have been demonstrated, such as the over-
lapping bull’s eye structure,17
SPP crystal structures for splitting SPPs of
different wavelengths into different
directions.19,20However, a practical plas-
monic demultiplexer needs to disperse
multiple-channel data streams at different
wavelengths spatially and focus them into
various SPP wavelength components.
In this work, we propose a plasmonic de-
multiplexer that can implement light?SPP
coupling, effective dispersion, and multiple-
channel SPP guiding. The resolution was as
high as 10 nm in the experiment, and fur-
ecent developments on optics based
on surface plasmon polaritons (SPPs)
or plasmonics, which allow manipu-
ther improvement is expected by using a
high refractive index superstrate. Besides
the applications in WDM systems, the plas-
monic demultiplexer can also be used as a
SPP spectroscope or filter.
RESULTS AND DISCUSSION
The plasmonic demultiplexer we pro-
posed is schematically shown in Figure 1a,
which is composed of concentric grooves
on a gold film with the center M and a ra-
dial width w. The projections of the grooves
on the y axis are spaced equidistantly and
form a grating with a period d. If a plane
wave illuminates the grooves normally
(along the z axis), SPPs are launched at the
grooves due to the momentum match be-
tween incoming light and SPPs through
scattering and propagate along the 2D
film.1It is expected that the diffracted SPPs
behave according to the standard grating
equation d sin ? ? m?SPP, where m is the or-
der number and an integer, ? is the dif-
fracted angle, and ?SPPis the wavelength of
the diffracted SPP. In order to focus the
SPPs of different wavelengths on a focal
circle (red circle in Figure 1a) with a diam-
eter R and touching the pole A of a grating
arc (green arc in Figure 1a) with a radius R,
the grooves need to be located on the
*Address correspondence to
Received for review June 14, 2010
and accepted September 29, 2010.
Published online October 6, 2010.
© 2010 American Chemical Society
KEYWORDS: surface plasmon polaritons · plasmonics · demultiplexers ·
strip waveguides · spectroscopy
www.acsnano.orgVOL. 4 ▪ NO. 11 ▪ 6433–6438 ▪ 2010
grating arc, which is a typical example of Rowland-
type mounting.21To blaze the SPPs at different wave-
length into intensive focal spots for the order number
m, the concentric structures are designed, and the ra-
dial difference between the adjacent grooves, which
determine the phase difference of the SPPs excited at
the adjacent grooves, is Dr? Rn?1? Rn? m?SPPC, where
?SPPCis the designed central SPP wavelength of the
plasmonic demultiplexer. As a result, the M point is the
focal spot of the SPPs with a wavelength of ?SPPC, and
the focal spot shifts along the circle when the wave-
length is changed. For different m values, the coordi-
nates of the groove center M are determined by xm?
R cos ? cos ?, ym? R cos ? sin ?, where ? ?
arcsin(m?SPPC/d). Therefore, the plasmonic demulti-
plexer can launch the input WDM signal to SPPs, split
various wavelength components spatially, and then fo-
cus them into individual SPP waveguides.
Three plasmonic demultiplexers with Dr? 813.5,
1627, and 2441 nm, which correspond to m ? 1, 2,
and 3 for ?SPPC? 813.5 nm, were fabricated (details
in the Methods below). All of those structures had
the same period d ? 4068 nm and total period num-
ber N ? 33, which corresponds to a total length
L ? 134 ?m along the y axis. The radial width of
the grooves is w ? 407 nm and R ? 100 ?m. The
scanning electron micrograph (SEM) of the plas-
monic demultiplexer with m ? 3 is shown in
Figure 1b. In order to increase the SPP intensity, an-
other two grooves were added to each period with a
separation of 813.5 nm in the radial direction (inset
of Figure 1b). The projections of the three grooves in
one period on the y axis are the same, so that they
do not break the periodical nature along the y axis.
In the experiment, leakage radiation microscopy22
(LRM) was used to detect the SPP images and the light
polarization was adjusted to be parallel to the line AM
(Figure 1a). The LRM SPP images of the plasmonic de-
multiplexers for m ? 1, 2, and 3 are shown in Figure
2a?c, respectively, and the parts of the corresponding
SEMs of the demultiplexers are shown in the insets.
