ArticlePDF Available

# On-Chip Biological and Chemical Sensing With Reversed Fano Lineshape Enabled by Embedded Microring Resonators

## Abstract and Figures

High-$Q$ microresonators play an important role in developing fully integrated, highly sensitive, and cost-effective bio/chemical sensors. The Fano effect in doubly resonant physical systems may be used to improve sensing performance. In this paper, we show that coupled optical resonators (sometimes termed photonic molecules) in an embedded configuration can significantly enhance the sensitivity and limit of detection (LOD) of on-chip sensors by producing a reversed Fano effect. Improvement of one order in sensitivity, as compared to a sensor based on conventional Fano effect, can be achieved using embedded high- $Q$ resonators on a CMOS-compatible platform. We estimate the LOD by taking into account thermal drift, optical losses (material absorption, scattering, substrate leakage and bending loss), laser intensity noise, linewidth and frequency jitter, and link and detector signal-to-noise ratio (SNR). The overall LOD is found to be as low as 3.24 × 10$^{-8}$ RIU. Moreover, in the proposed sensor based on embedded rings, intensity SNR is no longer the limiting factor of the LOD, which could be further lowered with better thermal control and laser frequency stability.
Content may be subject to copyright.
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 20, NO. 3, MAY/JUNE 2014 5200110
On-Chip Biological and Chemical Sensing With
Reversed Fano Lineshape Enabled by
Embedded Microring Resonators
Xiaoyan Zhou, Lin Zhang, Member, IEEE, Andrea M. Armani, Senior Member, IEEE, Daihua Zhang,
Xuexin Duan, Jing Liu, Hao Zhang, and Wei Pang
Abstract—High-Qmicroresonators play an important role in
developing fully integrated, highly sensitive, and cost-effective
bio/chemical sensors. The Fano effect in doubly resonant physical
systems may be used to improve sensing performance. In this paper,
we show that coupled optical resonators (sometimes termed pho-
tonic molecules) in an embedded conﬁguration can signiﬁcantly
enhance the sensitivity and limit of detection (LOD) of on-chip
sensors by producing a reversed Fano effect. Improvement of one
order in sensitivity, as compared to a sensor based on conventional
Fano effect, can be achieved using embedded high-Qresonators
on a CMOS-compatible platform. We estimate the LOD by taking
into account thermal drift, optical losses (material absorption, scat-
tering, substrate leakage and bending loss), laser intensity noise,
linewidth and frequency jitter, and link and detector signal-to-
noise ratio (SNR). The overall LOD is found to be as low as 3.24 ×
108RIU. Moreover, in the proposed sensor based on embedded
rings, intensity SNR is no longer the limiting factor of the LOD,
which could be further lowered with better thermal control and
laser frequency stability.
Index Terms—Biological and chemical sensor, double ring res-
onators, Fano effect, and microring resonators.
I. INTRODUCTION
THE development of biochemical sensors based on
integrated optical micro-resonators has been the focus
of signiﬁcant research efforts over the past few decades. The
motivation for pursuing this technology is the ability to si-
multaneously achieve high sensitivity, low power consumption,
and label-free detection [1]–[5]. Various device structures have
been explored, including Fabry–Perot (FP) resonators [6], [7],
Manuscript received August 8, 2013; revised September 18, 2013, November
21, 2013, and December 2, 2013; accepted December 2, 2013. This work was
supported in part by the Natural Science Foundation of China (NSFC No.
61006074 and No. 61176106) and in part by the NIH Director’s New Innovator
Award Program [1DP2OD007391-01].
X. Zhou, D. Zhang, X. Duan, J. Liu, H. Zhang, and W. Pang are with the
State Key Laboratory of Precision Measuring Technology and Instruments,
Tianjin University, Tianjin 300072, China (e-mail: weipang@tju.edu.cn).
L. Zhang is with the Microphotonics Center and Department of Materials
Science and Engineering, Massachusetts Institute of Technology, Cambridge,
MA 02139, USA (e-mail: linzhang@mit.edu).
A. M. Armani is with the Department of Chemical Engineering and Materials
Science, University of Southern California, Los Angeles, CA 90089, USA
(e-mail: armani@usc.edu).
Color versions of one or more of the ﬁgures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identiﬁer 10.1109/JSTQE.2013.2294465
microspheres [8], [9], microdisks/microtoroids [10], [11], mi-
crorings [12]–[20], and photonic crystal cavities [21], [22].
Among them, microring resonators exhibit compact size, fully
on-chip device operation, potential to be densely integrated,
multi-channel sensing capability, cost-effectiveness, straightfor-
ward fabrication methods, and the high refractive index con-
trast [20], [23], [24]. Additionally, there is signiﬁcant ﬂexibility
in cavity design when using microrings, which allows alterna-
tive conﬁgurations to be explored. For example, one can cas-
cade microrings in an embedded conﬁguration [25]–[27], as
shown in Fig. 1(a), which is difﬁcult to achieve using other
types of micro-resonators. Although embedded rings exhibit
some unique advantages in optical delay lines [25], modula-
tors [25], and nonlinear components [27], their application in
the sensing domain has not been explored.
The primary method for performing detection using optical
microresonators is to use the evanescent ﬁeld of the cavity to
sample the surrounding environment. As the refractive index
of the environment changes, due to global changes in the solu-
tion refractive index as an analyte binds to the device surface,
the resonant wavelength of the resonator shifts. By detecting
optical signals, one can quantify the amount of analyte in the
solution [3], [4]. Generally, there are two detection methods:
1) monitoring the resonant wavelength shift or 2) detecting the
optical intensity change at a ﬁxed wavelength. Resonance-shift-
based detection offers a larger dynamic range, but may require
a complex experiment setup since wavelength sweeping is typi-
cally involved [23]. On the other hand, a detection scheme based
on intensity monitoring has a simpler setup, enables real-time
readout, and, most importantly, shows potential advantages in
reaching a lower limit of detection (LOD) [13].
In both detection methods, the quality (Q)factor of a micro-
resonator is an important ﬁgure-of-merit. As the cavity Qin-
creases, the uncertainty in the measurement of the wavelength
decreases, improving the performance in wavelength-shift based
sensing schemes. On the other hand, in the intensity detection
approach, a larger Qor a sharper roll-off of the resonance line-
shape can be directly converted to higher sensitivity. Intensive
research efforts have been made to increase the intensity-based
sensitivity through reduction of loss [2], [13], adjusting the re-
lation of coupling and loss [19], and exploiting novel physi-
cal effects, such as electromagnetically induced transparency
(EIT) [28] and Fano [14], [16] effects.
A Fano resonance has a characteristic asymmetric sharp line-
shape, which can be leveraged to improve the sensitivity of
1077-260X © 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
5200110 IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 20, NO. 3, MAY/JUNE 2014
Fig. 1. (a) Schematic of the embedded-ring resonators. (b) Fano effect in embedded-ring resonators and detection method based on intensity change at aﬁxed
wavelength. (c) On-chip sensing scheme with integration of embedded-ring sensor and microﬂuidic channel.
intensity-based sensing. The asymmetry originates from the
interference of a continuum and a discrete state [29], [30].
Although different types of Fano spectral features have been
observed [31], [32], there is little reported on how these Fano
effects inﬂuence device performance, especially for sensing
applications. One can produce the Fano effect using various
micro-resonator structures, such as a micro-resonator incorpo-
rating two partially reﬂective elements serving as a FP cavity
[14], [16], [33], [34], a Mach–Zehnder interferometer coupled
with a microcavity [35], [36], coupled resonators [25], [37], and
a single resonator with two modes interacting [38]. However,
the two types of Fano spectral features are rarely shown and
discussed.
Inspired by the work in [27] where different EIT effects with
dramatically varied spectral phases are produced, we show that
an embedded-ring sensor exhibits two types of Fano lineshapes.
We analyze and compare the two Fano spectral features for
bio-chemical sensing applications. The reversed Fano lineshape
exhibits a much larger slope than the conventional Fano line-
shape. The sensing performance is signiﬁcantly enhanced using
the reversed Fano lineshape, with one order improvement in
sensitivity compared to the conventional Fano lineshape ap-
proach. The LOD of our sensor calculated with considerations
for various noise sources is on the order of 108RIU, which is
comparable to the best result in literature.
II. SENSITIVITY AND LIMIT OF DETECTION
The performance of biochemical sensors is often character-
ized by two metrics: sensitivity and LOD. The sensitivity is
deﬁned as the amount of wavelength shift or intensity change
when the sample refractive index is changed by one refractive
index unit (RIU) [1]. The sensitivity can be separated into two
terms: the device sensitivity and the waveguide sensitivity. The
device sensitivity is closely related to the conﬁguration of the
resonator, and the waveguide sensitivity is determined only by
the waveguide cross-section and related material properties.
