Professional Documents
Culture Documents
Series Editors: Chennupati Jagadish
Series Editors: Chennupati Jagadish
CHENNUPATI JAGADISH
Distinguished Professor
Department of Electronic Materials Engineering
Research School of Physics and Engineering
Australian National University
Canberra, ACT2601, Australia
ZETIAN MI
Professor
Department of Electrical Engineering and Computer Science
University of Michigan
1310 Beal Avenue
Ann Arbor, MI 48109
United States of America
Academic Press is an imprint of Elsevier
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
525 B Street, Suite 1650, San Diego, CA 92101, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
125 London Wall, London, EC2Y 5AS, United Kingdom
No part of this publication may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher. Details on how to seek permission, further
information about the Publisher’s permissions policies and our arrangements with organizations such
as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website:
www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume
any liability for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas
contained in the material herein.
ISBN: 978-0-12-817542-2
ISSN: 0080-8784
Isabel Barth
The Department of Physics, University of York, York, United Kingdom. (ch5)
Alexander Drayton
The Department of Physics, University of York, York, United Kingdom. (ch5)
Shanhui Fan
Ginzton Laboratory and Department of Electrical Engineering, Stanford University,
Stanford, CA, United States. (ch1,4,6)
Cheng Guo
Department of Applied Physics, Stanford University, Stanford, CA, United States. (ch4)
Yu Guo
Electrical Engineering and Ginzton Laboratory, Stanford University, Stanford, CA,
United States. (ch3)
Yuri S. Kivshar
Department of Physics and Engineering, ITMO University, St. Petersburg, Russia;
Nonlinear Physics Center, Australian National University, Canberra, ACT, Australia. (ch2)
Thomas F. Krauss
The Department of Physics, University of York, York, United Kingdom. (ch5)
Mikhail F. Limonov
Ioffe Institute; Department of Physics and Engineering, ITMO University, St. Petersburg,
Russia. (ch2)
Mikhail V. Rybin
Ioffe Institute; Department of Physics and Engineering, ITMO University, St. Petersburg,
Russia. (ch2)
Ken Xingze Wang
School of Physics, Huazhong University of Science and Technology, Wuhan, Hubei, China.
(ch3)
Zongfu Yu
University of Wisconsin, Department of Electrical and Computer Engineering, Madison,
WI, United States. (ch3)
Weidong Zhou
Department of Electrical Engineering, University of Texas at Arlington, Arlington, TX,
United States. (ch1,6,7)
ix
Preface
Contents
1. Introduction 1
2. Principles of Fano resonance in photonic crystal slabs 2
2.1 2D photonic crystal slabs 2
2.2 The presence of Fano resonance in photonic crystal slabs 3
2.3 Characteristics of Fano resonance in photonic crystal slabs 5
References 10
1. Introduction
In contrast to a conventional symmetric-shaped Lorentzian resonance,
Fano resonance is predominantly used to describe asymmetric resonance
(Fano, 1961; Ott et al., 2013), which arises from the constructive and
destructive interference of discrete resonance states by broadband continuum
states. First observed from the transmission spectra of gratings, Fano reso-
nance is responsible for one variety of Wood’s anomalies (Wood, 1902). This
phenomenon and the underlying mechanism, being common and ubiquitous
in many realms of physical sciences (Fano, 1961; Miroshnichenko et al.,
2010), are found in a wide variety of nanophotonic structures and quantum
systems, such as quantum dots, photonic crystals (PhCs), plasmonics, and
metamaterials. The asymmetric and steep dispersion of the Fano resonance
profile promises applicability in a wide range of photonic devices, such
as optical filters, switches, sensors, broadband reflectors, lasers, detectors,
slow-light, and nonlinear devices. With the advances in nanofabrication pro-
cesses and the development of integration techniques, Fano resonance-based
materials, structures, devices, and systems have been widely investigated and
reported over the last few years. There are also excellent reviews written
A B
Photonic crystal slab (PCS) ment
Sp Em
issi
nfine on
co co
on nt
ot PBG r
ol
Ph
Photonic
bandgap
En
e
GVA PDOS
tim
han
Group Photonic
interaction
velocity density of
ced absorption
anomaly states
PhC
0.8 Photonic
crystals
anced
Frequency (c/a)
Enh
refractive
Sup
plasmon
0.4 Photonic bandgap index
NLO
er
TM
n
Nonlinear
pr
tio
is
0.2 a M K a optical m
2r
er properties
en An
G cg iso
TE oni trop
0 Second harm y
G M K G
Fig. 1 Two-dimensional photonic crystal slabs (2D PCS): (A) Schematic representation of
triangular lattice PCS, with simulated dispersion plot for two (TE/TM) polarizations; and
(B) Overview of photonic crystal properties and applications. Reproduced with permission
from Zhou, W., Zhao, D., Shuai, Y.-C., Yang, H., Chuwongin, S., Chadha, A., Seo, J.-H.,
Wang, K. X., Liu, V., Ma, Z., 2014. Progress in 2D photonic crystal Fano resonance photonics.
Prog. Quantum Electron. 38 (1), 1–74.
where td and rd are the direct transmission and reflection coefficients, ω0 and
γ are the center frequency and width of the resonance, and f is the normal-
ized complex amplitude of the resonant mode. The plus and minus sign
correspond with even and odd resonant modes for structures with the mirror
symmetry. Energy conservation requires
f ¼ ðtd rd Þ (3)
For 0 < jtd j < 1, Eq. (1) gives a Fano resonance lineshape. For example,
we consider a system consisting of a square lattice of air holes introduced into
a dielectric slab (Fig. 2A). The intensity transmission spectrum is given in
Fig. 2B and C, exhibiting Fano resonance lineshapes with strong asymmetric
profiles. Using finite-difference time-domain (FDTD) simulations, one can
determine the parameters for Eq. (1) and compare the spectrum from this
theoretical derivation to that from the FDTD simulation (Fig. 2B and C).
Fano resonance principles in photonic crystal slabs 5
A B 1
Transmission
0.8
0.6
0.4
0.2
0
0.25 0.3 0.35 0.4 0.45
Frequency (c/a)
C
1
Transmission
0.8
0.6
0.4
0.2
0
0.36 0.38 0.4 0.42
Frequency (c/a)
Fig. 2 Fano resonances in 2D photonic crystal slabs: (A) Photonic crystal structure con-
sisting of a square lattice of air holes with a radius of r ¼ 0.2a in a dielectric slab with
dielectric constant nh ¼ 12 and a thickness (t) of 0.5a, where a is the lattice constant.
The arrow indicates the direction of the incident light. (B) The intensity transmission
spectrum through such a structure. The circles are the result of the finite-difference
time-domain simulations. The solid curve is determined by the analytic theory as repre-
sented by Eq. (1). (C) The same plot as (B), except the frequency range is now restricted
to [0.36(c/a), 0.42(c/a)] to exhibit further details of the resonance line shape. Reproduced
with permission from Fan, S., Suh, W., Joannopoulos, J., 2003. Temporal coupled-mode
theory for the Fano resonance in optical resonators. J. Opt. Soc. Am. 20 (3), 569–572.
coupled to two ports to better illustrate the essence of the physical process.
This two-port configuration could be guaranteed in the structure when the
light’s wavelength is greater than the lattice constant such that there is no
diffraction above or below the slab, as shown in Fig. 2A.
The dynamics of this lossless single-resonance double-port system is
described by the coupled-mode equations
da 1 s1 +
¼ iω0 a + ðκ 1 κ2 Þ (4)
dt τ s2 +
s1 s1 + d1
¼C +a (5)
s2 s2 + d2
where a is the amplitude of the resonant mode such that j aj2 represents the
electromagnetic energy stored in the resonance. ω0 and γ ¼ 1/τ are the cen-
ter frequency and the width of the resonance, respectively. s1+ and s2+ are
the incoming amplitudes of the plane waves on the two ports such that j s1+j2
and js2+j2 correspond with the intensities. s1 and s2 are the outgoing
amplitudes of the plane waves from the two ports such that j s1j2 and js2j2
correspond with the intensities. κ1 and κ2 are the coupling constants between
the incoming waves and the resonance. d1 and d2 are the coupling constants
between the outgoing waves and the resonance. C is the background scatter-
ing matrix.
The coupling constants are constrained by energy conservation and
time-reversal symmetry as follows (Suh et al., 2004):
2
jd1 j2 + jd2 j2 ¼ (6)
τ
κ 1 ¼ d1 ; κ2 ¼ d2 (7)
∗
d1 d1
wC ∗ ¼ (8)
d2 d2
By choosing the appropriate positions of the reference planes, without
loss of generality, the direct scattering matrix C can be expressed as
iϕ r it
C¼e (9)
it r
where ϕ is a phase factor, and r and t, both being real, are the reflection and
transmission coefficients of the direct process satisfying r2 + t2 ¼ 1 (Haus,
1984). Combining Eqs. (6–9), one can solve for the coupling constants in
terms of the decay rate and the elements in the direct scattering matrix
Fano resonance principles in photonic crystal slabs 7
(Wang et al., 2013), and the theory accounts for asymmetric photonic crystal
slabs (Rosenberg et al., 2005). For the structure with mirror symmetry in
Fig. 2A, one can further simplify the solution by requiring d1 ¼ d2, and
the intensity reflection coefficient R is therefore
A B
1
PML
Source 0.8 (i) High Q
Transmission
0.6
PT
0.4 (ii) Medium Q
C
PB
PBC 0.2
PB (iii) Low Q
0
1.44 1.49 1.54 1.59 1.64
PML Wavelength (µm)
C
1
(iii) Low Q
0.8
Reflection
0.4
(i) High Q
0.2
0
1.44 1.49 1.54 1.59 1.64
Wavelength (µm)
Fig. 3 Simulation and spectral control of Fano resonances in 2D PCS: (A) Schematic
of an unit cell used in the simulation based on 3D-FDTD technique; Simulated
(B) transmission and (C) reflection spectra for surface-normal Fano filters with different
quality factors (Qs) for either highly spectrally selective filters or broadband reflectors.
Reproduced with permission from Zhou, W., Ma, Z., Yang, H., Qiang, Z., Qin, G., Pang, H.,
Chen, L., Yang, W., Chuwongin, S., Zhao, D., 2009. Flexible photonic-crystal Fano filters
based on transferred semiconductor nanomembranes. J. Phys. D Appl. Phys. 42, 234007.
On the other hand, a larger r/a value can lead to a lower Q filters for broad-
band reflector design (r/a ¼ 0.28, t/a ¼ 0.347, a ¼ 980 nm for low Q case
here). A medium Q design was also presented with r/a ¼ 0.19, t/a ¼ 0.417,
and a ¼ 600nm.
The resonance modes were further verified with the field propagation
plots based on 3D FDTD simulations. Snapshots of field propagation for
the on- and the off-resonance modes are shown in Fig. 4A and B, respec-
tively. Note that, for the on-resonance mode (λ1), the surface-normal
incident light is reflected from the patterned PCS structure due to the coher-
ence (in-phase) reflection, which leads to a dip in the transmission spectra.
Fano resonance principles in photonic crystal slabs 9
–Max
Fig. 4 Snap shots of electric field distribution of Fano filters at (A) on-resonance and
(B) off-resonance conditions. Reproduced with permission from Zhou, W., Ma, Z.,
Yang, H., Qiang, Z., Qin, G., Pang, H., Chen, L., Yang, W., Chuwongin, S., Zhao, D., 2009.
Flexible photonic-crystal Fano filters based on transferred semiconductor nanomembranes.
J. Phys. D Appl. Phys. 42, 234007.
0.5
I II w3
Frequency (wa/2pc=a/l)
w2
0.4
w1
0.3
0.2 M
0.1 G X
0.0
G X M G
Fig. 5 Simulated dispersion characteristics for the square lattice Si PCS structure on low
index glass substrate. The lattice parameters are r/a ¼ 0.19, t/a ¼ 0.417, the refractive
indices of silicon and glass are 3.48 and 1.5, respectively. Reproduced with permission
from Qiang, Z., Yang, H., Chen, L., Pang, H., Ma, Z., Zhou, W., 2008. Fano filters based
on transferred silicon nanomembranes on plastic substrates. Appl. Phys. Lett. 93, 061106.
References
Boutami, S., Bakir, B.B., Hattori, H., Letartre, X., Leclercq, J.L., Rojo-Romeo, P.,
Garrigues, M., Seassal, C., Viktorovitch, P., 2006. Broadband and compact 2-D
photonic crystal reflectors with controllable polarization dependence. IEEE Photon.
Technol. Lett. 18 (7), 835–837.
Boutami, S., Bakir, B., Regreny, P., Leclercq, J., Viktorovitch, P., 2007. Compact 1.55 μm
room-temperature optically pumped VCSEL using photonic crystal mirror. Electron.
Lett. 43, 282.
Carletti, L., Malureanu, R., Mørk, J., Chung, I.S., 2011. High-index-contrast grating
reflector with beam steering ability for the transmitted beam. Opt. Express 19 (23),
23567–23572.
Chang-Hasnain, C.J., 2011. High-contrast gratings as a new platform for integrated
optoelectronics. Semicond. Sci. Technol. 26, 014043.
Chen, L., Huang, M.C.Y., Mateus, C.F.R., Chang-Hasnain, C.J., Suzuki, Y., 2006.
Fabrication and design of an integrable subwavelength ultrabroadband dielectric mirror.
Appl. Phys. Lett. 88, 031102.
Fan, S., 2002. Sharp asymmetric line shapes in side-coupled waveguide-cavity systems. Appl.
Phys. Lett. 80 (6), 908.
Fan, S., Joannopoulos, J.D., 2002. Analysis of guided resonances in photonic crystal slabs.
Phys. Rev. B 65 (23), 235112.
Fan, S., Suh, W., Joannopoulos, J., 2003. Temporal coupled-mode theory for the Fano
resonance in optical resonators. J. Opt. Soc. Am. 20 (3), 569–572.
Fano, U., 1961. Effects of configuration interaction on intensities and phase shifts. Phys. Rev.
124 (6), 1866.
Giannini, V., Francescato, Y., Amrania, H., Phillips, C.C., Maier, S.A., 2011. Fano
resonances in nanoscale plasmonic systems: a parameter-free modeling approach. Nano
Lett. 11 (7), 2835–2840.
Haus, H.A., 1984. Waves and Fields in Optoelectronics. Prentice-Hall, Englewood
Cliffs, NJ.
Joannopoulos, J.D., Johnson, S.G., Winn, J.N., Meade, R.D., 2008. Photonic Crystals:
Molding the Flow of Light, second ed. Princeton University Press.
Fano resonance principles in photonic crystal slabs 11
John, S., Toader, O., Chutinan, A., 2004. Photonic band gap architectures for
micro-fabrication and diffractionless optical networking. In: IEICE Transactions on
Electronics. vol. E87-C(3). The Institute of Electronics, Information and Communication
Engineers, pp. 266–273.
Johnson, K., 1981. Coupled scalar wave diffraction theory. Appl. Phys. 24 (3), 249–260.
Johnson, S.G., Shanhui, F., Villeneuve, P.R., Joannopoulos, J.D., Kolodziejski, L.A.,
1999. Guided modes in photonic crystal slabs. Phys. Rev. B Condens. Matter 60 (8),
5751–5758.
Johnson, S.G., Povinelli, M.L., Joannopoulos, J.D., 2001. New photonic-crystal
system for integrated optics. In: Active and Passive Optical Components for WDM
Communication. SPIE-Int. Soc. Opt. Eng, Denver, CO.
Kanamori, Y., Kitani, T., Hane, K., 2007. Control of guided resonance in a photonic crystal
slab using microelectromechanical actuators. Appl. Phys. Lett. 90, 031911.
Kanskar, M., Paddon, P., Pacradouni, V., Morin, R., Busch, A., Young, J.F., Johnson, S.R.,
MacKenzie, J., Tiedje, T., 1997. Observation of leaky slab modes in an air-bridged
semiconductor waveguide with a two-dimensional photonic lattice. Appl. Phys. Lett.
70 (11), 1438–1440.
Kikuta, H., Toyota, H., Yu, W., 2003. Optical elements with subwavelength structured
surfaces. Opt. Rev. 10 (2), 63–73.
Lin, C., Lu, Z., Shi, S., Jin, G., Prather, D.W., 2005. Experimentally demonstrated filters
based on guided resonance of photonic-crystal films. Appl. Phys. Lett. 87, 091102.
Liu, V., Fan, S., 2012. S4: a free electromagnetic solver for layered periodic structures.
Comput. Phys. Commun. 183 (10), 2233–2244.
Luk’yanchuk, B., Zheludev, N.I., Maier, S.A., Halas, N.J., Nordlander, P., Giessen, H.,
Chong, C.T., 2010. The Fano resonance in plasmonic nanostructures and metamaterials.
Nat. Mater. 9 (9), 707–715.
Magnusson, R., Shokooh-Saremi, M., 2008. Physical basis for wideband resonant reflectors.
Opt. Express 16 (5), 3456–3462.
Magnusson, R., Wang, S.S., 1992. New principle for optical filters. Appl. Phys. Lett.
61, 1022.
Mekis, A., Chen, J.C., Kurland, I., Fan, S., Villeneuve, P.R., Joannopoulos, J.D., 1996. High
transmission through sharp bends in photonic crystal waveguides. Phys. Rev. Lett.
77 (18), 3787–3790.
Miroshnichenko, A.E., Flach, S., Kivshar, Y.S., 2010. Fano resonances in nanoscale structures.
Rev. Mod. Phys. 82 (3), 2257.
Moharam, M., Gaylord, T., 1981. Rigorous coupled-wave analysis of planar-grating diffraction.
J. Opt. Soc. Am. 71 (7), 811–818.
Moharam, M., Grann, E.B., Pommet, D.A., Gaylord, T., 1995. Formulation for stable and
efficient implementation of the rigorous coupled-wave analysis of binary gratings. J. Opt.
Soc. Am. A 12 (5), 1068–1076.
Noda, S., Baba, T., 2003. Roadmap on Photonic Crystals. Springer.
Noda, S., Imada, M., Okano, M., Ogawa, S., Mochizuki, M., Chutinan, A., 2002.
Semiconductor three-dimensional and two-dimensional photonic crystals and devices.
IEEE J. Quantum Electron. 38 (7), 726–735.
Notomi, M., 2011. Strong light confinement with periodicity. Proc. IEEE 99 (10),
1768–1779.
Ott, C., Kaldun, A., Raith, P., Meyer, K., Laux, M., Evers, J., Keitel, C.H., Greene, C.H.,
Pfeifer, T., 2013. Lorentz meets Fano in spectral line shapes: a universal phase and its laser
control. Science 340 (6133), 716–720.
Qiang, Z., Yang, H., Chen, L., Pang, H., Ma, Z., Zhou, W., 2008. Fano filters based
on transferred silicon nanomembranes on plastic substrates. Appl. Phys. Lett. 93,
061106.
12 Weidong Zhou and Shanhui Fan
Rahmani, M., Luk’yanchuk, B., Hong, M., 2013. Fano resonance in novel plasmonic
nanostructures. Laser Photonics Rev. 7 (3), 329–349.
Rosenberg, A., Carter, M., Casey, J., Kim, M., Holm, R., Henry, R., Eddy, C.,
Shamamian, V., Bussmann, K., Shi, S., Prather, D.W., 2005. Guided resonances in
asymmetrical GaN photonic crystal slabs observed in the visible spectrum. Opt. Express
13 (17), 6564–6571.
Seassal, C.D.Y., Letartre, X., Grillet, C., Rojo-Romeo, P., Viktorovitch, P., Benyattou, T.,
2002. Optical coupling between a two-dimensional photonic crystal-based microcavity
and single-line defect waveguide on InP membranes. IEEE J. Quantum Electron. 38 (7),
811–815.
Shuai, Y., Zhao, D., Tian, Z., Seo, J., Plant, D.V., Ma, Z., Fan, S., Zhou, W., 2013. Double-
layer Fano resonance photonic crystal filters. Opt. Express 21 (21), 24582–24589.
Suh, W., Fan, S., 2004. All-pass transmission or flattop reflection filters using a single photonic
crystal slab. Appl. Phys. Lett. 84, 4905.
Suh, W., Wang, Z., Fan, S., 2004. Temporal coupled-mode theory and the presence of
non-orthogonal modes in lossless multimode cavities. IEEE J. Quantum Electron.
40 (10), 1511–1518.
Taflove, A., Hagness, S.C., 2000. Computational Electrodynamics: The Finite-Difference
Time-Domain Method. Artech House.
Taillaert, D., Bogaerts, W., Bienstman, P., Krauss, T., Van Daele, P., Moerman, I.,
Verstuyft, S., De Mesel, K., Baets, R., 2002. An out-of-plane grating coupler for efficient
butt-coupling between compact planar waveguides and single-mode fibers. IEEE J.
Quantum Electron. 38 (7), 949–955.
Tikhodeev, S.G., Yablonskii, A.L., Muljarov, E.A., Gippius, N.A., Ishihara, T., 2002.
Quasiguided modes and optical properties of photonic crystal slabs. Phys. Rev. B
66 (4), 45102.
Wang, K.X., Yu, Z., Sandhu, S., Fan, S., 2013. Fundamental bounds on decay rates in
asymmetric single-mode optical resonators. Opt. Lett. 38 (2), 100–102.
Wood, R., 1902. XLII. On a remarkable case of uneven distribution of light in a diffraction
grating spectrum. Lond. Edinb. Phil. Mag. 4 (21), 396–402.
Zhou, W., 2011. Photonic crystal surface-emitting lasers and infrared sensors. In: Nalwa, H.S.
(Ed.), Encyclopedia of Nanoscience and Nanotechnology. vol. 20. American Scientific
Publishers.
Zhou, W., Ma, Z., Yang, H., Qiang, Z., Qin, G., Pang, H., Chen, L., Yang, W.,
Chuwongin, S., Zhao, D., 2009. Flexible photonic-crystal Fano filters based on
transferred semiconductor nanomembranes. J. Phys. D Appl. Phys. 42, 234007.
Zhou, W., Zhao, D., Shuai, Y.-C., Yang, H., Chuwongin, S., Chadha, A., Seo, J.-H.,
Wang, K.X., Liu, V., Ma, Z., 2014. Progress in 2D photonic crystal Fano resonance
photonics. Prog. Quantum Electron. 38 (1), 1–74.
CHAPTER TWO
Contents
1. Introduction 13
2. Photonic crystals vs metamaterials 15
3. Dielectric rods arranged in a square lattice 19
4. Experimental observation of the photonic phase transitions 24
5. Phase diagram of periodic dielectric photonic structures 27
6. Dielectric metamaterials with near-zero effective parameters 33
7. Toward practical realization of metamaterials in the visible frequency range 35
8. Conclusion and outlook 41
Acknowledgments 42
References 42
1. Introduction
In the late 1980s, Yablonovich (1987) and John (1987) introduced a
new class of artificial dielectric media—photonic crystals—that are periodic
structures having a spatial modulation of the dielectric index with a period
of half a wavelength and larger. That time, advanced nanotechnologies
became mature enough to support the fabrication of such structures with
features of several 100 nm in size. The idea of that concept was to apply
the solid-state theory of electrons to the case of electromagnetic waves
and to develop a class of “semiconductors for light” materials, which would
allow many novel applications. These photonic crystals operate due to spa-
tially extended Bragg resonances that appear as a result of the structure
periodicity.
A B
cones do not mix each other, and only the main light cone at the origin k ¼ 0
is enough for a complete description of the electrodynamic problem. How-
ever, a nonzero dielectric contrast perturbs the states between different light
cones at the boundary of the Brillouin zone that results in opening of
bandgaps related to the certain direction of light. These bandgaps arise
due to Bragg scattering on the periodic structure. Fig. 1A sketches a pair
of unperturbed light cones with vertices at the Γ and Γ 0 points that cross
at the boundary of the Brillouin zone at the X point defining the lowest
intersection point for a square lattice. We consider the unique region of
wave vectors on the ΓX segment. The branch corresponding to the light
cone starts from the Γ point at a certain angle defining the effective refractive
index and ends at the X point. Near the X point the dispersion dependence
ω(k) of the lowest branch demonstrates a deviation from a linear function.
At high frequencies the second branch appears going from the X point to
the Γ point, and the Bragg bandgap is observed between the high- and
low-frequency branches.
Now we consider the band diagram of metamaterials, whose meta-atoms
support a local resonance at some frequency. Owing to the uncertainty
relation, the wave vector of the resonant state is delocalized in the reciprocal
space. At the same time, being localized in the real space, the resonance can
be considered as a quasi-discrete state with a certain frequency. Fig. 1B
exhibits such states by the horizontal orange dotted line, which means that
the frequency is well determined and the wave vector is uncertain. Beside
this, Fig. 1B demonstrates light cones corresponding to propagating waves
(gray dotted lines). A polariton-like feature appears in the band diagram due
to a mixing of the propagating waves with a localized state (Kaina et al.,
2015; Yves et al., 2017). The branch starting at the Γ point at ω ¼ 0 dem-
onstrates a linear dependence at low frequencies and a deviation from the
line as it approaches to the eigenfrequency of the localized mode and trans-
forms into a horizontal line merging with an unperturbed local state. With a
further increase in frequency, a flat branch emerges from the Γ point, which
merges in turn with the continuation of the light cone. As a result a bandgap
arises between the high- and low-frequency branches, which is associated
with a local resonance supported by the meta-atom. We notice the differ-
ence: the spectral position of the Bragg gap is related to the shape of the
Brillouin zone and therefore varies depending on the propagation direction
in the reciprocal space, whereas the position of the band gap associated with
a local resonance does not have dependence on k. It means that in a
metamaterial the resonant frequencies correspond to the weak spatial
18 Mikhail V. Rybin et al.
Hankel function of order n, and r and θ are the polar coordinates. The har-
monic amplitudes are determined by the resonant cylindrical Lorenz–Mie
20 Mikhail V. Rybin et al.
r /a-scenario -scenario
a′ (a) (b) a
(c)
a
a e=4 b e=9
0.8 TE11
TE01
0.6
TE01
0.4
Bragg Bragg
0.2
0.0
c e = 12.8 d e = 19.5
0.8 TE02
TE11
Frequency a /λ
TE21
0.6
Mie + Bragg
TE11
0.4 Mie + Bragg
TE01
TE01
0.2
0.0
e e = 25 f e = 40
0.8
TE02 TE12
TE31
0.6 TE21 TE02
TE21
0.4 TE11
TE11
TE01 Mie
0.2 TE01 Mie
0.0
0 1 2 3 4 0 1 0 1 2 3 4 X 0 1
Field intensity Wavevector Transmission Field intensity Wavevector Transmission
Fig. 3 (Left-hand sections in all panels) Calculated Mie scattering efficiency for an iso-
lated dielectric circular rod for TEn (n 0) modes. (Central sections) The band structure
for 2D square lattice of rods with r ¼ 0.25a in air for the TE polarization. The band struc-
ture is shown between the Γ point (wave vector k ¼ 0) and the X point (j k j ¼ π/a along
the x direction). (Right-hand panels) The transmittance calculated for 10 lattice layers of
the 2D square structure of rods in air for the TE polarization. The frequency and wave
vector are plotted in dimensionless units a/λ, where λ is vacuum wavelength and a
denotes the lattice constant. (A) ε ¼ 4, (B) ε ¼ 9, (C) ε ¼ 12.8, (D) ε ¼ 19.5, (E) ε ¼ 25,
and (F) ε ¼ 40. Reproduced from Rybin, M.V., Filonov, D.S., Samusev, K.B., Belov, P.A.,
Kivshar, Y.S., Limonov, M.F., 2015. Phase diagram for the transition from photonic crystals
to dielectric metamaterials. Nat. Commun. 6, 10102 under a Creative Commons License
(http://creativecommons.org/licenses/by/4.0/).
the Mie resonances are at the higher frequencies and the fundamental gap is
related to the Bragg resonance (compare Fig. 3A with Fig. 1A). With the
increase of ε the Bragg gap is affected by the Mie resonance TE01.
Fig. 3B demonstrates that the fundamental gap shifts down slightly, and both
the center and the low-frequency wing of Mie peak do not correspond to
any band gap. At ε ¼ 12.8 the fundamental gap degenerates (Fig. 3C) and
opens again until ε reaches 19.5 when the second band degenerates into a
horizontal line (Fig. 3D). For greater ε > 19.5, the band structure transforms
from the photonic crystal type to the metamaterial type with a pronounced
polariton-like feature (Figs. 1 and 3E and F). We notice that the bandgap in
the polariton-like feature matches the Mie resonance of isolated rod, which
means that it is a Mie gap. Thus, the square lattice of the dielectric rods trans-
forms from the photonic crystal phase to the metamaterial phase at critical
values ε ¼ 19.5 and r ¼ 0.25a.
The picture of metamaterial phase appearance in the photonic structure
becomes clear with an analysis of bandgap map. We plot in Fig. 4 the depen-
dence of the frequency gap for the ΓX direction as a function of the rod per-
mittivity. Also, we add the spectral frequencies of low-index Mie resonances
0.6 TE12
TE21 TE02
TE31
TE01
0.4
TE11
a/λ
TE0k
TE1k
Mie
0.2 TE2k
TE3k
Bragg
Photonic
crystal Metamaterial
(a) TE (b) TM (c)
0.0
10 20 30 40 50 60 G XG X
e Wave vector
Fig. 4 Bandgap diagram for the 2D square lattice of circular rods. (A) The bandgap dia-
gram obtained from the photonic band structures calculated for 1 ε 60 with steps of
Δε ¼ 1. The rods are embedded in air, r/a ¼ 0.25, and TE polarization is considered. The
Bragg and Mie gaps obtained from the band structure calculations are marked by dark
green. (B and C) The photonic band structure for the 2D square lattice of rods with
ε ¼ 60 for TE and TM polarization, respectively, for the ΓX scan of the wave vector k.
Reproduced from Rybin, M.V., Filonov, D.S., Samusev, K.B., Belov, P.A., Kivshar, Y.S.,
Limonov, M.F., 2015. Phase diagram for the transition from photonic crystals to dielectric
metamaterials. Nat. Commun. 6, 10102 under a Creative Commons License (http://
creativecommons.org/licenses/by/4.0/).
Transition from photonic crystals to metamaterials 23
TEnm to the map. The Bragg gap appears at a/λ ¼ 0.5 when the rod permit-
tivity is low. With the increase of the permittivity the TEnm Mie resonances
shift down to a/λ ¼ 0.5, and the Bragg gap becomes broad similar to the case
of the resonant photonic crystals (Ivchenko, 2005). The interference
between Mie and broaden Bragg resonances results in a nontrivial picture
of alternating allowed bands and band gaps. However at ε ¼ 19.5 the
TE01 Mie gap splits off from the broaden Bragg gap and for the higher per-
mittivities the fundamental (lowest) band gap is related to the Mie resonance
on the structure elements only.
Recently, Maslova et al. (2018) found conditions for metamaterials with
electric response operating for the TM polarization. The TM polarization
is known to demonstrate stronger effects with respect to the case of TE
polarization. For this reason the effect of the resonant Bragg gap broadening
is stronger, which disables metamaterial regime for the structure with and
electric metamaterials with r ¼ 0.25a (see Fig. 4C). However, sparse struc-
tures make is possible to observe metamaterials with electric response
(Fig. 5). When the rod permittivity is ε ¼ 30 that corresponds to Ge2Sb2Te5
0.7
0.6
0.5
0.4
a/l
0.3
0.2
0.1
(a) (b) (c)
0.0
X M X M –4 0 4
Wave vector e eff
Fig. 5 Band diagrams of dielectric rods arranged in a square lattice for TM polarization.
(A) Photonic crystal with ε ¼ 12 and r/a ¼ 0.1. (B) Metamaterial with ε ¼ 30 and r/a ¼ 0.1.
The second branches in the diagrams are marked by thick lines. Green circles correspond
to the bottom boundary of the second band, the red dotted lines mark frequencies of the
maximum field enhancement at a/λ ¼ 0.518 and a/λ ¼ 0.472 in a structure of 5 by 250
rods. (C) Evaluated effective permittivity εeff of the metamaterial with ε ¼ 30 and
r/a ¼ 0.1. Gray shading in (B and C) corresponds to the Mie gap. Reproduced with permis-
sion from Maslova, E.E., Limonov, M.F., Rybin, M.V., 2018. Dielectric metamaterials with
electric response. Opt. Lett. 43, 5516–5519. Copyright (2018) OSA.
24 Mikhail V. Rybin et al.
A B
y 2r
a
TE z a
H
y
E
x
x
z
C
Reciever
E
H k
Y
Transmitter Z
X
input port of the analyzer. The spectra were normalized to the transmission
of the free space.
Fig. 7 demonstrates the experimental and theoretical spectra of the
metacrystal as a function of the sample volume related to the r/a parameter
as well as its temperature, which determines the rod permittivity. During the
measurements for the r/a-scenario the temperature of water was stabilized
at 90°C while the lattice spacing of the square lattice was reduced from
a ¼ 212 mm (r/a ¼ 0.08) to 89 mm (r/a ¼ 0.19). The transmission dip
corresponding to the lowest Bragg gap shifts to the higher frequencies
and almost disappears at r/a 0.09 when it reaches to the frequency of
the lowest Mie dipole resonance TE01. For the shorter lattice spacing the
Bragg dip appears again at r/a 0.095, however at the higher frequency side
of the Mie resonance. As a result, the lowest Bragg band and the TE01 Mie
Fig. 7 Transmission spectra for a square lattice of circular rods. TE polarization. (A–C)
r/a-scenario, ε ¼ 62. The calculated (B) and experimentally measured (C) transmission
spectra of the metacrystal as a function of the filling ratio r/a. The calculated spectra
are shown in the range 0.05 r/a 0.21, the experimental spectra are shown in the
range 0.08 r/a 0.19. (D–F) ε-scenario, r/a ¼ 0.08. The calculated (E) and experimen-
tally measured (F) transmission spectra of the metacrystal as a function of the dielectric
permittivity. The calculated spectra are shown in the range 30 ε 94, the experimen-
tal spectra are shown in the range 62 ε 94. The spectra are shifted vertically by the
constant value. Reproduced from Rybin, M.V., Filonov, D.S., Samusev, K.B., Belov, P.A.,
Kivshar, Y.S., Limonov, M.F., 2015. Phase diagram for the transition from photonic crystals
to dielectric metamaterials. Nat. Commun. 6, 10102 under a Creative Commons License
(http://creativecommons.org/licenses/by/4.0/).
Transition from photonic crystals to metamaterials 27
band change their positions in the frequency scale indicating the transition
into the metamaterial phase.
The ε-scenario for the metacrystal with r/a ¼ 0.08 was realized by
varying of the temperature of water leading to modifying the rod permittiv-
ity. Increasing of the dielectric permittivity results in the spectral shift of the
Mie resonances, and the Bragg dip almost keeps its position. Fig. 7F shows a
significant proximity of the Mie and Bragg resonances, although their inter-
section was not achieved under the experimental conditions. In theory the
Mie resonance becomes the fundamental gap and the metacrystal turns into
the metamaterial regime at r/a ¼ 0.08 and the rod permittivity about
90 (Fig. 7E).
Let us discuss this transition of the metacrystal sample. We remind that
the fundamental Mie gap is associated with the negative effective magnetic
permeability μeff < 0 (e.g., see O’Brien and Pendry, 2002) and strong special
dispersion above the lowest Bragg gap disables μeff for the photonic crystal
regime. Thus, the continuous change in temperature or sample volume leads
to the appearance of the effective magnetic permeability deviating from
unity at a certain frequency interval. At the same time thermodynamic phase
transitions are known to be associated with dramatic changes in a generalized
susceptibility (e.g., Landau et al., 1980). Hence, the observed transition in
metacrystal (appearance of the nonzero magnetic susceptibility) owning
to the change in temperature or volume can be treated as a phase transition.
However, the changes of the thermodynamic quantities (temperature and
volume) in such phase transitions manifesting in the photonic response
are just means rather than physical causes. The physical reason for the pho-
tonic transition is the change of the water permittivity or the lattice spacing.
Therefore, the observed transitions are not thermodynamic phase transitions
but they are transitions of a new kind that should be referred to as photonic
phase transitions.
for a variety of values of r/a reveal the boundary of the metamaterial phase that
is shown in Fig. 8 presenting the phase diagram as a function of rod permit-
tivity and the radius normalized by the square lattice constant for the case of
TE polarization. The phase diagram unveils the structure parameters required
for the metamaterial regime. The left side of the phase boundary corresponds
to the r/a- and ε-scenarios described earlier in Fig. 2. However, there exists a
competing process resulting in the right side of the boundary. This process is
related to the interaction between the neighboring rods and collective modes
driving the strong spatial dispersion typical to the photonic crystals.
