Applied Physics Express

LETTERS Related content

- Photonic crystal resonances for sensing
Improving the sensitivity of compound waveguide and imaging

grating biosensor via modulated wavevector - Cavity-mediated interactions between the

guided mode resonance and surface
phonon polaritons in a hybrid
photonic–phononic system
To cite this article: Hai Lu et al 2018 Appl. Phys. Express 11 082202
- Surface plasmon differential ellipsometry
of aqueous solutions for bio-chemical
sensing

View the article online for updates and enhancements.

This content was downloaded from IP address 140.113.155.173 on 25/09/2020 at 07:48

Applied Physics Express 11, 082202 (2018)
https://doi.org/10.7567/APEX.11.082202

Improving the sensitivity of compound waveguide grating biosensor

via modulated wavevector
Hai Lu1*, Meng Huang1, Xiubao Kang2, Wenxing Liu3, Chao Dong1, Jun Zhang1, Shiqiang Xia1, and Xianzhou Zhang1
1
Engineering Laboratory for Optoelectronic Technology and Advanced Manufacturing, Henan Normal University, Xinxiang 453007, China
2
School of Science, Henan Institute of Engineering, Zhengzhou 451191, China
3
Department of Physics, Nanchang University, Nanchang 330031, China
*E-mail: [email protected]
Received May 20, 2018; accepted June 21, 2018; published online July 5, 2018

We report here on the improvement of the sensitivity of a compound waveguide grating (CWG) biosensor via a modulated wavevector. We first
address that the primary challenges to enhance the sensitivity of the CWG biosensor is the constant effective relative permittivity of the grating. We
then present guidelines on how to improve the sensitivity of the CWG biosensor. As a result, the sensitivity of the CWG biosensor in our example
doubles its original value without changing any of the structural parameters. In this work, we aim at paving the way for the development of efficient
CWG biosensors by offering new strategies to optimize the sensitivity through wavevector modulation.
© 2018 The Japan Society of Applied Physics

he ability to detect small changes in the refractive

T index of a biological analyte is very important to

fundamental studies in biological and chemical
sciences, as well as applications such as label-free photonic
detection.1–3) On the basis of this method, to achieve a highly
efficient label-free biosensor, various optical technologies
have been proposed. Typical examples include surface
plasmon resonance (SPR) and guided mode resonance
(GMR) biosensors.4–10)
As a kind of surface wave, the surface plasmon mode is
strongly localized near the interface. Because of the strong
Fig. 1. Scheme of CWG biosensor. The grating region is covered by the
overlap between the resonant mode and the biosolution and
biosolution. nB, nG, and ns are the refractive indices of the biosolution,
the fact that SPRs are excited at working wavelengths where grating, and substrate, respectively. nG = 3.48 (Si) and ns = 1.48 (SiO2); d is
metals offer strong dispersion, SPR biosensors can achieve the grating layer thickness, d = 0.16 µm; the grating constant is Λ = 0.862
high sensitivity S values (the sensitivity is defined as the µm; fa, fb, and fc are the filling factors. θ is the angle between the incident
resonance spectral shift over a perturbation in the refractive light and the z-axis, and φ is the angle between the in-plane projection of the
incident light and the x-axis.
index of the biosolution). The sensitivity of SPR biosensors
is generally on the order of several hundred nm=RIU (Δλ=
Δn).11,12) Accompanying the strong dispersion, the inevitable lation in CWG biosensors. We propose a practical imple-
loss of metal components leads to small quality factors for the mentation for rotating a longer-period CWG biosensor into
SPRs. The most widely adopted metric for the intrinsic spec- an oblique configuration with an optimized sensitivity as high
trum resolving power of an optical biosensor is the figure of as 345 nm=RIU that is about two times higher in magnitude
merit [FOM = S (nm=RIU)=Γ (nm), where Γ is the full width than a normal case (140 nm=RIU). We anticipate that such an
at half maximum (FWHM) of the resonance]. In the SPR improvement strategy could be generalized to other GMR-
biosensor design, the FOM is generally less than 100 owing based concepts and could allow for the development of
to the small quality factor caused by metal losses.13) efficient CWG biosensors.
On the other hand, in contrast to conventional SPR bio- This Letter is divided into three parts. We first demonstrate
sensors, practical implementations of GMR-based biosensors that it is not helpful to improve S by adjusting the fill factor,
have been elusive for two main reasons. First, the decreased even though the FOM can increase significantly. Next, we
FWHM of the GMR comes commonly at the price of in- address the S of a CWG biosensor by analyzing the relation-
creased layer thickness. Second, as the resonantly enhanced ship between S and the wavevector, with the goal of finding a
guided mode generally decays rapidly in the substrate, the guideline that improves the S in such a situation. Finally, we
sensitivity of a GMR sensor is then limited by the small apply this guideline to the oblique configuration for which we
overlap between the resonantly enhanced guided mode also consider the longer-period CWG biosensor case.
and the biosolution. Concerning the FWHM of the GMR, For a typical CWG biosensor in Fig. 1, the spatially
we have recently reported that an asymmetric resonance modulated permittivity in the grating region can be expanded
response (Fano-like resonance) can be realized in a com- into a Fourier series as ε(x) = ∑εn exp(i2nπx=Λ), where
pound waveguide grating (CWG) structure, whose FWHM the zero-order effective relative permittivity "0 ¼ ð1  fb 
can be tailored in an independent manner.14) Soon afterwards, fc Þn2G þ ð fb þ fc Þn2B , and the grating Fourier harmonics
Hu et al. have shown that a similar strategy can be used for "n ¼ ðn2G  n2B Þ=ðnÞ  fsin½nð1  fc Þ  sinðnfb Þg. In ac-
optical sensing.15) cordance with our discussion in Ref. 14, the grating Fourier
In this Letter, we address the second issue that concerns harmonics εn controls the amplitude of the evanescent
the improvement of sensitivity through wavevector modu- diffraction field and is responsible for the interaction of
082202-1 © 2018 The Japan Society of Applied Physics
Appl. Phys. Express 11, 082202 (2018) H. Lu et al.