Here, in intensity-based sensing scheme, the overall sensitiv-
ity can be expressed as:
S=S1·S2·S3=dI
dλ·dλ
dneﬀ ·dneﬀ
dnsam
(1)
where dI is the intensity change of optical signal, dλis the
resonant wavelength shift of the device, and dneﬀ and dnsam
Fig. 2. Two types of Fano lineshapes in embedded-ring resonators by differ-
entiating the parity of the difference between the mode numbers of two rings
(m1m2=even number or odd number). The ring-waveguide coupling co-
efﬁcients are t1=0.6 and t3=0.5. In (a) and (b), the ring–ring coupling coef-
ﬁcients are t2=t4=0.03, and m1=900, m2=680 0.07 for (a) and (b),
respectively. In (c) and (d), t2=t4=0.2, and m1=901, m2=680 0.07
for (c) and (d), respectively.
are the refractive index change of the guided mode and the
sample, respectively.
Based on the above analysis, S1·S2is the device sensitivity,
and S3is the waveguide sensitivity. S1is the ratio of the intensity
change caused by the resonant wavelength shift. The Fano effect
improves S1because of the high intensity contrast (dI) within
the extremely small wavelength range (dλ) which is enabled by
the asymmetric resonance lineshape, as shown in Fig. 1(b). S2
represents the resonant wavelength shift due to the change in
neﬀ . In sensing based on intensity detection, S2approximately
equals λ/neﬀ since a narrow band is considered [39]. S2is
almost a constant and is determined by the working wavelength,
waveguide material, and waveguide cross-section. S3is deﬁned
as the shift in neﬀ caused by the change in nsam , and can be
optimized through waveguide cross-section design.
The LOD represents the minimal detectable amount of ana-
lyte. In intensity-based detection, LOD can be expressed as [5]:
LOD =δI
S(2)
where δI is the overall intensity uncertainty of the sensor limited
by system noises.
III. TWO TYPES OF FANO LINESHAPES AND DEVICE
SENSITIVITY (S1)OPTIMIZATION
Although the Fano effect has been seen in various resonant
microstructures with a lineshape shown in Figs. 2(a) and (b)
ZHOU et al.: ON-CHIP BIOLOGICAL AND CHEMICAL SENSING WITH REVERSED FANO LINESHAPE ENABLED 5200110
[33]–[38], the reversed Fano lineshape [see Figs. 2(c) and (d)]
is rarely observed. We ﬁnd that the embedded ring resonators
in Fig. 1(a) exhibit two types of Fano spectral responses, when
varying the relationship between the mode numbers, m1and
m2, of an outer ring (Ring 1) and an inner ring (Ring 2). They
are coupled with each other in four regions, with amplitude
coupling coefﬁcients labeled in Fig. 1(a) as t1,t
2,t
3, and t4,
respectively. Light is launched into the embedded rings from
the “In” port, and signal is detected at the “Out” port, as shown
in Fig. 1(a).
To perform detection, the entire embedded ring structure
would be integrated within a microﬂuidic channel, as shown
in Fig. 1(c). This approach will allow efﬁcient delivery of an
analyte solution to the microring surface for high sensitivity
detection. In the present paper, the optical sensor is based on a
Si3N4-on-SiO2material platform to reduce the optical loss [40].
The working wavelength is 780 nm to optimize the relative ab-
sorption losses of water [41] and Si3N4, which is detailed in
Section IV. The mode numbers, m1and m2, are set to be ap-
proximately 900 and 681, respectively, corresponding to a radius
of 57 μm of Ring 2. Coupled mode theory [42] is used to ana-
lyze the transfer characteristics of the embedded rings (see [27]
for more details).
If m1and m2are both integers, the resonant wavelengths
of Ring 1 and Ring 2 are aligned, and EIT-like responses or
mode splitting are seen [25], [27]. When the resonant wave-
lengths of Ring 1 and Ring 2 are offset, we observe asymmetric
Fano lineshapes, as shown in Fig. 2. The Fano effect in the em-
bedded rings results from the interference between the optical
ﬁelds in Ring 1 and Ring 2, where the resonance of Ring 1
serves as the continuum and the resonance of Ring 2 is analo-
gous to a discrete state [29], [30]. For example, Fig. 2(a) shows
that a Fano lineshape occurs on the short wavelength side of the
wide Lorentzian resonance by setting m1=900 and m2=679.93.
Conversely, the Fano lineshape appears on the long wavelength
side of the Lorentzian with m2changed to 680.07, as shown
in Fig. 2(b). This indicates that the wide Lorentzian is the res-
onance of Ring 1 and the position of the Fano lineshape is
changed by Ring 2.
Two types of Fano spectral responses are obtained in the cases
of m1m2even number and m1m2odd number. In
Figs. 2(a) and (b), where m1m2even number, we ﬁrst see a
peak and then a valley in both Fano lineshapes looking from the
central wavelength (which is the resonant wavelength of Ring
1), indicating that Ring 1 interferes with Ring 2 constructively
between their resonant wavelengths. This type of Fano lineshape
is frequently reported in the literature [33]–[38], and we name it
Type I. In contrast, the Fano lineshapes in the case of m1m2
odd number is reversed from Type I, exhibiting ﬁrst a valley
and then a peak away from the central wavelength, as shown
in Figs. 2(c) and (d), where m1=901, and m2= 680 0.07,
respectively. This reversal means that Ring 1 and Ring 2 interfere
destructively near their resonances, and we call it Type II. Note
that we have adjusted the coupling coefﬁcients to show similar
linewidths of Type I and II in Fig. 2 for illustration purposes only.
With the same set of coupling coefﬁcients, the Type I and Type
II spectral responses show distinctly different slopes, which re-
Fig. 3. Device sensitivity (S1)as a function of mode number of Ring 1 (m1)
with loss coefﬁcients 0.1, 0.2, 0.3, 0.5, and 1.0 dB/cm considered. The mode
number of Ring 2 is m2=681. The ring–ring coupling coefﬁcients are t2=t4
=0.1. In (a), the ring-waveguide coupling coefﬁcients are t1=0.5 and t3=
0.6, and t1and t3are switched in (b) with t1=0.6 and t3=0.5.
sults in highly varied S1for sensing. The change in Fano line-
shape and sensitivity between the two types of Fano lineshapes
is analyzed by setting m2=681 and changing m1from 900 to
901. In this process, the Fano lineshape changes from Type II
to Type I. To ensure the sensor works in the linear range of the
Fano lineshape, we deﬁne dI as the intensity contrast between
the points near the Fano peak and valley, which have a slope
of 0.1 times that of the maximum Fano slope, and dλas the
corresponding wavelength range between the two points. The
coupling coefﬁcients are set to be t1=0.5, t3=0.6, t2=t4=
0.1. In Fig. 3(a), a family of curves are obtained with differ-
ent losses. By switching t1and t3, i.e., t1=0.6 and t3=0.5,
similar curves are plotted in Fig. 3(b), which generally show an
enhanced S1when compared to Fig. 3(a) in all loss cases. For a
small loss, the sensitivity, S1, of the Fano effect near m1=900
(m1m2odd number, i.e., Type II) is almost 15 times larger
than that near m1=901 (m1m2even number, i.e., Type I).
As the loss increases, the S1in both Fano types decreases, with
S1of Type II decreasing more rapidly. When the loss reaches
1 dB/cm, the curve becomes quite ﬂat at the left part, and the
advantage of using Type II Fano lineshape is less apparent.
Based on the discussion of the loss mechanisms (detailed in
Section IV), the overall loss coefﬁcient of our device is estimated
to be 0.2 dB/cm. Therefore, we chose the curve with 0.2 dB/cm
in Fig. 3(b) and illustrate the evolution of Fano lineshapes at
several points, denoted by A to F, between m1=900 and
m1=901 in Fig. 4. The background resonance absorption line
5200110 IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 20, NO. 3, MAY/JUNE 2014
Fig. 4. Fano lineshapes at points A–F in Fig. 3(b).
of Ring 1 clearly shifts as m1is varied, which also results in a
small shift of the Fano resonant wavelength. Additionally, the
Fano lineshape becomes broader as m1changes from 900 to
901, that is, dλincreases from Type II to Type I, which explains
the decrease of S1from Type II to Type I. On the other hand,
when m1approaches 900, the Fano effect shows a decreasing
intensity contrast (dI) (Fano lineshapes of points A and B),
which eventually overturns the trend of S1and results in a peak
near m1=900 in Fig. 3.
Based on this analysis, we chose Fano Type II with m1=900.1
and m2=681 (point B in Fig. 3) as the working point of our
sensor. This set of conditions is ideally suited for detection
because it is slightly on the right of the peak of S1curve, which
possesses both a large S1of 1045.08 /nm and a high intensity
contrast with dI =0.7. A Fano Type I sensor with a similar
conﬁguration (m1=900.9, point F in Fig. 3) shows a S1of
109.21 /nm. Comparing these two types of Fano effects, we note
that there is one order enhancement of S1in Type II over Type I.