We also construct the phase diagram for the metamaterials with electric
response in TM polarization. The metamaterial phase appears as a narrow
area at the left of Fig. 8 corresponding to sparse structures. The reason is
in the stronger electric multipole–multipole interaction relative to their
magnetic counterparts. Also the minimum rod permittivity for the electric
metamaterials with a square lattice is ε ¼ 27. We notice that the
metamaterials with magnetic response are of interest due to their effective
magnetic permeability μeff 6¼ 1, however a variety of applications demand
control of the electric response as well as the low-loss dielectric
metamaterials with engineered effective permittivity εeff can significantly
improve the light-matter interaction.
100
Electric met
Rod permittivity
80
Magnetic metamaterials
amaterials
60
40
20
Photonic crystals
0.0 0.1 0.2 0.3 0.4 0.5
r/a
Fig. 8 TM + TE photonic phase diagrams of dielectric rods arranged in a square lattice.
The region of the electric metamaterials in the TM polarization is marked by blue. The
region of the magnetic metamaterials in the TE polarization is marked by red. Circles are
phase boundary obtained from the band diagram analysis. Solid curves are guides for
eyes only. Reproduced with permission from Maslova, E.E., Limonov, M.F., Rybin, M.V.,
2018. Dielectric metamaterials with electric response. Opt. Lett. 43, 5516–5519. Copyright
(2018) OSA.
Transition from photonic crystals to metamaterials 29
Following the intuition one might think that transitions between the
photonic crystal and metamaterial regimes are gradual. For example, let
us consider the r/a parameter continuously decreasing from a value
corresponding to the metamaterial phase (see Fig. 8) through the boundary
to a photonic crystal value. The common believe is that the “quantity” of
metamaterial is becoming weaker until it disappears completely. However
it is not correct description of the transitions. In reality, the field distribution
(photonic subsystem) undergoes a dramatic transformation when the struc-
ture parameters (ε and r/a) cross the phase boundary. Fig. 9 illustrates the
field distribution under the photonic phase transition. We examine the filed
profiles in two prism-shaped structures consisting of dielectric rods (ε ¼ 25)
arranged in a square lattice. The systems are illuminated by a Gaussian beam
directed along the Γ-X (the first prism) and the Γ-M (the second prism) at
the lowest frequency of the second dispersion band. We are decreasing the
volume of samples by varying the lattice spacing from r/a ¼ 0.18 (photonic
crystal according to the phase diagram in Fig. 8) to r/a ¼ 0.20 (metamaterial).
Fig. 9 A dramatic change in the magnetic field pattern depending on the operational
photonic crystal or metamaterial regimes (or the corresponding phase). Simulated field
patterns for the lowest frequency of the second band vs nanorod radius. An incident
Gaussian beam propagates along the ΓX direction (A)–(D) or along the ΓM direction
(E)–(H). The structure boundary is marked by a gray dashed line; arrows show the inci-
dent direction, nanorod permittivity ε ¼ 25. Simulations are performed within the mul-
tiple scattering theory approach. Reproduced from Li, S.V., Kivshar, Y.S., Rybin, M.V., 2018.
Toward silicon-based metamaterials. ACS Photonics 5, 4751–4757 with permission, ACS.
30 Mikhail V. Rybin et al.
A B
However, event at a/λ ¼ 0.4242 the modes of another type located around
the X points coexist to the dimensionless modes around the Γ point that
indicates the strong special dispersion regime as well. We notice that the
modes around the X points on the Brillouin zone surface appear due to
the Bragg scattering on the periodic structure.
Now we consider the metamaterial with (r/a ¼ 0.190, ε ¼ 25) which is
described as only one isofrequency contour per frequency. Because the struc-
ture parameters are near the phase boundary most frequencies correspond to
the four-pointed star revealing the strong spatial dispersion. However for the
low-frequency interval below the a/λ ¼ 0.41033 the counters become circles
and the structure has the frequency interval with the weak spatial dispersion
indicating the metamaterial phase. Now we can describe the transition
through the boundary. In the isofrequency representation the Bragg reso-
nances manifest themselves as contours around points on the Brillouin zone
surface, whereas the Mie resonances exhibit circles around the Γ point similar
to the light cone (dispersion of free photons in homogeneous media). At some
frequencies the contours of Bragg modes are superimposed with the dis-
persionless Mie contours resulting in more the one mode for certain directions
of the wave vector. This regime is described by the strong spatial dispersion
and we consider it as photonic crystal regime because of the existence of the
Bragg scattering. When the frequency of Bragg resonance increases the con-
tours around the Brillouin zone surface disappears and the Mie contours
remain the exclusive resonant feature in the structure. Below some frequency
the contour becomes circular exhibiting the metamaterial properties of the
Mie resonant-driven dielectric structures.
Also we discuss an important consequence from the different nature of
the resonances defining the fundamental gap in photonic crystals and
metamaterials. In photonic crystals the Bragg scattering requires the period-
icity of the structure and the dielectric contrast is not a crucial factor. In con-
trast, for the metamaterial phase the Mie resonant modes play the decisive role,
while the periodicity is important only for the theoretical description of pho-
tonic properties by exploiting terms of the band theory. Therefore, a position
disorder has to influence on the photonic crystal Bragg bandgaps, unlike the
robust Mie bandgaps of the metamaterial regime. We analyze a transmission of
the TE-polarized plane wave through a photonic structure in different phases.
Fig. 11A shows the light incidence at a central frequency of the Bragg bandgap
(a/λ ¼ 0.3, ε ¼ 4, r/a ¼ 0.25) with the striped pattern along the exponential
decay that makes the sample opaque. The introduction of the position disor-
der breaks the periodicity, and the Bragg scattering disappears making the
32 Mikhail V. Rybin et al.
Fig. 11 Transmission of the plane wave through periodic photonic structure with and
without position disorder. (A and B) Photonic crystal regime ε ¼ 4, a/λ ¼ 0.3. (C and D)
Metamaterial regime ε ¼ 25, a/λ ¼ 0.45. All structures r/a ¼ 0.25, TE polarization. The
plane wave is incident from the bottom side. Transmission spectra of photonic crystal
(E) and metamaterial (F). Spectra of prefect structures are shown by black solid curve.
Spectra of structures with position disorder are shown by red dashed curve.
Bragg gap or as a specific TEnk Mie gaps. Fig. 11F uncovers that after the
photonic phase transition to the metamaterial regime the fundamental
Mie gap (below all Bragg gaps) is not affected by position disorder in contrast
to other gaps Bragg and Mie both. This difference becomes another impor-
tant feature of photonic structure appearing with the phase transition.
A B C D
Fig. 13 Electric field localization in the plates being the photonic crystal with ε ¼ 12 and
r/a ¼ 0.10 or the metamaterial with ε ¼ 30 and r/a ¼ 0.10. (A) Distribution of the electric
field intensity in the photonic crystal plate. (B) The electric field intensity normalized to
the intensity in the free space without account of the field inside the cylinders for the
photonic crystal structure. (C) The electric field intensity normalized to the intensity in
the free space without account of the field inside the cylinders for the metamaterial
regime. (D) Distribution of the electric field intensity in the metamaterial plate. Green
dashed lines show the plate boundaries. Black arrows show the Gaussian beam incident
direction. Reproduced with permission from Maslova, E.E., Limonov, M.F., Rybin, M.V.,
2018. Dielectric metamaterials with electric response. Opt. Lett. 43, 5516–5519. Copyright
(2018) OSA.
the near-zero regime, the optical size of the sample is extremely small rel-
ative to the wavelength. As a result the Gaussian wave excites a mode
extended over the entire structure and the I/I0 ratio increases over
50 (Fig. 13C and D). The peak frequency a/λ ¼ 0.4717 differs by a small
value from the lowest frequency of the second dispersion band (in Fig. 5)
due to the finite size of the sample (it contains only 5 rods along the beam
incidence).
Frequency a/l
1 0.4
Lossless
0.8 0.3
0.6
0.2
0.4
0.2 0.1
0 0
0 0.5 1 0 0.5 1
Wavevector 2pk/a Wavevector 2pk/a
C D
G X G G X G
0.5 0.25
Frequency a/l
Frequency a/l
0.24
With losses
0.48
0.23
0.46
0.22
0.44
0.21
0.42 0.2
0.475 0.5 0.525 0.4 0.5 0.6
Wavevector 2pk/a Wavevector 2pk/a
Fig. 14 Phase criteria for the lossless and lossy structures. (A and B) Band diagrams of a
square lattice of nanorods with r/a ¼ 0.3 and lossless permittivity ε ¼ 4 (the photonic crys-
tal phase) and 30 (the metamaterial phase). Data obtained by the direct method are
shown by black solid curves. Results of the inverse dispersion method are shown by cyan
circles. (C and D) Band diagrams calculated by the inverse dispersion method for the
similar structure as in (A and B) but with the nanorod permittivity having an additional
imaginary part ε00 ¼ 0.01ε0 (cyan solid curves). Black dashed curves show the results for loss-
less structures. Reproduced with permission from Li, S.V., Kivshar, Y.S., Rybin, M.V., 2018.
Toward silicon-based metamaterials. ACS Photonics 5, 4751–4757. Copyright (2018) ACS.
cross each other (Fig. 14C), while for the case of Mie bandgap the modes
exhibit avoided crossing (Fig. 14D).
Besides, we optimize the crystalline lattice taking into account two com-
peting conditions preventing the metamaterial phase discussed in Section 5:
the lowest Bragg frequency has to be higher than the Mie frequency and the
distance between the neighbor rods has to be large enough to preclude from
the collective modes formation. Having the densest packing, a hexagonal
lattice provides the smallest distance between the crystal planes among other
lattices, thus here we consider this type of lattice.
Fig. 15 compares photonic crystal–metamaterial phase diagrams of struc-
tures with the hexagonal and square lattices obtained by both considered
criteria. First we notice a significant decrease of the minimum permittivity
38 Mikhail V. Rybin et al.
required for the metamaterial phase: the value ε ¼ 14 for the hexagonal lat-
tice is more promising for a practical realization than ε ¼ 19.5 for the square
lattice. Besides, we determine another phase boundary (shown by circles in
Fig. 15) by the crossing–anticrossing criterion for photonic sutures
with losses. For this case we take complex permittivity by the formula
ε ¼ ε0 (1 + i0.01) and the axis of ordinates shows ε0 (the real part of ε). The
phase diagrams reveal the excellent agreement of the two criteria. The small
difference for the denser structures (r/a > 0.35) can be explained as differ-
ent numbers of plane waves involved in calculations. The Hermitian eigen-
value problem solved in the ω(k) method allows expansion over 128 128
plane waves, whereas the inverse dispersion method k(ω) solves the non-
Hermitian eigenvalue problem that limits us by resolution of 25 25 plane
waves. Another reason for deviation between the phase boundaries evalu-
ated by two approaches is in the permittivity real-to-imaginary parts ratio
ε00 /ε0 . We find that, when the imaginary part of permittivity grows, an error
in the phase boundary is increasing. In particular, for the imaginary part of
dielectric permittivity described by ε00 /ε0 ¼ 0.35, this deviation is about 5%.
Now we have a complete tool for constructing the phase diagram of
periodic photonic structures, which constituents have a complex
frequency-dependent permittivity. The phase diagram in Fig. 15 shows that
50
Metamaterial
40
Photonic crystal
Photonic crystal
Dielectric permitivity
30
20
10
Hexagonal lattice Square lattice
0
0 0.1 0.2 0.3 0.4 0.5
r /a
Fig. 15 Phase diagram that allows differentiating metamaterials and photonic crystals.
The boundary is calculated for both hexagonal (red line) and square (black dashed line)
lattices, the circles mark the data obtained by the direct and inverse dispersion methods,
respectively. Green and red shadings correspond to the photonic crystal and
metamaterial regimes of a hexagonal lattice, respectively. Reproduced with permission
from Li, S.V., Kivshar, Y.S., Rybin, M.V., 2018. Toward silicon-based metamaterials. ACS Pho-
tonics 5, 4751–4757. Copyright (2018) ACS.
Transition from photonic crystals to metamaterials 39
A B C
900
Wavelength (nm)
500
400
0.1 0.2 0.3 0.4 0.5 50 100 150 200 250 300 350 20 30 40 50 60 70 80 90 100
r/a Lattice constant (nm) Radius (nm)
Fig. 16 Phase diagram of a periodic silicon nanorod structure with a hexagonal lattice.
Red circles show the phase boundary, solid line is a guide for eyes only. Green and red
shadings correspond to the photonic crystal and metamaterial regimes, respectively.
The phase boundary for the square lattice is shown by a black dashed line. Phase dia-
gram is plotted in the difference axes: (A) ratio r/a vs wavelength; (B) lattice spacing a vs
wavelength; and (C) rod radius r vs wavelength. Reproduced with permission from Li, S.V.,
Kivshar, Y.S., Rybin, M.V., 2018. Toward silicon-based metamaterials. ACS Photonics 5,
4751–4757. Copyright (2018) ACS.
40 Mikhail V. Rybin et al.
A C
E
44° 63°
27°
B D
F
Fig. 17 Refraction of a Gaussian beam on a silicon-based metamaterial prism, for
a ¼ 167 nm and r ¼ 50 nm. The top panels are the magnetic field profiles (red are positive
values and blue are negative values of Hz); the bottom panels are the electromagnetic
field intensities. (A and B) Off-resonance case; λ ¼ 1500 nm. (C and D) the low-frequency
edge λ ¼ 700 nm; (E and F) the high-frequency edge λ ¼ 495 nm. Green dashed triangles
mark the prism. Black arrows in the top panels show the propagation direction of the
incident beam. The results are for the TE polarization. Reproduced with permission from
Li, S.V., Kivshar, Y.S., Rybin, M.V., 2018. Toward silicon-based metamaterials. ACS Photonics
5, 4751–4757. Copyright (2018) ACS.
with absorption, the first and second branches are merged into continuous
branches and the metamaterial phase can be distinguished by the criterion
based on crossing–anticrossing of the branches in complex bandgap dia-
grams. Although the considered photonic phase transitions are quite dissim-
ilar to conventional thermodynamic phase transitions, the phase transition
terminology is feasible because it describes abrupt changes in the photonic
patterns observed at the boundary between photonic crystal and
metamaterial regimes. Besides, more important is that the developed phase
diagrams as a function of structure parameters (permittivity and sizes)
become very useful in a design of metamaterials with desired properties.
Here we have reviewed several aspects of the metamaterial regimes includ-
ing near-zero parameters, stability against position disorder, and practical
realization of silicon-based metamaterials in the visible frequency range.
The study of photonic phase transitions has been started in the recent years,
so we anticipate the further development of this concept.
Acknowledgments
This work was supported by the Ministry of Education and Science of the Russian Federation
(Grant 3.1500.2017/4.6) and the Strategic Fund of the Australian National University.
References
Andryieuski, A., Kuznetsova, S.M., Zhukovsky, S.V., Kivshar, Y.S., Lavrinenko, A.V.,
2015. Water: promising opportunities for tunable all-dielectric electromagnetic
metamaterials. Sci. Rep. 5, 13535.
Bohren, C.F., Huffman, D.R., 1998. Absorption and Scattering of Light by Small Particles.
Wiley-VCH.
Decoopman, T., Tayeb, G., Enoch, S., Maystre, D., Gralak, B., 2006. Photonic crystal lens:
from negative refraction and negative index to negative permittivity and permeability.
Phys. Rev. Lett. 97, 073905.
Gorlach, M.A., Belov, P.A., 2014. Effect of spatial dispersion on the topological transition in
metamaterials. Phys. Rev. B 90 115136.
Green, M., Keevers, M., 1995. Optical properties of intrinsic silicon at 300 K. Prog. Photo-
volt. 3, 189–192.
Huang, X., Lai, Y., Hang, Z.H., Zheng, H., Chan, C.T., 2011. Dirac cones induced by acci-
dental degeneracy in photonic crystals and zero-refractive-index materials. Nat. Mater.
10, 582.
Ivchenko, E.L., 2005. Optical Spectroscopy of Semiconductor Nanostructures. Alpha
Science, Harrow, UK.
Joannopoulos, J.D., Johnson, S.G., Winn, J.N., Meade, R.D., 2008. Photonic Crystals:
Molding the Flow of Light, second ed. Princeton University Press.
John, S., 1987. Strong localization of photons in certain disordered dielectric superlattices.
Phys. Rev. Lett. 58, 2486–2489.
Johnson, S.G., Joannopoulos, J.D., 2001. Block-iterative frequency domain methods for
Maxwell’s equations in a plane-wave basis. Opt. Express 8, 173.
Transition from photonic crystals to metamaterials 43
Kaina, N., Lemoult, F., Fink, M., Lerosey, G., 2015. Negative refractive index and acoustic
superlens from multiple scattering in single negative metamaterials. Nature 525, 77.
Kuznetsov, A.I., Miroshnichenko, A.E., Brongersma, M.L., Kivshar, Y.S., Luk’yanchuk, B.,
2016. Optically resonant dielectric nanostructures. Science 354, aag2472.
Landau, L.D., Lifshitz, E.M., Pitaevskii, L.P., 1980. Statistical Physics, Part I. Pergamon
Press, Oxford.
Liberal, I., Engheta, N., 2017. Near-zero refractive index photonics. Nat. Photonics 11, 149.
Liu, Z., Lee, H., Xiong, Y., Sun, C., Zhang, X., 2007. Far-field optical hyperlens magnifying
sub-diffraction limited objects. Science 315, 1686.
Maslova, E.E., Limonov, M.F., Rybin, M.V., 2018. Dielectric metamaterials with electric
response. Opt. Lett. 43, 5516–5519.
Mirmoosa, M.S., Kosulnikov, S.Y., Simovski, C.R., 2015. Unbounded spatial spectrum of
propagating waves in a polaritonic wire medium. Phys. Rev. B 92, 075139.
O’Brien, S., Pendry, J.B., 2002. Photonic bandgap effects and magnetic activity in dielectric
composites. J. Phys. Condens. Matter 14, 4035.
Pendry, J.B., Holden, A.J., Robbins, D.J., Stewart, W.J., 1999. Magnetism from conductors
and enhanced nonlinear phenomena. IEEE Trans. Microwave Theory Technol.
47, 2075.
Pendry, J.B., Aubry, A., Smith, D.R., Maier, S.A., 2012. Transformation optics and sub-
wavelength control of light. Science 337, 549.
Reyes-Gómez, E., Cavalcanti, S.B., Oliveira, L.E., De Carvalho, C.A.A., 2014. Metric-
signature topological transitions in dispersive metamaterials. Phys. Rev. E 89, 033202.
Rybin, M.V., Limonov, M.F., 2016. Inverse dispersion method for calculation of complex
photonic band diagram and PT symmetry. Phys. Rev. B 93, 165132.
Rybin, M.V., Filonov, D.S., Samusev, K.B., Belov, P.A., Kivshar, Y.S., Limonov, M.F.,
2015. Phase diagram for the transition from photonic crystals to dielectric metamaterials.
Nat. Commun. 6, 10102.
Shelby, R.A., Smith, D.R., Schultz, S., 2001. Experimental verification of a negative index
of refraction. Science 292, 77–79.
Simovski, C.R., 2009. Material parameters of metamaterials. Opt. Spectrosc. 107, 726.
Simovski, C.R., 2018. Composite Media with Weak Spatial Dispersion. Pan Stanford.
Smith, D.R., Padilla, W.J., Vier, D.C., Nemat-Nasser, S.C., Schultz, S., 2000. Composite
medium with simultaneously negative permeability and permittivity. Phys. Rev. Lett.
84, 4184.
Staude, I., Schilling, J., 2017. Metamaterial-inspired silicon nanophotonics. Nat. Photonics
11, 274–284.
Veselago, V.G., 1968. The electrodynamics of substances with simultaneously negative
values of ε and μ. Sov. Phys. Usp. 10, 509.
Wuttig, M., Bhaskaran, H., Taubner, T., 2017. Phase-change materials for non-volatile pho-
tonic applications. Nat. Photonics 11, 465.
Xu, L., Chen, H., 2015. Conformal transformation optics. Nat. Photonics 9, 15.
Yablonovich, E., 1987. Inhibited spontaneous emission in solid-state physics and electronics.
Phys. Rev. Lett. 58, 2059–2062.
Yves, S., Fleury, R., Berthelot, T., Fink, M., Lemoult, F., Lerosey, G., 2017. Crystalline
metamaterials for topological properties at subwavelength scales. Nat. Commun. 8,
16023.
Zhang, B., Jie, J., Zhang, X., Ou, X., Zhang, X., 2017. Large-scale fabrication of silicon
nanowires for solar energy applications. ACS Appl. Mater. Interfaces 9, 34527–34543.
CHAPTER THREE
Contents
1. Introduction 45
2. Ray optics light trapping theory 47
3. Wave optics light trapping theory 50
4. Fundamental limit of light trapping 54
4.1 Light trapping in structures with wavelength-scale periodicity 55
4.2 Light trapping in thin films 57
5. Light trapping in photonic crystals 61
5.1 Light trapping in 2D photonic crystals 62
5.2 Light trapping in 3D photonic crystals 73
6. Absorption enhancement in ultrathin crystalline silicon solar cells with
antireflection and light-trapping nanocone gratings 77
6.1 Motivation 77
6.2 Absorption enhancement at Yablonovitch limit 79
7. Conclusion 86
References 86
1. Introduction
Photon management strategies, including light trapping, have been
tremendously successful in efficiency improvement and cost reduction in
practical solar cells (Mokkapati and Catchpole, 2012). Light trapping allows
cells to absorb sunlight using an active material layer that is much thinner
than the material’s intrinsic absorption length. This effect then reduces
the amount of materials used in photovoltaic cells, which cuts cell cost in
general, and moreover facilitates mass production of photovoltaic cells that
are based on less abundant materials. In addition, light trapping can improve
Aðω, θ, φÞ
f ðω, θ, φÞ ¼ : (2)
αdeff
Here, ω, θ, and φ are the frequency, the angle of incident light, and the
azimuthal angle of light, respectively. A is the absorption coefficient of the
Fig. 2 A dielectric slab structure with a randomly textured front surface and a back mir-
ror. The equivalent thickness is the same as in Fig. 1.
where f(ω, θ, φ) is defined in Eq. (3). For both the isotropic and anisotropic
cases, that is, f(ω, θ, φ) ¼ 4n2 and f ðω, θ, φÞ ¼ 4n2 = sin 2 θ0 θθ0 , respectively,
we have F3D ¼ 4πn2. Moreover, it has been shown that as an upper limit
F3D ¼ 4πn2 in fact applies to any light trapping structure of the form of
Fig. 2 with an arbitrary angular response of f(θ, φ) (Yu and Fan, 2011).
Since no actual physical light trapping structure has an ideal isotropic
angular response, to compare the performance of a physical light trapping
structure to the conventional limit, it is imperative that one performs angle
integration (Naqavi et al., 2014; Yu and Fan, 2011). Observing an enhance-
ment factor greater than 4n2 for a single angle of incidence, for example,
should not be taken as the evidence that one has overcome the conventional
limit (Ganapati et al., 2014). In this context, the 4πn2 limit, which is really
the conventional limit expressed in an angle-integrated fashion, is very useful
when one needs to compare the performance of a physical structure to the
ray optics limit.
The ray optics theory also describes the conventional light trapping
structure of Fig. 2 in 2D. For the isotropic case the enhancement factor is
et al., 2013a; Pala et al., 2013; Peretti et al., 2012; Sheng et al., 2011; Vynck
et al., 2012; Wang et al., 2012, 2013a, 2014a,b; Yu et al., 2012; Zeman et al.,
2013). Conventional light trapping theory, largely developed in the context
of c-Si solar cells in the 1980s, was initially developed using the tools of
ray optics. For solar cells with thicknesses or feature sizes that are comparable
to the wavelengths of sunlight, ray tracing is no longer applicable. It is there-
fore essential to develop a light trapping theory entirely from a wave optics
perspective.
The development of wave optics light trapping theory has a long history.
Yablonovitch, in one of his initial papers that elucidate the 4n2 limit
(Yablonovitch, 1982; Yablonovitch and Cody, 1982), used a blackbody
radiation argument that is directly related to the modal structure of the cell.
Sheng, Bloch, and Stepleman argued that the DOS plays an important role
and hinted at the possibility of modifying the ray optics limit in thin film
structures (Sheng et al., 1983). Stuart and Hall examined light trapping in
wavelength-scale thin films and concluded that such a wavelength-scale thin
film in fact has an absorption enhancement limit comparable to and some-
what below that of the bulk cell (Stuart and Hall, 1997). All these studies
have pointed out the importance of modal structures of the cells in the
understanding of the behaviors of light trapping.
As an important step toward a wave optics light trapping theory, Yu,
Raman, and Fan recently introduced a statistical temporal coupled mode
theory (TCMT) formalism to account for light trapping in the wave optics
domain (Yu et al., 2010a,b). This theory has completely reproduced all the
key results of conventional ray optics light trapping theory, including the
Yablonovitch 4n2 limit (Yu et al., 2010a), the correction to the 4n2 limit
when the single-pass absorption is nonnegligible (Yu et al., 2012), and
the correction to the 4n2 limit for an absorber with angular selective absorp-
tion (Yu and Fan, 2011; Yu et al., 2010a). Moreover, this wave optics theory
predicts that one can significantly overcome the conventional light trapping
limit in certain nanophotonic structures, a prediction that has been validated
with full-field simulations showing that the conventional limit can be over-
come for all incident angles and over a wide wavelength ranges (Wang et al.,
2014a; Yu et al., 2010a). In this section, we review this wave optics light
trapping theory.
A solar cell is modeled as a structure supporting a number of optical
modes that may couple to a number of free-space channels and thus become
resonances. Light trapping is described by the coupling of incident radiation
into the optical modes supported by the solar cell structure. Absorption is
52 Ken Xingze Wang et al.
e
γ m γ m, n
A m , n ð ωÞ ¼ , (9)
ðω ωm Þ + ðe
2
γ m + γ m Þ2 =4
where ω is the frequency, ωm is the resonant frequency of the mth mode, eγ m
P
is the intrinsic material absorption rate of the mth mode, γ m ¼ N n¼1 γ m, n is
the total external coupling rate of the mth mode, γ m, n is the decay rate of the
mth mode into the nth channel (due to reciprocity, it is also the in-coupling
rate of the mth mode from the nth channel), and
πω2 L 2
N ¼2 2 (10)
c 2π
is the total number of channels in real space with a period L. For incident
sunlight from the nth channel, the overall absorption coefficient, calculated
using the concept of spectral absorption cross section (Yu et al., 2010a), is
Z ω + Δω X M
1
An ðωÞ ¼ Am, n ðωÞω
Δω ω m¼1
M Z ω + Δω
1 X
¼ Am, n ðωÞω (11)
Δω m¼1 ω
M Z ∞
1 X
Am, n ðωÞω,
Δω m¼1 ∞
where
4πn3 Δðω3 Þ L 2 d
M ¼2 (12)
3c 3 2π 2π
is the total number of resonances in the spectral range (ω, ω + Δω) supported
by the structure with a thickness d and a refractive index n. The approxima-
tion on the limits of integration in the last step of Eq. (11) is justified since Δω
is chosen to be much greater than the linewidths of the resonances, which
have high qualify factors, while still being much smaller than ω. It remains to
justify the summation in Eq. (11), which holds only if the M modes are
orthogonal.
From Eq. (7), we can calculate the broadband absorption enhancement
by the mth resonance when light is incident from the nth channel. By sum-
ming over the contributions from all M resonances and N channels in the
frequency range of [ω, ω + Δω], and by comparing the absorption to the
single-pass absorption αdeff, we can calculate the angle-integrated absorption
54 Ken Xingze Wang et al.
where ρ(ω) ¼ M/Δω is the spectral DOS, and fm is the contribution to the
enhancement factor from the mth resonance. η is the volume or area fraction
P
of the absorbing material. The equality is approached if N n¼1 γ m, n ≫γ 0 for
all m, that is, all the resonances are in the over-coupling regime. Since
the material is weakly absorptive, the equality holds as long as all resonances
are coupled to external radiation. In other words, all photonic states in
the structure contribute to light trapping, and the averaging over the band-
width results in the appearance of DOS in the enhancement factor F. In
next section, we use Eq. (13) to show that the limit of 4n2 is only correct
in bulk structures and the absorption enhancement factor can go far beyond
this limit with proper design in the nanophotonic regime.
in the medium surrounding the cell. The theory can be generalized to the
case of a restricted emission cone and reproduces the standard result of
4n2/sin2θ (Yu et al., 2010a).
The analysis here also points to scenarios where the conventional limit is
no longer applicable. Eq. (10) is not applicable when the periodicity is com-
parable to the wavelength, whereas Eq. (12) is not valid when the film thick-
ness is much smaller than the wavelength. Below, we consider both of
these cases.
A ky B
12n2
8n2
k0 F
kx
4n2
2p/L 0
0 1 2 3 4 5 6 7 8
L/l
Fig. 3 Light trapping in periodic structures. (A) Blue dots represent channels in the k
space. Channels in the circle correspond to free-space propagating modes.
(B) Theoretical upper limit of the absorption enhancement factor using a light-trapping
scheme where a square-lattice periodic grating structure is introduced into a thin film.
Red area represents a spectral range where the upper limit of the absorption enhance-
ment factor Fiso is above 4n2.
56 Ken Xingze Wang et al.
Scattering
Cladding
Active
Mirror
Fig. 4 Structure for overcoming the conventional light-trapping limit. (A) A nano-
photonic light-trapping structure. The scattering layer consists of a square lattice of
air groove patterns with periodicity L ¼ 1200 nm. The thicknesses of the scattering,
cladding, and active layers is 80, 60, and 5 nm, respectively. The mirror layer is a perfect
electric conductor. (B) The profile of electric-field intensity for the fundamental wave-
guide mode. Fields are strongly confined in the active layer. To obtain the waveguide
mode profile, the scattering layer is modeled by a uniform slab with an averaged dielec-
tric constant.
absorbers in organic solar cells given their short exciton diffusion lengths of
about 3–10 nm (Huynh et al., 2002; Mayer et al., 2007; Yu et al., 1995).
In order to enhance the absorption in the active layer, we place a trans-
pffiffiffiffiffiffiffiffiffi
parent cladding layer nH ¼ 12:5 on top of the active layer. Such a cladding
layer serves two purposes. First, it enhances density of state. The overall
structure supports a fundamental mode with group index nwg close to nH,
which is much higher than that of the absorbing material. Second, the index
contrast between active and cladding layer provides nanoscale field confine-
ment. Fig. 4B shows the fundamental waveguide mode. The field is highly
concentrated in the low-index active layer, due to the well-known slot-
waveguide effect (Almeida et al., 2004). Thus, the geometry here allows
the creation of a broadband high-index guided mode, with its energy highly
concentrated in the active layer, satisfying the requirement in Eq. (16) for
high absorption enhancement.
In order to couple incident light into such nanoscale guided modes, we
introduce a scattering layer with a periodic pattern on top of the cladding
layer, with a periodicity L much larger than our wavelength range of inter-
est. Each unit cell consists of a number of air grooves. These grooves are
oriented along different directions to ensure that scattering strength does
not strongly depend on the angles and polarizations of the incident light.
We emphasize that there is no stringent requirement on these grooves as
long as the scattering strength dominates over resonance absorption rates.
We simulate the proposed structure by numerically solving Maxwell’s
equations (Fig. 5A). The device has a spectrally averaged absorption
enhancement factor of Fiso ¼ 119 (red line) for normally incident light.
(All the absorption spectra and enhancement factors are obtained by averag-
ing s and p polarized incident light.) This enhancement factor is well above
the conventional limit for both the active material (4n2L ¼ 10) and the clad-
ding material (4n2H ¼ 50). Moreover, the angular response is nearly isotropic
(Fig. 5C and D). Thus such enhancement cannot be attributed to the
narrowing of angular range in the emission cone, and instead is due entirely
to the nanoscale field confinement effect.
Using our theory, we calculate the theoretical upper limit of light-
trapping enhancement in this structure. For wavelength λ ¼ 500 nm, we
obtain an upper limit of Fiso ¼ 147. The enhancement factor observed in
the simulation is thus consistent with this predicted upper limit. The actual
enhancement factor obtained for this structure falls below the calculated the-
oretical upper limit because some of the resonances are not in the strong
overcoupling regime.
60 Ken Xingze Wang et al.
A B
100
10–1
10–2
10–3
10–4
400 500 600 400 500 600
wavelength (nm)
C 0 D
0
–30 30 –30 30
–60 60 –60 60
–90 90 90
–90
0 50 100 150 0 50 100 150
F F
j = 0° j = 45°
Fig. 5 Absorption with the light-trapping structures. (A) Absorption spectrum for nor-
mally incident light for the structure shown in Fig. 4. The spectrally averaged absorption
(red solid line) is much higher than both the single-pass absorption (light-gray dashed
line) and the absorption as predicted by the limit of 4n2L (dark-gray dashed line). The ver-
tical axis is the absorption coefficient. (B) Absorption spectrum without nanoscale light
confinement. The structure is the same as that of (A) except that the dielectric constant
of the active layer is now the same as the cladding layer. The dark-gray dashed line rep-
resents the absorption as predicted by the limit of 4n2H. (C and D) Angular dependence of
the spectrally averaged absorption enhancement factor for the structure in Fig. 4. Inci-
dent angles are labeled on top of the semicircles. Incident planes are oriented at 0
(C) and 45 (D) degrees (azimuthal angles) with respect to the [10] direction of the lattice.
The red circles represent the 4n2L limit.
including nanowire (Kayes et al., 2005; Law et al., 2005) and plasmonic struc-
tures (Atwater and Polman, 2010). In plasmonic structures, the presence of
nanoscale guided modes may also provide opportunities to overcome the
conventional limit.
the absorption enhancement factor in such a weak absorption limit, one can
derive the enhancement factor where the single-pass absorption is no longer
negligible (Green, 2002; Yu et al., 2012). Focusing on the weak absorption
limit therefore allows us to focus on illustrating the most prominent effects of
the DOS engineering in photonic crystals on light trapping absorption
enhancement.
Real solar cells are of course 3D structures. To simplify numerical cal-
culations, however, many theoretical literatures on light trapping in solar
cells have considered 2D structures (Chutinan and John, 2008; Chutinan
et al., 2009; Duche et al., 2011; Gomard et al., 2012; Park et al., 2009;
Yu et al., 2010b). By 2D, we mean structures that are uniform in the third
dimension. Moreover, we consider only light propagating within a 2D plane
perpendicular to the third dimension.
The wave optics theory in Section 3 points to several unique aspects of
light trapping in photonic crystals. We examine the most prominent devi-
ations of the optical properties of photonic crystals from those of bulk struc-
tures. First, the DOS of a photonic crystal (for example, in Fig. 6) can differ
significantly from that of the bulk structure (for example, in Fig. 2). Second,
for a bulk structure, if the thickness is greater than a few wavelengths, one
can assume complete overlap of the modes with the absorbing material. In
contrast, in a photonic crystal, the electromagnetic fields do not completely
overlap with the absorbing material, and therefore the modal absorption loss
rate is related to the photonic crystal geometries as well as material constants.
Third, to access all the modes in a bulk structure, one typically employs sur-
face roughness. In contrast, we will show that a photonic crystal by itself,
with appropriate configurations, provides complete access to all its modes
without surface disorders.
0.8
Frequency (c/a)
0.6
0.4
TE modes
M
0.2 K1 K2
X
G
0 TM modes
G X M G
Fig. 7 Band structure of the photonic crystal in Fig. 6. The first Brillouin zone of the
square lattice is shown in the inset. Γ denotes the k point (0, 0), X denotes (0, π/a),
and M denotes (π/a, π/a).
Fig. 8 Projected TM band structure obtained from Fig. 7. The lightline is given by ω ¼ ckx,
where kx is the projected wave vector.
(TM) polarization, but not for the transverse electric (TE) polarization. In
order to highlight the features in the DOS of a 2D photonic crystal, we
choose to work with the TM polarization for all subsequent discussions
in 2D structures.
The same photonic band structure in Fig. 7 can alternatively be presented
as a projected band diagram (Fig. 8), where we project the 2D band structure
ω(kx, kz) onto the ω kx plane. The shaded regions in Fig. 8 corresponds to
states in the photonic crystal, the unshaded region is the gap region. We
notice that a significant part of the first band and some parts of the second
64 Ken Xingze Wang et al.
band lie below the light line ω ¼ ckx. Such a projected band structure will be
useful for the discussion of mode coupling issues.
We can calculate the DOS of this photonic crystal by a uniform sampling
of all the k-points in the first Brillouin zone. In Fig. 9, we observe two
bandgaps at frequencies close to 0.3c/a and 0.7c/a, as well as a number of
van Hove singularities where the DOS is divergent (Bassani and
Parravicini, 1975; Van Hove, 1953).
For each mode, we can also calculate the overlap factor Γ between the
electric field and the dielectric rods using the Dirichlet-to-Neumann (DTN)
method (Liu and Fan, 2011). At any given frequency, we select two k points
from the segments connecting Γ to X to M and back to Γ. The selected k
points correspond to the two leftmost intersections of horizontal cuts with
the band structure plot in Fig. 7. Fig. 10 shows the overlap factors for such
Fig. 9 Density of TM states per area for the 2D photonic crystal of Fig. 6.