The incident wavelength was 830 nm, corresponding
to a SPP wavelength of 813.5 nm.23The directly trans-
mitted light had been spatially filtered to make the SPP
images clean.22The LRM images in the Fourier plane
with and without filters are shown in Supporting Infor-
mation Figure S1. In Figure 2, three focal spots are
clearly seen, and their locations agree with the discus-
sion above. The distances between the three focal spots
and the pole of the grating circle A are 95, 88, and 76
?m, and the experimental transverse full widths at half-
maximum (fwhm) of the focal spots are 645, 656, and
670 nm for m ? 1, 2, and 3, respectively. Then, the
wavelength of the incident light was changed, and the
corresponding SPP images were recorded. In order to
determine the resolutions of the plasmonic demulti-
plexers, two SPP images with different wavelength
were summed up. Three results are shown in the in-
sets of Figure 2d?f with wavelength differences of 33,
20, and 15 nm for m ? 1, 2, and 3, respectively. The in-
tensity distributions along the centers of the two focal
spots (dashed lines in the insets) are depicted in Figure
2d?f. The two focal spots with different wavelengths
are clearly separated for the three cases. As a result, the
experimental resolutions of ?? ? 33, 20, and 15 nm
are obtained for m ? 1, 2, and 3, respectively. It can be
seen that a higher resolution is obtained with a higher
order number m.
The distinctive characteristics of the plasmonic de-
multiplexers compared with those of optical demulti-
plexers include the following: they have intrinsically
two-dimensional nature; there is an enhanced out-of-
plane electronic field component, a higher propagation
loss, and inhomogeneous exciting efficiency along the
demultiplexer. However, the theoretical resolution of
the plasmonic demultiplexers can be estimated as that
of a conventional optical grating based on the Rayleigh
where neffis the effective index of SPPs, and ?Cis the
corresponding central wavelength in vacuum. Accord-
ing to eq 1, the estimated theoretical wavelength reso-
Figure 1. (a) Schematic of the plasmonic demultiplexer. (b)
the detail structures in the red rectangular box.
∆λ ) neff× ∆λSPP) neff× λSPPC/mN ) λC/mN
VOL. 4 ▪ NO. 11 ▪ ZHAO AND ZHANGwww.acsnano.org
lutions of the above plasmonic demultiplexers are ??
? 25, 13, and 8 nm for m ? 1, 2, and 3, respectively,
which are smaller than that of the experimental results.
Besides the differences between the optical grating
the criterion used in the experiment is more critical
than the Rayleigh criterion.
A more accurate calculation of the resolution was
obtained by using the Huygens?Fresnel principle. Fig-
ure 3a,b shows the calculated
SPP intensity and experimental
result, respectively. It can be seen
that the theoretical result agrees
with the experimental one well,
which validates our theoretical
method. Two calculated SPP im-
ages with different wavelengths
were summed up to determine
the resolution. By using this
method, the theoretical resolu-
tions of the above plasmonic de-
multiplexers were calculated to
be 30, 15, and 10 nm for m ? 1,
2, and 3, respectively, based on
the Rayleigh criterion (see Sup-
porting Information Figure S2).
The results are closer to the ex-
According to eq 1, the resolu-
tion is determined by the prod-
uct mN. In the case of L ? 134 ?m
with an air superstrate, the maxi-
mum mN was achieved when m
? 3. Although the resolution can
be improved by increasing the to-
tal length L, the size of the plas-
monic demultiplexer is also in-
creased. Here, we used a
superstrate with a higher refrac-
tive index to decrease the SPP
resolution. Water (refractive in-
dex of 1.33) was dropped on the gold film, and the
SPP wavelength reduced to 602 nm for a vacuum wave-
length of 830 nm. In this way, the maximum mN was
achieved with m ? 4 while the total number of peri-
ods N and the total length L are the same as that in the
case of air superstrate. The parameters of the fabri-
cated plasmonic demultiplexer were as follows: ?SPPC
? 602 nm, L ? 134 ?m, w ? 301 nm, m ? 4, and N ?
33. The total length was kept the
same. Due to the shortening of the
SPP wavelength, which results in
a higher order number, the theo-
retical resolution according to eq 1
is improved to 6 nm. The SPP im-
ages with different incident wave-
length were taken and summed
up to compare with that of the
case with an air superstrate and m
? 3. The main results are shown in
Figure 4. The intensity profiles
were obtained along the centers
of the two focal spots (dashed
lines in the insets). It can be seen
that the water superstrate evi-
Figure 2. SPP images for m ? 1 (a), 2 (b), and 3 (c). Insets are parts of the corre-
sponding SEM images. Intensity profiles for m ? 1 (d), 2 (e), and 3 (f) with wave-
length differences of 33, 20, and 15 nm, respectively, across the focus centers of the
two focal spots (dashed line in the insets). Insets are the corresponding summed
up SPP images around focal areas for two wavelengths.