It is important to emphasize that the S1of Type I can result in a
large overall sensitivity and a low LOD as analyzed in Section V.
This performance is comparable with works of others (Table I).
The improvement in S1in Fano Type II obtained in our paper
represents an opportunity to further advance this approach.
The changes in S1from the Fano Type II to the Type I un-
der different ring-ring coupling coefﬁcients are also analyzed in
Fig. 5 with a loss factor of 0.2 dB/cm. The ring-waveguide cou-
pling coefﬁcients are set to be t1=0.5 and t3=0.6 in Fig. 5(a),
and are switched in Fig. 5(b). Comparing Figs. 5(a) and (b) with
the same ring–ring coupling coefﬁcients, one ﬁnds that S1is im-
proved in Fig. 5(b) for all m1. Thus, a lager t1is preferred. When
t2and t4decrease from 0.14 to 0.08, S1for all m1increases, as
shown in both Figs. 5(a) and (b). This increase is the result of the
Fano linewidth narrowing as the ring-ring coupling coefﬁcients
decrease. As t2and t4further decrease from 0.08 to 0.06, S1
near m1=901 (Type I) continues to increase, but S1near m1
=900 (Type II) decreases. The reason for this difference can be
explained as follows. With the decrease of t2and t4, the Fano
linewidth narrows, and both types of Fano spectral responses are
becoming more sensitive to loss, resulting in a smaller intensity
contrast (dI). Since the linewidth of Type II is much narrower
than Type I, the decrease in dI for Type II is more severe than
that in Type I. This rapid reduction in dI eventually overturns the
increasing trend of S1created by the linewidth narrowing.
Therefore, it is important to choose a reasonable t2and t4to
avoid both small S1and dI.
IV. WAVEGUIDE CROSS-SECTION OPTIMIZATION
Waveguide structure design for ultra-sensitive optical sensors
requires general considerations for sensitivity (S3)improve-
ment, thermal shift noise control, and loss reduction. Obviously,
as a contributor to the overall sensitivity, S3is an important
ﬁgure-of-merit in waveguide design, and many previous works
concentrated on optimizing the waveguide to achieve a large
S3[15], [43]. Nevertheless, maintaining a low optical loss is
crucial to obtain a large device sensitivity (S1), and the thermal
noise could limit the LOD of the system. Our paper takes all
the three factors into consideration, with emphasis on obtaining
thermal stability and reducing optical loss.
A. Thermal Stability
For biosensors with high sensitivity, thermal noise could be a
dominant issue. Because of the thermo–optic effect, the change
in temperature will shift the resonant wavelength of the optical
resonator, which causes additional intensity changes, resulting
in false-positive signals. Therefore, it is highly desirable to make
the device athermal or temperature insensitive.
One way to reduce the thermal intensity shift is to construct
a waveguide using materials with opposite thermo–optic coefﬁ-
cients. The waveguide cross-section is shown in Fig. 6(a), with a
strip Si3N4waveguide on a SiO2layer. Since the thermo–optic
coefﬁcients of Si3N4(dnSi3N4/dT =4×105/K) and water
(dnwater/dT =8×105/K) are of opposite signs [44], [45],
the refractive index change resulting from a temperature shift
could be balanced to eliminate thermal drift by shrinking the
waveguide cross-section to squeeze about one third of light into
the sample solution.
Theoretically, numerous waveguide cross-section structures
could be constructed to realize this neutrality condition. We em-
ploy a high-aspect-ratio waveguide structure [46] with consid-
eration on scattering loss reduction, which is detailed in Section
IV-B. Five structures with heights varying from 60 to 90 nm and
widths varying from 1500 to 2500 nm are considered for ther-
mal stability optimizations. The material refractive indices used
are nSi3N4 =1.9974, nSiO2 =1.4537, nSi =3.6352, and nwater
=1.3298 at the wavelength of 780 nm. The thermal ﬂuctuation
(ΔT) range explored was 0 to 0.01 K. The quasi-TE mode is
considered, and the waveguide structure is optimized using a
ﬁnite-element-method (FEM) solver.
The change in the effective refractive index (neﬀ )of the
waveguide resulting from the thermal ﬂuctuations (ΔT) is il-
lustrated in Fig. 6(c). From this graph, we can determine that the
thermal stability is affected primarily by the waveguide height.
ZHOU et al.: ON-CHIP BIOLOGICAL AND CHEMICAL SENSING WITH REVERSED FANO LINESHAPE ENABLED 5200110
Fig. 5. Device sensitivity (S1)as a function of mode number of Ring 1 (m1)
with ring–ring coupling coefﬁcients t2=t4=0.06, 0.08, 0.1, 0.12, and 0.14
considered. The mode number of Ring 2 is m2=681. The loss coefﬁcient is
set to be 0.2 dB/cm. In (a), the ring-waveguide coupling coefﬁcients are t1=
0.5 and t3=0.6, and t1and t3are switched in (b) with t1=0.6 and t3=0.5.
Speciﬁcally, a waveguide with a larger height has an increased
change in neﬀ with temperature. This phenomenon also holds for
width variations, although the change rate in neﬀ is smaller. The
waveguide with cross-section of 70 ×2000 nm2(height ×
width) shows the best thermal stability, and the optical ﬁeld
distribution is balanced between the waveguide core and the
sampled solution. Taking into account both S3improvement (de-
tailed in Section IV-C) and thermal noise reduction, we choose
the waveguide structure to be 70 ×2000 nm2(height ×width)
and show the power distribution in Fig. 6(b). The thermal noise
in this case is 1.48 ×1010 RIU in neﬀ for ΔT=0.01 K.
Considering a fabrication error of ±20 nm in width and ±2nm
in height, the thermal noise is still quite low, with Δneﬀ =3.32
×109RIU for ΔT=0.01 K.
B. Optical Loss
Optical loss affects the sharpness of the Fano lineshape di-
rectly. As indicated in Fig. 3, performance of the sensor dete-
riorates quickly as optical loss increases, and thus maintaining
a low optical loss is critical. An important issue in waveguide
design is associated with reducing sidewall scattering, which is
the primary contributor to the overall optical loss.
For the thermal stabilization detailed in Section IV-A, about
one third of the optical power should be distributed in the sample.
Typically, this will unavoidably lead to an increased overlap of
the optical power with the sidewalls, resulting in signiﬁcant light
Fig. 6. (a) Waveguide cross-section structure of the embedded rings. (b) Opti-
cal intensity distribution of quasi-TE mode in the cross-section of the waveguide
with width of 2000 nm and height of 70 nm. (c) Effective index changes (Δneﬀ)
of the waveguide as a function of the thermal change (ΔT). The cross-sections
are denoted by height ×width.
scattering. Instead, we use a high-aspect-ratio waveguide with
the width much larger than the height to squeeze light into the
ambient solution mainly through the waveguide top surface [see
Fig. 6(b)] to maintain a low scattering loss.
We need to evaluate the scattering loss at 780 nm in the high-
aspect-ratio waveguides that were originally designed for loss
reduction in near infrared [46]. Scattering loss can be calculated
using the theoretical model developed in [47]. This model has
been experimentally veriﬁed by various groups [46], [48], [49].
We use the mean square deviation roughness and correlation
length of the roughness for sidewalls, given in [46], (σside,L
side)
=(3.16 nm, 50 nm), and use those for waveguide upper and
lower surfaces (σsurf ,L
surf )=(0.1 nm, 30 nm) [46]. Since
the model is for a planar optical waveguide, we employ an ef-
fective mode index method to obtain the effective indices to
replace the waveguide indices in the model for both sidewall
and surface scattering loss calculations. The overall scatter-
ing loss is 0.081 dB/cm for a waveguide with a cross-section
of 70 ×2000 nm2(height ×width). This value includes the
contributions from sidewall scattering and surface scattering of
0.031 dB/cm and 0.05 dB/cm, respectively.
There are other loss mechanisms contributing to the overall
loss coefﬁcient in the device, including absorption loss, leakage
loss and bending loss. By adding the negative imaginary parts
(k)to the material refractive indices in the previously developed
cross-section FEM model, the absorption loss can be obtained.
From [41] the material loss of water can be found: kwater =
1.434 ×107at the wavelength of 780 nm. The absorption
loss of a Si3N4strip waveguide is reported to be 0.055 dB/cm at
1550 nm [40], and a thin slab waveguide of Si3N4showslossless
than 0.1 dB/cm when operated in TE0mode at a wavelength of
632.8 nm [50]. Based on these values, the material loss of Si3N4
5200110 IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 20, NO. 3, MAY/JUNE 2014
Fig. 7. Effective index changes (Δneﬀ) of the waveguide as a function of the
sample index change (Δnsam ).
should not exceed 0.1 dB/cm at 780 nm, which corresponds to
kSi3N4 =1.429 ×107. The absorption loss of our waveguide
is calculated to be 0.051 dB/cm using an FEM mode solver.