Fig. 10 Overlap factor of the TM mode for two selected k points at each frequency. The
k points are selected from left to right, for example, at a frequency of 0.2c/a, the first two
k points are shown as K1 and K2, in Fig. 7. Inside the bandgap, the overlap factor is not
defined.
Light trapping in photonic structures 65
Z π=2
F2D ¼ dθf ðθÞ cos θ
π=2
Z (17)
1
¼ dkx f ðkx Þ
k0 |kx |<k0
Δk X
¼ fn (18)
k0 n
ð2πc Þ2 Γ
F2d < ρ ð ωÞ (19)
nηω 2D
where ρ2D(ω) is the DOS per unit area.
We calculated ρ2D(ω) (Fig. 9), as well as Γ(ω) (Fig. 10) for selected k
points. Since the overlap factors Γ are approximately equal for different k
points at the same frequency, we simply take an average of the two overlap
factors shown in Fig. 10, and use the averaged overlap factor Γ in Eq. (19).
With these information and other parameters of the structure, we can obtain
the right-hand side expression (2πc)2Γρ2D(ω)/nηω of Eq. (19), which gives
Light trapping in photonic structures 67
an upper bound for F2D. We plot this bound for a range of frequencies in
Fig. 11, referring to it as the “Theoretical Bound” curve.
Fig. 11 verifies our theoretical result on the correspondence between the
absorption enhancement and the theoretical bound. We note in particular
the following features. First, over a few narrow ranges of frequencies, the
Yablonovitch limit is surpassed due to the van Hove singularities in the pho-
tonic DOS. Second, overall the absorption enhancement is below the con-
ventional limit. Third, the photonic crystal absorber, without grating or
surface roughness, is itself a very efficient scatterer. Over most frequency
ranges above 0.3c/a, there is a close match between the DOS and the absorp-
tion curves, implying that most accessible resonances are excited. Unlike the
conventional case of uniform slab, there is no need to use a surface grating to
achieve light trapping for photonic crystals. However, on the other hand, for
the first band from a frequency of 0 to approximately 0.3c/a, the numerically
obtained light trapping absorption enhancement is significantly lower than
its upper bound imposed by the DOS. This is because many modes are
below the light line and decoupled from external radiation, as we have
eluded to in the discussion of Fig. 8.
singularities. In this photonic crystal, the theoretical bound exceeds the con-
ventional limit in the frequency range approximately from 0.22c/a to 0.27c/a,
due to the van Hove singularity in DOS. Assuming that the frequency range
is centered at a free space wavelength of 1000 nm, this frequency range then
corresponds to a wavelength range of 907 to 1114 nm. Therefore, it is pos-
sible to use van Hove singularities to achieve light trapping limit above the
conventional limit over a bandwidth that is relevant for solar cell light trap-
ping at the crystalline silicon bandedge. We may also stack multiple photonic
crystals to use multiple van Hove singularities for an even broader bandwidth
of light trapping absorption enhancement.
modes in the crystal are above the light line. Many other different trunca-
tions are possible.
In the simulation results shown in Fig. 14, we indeed observe significant
improvement of light trapping enhancement in the first band, as compared
to the structure in Fig. 6. Also, the light trapping enhancement in the upper
bands remain essentially unchanged as we vary the crystal orientation, as
expected from the mode coupling argument.
Fig. 13 A different truncation of the 2D photonic crystal in Fig. 6. The structure is rotated
pffiffiffi
by 26.565∘ from the original structure in Fig. 6. The effectively period is 5a, where a is
the lattice constant of the square lattice.
Fig. 15 The photonic crystal in Fig. 6 with one additional scattering layer on top. The
dielectric rods in the first layer on top are transparent while the other rods are weakly
absorptive. The real part of the dielectric constant is 12. The radius of each absorptive
rod is 0.2a, where a is the lattice constant. The radii of the transparent rods are alter-
nating in every five rods, being 0.2a, 0.18a, 0.22a, 0.16a, and 0.24a, thus forming a
scattering layer.
74 Ken Xingze Wang et al.
Z π=2 Z 2π
F3D ¼ dθ dφf ðθ, φÞcos θ sinθ,
0
Z Z 0 (23)
1
¼ 2 dkx dky f kx , ky ,
k0 k2x + k2y <k20
Δk2 X
¼ 2 fn (24)
k0 n
kx ¼ k0 sin θ cos φ,
ky ¼ k0 sinθ sinφ,
ω
k0 ¼ ,
c
2π
Δk ¼ :
L
Similar to Eq. (19), we have
ð2πc Þ3 Γ
F3D < ρ ðωÞ (25)
nηω2 3D
where ρ3D(ω) is the DOS per unit volume.
Here, we highlight several notable differences between the 3D and 2D
photonic crystals. In 3D, one can show that in the long wavelength limit, for
a high index absorbing material embedded in a low index transparent back-
ground, the same index contrast and volume fraction would cause greater
Light trapping in photonic structures 75
DOS reduction than in 2D. The result of the effective medium theory for
this case gives
3
F3D ε2 2
< η + ð1 ηÞ (26)
4πn2 ε1
Fig. 18 3D photonic crystal. The woodpile structure consists of four alternating layers.
Each layer has a thickness of 0.25a and consists of a 1D array of infinitely long square rods
with widths of 0.25a, where a is the lattice constant. The horizontal locations of the first
and the second layers are shifted laterally by 0.5a from the horizontal locations of the
third and the fourth layers. The lattice orientations of the first and the third layers are
orthogonal to the lattice orientations of the second and the fourth layers.
We obtain ρ3D(ω) by MPB for an infinite and lossless structure, and F3D
by S4 for the corresponding finite and lossy structure. We plot the right-
hand side expression (2πc)3ρ3D(ω)/nω2 in Eq. (25) as the “Theoretical
Bound” curve in Fig. 19, assuming that the overlap factor Γ is equal to
the volume fraction η. We plot F3D as the “Absorption Calculation” curve
in Fig. 19. The absorption enhancement follows the DOS in the same man-
ner as in the 2D case. We observe that both the theoretical bound and the
actual angle-integrated absorption enhancement factor fall significantly
below the conventional limit of 4πn2 ¼ 150 in this case.
To summarize this section, we have considered light trapping in pho-
tonic crystals, where the photonic crystals themselves operate as inhomoge-
neous absorbing media. We show that the DOS of photonic crystals strongly
influences the light trapping behaviors. We also note the importance of
modal overlap and the need for efficient mode coupling. Our numerical
results show that, in a 2D photonic crystal, with the use of van Hove singu-
larities in the DOS, the angle-integrated light trapping absorption enhance-
ment factor can exceed the conventional limit over substantial bandwidths.
In a 3D photonic crystal, it is more difficult to use photonic crystals to
overcome the conventional limit, due to the weakening of the van Hove
singularity and the fact that embedding a high index absorbing material in
a low index background reduces the overall DOS away from van Hove
singularities.
Light trapping in photonic structures 77
6.1 Motivation
There is significant recent interest in designing ultrathin c-Si solar cells with
active layer thickness of a few microns (Biswas and Xu, 2011; Dewan et al.,
2011; Fahr et al., 2011; Gjessing et al., 2011; Grandidier et al., 2011; Han
and Chen, 2010; Kelzenberg et al., 2010; Li et al., 2010; Mallick et al.,
2010; Nagel and Scarpulla, 2010; Paetzold et al., 2011; Park et al., 2009;
Sheng et al., 2011; Wang et al., 2010; Zhang et al., 2011). In Fig. 20, we
plot the complex dielectric constant of c-Si in comparison with the Air Mass
1.5 (AM 1.5) solar irradiance spectrum. The AM 1.5 spectrum is broadband,
hence it is beneficial to increase solar absorption over the entire optical spec-
trum. The real part of the dielectric constant is much greater than unity,
which implies that there is a strong optical impedance mismatch at the air
78 Ken Xingze Wang et al.
0.5
50
Dielectric constants of crystalline silicon
40 er
30 ei
20
10
and c-Si interface, and as a result antireflection is desirable. Close to the c-Si
bandedge, the imaginary part of the dielectric constant is negligible, which
implies that the absorption depth of infrared photons is extremely large, and
as a result light trapping should be performed.
Efficient light absorption in such thin films therefore requires both
broadband antireflection coatings and effective light trapping techniques,
which often have different design considerations. In this section, we show
that by employing a double-sided grating design, we can separately optimize
the geometries for antireflection and light trapping purposes to achieve
broadband light absorption enhancement. The photocurrent generated by
the proposed thin film absorber is able to reach the Yablonovitch limit
(Green, 2002; Yablonovitch, 1982; Yablonovitch and Cody, 1982).
We use nanocones as the basic building elements for the grating geom-
etry because of their unique optical properties and compatibility with inex-
pensive fabrication techniques (Zhu et al., 2009, 2010a,b). The structure we
consider, as shown in Fig. 21, contains a c-Si thin film with nanocone grat-
ings also made of c-Si. The circular nanocones form 2D square lattices on
both the front and the back surfaces. The film is placed on a mirror. As a
starting point we assume the mirror is made of a PEC.
Light trapping in photonic structures 79
Fig. 21 3D c-Si thin film structures in air. Blue represents c-Si, gray represents PEC, and
yellow represents nonabsorbing silicon nitride. The nanocones are made of c-Si, as is the
uniform layers, and they are placed in a 2D square lattice on both surfaces of the film.
Fig. 22 3D c-Si thin film structures in air. In all panels, blue represents silicon, gray rep-
resents a PEC, and yellow represents nonabsorbing silicon nitride. The nanocones are
made of silicon, as is the uniform layers, and they are placed in a 2D square lattice either
on the front or on the back surface of the film. (A) The optimized double-sided nano-
structure, (B) The optimized top-only nanostructure. (C) The optimized bottom-only
nanostructure with a thin layer of nonabsorbing silicon nitride on top. (D) The flat film
with a thin layer of nonabsorbing silicon nitride on top.
through the material only once, and assuming normal incidence, the single-
pass absorption spectrum (the green curves in Fig. 23) in a thin film with
thickness d is given by
Fig. 23 Absorption spectra under normal incidence from AM 1.5 solar irradiance. The
red curves represent the Yablonovitch limit given by Eq. (28), the green curves represent
the single-pass absorption spectra given by Eq. (27), and the thick black curves are run-
ning averages of the absorption spectra for the corresponding structures in Fig. 22. The
kinks at the longer wavelength range in the curves are due to the discontinuity of the
material constant from the reference book (Palik, 1997). (A) The absorption spectrum of
the double-sided structure in Fig. 22A. (B) The absorption spectrum of the “top-only”
structure in Fig. 22B. (C) The absorption spectrum of the “bottom-only” structure in
Fig. 22C. (D) The absorption spectrum of the flat thin film in Fig. 22D. (E) Comparison
of the short-circuit currents generated by the four structures (gray bars), the
Yablonovitch limit (red line), the single-pass absorption (green line), and the full absorp-
tion (blue line).
The device physics for the absorption enhancement strategy in our struc-
ture is as follows: The usable solar spectrum for a c-Si cell extends from 300
to 1100 nm. To achieve significant absorption enhancement, one needs
broad-band antireflection over the entire usable solar spectrum due to the
strong impedance mismatch between silicon and air. One also needs efficient
82 Ken Xingze Wang et al.
light trapping from roughly 800 to 1100 nm where a silicon cell with an
equivalent thickness of a few microns is weakly absorbing. Our strategy is
to use the front surface grating for the goal of antireflection, and the back
surface grating for the goal of light trapping. Doing so allows us to separately
address the different structural requirements in order to achieve these two
separate goals. Below, we illustrate this strategy by comparing our optimized
structure to optimized “top-only” and “bottom-only” grating structures. All
these structures have an equivalent thickness of 2 μm. In the study of these
optimized “top-only” or “bottom-only” structures, we will compare their
performance to a flat thin film structure with the same equivalent thickness
and with a nonabsorbing silicon nitride antireflection coating on top, as
shown in Fig. 22D.
The optimized “top-only” structure and its absorption spectrum are
shown in Figs. 22B and 23B, respectively. The periodicity of the nanocone
grating is 500 nm, and the height is 900 nm, four times the base radius of
225 nm. The planar part of the structure has a thickness of 1809 nm. Com-
paring the absorption spectrum of such an optimized “top-only” structure
(see Fig. 23B) to that of the unpatterned flat thin film in Fig. 23D, we
observe substantial absorption enhancement over the entire usable solar
spectrum. The short-circuit photocurrent for the “top-only” structure is
31.7 mA/cm2, compared to 19.7 mA/cm2 for the flat thin film. From 300
to 700 nm, the absorption curve closely follows the Yablonovitch limit.
The contribution for light absorption enhancement in Fig. 23B, com-
pared to the flat thin film absorption in Fig. 23D, originates mainly from
antireflection. Nanocone arrays suppress reflection because the cone geom-
etry provides an averaged, graded index from air to silicon as the radius of its
cross section increases from zero to its maximum at the planar film surface.
The reflection suppression is broadband since the index-matching is largely
independent of wavelength. To achieve effective antireflection, the period-
icity of the array has to be in the subwavelength regime for the incoming
light to see an effective averaged index. In addition, a high aspect ratio is
preferred to provide a smooth index transition from air to silicon. These
structural aspects are precisely what we see in the optimized “top-only”
structure. However, for longer wavelengths, between 700 to 1100 nm,
the absorption of the optimized “top-only” structure falls significantly below
the Yablonovitch limit. Therefore, the structural feature of a nanocone that
is optimal for antireflection purposes is suboptimal for light trapping. Nev-
ertheless, such a nanocone structure for antireflection has excellent perfor-
mance in a wide range of light incident angles since the gradual change of
refractive index is maintained over this range (Zhu et al., 2009, 2010b).
Light trapping in photonic structures 83
shows that for ultrathin films, our approach can consistently outperform
both the “top-only” and “bottom-only” grating designs. The photocurrents
from the optimized double-sided grating structures are very close to the
theoretical Yablonovitch limit for a range of thicknesses in the few
microns range.
To characterize the loss of real mirror, we replace the PEC mirror by a
flat silver layer for the optimized structure in Fig. 22A. A thin layer of silicon
dioxide is placed between the silicon film and silver as a spacer to reduce
metallic loss. We observe only a small reduction in short-circuit photocur-
rent, from 34.63 to 33.86 mA/cm2 (see Fig. 25A). Since the grating at the
bottom of the silicon layer is sufficiently far from the silver surface, there is
no surface plasmon excitation (see Fig. 25B and C). Therefore, our flat silver
back-reflector induces no significant local or long-range surface plasmon
excitation and its resultant substantial parasitic loss (Biswas and Xu, 2011).
When we extend our optimization to include nanocone gratings made of
nonabsorbing dielectric materials instead of silicon, we observe similar geo-
metrical configurations and absorption enhancements. Therefore, the strat-
egy has great flexibility in the nanocone grating design regarding material
choice for either the grating or the back-reflector.
As a final remark, the doubled-sided structure could be fabricated
by applying the Langmuir–Blodgett (LB) assembly method of silica
nanospheres and the reactive ion etching (RIE) to each side sequentially
(Hsu et al., 2008; Zhu et al., 2009, 2010a). One could use the size of the
silica nanoparticles to control the periodicity, and the RIE to control the
shapes of the nanocones. This fabrication process could also be applied to
Light trapping in photonic structures 85
Fig. 25 (A) Comparison of the absorption spectra with PEC and with silver (plus a spacer
between the nanostructure and the silver back-reflector). The red curve is the absorption
spectrum of the structure in Fig. 22A, and the blue curve is the absorption spectrum of
the same structure, except that the PEC is replaced with real silver (Palik, 1997), and a
silicon dioxide layer of thickness 2 μm is placed between silicon and silver as a spacer.
(B) and (C) The double-sided grating structure and the electric field profile at the wave-
length of 950 nm.
a large scale (Hsu et al., 2008; Zhu et al., 2010a). Since our structure involves
only a slight surface modification of a silicon film with nanocones of rela-
tively low aspect ratio, the issues associated with surface recombination
and degradation of electronic properties should be less severe than those
of other high aspect ratio nanoscale structures. The double-sided nanocone
grating design provides an experimentally realistic strategy in efficiency
improvement and cost reduction for c-Si solar cells. Although we focus
on c-Si, many of the strategies can be applied to other materials (Liang
et al., 2012).
86 Ken Xingze Wang et al.
7. Conclusion
Light trapping is desirable for weak absorbers because the single-pass
absorption is usually too small for a significant percentage of light to be
absorbed by the solar absorber. In this chapter, we describe a statistical
TCMT formalism to construct a wave optics light trapping theory, which
reduces to the conventional light trapping theory in the ray optics regime.
We show that the DOS, especially the van Hove singularities, of photonic
crystals strongly influences the light trapping behaviors. We validate the
wave optics light trapping theory and show that nanophotonic light trapping
performance can surpass the conventional benchmarks, such as the
Yablonovitch 4n2 limit, even after a complete angle integration. We apply
light trapping with antireflection to practical photovoltaic cells. Using
jointly optimized nanocone gratings on both sides of ultrathin c-Si solar
cells, we showed that it is possible to reach the 4n2 limit. These results pro-
vide a theoretical guidance for the design of optical absorbers that are
inhomogeneous in general, including nanostructured photovoltaic and
photoelectrochemical cells. We show that to design and experimentally real-
ize high-efficiency photonic crystal solar cells, one should include consider-
ations on the DOS including the effects of the effective medium and van
Hove singularities, the overlap factor, the mode coupling, and antireflection
designs.
References
Agrawal, M., Peumans, P., 2008. Broadband optical absorption enhancement through
coherent light trapping in thin-film photovoltaic cells. Opt. Express 16 (8), 5385–5396.
Almeida, V.R., Xu, Q., Barrios, C.A., Lipson, M., 2004. Guiding and confining light in void
nanostructure. Opt. Lett. 29 (11), 1209–1211.
Atwater, H.A., Polman, A., 2010. Plasmonics for improved photovoltaic devices. Nat.
Mater. 9, 205–213.
Barnett, S.M., Loudon, R., 1996. Sum rule for modified spontaneous emission rates. Phys.
Rev. Lett. 77 (12), 2444–2446.
Bassani, F., Parravicini, G.P., 1975. Electronic States and Optical Transistions in Solids.
Pergamon Press.
Battaglia, C., Hsu, C.-M., S€ oderstr€om, K., Escarre, J., Haug, F.-J., Charrière, M.,
Boccard, M., Despeisse, M., Alexander, D.T.L., Cantoni, M., Yi, C., Ballif, C.,
2012a. Light trapping in solar cells: can periodic beat random? ACS Nano 6 (3),
2790–2797.
Battaglia, C., Boccard, M., Haug, F.-J., Ballif, C., 2012b. Light trapping in solar cells: when
does a Lambertian scatterer scatter Lambertianly? J. Appl. Phys. 112, 094504.
Bergman, D.J., 1978. The dielectric constant of a composite material—a problem in classical
physics. Phys. Rep. 43 (9), 377–407.
Light trapping in photonic structures 87
Bergman, D.J., 1980. Exactly solvable microscopic geometries and rigorous bounds for the
complex dielectric constant of a two-component composite material. Phys. Rev. Lett.
44, 1285.
Bermel, P., Luo, C., Zeng, L., Kimerling, L.C., Joannopoulos, J.D., 2007. Improving thin-
film crystalline silicon solar cell efficiencies with photonic crystals. Opt. Express 15 (25),
16986–17000.
Biswas, R., Xu, C., 2011. Nano-crystalline silicon solar cell architecture with absorption at
the classical 4n2 limit. Opt. Express 19 (S4), A664–A672.
Bozzola, A., Liscidini, M., Andreani, L.C., 2012. Photonic light-trapping versus Lambertian
limits in thin film silicon solar cells with 1D and 2D periodic patterns. Opt. Express
20 (S2), A224–A244.
Brongersma, M.L., Cui, Y., Fan, S., 2014. Light management for photovoltaics using high-
index nanostructures. Nat. Mater. 13, 451–460.
Callahan, D.M., Munday, J.N., Atwater, H.A., 2012. Solar cell light trapping beyond the ray
optic limit. Nano Lett. 12, 214–218.
Chutinan, A., John, S., 2008. Light trapping and absorption optimization in certain thin-film
photonic crystal architectures. Phys. Rev. A 78, 023825.
Chutinan, A., Kherani, N.P., Zukotynski, S., 2009. High-efficiency photonic crystal solar
cell architecture. Opt. Express 17 (11), 8871–8878.
Deckman, H.W., Roxlo, C.B., Yablonovitch, E., 1983. Maximum statistical increase of
optical absorption in textured semiconductor films. Opt. Lett. 8 (9), 491–493.
Dewan, R., Vasilev, I., Jovanov, V., Knipp, D., 2011. Optical enhancement and losses of
pyramid textured thin-film silicon solar cells. J. Appl. Phys. 110, 013101.
Du, Q.G., Kam, C.H., Demir, H.V., Yu, H.Y., Sun, X.W., 2011. Enhanced optical absorp-
tion in nanopatterned silicon thin films with a nano-cone-hole structure for photovoltaic
applications. Opt. Lett. 36 (9), 1713–1715.
Duche, D., Escoubas, L., Simon, J.-J., Torchio, P., Vervisch, W., Flory, F., 2008. Slow
Bloch modes for enhancing the absorption of light in thin films for photovoltaic cells.
Appl. Phys. Lett. 92, 193310.
Duche, D., Drouard, E., Simon, J.J., Escoubas, L., Torchio, P., Le Rouzo, J., Vedraine, S.,
2011. Light harvesting in organic solar cells. Sol. Energy Mater. Sol. Cells 95, S18–S25.
Eyderman, S., John, S., Deinega, A., 2013. Solar light trapping in slanted conical-pore pho-
tonic crystals: beyond statistical ray trapping. J. Appl. Phys. 113, 154315.
Fahr, S., Kirchartz, T., Rockstuhl, C., Lederer, F., 2011. Approaching the Lambertian limit
in randomly textured thin-film solar cells. Opt. Express 19 (S4), A865–A874.
Fan, S., Joannopoulos, J.D., 2002. Analysis of guided resonances in photonic crystal slabs.
Phys. Rev. B 65, 235112.
Ganapati, V., Miller, O.D., Yablonovitch, E., 2014. Light trapping textures designed by elec-
tromagnetic optimization for subwavelength thick solar cells. IEEE J. Photovoltaics 4 (1),
175.
Gaponenko, S.V., 2010. Introduction to Nanophotonics. Cambridge University Press.
Gjessing, J., Sudbø, A.S., Marstein, E.S., 2011. Comparison of periodic light-trapping struc-
tures in thin crystalline silicon solar cells. J. Appl. Phys. 110, 033104.
Goetzberger, A., 1981. Optical confinement in thin Si-solar cells by diffuse back reflectors.
In: 15th Photovoltaic Specialists Conference. IEEE, pp. 867–870.
Gomard, G., Meng, X., Drouard, E., El Hajjam, K., Gerelli, E., Peretti, R., Fave, A.,
Orobtchouk, R., Lemiti, M., Seassal, C., 2012. Light harvesting by planar photonic crys-
tals in solar cells: the case of amorphous silicon. J. Opt. 14 (2), 024011.
Gomard, G., Peretti, R., Drouard, E., Meng, X., Seassal, C., 2013. Photonic crystals and
optical mode engineering for thin film photovoltaics. Opt. Express 21 (8), A515.
Grandidier, J., Callahan, D.M., Munday, J.N., Atwater, H.A., 2011. Light absorption
enhancement in thin-film solar cells using whispering gallery modes in dielectric
nanospheres. Adv. Mater. 23, 1272–1276.
88 Ken Xingze Wang et al.
Green, M.A., 2002. Lambertian light trapping in textured solar cells and light-emitting
diodes: analytical solutions. Prog. Photovolt. Res. Appl. 10 (4), 235–241.
Green, M.A., 2011. Enhanced evanescent mode light trapping in organic solar cells and other
low index optoelectronic devices. Prog. Photovolt. Res. Appl. 19 (4), 473–477.
Guldin, S., H€ uttner, S., Kolle, M., Welland, M.E., M€ uller-Buschbaum, P., Friend, R.H.,
Steiner, U., Tetreault, N., 2010. Dye-sensitized solar cell based on a three-dimensional
photonic crystal. Nano Lett. 10 (7), 2303–2309.
Han, S.E., Chen, G., 2010. Toward the Lambertian limit of light trapping in thin nanostruc-
tured silicon solar cells. Nano Lett. 10 (11), 4692–4696.
Haug, F.-J., S€ oderstr€
om, K., Naqavi, A., Ballif, C., 2011. Resonances and absorption
enhancement in thin film silicon solar cells with periodic interface texture. J. Appl. Phys.
109, 084516.
Heine, C., Morf, R.H., 1995. Submicrometer gratings for solar energy applications. Appl.
Opt. 34 (14), 2476–2482.
Ho, K.M., Chan, C.T., Soukoulis, C.M., Biswas, R., Sigalas, M., 1994. Photonic band gaps
in three dimensions: new layer-by-layer periodic structures. Solid State Commun. 89 (5),
413–416.
Hoppe, H., Sariciftci, N.S., 2004. Organic solar cells: an overview. J. Mater. Res. 19 (7),
1924–1945.
Hsu, C.-M., Connor, S.T., Tang, M.X., Cui, Y., 2008. Wafer-scale silicon nanopillars and
nanocones by Langmuir-Blodgett assembly and etching. Appl. Phys. Lett. 93, 133109.
Huynh, W.U., Dittmer, J.J., Alivisatos, P., 2002. Hybrid nanorod-polymer solar cells.
Science 295 (5564), 2425–2427.
Joannopoulos, J.D., Johnson, S.G., Winn, J.N., Meade, R.D., 2008. Photonic Crystals:
Molding the Flow of Light, second ed. Princeton University Press.
John, S., 2012. Why trap light? Nat. Mater. 11, 997–999.
Johnson, S.G., Joannopoulos, J.D., 2001. Block-iterative frequency-domain methods for
Maxwell’s equations in a planewave basis. Opt. Express 8 (3), 173–190.
Kayes, B.M., Atwater, H.A., Lewis, N.S., 2005. Comparison of the device physics principles
of planar and radial p-n junction nanorod solar cells. J. Appl. Phys. 97 (11), 114302.
Kelzenberg, M.D., Boettcher, S.W., Petykiewicz, J.A., Turner-Evans, D.B., Putnam, M.C.,
Warren, E.L., Spurgeon, J.M., Briggs, R.M., Lewis, N.S., Atwater, H.A., 2010.
Enhanced absorption and carrier collection in Si wire arrays for photovoltaic applica-
tions. Nat. Mater. 9, 239–244.
Ko, D.-H., Tumbleston, J.R., Zhang, L., Williams, S., DeSimone, J.M., Lopez, R.,
Samulski, E.T., 2009. Photonic crystal geometry for organic solar cells. Nano Lett.
9 (7), 2742–2746.
Kuang, P., Deinega, A., Hsieh, M.-L., John, S., Lin, S.-Y., 2013. Light trapping and near-
unity solar absorption in a three-dimensional photonic-crystal. Opt. Lett. 38 (20),
4200–4203.
Law, M., Greene, L.E., Johnson, J.C., Saykally, R., Yang, P., 2005. Nanowire dye-sensitized
solar cells. Nat. Mater. 4 (6), 455.
Le, K.Q., John, S., 2014. Synergistic plasmonic and photonic crystal light-trapping: architec-
tures for optical up-conversion in thin-film solar cells. Opt. Express 22 (S1), A1–A12.
Li, L., 1997. New formulation of the Fourier modal method for crossed surface-relief grat-
ings. J. Opt. Soc. Am. A 14 (10), 2758–2767.
Li, J., Yu, H.Y., Wong, S.M., Zhang, G., Lo, G.-Q., Kwong, D.-L., 2010. Si nanocone array
optimization on crystalline Si thin films for solar energy harvesting. J. Phys. D. Appl.
Phys. 43 (25), 255101.
Liang, D., Huo, Y., Kang, Y., Wang, K.X., Gu, A., Tan, M., Yu, Z., Li, S., Jia, J., Bao, X.,
Wang, S., Yao, Y., Wong, H.-S.P., Fan, S., Cui, Y., Harris, J.S., 2012. Optical absorp-
tion enhancement in freestanding GaAs thin film nanopyramid arrays. Adv. Energy
Mater. 2 (10), 1254–1260.
Light trapping in photonic structures 89
Lin, C., Chen, C., Sharkawy, A., Schneider, G.J., Venkataraman, S., Prather, D.W., 2005.
Efficient terahertz coupling lens based on planar photonic crystals on silicon on insulator.
Opt. Lett. 30 (11), 1330–1332.
Liu, V., Fan, S., 2011. Efficient computation of equifrequency surfaces and density of states in
photonic crystals using Dirichlet-to-Neumann maps. J. Opt. Soc. Am. B 28 (8), 1837.
Liu, V., Fan, S., 2012. S4: a free electromagnetic solver for layered periodic structures. Com-
put. Phys. Commun. 183, 2233–2244.
Mallick, S.B., Agrawal, M., Peumans, P., 2010. Optimal light trapping in ultra-thin photonic
crystal crystalline silicon solar cells. Opt. Express 18 (6), 5691–5706.
Mallick, S.B., Agrawal, M., Wangperawong, A., Barnard, E.S., Singh, K.K., Visser, R.J.,
Brongersma, M.L., Peumans, P., 2012. Ultrathin crystalline-silicon solar cells with
embedded photonic crystals. Appl. Phys. Lett. 100, 053113.
Mann, S.A., Garnett, E.C., 2013. Extreme light absorption in thin semiconductor films
wrapped around metal nanowires. Nano Lett. 13, 3173–3178.
Mayer, A.C., Scully, S.R., Hardin, B.E., Rowell, M.W., McGehee, M.D., 2007. Polymer-
based solar cells. Mater. Today 10 (11), 28–33.
McPhedran, R.C., Botten, L.C., McOrist, J., Asatryan, A.A., de Sterke, C.M.,
Nicorovici, N.A., 2004. Density of states functions for photonic crystals. Phys. Rev.
E 69, 016609.
Mihi, A., Mı́guez, H., 2005. Origin of light-harvesting enhancement in colloidal-photonic-
crystal-based dye-sensitized solar cells. J. Phys. Chem. B 109 (33), 15968–15976.
Milton, G.W., 1981. Bounds on the complex permittivity of a two-component composite
material. J. Appl. Phys. 52, 5286.
Milton, G.W., 2002. The Theory of Composites. Cambridge University Press.
Mokkapati, S., Catchpole, K.R., 2012. Nanophotonic light trapping in solar cells. J. Appl.
Phys. 112, 101101.
Munday, J.N., 2012. The effect of photonic bandgap materials on the shockley-queisser
limit. J. Appl. Phys. 112 (6), 064501.
Munday, J.N., Callahan, D.M., Atwater, H.A., 2012. Light trapping beyond the 4n2 limit in
thin waveguides. Appl. Phys. Lett. 100 (121121).
Mutitu, J.G., Shi, S., Chen, C., Creazzo, T., Barnett, A., Honsberg, C., Prather, D.W., 2008.
Thin film silicon solar cell design based on photonic crystal and diffractive grating struc-
tures. Opt. Express 16 (19), 15238–15248.
Nagel, J.R., Scarpulla, M.A., 2010. Enhanced absorption in optically thin solar cells by scat-
tering from embedded dielectric nanoparticles. Opt. Express 18 (S2), A139–A146.
Naqavi, A., Haug, F.-J., Battaglia, C., Herzig, H.P., Ballif, C., 2013a. Light trapping in solar
cells at the extreme coupling limit. J. Opt. Soc. Am. B 30 (1), 13.
Naqavi, A., Haug, F.-J., Ballif, C., Scharf, T., Herzig, H.P., 2013b. Limit of light coupling
strength in solar cells. Appl. Phys. Lett. 102, 131113.
Naqavi, A., Haug, F.-J., S€ oderstr€
om, K., Battaglia, C., Paeder, V., Scharf, T., Herzig, H.P.,
Ballif, C., 2014. Angular behavior of the absorption limit in thin film silicon solar cells.
Prog. Photovolt. Res. Appl. 22 (11), 1147–1158.
Narasimhan, V.K., Cui, Y., 2013. Nanostructures for photon management in solar cells.
Nanophotonics 2 (3), 187–210.
Oskooi, A., Tanaka, Y., Noda, S., 2014. Tandem photonic-crystal thin films surpassing Lam-
bertian light-trapping limit over broad bandwidth and angular range. Appl. Phys. Lett.
104, 091121.
Paetzold, U.W., Moulin, E., Michaelis, D., B€ ottler, W., W€achter, C., Hagemann, V.,
Meier, M., Carius, R., Rau, U., 2011. Plasmonic reflection grating back contacts for
microcrystalline silicon solar cells. Appl. Phys. Lett. 99, 181105.
Pala, R.A., Liu, J.S.Q., Barnard, E.S., Askarov, D., Garnett, E.C., Fan, S.,
Brongersma, M.L., 2013. Optimization of non-periodic plasmonic light-trapping layers
for thin-film solar cells. Nat. Commun. 4, 2095.
90 Ken Xingze Wang et al.
Palik, E.D. (Ed.), 1997. Handbook of Optical Constants of Solids. Elsevier Inc.
Park, Y., Drouard, E., El Daif, O., Letartre, X., Viktorovitch, P., Fave, A., Kaminski, A.,
Lemiti, M., Seassal, C., 2009. Absorption enhancement using photonic crystals for sil-
icon thin film solar cells. Opt. Express 17 (16), 14312–14321.
Peretti, R., Gomard, G., Seassal, C., Letartre, X., Drouard, E., 2012. Modal approach for
tailoring the absorption in a photonic crystal membrane. J. Appl. Phys. 111, 123114.
Polman, A., Atwater, H.A., 2012. Photonic design principles for ultrahigh-efficiency pho-
tovoltaics. Nat. Mater. 11, 174–177.
Raut, H.K., Ganesh, V.A., Nair, A.S., Ramakrishna, S., 2011. Anti-reflective coatings: a
critical, in-depth review. Energy Environ. Sci. 4, 3779.
Saeta, P.N., Ferry, V.E., Pacifici, D., Munday, J.N., Atwater, H.A., 2009. How much can
guided modes enhance absorption in thin solar cells? Opt. Express 17 (23), 20975–20990.
Sandhu, S., Yu, Z., Fan, S., 2013. Detailed balance analysis of nanophotonic solar cells. Opt.
Express 21 (1), 1209–1217.
Scheel, S., 2008. Sum rule for local densities of states in absorbing dielectrics. Phys. Rev. A
78 (013841).
Sheng, P., Bloch, A.N., Stepleman, R.S., 1983. Wavelength-selective absorption enhance-
ment in thin-film solar cells. Appl. Phys. Lett. 43 (6), 579.
Sheng, X., Johnson, S.G., Michel, J., Kimerling, L.C., 2011. Optimization-based design of
surface textures for thin-film Si solar cells. Opt. Express 19 (S4), A841–A850.
S€
oz€uer, H.S., Dowling, J.P., 1994. Photonic band calculations for woodpile structures.
J. Mod. Opt. 41 (2), 231–239.
Stuart, H.R., Hall, D.G., 1997. Thermodynamic limit to light trapping in thin planar struc-
tures. J. Opt. Soc. Am. A 14 (11), 3001–3008.
Tikhodeev, S.G., Yablonskii, A.L., Muljarov, E.A., Gippius, N.A., Ishihara, T., 2002.
Quasiguided modes and optical properties of photonic crystal slabs. Phys. Rev. B
66, 045102.
Tobias, I., Luque, A., Marti, A., 2008. Light intensity enhancement by diffracting structures
in solar cells. J. Appl. Phys. 104, (3), 034502.
Tsakalakos, L., Balch, J., Fronheiser, J., Shih, M.-Y., LeBoeuf, S.F., Pietrzykowski, M.,
Codella, P.J., Korevaar, B.A., Sulima, O.V., Rand, J., Davuluru, A., Rapol, U.,
2007. Strong broadband optical absorption in silicon nanowire films.
J. Nanophotonics 1 (1), 013552.
Tumbleston, J.R., Ko, D.-H., Samulski, E.T., Lopez, R., 2009a. Electrophotonic enhance-
ment of bulk heterojunction organic solar cells through photonic crystal photoactive
layer. Appl. Phys. Lett. 94, 043305.
Tumbleston, J.R., Ko, D.-H., Samulski, E.T., Lopez, R., 2009b. Absorption and
quasiguided mode analysis of organic solar cells with photonic crystal photoactive layers.
Opt. Express 17 (9), 7670–7681.
Van Hove, L., 1953. The occurence of singularities in the elastic frequency distribution of a
crystal. Phys. Rev. 89, 1189.
Vynck, K., Burresi, M., Riboli, F., Wiersma, D.S., 2012. Photon management in two-
dimensional disordered media. Nat. Mater. 11, 1017.