Figure 3. Comparison between the theoretical SPP image (a) and experimental re-
sult (b) for a incident wavelength of 830 nm and m ? 3 around the focal area.
www.acsnano.org VOL. 4 ▪ NO. 11 ▪ 6433–6438 ▪ 2010
dently improves the resolution. A resolution of 10 nm
was obtained, which is higher than the resolution with
an air superstrate. When the wavelength difference was
increased to 15 nm, the two spots were completely
separated, as shown in Figure 4d. At the same time,
the fwhm of the focal spot decreased to 500 nm for the
vacuum wavelength of 830 nm. In a practical plas-
monic demultiplexer, a superstrate with a higher refrac-
tive index, such as InP that is widely used in the tele-
communication system, can be used to obtain a higher
In a telecommunication system, the WDM signal
that is spatially dispersed with respect to wavelength
by the demultiplexer should be coupled to various
waveguides. Metal strip waveguides have emerged for
the advantages of confining light energy on subwave-
length scales.24?27In the experi-
ment, five strip waveguides with a
width of 2 ?m were fabricated us-
ing FIB and located on the focal
circle of the above demultiplexer
with an air superstrate and m ? 3
(SEM image can be found in Sup-
porting Information Figure S3). The
detailed SEM of the waveguides is
shown in Figure 5a. The separation
between two adjacent strip
waveguides was 1 ?m. The orienta-
tions of the strip waveguides were
aligned to be parallel to the line AM
(Figure 1a), and the entrance of
each waveguide was located on
the focal circle. Figure 5b shows the
SPP image for a vacuum wave-
length of 847 nm. Here, the di-
rectly transmitted light was not
blocked for the sake of having a
clear image of the waveguides (the
image of SPP guiding in the
waveguide with directly transmit-
ted light blocked is shown in Sup-
porting Information Figure S4). The fringes came from
the interference between the SPPs and the directly
transmitted light. The SPP images for the vacuum wave-
lengths of 806 and 847 nm are shown in Figure 5c,d, re-
spectively. It is clearly seen that the SPPs with two dif-
ferent wavelengths were fed into different waveguides
and propagated along them. The fringes in the areas
where gold was removed were caused by the interfer-
ence between the directly transmitted light and the
scattering light of SPPs by the gold remnant in these ar-
eas after FIB fabrication. The linear dispersion, ?l, of
the plasmonic demultiplexer can be obtained by using
the grating equation as
where ?? is the wavelength difference of the
incident light. In the experiment, ?l ? 3 ?m,
then ?? ? 33 nm, which agrees well with the
experimental result of 41 nm.
We have proposed and experimentally
demonstrated plasmonic demultiplexers by
controlling the phase of SPPs excited at
concentric grooves fabricated at the dielec-
tric/metal interface. Besides dispersing and
multiple-channel SPP guiding, 3D propagat-
ing light mode to SPP mode coupling was
also implemented simultaneously. There-
fore, the plasmonic demultiplexers can act
Figure 4. Comparison between the cases of air and water superstrates. Intensity
profiles with wavelength differences of 10 (a) and 15 nm (b) for the case of air su-
perstrate across the focus centers of the two focal spots (dashed line in the in-
sets). Intensity profiles with wavelength differences of 10 (c) and 15 nm (d) for the
case of water superstrate across the focus centers of the two focal spots (dashed
Figure 5. (a) SEM image of the strip waveguides. (b) SPP image for a
incident wavelength of 847 nm. Enlarged SPP images around the
waveguides for incident wavelengths of (c) 806 and (d) 847 nm.
VOL. 4 ▪ NO. 11 ▪ ZHAO AND ZHANGwww.acsnano.org
as converters to connect 3D conventional
diffraction-limited photonic devices and 2D SPP de-
vices in a WDM system. A resolution as high as 10 nm
was obtained in the experiment, and further im-
provement can be realized by using a superstrate
with high refractive index. Due to the linear disper-
sion characteristic, the plasmonic demultiplexer can
be used as a plasmonic spectroscopy or filter, and
more applications, such as optical manipulation,28
are also expected.