Leakage loss results from the leakage of optical power into
the Si substrate, which can be made negligible by increasing the
thickness of the SiO2layer. Our simulation shows that with a
SiO2layer thickness greater than 3.5 μm, leakage loss is on the
order of 105dB/cm. Bending loss arises from radiation loss
at the bends in the waveguide and decreases quickly when the
bending radius is increased. Previous work has reported that a
Si3N4microring encapsulated in SiO2layers with a radius of
20 μm, height of 644 nm, and width of 900 nm has a loss of
0.065 dB/cm, including scattering loss and bending loss near
1550 nm [40]. Our microring, with a radius of 57 μm, should
have a bending loss at 780 nm smaller than or comparable to that
in [40]. Here, the bending loss of our waveguide is estimated to
be 0.065 dB/cm.
Taking all of the loss mechanisms into account, the overall
loss coefﬁcient in our design is 0.2 dB/cm. Further discussion
on the role of the optical loss can be found in Section V.
C. Waveguide Sensitivity Optimization
In order to achieve a large waveguide sensitivity (S3)in the
embedded-ring sensor, the ﬁve structures in Fig. 6(c) are con-
sidered. The neﬀ of the waveguide is calculated for a sample
analyte refractive index (nsam)changed from nsam =nwater =
1.3298 to nsam =nwater +1×106=1.329801, as shown in
Fig. 7. The change in neﬀ as a function of changes in nsam is
plotted, and the response is very linear over this refractive index
range. S3is deﬁned as the slope of the lines. Generally, the
difference in S3for various structures is very small. From the
inset in Fig. 7, we note that the effect of the waveguide height on
S3is stronger than that of the width, and decreasing the height
from 90 to 70 nm results in an increase in S3. Reducing the
waveguide width also helps to improve S3slightly, as shown in
Fig. 7. This improvement is because as the cross-section of the
waveguide is reduced, the optical ﬁeld is less conﬁned within the
waveguide, which increases the interaction of the optical ﬁeld
with the environment. Further decreasing the height from 70 to
60 nm results in a decrease in S3, which is attributed to quickly
increased ﬁeld extension into the SiO2. Among all of the cases,
the waveguides with height ×width =70 ×2000 nm2and 70
×2500 nm2offer the largest S3of 0.212.
V. OVERALL SENSITIVITY AND LIMIT OF DETECTION
Combining the design parameters with optimized S1and S3
in Sections III and IV, we have an embedded microring sensor
with a cross-section of 70 ×2000 nm2(height ×width) working
in Fano Type II regime (m1=900.1 and m2=681, point B in
Fig. 4). This structure shows a high S1of 1045.08 /nm and a S3
of 0.212. The S2of this sensor is calculated using S2=λ/neﬀ .
Therefore, the overall sensitivity of our device is:
S=S1·S2·S3= 1045.08 ·780
1.488 ·0.212
=1.16 ×105/RIU
In contrast, using embedded rings working in Fano Type I
regime with similar conﬁguration (m1=900.9 and m2=681,
point F in Fig. 4), the overall sensitivity is only 1.21×104/RIU,
which is one order smaller than that in Fano Type II.
There are intrinsic (device) and extrinsic (detection system)
noise sources in an optical cavity that limit the LOD of the
sensor. For example, extrinsic noise includes temperature shifts,
relative intensity noise (RIN) and phase noise of the laser, and
shot and Johnson noises of the detector, whereas intrinsic noise
is related to the temperature variation of the cavity [51]. Closely
related to the speciﬁcs of the sensing system, the LOD can vary
greatly even with similar sensitivity. Using lasers and detectors
with high resolution can reduce the LOD of the system. In
addition, novel detection methods and statistical analysis have
been proposed to minimize the inﬂuence of the laser noise [52],
[53].
Generally speaking, these noise sources could be categorized
into three types based on the mechanisms through which they
affect and limit the LOD of the sensor, namely the thermal drift
noise, intensity noise, and spectral noise.
The thermal shift noise has been discussed in detail in Section
IV, and the resultant neﬀ error can be converted to the LOD
though
LODthermal =Δneﬀ therm al
S3
(3)
where Δneﬀ thermal is the shift in neﬀ due to temperature ﬂuc-
tuation, which is 3.32 ×109RIU. Thermally limited LOD in
our structure LODthermal =1.57 ×108RIU for ΔT=0.01 K.
Intensity noise includes the RIN noise of the laser and signal-
to-noise ratio (SNR) of the photonic links and the detector,
affecting the intensity uncertainty (δI in (2)) directly by
LODintensity =δIintensity
S.(4)
We calculate the LOD of the sensor under 30-dB SNR
[16], and intensity-noise-limited LOD for this sensing system
LODintensity =8.62 ×109RIU for Fano type II.
Spectral noise consists mainly of the laser linewidth and the
laser frequency jitter. Since the laser linewidth can be reduced
down to 1.8 kHz (0.02 fm) [54], [55] with a recently proposed
phase-noise cancellation scheme for semiconductor lasers, it is
ZHOU et al.: ON-CHIP BIOLOGICAL AND CHEMICAL SENSING WITH REVERSED FANO LINESHAPE ENABLED 5200110
TAB L E I
LOD COMPARISON BETWEEN THIS PAPER AND PUBLISHED PAPERS
negligible compared with laser frequency jitter, which is on the
order of 1–10 fm. The spectral noise can be converted to the
intensity uncertainty in (2) through δIspectral λ/S1, where
Δλis the spectral uncertainty, and the LOD limited by spectral
noise
LODspectral =δIspectral
S=Δλ
S2·S3
.(5)
LODspectral is 2.7 ×108RIU for Δλ=3fm.
The overall LOD can be calculated by taking the root mean
square of each noise components [51]
LOD = LOD2
thermal +LOD
2
spectral +LOD
2
intensity .(6)
Based on this analysis, the ﬁnal LOD for a sensor based on
a Fano Type II structure is 3.24 ×108RIU with (ΔT,Δλ,
SNR) =(0.01 K, 3 fm, 30 dB). Note that LODintensity is rather
low compared with LODthermal and LODspectral, which is at-
tributed to the high sensitivity brought by reversed Fano effect.
Under the same noise level, the LOD for Fano Type I sensor
is 8.81 ×108RIU, which is more than twice that of Type
II. LODthermal and LODspectral for the two Fano types are the
same, but the LODintensity for a Fano type I sensor is 8.24 ×
108RIU due to a smaller Sand becomes the limiting factor of
the overall LOD.
These results are compared with published works in Table I.
It is important to note that previous works with intensity-based
detection include only the intensity noise when calculating the
LOD. In our paper, with many other noise sources considered,
the obtained LOD using Fano Type II can be as low as the best
result in previously published results [16].
Reducing the laser spectral noise and improving the thermal
control ability can further lower the overall LOD of the sensor.
Fig. 8 shows the LOD using Fano Types I and II for spectral
noise ranging from 1 to 10 fm. Thermal ﬂuctuations of 0.01
and 0.005 K are considered. The overall LOD decreases when
spectral noise is reduced for both types, but the LOD for Type
I is limited to be above 8 ×108RIU by the intensity noise.
The effect of thermal noise becomes dominant at low spectral
noise level for Type II. From Fig. 8, one can see that using
the proposed Fano Type II resonator signiﬁcantly reduces the
inﬂuence of the intensity noise, making it possible to further
improve the overall LOD by using better lasers and temperature
Fig. 8. LOD of Fano type II for various spectral noise values with thermal
ﬂuctuation ΔT=0.01 K and ΔT=0.005 K. SNR is 30 dB.
Fig. 9. Sensitivity and LOD as a function of loss in two types of Fano line-
shapes with (ΔT,Δλ, SNR) =(0.01 K, 3 fm, 30 dB).
controllers. The LOD calculated in our paper is pointed out with
dash lines. A low LOD of 1.47 ×108RIU can be obtained
with (ΔT,Δλ)=(0.005 K, 1 fm) for Fano Type II.
The dependence of the overall sensitivity and LOD on the loss
coefﬁcients for the two types of Fano lineshapes is illustrated in
Fig. 9. As the loss increases, both the sensitivity and the LOD
deteriorate, with the sensitivity of Type II decreasing at a faster
rate. The LOD of Type II only increases quicky at high loss
levels, i.e., loss =0.5 and 1 dB/cm, because the LODintensity
becomes dominant due to decreased sensitivity when loss is
high. As a result, the advantage for Fano Type II over Type I is
less obvious as the optical loss increases. This demonstrates the
importance of loss analysis and control for a biosensor based on
the reversed Fano spectral shape.
5200110 IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 20, NO. 3, MAY/JUNE 2014
VI. DISCUSSION
The integration of the entire sensing system on-chip will en-
able the development of a portable and in situ testing system.
