68 Smart Sensors for Industrial Applications

70 100
90
60
80 σ = 0.9990
50 70

. 99
t=0
40 60 σ = 0.9985
50
S3m

S3m
30 σ = 0.9980
40
σ = 0.9970
20 7 30
t = 0.9 σ = 0.9960
20
10 t = 0.95 t= σ2
t =0.9 10
0 0
0.98 0.985 0.99 0.995 1 0.98 0.985 0.99 0.995 1
(a) Transmission coefficient (σ) (b) Self-coupling coefficient (t)

FIGURE 5.3 (a) Maximum sensitivity to the transmission coefficient and (b) sensitivity to the self-coupling
coefficient. (From Xia, Z. et al., IEEE J. Quantum Electron., 44(1), 100, 2008. With permission.)

for intensity change sensing, where ∆I is the intensity change and ∆n c is the index change of
the analyte:

∆λ 0
s= (5.4)
∆nc

for wavelength shift sensing, where ∆λ0 is the resonance wavelength change.
We have rigorously analyzed the device sensitivity for dual waveguide–coupled microring
resonator sensors based on intensity detection. The analysis and simulation results indicate that
the device sensitivity increases monotonically as the transmission coefficient approaches unity as
shown in Figure 5.3a, while the optimum of self-coupling coefficient t equals σ2, at which the
maximal sensitivity is obtained (Figure 5.3b). Moreover, a wavelength offset is also required so as
to operate at the best sensitivity, and this offset is determined by t and σ, together with the specific
mode number.

5.3 MICRORING SENSOR PROPERTIES

Sensitivity, selectivity, dynamic range, and thermal stability are most important properties for
microring optical sensor. Several works on these properties are presented here. Novel sensing
mechanisms, such as Fano resonance, coupling-induced intensity sensing, are adopted to obtain
higher sensitivity. The microring arrays with different probes and measurement strategies are
used to increase the sensing selectivity. Multimicroring resonators, such as interferencing micror-
ings and cascaded microrings, are applied to enlarge the dynamic range. To overcome one of
the strongest influences from environment, temperature, the athermal microring sensor is also
investigated.

5.3.1 Sensitivity
The conventional silicon microring sensor is based on single microring, where only the single sym-
metric Lorentzian resonance lineshape was used to respond to the index change. The fundamental
sensing theory has several limitations for the higher sensitivity, such as the limited Q factor and
limited waveguide index change to the light–matter interactions.
To overcome these limitations, a highly sensitive Fano resonance single microring sensor was
proposed. The spectra of silicon microring resonators coupled with waveguides possessing an
Silicon Microring Sensors 69

End facet End facet

Input Output
Reflection Reflection

FIGURE 5.4 High-sensitivity Fano resonance single microring sensor. (From Yi, H. et al., Opt. Express,
18(3), 2967, 2010. With permission.)

1.0

0.8

0.6

0.4

0.2

0.0
1536.36 1536.38 1536.40 1536.42 1536.44
(a) (b) Wavelength (nm)

FIGURE 5.5 (a) SEM of single microring resonator and (b) asymmetric spectrum in experiment. (From Yi, H.
et al., Opt. Express, 18(3), 2967, 2010. With permission.)

end-facet reflections are discussed theoretically and experimentally [28]. The end-facet reflection
in a silicon waveguide forms a Fabry–Perot (FP) resonator that couples to the microring resonance,
thus changing the symmetric Lorentzian resonance lineshape to a strongly asymmetric shape shown
in Figure 5.4. The physics underlying this change in lineshape is closely related to the Fano line-
shape [29], which results from a discrete resonance coupled to a continuum—here the quasicon-
tinuum of densely spaced FP resonances of the waveguide.
The spectral slope is demonstrated to become steeper when the resonances of FP and micror-
ing have a π-phase difference as shown in Figure 5.5. This requirement is easy to meet because the
long cavity of FP resonance leads to dense collection of resonances with small FSR. For applica-
tions that have much shorter length, an asymmetric resonance can also be obtained but by judicious
choice of the phase difference. Therefore, our device can provide asymmetric resonance with easier
design. Because a steeper slope is obtained in asymmetric resonance shape, Si microring resonators
can provide enhanced sensitivity in chemical detection application. Thus, this feature can reduce
the stringent requirement of a high-quality factor in microring resonators as in our demonstration
experiment. With regard to device fabrication, this means that there can be greater tolerance to
imperfections in the microring and in the waveguide end facets. This is an effective method to pro-
duce inexpensive and easily fabricated chemical sensors.
Other novel microring sensors were also proposed to permit highly sensitive intensity detec-
tion without necessitating monitoring of the entire resonance spectrum [30]. The physical mech-
anism underlying their superior performance is the significant coupling coefficient change that
can change the critical coupling condition, resulting in a large change in the output intensity [31].
70 Smart Sensors for Industrial Applications

1.0
Sensing region
0.8

0.6
Critical κ1 SMR κ1 SMR
I0

coupling
0.4
κ1
0.2 κc κc
κc κc
0.0 RMR RMR
0.0 0.2 0.4 0.6 0.8 1.0
(a) κ (b) (c)

FIGURE 5.6 (a) Output intensity change to different coupling coefficient κ, (b) Type I sensor, and (c) Type II
sensor. (From Yi, H. et al., J. Soc. Am. B, 28(7), 1611, 2011. With permission.)

1.0 1.0
Type I
0.8 0.8 Type II

Type I Type II
0.6 0.6 Critical
coupling
I0
κ

0.4 0.4

0.2 0.2

0.0 0.0
1.9059 1.9060 1.9061 1.9059 1.9060 1.9061
(a) neff (b) neff

FIGURE 5.7 (a) Coupling coefficient change to different effective index and (b) the corresponding intensity
to different effective index. (From Yi, H. et al., J. Soc. Am. B, 28(7), 1611, 2011. With permission.)

At the resonance wavelength λ0, the output intensity I 0 is sensitive to κ, especially within the
small κ region as in Figure 5.6a. Based on this property, two types of microring-assisted MZI
coupler are investigated and applied in the resonator to function as an intensity sensor, shown
in Figure 5.6b and c.
The sensitivity of the Type II sensor without phase bias is predicted to be δn ∼ 4.9 × 10 −8 RIU
from Figure 5.7, which is one order of magnitude smaller than that of conventional microring
sensors. By comparing the intensity curves of the conventional microring sensor and the cou-
pling-induced sensors, a higher sensitivity is observed in the latter due to the steeper slope of
the curve for I 0 versus n eff. Thus, the coupling-induced intensity sensor based on the microring
is demonstrated to provide a different and superior approach to intensity sensing by providing
enhanced sensitivity.

5.3.2 Selectivity
Typically, there are multiple analytes at the same one measurement. One microring sensor with a
specific probe is only able to sense one analyte. Multimicroring sensors are essential to distinguish
different analytes from each others.