Wang, P., Menon, R., 2014. Optimization of generalized dielectric nanostructures for
enhanced light trapping in thin-film photovoltaics via boosting the local density of opti-
cal states. Opt. Express 22 (S1), A99–A110.
Wang, W., Wu, S., Reinhardt, K., Lu, Y., Chen, S., 2010. Broadband light absorption
enhancement in thin-film silicon solar cells. Nano Lett. 10 (6), 2012–2018.
Wang, K.X., Yu, Z., Liu, V., Cui, Y., Fan, S., 2012. Absorption enhancement in ultrathin
crystalline silicon solar cells with antireflection and light-trapping nanocone gratings.
Nano Lett. 12 (3), 1616–1619.
Light trapping in photonic structures 91
Wang, S., Weil, B.D., Li, Y., Wang, K.X., Garnett, E., Fan, S., Yi, C., 2013a. Large-area
free-standing ultrathin single-crystal silicon as processable materials. Nano Lett. 13 (9),
4393–4398.
Wang, C., Yu, S., Chen, W., Sun, C., 2013b. Highly efficient light-trapping structure design
inspired by natural evolution. Sci. Rep. 3, 1025.
Wang, K.X., Yu, Z., Liu, V., Raman, A., Cui, Y., Fan, S., 2014a. Light trapping in photonic
crystals. Energy Environ. Sci. 7, 2725–2738.
Wang, K.X., Zongfu, Y., Liu, V., Brongersma, M.L., Jaramillo, T.F., Fan, S., 2014b. Nearly
total solar absorption in ultrathin nanostructured iron oxide for efficient photo-
electrochemical water splitting. ACS Photonics 1 (3), 235–240.
Wiesendanger, S., Zilk, M., Pertsch, T., Rockstuhl, C., Lederer, F., 2013. Combining ran-
domly textured surfaces and photonic crystals for the photon management in thin film
microcrystalline silicon solar cells. Opt. Express 21 (103), A450–A459.
Wong, B.M., Morales, A.M., 2009. Enhanced photocurrent efficiency of a carbon nanotube
p-n junction electromagnetically coupld to a photonic structure. J. Phys. D. Appl. Phys.
42, 055111.
Yablonovitch, E., 1982. Statistical ray optics. J. Opt. Soc. Am. 72 (7), 899–907.
Yablonovitch, E., Cody, G.D., 1982. Intensity enhancement in textured optical sheets for
solar cells. IEEE Trans. Electron Devices ED-29 (2), 300.
Yu, Z., Fan, S., 2011. Angular constraint on light-trapping absorption enhancement in solar
cells. Appl. Phys. Lett. 98, 011106.
Yu, G., Gao, J., Hummelen, J.C., Wudl, F., JHeeger, A., 1995. Polymer photovoltaic cells:
enhanced efficiencies via a network of internal donor-acceptor heterojunctions. Science
270 (5243), 1789–1791.
Yu, Z., Raman, A., Fan, S., 2010a. Fundamental limit of light trapping in solar cells. Proc.
Natl. Acad. Sci. U. S. A. 107 (41), 17491–17496.
Yu, Z., Raman, A., Fan, S., 2010b. Fundamental limit of light trapping in grating structures.
Opt. Express 18 (S3), A366.
Yu, Z., Raman, A., Fan, S., 2012. Thermodynamic upper bound on broadband light cou-
pling with photonic structures. Phys. Rev. Lett. 109, 173901.
Yu, Z., Sandhu, S., Fan, S., 2014. Efficiency above the Shockley-Queisser limit by using
nanophotonic effects to create multiple effective bandgaps with a single semiconductor.
Nano Lett. 14 (1), 66–70.
Zeman, M., Isabella, O., Solntsev, S., J€ager, K., 2013. Modeling of thin-film silicon solar
cells. Sol. Energy Mater. Sol. Cells 119, 94–111.
Zhang, R.Y., Shao, B., Dong, J.R., Zhang, J.C., Yang, H., 2011. Absorption enhancement
analysis of crystallline Si thin film solar cells based on broadband antireflection nanocone
grating. J. Appl. Phys. 110, 113105.
Zhu, J., Yu, Z., Burkhard, G.F., Hsu, C.-M., Connor, S.T., Xu, Y., Wang, Q.,
McGehee, M., Fan, S., Cui, Y., 2009. Optical absorption enhancement in amorphous
silicon nanowire and nanocone arrays. Nano Lett. 9 (1), 279–282.
Zhu, J., Yu, Z., Fan, S., Cui, Y., 2010a. Nanostructured photon management for high per-
formance solar cells. Mater. Sci. Eng. R 70, 330–340.
Zhu, J., Hsu, C.-M., Yu, Z., Fan, S., Cui, Y., 2010b. Nanodome solar cells with efficient
light management and self-cleaning. Nano Lett. 10 (6), 1979–1984.
CHAPTER FOUR
Contents
1. Introduction 93
2. Filtering in the wavevector domain using guided resonances 95
3. Theoretical analysis of isotropic filters using a photonic crystal slab 97
4. Numerical demonstration of the isotropic filters 104
4.1 Isotropic high-pass filter (the Laplacian) 105
4.2 Isotropic low-pass filter 105
4.3 Isotropic band-reject filter 108
4.4 Isotropic band-pass filter 108
5. Conclusion 111
Acknowledgments 111
References 112
1. Introduction
In this chapter we will discuss the use of photonic crystal slab to
accomplish a number of imaging processing tasks, including edge detection,
image smoothing, white noise suppression, and suppression or extraction of
periodic features. All these tasks involve filtering in the wavevector domain.
These tasks are of broad applications ranging from microscopy and medical
imaging to industrial inspection and object detection (Abràmoff et al., 2004;
Brosnan and Sun, 2004; Dalal and Triggs, 2005; Gonzalez and Woods, 2008;
Markham et al., 1963).
All these image processing operations, as mentioned above, can certainly
be implemented using conventional electronic computing. However,
nowadays there are many big-data applications that require real-time and
high-throughput image processing, for which it is of interest to develop
γðkÞ
tðω, kÞ ¼ td + f , (1)
i½ω ωðkÞ + γðkÞ
Fig. 1 Operating principle of filtering in the wavevector domain using guided reso-
nances. (A) Dispersion for a single band of guided resonance ω(k). (B) Transmission
spectrum t(ω) at k ¼ 0. (C) k-dependent transmission coefficient t(k) at the resonant
frequency at k ¼ 0: ω ¼ ω0. (D) Transmission spectrum t(ω) at k 6¼ 0.
f ¼ td ¼ 1 (2)
γðkÞ
tðω, kÞ ¼ 1 : (3)
i½ω ωðkÞ + γðkÞ
ω ¼ ωðkÞ: (4)
∂t ∂t
tðω0 , kÞ ¼ 0 + δωðkÞ + δγðkÞ, (5)
∂ωðkÞΓ ∂γðkÞΓ
where
δωðkÞ ¼ ωðkÞ ω0 , δγðkÞ ¼ γðkÞ γ 0 , (6)
∂t i ∂t
¼ , ¼ 0, (7)
∂ωðkÞΓ γ0 ∂γðkÞΓ
therefore,
i
tðω0 , kÞ ¼ δωðkÞ: (8)
γ0
In this special case, t(k) is simply proportional to the band dispersion δω(k)
near k ¼ 0. If δωðkÞ∝jkj2 , then tðkÞ∝ jkj2 as well.
The analysis above indicates that to use guided resonances to perform
second-order differentiation, it is sufficient to satisfy the following three
conditions:
1. td ¼ 1.
2. Only one guided resonance band is coupled.
3. The band satisfies the dispersion ðωðkÞ ω0 Þ∝ jkj2 .
While the discussion above is specific to differentiation, it is relevant for
understanding of some of the other filter functionalities as well. Based
on the discussion here, in the next section we will illustrate how to use a
photonic crystal slab to achieve these filter functionalities.
Fig. 2 Geometry of the device, which consists of a photonic crystal slab with a square
lattice of air holes, separated from a uniform dielectric slab by an air gap. The slabs
have a dielectric constant E ¼ 4.67. The geometry parameters are: d ¼ 0.50a, r ¼ 0.11a,
ds ¼ 0.12a, dg ¼ 1.93a, where a is the lattice constant. The coordinate system is shown
above the device. The red arrow indicates the direction of the incident light. The electric
field directions of the S and P-polarized light are also indicated. Reprinted with permission
from Guo, C., Xiao, M., Minkov, M., Shi, Y., Fan, S., 2018a. Isotropic wavevector domain image
filters by a photonic crystal slab device. J. Opt. Soc. Am. A Opt. Image Sci. Vis. 35 (10),
1685–1691, the Optical Society.
radii r ¼ 0.11a, where a is the lattice constant. The thickness d and radius r are
chosen to realize isotropic bands of guided resonances. A uniform dielectric
slab with a thickness ds ¼ 0.12a is placed in the vicinity of the photonic crystal
slab. The air gap between the two slabs has a thickness dg ¼ 1.93a. As a side
note, the two-layer structure we use in the chapter is quite compact. For
a resonant wavelength λ ¼ 500 nm and a ¼ 650 nm the total thickness is
d + dg + ds ¼ 1.66 μm. Such structure can be fabricated using focused ion
beam-assisted laser interference lithography (Vogelaar et al., 2001).
To satisfy the first criterion in Section 2, we note that the direct trans-
mission coefficient td is related to the nonresonant transmission pathway
(Ochiai and Sakoda, 2001). Hence, it is possible to realize td ¼ 1 by chang-
ing the thickness of the slab. In the structure as shown in Fig. 2, we achieve
td ¼ 1 by placing a uniform dielectric slab in the vicinity of the photon crystal
Optical image processing using photonic crystal slab 99
slab, and by tuning the distance between the slabs. This has the advantage
that we can tune td without significantly affecting the band structure of
the photonic crystal slab.
To satisfy the second and third criterion in Section 2, one will need to
design the band structure of the guided resonances. The design here is in fact
quite nontrivial due to the vectorial nature of electromagnetic waves. Since
the required filter transfer function H(k) is isotropic in k-space, it is natural to
consider a photonic crystal slab structure that possesses rotational symmetry.
As an illustration, here we consider a slab structure with a square lattice of air
holes that has C4v symmetry. For concreteness, we consider circular holes;
the same analysis will apply to other hole shapes that preserve C4v symmetry.
For such a slab, it is known that at the Γ point, which corresponds to jkj ¼ 0,
the only modes that can couple to external plane waves must be twofold
degenerate, belonging to the two-dimensional irreducible representation
of the C4v group (Fan and Joannopoulos, 2002; Ochiai and Sakoda,
2001). Near such modes, in the vicinity of the Γ point, from the k p analysis,
the band structure in general can be described by the following 2 2
effective Hamiltonian. (As we consider only small wavevector range near
the Γ point, it is sufficient to include only the two modes at the Γ point
in our k p analysis.)
^
HðkÞ ¼ ðω0 iγ 0 + ajkj2 Þ^I + bðk2x k2y Þ^
σ z + ckx ky σ^ x , (9)
where a, b, c are three complex coefficients and the σ’s are the Pauli matrices.
This Hamiltonian has two eigenvalues of
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ω ðkÞ iγ ðkÞ ¼ ω0 iγ 0 + ajkj2 b2 jkj4 + ðc 2 4b2 Þk2x k2y : (10)
Fig. 3 (A, C, D) Nearly isotropic photonic band structure of the photonic crystal slab
shown in Fig. 1 near the frequency ω0 ¼ 0.77098 2πc/a. (A) Band dispersions along
Γ–X and Γ–M directions. (C) Constant frequency contours of the lower band.
(D) Constant frequency contours of the upper band. (B) Scheme of multiple filtering func-
tions of the device. When the light frequency is on resonance at the normal incidence
(ω ¼ ω0, labeled by red arrows in (B) and red horizontal lines in (E,F)), the device realizes
isotropic high-pass filter (Laplacian) at transmission and low-pass filter at reflection. When
the light frequency is detuned from the resonance at the normal incidence (ω ¼ ωk,
labeled by green arrows in (B) and green horizontal lines in (E,F)), the device realizes isotro-
pic band-reject filter at transmission and band-pass filter at reflection. (E) Transmittance jtj
for S polarized light as a function of ω and jkj near ω0 ¼ 0.77098 2πc/a along a general
wavevector direction (φ ¼ 14°). Due to the isotropic band structure, S light only excites
the upper band, and the transmission spectra are almost identical along any wavevector
direction φ. jtj ¼ 0 when ω ¼ ω(k). (F) Reflectance jrj for S polarized light as a function of
ω and jkj near ω0 ¼ 0.77098 2πc/a along a general wavevector direction (φ ¼ 14°). The
reflection spectra are identical along any wavevector direction. jrj ¼ 1 when ω ¼ ω(k). In all
the plots, frequency (ω ω0) is in the units of 104 2πc/a, while jkj, kx, and ky are in the
units of 103 2π/a. Reprinted with permission from Guo, C., Xiao, M., Minkov, M., Shi, Y.,
Fan, S., 2018a. Isotropic wavevector domain image filters by a photonic crystal slab device.
J. Opt. Soc. Am. A Opt. Image Sci. Vis. 35 (10), 1685–1691, the Optical Society.
102 Cheng Guo and Shanhui Fan
γ ðkÞ
r ðω, kÞ ¼ eiϕ , (19)
i½ω ω ðkÞ + γ ðkÞ
where the upper (lower) sign corresponds to S(P) polarized light and upper
(lower) band; ω is the incident light frequency. Therefore on resonance,
t ðω ðkÞ,kÞ ¼ 0, r ðω ðkÞ, kÞ ¼ eiϕ : (20)
The numerically determined transmission and reflection spectra for S
polarized light are plotted in Fig. 3E and F. Due to the effect of single-band
excitation, S-polarized light only excites the upper band of guided reso-
nances. Moreover, as expected from Eqs. (12), (18), and (19), the resultant
transmission and reflection spectra are isotropic, i.e., the spectra are identical
along any wavevector direction as defined by the angle φ in Fig. 2. When
ω ¼ ω(k), the transmittance exhibits sharp dips with jtj ¼ 0, while the reflec-
tance exhibits peaks with jrj ¼ 1, as expected from Eq. (20).
Depending on the operating conditions, the structure as shown in Fig. 2
can be used to perform several very useful image-processing functionalities
(Guo et al., 2018a):
1. Isotropic high-pass filter
At the frequency ω ¼ ω0 ω(k ¼ 0), the k-dependent transmittance is:
jω ðkÞ ω0 j
jt ðω0 ,kÞj ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
½ω ðkÞ ω0 2 + γ ðkÞ2
(21)
jω ðkÞ ω0 j A 2
¼ jkj :
γ0 γ0
This transmittance realizes the Laplacian, a special isotropic high-pass
filter (Guo et al., 2018b).
2. Isotropic low-pass filter
At the frequency ω ¼ ω0, if one considers instead the reflected light,
the transfer function is:
1
jr ðω0 , kÞj ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 + ½ω ðkÞ ω0 2 =γ ðkÞ2
(22)
1
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi :
1 + A2 jkj4 =γ 20
This transfer function realizes an isotropic low-pass filter with reflection
peak jrj ¼ 1 at the Γ point.
104 Cheng Guo and Shanhui Fan
profile, S out ðkx ,ky Þ ¼ Hðkx , ky ÞS in ðkx , ky Þ, where H(kx, ky) is the transfer
function. (3) Obtain the output field profile Sout(x, y) by inverse Fourier
transform. Calculate the output image jSout(x, y)j2.
Fig. 4 Isotropic high-pass filter (Laplacian). (A) Transmittance for S-polarized light
jtsj(kx, ky) at the frequency ω0 ¼ 0.77098 2πc/a. (B) jtsj as a function of jkj along a
general wavevector direction (φ ¼ 14°), and the fitting result of Eq. (21). The fit is very
good for jkj up to 0.6 102 2π/a. (C) Log plot of the Fourier transform of the field
profile for the incident image (E): log jS in jðkx ,ky Þ. (D) Log plot of the Fourier transform
of the field profile for the reflected image (F): log jS out jðkx , ky Þ. The low wavevector
components have been filtered out. (E) Incident image jSinj2 of the Stanford emblem.
The image size is 5220a 3456a. (F) Calculated transmitted image jSoutj2, which shows
the edges with different orientations. jkj, kx, and ky are in the units of 102 2π/a.
Reprinted with permission from Guo, C., Xiao, M., Minkov, M., Shi, Y., Fan, S., 2018a.
Isotropic wavevector domain image filters by a photonic crystal slab device. J. Opt.
Soc. Am. A Opt. Image Sci. Vis. 35 (10), 1685–1691, the Optical Society.
Optical image processing using photonic crystal slab 107
Fig. 5 Isotropic low-pass filter. (A) Reflectance for S-polarized light jrsj(kx, ky) at the
frequency ω0 ¼ 0.77098 2πc/a. (B) jrsj as a function of jkj along a general wavevector
direction (φ ¼ 14°), and the fitting result of Eq. (22). (C) Log plot of the Fourier transform of
the field profile for the incident image (E): log jS in jðkx ,ky Þ. (D) Log plot of the Fourier trans-
form of the field profile for the reflected image (F): logjS out jðkx , ky Þ. The high wavevector
components have been filtered out. (E) Incident image jSinj2 of the Stanford emblem with
substantial white noise component. The image size is 5220a 3456a. (F) Calculated
reflected image jSoutj2, which reduces the white noise by image smoothing. jkj, kx,
and ky are in the units of 102 2π/a. Reprinted with permission from Guo, C., Xiao, M.,
Minkov, M., Shi, Y., Fan, S., 2018a. Isotropic wavevector domain image filters by a photonic
crystal slab device. J. Opt. Soc. Am. A Opt. Image Sci. Vis. 35 (10), 1685–1691, the Optical
Society.
108 Cheng Guo and Shanhui Fan
transform of the field profile for this incident image. Fig. 5D shows the
calculated Fourier spectrum for the reflected image, which is obtained by
a pointwise product of Fig. 5A and C. The high wavevector components
have been filtered out. Fig. 5F shows the calculated reflected image, where
the noise component has indeed been significantly reduced, demonstrating
image smoothing.
Fig. 6 Isotropic band-reject filter. (A) Transmittance for S-polarized light jtsj(kx, ky) at the
frequency ωq ¼ 0.77130 2πc/a. (B) jtsj as a function of jkj along a general wavevector
direction (φ ¼ 14°), and the fitting result of Eq. (23) along with Eq. (12) and γ(k) γ 0.
(C) Log plot of the Fourier transform of the field profile for the incident image
(E): log jS in jðkx , ky Þ. The sinusoidal noise appears as peaks in the spectrum which lie
on an approximate circle around the origin. (D) Log plot of the Fourier transform of
the field profile for the transmitted image (F): log jS out jðkx , ky Þ. The wavevector compo-
nents corresponding to the periodic noise have been filtered out. (E) Incident image
jSinj2 of the Stanford emblem with periodic noise. The image size is 5220a 3456a.
(F) Calculated transmitted image jSoutj2, which eliminates the periodic noise. jkj,
kx, and ky are in the units of 102 2π/a. Reprinted with permission from Guo, C.,
Xiao, M., Minkov, M., Shi, Y., Fan, S., 2018a. Isotropic wavevector domain image filters
by a photonic crystal slab device. J. Opt. Soc. Am. A Opt. Image Sci. Vis. 35 (10),
1685–1691, the Optical Society.
110 Cheng Guo and Shanhui Fan
Fig. 7 Isotropic band-pass filter. (A) Reflectance for S-polarized light jrsj(kx, ky) at the
frequency ωq ¼ 0.77130 2πc/a. (B) jrsj as a function of jkj along a general wavevector
direction (φ ¼ 14°), and the fitting result of Eq. (24) together with Eq. (12) and γ(k) γ 0.
(C) Log plot of the Fourier transform of the field profile for the incident image
(E): log jS in jðkx , ky Þ. The sinusoidal noise appears as impulses in the spectrum which
lie on an approximate circle around the origin. (D) Log plot of the Fourier transform
of the field profile for the reflected image (F): log jS out jðkx , ky Þ. (E) Incident image jSinj2
of the Stanford emblem with periodic noise. The image size is 5220a 3456a.
(F) Calculated reflected image jSoutj2, which isolates the periodic noise and simplifies
its analysis. jkj, kx, and ky are in the units of 102 2π/a. Reprinted with permission from
Guo, C., Xiao, M., Minkov, M., Shi, Y., Fan, S., 2018a. Isotropic wavevector domain image
filters by a photonic crystal slab device. J. Opt. Soc. Am. A Opt. Image Sci. Vis. 35 (10),
1685–1691, the Optical Society.
Optical image processing using photonic crystal slab 111
5. Conclusion
Our design of isotropic wavevector domain image filters using a pho-
tonic crystal slab is based on the guided resonances with isotropic band struc-
ture. The same idea can extend to other photonic structures that host resonant
modes with isotropic band structure. In particular, the phase-shifted Bragg
grating in Bykov et al. (2014) can also perform the four filtering functionalities
of our device, but with the transmission/reflection mode flipped.
In conclusion, we have shown that isotropic high-pass, low-pass, band-
reject, and band-pass filters can be implemented at visible wavelengths using
one photonic crystal slab device. Such a simple but multifunctional photonic
device may find various applications involving image processing. For instance,
it can be used as the first layer of image recognition systems. As such our work
indicates the new opportunities for compact, high-throughput, and low-
energy-consumption optical analog computing as provided by nanophotonic
structures.
Acknowledgments
We acknowledge Dr. Meng Xiao, Dr. Momchil Minkov, and Dr. Yu Shi for collaboration,
and Dr. Bo Zhao, Dr. Alexander Cerjan, and Yu Guo for helpful discussion. This work is
supported by Samsung Electronics and US Air Force (USAF) (FA9550-17-1-0002).
112 Cheng Guo and Shanhui Fan
References
AbdollahRamezani, S., Arik, K., Khavasi, A., Kavehvash, Z., 2015. Analog computing using
graphene-based metalines. Opt. Lett. 40 (22), 5239. ISSN 0146-9592. https://doi.org/
10.1364/OL.40.005239. https://www.osapublishing.org/abstract.cfm?URI¼ol-40-22-
5239.
Abràmoff, M.D., Magalhães, P.J., Ram, S.J., 2004. Image processing with ImageJ.
Biophotonics Int. 11 (7), 36–42. ISSN 1081-8693.
Aizenberg, I.N., Butakoff, C., 2002. In: Frequency domain medianlike filter for periodic
and quasi-periodic noise removal. Proceedings of SPIE, vol. 4667. San Jose, http://
proceedings.spiedigitallibrary.org/proceeding.aspx?articleid¼876767. http://link.aip.
org/link/?PSI/4667/181/1&Agg¼doi.
Banham, M.R., Katsaggelos, A.K., 1997. Digital image restoration. IEEE Signal Process.
Mag. 14 (2), 24–41. ISSN 10535888. https://doi.org/10.1109/79.581363. http://
ieeexplore.ieee.org/document/581363/.
Bracewell, R.N., 1986. The Fourier Transform and Its Applications. McGraw-Hill,
New York. ISBN 0073039381.
Brosnan, T., Sun, D.W., 2004. Improving quality inspection of food products by computer
vision—a review. J. Food Eng. 61 (1), 3–16. ISSN 02608774. https://doi.org/10.1016/
S0260-8774(03)00183-3.
Buades, A., Coll, B., Morel, J.M., 2005. A review of image denoising algorithms, with a new
one. Multiscale Model. Simul. 4 (2), 490–530. ISSN 1540-3459. https://doi.org/
10.1137/040616024. http://epubs.siam.org/doi/10.1137/040616024.
Bykov, D.A., Doskolovich, L.L., Bezus, E.A., Soifer, V.A., 2014. Optical computation
of the Laplace operator using phase-shifted Bragg grating. Opt. Express 22 (21),
25084. ISSN 1094-4087. https://doi.org/10.1364/OE.22.025084. https://www.
osapublishing.org/oe/abstract.cfm?uri¼oe-22-21-25084.
Bykov, D.A., Doskolovich, L.L., Morozov, A.A., Podlipnov, V.V., Bezus, E.A., Verma, P.,
Soifer, V.A., 2018. First-order optical spatial differentiator based on a guided-mode
resonant grating. Opt. Express 26 (8), 10997–11006. ISSN 1094-4087. https://doi.
org/10.1364/OE.26.010997. https://www.osapublishing.org/abstract.cfm?URI¼oe-
26-8-10997.
Chang, T.Y., Hong, J.H., Yeh, P., 1990. Spatial amplification: an image-processing
technique using the selective amplification of spatial frequencies. Opt. Lett. 15 (13), 743.
ISSN 0146-9592. https://doi.org/10.1364/OL.15.000743. https://www.osapublishing.
org/abstract.cfm?URI¼ol-15-13-743.
Chizari, A., Abdollahramezani, S., Jamali, M.V., Salehi, J.A., 2016. Analog optical computing
based on a dielectric meta-reflect array. Opt. Lett. 41 (15), 3451. ISSN 0146-9592.
https://doi.org/10.1364/OL.41.003451. https://www.osapublishing.org/abstract.cfm?URI
¼ol-41-15-3451. https://arxiv.org/abs/1605.07150%0Ahttp://doi.org/10.1364/OL.
41.003451.
Dalal, N., Triggs, B., 2005. Histograms of oriented gradients for human detection. In: 2005
IEEE Computer Society Conference on Computer Vision and Pattern Recognition
(CVPR’05) IEEE, pp. 886–893.
Dong, Z., Si, J., Yu, X., Deng, X., 2018. Optical spatial differentiator based on subwavelength
high-contrast gratings. Appl. Phys. Lett. 112 (18), 181102. ISSN 0003–6951. https://doi.
org/10.1063/1.5026309. http://aip.scitation.org/toc/apl/112/18. http://aip.scitation.
org/doi/10.1063/1.5026309. https://doi.org/10.1063/1.5026309.
Eu, J.K.T., Liu, C.Y.C., Lohmann, A.W., 1973. Spatial filters for differentiation.
Opt. Commun. 9 (2), 168–171. ISSN 00304018. https://doi.org/10.1016/0030-4018
(73)90251-4. http://linkinghub.elsevier.com/retrieve/pii/0030401873902514.
Fan, S., Joannopoulos, J.D., 2002. Analysis of guided resonances in photonic crystal slabs.
Phys. Rev. B 65 (23), 235112. ISSN 0163-1829. https://doi.org/10.1103/
PhysRevB.65.235112.
Optical image processing using photonic crystal slab 113
Fang, Y., Ruan, Z., 2018. Optical spatial differentiator for a synthetic three-dimensional
optical field. Opt. Lett. 43 (23), 5893–5896.
Fang, Y., Lou, Y., Ruan, Z., 2017. On-grating graphene surface plasmons enabling spatial
differentiation in the terahertz region. Opt. Lett. 42 (19), 3840. ISSN 0146-9592.
https://doi.org/10.1364/OL.42.003840. https://www.osapublishing.org/abstract.cfm?
URI¼ol-42-19-3840.
Golovastikov, N.V., Bykov, D.A., Doskolovich, L.L., 2014. Resonant diffraction gratings
for spatial differentiation of optical beams. Quantum Electron. 44 (10), 984–988. ISSN
1063-7818. https://doi.org/10.1070/QE2014v044n10ABEH015477. http://stacks.iop.
org/1063-7818/44/i¼10/a¼984?key¼crossref.8bee7b769ad2412f3500ff65fffe3a5e%5Cn.
http://iopscience.iop.org/1063-7818/44/10/984. http://stacks.iop.org/1063-7818/44/i¼
10/a¼984?key¼crossref.8bee7b769ad2412f3500ff65fffe3a5e.
Gonzalez, R., Woods, R., 2008. Digital Image Processing, third ed. Prentice Hall, Upper
Saddle River, NJ. ISBN 9780131687288.
Goodman, J.W., 2005. Introduction to Fourier Optics, third ed. Roberts and Company,
Englewood. ISBN 0974707724.
G€orlitz, D., Lanzl, F., 1975. A holographic spatial filter for direction independent differen-
tiation. Jpn. J. Appl. Phys. 14 (S1), 223. ISSN 0021-4922. https://doi.org/10.7567/
JJAPS.14S1.223. http://stacks.iop.org/1347-4065/14/223.
Guo, C., Xiao, M., Minkov, M., Shi, Y., Fan, S., 2018a. Isotropic wavevector domain image
filters by a photonic crystal slab device. J. Opt. Soc. Am. A Opt. Image Sci. Vis. 35 (10),
1685–1691.
Guo, C., Xiao, M., Minkov, M., Shi, Y., Fan, S., 2018b. Photonic crystal slab Laplace operator
for image differentiation. Optica 5 (3), 251. ISSN 2334-2536. https://doi.org/10.1364/
OPTICA.5.000251. https://www.osapublishing.org/abstract.cfm?URI¼optica-5-3-251.
Hwang, Y., Davis, T.J., Lin, J., Yuan, X.C., 2018. Plasmonic circuit for second-order spatial
differentiation at the subwavelength scale. Opt. Express 26 (6), 7368. ISSN 1094-4087.
https://doi.org/10.1364/OE.26.007368. https://www.osapublishing.org/abstract.cfm?
URI¼oe-26-6-7368.
Kwon, H., Sounas, D., Cordaro, A., Polman, A., Alù, A., 2018. Nonlocal metasurfaces for
optical signal processing. Phys. Rev. Lett. 121 (17), 173004.
Markham, R., Frey, S., Hills, G.J., 1963. Methods for the enhancement of image detail
and accentuation of structure in electron microscopy. Virology 20 (1), 88–102. ISSN
00426822. https://doi.org/10.1016/0042-6822(63)90143-0. http://linkinghub.elsevier.
com/retrieve/pii/0042682263901430.
Nejad, F.Z., Fleury, R., 2018. Performing mathematical operations using high-index
acoustic metamaterials. New J. Phys. 20, 073001.
Ochiai, T., Sakoda, K., 2001. Dispersion relation and optical transmittance of a hexagonal
photonic crystal slab. Phys. Rev. B 63 (12), 125107. ISSN 0163-1829. https://doi.
org/10.1103/PhysRevB.63.125107.
Pors, A., Nielsen, M.G., Bozhevolnyi, S.I., 2015. Analog computing using reflective plasmonic
metasurfaces. Nano Lett. 15 (1), 791–797. ISSN 1530-6984. https://doi.org/10.1021/
nl5047297.
Roberts, A., Gómez, D.E., Davis, T.J., 2018. Optical image processing with metasurface
dark modes. J. Opt. Soc. Am. A Opt. Image Sci. Vis. 35 (9), 1575–1584.
Sakoda, K., 2005. Optical properties of photonic crystals, Springer Series in Optical Sciences,
vol. 80. Springer-Verlag, Berlin, Heidelberg. ISBN 3-540-20682-5.
Silva, A., Monticone, F., Castaldi, G., Galdi, V., Alu, A., Engheta, N., 2014. Performing
mathematical operations with metamaterials. Science. 343 (6167), 160–163. ISSN
0036-8075. https://doi.org/10.1126/science.1242818.
Sirohi, R.S., Mohan, V.R., 1977. Differentiation by spatial filtering. Opt. Acta 24 (11),
1105–1113. ISSN 0030-3909. https://doi.org/10.1080/713819447. http://www.
tandfonline.com/doi/abs/10.1080/713819447.
114 Cheng Guo and Shanhui Fan
Solli, D.R., Jalali, B., 2015. Analog optical computing. Nat. Photonics 9 (11), 704–706.
ISSN 1749-4885. https://doi.org/10.1038/nphoton.2015.208. http://www.nature.com/
doifinder/10.1038/nphoton.2015.208.
Tikhodeev, S.G., Yablonskii, A.L., Muljarov, E.A., Gippius, N.A., Ishihara, T., 2002.
Quasiguided modes and optical properties of photonic crystal slabs. Phys. Rev. B
66 (4), 045102. ISSN 0163-1829. https://doi.org/10.1103/PhysRevB.66.045102.
Vogelaar, L., Nijdam, W., van Wolferen, H.A.G.M., de Ridder, R.M., Segerink, F.B., Fl€ uck, E.,
Kuipers, L., van Hulst, N.F., 2001. Large area photonic crystal slabs for visible light with
waveguiding defect structures: fabrication with focused ion beam assisted laser interference
lithography. Adv. Mater. 13 (20), 1551. ISSN 09359648. https://doi.org/10.1002/1521-
4095(200110)13:20¡1551::AID-ADMA1551¿3.0.CO;2-V. http://doi.wiley.com/10.1002/
1521-4095%28200110%2913%3A20%3C1551%3A%3AAID-ADMA1551%3E3.0.CO%
3B2-V.
Wang, K.X., Yu, Z., Sandhu, S., Fan, S., 2013. Fundamental bounds on decay rates in asym-
metric single-mode optical resonators. Opt. Lett. 38 (2), 100. ISSN 0146-9592. https://
doi.org/10.1364/OL.38.000100. https://www.osapublishing.org/abstract.cfm?URI¼ol-
38-2-100.
Wu, W., Jiang, W., Yang, J., Gong, S., Ma, Y., 2017. Multilayered analog optical differen-
tiating device: performance analysis on structural parameters. Opt. Lett. 42 (24), 5270.
ISSN 0146-9592. https://doi.org/10.1364/OL.42.005270. https://www.osapublishing.
org/abstract.cfm?URI¼ol-42-24-5270.
Youssefi, A., Zangeneh-Nejad, F., Abdollahramezani, S., Khavasi, A., 2016. Analog
computing by Brewster effect. Opt. Lett. 41 (15), 3467. ISSN 0146-9592. https://
doi.org/10.1364/OL.41.003467. https://www.osapublishing.org/abstract.cfm?URI¼ol-
41-15-3467.
Yu, F.T.S., Jutamulia, S., Yin, S., 2001. Introduction to Information Optics. Academic Press,
San Diego. ISBN 0127748113.
Zhou, W., Zhao, D., Shuai, Y.C., Yang, H., Chuwongin, S., Chadha, A., Seo, J.H.,
Wang, K.X., Liu, V., Ma, Z., Fan, S., 2014. Progress in 2D photonic crystal Fano res-
onance photonics. Prog. Quantum Electron. 38 (1), 1–74. ISSN 00796727. https://doi.
org/10.1016/j.pquantelec.2014.01.001. http://linkinghub.elsevier.com/retrieve/pii/
S0079672714000020. https://doi.org/10.1016/j.pquantelec.2014.01.001.
Zhu, T., Zhou, Y., Lou, Y., Ye, H., Qiu, M., Ruan, Z., Fan, S., 2017. Plasmonic computing
of spatial differentiation. Nat. Commun. 8, 15391. ISSN 2041-1723. https://doi.org/
10.1038/ncomms15391. http://www.nature.com/doifinder/10.1038/ncomms15391.
https://doi.org/10.1038/ncomms15391.
CHAPTER FIVE
Contents
1. Introduction 116
1.1 A historical note 117
2. Figures of merit for sensing 118
2.1 Photonic factors 119
2.2 Electronic factors 119
2.3 Biochemical factors 120
3. Operation of GMR sensors 121
3.1 Optical signature 121
3.2 Polarization dependence 122
3.3 Q-factor and sensitivity 123
3.4 Bulk sensitivity 124
3.5 Surface sensitivity 125
4. Comparison between GMR sensors and other sensing modalities 128
4.1 Surface plasmon resonance 128
4.2 Nanohole arrays 129
4.3 Microrings 130
4.4 Bimodal waveguide 131
4.5 Concluding remarks 132
5. Resonant hyperspectral imaging 132
5.1 Sensing and imaging 135
5.2 Multiparameter imaging 137
5.3 Plasmonic nanohole arrays 137
6. Spatial-spectral equivalence 139
7. Phase-sensitive measurement 142
7.1 Realization of phase-sensitive measurements 143
8. Conclusion 145
References 146
1. Introduction
This chapter considers applications in the area of sensing. Since the
very nature of the resonance is less relevant in this context, we shall refer
to it more generally as a guided mode resonance, GMR, which applies
whether the Fano nature of the resonance is apparent or not.
The key property of a sensor is its ability to translate a quantity that is
difficult to measure into one that can be measured easily, so it acts as a trans-
ducer. In the case of GMR sensors, the fact that a GMR is a surface wave that
interacts closely with its environment is exploited, so GMR sensors fall into
the general class of surface affinity sensors. Surface affinity sensors are com-
monly used in biomedical applications and they detect target molecules
by their selective binding to a capture molecule such as an antibody, the anti-
body being anchored to the surface. The target molecule may be a marker
for cancer or some other disease. When the target binds to the antibody, the
optical mass on the surface increases and so the refractive index on the sur-
face increases. This local increase in refractive index changes the effective
index of the guided mode and thereby its resonance wavelength. By mea-
suring the change in wavelength, one can directly infer back to the change in
refractive index and the concentration of target molecules in the biological
medium, i.e., the analyte. Ultimately, a GMR sensor is therefore a refractive
index sensor and its performance is best described by the smallest refractive
index change it can detect; this principle is illustrated in Fig. 1, and the rel-
evant performance metrics will be discussed in Section 3.