The simple detection scheme offered by monitoring the intensity
change allows the possibility for integrating both the light source
and the photodetector, together with the resonator. Various semi-
conductor lasers have been proposed for on-chip operation, and
the AlGaAs/GaAs lasers offer emission around 800 nm [56],
which is a suitable light source for our sensor. We note that com-
peting hybrid bonding methods to integrate the AlGaAs/GaAs
laser on a silicon wafer have also been demonstrated [57], [58].
Using the recently proposed phase-noise cancellation scheme
for semiconductor lasers, one can reduce the laser linewidth
down to the order of kHz [53], [54], enabling the realization
of the high-performance on-chip laser required in sensitive bio-
chemical detection applications. On the other hand, a Ge-on-Si
detector working in wavelengths ranging from 650 to 1605 nm
has been proposed [59], which can be seamlessly integrated with
our silicon nitride resonators.
For this highly sensitive biosensor based on embedded rings,
tuning structures are needed to precisely align the resonance
wavelengths of the two rings relative to the laser. This might
add to the device design complexity. Since the rings used in our
sensor are relatively large in radius and thus have a small free
spectral range, the wavelength alignment could be realized with
a small tuning voltage or thermal change.
VII. CONCLUSION
We have observed and analyzed two types of Fano spec-
tral features in embedded-ring resonators. A detailed theoretical
comparison between these two Fano types for applications in
biodetection is performed. Based on the reversed Fano line-
shape, we designed a highly sensitive biochemical sensor ex-
hibiting a high sensitivity of 1.16 ×105/RIU. This is one order
of enhancement than the performance of sensors operating in
the normal Fano spectral regime. Based on a systematic noise
analysis, our sensor shows a low limit of detection of 3.24 ×
108RIU when noise (ΔT,Δλ,SNR)=(0.01K,3fm,30dB)is
included, which could be further reduced with improved system
noise control methods.
REFERENCES
[1] X. Fan, I. M. White, S. I. Shopova, H. Zhu, J. D. Suter, and Y. Sun,
“Sensitive optical biosensors for unlabeled targets: A review,” Anal. Chim.
Acta, vol. 620, no. 1–2, pp. 8–26, Jul. 2008.
[2] F. Vollmer and S. Arnold, “Whispering-gallery-mode biosensing: Label-
free detection down to single molecules,” Nat. Methods, vol. 5, pp. 591–
596, Jul. 2008.
[3] H. K. Hunt and A. M. Armani, “Label-free biological and chemical sen-
sors,” Nanoscale, vol. 2, pp. 1544–1559, Jun. 2010.
[4] M. S. Luchansky and R. C. Bailey, “High-Q optical sensors for chemical
and biological analysis,” Anal. Chem., vol. 84, pp. 793–821, Jan. 2012.
[5] C. Ciminelli, C. M. Campanella, F. Dell’Olio, C. E. Campanella, and
M. N. Armenise, “Label-free optical resonant sensors for biochemical
applications,” Prog. Quan. Electron., vol. 37, no. 2, pp. 51–107, Mar.
2013.
[6] W. Liang, Y. Huang, Y. Xu, R. K. Lee, and A. Yariv, “Highly sensitive
ﬁber Bragg grating refractive index sensors,” Appl. Phys. Lett., vol. 86,
no. 15, p. 151122, Apr. 2005.
[7] T. Wei, Y. Han, H. L. Tsai, and H. Xiao, “Miniaturized ﬁber inline Fabry–
Perot interferometer fabricated with a femtosecond laser,” Opt. Lett.,
vol. 33, no. 6, pp. 536–538, Mar. 2008.
[8] F. Vollmer, S. Arnold, and D. Keng, “Single virus detection from the
reactive shift of a whispering-gallery mode,” in Proc. Nat. Acad. Sci.,
Dec. 2008, vol. 105, pp. 20701–20704.
[9] N. M. Hanumegowda, C. J. Stica, B. C. Patel, I. White, and X. Fan, “Re-
fractometric sensors based on microsphere resonators,” Appl. Phys. Lett.,
vol. 87, no. 20, p. 201107, Nov. 2005.
[10] R. W. Boyd and J. E. Heebner, “Sensitive disk resonator photonic biosen-
sor,” Appl. Opt., vol. 40, no. 31, pp. 5742–5747, Nov. 2001.
[11] J. Zhu, S. K. Ozdemir, Y. F. Xiao, L. Li, L. He, D. R. Chen, and L. Yang,
“On-chip single nanoparticle detection and sizing by mode splitting in an
ultrahigh-Q microresonator,” Nat. Photon., vol. 4, no. 1, pp. 46–49, Jan.
2010.
[12] A. Yalcin, K. C. Popat, O. C. Aldridge, T. A. Desai, J. Hryniewicz,
N. Chbouki, B. E. Little, O. King, V. Van, S. Chu, D. Gill, M. Anthes-
Washburn, M. S. Unlu, and B. B. Goldberg, “Optical sensing of
biomolecules using microring resonators,” IEEE J. Sel. Top. Quantum
Electron., vol. 12, no. 1, pp. 148–155, Jan./Feb. 2006.
[13] C.-Y. Chao, W. Fung, and L. J. Guo, “Polymer microring resonators for
biochemical sensing applications,” IEEE J. Sel. Top. Quantum Electron.,
vol. 12, no. 1, pp. 134–142, Jan./Feb. 2006.
[14] C.-Y. Chao and L. J. Guo, “Biochemical sensors based on polymer mi-
crorings with sharp asymmetrical resonance,” Appl. Phys. Lett., vol. 83,
no. 8, pp. 1527–1529, Aug. 2003.
[15] C. A. Barrios, K. B. Gylfason, B. S´
anchez, A. Griol, H. Sohlstr¨
om, and
M. Holgado, “Slot-waveguide biochemical sensor,” Opt. Lett., vol. 32,
no. 21, pp. 3080–3082, Nov. 2007.
[16] H. Yi, D. S. Citrin, and Z. Zhou, “Highly sensitive silicon microring sensor
with sharp asymmetrical resonance,” Opt. Exp., vol. 18, no. 3, pp. 2967–
2972, Feb. 2010.
[17] H. Yi, D. S. Citrin, and Z. Zhou, “Highly sensitive athermal optical mi-
croring sensor based on intensity detection,” IEEE J. Quantum Electron.,
vol. 47, no. 3, pp. 354–358, Mar. 2011.
[18] V. M. N. Passaro, B. Troia, and F. De Leonardis, “A generalized approach
for design of photonic gas sensors based on Vernier-effect in mid-IR,”
Sensor. Actuat. B-Chem., vol. 168, pp. 402–420, Jun. 2012.
[19] Z. Xia, Y. Chen, and Z. Zhou, “Dual waveguide coupled microring res-
onator sensor based on intensity detection,” IEEE J. Quantum Electron.,
vol. 44, no. 1, pp. 100–107, Jan. 2008.
[20] A. Ramachandran, S. Wang, J. Clarke, S. Ja, D. Goad, L. Wald, E. Flood,
E. Knobbe, J. Hryniewicz, S. Chu, D. Gill, W. Chen, O. King, and B. Little,
“A universal biosensing platform based on optical micro-ring resonators,”
Biosens. Bioelectron., vol. 23, no. 7, pp. 939–944, Feb. 2008.
[21] E. Chow, A. Grot, L. W. Mirkarimi, M. Sigalas, and G. Girolami, “Ultra-
compact biochemical sensor built with two-dimensional photonic crystal
microcavity,” Opt. Lett., vol. 29, no. 10, pp. 1093–1095, May 2004.
[22] M. Loncar, A. Scherer, and Y. Qiu, “Photonic crystal laser sources for
chemical detection,” Appl. Phys. Lett., vol. 82, no. 26, pp. 4648–4650,
Jun. 2003.
[23] N. Jokerst, M. Royal, S. Palit, L. Luan, S. Dhar, and T. Tyler, “Chip scale
integrated microresonator sensing systems,” J. Biophoton., vol. 2, no. 4,
pp. 212–226, Apr. 2009.
[24] Y. Sun and X. Fan, “Optical ring resonators for biochemical and chemical
sensing,” Anal. Bioanal. Chem., vol. 399, no. 1, pp. 205–211, Jan. 2011.
[25] L. Zhang, M. Song, T. Wu, L. Zou, R. G. Beausoleil, and A. E. Willner,
“Embedded ring resonators for microphotonic applications,” Opt. Lett.,
vol. 33, no. 17, pp. 1978–1980, Sep. 2008.
[26] L. Zhang, M. Song, R. G. Beausoleil, and A. E. Willner, “Embedded ring
resonators,” IEEE LEOS Newsletters, vol. 22, no. 6, pp. 21–22, Dec. 2008.
[27] X. Zhou, L. Zhang, A. M. Armani, R. G. Beausoleil, A. E. Willner, and
W. Pang, “Power enhancement and phase regimes in embedded microring
resonators in analogy with electromagnetically induced transparency,”
Opt. Exp., vol. 21, no. 17, pp. 20179–20186, Aug. 2013.