Experiment. Gold film with a thickness of 50 nm was ther-
mally evaporated onto a 2 cm ? 2 cm glass substrate. Then the
plasmonic demultiplexers were fabricated on the gold film using
a focused ion-beam (FIB) milling system. A laser beam from a
tunable cw Ti:sapphire laser was focused by a cylindrical lens to
illuminate the plasmonic demultiplexer. The polarization of the
light source was tuned by a half-wave plate. A leakage radiation
microscopy (LRM) was used to detect the SPPs, which consisted
of an oil-immersion objective (100?, numerical aperture of 1.4),
three lenses, and a spatial filter in front of Lens2 for spatially
blocking the directly transmitted laser light through the gold
film. The SPP intensity distribution was recorded by a charge-
coupled device (CCD) camera. The details of the LRM can be
found in ref 22 (for the experiment setup, see the Supporting In-
formation Figure S5).
Theory. The theoretical calculation of the SPP field was ob-
tained by using the Huygens?Fresnel principle. Each point of
the plasmonic demultiplexer was considered as a secondary SPP
point source. The resulting SPP field can be obtained by adding
up all of the field of the SPP point source and taking into account
its phase and amplitude. A complex phasor form of a SPP point
source, which located at the origin and x-polarized is29
E(F ˆ,? ˆ,z ˆ)SPP) A(z ˆ -
where A is a constant, kSPPis the wave vector of SPP, and
was considered in the calculation.
(1)(kSPP?) is the m ? 1 Hankel function. The propagation loss
Acknowledgment. This work was supported by the National
Natural Science Foundation of China under Grants 61036005
and 11074015, the Research Fund for the Doctoral Program of
Higher Education under Grant 20090001110010, and the Na-
tional Basic Research Program of China under Grants
2009CB623703 and 2007CB307001.
Supporting Information Available: Figure S1 shows the Fou-
rier plane LRM images without and with filter. Figure S2 shows
theoretical intensity profiles for m ? 1, 2, and 3 across the focus
centers of the two focal spots. Figure S3 shows the SEM image
of a plasmonic demultiplexer with m ? 3 integrated with five
strip waveguides. Figure S4 shows the images of SPP guiding in
one of the strip waveguides without and with filter. Figure S5 is a
schematic of the experiment setup. This material is available
free of charge via the Internet at http://pubs.acs.org.
REFERENCES AND NOTES
1. Barnes, W. L.; Dereux, A.; Ebbesen, T. W. Surface Plasmon
Subwavelength Optics. Nature 2003, 424, 824–830.
2. Gramotnev, D. K.; Bozhevolnyi, S. I. Plasmonics beyond the
Diffraction Limit. Nat. Photonics 2010, 4, 83–91.
3. Schuller, J. A.; Barnard, E. S.; Cai, W.; Jun, Y. C.; White, J. S.;
Brongersma, M. L. Plasmonics for Extreme Light
Concentration and Manipulation. Nat. Mater. 2010, 9, 193–
4. Zayats, A. V.; Smolyaninov, I. I.; Maradudin, A. A. Nano-
optics of Surface Plasmon Polaritons. Phys. Rep. 2005, 408,
5. Liu, Z.; Steele, J. M.; Srituravanich, W.; Pikus, Y.; Sun, C.;
Zhang, X. Focusing Surface Plasmons with a Plasmonic
Lens. Nano Lett. 2005, 5, 1726–1729.
6. Liu, Z.; Steele, J. M.; Lee, H.; Zhang, X. Tuning the Focus of
a Plasmonic Lens by the Incident Angle. Appl. Phys. Lett.
2005, 88, 171108/1–171108/3.
7. Yin, L.; Vlasko-Vlasov, V. K.; Pearson, J.; Hiller, J. M.; Hua, J.;
Welp, U.; Brown, D. E.; Kimball, C. W. Subwavelength
Focusing and Guiding of Surface Plasmons. Nano Lett.
2005, 5, 1399–1402.
8. Radko, I. P.; Bozhevolnyi, S. I.; Evlyukhin, A. B.; Boltasseva,
A. Surface Plasmon Polariton Beam Focusing with
Parabolic Nanoparticle Chains. Opt. Express 2007, 15,
9. Evlyukhin, A. B.; Bozhevolnyi, S. I.; Stepanov, A. L.; Kiyan, R.;
Reinhardt, C.; Passinger, S.; Chichkov, B. N. Focusing and
Directing of Surface Plasmon Polaritons by Curved Chains
of Nanoparticles. Opt. Express 2007, 15, 16667–16680.