A key point to note at the very beginning of this chapter is that GMR
sensors are not the most highly performing of all available sensors; other
types of sensors offer lower limits of detection. Nevertheless, as the chapter
develops, it will become apparent that GMRs offer the most versatile
approach toward sensing. They can be readily incorporated into low-cost
point-of-care devices because of their ease of coupling, they can perform
both the sensing and the imaging function, and they offer a multiparameter
approach via the wavelength, the amplitude, the Q-factor, and the phase of
the resonance. All of these effects will be discussed in detail in this chapter.
The driver behind much of the discussion is therefore not the quest for the
highest possible performance, but the desire for versatility, i.e., suitability for
a variety of applications, and the highest performance that can be built at the
lowest cost, which is much needed by healthcare providers both in the
developed and in the developing world.
Biosensing and imaging 117
Fig. 1 Sketch of the operation of a GMR surface affinity sensor, which translates the
binding of target molecules into a shift of the resonance wavelength.
Eq. (1) tells us that the diffraction angle θm ¼ 90 degrees occurs when the
wavelength λ equals the grating period g, which is the point at which the
sharp drop in intensity observed by Wood occurs. The refractive index n
118 Alexander Drayton et al.
being used. Many of these noise sources can be mitigated, e.g., by using
advanced algorithms or simply by incrementing the integration time, but
it is often difficult to quantify the overall impact of these factors. Several
authors have suggested that optical sensing methods are ultimately limited
by the photon shot noise, but in practice, all the other sources of noise tend
to dominate well before the shot noise limit is reached.
Because noise sources are variable and very much depend on the partic-
ular setup being used, including the stability of the laboratory equipment,
the FOM is typically not expressed as LOD ¼ k Q S, with k some pro-
portionality factor that accounts for the electronic noise factors. In fact, ear-
lier work by White and Fan (2008) attempted to define such a universal
FOM, but we note that their figure is not used in the field. Instead,
researchers tend to quote Q-factor, sensitivity and LOD separately.
As a general rule, high performance sensors achieve LOD values of
Δn 105–106. Putting these values into context with the refractive index
variation of water, which is of order Δn ¼ 104/°C (Bashkatov and Genina,
2003), we note that achieving an LOD of Δn ¼ 106 requires a temperature
stability of ΔT ¼ 102 degrees. Even though LODs of Δn 107–108 have
been reported, it is clear that such values can only be achieved with highly
stabilized laboratory equipment and by using liquid analytes that have been
preconditioned.
Fig. 2 Examples for the reflectance (A) and spatial distribution of the dominant electric
field component at resonance (B, C) for TE and TM modes, respectively. The simulated
structure is a Si3N4-grating (nSi3N4 ¼ 2) with a thickness of 150 nm, a period of 555 nm and
a filling factor of 80% on a SiO2 substrate (nSiO2 ¼ 1.45). Water is assumed to fill the grat-
ing grooves and the superstrate (nliquid ¼ 1.333) at normal incidence from the
substrate layer.
Biosensing and imaging 123
Fig. 3 (A) Resonance map for TM mode showing reflectance as a function of the inci-
dence wavelength and refractive index of the cover layer. Incidence is from the sub-
strate layer. (B) Linear range of sensitivity around the baseline index of water as
typically used in biosensing. (C,D) Field plots illustrating the mode confinement for
the extremes of lowest and highest refractive index of the cover layer.
Table 1 Parameter comparison between Si and Si3N4 gratings as displayed in Fig. 4 and
their resulting bulk and surface sensitivities
Grating Refractive Effective Bulk sensitivity Surface sensitivity
material index index (nm/RIU) (nm; n 5 1.4, t 5 10 nm)
Silicon 3.56 2.63 187 1.50
Silicon 2.00 1.75 176 0.87
nitride
Fig. 4 (A) Electric field profiles of TM modes at resonance for two different dielectric
materials and (B) the resulting GMR spectra. All parameters are the same as in Fig. 2
except for the material refractive index.
This difference is a direct consequence of Eq. (3); since the effective index in
the silicon structure is higher, the evanescent tail is shorter and the overlap
with the biolayer is larger.
An alternative way of highlighting the difference in surface sensitivity is
shown in Fig. 5, which is a plot of wavelength shift as a function of biolayer
thickness, similar to the study performed in Otte et al. (2009). Using the
same structures as in Fig. 4, the graph illustrates the higher surface sensitivity
of the higher index silicon grating, especially for a small biolayer thickness.
For a large thickness, the two curves converge, as is expected for the sim-
ilarity in bulk sensitivity.
128 Alexander Drayton et al.
Fig. 5 Simulated resonance wavelength shift for varying biolayer thicknesses assuming
a constant refractive index nbiolayer ¼ 1.4 on the sensing surface for the two gratings
defined in Fig. 4
Fig. 6 (A) Operation principle of a surface plasmon resonant sensor and (B) observation
of the characteristic dip upon coupling.
and they do match, the light does couple and the plasmon is excited. Since the
plasmon is a lossy mode, a lot of the light is dissipated before it couples back, so
a characteristic dip is observed in reflection (Fig. 6B). It is the traveling wave
nature of the surface plasmon that gives it its exquisite sensitivity, because the
smallest changes in effective index of the plasmon mode will lead to a reso-
nance shift (Yu and Fan, 2011). As a result, sensitivities of several
1000 nm/RIU have been reported (Pfeifer et al., 1999).
Because of the lossy nature of the plasmon mode, SPR systems have
very low quality factors, however, i.e., of the order of Q ¼ 10–20 which
is the lowest of the sensing modalities we explore here. The low Q factor
is offset by the high sensitivity, however, so the overall FOM S Q
104–105 nm/RIU which is comparable to most biosensing platforms includ-
ing the GMR. As a result, the typical implementation of SPR, such as in
Biacor’s SPR system, exhibits a LOD in the ng/mL range (Altintas et al.,
2011; Homola, 2008), making it useful for immunoassays, while it is not
the most sensitive of all available biosensors.
holes and couples back to a free-space mode on the other side (Genet and
Ebbesen, 2007). The conversion to plasmons explains how light can transmit
through tiny holes that are well below any classical cutoff (Liu and Lalanne,
2008; Martı́n-Moreno et al., 2001), which is why the effect is termed
“extraordinary” optical transmission.
The recognition that nanohole arrays can be used for sensing came later
with key demonstrations by Altug et al. who explored the resonance char-
acteristics and relevant sensing applications (Cetin et al., 2015).
The resonances they studied usually exhibit Q factors of the order of
Q 100 when working with liquids (Yanik et al., 2011) and sensitivities
of up to 650 nm/RIU which yields a FOM of S Q 104–105, which
is comparable albeit slightly higher than SPR and GMR structures. As a
result, nanoholes have exhibited impressive detection limits down to
500 pg/mL for the real-time molecular binding of IgG and 145 pg/mL
for vascular endothelial growth factor (VEGF) (Li et al., 2017).
These values demonstrate the strength of the nanohole array method as it
combines the benefits of GMR with its ease of coupling with the sensitivity
of SPR. The only downside is that the transmission is relatively low
(T 10%–20%) which results in weak signals. Since nanoholes only work
in transmission, it also means that the collection optics interferes with the
fluidic access, i.e., detection electronics and fluidics are on the same side
of the device. As stated previously, the fact that GMR and SPR can operate
in reflection means that the optics and the electronics can be placed on one
side of the device and the fluidics on the other, which is much more elegant
for practical implementations.
4.3 Microrings
Microrings are guided-mode devices where light propagates in the form of a
whispering gallery mode which is resonant when an integer number of
wavelengths fit into the circumference of the ring. The mode is strongly
confined to the waveguide (Fig. 7) and is a proper long-lived guided mode
(unlike the leaky mode of the GMR or the very lossy mode of the SPR), so
very high Q-factors can be achieved. In fact, Q-factors into the Q 108–109
have been achieved with microdisk resonators (Armani et al., 2003) but
these are not typically used for sensing as they are impractically high; as
explored by Hu et al. (2009), Q-factors above 105 are not useful in this con-
text because the wavelength noise of typical lasers then limits the accuracy of
determining the resonance. In terms of sensitivity, microrings tend to be
Biosensing and imaging 131
Fig. 7 Operation principle of microring resonator sensor with the top view also showing
the access waveguide.
worse than the other modalities, because the same strong confinement to the
waveguide that is needed for the high Q-factor tends to limit the overlap of
the mode with the analyte, such that sensitivities of S 50–80 nm/RIU are
typically being achieved (Bogaerts et al., 2012; Juan-Colás et al., 2017b).
Nevertheless, due to the high Q-factor, the overall FOM tends to be quite
high, i.e., S Q 105–106, which is higher than the other modalities
explored so far.
the interference pattern represents a direct measure for the relative phase
change between the two modes (Dante et al., 2015). Bimodal waveguides
are an elegant solution because they are less complex than conventional inter-
ferometers and they have demonstrated detection limits of 2.5 107 RIU
(Herranz et al., 2017).
Fig. 12 Schematic of combining both imaging and sensing on the same GMR platform.
Cells attached to the surface change the resonance wavelength of the hyperspectral
image from blue to red, as in Fig. 11. When suitably stimulated, they secrete protein bio-
markers that are detected via complementary antibodies immobilized on the surface;
on binding, the resulting phase change is detected via a change in the hyperspectral
image from blue to yellow.
Fig. 13 Comparison of resonant imaging vs phase contrast imaging. All three images
show a baby hamster kidney (BHK) cell secreting thrombopoietin (TPO). Panel
(A) represents the resonant image at the beginning of the experiment, when only
the cell is present (t ¼ 0 min). Panel (B) shows a resonant image combining the cell
and the secreted proteins at t ¼ 85 min. Panel (C) shows a phase contrast micrograph
of the cell for reference, which remains unchanged in the presence of the proteins. It
is clear that the resonant image can see both the cell and the proteins it secrets while
the phase contrast image can only see the cell.
A Common cell-matrix
≡
Increase in
adhesion components refractive index
Increase in refractive
Focal adhesions (FAs) ≡
index and scattering
B Δλ (PWS)
1
Reflectance (AU)
BG Sh Δ/ (PIS)
1 ift
Cell 0.9
0.8
Reflectance (AU)
0.8
0.6
624 626 628
0.4
Wavelength (nm)
0.2
0
610 620 630 640 650
Wavelength (nm)
Fig. 14 Illustration of the separate effects of a cell on the optical resonance; while the
cell body increases the effective index of the resonant mode and shifts the resonance
compared to the background, the additional scattering at the edges of the cell respon-
sible for adhesion (focal adhesions) can be separately identified via a reduction in the
amplitude of the resonance peak. (A) Illustration of the scattering at focal adhesions
due to their rapid change in density; (B) Corresponding experimental evidence
for the refractive index change and additional scattering loss. PWS refers to peak
wavelength shift and PIS to peak intensity shift. Panel (A) taken from Juan-Colás,
J. & Krauss, T. F., 2018. Multiparameter resonant imaging for studying cell interactions.
Light Sci. Appl., 7. Panel (B) taken from fig. 1c of Zhuo, Y. et al., 2018. Quantitative anal-
ysis of focal adhesion dynamics using photonic resonator outcoupler microscopy (PROM).
Light Sci. Appl., 7(1), 9.
Biosensing and imaging 139
nanohole approach has demonstrated imaging with high sensitivity and high
spatial resolution. Plasmonic nanohole arrays fabricated in optically thick
metal films and they support the EOT phenomenon, which is attributed
to the excitation of plasmons by grating coupling of light at normal incidence;
these plasmons result in a strong field enhancement in the vicinity of the
nanoholes. This strong light confinement enables nanohole arrays to be highly
sensitive to minute local refractive index changes, such as those induced by
the binding of biomolecules to the sensor surface. In addition, the localization
of light around the nanohole provides high spatial resolution. A good example
of this capability is provided by Li et al. (2017) where the authors demonstrate
high resolution imaging (spatial resolution not quantified) and high sensitivity
sensing of VEGF, a signaling protein, directly in complex cell media. The
VEGF was secreted from cancer cells attached to the sensor surface, providing
a similar functionality as the resonant image shown in Fig. 13.
Nevertheless, the nanohole approach also has its downsides. Intrinsically,
even though the transmission is considered “extraordinary,” it is relatively
low in absolute terms, i.e., around 10%, which limits the signal to noise ratio.
In terms of practicality, EOT being observed in transmission also means that
the fluidics needs to be integrated into the optical system, i.e., source and
detector are on opposite sides of the microfluidic circuitry. In contrast, the
GMR, preferentially working in reflection, allows a clean separation between
the optical system on one side of the setup and the fluidics on the other;
working in reflection therefore mimics the inverted microscope arrangement
favored by biologists. In terms of substrate, the nanohole array tends to be
fabricated on free-standing silicon nitride membranes (Li et al., 2017), which
are fragile and expensive and therefore not suited for low-cost point-of-care
applications. Clearly, when identifying the most suitable approach for a given
application, it is important to consider all aspects ( Juan-Colás et al., 2017a;
Pitruzzello and Krauss, 2018) (Table 2).
6. Spatial-spectral equivalence
It is intrinsic to a resonance-based optical sensor that its response is
qualified as a function of wavelength; in fact, most of the figures shown
in this chapter use wavelength as the parameter to describe a particular aspect
of sensor performance. While it is easy to perform spectral characterization
in a research laboratory, doing so in the field is more difficult and it means
that a spectrometer has to be included into the system in order to provide the
readout. Spectrometers are not cheap; despite the efforts of a number of
Table 2 Comparison of different sensing modalities
Bulk sensitivity
Structure Analyte Q factor (nm/RIU) FOM (nm/RIU) LOD References
SPR (Biacore) CEA antigen 20 10 3
2 10 4
3 ng/mL Altintas et al. (2011)
GMR Different 85 510 4.3 10 4
— Huang et al. (2009)
Solutions
GMR Streptavi-din 150 88 1.3 104 1 ng/mL Cunningham et al. (2002)
1
Microrings MCP-1 2 10 4
— — 0.5 pg mL Valera et al. (2016)
and Washburn et al. (2009)a
Bmodal waveguide HCL Solutions — 6 102 πrad/RIU — 2.5 107 RIU Herranz et al. (2017)
Bimodal Waveguide hGH — 6 102 πrad/RIUb — 10 pg mL1 González-Guerrero et al.
(2017)
Nanobeam Cavity CEA antigen 1 104 70 7105 0.1 pg mL1 Liang et al. (2013)
1
Nanohole Array IgG <100 700 7 10 4
145 pg mL Li et al. (2017)
a
Multiple references used as specific details on Genalyte Maverick™ systems are not openly available.
b
Assumed to be the same as other work(s) from the same research group due to the same structure being used.
Biosensing and imaging 141
Fig. 15 Illustration of the chirped GMR concept which translates spectral dimension
into spatial dimension and combines the sensing and the spectral readout function
in a single structure. (A) Design of grating with spatially varying period in one direction.
(B) Hyperspectral image illustrating the continuous change of the resonance wave-
length as a color code. (C) Illumination with a single wavelength source yields a single
line of resonance.
142 Alexander Drayton et al.
7. Phase-sensitive measurement
All the modalities described so far record the amplitude of the light
interacting with the biological material, i.e., the cells and proteins. We
record the amplitude of the transmission or reflection, the amplitude of
the resonance shift or the Q-factor, which is the spectral width of the
amplitude response. Nevertheless, amplitude describes only one aspect
of the optical interaction and arguably, most advanced photonic devices
make use of the phase for advanced functionality.
Recording phase requires an interferometric approach, which can
indeed yield excellent results. For example, some of the most sensitive
photonic biosensors are based on interferometry, such as the bimodal
waveguide (Section 4.3), also (Schmitt et al., 2007) which exploits the
interference between two different modes in a waveguide, or the Farfield
approach (Cross et al., 2003). These approaches can achieve exquisite
LODs of Δn < 108 RIU. Is it possible to also benefit from the phase
approach in the context of GMR?
In a way, the guided mode resonance is already a result of the in-phase
interaction between multiply scattered waves, but it is still susceptible to
amplitude fluctuations, so is ultimately limited by amplitude noise. Instead,
a truly phase-sensitive measurement is limited only by phase noise.
Phase noise tends to be significantly lower than amplitude noise for common
light sources used in biosensing, including low-cost VCSELs. The phase
noise of lasers is inversely proportional to their coherence length lc:
Δϕ λ
¼ (5)
ϕ lc
Biosensing and imaging 143
Fig. 16 Reflectance and phase response of GMR grating proposed in Sahoo et al. (2017).
Biosensing and imaging 145
8. Conclusion
Having now reviewed the various properties of guided mode reso-
nances for sensing applications, one can conclude that GMR arguably pro-
vides the most versatile of all optical sensing approaches. It affords high
sensitivity, albeit not the ultrahigh sensitivity achieved by other sensor
modalities, while also offering a range of other functionalities, such as imag-
ing and multiparameter sensing which are much more difficult or not pos-
sible to achieve with other modalities. Similarly, the idea of mapping the
spectral response onto the spatial domain as exemplified by the chirped
GMR approach is a unique feature that allows combining the sensor element
with the wavelength-selective readout element in the same structure; most
other optical sensing modalities require an external spectrometer. Together
with the ease of coupling to the GMR directly with a collimated beam, the
versatility and inherent simplicity make the GMR very suitable for high per-
formance, yet low-cost diagnostic devices that may be deployed in the field
and that are much needed by healthcare providers both in the developed and
in the developing world.
146 Alexander Drayton et al.
References
Altintas, Z., Uludag, Y., Gurbuz, Y., Tothill, I.E., 2011. Surface plasmon resonance based
immunosensor for the detection of the cancer biomarker carcinoembryonic antigen.
Talanta 86, 377–383. 30 10.
Armani, D.K., Kippenberg, T.J., Spillane, S.M., Vahala, K.J., 2003. Ultra-high-Q toroid
microcavity on a chip. Nature 421 (6926), 925.
Armbruster, D.A., Pry, T., 2008. Limit of blank, limit of detection and limit of quantitation.
Clin. Biochem. Rev. 29 (Suppl. 1), S49.
Barrios, C.A., 2012. Integrated microring resonator sensor arrays for labs-on-chips. Anal.
Bioanal. Chem. 403 (6), 1467–1475.
Bashkatov, A.N., Genina, E.A., 2003. Water refractive index in dependence on temperature
and wavelength: a simple approximation. Optical Technologies in Biophysics and Med-
icine IV. vol. 5068. pp. 393–396.
Block, I.D., Ganesh, N., Lu, M., Cunningham, B.T., 2008. A sensitivity model for predicting
photonic crystal biosensor performance. IEEE Sensors J. 8 (3), 274–280.
Block, I.D., et al., 2009. Optimizing the spatial resolution of photonic crystal label-free imag-
ing. Appl. Opt. 48 (34), 6567–6574.
Bogaerts, W., et al., 2012. Silicon microring resonators. Laser Photonics Rev. 6 (1), 47–73.
Cetin, A.E., et al., 2015. Plasmonic nanohole arrays on a robust hybrid substrate for highly
sensitive label-free biosensing. ACS Photonics 2 (8), 1167–1174.
Chen, W., et al., 2013. Photonic crystal enhanced microscopy for imaging of live cell adhe-
sion. Analyst 138 (20), 5886–5894.
Cross, G.H., et al., 2003. The metrics of surface adsorbed small molecules on the Young’s
fringe dual-slab waveguide interferometer. J. Phys. D. Appl. Phys. 37 (1), 74.
Cunningham, B., Li, P., Lin, B., Pepper, J., 2002. Colorimetric resonant reflection as a direct
biochemical assay technique. Sensors Actuators B Chem. 81 (2–3), 316–328.
Dante, S., et al., 2015. Linear readout of integrated interferometric biosensors using a peri-
odic wavelength modulation. Laser Photonics Rev. 9 (2), 248–255.
Dudzik, G., Rzepka, J., 2010. Single frequency stable VCSEL as a compact source for inter-
ferometry and vibrometry. AIP Conference Proceedings. vol. 1253 (1), pp. 317–320.
Fischer, M., Triggs, G.J., Krauss, T.F., 2016. Optical sensing of microbial life on surfaces.
Appl. Environ. Microbiol. 82 (5), 1362–1371.
Genet, C., Ebbesen, T.W., 2007. Light in tiny holes. Nature 7123, 39.
González-Guerrero, A.B., et al., 2017. Direct and label-free detection of the human growth
hormone in urine by an ultrasensitive bimodal waveguide biosensor. J. Biophotonics
10 (1), 61–67.
Herranz, S., Gavela, A.F., Lechuga, L.M., 2017. Label-free biosensors based on bimodal
waveguide (BiMW) interferometers. In: Methods in Molecular Biology. Springer,
pp. 161–185.
Hessel, A., Oliner, A.A., 1965. A new theory of Wood’s anomalies on optical gratings. Appl.
Opt. 4 (10), 1275–1297.
Homola, J., 2008. Surface plasmon resonance sensors for detection of chemical and biological
species. Chem. Rev. 108 (2), 462–493.
Hu, J., Sun, X., Agarwal, A., Kimerling, L.C., 2009. Design guidelines for optical resonator
biochemical sensors. J. Opt. Soc. Am. B 25 (5), 1032–1041.
Biosensing and imaging 147
Huang, M., Yanik, A.A., Chang, T.-Y., Altug, H., 2009. Sub-wavelength nanofluidics in
photonic crystal sensors. Opt. Express 17 (26), 24224–24233.
Juan-Colás, J., Krauss, T.F., 2018. Multiparameter resonant imaging for studying cell inter-
actions. Light Sci. Appl. 7, 45.
Juan-Colás, J., Johnson, S., Krauss, T.F., 2017a. Dual-mode electro-optical techniques for
biosensing applications: a review. Sensors (Switzerland) 17 (9), 1424–8220.
Juan-Colás, J., Krauss, T.F., Johnson, S.D., 2017b. Real-time analysis of molecular confor-
mation using silicon electrophotonic biosensors. ACS Photonics 4 (9), 2320–2326.
Juan-Colas, J., et al., 2018. Quantifying single cell secretion in real time using resonant hyper-
spectral imaging. Proc. Natl. Acad. Sci. U. S. A. 115 (52), 13204–13209.
Kabashin, A.V., Patskovsky, S., Grigorenko, A.N., 2009. Phase and amplitude sensitivities in
surface plasmon resonance bio and chemical sensing. Opt. Express 17 (23),
21191–21204.
Lalanne, P., Hugonin, J.P., Chavel, P., 2006. Optical properties of deep lamellar gratings: a
coupled Bloch-mode insight. J. Lightwave Technol. 24 (6), 2442–2449.
Li, X., et al., 2017. Plasmonic nanohole array biosensor for label-free and real-time analysis of
live cell secretion. Lab Chip 17 (13), 2208–2217.
Liang, F., et al., 2013. Scalable photonic crystal chips for high sensitivity protein detection.
Opt. Express 21 (26), 32306–32312.
Liedberg, B., Nylander, C., Lunstr€ om, I., 1983. Surface plasmon resonance for gas detection
and biosensing. Sensors Actuators 4, 299–304.
Liu, H., Lalanne, P., 2008. Microscopic theory of the extraordinary optical transmission.
Nature 452 (7188), 728.
Magnusson, R., Wang, S.S., 1992. New principle for optical filters. Appl. Phys. Lett. 61 (9),
1022–1024.
Magnusson, R., Lee, K.J., Wawro, D., 2004. Guided- mode resonance biosensors employing
phase detection. In: Diffractive Optics and Micro-Optics (p. DTuC2). Optical Society
of America.
Magnusson, R., Wawro, D., Zimmerman, S., Ding, Y., 2011. Resonant photonic biosensors
with polarization-based multiparametric discrimination in each channel. Sensors 11 (2),
1476–1488.
Martens, D., et al., 2018. Low-cost integrated biosensing platform based on SiN
nanophotonics for biomarker detection in urine. Anal. Methods 10, 3066–3073.
Martı́n-Moreno, L., et al., 2001. Theory of extraordinary optical transmission through sub-
wavelength hole arrays. Phys. Rev. Lett. 86 (6), 1114.
Maystre, D., 2012. Theory of Wood’s anomalies. In: Plasmonics. Springer, pp. 39–83.
Nazirizadeh, Y., et al., 2013. Material-based three-dimensional imaging with nanostructured
surfaces. Appl. Phys. Lett. 102 (1), 11116.
Otte, M.A., et al., 2009. Identification of the optimal spectral region for plasmonic and
nanoplasmonic sensing. ACS Nano 4 (1), 349–357.
Pfeifer, P., et al., 1999. Real time sensing of specific molecular binding using surface plasmon
resonance spectroscopy. Sensors Actuators B Chem. 54 (1–2), 166–175.
Pitruzzello, G., Krauss, T.F., 2018. Photonic crystal resonances for sensing and imaging.
J. Opt. 20 (7), 73004.
Pruijmboom, A., Schemmann, M., Hellmig, J., Schutte, J., Moench, H., Pankert, J., 2008.
VCSEL-based miniature laser-Doppler interferometer. In: Vertical-Cavity Surface-
Emitting Lasers XII. vol. 6908. International Society for Optics and Photonics, p. 69080I.
Rayleigh, L., 1907. III. Note on the remarkable case of diffraction spectra described by Prof.
Wood. In: The London, Edinburgh, and Dublin Philosophical Magazine and Journal of
Science. vol. 14. Taylor & Francis, pp. 60–65. 79.
Rosenblatt, D., Sharon, A., Friesem, A.A., 1997. Resonant grating waveguide structures.
IEEE J. Quantum Electron. 33 (11), 2038–2059.
148 Alexander Drayton et al.
Sahoo, P.K., Sarkar, S., Joseph, J., 2017. High sensitivity guided-mode-resonance optical
sensor employing phase detection. Sci. Rep. 7 (1), 7607.
Schmitt, K., et al., 2007. Interferometric biosensor based on planar optical waveguide sensor
chips for label-free detection of surface bound bioreactions. Biosens. Bioelectron.
22 (11), 2591–2597.
Su, Y.-D., Chen, S.-J., Yeh, T.-L., 2005. Common-path phase-shift interferometry surface
plasmon resonance imaging system. Opt. Lett. 30 (12), 1488–1490.
Thorlabs Inc., 2019a. HeNe Tutorial. [Online] Available at: https://www.thorlabs.
com/tutorials.cfm?tabID¼9F3258E0-FB6B-46D4-944D-B31A84EF1DAA [Accessed
23 11 2018].
Thorlabs Inc., 2019b. HeNe Lasers: Red. [Online] Available at: https://www.thorlabs.com/
newgrouppage9.cfm?objectgroup_id¼1516 [Accessed 23 11 2018].
Thorlabs Inc., 2019c. VCSEL Diode Module (CPS850V). [Online] Available at: https://
www.thorlabs.com/thorproduct.cfm?partnumber¼CPS850V.
Tittl, A., et al., 2018. Imaging-based molecular barcoding with pixelated dielectric meta-
surfaces. Science 360 (6393), 1105–1109.
Triggs, G.J., et al., 2015. Spatial resolution and refractive index contrast of resonant photonic
crystal surfaces for biosensing. IEEE Photonics J. 7 (3), 6801810.
Triggs, G.J., et al., 2017. Chirped guided-mode resonance biosensor. Optica 4 (2), 229.
Valera, E., Shia, W.W., Bailey, R.C., 2016. Development and validation of an immuno-
sensor for monocyte chemotactic protein 1 using a silicon photonic microring resonator
biosensing platform. Clin. Biochem. 49 (1), 121–126.
Wang, S.S., Magnusson, R., Bagby, J.S., Moharam, M.G., 1990. Guided-mode resonances
in planar dielectric-layer diffraction gratings. J. Opt. Soc. Am. A 7 (8), 1470–1474.
Washburn, A.L., Gunn, L.C., Bailey, R.C., 2009. Label-free quantitation of a cancer bio-
marker in complex media using silicon photonic microring resonators. Anal. Chem.
81 (22), 9499–9506.
White, I.M., Fan, X., 2008. On the performance quantification of resonant refractive index
sensors. Opt. Express 16 (2), 1020.
Wood, R.W., 1902. XLII. On a remarkable case of uneven distribution of light in a diffrac-
tion grating spectrum. In: The London, Edinburgh, and Dublin Philosophical Magazine
and Journal of Science. vol 4. Taylor & Francis, pp. 396–402. (21).
Yanik, A.A., et al., 2011. Seeing protein monolayers with naked eye through plasmonic Fano
resonances. Proc. Natl. Acad. Sci. U. S. A. 108 (29), 11784–11789.
Yariv, A., 1991. Optical Electronics. Saunders College Publishings.
Young, G., et al., 2018. Quantitative mass imaging of single biological macromolecules.
Science 360 (6387), 423–427.
Yu, Z., Fan, S., 2011. Extraordinarily high spectral sensitivity in refractive index sensors using
multiple optical modes. Opt. Express 19 (11), 10029–10040.
Zhuo, Y., et al., 2017. Label-free imaging of stem cell adhesion and dynamic tracking of
boundary evolution using photonic crystal enhanced microscopy (PCEM). Microsc.
Microanal. 23 (S1), 1142–1143.
Zhuo, Y., et al., 2018. Quantitative analysis of focal adhesion dynamics using photonic res-
onator outcoupler microscopy (PROM). Light Sci. Appl. 7 (1), 9.
CHAPTER SIX
Contents
1. Fano resonance photonic crystal filters 149
1.1 Single-layer filters 149
1.2 Coupled double-layer filters 152
1.3 Fano resonance filter configurations 157
2. Fano resonance photonic crystal cavity modulators 168
3. Fano resonance photonic crystal cavity optomechanics 177
3.1 Optical forces in nanoscale cavities 177
3.2 Optical forces in coupled Fano resonance PCSs 178
3.3 Optomechanical control and MOMES applications 184
Acknowledgments 186
References 186
A B C
1
Transmission
0.5
0
–10 –5 0 5 10 –10 –5 0 5 10 –10 –5 0 5 10
(w –w 0) (w –w 0) (w –w 0)
Fig. 1 Single-pole filter lineshapes. ω0 and γ ¼ 1/τ are the center frequency and width
of the resonance, respectively. (A) A bandpass filter constructed by a resonance
imposed on a total reflection background. (B) A notch filter constructed by a reso-
nance imposed on a zero reflection background. (C) A Fano filter constructed by a
resonance imposed on a partial reflection background. Reproduced with permission
from Zhou, W., Zhao, D., Shuai, Y.-C., Yang, H., Chuwongin, S., Chadha, A., Seo, J.-H.,
Wang, K. X., Liu, V., Ma, Z., 2014. Progress in 2D photonic crystal Fano resonance
photonics. Prog. Quantum Electron. 38 (1), 1–74.
A B
1.0 1.0
even–odd even–even
Transmission
0.8 Δw = 6g 0.8 Δw = 6g
0.6 0.6
0.4 0.4
0.2 0.2
0.0
–4 –2 0 2 4 0.0
–4 –2 0 2 4
1.0 1.0
even–even
even–odd
Transmission
0.8 Δw = 2g 0.8 Δw = 2g
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
–4 –2 0 2 4
–4 –2 0 2 4
1.0 1.0
even–even
even–odd
Δw = g
Transmission
0.8 Δw = g 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
–4 –2 0 2 4 –4 –2 0 2 4
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
–4 –2 0 2 4 –4 –2 0 2 4
(w – w 0)/g (w – w 0)/g
Fig. 2 Double-pole filter lineshapes resulting from two resonances. ω0 and γ ¼ 1/τ are
the center frequency and width of the resonances, respectively. Δω is the frequency
difference between the two resonances. (A) Transmission spectra through two orthog-
onal, even and odd, resonances. When Δω ≫ γ, the system behaves in essence as two
separate Lorentzian filters. At Δω ¼ 2γ, the structure is a flattop filter. When Δω ≪ γ, the
structure is an all-pass filter. (B) Transmission spectra through two nonorthogonal, both
even, resonances. The structure exhibits electromagnetically induced transparency (EIT)
lineshapes. Reproduced with permission from Suh, W., Wang, Z., Fan, S., 2004. Temporal
coupled-mode theory and the presence of non-orthogonal modes in lossless multimode
cavities. IEEE J. Quantum Electron. 40 (10), 1511–1518.
Because the resonances are orthogonal, similar to Eq. (1), the complex
amplitude transmission coefficient can be written as
γ even γ odd
t ¼1 (3)
iω iωeven + γ even iω iωodd + γ odd
152 Weidong Zhou and Shanhui Fan
Fig. 3 (A) Photonic crystal slab structure consisting of a square lattice of air holes
introduced into a high-index dielectric slab. The slab is 2.05a, thick and the radius of
the holes is 0.12a, where a is the lattice constant. The arrow represents the direction
of the incident light. (B) Schematic of a theoretical model for a resonator system that
supports two resonant modes with opposite symmetry with respect to the mirror plane
perpendicular to the incident light. Reproduced with permission from Suh, W., Fan, S.,
2004. All-pass transmission or flattop reflection filters using a single photonic crystal slab.
Appl. Phys. Lett. 84, 4905.
When γ even ¼ γ odd ¼ γ, ωeven ¼ ωodd ¼ ω0, Eq. (3) can be simplified to:
iðω ω0 Þ γ
t¼ (4)
iðω ω0 Þ + γ
and the photonic crystal slab becomes an all-pass filter. The intensity trans-
mission is uniform for both on and off resonances. However, the phase delay
varies from 0 to 2π in the vicinity of the resonance. This can be realized in
the structure if we choose a dielectric constant of 10.07, as shown in Fig. 3.
The all-pass transmission and the phase delay spectra are shown in Fig. 4.
The theoretical derivation (Eq. 4) is verified by FDTD simulations (Suh
and Fan, 2004).
When γ even ¼ γ odd ¼ γ, j ωeven ωodd j ¼ 2γ, Eq. (3) becomes
ðω ω0 Þ4
T ¼ jtj2 ¼ (5)
ðω ω0 Þ4 + 4γ 4
where ω0 ¼ (ωeven + ωodd)/2 and flattop reflection is achieved in the vicinity
of the resonance. This can also be realized in the structure shown in Fig. 3, if
we choose a dielectric constant of 10.9. In Fig. 5, we plot the transmission
spectrum. The theoretical derivation (Eq. 5) is again verified against FDTD
simulations (Suh and Fan, 2004).
Fig. 4 Spectral response functions (intensity transmission and group delay) for the one-
slab structure in Fig. 3 with a dialectic constant of 10.07. Reproduced with permission
from Zhou, W., Zhao, D., Shuai, Y.-C., Yang, H., Chuwongin, S., Chadha, A., Seo, J.-H.,
Wang, K. X., Liu, V. & Ma, Z., 2014. Progress in 2D photonic crystal Fano resonance photon-
ics. Prog. Quantum Electron. 38 (1), 1–74.
Fig. 5 Intensity transmission for the one-slab structure in Fig. 3 with a dialectic constant
of 10.9. Reproduced with permission from Zhou, W., Zhao, D., Shuai, Y.-C., Yang, H.,
Chuwongin, S., Chadha, A., Seo, J.-H., Wang, K. X., Liu, V. & Ma, Z., 2014. Progress in 2D
photonic crystal Fano resonance photonics. Prog. Quantum Electron. 38 (1), 1–74.
where the top signs are used when both modes are even, and the bottom
signs are used when both modes are odd. In Fig. 2B, we assume that both
modes are even, and that ω1 ¼ ω0 Δω/2, ω2 ¼ ω0 + Δω/2, and γ 1 ¼ γ 2 ¼ γ.
The transmission behavior in Fig. 2B resembles that of electromagnetically
induced transparency (EIT). At the resonant frequencies ω1 and ω2, the
transmission is zero regardless of the spacing between the two resonant
frequencies; while at ω0, the transmission is at unity. This configuration
was demonstrated using coupled PCSs (Suh and Fan, 2003).
In coupled PCSs, there exists extra opportunity to engineer dark reso-
nances (Liu et al., 2009). A dark resonance is a guided resonance decoupled
from particular free space radiation modes due to symmetry mismatch.
We also refer to externally excited guided resonances as bright resonances.
A dark resonance in coupled PCSs can arise from the coupling of dark
resonances or bright resonances in individual slabs.
For two PCSs that each possess a dark resonance, bringing them close to
each other allows their resonances to couple. If both slabs are aligned such
that the proper symmetries are preserved, these coupled dark resonances
remain dark. If we shift the relative position of these two slabs to break
the symmetry, these coupled guided resonances are externally excited and
become bright. This will result in two Fano resonances in close proximity,
and the quality factors depend strongly on the lateral displacement. This type
of dark resonance is illustrated in Fig. 6.
It is also possible to couple two bright resonances together to make them
dark. For a pair of PCSs each with a bright resonance, we can decouple them
from radiation modes by placing a Fabry-Perot resonance between the two
slabs. At a frequency where the reflectivity of each individual slab is at unity,
the bright resonances lose access to external radiation. This type of dark res-
onance is illustrated in Fig. 7.