[28] Y.-F. Xiao, L. He, J. Zhu, and L. Yang, “Electromagnetically induced
transparency-like effect in a single polydimethylsiloxane-coated silica mi-
crotoroid,” Appl. Phys. Lett., vol. 94, no. 23, p. 231115, Jun. 2009.
[29] U. Fano, “Effects of conﬁguration interaction on intensities and phase
shifts,” Phys. Rev., vol. 124, no. 6, pp. 1866–1878, Dec. 1961.
[30] A. E. Miroshnichenko, S. Flach, and Y. S. Kivshar, “Fano resonances in
nanoscale structures,” Rev. Mod. Phys., vol. 82, no. 3, pp. 2257–2298,
Aug. 2010.
[31] K. Kobayashi, H. Aikawa, S. Katsumoto, and Y. Iye, “Tuning of the Fano
effect through a quantum dot in an Aharonov-Bohm interferometer,” Phys.
Rev. Lett., vol. 88, no. 25, p. 256806, Jun. 2002.
ZHOU et al.: ON-CHIP BIOLOGICAL AND CHEMICAL SENSING WITH REVERSED FANO LINESHAPE ENABLED 5200110
[32] H.Lu,R.L
¨
u, and B. F. Zhu, “Tunable fano effect in parallel-coupled
double quantum dot system,” Phys. Rev. B, vol. 71, no. 23, p. 235320, Jun.
2005.
[33] S. Fan, “Sharp asymmetric line shapes in side-coupled waveguide-cavity
systems,” Appl. Phys. Lett., vol. 80, no. 6, pp. 908–910, Feb. 2002.
[34] W. Liang, L. Yang, J. K. Poon, Y. Huang, K. J. Vahala, and A. Yariv,
“Transmission characteristics of a Fabry–Perot etalon-microtoroid res-
onator coupled system,” Opt. Lett., vol. 31, no. 4, pp. 510–512, Feb. 2006.
[35] Y. Lu, J. Yao, X. Li, and P. Wang, “Tunable asymmetrical Fano reso-
nance and bistability in a microcavity-resonator-coupled Mach–Zehnder
interferometer,” Opt. Lett., vol. 30, no. 22, pp. 3069–3071, Nov. 2005.
[36] L. Zhou and A. W. Poon, “Fano resonance-based electrically reconﬁg-
urable add-drop ﬁlters in silicon microring resonator-coupled Mach–
Zehnder interferometers,” Opt. Lett., vol. 32, no. 7, pp. 781–783, Apr.
2007.
[37] M. Tomita, K. Totsuka, R. Hanamura, and T. Matsumoto, “Tunable
Fano interference effect in coupled-microsphere resonator-induced trans-
parency,” J. Opt. Soc. Amer. B, vol. 26, no. 4, pp. 813–818, Apr. 2009.
[38] B.-B. Li, Y.-F. Xiao, C.-L. Zou, Y.-C. Liu, X.-F. Jiang, Y.-L. Chen, Y. Li,
and Q. Gong, “Experimental observation of Fano resonance in a sin-
gle whispering-gallery microresonator,” Appl. Phys. Lett., vol. 98, no. 2,
p. 021116, Jan. 2011.
[39] C. A. Barrios, K. B. Gylfason, B. S´
anchez, A. Griol, H. Sohlstr¨
om,
M. Holgado, and R. Casquel, “Slot-waveguide biochemical sensor: Er-
ratum,” Opt. Lett., vol. 33, no. 21, pp. 2554–2555, Nov. 2008.
[40] A. Gondarenko, J. S. Levy, and M. Lipson, “High conﬁnement micron-
scale silicon nitride high Q ring resonator,” Opt. Exp., vol. 17, no. 14,
pp. 11366–11370, Jul. 2009.
[41] K. F. Palmer and D. Williams, “Optical properties of water in the near
infrared,” J. Opt. Soc. Amer., vol. 64, no. 8, pp. 1107–1110, Aug. 1974.
[42] A. Yariv, “Universal relations for coupling of optical power between mi-
croresonators and dielectric waveguides,” Electron. Lett., vol. 36, no. 4,
pp. 321–322, Feb. 2000.
[43] F. Dell’Olio and V. M. Passaro, “Optical sensing by optimized silicon slot
waveguides,” Opt. Exp., vol. 15, no. 8, pp. 4977–4993, Apr. 2007.
[44] F. Qiu, A. M. Spring, F. Yu, and S. Yokoyama, “Complementary metal–
oxide–semiconductor compatible athermal silicon nitride/titanium dioxide
hybrid micro-ring resonators,” Appl. Phys. Lett., vol. 102, no. 5, p. 051106,
Feb. 2013.
[45] X. Tu, J. Song, T.-Y. Liow, M. K. Park, J. Q. Yiying, J. S. Kee, M. Yu, and
G.-Q. Lo, “Thermal independent Silicon-Nitride slot waveguide biosensor
with high sensitivity,” Opt. Exp., vol. 20, no. 3, pp. 2640–2648, Jan. 2012.
[46] J. F. Bauters, M. J. R. Heck, D. John, D. Dai, M.-C. Tien, J. S. Barton,
A. Leinse, R. G. Heideman, D. J. Blumenthal, and J. E. Bowers, “Ultra-
low-loss high-aspect-ratio Si3N4waveguides,” Opt. Exp., vol. 19, no. 4,
pp. 3163–3174, Feb. 2011.
[47] F. P. Payne and J. P. R. Lacey, “A theoretical analysis of scattering loss
from planar optical waveguides,” Opt. Quantum Electron., vol. 26, no. 10,
pp. 977–986, Oct. 1994.
[48] K. K. Lee, D. R. Lim, L. C. Kimerling, J. Shin, and F. Cerrina, “Fabrication
of ultralow-loss Si/SiO2waveguides by roughness reduction,” Opt. Lett.,
vol. 26, no. 23, pp. 1888–1890, Dec. 2001.
[49] K. K. Lee, D. R. Lim, H.-C. Luan, A. Agarwal, J. Foresi, and
L. C. Kimerling, “Effect of size and roughness on light transmission in a
Si/SiO2 waveguide: Experiments and model,” Appl. Phys. Lett., vol. 77,
no. 11, pp. 1617–1619, Sep. 2000.
[50] W. Stutius and W. Streifer, “Silicon nitride ﬁlms on silicon for optical
waveguides,” Appl. Opt., vol. 16, no. 12, pp. 3218–3222, Dec. 1977.
[51] I. M. White and X. Fan, “On the performance quantiﬁcation of resonant
refractive index sensors,” Opt. Exp., vol. 16, no. 2, pp. 1020–1028, Jan.
2008.
[52] M. I. Cheema, S. Mehrabani, A. A. Hayat, Y.-A. Peter, A. M. Armani,
and A. G. Kirk, “Simultaneous measurement of quality factor and wave-
length shift by phase shift microcavity ring down spectroscopy,” Opt.
Exp., vol. 20, no. 8, pp. 9090–9098, Apr. 2012.
[53] M. I. Cheema, U. A. Khan, A. M. Armani, and A. G. Kirk, “Towards more
accurate microcavity sensors: maximum likelihood estimation applied to
a combination of quality factor and wavelength shifts,” Opt. Exp., vol. 21,
no. 19, pp. 22817–22828, Sep. 2013.
[54] M. Bagheri, F. Aﬂatouni, A. Imani, A. Goel, and H. Hashemi, “Semi-
conductor laser phase-noise cancellation using an electrical feed-forward
scheme,” Opt. Lett., vol. 34, no. 19, pp. 2979–2981, Oct. 2009.
[55] F. Aﬂatouni and H. Hashemi, “Wideband tunable laser phase noise reduc-
tion using single sideband modulation in an electro-optical feed-forward
scheme,” Opt. Lett., vol. 37, no. 2, pp. 196–198, Jan. 2012.
[56] Y. Suematsu and K. Iga, “Semiconductor lasers in photonics,” J. Lightw.
Technol., vol. 26, no. 9, pp. 1132–1144, May 2008.
[57] A. W. Fang, H. Park, O. Cohen, R. Jones, M. J. Paniccia, and J. E. Bowers,
“Electrically pumped hybrid AlGaInAs-silicon evanescent laser,” Opt.
Exp., vol. 14, no. 20, pp. 9203–9210, Oct. 2006.
[58] M. E. Groenert, C. W. Leitz, A. J. Pitera, V. Yang, H. Lee, R. Ram,
and E. A. Fitzerald, “Monolithic integration of room-temperature cw
GaAs/AlGaAs lasers on Si substrates via relaxed graded GeSi buffer lay-
ers,” J. Appl. Phys., vol. 93, no. 1, pp. 362–367, Jan. 2003.
[59] J. Liu, J. Michel, W. Giziewicz, D. Pan, K. Wada, D. D. Cannon, and
J. Yasaitis, “High-performance, tensile-strained Ge p-i-n photodetectors
on a Si platform,” Appl. Phys. Lett., vol. 87, no. 10, p. 103501, Sep. 2005.