10. Wang, J.; Zhang, J.; Wu, X.; Luo, H.; Gong, Q.
Subwavelength-Resolved Bidirectional Imaging between
Two and Three Dimensions Using a Surface Plasmon
Launching Lens. Appl. Phys. Lett. 2009, 94,
11. Ditlbacher, H.; Hohenau, A.; Wagner, D.; Kreibig, U.; Rogers,
M.; Hofer, F.; Aussenegg, F. R.; Krenn, J. R. Silver Nanowires
as Surface Plasmon Resonators. Phys. Rev. Lett. 2005, 95,
12. Liu, L.; Han, Z.; He, S. Novel Surface Plasmon Waveguide
for High Integration. Opt. Express 2005, 13, 6645–6650.
13. Ditlbacher, H.; Krenn, J. R.; Schider, G.; Leitner, A.;
Aussenegg, F. R. Two-Dimensional Optics with Surface
Plasmon Polaritons. Appl. Phys. Lett. 2002, 81, 1762–1764.
14. Bozhevolnyi, S. I.; Volkov, V. S.; Deveaux, E.; Laluet, J. Y.;
Ebbesen, T. W. Channel Plasmon Subwavelength
Waveguide Components Including Interferometers and
Ring Resonators. Nature 2006, 440, 508–511.
15. Drezet, A.; Koller, D.; Hohenau, A.; Leitner, A.; Aussenegg,
F. R.; Krenn, J. R. Surface Plasmon Polariton Microscope
with Parabolic Reflectors. Opt. Lett. 2007, 32, 2414–2416.
16. Zia, R.; Schuller, J. A.; Chandran, A.; Brongersma, M. L.
Plasmonics: The Next Chip-Scale Technology. Mater. Today
2006, 9, 20–27.
17. Laux, E.; Genet, C.; Skauli, T.; Ebbesen, T. W. Plasmonic
Photon Sorters for Spectral and Polarimetric Imaging. Nat.
Photonics 2008, 2, 161–164.
18. Kenneth, D.; Jennifer, A. D.; Merrielle, S.; Harry, A. A.
Tunable Color Filters Based on Metal?Insulator?Metal
Resonators. Nano Lett. 2009, 9, 2579–2583.
19. Drezet, A.; Koller, D.; Hohenau, A.; Leitner, A.; Aussenegg,
F. R.; Krenn, J. R. Plasmonic Crystal Demultiplexer and
Multiports. Nano Lett. 2007, 7, 1697–1700.
20. Mikhailov, V.; Wurtz, G. A.; Elliott, J.; Bayvel, P.; Zayats, A. V.
Dispersing Light with Surface Plasmon Polaritonic Crystals.
Phys. Rev. Lett. 2007, 99, 083901/1–083901/4.
21. Hutley, M. C. Diffraction Gratings; Academic: London, 1982.
22. Drezet, A.; Hohenau, A.; Stepanov, A. L.; Ditlbacher, H.;
Steinberger, B.; Galler, N.; Aussenegg, F. R.; Leitner, A.;
Krenn, J. R. How To Erase Surface Plasmon Fringes. Appl.
Phys. Lett. 2006, 89, 091117/1–091117/3.
23. Raether, H. Surface Plasmons on Smooth and Rough
Surfaces and on Grating; Springer: Berlin, 1988; pp
24. Zia, R.; Selker, M. D.; Brongersma, M. L. Leaky and Bound
Modes of Surface Plasmon Waveguides. Phys. Rev. B. 2005,
25. Zia, R.; J. Schuller, A.; Brongersma, M. L. Near-Field
Characterization of Guided Polariton Propagation and
www.acsnano.orgVOL. 4 ▪ NO. 11 ▪ 6433–6438 ▪ 2010
Cutoff in Surface Plasmon Waveguides. Phys. Rev. B 2006, Download full-text
26. Weeber, J. C.; Krenn, J. R.; Dereux, A.; Lamprecht, B.;
Lacroute, Y.; Goudonnet, J. P. Near-Field Observation of
Surface Plasmon Polariton Propagation on Thin Metal
Stripes. Phys. Rev. B 2001, 64, 045411/1–045411/9.
27. Weeber, J. C.; Lacroute, Y.; Dereux, A. Optical Near-Field
Distributions of Surface Plasmon Waveguide Modes. Phys.
Rev. B 2003, 68, 115401/1–115401/10.
28. Quidant, R; Girard, C. Surface-Plasmon-Based Optical
Manipulation. Laser Photon. Rev. 2008, 2, 47–57.
29. Chang, S.-H.; Gray, S. K.; Schatz, G. C. Surface Plasmon
Generation and Light Transmission by Isolated Nanoholes
and Arrays of Nanoholes in Thin Metal Films. Opt. Express
2005, 13, 3150–3165.
VOL. 4 ▪ NO. 11 ▪ ZHAO AND ZHANGwww.acsnano.org