Fano resonance photonic crystal filters and modulators 155
A B
1
Transmission
0
0.3 0.4 0.5 0.6 0.7
w (2pc/a)
C
1
D
Transmission
0
0.3 0.4 0.5 0.6 0.7
w (2pc/a)
E
1
F
Transmission
0
0.3 0.4 0.5 0.6 0.7
w (2pc/a)
Fig. 6 Dark states in a two photonic crystal slab system that arise from coupled dark
resonances from single slabs. (A) Single-slab transmission spectrum. (B) Mode profile
at the dark state frequency (top) and field profile at the frequency indicted by the arrow
in (A). (C) Double-slab transmission spectrum for d ¼ 0.5a, where d is the spacing and a is
the lattice constant. (D) Field profile at frequency indicated by the arrow in (C).
(E) Double-slab transmission spectrum for Δx ¼ 0.15a, where Δx is the lateral displace-
ment. (F) Field profile at lowest frequency resonance from (E). Reproduced with permis-
sion from Liu, V., Povinelli, M., Fan, S., 2009. Resonance-enhanced optical forces between
coupled photonic crystal slabs. Opt. Express 17 (24), 21897–21909.
156 Weidong Zhou and Shanhui Fan
A B
1
Transmission
0
0.3 0.4 0.5 0.6 0.7
w (2pc/a)
C
1
Transmission
0
0.3 0.4 0.5 0.6 0.7
w (2pc/a)
E
1
Transmission
0
0.3 0.4 0.5 0.6 0.7
w (2pc/a)
Fig. 7 Dark states in a two photonic crystal slab system that arise from coupled bright
resonances from single slabs. (A) Single-slab transmission spectrum. (B) Mode profile at
the dark state frequency (top) and field profile at the frequency indicted by the arrow in
(A). (C) Double-slab transmission spectrum for d ¼ 0.5a, where d is the spacing and a is
the lattice constant. (D) Field profile at frequency indicated by the arrow in (C).
(E) Double slab transmission spectrum for d ¼ 0.65a. (F) Field profile at peak of reso-
nance in (E). Reproduced with permission from Liu, V., Povinelli, M., Fan, S., 2009.
Resonance-enhanced optical forces between coupled photonic crystal slabs. Opt. Express
17 (24), 21897–21909.
Fano resonance photonic crystal filters and modulators 157
Fig. 8 Schematics of Fano resonance PCS filter configurations: (A) Single-layer PCS on
glass (top) or flexible (bottom) substrates; (B) double-PCS Fano filter 3D sketch; (C) Key
parameters defined for the square lattice double-layer PCS filter including lattice con-
stant (a), air hole radius (r), thin oxide buffer layer thickness (tb), and two single layer PCS
layer thicknesses (t1 and t2); (D) 3D close-up view of double-PCS Fano filters with lattice
displacement; and (E) Illustration and definition of key design parameters, including lat-
tice displacement (Δx, Δy) between two layers.
158 Weidong Zhou and Shanhui Fan
thicknesses for the top (t1) and bottom Si PhC layers (t2), and the oxide
buffer layer in between (tb).
Yang et al. (Qiang et al., 2008) reported the first experimental surface-
normal Fano filters based on transferred silicon nanomembranes on glass
and on flexible PET substrates. Subsequently, with improved designs
and processes, single-layer PCS Fano filters have been demonstrated with
much improved performance. Shown in Fig. 9A and B are scanning
electron microscope (SEM) images of the fabricated single-layer Fano
filters on glass substrate. The measured (blue solid line) and simulated
(red dashed line) transmission spectra are shown in Fig. 9C, with enlarged
images shown in Fig. 9D. Two transmission dips were found at 1529.88
and 1564.62 nm. For the 1564.62 nm dip, the Q value was 1737, with a
26 dB extinction ratio.
A B
C D
Transmission power (dBm)
–25 –25
Transmission power (dBm)
–30 –30
–35 –35
–40 –40
–45 –45
–50 –50
Measured Measured
–55 –55
Simulated Simulated
1520 1530 1540 1550 1560 1570 1580 1560 1562 1564 1566 1566
Wavelength (nm) Wavelength (nm)
Fig. 9 Experimental results for Design S1: (A) Top and (B) cross-section views of fabri-
cated single-layer PhC Fano resonance filters on oxide buffer; (C) Measured (blue solid
line) and simulated (red dash line) transmission spectra for the fabricated single-layer
PhC Fano resonance filter transferred on glass substrates; and (D) Enlargement of
(C) over the second dip (λ ¼ 1564.62 nm) region. Reproduced with permission from
Shuai, Y., Zhao, D., Tian, Z., Seo, J., Plant, D. V., Ma, Z., Fan, S., Zhou, W., 2013b. Double-
layer Fano resonance photonic crystal filters. Opt. Express 21 (21), 24582–24589.
Fano resonance photonic crystal filters and modulators 159
The angle and polarization properties of these filters have been investi-
gated in detail by Qiang et al.(Qiang et al., 2008) and Li et al. (Chen et al.,
2009), focusing on measurement and simulated dispersion properties.
Detailed analysis of the angle-dependent Fano resonant mode dispersion
properties is presented in Fig. 10. Following the conventional definition,
the light incident angles are specified by two polar angles, the colatitude
angle θ (an angle from the surface-normal direction) and the azimuth
angle ϕ (an angle from the positive x-axis to the orthogonal projection of
the incident beam in the x–y plane) (Lousse et al., 2004). The incident
beam with the wave-vector k was kept within the Γ–X direction in the
k-space, i.e., for the case shown in Fig. 10A with ϕ ¼ 0°, the momentum
A B
q = 0° 10° 20° 30° 40° 1400 1
Normalized frequency (a/l)
0.42
Wavelength (nm)
1450 0.8
1500 0.6
0.40
1550 0.4
0.38 0.2
Experiment 1600
Simulation
1650 0
0.36 0 5 10 15 20
0 0.08 0.16 0.24 0.32
Incident angle (degree)
Kx (2p /a, G → X)
C D
q = 40° 30° 20° 10° 0° 1400 1
Normalized frequency (a/l)
0.42
Wavelength (nm)
1450 0.8
1500 0.6
0.40
1550 0.4
0.38
Experiment 1600 0.2
Simulation
0.36 1650 0
20 15 10 5 0
1.40 1.46 1.52 1.58 1.65 1.71
Incident angle (degree)
Kx (2p /a, M → G )
Fig. 10 (A, C) Simulated dispersion plot (red lines) zoomed-in along (Γ–X, Γ–M) direc-
tions, along with measured Fano resonances (blue dots) for different incident angles θ
with (A) ϕ ¼ 0° and Ψ ¼ 0°, and (C) ϕ ¼ 45° and Ψ ¼ 45°. The straight gray lines represent
the relationships between the normalized frequency and the normalized wave-vectors
along (Γ–X, Γ–M) directions. (B, D) Measured transmission intensity contour plots for
angle-dependent transmissions at different incident angles θ with (B) ϕ ¼ 0° and
Ψ ¼ 0°, and (D) ϕ ¼ 45° and Ψ ¼ 45°. Reproduced with permission from Qiang, Z.,
Yang, H., Chen, L., Pang, H., Ma, Z., Zhou, W., 2008. Fano filters based on transferred silicon
nanomembranes on plastic substrates. Appl. Phys. Lett. 93, 061106.
160 Weidong Zhou and Shanhui Fan
vector kx relates to the normalized momentum kΓ–X along the Γ–X direction
as kx ¼ kΓx 2πa cosϕ. Thus, the normalized frequency (a/λ) can be related
to the normalized momentum kΓ–X in the dispersion plot as kΓx ¼ λa sinθ.
A set of straight lines can be superimposed on the normalized dispersion
plot along the Γ–X direction, as shown in Fig. 10A. The measured Fano
resonant modes were normalized and plotted in Fig. 10A, blue circles.
Excellent agreement was achieved between the experimental and simulation
results. The transmission intensity contour plots for the measured angle-
dependent transmission spectra, with ϕ ¼ 0°, are also shown in Fig. 10B,
where the intensity dips/peaks are shown in blue and red. The intensity con-
tour plot clearly resembles the spectra shifting behavior shown in Fig. 10A.
Following a similar procedure, the measured transmission results shown
in Fig. 10B, with ϕ ¼ 45°, were analyzed such that wave vector k lies in
the Γ–M plane, with results shown in Fig. 10C and D. Again, the experi-
mental and simulation results agree very well. It was confirmed that
angle-independent filters are feasible based on the dispersion engineering
of the Fano resonances in PCS. Details can be found in a short review by
Zhou et al. (2009).
A B
100 100
Transmission (%)
Transmission (%)
80 80
60 60
40 40
20 20
S3 S3 Q = 4500
0 0
1350 1450 1550 1650 1570 1580 1590 1600
Wavelength (nm) Wavelength (nm)
C D
100 100
Transmission (%)
Transmission (%)
80 80
60 60
40 40
20 20
D3 D3 Q = 98,000
0 0
1450 1500 1550 1600 1650 1566.1 1566.6 1567.1 1567.6
Wavelength (nm) Wavelength (nm)
Fig. 11 Simulated transmission spectra for (A, B) single- and (C, D) double-layer Fano
resonance PhC filters, where (B) and (D) are enlarged plots of (A) and (C), respectively.
The design parameters are summarized in Table 1 for Case S3 and D3, respectively.
Reproduced with permission from Shuai, Y., Zhao, D., Tian, Z., Seo, J., Plant, D. V.,
Ma, Z., Fan, S., Zhou, W., 2013b. Double-layer Fano resonance photonic crystal filters.
Opt. Express 21 (21), 24582–24589.
162 Weidong Zhou and Shanhui Fan
A B
7
tb = 0 nm Q ~ 5K 1600 l 10
8
Q ~ 22K Q
Resonant l (nm)
20 nm
1570 106
Transmission
Q ~ 110K 40 nm
6
Q ~ 1e7
Q factor
60 nm
1540
4
Q ~ 125K 80 nm 105
Q ~ 37K 100 nm 1510
Q ~ 21K 120 nm
2 104
Q ~ 15K 140 nm 1480
Q ~ 12K 160 nm
0 1450 103
1450 1500 1550 1600 1650 0 40 80 120 160
Wavelength (nm) Buffer thickness (nm)
C D
1
tb = 20 nm
0.8 l = 1559 nm
Transmission
0.6
tb = 60 nm
0.4 l = 1515 nm
0.2
tb = 60 nm Q = 107 tb = 100 nm
0
1515.277 1515.279 l = 1489 nm
Wavelength (nm)
Fig. 12 Simulation results for Design D2 with different buffer layer thicknesses tb:
(A) Transmission spectra with different tb from 0 to 160 nm; (B) High Q resonant wave-
lengths and the corresponding Q values for different buffer thicknesses tb; (C) Zoom-in
spectrum for the buffer thickness tb ¼ 60 nm and filter Q of 10,000,000; and
(D) Simulated E-field intensity profile at resonant wavelengths for three different tb
values, where tb ¼ 60 nm representing the highest Q condition for this design.
Reproduced with permission from Shuai, Y., Zhao, D., Tian, Z., Seo, J., Plant, D. V.,
Ma, Z., Fan, S., Zhou, W., 2013b. Double-layer Fano resonance photonic crystal filters.
Opt. Express 21 (21), 24582–24589.
Fano resonance photonic crystal filters and modulators 163
SOI, a low index oxide buffer layer was first formed using thermal oxidation
of a single-crystalline Si layer on the SOI substrate, followed by a low pres-
sure chemical vapor deposition (LPCVD) poly-Si deposition process, to
form a poly-Si/thermal SiO2/crystalline-Si double-Si-layer structure. The
other double-layer PhC Fano filters on quartz substrates were formed by
two rounds of the LPCVD poly-Si deposition process, with a plasma-
enhanced chemical vapor deposited (PECVD) SiO2 layer sandwiched in
between the two LPCVD poly-Si layers. Shown in Figs. 13A and 5C are
cross-sectional SEM images for the double-layer poly-Si/SiO2/c-Si filter
structures, with different thermal oxide thicknesses. The same E-beam pat-
terning and etching processes were utilized for the 2D-PhC patterning. An
SEM image is shown in Fig. 13D.
Fig. 13 Cross-sectional SEM images for fabricated double-layer PhC Fano resonance
filters based on Design D2 parameters: (A–C) Double-layer PhC structure was formed
by poly-Si deposition on top of the SOI substrates; and (D) Double-layer PhC structure
was formed by two steps of poly-Si deposition on quartz substrates. Notice the oxide
buffer thicknesses are 20, 160, and 20 nm for cases (B), (C), and (D), respectively.
Reproduced with permission from Shuai, Y., Zhao, D., Tian, Z., Seo, J., Plant, D. V.,
Ma, Z., Fan, S., Zhou, W., 2013b. Double-layer Fano resonance photonic crystal filters.
Opt. Express 21 (21), 24582–24589.
164 Weidong Zhou and Shanhui Fan
–10 –10
–11 –11
Measured Measured
–12 –12 On quartz
Simulated On quartz Simulated
1530 1540 1550 1560 1543.7 1544.7 1545.7 1546.7
Wavelength (nm) Wavelength (nm)
C D
–12 –12
Measured
Reflection power (dBm)
On SOI
Reflection power (dBm)
Transmission (%)
80 80
60 60 100
100
80
80
40 60
40 60
40
40
20 T 20 20
T
20
0
0 1559.3764 1576.3765
1559 1559.5 15560
0 0
1450 1500 1550 1600 1650 1450 1500 1550 1600 1650
(A) Wavelength (nm) (B) Wavelength (nm)
80
80
60
R 80 60
Reflection (%)
Reflection (%)
40
R
40
20 20
60 0
60
0
1559 1559.5 15560 1576.3764 1576.3765
40 40
20 20 y = 90°
0 y = 0°
0
1450 1500 1550 1600 1650 1450 1500 1550 1600 1650
(C) Wavelength (nm) (D) Wavelength (nm)
Fig. 15 Simulated transmission and reflection spectra for double-layer stacked
Fano filters with (B, D) and without (A, C) lattice offset (Δx ¼ 0.2a, Δy ¼ 0.5a), with
the insets showing enlarged plots for the dominant (high Qs) resonance for each cases.
Also shown in the insets are the schematics of the structures considered in the sim-
ulation. For the transmission spectra shown in (A, B), glass substrate was considered.
For the reflection spectra shown in (C, D), SOI substrate was considered. Ψ represents
the incident beam polarization (E-vector) defined as the angle from the positive x-axis
to the polarization direction. Reproduced with permission from Shuai, Y., Zhao, D.,
Chadha, A. S., Seo, J.-H., Yang, H., Fan, S., Ma, Z., Zhou, W., 2013a. Coupled double-layer
Fano resonance photonic crystal filters with lattice-displacement. Appl. Phys. Lett. 103
(24), 241106.
Fig. 16 Fabricated Fano filter scanning electron micrograph (SEM) images: (A, B) Top
view and (C) cross-sectional view of double-layer Fano filter on SOI, with controlled lat-
tice misalignment offset (Δx ¼ 0.2a, Δy ¼ 0). Reproduced with permission from Shuai, Y.,
Zhao, D., Chadha, A. S., Seo, J.-H., Yang, H., Fan, S., Ma, Z., Zhou, W., 2013a. Coupled double-
layer Fano resonance photonic crystal filters with lattice-displacement. Appl. Phys. Lett. 103
(24), 241106.
Fig. 17 are the results obtained for the coupled double-layer PCSs, with large
lattice displacement (Δx ¼ 0.495a, Δy ¼ 0.45a). A measured reflection spec-
trum is shown in Fig. 17A, along with a top SEM image shown in the inset.
We obtained two resonant locations, around 1548.6 and 1567 nm, with a
magnified plot around 1548.6nm resonant shown in Fig. 17B. Again, simu-
lation and Fano fit were carried out and the results match the experimental
results well. In this case, we obtained a Q factor of 80,000. This value is the
highest experimental Q factor reported so far, for all the different cases we fab-
ricated. While the value is still far from the optimal Q factor of 2.11 108, the
results demonstrate the great potential of coupled double-layer PCS structures
with controlled lattice displacement. During testing, we also noticed that the
performance of these filters is very sensitive to incident beam direction, colli-
mation, and polarization orientation. We anticipate much higher Q-factors
can be obtained with precisely controlled displacement and an optimized
fabrication process.
168 Weidong Zhou and Shanhui Fan
A B
0
–4
Y = 0° Y = 0°
–2
Reflection (dB)
Reflection (dB)
–6
–4
–8
–6
–10 –8 Measured
1 mm Δx = 0.495a Simulated
–12 Δy = 0.45a Fano fitted
–10
1540 1550 1560 1570 1548 1548.5 1549
Wavelength (nm) Wavelength (nm)
Fig. 17 Measured displaced Fano filter (Δx ¼ 0.495a, Δy ¼ 0.45a) reflection spectra for
p-polarization state, with theoretical and experimental Q-factors of 115,000 and 72,000,
respectively. Shown in the inset is the magnification of the measured reflection spec-
trum. Reproduced with permission from Shuai, Y., Zhao, D., Chadha, A. S., Seo, J.-H.,
Yang, H., Fan, S., Ma, Z., Zhou, W., 2013a. Coupled double-layer Fano resonance photonic
crystal filters with lattice-displacement. Appl. Phys. Lett. 103 (24), 241106.
A
R
T
X1 (Center) X2 (Edge)
C
B d
1 Cathode
60 nm N+ = 1 × 1019 cm–3 X
a 2r y2
170 nm N = 1 × 10 17
cm –3
y1
y1
tSiO2 P = 1 × 10 cm
17 –3
y2 170 nm
Ṽ 60 nm P+ = 1 × 1019 cm–3
–1 Anode Anode
D
Y
The SiO2 layer provides electrical isolation between the doped layers, so that
the device functions electrically as a parallel-plate capacitor. When a poten-
tial difference is imposed between the slabs, the accompanying charge accu-
mulation changes their refractive indices, modifying the optical frequency of
the high-Q mode. For a fixed optical frequency, the transmission and reflec-
tion are both affected, with slope efficiency proportional to the optical qual-
ity factor. We modeled the electrical response of our device in 2D (X–Y)
using Taurus Medici (Synopsys) (Synopsys, 2016), a semiconductor device
simulator. In the simulations, the doping level was taken to 1019 cm3 at the
60 nm outer regions of the silicon slabs and 1017 cm3 at the 170 nm inner
regions (Fig. 18C). The device width was taken to be D ¼ 20 μm limited by
170 Weidong Zhou and Shanhui Fan
the allowed number of grid nodes, and the electrode width to be d ¼ 5 μm, as
illustrated in Fig. 18C. The nonuniformity in the electric charge distribution
owing to the etched holes was ignored.
Optimized device geometry and Si/SiO2/Si layers’ thicknesses are deter-
mined by systematic investigation and analysis considering many factors in
our device model. In the proposed Si/SiO2/Si PIN structure, different sand-
wiched SiO2 thicknesses and Si doping levels are compared by combination
of optical and electrical analysis. Sandwiched SiO2 thickness of 20 nm was
chosen to make a balance between optical and electrical effects. Multiple fac-
tors were considered such as: its desired optical Q value, stacked Si/SiO2/Si
layers RC effect, free carrier absorption in Si, effective refractive index
changes, sufficient electrical isolation between Si layers as well as process
control preference.
In our optical study, we started from a transmission configuration as in
Fig. 19A. Different SiO2 thicknesses offer Q-factors varying from 5000 to
15,000 as shown in Fig. 19B. Two SiO2 thickness cases are compared in
Fig. 19C and D when applying forward bias, the 10 nm oxide case shows
larger spectral shifting than 20 nm oxide case; however, strong free carrier
absorption starts to occur with decreased oxide thickness, resulting in
reduced Q factor and extinction ratio. To avoid unintended free carrier
absorption and possible thermal effect, 20 nm oxide thickness is determined
for later reflection configuration design. It was noticed that, in Fig. 19E, this
interfaces accumulation (as “y1” positions, in Fig. 18C) produces a larger
blue shifting on spectra for the resonance mode when comparing to free car-
rier accumulation appearing in the middle of each Si layer (as “y2” positions,
in Fig. 18C). Owing to the semiconductor carriers equilibrium in these two
stacked Si PCS, our proposed P/N doping structure (Fig. 18C) can be mod-
ulated by either forward or reverse bias for blue or red spectral shift, respec-
tively, as shown in Fig. 19F. By applying the carrier refraction (Soref and
Bennett, 1987) and semiconductor carriers equilibrium np ¼ ni2, one could
imagine there would be a returning point of red-shifting with increased
reverse bias voltage and this returning point was observed during device
characterization.
A reflection configuration was further investigated by optical simulation.
In Fig. 20A, a 2D sketch shows the device geometry design for two kinds of
device sizes. It shows our electrical bias convention. We applied 1 μm buffer
oxide layer and 380 nm PECVD oxide for effective reflection spectra as well
as for electrical interconnection purposes. Simulation results show a dra-
matic spectral shifting in Fig. 20B when applying electrical bias. It reveals
A B C
Transmission configuration 100
tSiO2 = 20 nm Q ~ 15 K
Transmission (%)
4 80
tSiO2 = 15 nm Q ~ 15 K
Transmission
Si
60
SiO2 tSiO2 = 10 nm Q ~ 12 K
Q ~ 30 K
Si 2 40 tSiO2 = 10 nm
tSiO2 = 5 nm Q ~ 10 K
Q ~ 15 K 0V
20 +5 V
SiO2 substrate tSiO2 = 0 nm Q ~ 10 K Q~5 K
+10 V
0 0
1450 1500 1550 1600 1650 1572 1573 1574 1575
Wavelength (nm) Wavelength (nm)
D E F
100 100 100
Transmission (%)
Transmission (%)
60 0V 60 60 0V
+5 V +5 V
40 +10 V 40 Δn<0 40 –5 V
0V
(y1) y1
20 20 20
Δn<0 y2
Δn=0
(y2)
0 0 0
1558 1559 1560 1561 1558 1559 1560 1561 1558 1559 1560 1561
Wavelength (nm) Wavelength (nm) Wavelength (nm)
Fig. 19 Device electro-optical (EO) study in transmission configuration: a 2D sketch of the simulation lattice is shown in (A); for lattice constant
“a” ¼ 1 μm and “r/a” ¼ 0.08, the transmission spectral centered at 1550 nm are shown in (B) with different sandwiched SiO2 thicknesses varying
from 0 to 20 nm. Two different sandwiched SiO2 thicknesses, 10 nm in (C) and 20 nm in (D), are also compared for EO spectral shifting; Spectral
shifting by free carrier accumulation occurring at either middle or interfaces of Si layers is compared in (E), by 5 V forward bias. Spectral shifting
is also compared by 5 V forward/reverse bias, as shown in (F). Reproduced with permission from Shuai, Y.-C., Zhao, D., Liu, Y., Stambaugh, C.,
Lawall, J., Zhou, W., 2017. Coupled bilayer photonic crystal slab electro-optic spatial light modulators. IEEE Photonics J. 9 (2), 1–11.
172 Weidong Zhou and Shanhui Fan
A
~
350 or 80 µm V
300 or 50 µm
poly-Si n
380 nm
PECVD p
c-
oxide
Si
Buffer oxide
Silicon substrate
B
60
0V Q ~ 15 K
50 +5 V
+10 V
Reflection (%)
40
–5 V
30
20
10
0
1560 1561 1562
Wavelength (nm)
Fig. 20 (A) A sketch of device geometry cross-sectional view shows device design with
two kinds of lateral dimension (350 μm mesa with 300 μm PC; 80 μm mesa with 50 μm
PC) that are investigated in experiments. It has the same sandwiched SiO2 thickness, “r,”
“a,” Si layer thickness and doping profiles as those applied in transmission configuration
(Fig. 19D and F). The simulated reflection spectra with different biases are studied in
(B) with correspondence to transmission configuration. Reproduced with permission from
Shuai, Y.-C., Zhao, D., Liu, Y., Stambaugh, C., Lawall, J., Zhou, W., 2017. Coupled bilayer
photonic crystal slab electro-optic spatial light modulators. IEEE Photonics J. 9 (2), 1–11.
consumed a 30 nm layer of silicon from the surface, leaving 230 nm. The
thermal grown SiO2 layer was subsequently thinned down to 20 nm by
means of a HF wet etching. A polycrystalline silicon layer of 230 nm thick-
ness was deposited atop the thermal oxide layer by means of low-pressure
chemical vapor deposition. The stacked bilayers were subsequently
boron/phosphorous (p/n) doped by ion implantation into different depths,
followed by rapid thermal activation. The array of holes for the photonic
crystals was then patterned by electron-beam lithography on a 60 nm Cr
layer that is used as a hard mask. The stacked poly-Si/SiO2/Si layers were
etched through to the bottom buffer oxide layer by reactive ion etching,
after which the Cr mask was chemically removed. Finally, a 380 nm SiO2
layer was deposited on top by PECVD for interconnection.
Fig. 21A shows a microscope image of a 6 6 array of devices fabricated
with 350 μm 350 μm mesa area. The optical resonance mode reflection
measurement for an individual 350 μm 350 μm mesa area device shows
a good match to the simulation as shown in Fig. 21B. The measured optical
power for each wavelength was collected after passing through two crossed
polarizers to improve the extinction ratio (Nazirizadeh et al., 2008). Its mea-
sured reflection resonance peak wavelengths were recorded for applied elec-
trical biases varying from 10 to +10 V. Spectral blue-shifting and red-
shifting were found when applying forward and reverse bias, respectively.
A maximum red-shifting returning point was observed at (5 1) V as
shown in Fig. 21C. This spectral shifting is induced by overall free carrier
concentrations variation in two Si layers when electrical charges were
applied onto this capacitor like structure. Both spectral blue- and red-shift
by forward and reverse bias were obtained in the p/n-type stacking silicon
layers.
A 6 6 array of 80 μm 80 μm mesa devices was also fabricated for high
frequency EO modulation (Fig. 22). It was attached to a dielectric substrate
with a coplanar waveguide (CPW) line that was designed for 10 GHz oper-
ation. A single 80 μm 80 μm mesa area device was then wire-bonded. The
top and bottom electrodes were connected to 150 μm ground/signal pads for
wire-bonding. A scanning electron microscope (SEM) top view of a device
region with 80 μm 80 μm mesa area and a zoomed-in SEM image of the
50 μm 50 μm e-beam patterned array of holes are shown in Fig. 22B
and C.
While a modulator such as the one described here can in principle func-
tion in either transmission or reflection, we focus on the reflection mode of
operation in this work. A schematic diagram of the setup used for tests is
174 Weidong Zhou and Shanhui Fan
A B
70 3.5
–10 (dBm)
Simulation
60 Measurement –11 3
–12
50 2.5
Reflection (%)
Power (μW)
1520 1550
40 2
30 1.5
20 1
10 0.5
0 0
1500 1550 1600
Wavelength (nm)
C
1553.95
Wavelength (nm)
1553.9
1553.85
Red-shifting
Return point
1553.8
1553.75
–10 –8 –6 –4 –2 0 2 4 6 8 10
Voltage (V)
Fig. 21 A microscope top-view of a 6 6 array of 350 μm 350 μm mesa area devices is
shown in (A). Optical resonance mode reflection spectral for a single device is measured
in (B) with crossed-polarizers method to improve the extinction ratio (Nazirizadeh et al.,
2008). As shown in (C), continuous resonance mode spectral peak shifting was observed
by applying electrical bias from 10 to +10 V. The measurement points in blue circles
were connected by line, which shows a red-shifting return point between 4 and
6 V of reverse bias. Reproduced with permission from Shuai, Y.-C., Zhao, D., Liu, Y.,
Stambaugh, C., Lawall, J., Zhou, W., 2017. Coupled bilayer photonic crystal slab electro-
optic spatial light modulators. IEEE Photonics J. 9 (2), 1–11.
diverted by the circulator for detection, or with its fast axis at 45° to the inci-
dent polarization, so that light falling on the device is circularly polarized and
the reflected light is deflected by the PBS and subsequently detected.
We supplemented the spectral shifting measurements with frequency-
domain EO modulation measurements using a network analyzer. RF ampli-
fier and a high-speed receiver were also used for device EO modulation fre-
quency domain characterization. The predicted optical resonance was found
near 1497.5 nm for an 80 μm 80 μm mesa area device. For subsequent fre-
quency domain EO modulation tests, the laser wavelength was fixed at
1499 nm where the slope of the reflection spectrum takes its highest value.
Fig. 23B shows the measured frequency response of the particular
80 μm 80 μm mesa device, where 0 dB corresponds to the static (low-
frequency) response. As shown in the figure, the response falls off by 3 dB
at a frequency of (200 20) MHz.
In summary, we have experimentally demonstrated a novel electro-optic
intensity modulator exploiting a Fano resonance in a bilayer photonic crystal
architecture. Both blue and red spectral shifts in the resonance were
observed, by applying opposite electrical biases. The 3 dB bandwidth of
one of our devices, with a lateral size of 80 μm 80 μm, was measured to
be (200 20) MHz. These spatial light modulators are polarization indepen-
dent owing to highly symmetric 2D square lattice photonic crystal structure.
Calculations suggest that bandwidths in the GHz range can be achievable if
176 Weidong Zhou and Shanhui Fan
Fig. 23 A characterization setup shown in (A) was used for device optical reflection
spectral collection and electro-optical (EO) modulation characterization. A network
analyzer, RF amplifier, and a high-speed receiver were used for device EO modulation
frequency domain characterization. As shown in (B), a 200 MHz 3 dB cutoff fre-
quency, which is about 100 times of a 350 μm 350 μm mesa area device’s perfor-
mance, was found for the device with 80 μm 80 μm mesa area. Reproduced with
permission from Shuai, Y.-C., Zhao, D., Liu, Y., Stambaugh, C., Lawall, J., Zhou, W.,
2017. Coupled bilayer photonic crystal slab electro-optic spatial light modulators. IEEE
Photonics J. 9 (2), 1–11.
Fano resonance photonic crystal filters and modulators 177
the device size is reduced. High modulation depth should be possible as well,
by using a higher optical Q factor PCS stacking structure (Shuai et al., 2013a,
2013b) with process-controlled PCS hole size. These bilayer coupled PCS-
based modulators offer novel opportunities in vertical photonic integration,
including free-space communications, 3D optical interconnect, and spatial
control of light via extended horizontal arrays.
intrinsic resonance linewidth with low optical power (Li et al., 2009).
Micrometer-scale displacements of the waveguide are observed for
milliwatt-level optical input powers in a structure with a high-Q optical
microresonator coupled with a micrometer-scale waveguide (Roels et al.,
2009). Seamless wavelength routing over a range of 3000 times the intrinsic
channel width was achieved in a specially designed nanooptomechanical sys-
tem (Wiederhecker et al., 2009). Eichenfield et al. proposed the idea of
“optomechanical crystals,” i.e., that in a photonic crystal, the periodicity
of the host medium is used to manipulate the properties of light, whereas
a phononic crystal uses periodicity to manipulate mechanical vibrations
(Eichenfield et al., 2007). A picogram- and nanometer-scale in a
photonic-crystal optomechanical cavity was measured in a pair of specially
patterned nanoscale beams, resulting in a scale of the per-photon force and
the mass of the structure (Eichenfield et al., 2009).
All of these results indicate that there is potential for radiation pressure
actuated devices to be used in a variety of photonic applications, such as
channel routing/switching, buffering, dispersion compensation, pulse trap-
ping/release, widely tunable lasers, and other light signal manipulations on
chip (Ma and Povinelli, 2012).
y x
a
Δx
Fig. 24 Schematic of the double-slab system. The red arrows indicate the directions of
the incident light. Each slab consists of an array of high index rods (ε ¼ 12) surrounded
by air (ε ¼ 1). The empty rectangle indicates the surface over which the integration of the
stress tensor is performed. In the right panel, the bottom slab is shifted relative to the
first slab by a distance of Δx. Reproduced with permission from Liu, V., Povinelli, M., Fan, S.,
2009. Resonance-enhanced optical forces between coupled photonic crystal slabs. Opt.
Express 17 (24), 21897–21909.
d¥
0.635
w
Resonance frequency FWHM
0.63 w¥
0.625
0.62
0.615
0.6 0.65 0.7 0.75 0.8
Slab spacing (d/a)
Fig. 25 Resonance peak and linewidth in the vicinity of the dark state arising from
coupled bright resonances as in Fig. 7 (Case 1). Reproduced with permission from
Liu, V., Povinelli, M., Fan, S., 2009. Resonance-enhanced optical forces between coupled
photonic crystal slabs. Opt. Express 17 (24), 21897–21909.
106 107
102 ´ Fz (w = 0.630)
105 Fz 106
Q
104 105
Force (l0 /c)
103 104
Q
102 103
10 102
0.584
Resonance frequency
0.582
w
FWHM
0.58
0.578
106 107
Q
105 Fz 106
Fx
104 105
Force (l0/c)
103 104
Q
102 103
10 102
other by 0.2a, where a is the lattice constant. A dark state still exists, as seen in
Fig. 29. As the slab separation is changed, both the normal force, Fz, and the
lateral force, Fx, vary in the same way as Q, inversely proportional to the
square of the separation (Fig. 30).
In conclusion, in a coupled photonic crystal slab system, the optical
forces are strongly influenced by the symmetries of the system as well as
its resonance properties. It is worth noting that these proposed structures
182 Weidong Zhou and Shanhui Fan
d∞
0.64
w
FWHM
Resonance frequency 0.635
w∞
0.63
0.625
0.62
0.56 0.6 0.64 0.68 0.72
Slab spacing (d/a)
Fig. 29 Resonance peak and linewidth in the vicinity of the dark state for Case 3 at
Δx ¼ 0.2a. Reproduced with permission from Liu, V., Povinelli, M., Fan, S., 2009.
Resonance-enhanced optical forces between coupled photonic crystal slabs. Opt. Express
17 (24), 21897–21909.
105 106
Q
104 Fx 105
Fz
103 104
Force (l0/c)
102 103
Q
10 102
1 10
(Notomi et al., 2006), Notomi et al. (Roh et al., 2010) reported an exper-
imental demonstration of strong optomechanical interactions in InP-based
defect-free bilayer PCS cavities, based on bandedge or Fano resonance
modes. Significant radiation force of the stored light (F) should appear as
the electromagnetic energy stored in the cavity (U) changes with the change
in spatial separations (z) between two coupled PCS slabs, as shown in the
following equation:
dU U dω PQ dω
F ¼ ¼ ¼
dz ω dz ω2 dz
Fig. 31 (A) Schematics of the experiment. The structure is fully clamped (all-side
clamped); (B) Top-view scanning electron microscope image; and (C) Bird’s-eye view
(upper) and cross section (lower) of a bilayer PCS. The period is 750 nm and the hole
radius at the center is 270 nm. Reproduced with permission from Roh, Y.-G., Tanabe, T.,
Shinya, A., Taniyama, H., Kuramochi, E., Matsuo, S., Sato, T., Notomi, M., 2010. Strong
optomechanical interaction in a bilayer photonic crystal. Phys. Rev. B 81 (12), 121101.
184 Weidong Zhou and Shanhui Fan
A B
a 2R
h1
s
h2
a
Fig. 32 (A, B) Schematic and (C) SEM image of single-membrane (asymmetric) structure:
a photonic-crystal (holey) membrane consisting of a square-lattice of air holes on silicon
is suspended (separation s) on top of an unpatterned (homogeneous) silicon slab (thick-
ness h2 ¼ 0.2a) sitting on top of a semiinfinite silica substrate. Light is apparent on the
membrane from the normal direction (top). Key parameters are: a ¼ 920 nm, R ¼ 207 nm,
h1 ¼ h2 ¼ 185 nm, and s ¼ 265 nm. Reproduced with permissions from Rodriguez, A. W.,
McCauley, A. P., Hui, P.-C., Woolf, D., Iwase, E., Capasso, F., Loncar, M., Johnson, S. G.,
2011. Bonding, antibonding and tunable optical forces in asymmetric membranes. Opt.
Express 19 (3), 2225–2241; Hui, P.-C., Woolf, D., Iwase, E., Sohn, Y.-I., Ramos, D.,
Khan, M., Rodriguez, A. W., Johnson, S. G., Capasso, F., Loncar, M., 2013. Optical bistability
with a repulsive optical force in coupled silicon photonic crystal membranes. Appl. Phys.
Lett. 103 (2), 021102 (021102-5).
such as adaptive filtering of laser lines and optical clock recovery). MEMS’s are
the promising mechanisms for introducing variable functions into PC devices.
Although several devices with variable structures of PC slabs combined with
MEMS were considered, there have been very few reports on the experimen-
tal studies because of the difficulty in fabricating hybrid structures (Hui et al.,
2013; Rodriguez et al., 2011). A few experimental results, as well as principle
design, have been reported (Lu and Lee, 2009; Stomeo et al., 2010). In order
to introduce variable function to the Fano resonances in PC slabs, Suh et al.