Xiaoyan Zhou received the B.S. degree from Tianjin
University, Tianjin, China, in 2011. She is currently
working toward the Ph.D. degree at Tianjin Univer-
sity. Her research interests include coupled micror-
ing resonators, optical delay, optical nonlinearity, and
biosensor. She is a Student Member of the Optical So-
ciety of America.
Lin Zhang (M’11) received the B.S. and M.S. de-
grees with honors from Tsinghua University, Bei-
jing, China, in 2001 and 2004, respectively. He re-
ceived the Ph.D. degree from the Universityof South-
ern California, Los Angeles, CA, USA, in 2011. He
is currently working as a Postdoctoral Researcher
with the Massachusetts Institute of Technology, Cam-
bridge, MA, USA. His research interests include on-
chip nonlinear optics, chip-scale optical interconnec-
tion, micro-resonator devices and system applica-
tions, slow light, and photonic crystal ﬁbers. He has
published over 125 peer-reviewed journal articles and conference papers, includ-
ing ten invited papers, and two book chapters. He has two patents issued. He is
a Member of the IEEE Photonics Society and the Optical Society of America.
Dr. Zhang was cited as one of the 2003 Top-Ten Outstanding Graduate Students
at Tsinghua University. He received the Best Research Paper Award from the
Department of Electrical Engineering at USC. He is one of the recipients of
the 2008 IEEE LEOS Graduate Student Fellowship and Chinese Government
Award for Outstanding Self-Financed Students Abroad.
Andrea M. Armani (M’06–SM’11) received the
B.A. degree in physics from the University of
Chicago, Chicago, IL, USA, and the Ph.D. degree
in applied physics from the California Institute of
Technology, Pasadena, CA, USA. She is currently the
Fluor Early Career Chair of Engineering and an As-
sistant Professor of Chemical Engineering and Mate-
rials Science at the University of Southern California,
Los Angeles, CA, USA. Her research group focuses
on the synthesis of novel optical materials for the de-
velopment of new integrated photonic devices with
applications in telecommunications and biodetection. She has received several
awards, including the ONR Young Investigator Award, the NIH New Innovator
Award, and the PECASE. She is a Senior Member of SPIE, and an Associate
Editor of Optics Letters.
5200110 IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 20, NO. 3, MAY/JUNE 2014
Daihua Zhang, photograph and biography not available at the time of
publication.
Xuexin Duan, photograph and biography not available at the time of
publication.
Jing Liu, photograph and biography not available at the time of
publication.
Hao Zhang, photograph and biography not available at the time of
publication.
Wei Pang received the B.S. degree from Tsinghua University, Beijing, China,
in 2001, and the Ph.D. degree in electrical engineering from the University
of Southern California, Los Angeles, CA, in 2006. He worked in the wireless
semiconductor division of Avago Technologies. He is currently a Professor with
the College of Precision Instrument and Optoelectronics Engineering, Tianjin
University, Tianjin, China. He has published 60 refereed papers in the areas
of MEMS for wireless communications, biological detection, wireless sensor
platforms, and medical ultrasound. He is an author of six issued and six pending
U.S. patents.
... As a weak interaction, Fano resonances are exceptionally sensitive to structural parameters and the surrounding environment, and they can be used to design ultra-sensitive sensors [19,20]. Specially, for the sensor on lab-on-a-chip based on the MIM-waveguide-coupled resonator structures will help in developing the chip sensor system in the fields of material [21], biological [22], and chemical [22] sciences. Thus, the plasmonic refractive index sensors based on various MIM-waveguidecoupled resonator systems with Fano resonance are proposed and investigated. ...
... As a weak interaction, Fano resonances are exceptionally sensitive to structural parameters and the surrounding environment, and they can be used to design ultra-sensitive sensors [19,20]. Specially, for the sensor on lab-on-a-chip based on the MIM-waveguide-coupled resonator structures will help in developing the chip sensor system in the fields of material [21], biological [22], and chemical [22] sciences. Thus, the plasmonic refractive index sensors based on various MIM-waveguidecoupled resonator systems with Fano resonance are proposed and investigated. ...
Article
Full-text available
A metal–insulator–metal (MIM) waveguide with tooth cavity-coupled ring splitting cavity is proposed. Transmission characteristics and refractive index sensitivity are investigated by using finite-element method. A Fano-like line is observed in the transmission spectrum, and it is caused by the coherent superposition of the narrow discrete and wide continuous states. A maximum sensitivity of 1200 nm/RIU is achieved based on the Fano resonance effect when the air in the MIM waveguide and cavity is replaced by an insulator with a different refractive index. In addition, the derived structures of the plasmonic system are studied, and multiple Fano-like resonances are observed in the transmission spectrum. The effects of the structural parameters of plasmonic system on the Fano resonance are also investigated.
... Induced transparency with pulse delay and induced attenuation with pulse advancement in optical microresonator systems have the potential to be exploited for various applications, as in signal processing and several types of optical sensing [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16]. In addition, the importance of polarization effects in whispering-gallery microresonators is increasingly becoming recognized and studied [17][18][19][20][21][22]. ...
Article
Full-text available
Induced transparency and attenuation effects are observed in the throughput of a single whispering-gallery microresonator due to mode coupling between two coresonant orthogonally polarized whispering-gallery modes of very different quality factors. Intracavity cross-polarization coupling, occurring when either the transverse electric (TE) mode or the transverse magnetic (TM) mode is driven, results in coupled-mode induced transparency or coupled-mode induced attenuation. Coresonance between the TE and TM modes is obtained by strain tuning, and the cross-polarization coupling is produced by polarization rotation due to optical spin-orbit interaction in a slightly asymmetric resonator. The observed behavior enables slow light and fast light, i.e., the delay or advancement of an incident resonant pulse. Experimental results representative of several different types of behavior are presented here. Induced transparency is seen to be accompanied by pulse delay, whereas induced attenuation can involve pulse advancement or delay. The results are analyzed and explained by analytical modeling and by comparison to the output of a more detailed numerical model describing these effects. Delays of up to 170 ns and advancements of up to 14 ns are found. The observed range of cross-polarization coupling strengths (probability of polarization change per round trip), namely, 10−10–10−7, is in agreement with theory.
... Various waveguide-coupled resonator devices based on the MDM waveguide have been proposed and demonstrated, such as filters [3], switches [4], and sensors [5]. Among these devices, plasmonic sensors for refractive index sensing have attracted intense interest due to their promising applications in the biological and chemical detection [6]. The principle of resonator-type refractive index sensors is based on the strong dependence of the resonance frequency on the environment. ...
Article
Full-text available
In this paper, we propose a compact plasmonic sensor structure comprised of a metal-dielectric-metal (MDM) waveguide, and a baffle plate in waveguide core and two side-coupled rectangular cavities. In this structure, two Fano resonances are achieved and can be tuned independently by changing the structural parameters of the cavities. Especially, when the resonant wavelengths of the two Fano resonances are the same, the sensing sensitivity can be enhanced by coupling between two Fano resonances. By investigating the transmission spectrum, the effect of structural parameters on Fano resonances and the refractive index sensitivity of the sensor structure are analyzed in detail. The numerical simulations demonstrate a sensitivity as high as 1295nm/RIU and a figure of merit of 1647.
Article
To improve the integration of Fano devices, we design a T-shaped waveguide coupling micro-ring resonator (MRR) structure to achieve a single cavity with Fano resonance in the whole spectral bands. The mathematical relationship between the phase factor, the coupling coefficient of the bus waveguide, and the Fano resonance slope extinction ratio (ER) is established. The electron beam exposure process is used to obtain a device with an insertion loss of ∼3 dB. The maximum ER of the Fano lineshape exceeds 15 dB, and the slope ratio (SR) is 251.3 dB/nm. This design improves the compactness of the Fano resonant device.
Chapter
We demonstrate the simulation for all-pass microring resonators with integrated Fano resonance to obtain high Free Spectral Range (FSR). The simulation mainly performed to generate high FSR by varying 2 parameters that affect the system which are radius of the ring and distance between end-facet. Furthermore, the analysis is proceeded by a variation of coupling coefficient. The impact to the output spectrum due to variations of the parameters was observed and analyzed, focusing on the FSR. We also investigate the effect of coupling coefficient variation and its contribution to the optical performance of the microring. The highest FSR obtained from this simulation was 273.752 nm.
Chapter
We have proposed and experimentally demonstrated a broadband thermo-optic switch in silicon-on-insulator. The switch is based on a ring-bus-ring-bus system, which can generate reversed Fano lineshapes in the transmission spectra. Then, the fabricated device exhibits a 3-dB-bandwidth more than 280 GHz, a maximum extinction ratio of 28 dB, and a slope rate of 60.8 dB/nm. In addition, the simulation results reveal that the switch is capable of switching data rates as high as 320 Gb/s with negligible signal distortion.
Article
Silicon photonics has shown great potential in integrated biochemical sensing. However, it is challenging to detect ions selectively, which is crucial in several biochemical application. In this paper, we propose and demonstrate a newly designed ion sensor by combining two-dimensional (2D) plasmonic and sub-nanoporous K <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> x </sub> MoO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> with a silicon waveguide. A T-shaped waveguide coupled with a microring resonator (MRR) generates Fano resonance at all resonance modes. Sharp Fano-like spectra were optimized by numerically simulating the structure of the waveguide in terms of the coupling coefficients, phase factor, and MRR power loss. In our design, a Fano resonance wavelength shift is caused by a refractive index change based on the interaction between 2D near-infrared plasmonic K <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> x </sub> MoO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> and K <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> . Moreover, the alteration in plasmonic absorption leads to a variation in transmission power. This dual-sensing output method is unique compared with existing methods that utilize optical ion sensors. Our results demonstrate that the micro-scale silicon waveguide combined with a 2D plasmonic material exhibits excellent potential as a chemical sensor.
Article
The coupling mechanism of a system in which two identical microresonators with slot are coupled through a tapered fiber is theoretically investigated and its transmission rate is numerically calculated. Physical parameters are evaluated for their impacts on the transmission spectra. Fano-like lineshape is observed under certain condition ${(\gamma_{c}\gt 2k)}$ and optimization of dominant parameters ${(\kappa_{1}/\kappa_{0},kL,\gamma_{c}/\kappa_{0} \ \text{and}\ Q_{0})}$ is applied to make the lineshape more steepened and sharper. This structure is promising for application in optical switch, sensing and research on Fano resonance.
Article
Nonreciprocity of the counterpropagating waves in a ring resonator induced by the rotation rate results in a measurable frequency shift. A self-reference measurement using mode broadening induced by backscattering is proposed to detect the rotation rate in a whispering gallery mode resonator with a cavity-made slot filled with atomic vapor. Through detuning an optical pump rate and a strong driving field coupled to a three-level atomic vapor, the backscattering generated by the cavity-made slot becomes sensitive to the rotation. Degenerate clockwise and counterclockwise modes couple to each other and create two new eigenmodes via manipulating the backscattering of the cavity-made slot. Detecting the mode broadening induced by the rotation rate enables the gyroscope's sensitivity to be enhanced at least four orders of magnitude in a low Q factor microresonator.
Article
Full-text available
We demonstrate a high-performance, tensile-strained Ge p-i-n photodetector on Si platform with an extended detection spectrum of 650–1605 nm and a 3 dB bandwidth of 8.5 GHz measured at λ = 1040 nm. The full bandwidth of the photodetector is achieved at a low reverse bias of 1 V, compatible with the low driving voltage requirements of Si ultralarge-scale integrated circuits. Due to the direct bandgap shrinkage induced by a 0.20% tensile strain in the Ge layer, the device covers the entire C band and a large part of the L band in telecommunications. The responsivities of the device at 850, 980, 1310, 1550, and 1605 nm are 0.55, 0.68, 0.87, 0.56, and 0.11 A/W, respectively, without antireflection coating. The internal quantum efficiency in the wavelength range of 650–1340 nm is over 90%. The entire device was fabricated using materials and processing that can be implemented in a standard Si complementary metal oxide semiconductor (CMOS) process flow. With high speed, a broad detection spectrum and compatibility with Si CMOS technology, this device is attractive for applications in both telecommunications and integrated optical interconnects.
Article
We have realized photonic crystal lasers that permit the introduction of analyte within the peak of the optical field of the lasing mode. We have explored the design compromises for developing such sensitive low-threshold spectroscopy sources, and demonstrate the operation of photonic crystal lasers in different ambient organic solutions. We show that nanocavity lasers can be used to perform spectroscopic tests on femtoliter volumes of analyte, and propose to use these lasers for high-resolution spectroscopy with single-molecule sensitivity.
Article
We observe Fano resonance in a single microresonator, in which two modes are excited simultaneously through a fiber taper. Our analysis reveals that the Fano resonance originates from an indirect-coupling of two originally orthogonal modes.
Article
Micro-ring resonators have been widely utilized in silicon photonics. However they often exhibit a high sensitivity to ambient temperature fluctuations. In this letter, we have demonstrated a complementary metal–oxide–semiconductor compatible athermal micro-ring resonator made from titanium dioxide (TiO2) and silicon nitride (SiNx). We have exploited the negative thermo–optic coefficient of TiO2 to counterbalance the positive coefficient of SiNx. By a precise control over the TiO2 layer thickness, an athermal condition remarkably consistent with the simulation can be achieved. Therefore, a SiNx–TiO2 hybrid micro-ring resonator with a temperature dependent wavelength shift of 0.073 pm/ °C has been realized.
Article
Coupled microresonators exhibit great potential for nonlinear applications. In the present work, we explore the nonlinear performance of an embedded ring resonator analogous to an electromagnetically induced transparency (EIT) medium, also known as coupled resonator induced transparency (CRIT). Interestingly, an EIT-like amplitude response can have a remarkably different power enhancement factor that varies by more than one order of magnitude, which is attributed to the different phase regimes of the embedded micro-ring resonators. In addition to the non-monotonic phase profile reported in atomic EIT systems, the phase responses featuring 2π and 4π monotonic transitions are identified and analyzed. We also present an interesting phenomenon, in which the power enhancement changes greatly, even with the same transfer function (both intensity and phase responses). This reveals that wisely choosing the operating regime is critical to optimize nonlinear performance of the embedded double resonator system, without adding to design or fabrication difficulty.
Article
Optical microcavities are widely used for biological and chemical sensing applications. In these devices, a sensing event is estimated by measuring the shift in the resonant wavelength, or in the quality factor of the microcavity. However, all published works to date only use one of these measures to estimate the sensing event. Here, we show that the estimation accuracy of a sensing event can be improved by employing a combination of both the quality factor and the resonant wavelength measurements in a microcavity sensor. We further demonstrate an experimental application of this model by introducing a refractive index change for a microtoroidal cavity sensor immersed in a liquid. By further using the finite element method simulations in conjunction with the estimator model, we show the existence of three distinct measurement regimes as a function of the quality factor of the microcavity. Finally, the estimator model is extended to develop a sensing metric to compare performance of optical or non-optical sensors.
Article
For a number of years, the scientific community has been paying growing attention to the monitoring and enhancement of public health and the quality of life through the detection of all dangerous agents for the human body, including gases, proteins, virus, and bacterial agents. When these agents are detected through label-free biochemical sensors, the molecules are not modified structurally or functionally by adding fluorescent or radioactive dyes. This work focuses on label-free optical ring resonator-based configurations suited for bio-chemical sensing, highlighting their physical aspects and specific applications. Resonant wavelength shift and the modal splitting occurring when the analyte interacts with microresonant structures are the two major physical aspects analyzed in this paper. Competitive optical platforms proposed in the literature are also illustrated together with their properties and performance.
Article
We have developed a highly sensitive refractometric sensor based on fused silica microsphere resonators. The spectral position of the whispering gallery mode (WGM) of a sphere shifts in response to the refractive index change in the surrounding medium. The strong light-matter interaction due to the extremely high Q factor associated with the WGM results in a sensitivity of approximately 30 nm∕RIU (refractive index units). This, together with the high spectral resolution of our sensor system (∼0.01 pm), yields a detection limit of refractive index change on the order of 10−7 RIU. Theoretical calculation is also performed and agrees well with the experimental data.
Article
We investigated interference effect between two whispering gallery modes in a system of ultrahigh-Q silica microspheres in which the resonance frequencies of spheres were precisely controlled through thermal tuning. A symmetric transmission peak of coupled—resonator—induced transparency reshaped into a sharp asymmetric spectrum similar to Fano effect in atomic system as the resonance frequency of the second sphere was detuned. The resonance modes showed frequency shifts as a function of the coupling strength between the two spheres, indicating that two whispering gallery modes were configurationally mixed. The observations were compared with calculations and discussed using double-spiral structures in the phase space in the transmitted field.
Article
We demonstrate 0.8-dB/cm transmission loss for a single-mode, strip Si/SiO2 waveguide with submicrometer cross-sectional dimensions. We compare the conventional waveguide-fabrication method with two smoothing technologies that we have developed, oxidation smoothing and anisotropic etching. We observe significant reduction of sidewall roughness with our smoothing technologies, which directly results in reduced scattering losses. The rapid increase in the scattering losses as the waveguide dimension is miniaturized, as seen in conventionally fabricated waveguides, is effectively suppressed in the waveguides made with our smoothing technologies. In the oxidation smoothing case, the loss is reduced from 32 dB/cm for the conventional fabrication method to 0.8 dB/cm for the single-mode waveguide width of 0.5 mum. This is to our knowledge the smallest reported loss for a high-index-difference system such as a Si/SiO2 strip waveguide.