(2003) studied two PC slabs facing each other analytically for a wavelength-
tunable PC filter. Kanamori et al. (2007) experimentally demonstrated a
wavelength-selective variable reflection filter controlled by changing the
gap with a microelectromechanical actuator (Fig. 33).
Fig. 33 (A) Cross-section diagram of the MEMS tunable Fano filter design and (B) SEM
image of the fabricated Fano filter. The inset shows the magnified view of the fabricated
PC slab. Reproduced with permission from Kanamori, Y., Kitani, T., Hane, K., 2007. Control of
guided resonance in a photonic crystal slab using microelectromechanical actuators. Appl.
Phys. Lett. 90, 031911.
186 Weidong Zhou and Shanhui Fan
Acknowledgments
The authors acknowledge contributions from their current and former students, as well as
collaborators involved in the research work reviewed here. We also thank funding
agencies and program directors for their instrumental support from US Air Force Office
of Scientific Research (AFOSR, Dr. Gernot Pomrenke), US Army Research Office
(ARO, Dr. Michael Gerhold), US National Science Foundation (NSF, Drs. John Zavada
and Dominique Dagenais). We also thank the fabrication support from the University of
Texas at Arlington Nanotechnology Research Center, and the University of Texas at
Austin, part of NSF NNIN.
References
Chen, L., Qiang, Z., Yang, H., Pang, H., Ma, Z., Zhou, W.D., 2009. Polarization and angu-
lar dependent transmissions on transferred nanomembrane Fano filters. Opt. Express
17 (10), 8396–8406.
Chmielak, B., Waldow, M., Matheisen, C., Ripperda, C., Bolten, J., Wahlbrink, T.,
Nagel, M., Merget, F., Kurz, H., 2011. Pockels effect based fully integrated, strained
silicon electro-optic modulator. Opt. Express 19 (18), 17212–17219.
Crozier, K.B., Lousse, V., Kilic, O., Kim, S., Fan, S., Solgaard, O., 2006. Air-bridged pho-
tonic crystal slabs at visible and near-infrared wavelengths. Phys. Rev. B 73 (11), 115126.
Eichenfield, M., Michael, C.P., Perahia, R., Painter, O., 2007. Actuation of micro-
optomechanical systems via cavity-enhanced optical dipole forces. Nat. Photonics
1 (7), 416–422.
Eichenfield, M., Camacho, R., Chan, J., Vahala, K.J., Painter, O., 2009. A picogram- and
nanometre-scale photonic-crystal optomechanical cavity. Nature 459 (7246), 550–555.
Fan, S., Joannopoulos, J.D., 2002. Analysis of guided resonances in photonic crystal slabs.
Phys. Rev. B 65 (23), 235112.
Hui, P.-C., Woolf, D., Iwase, E., Sohn, Y.-I., Ramos, D., Khan, M., Rodriguez, A.W.,
Johnson, S.G., Capasso, F., Loncar, M., 2013. Optical bistability with a repulsive optical
force in coupled silicon photonic crystal membranes. Appl. Phys. Lett. 103 (2), 021102,
021102-5.
Jacobsen, R.S., Andersen, K.N., Borel, P.I., Fage-Pedersen, J., Frandsen, L.H., Hansen, O.,
Kristensen, M., Lavrinenko, A.V., Moulin, G., Ou, H., 2006. Strained silicon as a new
electro-optic material. Nature 441 (7090), 199–202.
Kanamori, Y., Kitani, T., Hane, K., 2007. Control of guided resonance in a photonic crystal
slab using microelectromechanical actuators. Appl. Phys. Lett. 90, 031911.
Kanskar, M., Paddon, P., Pacradouni, V., Morin, R., Busch, A., Young, J.F., Johnson, S.R.,
MacKenzie, J., Tiedje, T., 1997. Observation of leaky slab modes in an air-bridged semi-
conductor waveguide with a two-dimensional photonic lattice. Appl. Phys. Lett. 70 (11),
1438–1440.
Kilic, O., Digonnet, M., Kino, G., Solgaard, O., 2007. External fibre Fabry-Perot acoustic
sensor based on a photonic-crystal mirror. Meas. Sci. Technol. 18 (10), 3049–3054.
Kippenberg, T., Vahala, K., 2008. Cavity optomechanics: back-action at the mesoscale.
Science 321 (5893), 1172–1176.
Li, M., Pernice, W., Xiong, C., Baehr-Jones, T., Hochberg, M., Tang, H., 2008. Harnessing
optical forces in integrated photonic circuits. Nature 456 (7221), 480–484.
Li, M., Pernice, W., Tang, H., 2009. Tunable bipolar optical interactions between guided
lightwaves. Nat. Photonics 3 (8), 464–468.
Lin, C., Lu, Z., Shi, S., Jin, G., Prather, D.W., 2005. Experimentally demonstrated filters
based on guided resonance of photonic-crystal films. Appl. Phys. Lett. 87, 091102.
Fano resonance photonic crystal filters and modulators 187
Liu, V., Fan, S., 2012. S 4: a free electromagnetic solver for layered periodic structures. Com-
put. Phys. Commun. 183 (10), 2233–2244.
Liu, V., Povinelli, M., Fan, S., 2009. Resonance-enhanced optical forces between coupled
photonic crystal slabs. Opt. Express 17 (24), 21897–21909.
Lousse, V., Suh, W., Kilic, O., Kim, S., Solgaard, O., Fan, S.H., 2004. Angular and polar-
ization properties of a photonic crystal slab mirror. Opt. Express 12 (8), 1575–1582.
Lu, T.-W., Lee, P.-T., 2009. Ultra-high sensitivity optical stress sensor based on double-
layered photonic crystal microcavity. Opt. Express 17 (3), 1518–1526.
Ma, J., Povinelli, M.L., 2012. Applications of optomechanical effects for on-chip manipu-
lation of light signals. Curr. Opin. Solid State Mater. Sci. 16 (2), 82–90.
McKerracher, I., Fu, L., Tan, H., Jagadish, C., 2013. Integration of bandpass guided-mode
resonance filters with mid-wavelength infrared photodetectors. J. Phys. D Appl. Phys.
46 (9), 095104.
Nazirizadeh, Y., M€ uller, J.G., Geyer, U., Schelle, D., Kley, E.-B., T€ unnermann, A.,
Lemmer, U., Gerken, M., 2008. Optical characterization of photonic crystal slabs using
orthogonally oriented polarization filters. Opt. Express 16 (10), 7153–7160.
Notomi, M., Taniyama, H., Mitsugi, S., Kuramochi, E., 2006. Optomechanical wavelength
and energy conversion in high-Q double-layer cavities of photonic crystal slabs. Phys.
Rev. Lett. 97 (2), 023903.
Nxcnons, E., Huu, G., 1901. A preliminary communication on the pressure of heat and light
radiation. Phys. Rev. 13, 307.
Peng, S., Morris, G., 1996. Resonant scattering from two-dimensional gratings. J. Opt. Soc.
Am. A 13 (5), 993–1005.
Povinelli, M., Johnson, S., Lonèar, M., Ibanescu, M., Smythe, E., Capasso, F.,
Joannopoulos, J., 2005a. High-Q enhancement of attractive and repulsive optical forces
between coupled whispering-gallery-mode resonators. Opt. Express 13 (20),
8286–8295.
Povinelli, M.L., Loncar, M., Ibanescu, M., Smythe, E.J., Johnson, S.G., Capasso, F.,
Joannopoulos, J.D., 2005b. Evanescent-wave bonding between optical waveguides.
Opt. Lett. 30 (22), 3042–3044.
Qiang, Z., Yang, H., Chen, L., Pang, H., Ma, Z., Zhou, W., 2008. Fano filters based on
transferred silicon nanomembranes on plastic substrates. Appl. Phys. Lett. 93, 061106.
Rakich, P.T., Popovic, M.A., Soljacic, M., Ippen, E.P., 2007. Trapping, corralling and spec-
tral bonding of optical resonances through optically induced potentials. Nat. Photonics
1 (11), 658–665.
Rodriguez, A.W., McCauley, A.P., Hui, P.-C., Woolf, D., Iwase, E., Capasso, F.,
Loncar, M., Johnson, S.G., 2011. Bonding, antibonding and tunable optical forces in
asymmetric membranes. Opt. Express 19 (3), 2225–2241.
Roels, J., De Vlaminck, I., Lagae, L., Maes, B., Van Thourhout, D., Baets, R., 2009. Tunable
optical forces between nanophotonic waveguides. Nat. Nanotechnol. 4 (8), 510–513.
Rogers, J.A., Lagally, M.G., Nuzzo, R.G., 2011. Synthesis, assembly and applications of
semiconductor nanomembranes. Nature 477 (7362), 45–53.
Roh, Y.-G., Tanabe, T., Shinya, A., Taniyama, H., Kuramochi, E., Matsuo, S., Sato, T.,
Notomi, M., 2010. Strong optomechanical interaction in a bilayer photonic crystal.
Phys. Rev. B 81 (12), 121101.
Rosenberg, A., Carter, M., Casey, J., Kim, M., Holm, R., Henry, R., Eddy, C.,
Shamamian, V., Bussmann, K., Shi, S., Prather, D.W., 2005. Guided resonances in
asymmetrical GaN photonic crystal slabs observed in the visible spectrum. Opt. Express
13 (17), 6564–6571.
Shuai, Y., Zhao, D., Singh Chadha, A., Seo, J.-H., Yang, H., Fan, S., Ma, Z., Zhou, W.,
2013a. Coupled double-layer Fano resonance photonic crystal filters with lattice-
displacement. Appl. Phys. Lett. 103 (24), 241106.
188 Weidong Zhou and Shanhui Fan
Shuai, Y., Zhao, D., Tian, Z., Seo, J.-H., Plant, D.V., Ma, Z., Fan, S., Zhou, W., 2013b.
Double-layer Fano resonance photonic crystal filters. Opt. Express 21 (21),
24582–24589.
Shuai, Y.-C., Zhao, D., Liu, Y., Stambaugh, C., Lawall, J., Zhou, W., 2017. Coupled bilayer
photonic crystal slab electro-optic spatial light modulators. IEEE Photonics J. 9 (2), 1–11.
Soref, R., Bennett, B., 1987. Electrooptical effects in silicon. IEEE J. Quantum Electron.
23 (1), 123–129.
Stomeo, T., Grande, M., Rainò, G., Passaseo, A., D’Orazio, A., Cingolani, R., Locatelli, A.,
Modotto, D., De Angelis, C., De Vittorio, M., 2010. Optical filter based on two coupled
PhC GaAs-membranes. Opt. Lett. 35 (3), 411–413.
Suh, W., Fan, S., 2003. Mechanically switchable photonic crystal filter with either all-pass
transmission or flat-top reflection characteristics. Opt. Lett. 28 (19), 1763–1765.
Suh, W., Fan, S., 2004. All-pass transmission or flattop reflection filters using a single pho-
tonic crystal slab. Appl. Phys. Lett. 84, 4905.
Suh, W., Yanik, M.F., Solgaard, O., Fan, S., 2003. Displacement-sensitive photonic crystal
structures based on guided resonance in photonic crystal slabs. Appl. Phys. Lett. 82 (13),
1999.
Suh, W., Wang, Z., Fan, S., 2004. Temporal coupled-mode theory and the presence of non-
orthogonal modes in lossless multimode cavities. IEEE J. Quantum Electron. 40 (10),
1511–1518.
Suh, W., Solgaard, O., Fan, S., 2005. Displacement sensing using evanescent tunneling
between guided resonances in photonic crystal slabs. J. Appl. Phys. 98, 033102.
Synopsys, 2016. Taurus Medici [Online]. Sysnopsys.
Thurman, S.T., Morris, G.M., 2003. Controlling the spectral response in guided-mode res-
onance filter design. Appl. Optics 42 (16), 3225–3233.
Van Thourhout, D., Roels, J., 2010. Optomechanical device actuation through the optical
gradient force. Nat. Photonics 4 (4), 211–217.
Wiederhecker, G.S., Chen, L., Gondarenko, A., Lipson, M., 2009. Controlling photonic
structures using optical forces. Nature 462 (7273), 633–636.
Woolf, D., Hui, P.-C., Iwase, E., Khan, M., Rodriguez, A.W., Deotare, P., Bulu, I.,
Johnson, S.G., Capasso, F., Loncar, M., 2013. Optomechanical and photothermal inter-
actions in suspended photonic crystalmembranes. Opt. Express 21 (6), 7258–7275.
Yang, H., Zhao, D., Chuwongin, S., Seo, J.H., Yang, W., Shuai, Y., Berggren, J.,
Hammar, M., Ma, Z., Zhou, W., 2012. Transfer-printed stacked nanomembrane lasers
on silicon. Nat. Photonics 6 (9), 615–620.
Zhang, K., Seo, J.H., Zhou, W., Ma, Z., 2012. Fast flexible electronics using transferrable
silicon nanomembranes (topical review). J. Phys. D Appl. Phys. 45 (14), 143001.
Zhou, W., Ma, Z., Yang, H., Qiang, Z., Qin, G., Pang, H., Chen, L., Yang, W.,
Chuwongin, S., Zhao, D., 2009. Flexible photonic-crystal Fano filters based on trans-
ferred semiconductor nanomembranes. J. Phys. D Appl. Phys. 42, 234007.
CHAPTER SEVEN
Contents
1. Introduction 189
2. Overview of photonic crystal lasers 190
3. Photonic crystal MR-VCSEL cavity design 196
3.1 Photonic crystal MRs 196
3.2 Energy and phase penetration properties in MRs 198
3.3 MR-VCSEL 200
4. Photonic crystal bandedge surface-emitting laser cavity design 201
4.1 Bandedge mode design 203
4.2 Lateral cavity scaling 204
5. Printed membrane laser cavities 206
6. MR-VCSELs based on 1D PCS/grating mirrors 208
7. MR-VCSEL characterization 209
8. PCSEL characterization 212
9. PCSEL on bulk silicon substrate 213
10. PCSEL thermal performance analysis 215
11. Electrically pumped PCSEL characterization 218
12. Summary 220
Acknowledgments 221
References 221
1. Introduction
Silicon-based light sources have witnessed significant advances
over the last decade, with one of the most promising approaches being
heterogeneous integration of direct-bandgap group III–V materials on various
Si-based laser cavities on silicon-on-insulator (SOI) substrates (Bakir et al.,
2006; Balakrishnan et al., 2006; Chen et al., 2011; Liang et al., 2010; Mi
et al., 2005; Park et al., 2006; Roelkens et al., 2010; Stankovic et al., 2011;
• Efficient electrical
injection
• Simplified process
development
Continued
Table 1 Photonic crystal surface-emitting lasers (PCSELs)—cont’d
Configurations PhC functions Characteristics References
DBR- Photonic crystal membrane reflector Pros: Sciancalepore et al. (2012) and Yang et al.
free VCSELs: (2012)
VCSEL • 1D high contrast grating or 2D • DBR free with simple
photonic crystal-based single- wafer growth (LED
layer reflectors like structure)
• Either one or both MRs are • Spontaneous emission
used to replace DBRs control with high
spontaneous emission
factor
• Ultracompact cavity
• Simple cavity
resonance tuning with
patterning process
Cons/Challenges:
• Efficient electrical
injection
• Simplified process
development
Photonic crystal surface-emitting lasers 195
the active light emission region is placed inside the PBG defect cavity
(Noda et al., 2007; Yablonovitch, 1987). With extremely high
Q/Vmode (Q and Vmode being the corresponding quality factor and
modal volume, respectively), strong spontaneous emission control
is feasible for ultralow threshold single-mode lasers. Typically these
defect modes are located below lightline. This configuration can lead
to potentially thresholdless lasers (or single-mode LEDs) and single-
photon sources (Bhattacharya et al., 2005; Colombelli et al., 2003;
Ellis et al., 2011; Hwang et al., 2000; Painter, 1999; Park et al.,
2004; Zhou et al., 2001). Another architecture reported recently
(Wu et al., 2015) is the coupling of unpatterned monolayer tungsten
diselenide monolayer gain medium with the defect-mode photonic
crystal cavity. This configuration can lead to higher efficiency lasers
as no-patterned or etching is done on the gain medium. Challenges
may lie on the trade-offs in the desired high Q cavity inside the
PBG and the high coupling efficiency between the high index gain
medium and photonic crystal cavities.
(III) PBELs: The bandedge effect in defect-free photonic crystal structures,
where the group velocity is close to zero, has advantages in achieving
high performance semiconductor surface-emitting lasers, including
low lasing threshold, single longitudinal and transverse mode over a
large lasing area, narrow linewidth, high power output, small beam
divergence angle, polarization control, and output beam pattern con-
trol (Chua et al., 2014; Hirose et al., 2014; Imada et al., 1999; Meier
et al., 1999b; Miyai et al., 2006; Noda et al., 2001). Based on the
bandedge effect, high performance lasers on GaAs have been demon-
strated, including a very recent demonstration of high power, single
mode, and large-area PC bandedge lasers with unpatterned QW
active region using a regrowth process (Hirose et al., 2014). Different
from PBDL discussed above, the modes considered here are typically
located above lightline, thus these lasers are intrinsically surface emit-
ting with excellent beam qualities.
(IV) Membrane reflector VCSEL (MR-VCSELs): Based on the Fano reso-
nance principles, broadband reflection can be achieved in single-layer
MRs based on 1D or 2D PCS. These ultracompact high performance
MR structures have been proposed to replace the bulky DBR struc-
tures, to create ultracompact MR-VCSELs (Sciancalepore et al.,
2012; Yang et al., 2012). Taking advantage of Fano resonance prin-
ciples, ultracompact high performance VCSEL cavities can be formed
196 Weidong Zhou
on any substrate, such as III–V and Si, because the mirror cavity and
the gain medium can be heterogeneously integrated with different
material systems. This is one of the most promising ways to form lasers
on Si, and to form high performance long wavelength VCSELs, in
which DBR remains a great challenge.
Among four configurations discussed earlier, PBDL utilizes the localized or
defect mode in the PBG. On the other hand, for the latter two configura-
tions (PBEL and MR-VCSEL), it is not essential to design PCS’s with PBG,
as both utilize propagating modes in defect-free PCS structures. While
PBEL can utilize the mode both below and above the light lines,
MR-VCSEL’s only use the modes above the light lines. Fano resonance
modes can be designed in PBEL, where lasing is feasible due to distributed
feedback (Chua et al., 2011). Most notably, the other type of Fano
resonance-based laser is the recently demonstrated MR-VCSEL, where
Fano resonance MRs are incorporated in the VCSEL cavity design to form
ultracompact VCSEL’s and lasers on Si. It is also worth mentioning that since
PBG is not required in PBEL, the requirement on high index contrast can be
relaxed. Lasing cavities can be built directly on patterned bulk semiconduc-
tor InP or Si substrates.
In this chapter, we will focus on detailed discussions related to on-chip
photonic crystal lasers where the lasing cavity is formed directly on silicon
substrates, the active gain medium is a conventional compound semicon-
ductor quantum well (QW) structures.
A B
100 100
l = 1500 nm
80
Reflection (%)
Transmission
60
40
20
R
T (Log)
0
1400 1450 1500 1550 1600
Wavelength (nm)
Fig. 1 Fano resonance membrane reflector (MR) design and modal coupling:
(A) simulated reflection (blue) and transmission (red) spectra for the Fano MR and
(B) simulated field distribution profiles for the design shown in (A) at one Fano reso-
nance location (λ ¼ 1500 nm).
198 Weidong Zhou
A B
5 0.4
Lp,DBR Lp,1DSWG Lp,2DPCS Le,DBR Le,1DSWG Le,2DPCS
4
0.3
Le (μm)
Lp (μm)
3
0.2
2
0.1
1
0 0
1.5 1.55 1.6 1.5 1.55 1.6
Wavelength (μm) Wavelength (μm)
Fig. 2 (A) Calculated phase penetration depths for three types of reflectors and (B) the
energy penetration depths for three types of reflectors. Reproduced with permission from
Zhao, D., Ma, Z., Zhou, W., 2010. Field penetrations in photonic crystal Fano reflectors. Opt.
Express 18 (13), 14152–14158.
3.3 MR-VCSEL
Once two good MRs with high reflectivity and broader overlapped spectral
band are obtained, a multilayer Si-MR/QW/Si-MR (MR-VCSEL) cavity
can be build up (Fig. 3A). Here, we need to decide the cavity resonant (lon-
gitude) mode by selecting a suitable cavity length, i.e., the thickness of the
buffer layers between Si-MR and QW. Fig. 3B shows the reflection spectrum
(blue line) of one designed MR-VCSEL cavity. To further confirm this cavity
mode, the phase resonant condition (total phase change of one round-trip in
cavity is equal to an integer times of 2π) is also calculated according the reflec-
tion phase change ϕ of the top and bottom MRs, which is plotted as red-
dotted line in Fig. 3B. One can find the mode located at 1478 nm has a
m2π phase change, which means 1478 nm is a real cavity mode.
By employing FDTD, a short temporal Gaussian pulse is used to excite
the cavity modes and the quality factor of the cavity mode at 1478 nm is cal-
culated to be 4300 according to Q ¼ Re(ω)/2Im(ω) (Oskooi et al., 2010).
Then a longer temporal Gaussian pulse is used to excite the single cavity
mode and the stable field is recorded. The E-field distribution is demon-
strated in Fig. 3C with the red line, where the cavity index profile is also
plotted with the blue line.
Another design concern that needs to be taken care of is the separation
the lasing cavity modes from the QW layer as mentioned in Zhao et al.
(2012). Despite the relatively large field concentration in Si-MR layers,
the lasing mode field confinement factor is 5.6% for MR-VCSELs, which
is comparable to DBR-based VCSELs. This critical feature enables low
threshold lasing in the MR-VCSEL cavities.
Photonic crystal surface-emitting lasers 201
Si substrate
Lcav
B C
100 2 4
Refractive index
1.5 2
Reflection (%)
80 1 1
0.8 0
1
0.6 –1
60 0.4 –2
t2 = 400 nm 0.5
t4 = 400 nm 0.2 –3
0 –4
40 0 –3 –2 –1 0 1 2 3
1410 1440 1470 1500 1530
Position in cavity (μm)
Wavelength (nm)
Fig. 3 Characteristics of designed MR-VCSEL cavity: (A) schematic of MR-VCSEL;
(B) calculated cavity resonance mode based on cavity reflection and phase resonant
condition; and (C) field distribution of cavity mode. Reproduced with permission from
Zhao, D., Yang, H., Chuwongin, S., Seo, J., Ma, Z., Zhou, W., 2012. Design of photonic crystal
membrane-reflector-based VCSELs. IEEE Photonics J. 4 (6), 2169–2175.
A B
1540 D
Wavelength (nm)
C
1500 B
MQW A
1460
z Si PC
TE(s) TM(p)
q y 1420
SiO2 0 0.005 0.01
x SOI substrate kx (2p/a) → G – X
C D
4
3 Refractive index
1 2
1
|E|2
0.8 QW
y
0.6 0
Si–PC
0.4 –1
0.2
–2
0
0.5 1.5 2.5 3.5 x
Position in cavity (a)
Fig. 4 (A) Schematic of PC bandedge membrane laser on SOI. (B) Photonic band dia-
gram along Γ–X direction of Si-PC bandedge membrane laser cavity structure, with
the transferred InGaAsP QW mesas on top. (C) Field distribution of cavity mode and
(D) simulated field distributions. Reproduced with permission from Liu, S.-C., Zhao, D.,
Liu, Y., Yang, H., Sun, Y., Ma, Z., Reuterskio€ld-Hedlund, C., Hammar, M., Zhou, W., 2017.
Photonic crystal bandedge membrane lasers on silicon. Appl. Optics 56 (31), H67–H73.
structures, the cavity mode can lase over a large 2D area and gives a high
power output (Chua et al., 2014; Hirose et al., 2014).
A schematic of the PC bandedge membrane laser on Si is shown in
Fig. 4A. The laser cavity consists of an InGaAsP MQW membrane
heterostructure, and a Si-PC slab patterned on SOI with the BOx layer
thickness of 400 nm. The InGaAsP MQW layer design is based on the pre-
viously reported low temperature (LT) PC bandedge membrane laser,
which consists of a bottom InGaAs contact layer, top and bottom InP clad-
ding layers, and eight pairs of strain-compensated In0.76Ga0.24As0.83P0.17/
In0.485Ga0.515As0.83P0.17 with the center emission wavelength of 1535 nm
at room temperature (Zhao et al., 2016). In order to achieve room temper-
ature PC bandedge membrane lasers in such a cavity, it is highly desirable to
design a cavity with the following characteristics: (1) two dimension square
lattices Si-PC structure is designed to have a lattice constant of 480 nm,
based on the second-order Bragg diffraction condition (Chua et al.,
2014; Hirose et al., 2014; Imada et al., 1999, 2002; Martınez et al., 2009;
Photonic crystal surface-emitting lasers 203
Meier et al., 1999; Noda et al., 2001). (2) The cavity mode wavelength
should match with MQW emission peak, with the hole radius and thickness
of Si-PC slab designed to be 72 nm (r ¼ 0.15a) and 190 nm, respectively. (3)
The separation between the InGaAsP MQW layer and Si-PC slab is reduced
to enhance the confinement factors and the in-plane coupling efficiency
(Sakai et al., 2006; Sato et al., 2012).
The quality factor and the field distribution of this cavity mode are also
investigated. Their quality factors are obtained by Fano fitting the guided
resonances in the reflection spectra (Luk’yanchuk et al., 2010). In our cal-
culations, only the vertical radiation loss is considered as the in-plane loss and
the internal loss are negligible (Hirose et al., 2014). Hence, only the radiation
loss is considered for the infinite cavity using the following relationship:
2π
α¼ (5)
Q∙a
Therefore, the gain threshold can be derived as follows:
α
gth ¼ (6)
Г QW
The E-field along the center of cavity (left) and the cross-sectional views
of the E-field intensity of modes (right) are shown in Fig. 4D. One can see
the field is strongly confined inside the cavity. Based on the above funda-
mental equations, the confinement factor, quality factor, and radiation loss
of mode D (λ ¼ 1540 nm) is calculated to be 16.83%, 1.9 105 and
0.68 cm1, respectively. Single-mode lasing can be predicted to occur at
mode D which has the lowest gain threshold of 4.04 cm1.
A B
Layer Thickness
p-InP 130 nm
InGaAsP MQW 127 nm
n-InP 30 nm
z Si–PC
SiO 2 n-InGaAs 40 nm
q y
te 190 nm, a = 480 nm, and
bstra
Si su Si–PC
r/a = 0.14
x
SiO2 400 nm
Si substrate 525 μm
C
D
Etch
depth
z z
q y q y
x x
Fig. 5 (A) Schematic of PCSEL with 600 × 600 μm2 size. (B) Cavity layer definitions.
(C) Schematic of PCSEL with variation in lateral dimension (L). (D) Schematic of PCSEL
with variation in vertical etch depth. Reproduced with permission from Liu, S.-C., Zhao, D.,
Ge, X., Reuterskio€ld-Hedlund, C., Hammar, M., Fan, S., Ma, Z., Zhou, W., 2018. Size scaling of
photonic crystal surface emitting lasers on silicon substrates. IEEE Photonics J. 10 (3), 1–6.
A B
350 7
1600 30
Gain threshold (cm–1)
confinement scheme by controlling the etch depth of the lasing cavity (lat-
eral index contrast). The result is shown in Fig. 6B for a PCSEL on Si with a
lateral cavity dimension of 48 μm (L ¼ 100 a). By increasing the etch depth
from 0 to 500 nm, which corresponds to the complete removal of the
InGaAsP QW heterostructure and Si-PC cavity (Fig. 5D), the cavity Q fac-
tor increases from 2300 to 6800. We attribute this to the increased lateral
confinement due to the increased lateral index contrast.
QW layer and Si-PC cavity layer in order to obtain a high confinement fac-
tor which is critical to low threshold lasing.
An SOI wafer with 230 nm top Si layer and 400 nm BOx layer was used
here. The top Si layer was first thinned down to 190 nm by thermal oxida-
tion and selective wet etching processes. Next, the 2D square lattice PC pat-
tern with a size of 1 1 mm2 was defined with ZEP-520A resist by using the
standard electron beam lithography (EBL). The sample was then etched with
the ZEP-520A resist pattern as the mask to create a high index contrast
between Si and air by using the dry etching process. The scanning electron
microscopic (SEM) top view of Si-PC on SOI substrate is shown in Fig. 7A.
The lattice constant a and air hole radius r of the fabricated PC are 480 and
72 nm, respectively.
Fig. 7 (A) Zoomed-in SEM top view of the defect-free Si-PC cavity. (B) An SEM image of
InGaAsP MQW mesa transferred onto a Si-PC. (C) Cross-sectional view of the cavity and
(D) micrograph image of the transferred mesa array on a Si-PC. Reproduced with permis-
sion from Liu, S.-C., Zhao, D., Liu, Y., Yang, H., Sun, Y., Ma, Z., Reuterskio€ld-Hedlund, C.,
Hammar, M., Zhou, W., 2017. Photonic crystal bandedge membrane lasers on silicon. Appl.
Optics 56 (31), H67–H73.
208 Weidong Zhou
Fig. 8 Hybrid top MR/bottom DBR VCSEL configuration: suspended GaAs high contract
grating (HCG) reflector monolithically grown on GaAs substrate. Reproduced with per-
mission from Huang, M.C.Y., Zhou, Y., Chang-Hasnain, C.J., 2007. A surface-emitting laser
incorporating a high-index-contrast subwavelength grating. Nat. Photonics 1 (2), 119–122.
7. MR-VCSEL characterization
Optically pumped devices are mounted inside a cryostat and charac-
terized with a monochromator-based microphotoluminescence (μ-PL) set
up using quasi-continuous wave (c.w.) 532 nm laser pumping (with 50%
duty cycle).
Shown in Fig. 9A is the L–L plot (light output for different pump pow-
ers) and the corresponding spectral linewidths measured at T ¼ 50 K. The
threshold pump power is 8 mW, or 0.32 kW/cm2. The measured spectral
linewidths reduce from 30 nm below threshold to 0.6–0.8 nm above thresh-
old. The measured spectral outputs are shown in Fig. 9B, for pump powers
below, at, and above thresholds (points (i, ii, iii, iv) at L–L curve). The lasing
spectral linewidth is 8 Å, which is limited by the measurable resolution of
monochromator. The relative peak location shift shown in Fig. 9B from bias
levels (ii) and (iii, iv) is mostly related to mode hopping and temperature rise
inside the active region at higher pump power levels.
210 Weidong Zhou
A B
80 100 80
T = 50 K, D = 100 µm iv
Linewidth (nm)
60 60 iv
10
40 40
iii
20 1 20 iii
i ii ii
i
0 0
0 4 8 12 16 20 1430 1450 1470 1490 1510
Pump power (mW) Wavelength (nm)
Fig. 9 (A) The L–L plot (light output for different pump powers) and the corresponding
spectral linewidths at T ¼ 50 K and (B) the measured spectral outputs for pump powers
below, at, and above thresholds. Reproduced with permission from Yang, H., Zhao, D.,
Chuwongin, S., Seo, J.-H., Yang, W., Shuai, Y., Berggren, J., Hammar, M., Ma, Z.,
Zhou, W., 2012. Transfer-printed stacked nanomembrane lasers on silicon. Nat. Photonics
6 (9), 615–620.
A B
20
Power (a.u.)
Lth (mW)
1
15
0.5
0 120 10
1520 90 T0 = 125 K
)
70 (K
Wa 1500 ure 5
vele 1480 50 t
ngth p era 20 40 60 80 100 120
(nm 1460 15 m Temperature (K)
) Te
Fig. 10 (A) Measured lasing spectral at different temperatures under 532 nm laser
pumping and (B) lasing threshold as a function of temperature. Reproduced with permis-
sion from Yang, H., Zhao, D., Chuwongin, S., Seo, J.-H., Yang, W., Shuai, Y., Berggren, J.,
Hammar, M., Ma, Z., Zhou, W., 2012. Transfer-printed stacked nanomembrane lasers on
silicon. Nat. Photonics 6 (9), 615–620.
which match well with the cavity resonances. Fig. 10B shows the lasing
threshold (Lth) as a function of T. Lth lineally increases with T, which
matches the QW PL efficiency linearly decreases as T.
In order to estimate the lasing gain threshold in MR-VCSEL cavity,
detailed investigation was also carried out to estimate the associated phase
discontinuities in MR mirrors, as well the field penetration depth and effec-
tive cavity length in the optimized MR-VCSEL cavity. The results are sum-
marized in Table 2. Based on lasing gain threshold conditions,
Finally, lasing gain threshold can then be estimated, based on the cavity
gain and loss characteristics reported earlier from similar QW structures
(Rapp et al., 1999). Based on our design, MR-VCSEL can lase with a gain
threshold of 840–1300 cm1, depending on the MR reflectivity.
The cavity design here can definitely lead to high performance MR-
VCSELs with very low threshold and high efficiency, as shown in Table 2.
For comparison, also shown in Table 2 are the typical parameters for long
wavelength VCSELs based on conventional InGaAsP/InP DBRs. As can
be seen, the MR-VCSEL has a very similar lasing gain threshold with con-
ventional InGaAsP/InP DBR-based VCSEL, even though MR-VCSELs
have smaller cavities with much shorter effective cavity lengths (2.4 μm vs
5–20 μm) (Karim et al., 2000; Rapp et al., 1997).
A B
100
80 On SOI 1 P = 0.32 kW/cm2
T = 300 K T = 300 K
Peak intensity (a.u.)
c.w. 0.8
Intensity (a.u.)
Linewidth (nm)
60
10
0.6
40
0.4
20 1
0.2
0 0
0.1 0.2 0.3 0.4 0.5 0.6 1510 1530 1550 1570
Power density (kW/cm2) Wavelength (nm)
Fig. 11 (A) Lasing power and linewidth vs input power density at room temperature
and (B) emission spectrum of the laser above threshold. Reproduced with permission
from Liu, S.-C., Zhao, D., Liu, Y., Yang, H., Sun, Y., Ma, Z., Reuterskio€ld-Hedlund, C.,
Hammar, M., Zhou, W., 2017. Photonic crystal bandedge membrane lasers on silicon. Appl.
Optics 56 (31), H67–H73.
8. PCSEL characterization
PCSEL devices were optically pumped from the surface-normal direc-
tion via a long working distance objective lens (10) by a c.w. green (532 nm)
laser. The diameter of the excitation spot is around 110 μm. The optical char-
acteristics of this PC bandedge membrane laser device were then measured by
a monochromator-based μ-PL set-up. Fig. 11A shows the light output and
linewidth as a function of the pump power measured on the surface of the
focusing plane. A threshold power density of 0.25 kW/cm2 was obtained.
The measured spectral linewidth was reduced from 14nm (below threshold)
to 0.54nm (above threshold). Notice a thermal roll-over was observed at opti-
cal pumping intensities higher than 0.4 kW/cm2. A spectrum of the laser
above the threshold is shown in Fig. 11B where the lasing wavelength is
1542 nm, very close to the predicted value of mode D in the simulation.
The lower threshold pump power density is a result of the high Q of mode
D. All the other modes are possible to get lasing with sufficient pump energy.
Fig. 12 shows the lasing spectrum measured at a pump power density of
0.32 kW/cm2. A clear single-mode operation was observed. A zoomed-in
lasing spectrum plotted in semilog scale is shown in Fig. 12A. The peak
wavelength, full width at half maximum (FWHM) of the spectrum, and
SMSR are 1542 nm, 0.54 nm, and 31.8 dB, respectively. Fig. 12B is a mea-
sured far-field lasing profile with single-mode output at the focusing plane of
the collimator.
Photonic crystal surface-emitting lasers 213
A B
0 P = 0.32 kW/cm
2 3584
3328
3072
T = 300 K 2816
Relative intensity (dB)
SMSR = 31.8 dB
2048
–10 nm 1792
1536
1280
–15 1024
768
512
–20 256
0
–25
–30
–35
–40
1525 1535 1545 1555
Wavelength (nm)
Fig. 12 (A) Lasing spectral output plotted in semilog scale above the pumping thresh-
old and (B) 3D far-field laser profile. Reproduced with permission from Liu, S.-C., Zhao, D.,
Liu, Y., Yang, H., Sun, Y., Ma, Z., Reuterskio€ld-Hedlund, C., Hammar, M., Zhou, W., 2017.
Photonic crystal bandedge membrane lasers on silicon. Appl. Optics 56 (31), H67–H73.
A B
MQW
MQW
Si PC Si–PC
z Bulk Si
substrate Bulk Si
y
200 nm
x
C D
4
3
Refractive index
1 2
QW
0.8 1 y
|H|2
0.6 0 Si–PC
0.4 –1
0.2
–2
0 x
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Position in cavity (μm)
Fig. 13 (A) Schematic of PC bandedge membrane laser on bulk Si substrate. (B) Cross-
sectional view of the Si-PC bandedge membrane laser cavity structure on bulk Si sub-
strate, with the transferred InGaAsP QW mesas on top. (C) Field distribution of the cavity
mode and (D) simulated in-plane and cross-sectional field distributions. Reproduced with
permission from Liu, S.-C., Zhao, D., Liu, Y., Yang, H., Sun, Y., Ma, Z., Reuterskio€ld-Hedlund,
C., Hammar, M., Zhou, W., 2017. Photonic crystal bandedge membrane lasers on silicon.
Appl. Optics 56 (31), H67–H73.
in Fig. 13C, where the integrated H-field intensity of the whole cavity along
the z-axis direction is plotted together with the cavity index profile. The
cross-sectional view of H-field intensity of lasing mode is shown in
Fig. 13D. Note here that the large air hole radius was designed to ensure
the high refractive index contrast. One can see the field was confined inside
the thin cavity due to high index contrast above and below the laser cavity.
Moreover, the thickness of the Si-PC slab was chosen to be 400 nm to obtain
high Q value by reducing the vertical radiation loss. Finally, The Q value of
the lasing mode can be theoretically calculated around 3000, which would
be able to achieve the single-mode lasing in the 160 K operation temperature
(Zhao et al., 2016).
To evaluate the laser performance experimentally, the PC bandedge
membrane laser on bulk Si substrate was mounted on a temperature-
controlled stage inside a cryostat and characterized by the same μ-PL set-
up system in our previous works (Yang et al., 2012; Zhao et al., 2016).
Fig. 14A shows the light output and linewidth as a function of the pump
Photonic crystal surface-emitting lasers 215
A B
3 100
On bulk Si 1 P = 0.5 kW/cm2
2.5 T = 160 K T = 160 K
Peak intensity (a.u.)
c.w. 0.8
Linewidth (nm)
Intensity (a.u.)
2 10
0.6
1.5
1 0.4
1
0.5 0.2
0 0
0.2 0.3 0.4 0.5 0.6 0.7 1420 1440 1460 1480
Power density (kW/cm2) Wavelength (nm)
Fig. 14 (A) Lasing intensity and linewidth vs input power density at T ¼ 160 K for PCSEL
on bulk Si substrate and (B) emission spectrum of the laser above threshold. Reproduced
with permission from Liu, S.-C., Zhao, D., Liu, Y., Yang, H., Sun, Y., Ma, Z., Reuterskio€ld-
Hedlund, C., Hammar, M., Zhou, W., 2017. Photonic crystal bandedge membrane lasers
on silicon. Appl. Optics 56 (31), H67–H73.
The bulk Si substrate is expected to dissipate more heat along the vertical
direction from the surface of the cavity to the bottom substrate and heat sink.
A figure of merit to quantify the heat dissipation out of the cavity is the ther-
mal resistance (Ferrara et al., 2015; Flick et al., 2011). By extracting the shift
of emission peak wavelength with varying the temperature and pump
power, the thermal resistance of the laser can be mathematically defined
by the following ratio:
Δλ Δλ
|T ¼ const |T ¼ const
ΔPheat ΔPp
Rth ¼ (8)
Δλ Δλ
|P ¼ const |P ¼ const
ΔT ΔT
Here, the ΔPheat is the heat generated from the cavity and ΔPp is the total
pump power. We assume heat generation is primarily in the QW section
(Lindberg et al., 2005), and it is linearly proportional to the optical pump
power for the operation window between threshold and roll-over. We first
characterize the PC bandedge membrane laser on bulk Si substrate. The
wavelength spectrum dependence of the pumping power from the PC
bandedge membrane laser on bulk Si substrate was measured as shown in
Fig. 15A. The peak of the emission wavelength slightly shifted toward lon-
ger wavelengths when the pump power increased from 25 to 50 mW
at 160 K. In our measurement, the spectral shift vs power ratio presents
the spectral shift vs power generated from the QW region. Second, the
spectral dependence of the operation temperature varying from 120 to
200 K at a constant incident power was also measured and shown in
Fig. 15B. Please note the cavity design of the PC bandedge membrane laser
Photonic crystal surface-emitting lasers 217
A B
T = 160 K P = 40 mW
P = 40 mW T = 160 K
0.8 P = 50 mW 0.8 T = 180 K
T = 200 K
0.6 0.6
0.4 0.4
0.2 0.2
0 0
1440 1445 1450 1455 1460 1440 1445 1450 1455 1460
Wavelength (nm) Wavelength (nm)
Fig. 15 (A) Lasing spectra at different pump power above threshold at constant
160 K. (B) Lasing spectra at different operation temperature with constant 40 mW
pump power. Reproduced with permission from Liu, S.-C., Zhao, D., Liu, Y., Yang, H.,
Sun, Y., Ma, Z., Reuterskio€ld-Hedlund, C., Hammar, M., Zhou, W., 2017. Photonic crystal
bandedge membrane lasers on silicon. Appl. Optics 56 (31), H67–H73.
A B
1 Laser on SOI Laser on SOI
5
Laser on bulk Si Laser on bulk Si
0.8 4
Peak shift (nm)
0.6 3
0.4 2
0.2 1
0 0
25 30 35 40 45 50 120 140 160 180 200
Pump power (mW) Temperature (K)
Fig. 16 (A) Emission peak wavelength shift vs pump powers for PC bandedge mem-
brane laser on SOI and on bulk Si substrate. (B) Emission peak wavelength shift vs oper-
ation temperatures for PC bandedge membrane laser on SOI and on bulk Si substrate.
Reproduced with permission from Liu, S.-C., Zhao, D., Liu, Y., Yang, H., Sun, Y., Ma, Z.,
Reuterskio€ld-Hedlund, C., Hammar, M., Zhou, W., 2017. Photonic crystal bandedge mem-
brane lasers on silicon. Appl. Optics 56 (31), H67–H73.
on bulk Si substrate was optimized for device operation around 160 K. Single-
mode lasing operation was obtained at the 120–200 K range for the pump
power at 40 mW. The wavelength shift vs pumping power is calculated to
be 0.0125 nm/mW, and the shift vs temperature was 0.0412 nm/K.
Fig. 16A and B shows the measured emission wavelength peak shift at
different pump powers and varying operation temperatures for both PC
bandedge membrane lasers on bulk Si substrate and on SOI substrate
218 Weidong Zhou
Fig. 17 Electrically pumped PCSELs on SOI substrate: (A) schematic and (B) detailed
heterostructure and cavity design parameters.
Photonic crystal surface-emitting lasers 219
A B
1 10–1
10–2
0.8 10–3
Transmission
Current (mA)
10–4
0.6
10–5
0.4 10–6
10–7
0.2
TE 10–8
TM
0 10–9
1480 1500 1520 1540 –1 0 1 2 3 4
Wavelength (nm) Voltage (V)
Fig. 18 Electrically pumped PCSELs on SOI substrate: (A) simulated transmission spectra
for both TE and TM modes and (B) measured IV curve for the electrically pumped PCSEL.
transmission spectra of the cavity, with blue lines for the transverse electric
(TE, or s) polarization, and red lines for the transverse magnetic (TM, or p)
polarization. Substrate removal process is used in the device fabrication for
the release of top large-area InGaAsP MQW membrane from the native InP
substrate. The membrane is then transferred onto Si-PC cavity. Top n-type
ring contact is formed by e-beam evaporation, with alloy contact consisting
of Ni/Ge/Au/Ti/Au (10/20/30/10/200 nm). Mesa etching is then carried
out down to the bottom n-contact layer, followed with another e-beam evap-
oration process for the bottom n-contact formation. A plasma-enhanced
chemical vapor deposition SiO2 layer is deposited for laser planarization,
followed by interconnect layer formation. The complete device is then
mounted on a chip carrier with wire bonding process for device characteri-
zation. The current–voltage (IV) characteristics were measured and evaluated
for different BTJ sizes, with one curve plotted in semilog scale for electrically
pumped PCSEL shown in Fig. 18B. Excellent IV characteristics were shown
here, essential for efficient charge injection and lasing.
The device characterization was carried out with c.w. current injection
at room temperature. The output was measured in the direction normal to
the surface via a long working distance objective lens and sent into a
monochromator. Fig. 19A shows the measured electroluminescence (EL)
spectrum at various biased currents for a PCSEL with an 80 80 μm2 BTJ
mesa region. A dominant single-mode lasing emission peak at 1504 nm
appears above the threshold current. Fig. 19B shows the light–current–voltage
(LIV) characteristics. The turn on voltage is about 0.8 V. The threshold cur-
rent is about 2 mA.
220 Weidong Zhou
A B
8 4 4
I = 15 mA
I = 5 mA
EL intensity (a.u.)
6 I = 2.5 mA 3 3
Intensity (a.u.)
Voltage (V)
4 2 2
2 1 1
0 0 0
1400 1450 1500 1550 1600 0 2 4 6 8 10
Wavelength (nm) Current (mA)
Fig. 19 (A) The emission spectra at different current injection levels measured at room
temperature and (B) the light–current–voltage (LIV) characteristics for an 80 80 μm2
electrically pumped PCSEL on Si.
12. Summary
In summary, we review here recent progresses on photonic crystal
surface-emitting membrane lasers for on-chip integration with CMOS elec-
tronics. Defect-free photonic crystal structures are incorporated into laser
cavities as ultracompact MRs for DBR-free VCSELs and as bandedge cav-
ities evanescently coupled with gain medium for PCSELs. Lateral cavity size
effect was also investigated for cavity scaling. Electrically injected PCSELs
with BTJ-based InGaAsP QW heterostructures were also demonstrated.
Further research is needed to address device efficiencies. While high effi-
ciency VCSELs and PCSELs on native III–V substrates have been reported
earlier with wall-plug efficiencies greater than 50%, the efficiencies in the
heterogeneously integrated DBR-free MR-VCSELs and Si-based PCSELs
are largely impacted by the cavity coupling and thin membrane surface and
interface qualities. Device scaling in both vertical and lateral directions can
also lead to improved device modulation speed and operation data transfer
efficiency, with ultimate objective of subfemto joule per bit energy
efficiency.
Thermal and reliability are another set of issues to be addressed. Since
PBG is not essential in these photonic crystal structures reported here,
relaxed index contrast is possible to have photonic crystal lasing cavities built
on low index contract material systems. However certain index contract is
still needed, which may result in the need of low index buffers in the lasing
Photonic crystal surface-emitting lasers 221
cavity. This issue needs to be further addressed to avoid the use of low index
dielectric (e.g., SiO2, Si3N4) due to the poor thermal conductivities and the
strain-induced reliabilities issues.
On the other hand, new functional properties should also be expected
from photonic crystal surface-emitting membrane lasers, as the modulation
of photonic crystal cavity and the gain/loss modulation of the active media
can result in lasers with unique spectral, spatial, and beam properties.
Engineered beam shapes can be designed. Large-area single-mode lasers
are also feasible. Laser beams can also be steered with certain angles. With
the proper dispersion engineering, it is possible to have beam emission
in-plane.
Acknowledgments
The authors acknowledge contributions from their current and former students, as well as
collaborators involved in the research work reviewed here. Special thanks to Mattias
Hammar from Royal Institute of Technology who provided QW materials for the work.
We also thank the fabrication support from the University of Texas at Arlington
Nanotechnology Research Center, and the University of Texas at Austin, part of
NSF NNIN.
References
Altug, H., Englund, D., Vuckovic, J., 2006. Ultrafast photonic crystal nanocavity laser. Nat.
Phys. 2 (7), 484–488.
Babic, D., Corzine, S., 1922. Analytic expressions for the reflection delay, penetration depth,
and absorptance of quarter-wave dielectric mirrors. IEEE J. Quantum Electron.
28, 514–524.
Bakir, B.B., Seassal, C., Letartre, X., Viktorovitch, P., Zussy, M., Di Cioccio, L.,
Fedeli, J.M., 2006. Surface-emitting microlaser combining two-dimensional photonic
crystal membrane and vertical Bragg mirror. Appl. Phys. Lett. 88, 081113.
Balakrishnan, G., Jallipalli, A., Rotella, P., Huang, S., Khoshakhlagh, A., Amtout, A.,
Krishna, S., Dawson, L.R., Huffaker, D.L., 2006. Room-temperature optically pumped
(Al) GaSb vertical-cavity surface-emitting laser monolithically grown on an Si (100) sub-
strate. IEEE J. Sel. Top. Quantum Electron. 12 (6), 1636–1641.
Bhattacharya, P., Sabarinathan, J., Topol’ancik, J., Chakravarty, S., Yu, P., Zhou, W., 2005.
Quantum dot photonic crystal light sources. Proc. IEEE 93 (10), 1825–1838.
Boutami, S., Bakir, B., Regreny, P., Leclercq, J., Viktorovitch, P., 2007. Compact 1.55 μm
room-temperature optically pumped VCSEL using photonic crystal mirror. Electron.
Lett. 43, 282.
Chen, R., Tran, T.T.D., Ng, K.W., Ko, W.S., Chuang, L.C., Sedgwick, F.G.,
Chang-Hasnain, C., 2011. Nanolasers grown on silicon. Nat. Photonics 5 (3), 170–175.
Chua, S.L., Chong, Y., Stone, A.D., Soljacic, M., Bravo-Abad, J., 2011. Low-threshold las-
ing action in photonic crystal slabs enabled by Fano resonances. Opt. Express
19, 1539–1562.
Chua, S.L., Lu, L., Bravo-Abad, J., Joannopoulos, J.D., Soljacic, M., 2014. Larger-area
single-mode photonic crystal surface-emitting lasers enabled by an accidental Dirac
point. Opt. Lett. 39 (7), 2072–2075.
222 Weidong Zhou
Coldren, L., Corzine, S., 1995. Diode Lasers and Photonic Integrated Circuits. Wiley,
New York.
Colombelli, R., Srinivasan, K., Troccoli, M., Painter, O., Gmachl, C.F., Tennant, D.M.,
Sergent, A.M., Sivco, D.L., Cho, A.Y., Capasso, F., 2003. Quantum cascade surface-
emitting photonic crystal laser. Science 302 (5649), 1374–1377.
Dowling, J.P., Scalora, M., Bloemer, M.J., Bowden, C.M., 1994. The photonic band edge
laser: a new approach to gain enhancement. J. Appl. Phys. 75 (4), 1896–1899.
Ellis, B., Mayer, M.A., Shambat, G., Sarmiento, T., Harris, J., Haller, E.E., Vuckovic, J.,
2011. Ultralow-threshold electrically pumped quantum-dot photonic-crystal nanocavity
laser. Nat. Photonics 5 (5), 297–300.
Fan, S., Joannopoulos, J.D., 2002. Analysis of guided resonances in photonic crystal slabs.
Phys. Rev. B 65 (23), 235112.
Ferrara, J., Yang, W., Zhu, L., Qiao, P., Chang-Hasnain, C.J., 2015. Heterogeneously inte-
grated long-wavelength VCSEL using silicon high contrast grating on an SOI substrate.
Opt. Express 23, 2512–2523.
Flick, T., Becks, K.H., Dopke, J., M€attig, P., Tepel, P., 2011. Measurement of the thermal
resistance of VCSEL devices. J. Instrum. 6, C01021.
Furukawa, A., Sasaki, S., Hoshi, M., Matsuzono, A., Moritoh, K., Baba, T., 2004. High-
power single-mode vertical-cavity surface-emitting lasers with triangular holey structure.
Appl. Phys. Lett. 85, 5161–5163.
Hirose, K., Liang, Y., Kurosaka, Y., Watanabe, A., Sugiyama, T., Noda, S., 2014. Watt-class
high-power, high-beam-quality photonic-crystal lasers. Nat. Photonics 8 (5), 406–411.
Huang, M.C.Y., Zhou, Y., Chang-Hasnain, C.J., 2007. A surface-emitting laser incorporat-
ing a high-index-contrast subwavelength grating. Nat. Photonics 1 (2), 119–122.
Hwang, J., Ryu, H., Song, D., Han, I., Park, H., Jang, D., Lee, Y., 2000. Continuous room-
temperature operation of optically pumped two-dimensional photonic crystal lasers at
1.6 um. IEEE Photon. Technol. Lett. 12 (10), 1295–1297.
Iga, K., 2000. Surface-emitting laser-its birth and generation of new optoelectronics field.
IEEE J. Sel. Top. Quantum Electron. 6 (6), 1201–1215.
Imada, M., Noda, S., Chutinan, A., Tokuda, T., Murata, M., Sasaki, G., 1999. Coherent
two-dimensional lasing action in surface-emitting laser with triangular-lattice photonic
crystal structure. Appl. Phys. Lett. 75, 316.
Imada, M., Chutinan, A., Noda, S., Mochizuki, M., 2002. Multidirectionally distributed
feedback photonic crystal lasers. Phys. Rev. B 65 (19), 195306.
Karagodsky, V., Sedgwick, F.G., Chang-Hasnain, C.J., 2010. Theoretical analysis of
subwavelength high contrast grating reflectors. Opt. Express 18 (16), 16973–16988.
Karim, A., Bjorlin, S., Piprek, J., Bowers, J., 2000. Long-wavelength vertical-cavity lasers
and amplifiers. IEEE J. Sel. Top. Quantum Electron. 6 (6), 1244–1253.
Kim, J., Chrostowski, L., Bisaillon, E., Plant, D., 2007. DBR, sub-wavelength grating, and
photonic crystal slab Fabry–Perot cavity design using phase analysis by FDTD. Opt.
Express 15 (16), 10330–10339.
Kurosaka, Y., Iwahashi, S., Sakai, K., Miyai, E., Kunishi, W., Ohnishi, D., Noda, S., 2009.
Band structure observation of 2D photonic crystal with various V-shaped air-hole
arrangements. IEICE Electron. Exp. 6, 966–971.
Kwon, S.H., Kim, S.H., Kim, S.K., Lee, Y.H., 2004. Small, low-loss heterogeneous
photonic bandedge laser. Opt. Express 12 (12), 5356–5361.
Lee, J., Zhen, B., Chua, S.-L., Qiu, W., Joannopoulos, J.D., Soljacic, M., Shapira, O., 2012.
Observation and differentiation of unique high-Q optical resonances near zero wave
vector in macroscopic photonic crystal slabs. Phys. Rev. Lett. 109, 067401.
Liang, D., Roelkens, G., Baets, R., Bowers, J., 2010. Hybrid integrated platforms for silicon
photonics. Materials 3 (3), 1782–1802.
Photonic crystal surface-emitting lasers 223
Liang, Y., Peng, C., Sakai, K., Iwahashi, S., Noda, S., 2012. Three-dimensional coupled-
wave analysis for square-lattice photonic crystal surface emitting lasers with
transverse-electric polarization: finite-size effects. Opt. Express 20 (14), 15945–15961.
Lindberg, H., Strassner, M., Gerster, E., Bengtsson, J., Larsson, A., 2005. Thermal manage-
ment of optically pumped long-wavelength InP-based semiconductor disk lasers. IEEE J.
Sel. Top. Quantum Electron. 11, 1126–1134.
Liu, V., Fan, S., 2012. S4: a free electromagnetic solver for layered periodic structures.
Comput. Phys. Commun. 183 (10), 2233–2244.
Liu, S.-C., Zhao, D., Liu, Y., Yang, H., Sun, Y., Ma, Z., Reuterski€ old-Hedlund, C.,
Hammar, M., Zhou, W., 2017. Photonic crystal bandedge membrane lasers on silicon.
Appl. Optics 56 (31), H67–H73.
Liu, S.-C., Zhao, D., Ge, X., Reuterski€ old-Hedlund, C., Hammar, M., Fan, S., Ma, Z.,
Zhou, W., 2018. Size scaling of photonic crystal surface emitting lasers on silicon sub-
strates. IEEE Photonics J. 10 (3), 1–6.
Luk’yanchuk, B., Zheludev, N.I., Maier, S.A., Halas, N.J., Nordlander, P., Giessen, H.,
Chong, C.T., 2010. The Fano resonance in plasmonic nanostructures and metamaterials.
Nat. Mater. 9, 707–715.
Lundeberg, L.D., Boiko, D.L., Kapon, E., 2005. Coupled islands of photonic crystal hetero-
structures implemented with vertical-cavity surface-emitting lasers. Appl. Phys. Lett.
87, 241120.
Magnusson, R., Shokooh-Saremi, M., 2008. Physical basis for wideband resonant reflectors.
Opt. Express 16 (5), 3456–3462.
Martınez, L.J., Alen, B., Prieto, I., Galisteo-Lopez, J.F., Galli, M., Andreani, L.C.,
Seassal, C., Viktorovitch, P., Postigo, P.A., 2009. Two-dimensional surface emitting
photonic crystal laser with hybrid triangular-graphite structure. Opt. Express
17, 15043–15051.
Meier, M., Mekis, A., Dodabalapur, A., Timko, A., Slusher, R.E., Joannopoulos, J.D.,
Nalamasu, O., 1999. Laser action from two-dimensional distributed feedback in
photonic crystals. Appl. Phys. Lett. 74, 7–9.
Mi, Z., Bhattacharya, P., Yang, J., Pipe, K., 2005. Room-temperature self-organised
In0.5Ga0.5As quantum dot laser on silicon. Electron. Lett. 41 (13), 742–744.
Miyai, E., Sakai, K., Okano, T., Kunishi, W., Ohnishi, D., Noda, S., 2006. Photonics: lasers
producing tailored beams. Nature 441 (7096), 946.
Monat, C., Seassal, C., Letarte, X., Regreny, P., Gendry, M., Romeo, P.R.,
Viktorovitch, P., D’Yerville, M.L.V., Cassagne, D., Albert, J.P., Jalaguier, E.,
Pocas, S., Aspar, B., 2003. Two-dimensional hexagonal-shaped microcavities formed
in a two-dimensional photonic crystal on an InP membrane. J. Appl. Phys. 93 (1), 23.
Noda, S., 2010. Photonic crystal lasers—ultimate nanolasers and broad-area coherent lasers
[invited]. JOSA B 27 (11), B1–B8.
Noda, S., Yokoyama, M., Imada, M., Chutinan, A., Mochizuki, M., 2001. Polarization
mode control of two-dimensional photonic crystal laser by unit cell structure design.
Science 293 (5532), 1123–1125.
Noda, S., Fujita, M., Asano, T., 2007. Spontaneous-emission control by photonic crystals
and nanocavities. Nat. Photonics 1 (8), 449–458.
Oskooi, A.F., Roundy, D., Ibanescu, M., Bermel, P., Joannopoulos, J., Johnson, S.G., 2010.
MEEP: a flexible free-software package for electromagnetic simulations by the FDTD
method. Comput. Phys. Commun. 181 (3), 687–702.
Painter, O., 1999. Two-dimensional photonic band-gap defect mode. Science 284 (1819),
1999.
Painter, O., Lee, R.K., Scherer, A., Yariv, A., O’Brien, J.D., Dapkus, P.D., Kim, I., 1999.
Two-dimensional photonic band-gap defect mode laser. Science 284 (5421), 1819–1821.
224 Weidong Zhou
Park, H.-G., Kim, S.-H., Kwon, S.-H., Ju, Y.-G., Yang, J.-K., Baek, J.-H., Kim, S.-B.,
Lee, Y.-H., 2004. Electrically driven single-cell photonic crystal laser. Science
305 (5689), 1444–1447.
Park, H., Fang, A.W., Cohen, O., Jones, R., Paniccia, M.J., Bowers, J.E., 2006. Design and
fabrication of optically pumped hybrid silicon-AlGaInAs evanescent lasers. IEEE J. Sel.
Top. Quantum Electron. 12 (6), 1657–1663.
Qiang, Z., Yang, H., Chuwongin, S., Zhao, D., Ma, Z., Zhou, W., 2010. Design of Fano
broadband reflectors on SOI. IEEE Photon. Technol. Lett. 22 (15), 1108–1110.
Rapp, S., Piprek, J., Streubel, K., Andre, J., Wallin, J., 1997. Temperature sensitivity of
1.54-m vertical-cavity lasers with an InP-based Bragg reflector. IEEE J. Quantum
Electron. 33 (10), 1839.
Rapp, S., Salomonsson, F., Streubel, K., Mogg, S., Wennekes, F., Bentell, J., Hammar, M.,
1999. All-epitaxial single-fused 1.55μm vertical cavity laser based on an InP Bragg
reflector. Jpn. J. Appl. Phys 38, 1261–1264.
Roelkens, G., Liu, L., Liang, D., Jones, R., Fang, A., Koch, B., Bowers, J., 2010. III–V/
silicon photonics for on-chip and intra-chip optical interconnects. Laser Photonics
Rev. 4 (6), 751–779.
Sakai, K., Miyai, E., Sakaguchi, T., Ohnishi, D., Okano, T., Noda, S., 2005. Lasing
band-edge identification for a surface-emitting photonic crystal laser. IEEE J. Sel. Areas
Commun. 23 (7), 1335–1340.
Sakai, K., Miyai, E., Noda, S., 2006. Coupled-wave model for square-lattice two-
dimensional photonic crystal with transverse-electric-like mode. Appl. Phys. Lett.
89, 021101.
Sato, Y., Tanaka, Y., Upham, J., Takahashi, Y., Asano, T., Noda, S., 2012. Strong coupling
between distant photonic nanocavities and its dynamic control. Nat. Photonics 6, 56–61.
Sauvan, C., Hugonin, J., Lalanne, P., 2009. Difference between penetration and damping
lengths in photonic crystal mirrors. Appl. Phys. Lett. 95, 211101.
Sciancalepore, C., Bakir, B.B., Letartre, X., Harduin, J., Olivier, N., Seassal, C., Fedeli, J.,
Viktorovitch, P., 2012. CMOS-compatible ultra-compact 1.55-um emitting VCSELs
using double photonic crystal mirrors. IEEE Photon. Technol. Lett. 24 (6), 455–457.
Song, D.S., Kim, S.H., Park, H.G., Kim, C.K., Lee, Y.H., 2002. Single-fundamental-mode
photonic-crystal vertical-cavity surface-emitting lasers. Appl. Phys. Lett. 80 (21),
3901–3903.
Stankovic, S., Jones, R., Sysak, M.N., Heck, J.M., Roelkens, G., Van Thourhout, D., 2011.
1310-nm hybrid III–V/Si Fabry–Perot laser based on adhesive bonding. IEEE Photon.
Technol. Lett. 23 (23), 1781–1783.
Susa, N., 2001. Threshold gain and gain-enhancement due to distributed-feedback in
two-dimensional photonic-crystal lasers. J. Appl. Phys. 89 (2), 815–823.
Van Campenhout, J., Rojo Romeo, P., Regreny, P., Seassal, C., Van Thourhout, D.,
Verstuyft, S., Di Cioccio, L., Fedeli, J.M., Lagahe, C., Baets, R., 2007. Electrically
pumped InP-based microdisk lasers integrated with a nanophotonic silicon-on-insulator
waveguide circuit. Opt. Express 15 (11), 6744–6749.
Wu, S., Buckley, S., Schaibley, J.R., Feng, L., Yan, J., Mandrus, D.G., Hatami, F., Yao, W.,
Vuckovic, J., Majumdar, A., Xu, X., 2015. Monolayer semiconductor nanocavity lasers
with ultralow thresholds. Nature 520, 69–72. https://doi.org/10.1038/nature14290(.
Yablonovitch, E., 1987. Inhibited spontaneous emission in solid-state physics and electron-
ics. Phys. Rev. Lett. 58 (20), 2059–2062.
Yang, H., Chuwongin, S., Qiang, Z., Chen, L., Pang, H., Ma, Z., Zhou, W., 2009. Res-
onance control of membrane reflectors with effective index engineering. Appl. Phys.
Lett. 95, 023110.
Yang, H., Zhao, D., Chuwongin, S., Seo, J.-H., Yang, W., Shuai, Y., Berggren, J.,
Hammar, M., Ma, Z., Zhou, W., 2012. Transfer-printed stacked nanomembrane lasers
on silicon. Nat. Photonics 6 (9), 615–620.
Photonic crystal surface-emitting lasers 225
Yokouchi, N., Danner, A., Choquette, K.D., 2003. Two-dimensional photonic crystal con-
fined vertical-cavity surface-emitting lasers. IEEE J. Sel. Top. Quantum Electron. 9 (5),
1439–1445.
Yoshie, T., Scherer, A., Hendrickson, J., Khitrova, G., Gibbs, H., Rupper, G., Ell, C.,
Shchekin, O., Deppe, D., 2004. Vacuum Rabi splitting with a single quantum dot in
a photonic crystal nanocavity. Nature 432 (7014), 200–203.
Zhao, D., Ma, Z., Zhou, W., 2010. Field penetrations in photonic crystal Fano reflectors.
Opt. Express 18 (13), 14152–14158.
Zhao, D., Yang, H., Chuwongin, S., Seo, J., Ma, Z., Zhou, W., 2012. Design of photonic
crystal membrane-reflector-based VCSELs. IEEE Photonics J. 4 (6), 2169–2175.
Zhao, D., Liu, S., Yang, H., Ma, Z., Reuterski€ old-Hedlund, C., Hammar, M., Zhou, W.,
2016. Printed large-area single-mode photonic crystal bandedge surface-emitting lasers
on silicon. Sci. Rep. 6.
Zhou, W.D., Sabarinathan, J., Kochman, B., Berg, E., Qasaimeh, O., Pang, S.,
Bhattacharya, P., 2000. Electrically injected single-defect photonic bandgap surface-
emitting laser at room temperature. Electron. Lett. 36 (18), 1541–1542.
Zhou, W., Sabarinathan, J., Bhattacharya, P., Kochman, B., Berg, E.W., Yu, P.-C.,
Pang, S.W., 2001. Characteristics of a photonic bandgap single defect microcavity
electroluminescent device. IEEE J. Quantum Electron. 37 (9), 1153–1160.
Zhou, W., Zhao, D., Shuai, Y.-C., Yang, H., Chuwongin, S., Chadha, A., Seo, J.-H.,
Wang, K.X., Liu, V., Ma, Z., 2014. Progress in 2D photonic crystal Fano resonance
photonics. Prog. Quantum Electron. 38 (1), 1–74.
Index
Note: Page numbers followed by “f ” indicate figures, and “t” indicate tables.
A C
Amplitude, 6, 52, 95–97, 132, 142–143, Carcinoembryonic antigen (CEA), 120
151–152, 154 Chirped GMR approach, 141–142, 141f,
Angle-integrated light trapping absorption 145
enhancement factor, 49–50, 65–67, Common-path interferometry, 145
66f, 69–70, 70f, 72–76, 72f, 74–75f Controlled lattice displacement,
Artificial magnetic permeability, 19 165–167
Artificial structures, 15–16, 33 Coplanar waveguide (CPW), 173
Asymmetric photonic crystal slab, 6–7 CST Microwave Studio software, 20–21
Cylindrical Lorenz–Mie coefficients, 24
B
Bandgaps vs. rod permittivity, 27–28
Band structure, 2D photonic crystal, 63f D
density of TM states, 64, 64f Density of states (DOS), light trapping,
MPB package, 62–63 46–47, 51, 53–54
overlap factor of TM mode, 64–65, 64f elevated local DOS, 46
projected TM band structure, 63–64, 63f 3D photonic crystals, 74–76
Barcoding, 141–142 2D photonic crystals, 62–63
Bimodal waveguide sensor, 121, 128, absorption enhancement factor,
131–132, 140t 65–67
Biosensing and imaging, GMR sensor. density of TM states, 64, 64f
See Guided mode resonance (GMR) long wavelength limit, 68–70
sensor van Hove singularities, 67–68
Bloch theorem, 16–17 Dielectric metamaterials, 31–32
Bragg bandgaps, 16–17, 31–32, 36–37 benefit, 41–42
Bragg gaps, 17–18, 21–23, 26–27 one-dimensional structures, 18
Mie gaps, 32–33 with near-zero effective parameters,
resonant Bragg gap broadening, 23–24 33–35, 34–35f
swap, 19 with negative magnetic response, 14
symmetric, 32–33 Dielectric permittivity, 14, 27,
width, 18 37–38
Bragg mirror theory, 132 Dirichlet-to-Neumann (DTN) method,
Bragg resonances, 13–14, 20–23, 27, 31 64–65
Bragg scattering, 31 Double-layer PhC Fano resonance filters
on periodic structure, 16–17, 30–31 amplitude transmission coefficient, 154
periodicity, 31–32 controlled lattice displacement,
wavelength, 18 165–167
Brillouin zone, 16–17 coupled bright resonances, dark resonance
Bragg gap, 17–18 from, 154, 156f
minimum frequency, 18 coupled dark resonances, dark resonance
of square lattice, 20 from, 154, 155f
surface, 30–31 design parameters and Q’s, 160, 161t
Buried tunnel junction (BTJ), 218–219, 218f lattice parameters, 157–158, 157f
227
228 Index
F G
Fano resonance, in photonic crystal slabs Gaussian beam, 29–30, 29f, 34, 34f, 40, 41f
coupled double-layer filters, 152–153 Ge-Sb-Te (GST) alloy, 33
amplitude transmission coefficient, 154 Guided mode resonance (GMR) sensor, 2
controlled lattice displacement, advantages, 128, 132
165–167 vs. bimodal waveguides, 121, 128,
coupled bright resonances, dark 131–132, 140t
resonance from, 154, 156f Bragg mirror theory, 132
coupled dark resonances, dark chirped GMR, 141–142, 141f
resonance from, 154, 155f figures of merit
design parameters and Q’s, 160, 161t biochemical factors, 120–121
lattice parameters, 157–158, 157f electronic factors, 119–120
quartz substrates, 162–164, 163–164f limit of detection, 118–119
simulated transmission spectra, 160, photonic factors, 119
161–162f, 162 history, 117–118
Index 229
vs. microrings, 121, 128, 130–131, 140t incident and output field profile,
vs. nanohole arrays, 129–130, 140t Fourier transform of, 104–105
operation of isotropic band-pass filter, 108–111,
bulk sensitivity, 124, 127t 110f
optical signature, 121 isotropic band-reject filter, 108, 109f
polarization dependence, 122–123 isotropic high-pass filter, 105, 106f
Q-factor, 123–124 isotropic low-pass filter, 105–108, 107f
surface sensitivity, 124–127, 127t photonic crystal slab, guided resonance
phase-sensitive measurement dielectric constant, 97–98, 98f
phase noise, 142–143 direct transmission coefficient, 98–99
realization of, 143–145 filter transfer function, 99
reflectance and phase response, geometry parameters, 97–100, 98f
143–144, 144f Hamiltonian, 99
resonant hyperspectral imaging, 134 isotropic band-pass filter, 104
biofilms, early detection of, 134 isotropic band-reject filter, 104
hyperspectral cube, 134 isotropic high-pass filter, 103–104
multiparameter imaging, 137 isotropic low-pass filter, 103–104
plasmonic nanohole arrays, 137–139 nearly isotropic photonic band
sensing and imaging, 135–136 structure, 100–103, 101f
scanning illumination wavelength, single-band excitation effect, 100–103
133–134 square lattice of air holes, 97–99, 98f
spatial resolution dependence transmission and reflection coefficients,
on orientation, 133, 133f 102–103
refractive index contrast, 133 uniform dielectric slab, 97–99, 98f,
vs. SPR sensor, 128–129, 140t 102–103
surface affinity sensor, operation of, 116,
117f L
2D pixelated metasurface, 141–142 Lambertian limit, 46
Lateral cavity size scaling, 204–206
H Lattice constant vs. wavelength, 38–39
Hankel function, 19–20 Light-emitting devices, 191
Harmonic amplitude, 19–20 Light trapping, 45–47
Hermitian eigenvalue problem, 37–38 in bulk structure, 54–55
High-index-contrast gratings (HCGs), 2 cell efficiency, 45–46
High-index dielectrics meta-atoms, 14 conventional limits, in 2D and 3D
Hyperspectral imaging, GMR sensor. structures, 50, 50t
See Resonant hyperspectral imaging, in periodic structures, 55–56, 55f
GMR sensor in plasmonic structures, 60–61
ray optics theory, 47–50
I in thin films, 57–61
Image processing, 93–94. See also Optical in 3D photonic crystals
image processing “absorption calculation” curve, 76
Infrared absorbtion line analysis, 34 angle-integrated absorption
Interferometry, 142–145 enhancement factors, 73–76, 75f, 77f
Ion beam-assisted laser interference “theoretical bound” curve, 76
lithography, 97–98 vs. 2D photonic crystals, 74–75
Isotropic filters woodpile structure, 75, 76f
numerical demonstration in 2D photonic crystals
230 Index
W
V Wafer bonding technique, 208–209
van Hove singularity, 67–71, 73–76 Wave optics light trapping theory, 50–54
Vascular endothelial growth factor (VEGF), Wax protection process, 208
130 Wire bonding process, 218–219
Vertical-cavity surface-emitting laser Wood anomaly, 1–2, 117–118
(VCSEL), 191, 192–194t
broadband single-layer membrane Y
reflectors, 189–190 Yablonovitch limit, 46–50, 53–54, 67,
challenges, 190–191 79–85, 81f, 84f