(a)

(b)

(c)

(d)

Fig. 2. Spectrum shifts when there are small variations in the refractive
index of the biosolution for CWG biosensors with different filling factor fb
values of (a) 0, (b) 0.1, and (c) 0.14. (d) The peak wavelength changes
against the refractive index of the biosolution. The slopes of the curves Fig. 3. Distribution of electric field at the wavelength of the resonance
show that refractive index sensitivities corresponding to (a)–(c) were all peak for filling factor fb values of (a) 0 and (b) 0.14. Different colors
140 nm=RIU. represent different electric field densities. Black lines indicate the CWG
biosensor structure.

evanescent diffraction fields.14) In principle, ∣εn∣ is highly

dependent on the filling factor fb, which means that the recording high FOM of the biosensors based on plasmonic
coupling can be controlled by adjusting the filling factor fb. nanohole arrays.11,12)
Here, the total width of the two grating grooves in each period However, the performance of a biosensor depends on not
is kept constant, that is, fb + fc = 0.3. Generally, a smaller ∣ε1∣ only the FOM of the structure, but also the sensitivity S. These
means a weaker coupling, which, in turn, yields a decreased asymmetric resonance responses belong to a large family of
spectral width. With an appropriate design for waveguide characteristically asymmetric Fano resonances, which occur
gratings, leaky guided modes can be excited solely through from interference between a discrete GMR in waveguide
the first evanescent diffracted order of the grating. In addition, grating structures and a continuum Fabry–Perot etalon reso-
the spectral lineshapes and sideband levels of the GMR are nance of equivalent homogeneous structures.18,19) Because the
mainly determined by the equivalent homogeneous layers, zero-order effective relative permittivity of the grating is kept
and the refractive index of the grating layer could be constant for different filling factor fb values, the equivalent
represented using an effective medium theory approximation, homogeneous slab waveguides corresponding to the CWG
where the value of the effective medium was mainly deter- structures remain almost unchanged. This leads to the
mined by the zero-order effective relative permittivity ε0.16) dispersion relations of the leaky guided modes supported by
It is not difficult to find that ε0 is fixed at a certain value the structures, which are nearly identical. Therefore, as can be
(i.e., ε0 = 7.6662) for different filling factor fb values, indi- seen in Fig. 2(d), the same S is nearly preserved for different
cating that S almost remains constant when the filling factor fb structures.
changes. In addition, since the biosolution always covers the surface
On the basis of the above analysis method, we illustrate the grating region of the CWG biosensor, S should also be
spectrum of the CWG biosensor with various filling factor fb affected by the sum of the electromagnetic fields in the
values for the normal incidence TE-polarized light (electric solution. In Fig. 3, we show the electric field distribution.
field parallel to the y-direction, θ = 0°, and φ = 0°), as shown One can clearly find that the electric field is strongly localized
in Fig. 2. A scattering matrix method based on rigorous in the groove whether in the CWG biosensor or in the normal
coupled wave analysis (RCWA) is used here and in the GMR biosensor. As mentioned earlier, the zeroth-order field
following calculations.17) In Fig. 2, one can clearly find that, remains unchanged in the CWG biosensors with different fb
when the refractive index of the biosolution changes evenly values and does not contribute to intensity enhancement,
from 1.33 to 1.35, the peaks of the resonances experience owing to the fact that the zeroth-order effective relative
redshifts in all three cases. In particular, in the case of permittivity of the grating is kept constant for different fb
Fig. 2(c), the FWHM is as small as 0.53 nm, which leads to a values. On the other hand, the high-order evanescent field,
FOM as large as 2622. All these features indicate that, for which is responsible for intensity enhancement in the
CWG biosensors, the linewidth of the guided mode is structure, is bound in the groove. Although the electric field
obviously narrower than that of the classical GMR structure intensity localized in the CWG biosensor is higher than that
(e.g., when fb = 0, the CWG biosensors become the classical localized in the normal GMR structure owing to the larger
GMR structure. In this case, the FWHM and corresponding filling factor fb, S depends on not only the electric field within
FOM are just 80 nm and 17.5, respectively). Benefitting the groove but also the total electric field in the biosolution.
from the lossless configuration, such a FOM is 24 times Therefore, the change in electric field in the groove negli-
larger than the theoretical upper limits (FOM = 108) for a gibly affects S. In other words, the S values of the CWG
standard gold SPP sensor and almost as large as 37 times the biosensors with different fb values may be similar.
082202-2 © 2018 The Japan Society of Applied Physics
Appl. Phys. Express 11, 082202 (2018) H. Lu et al.

As known, the S of a biosensor represents its ability to (a)

detect small changes in the refractive index of a biological
analyte, while the wavevector is closely related to the
refractive index of the equivalent homogeneous grating layer
in a CWG biosensor. For the oblique configuration, the
normal wavevector of the Fano-like resonance for the CWG
biosensor can be written as
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
#ffi
u "  2
u m
kz ¼ tk20 n2B  sin  cos  þ  ðsin  sin Þ2 ; ð1Þ (b)


where k0 = 2π=λ and m ¼ 0; 1; 2; . . . ; and then, consid-

ering the small changes in the wavevector with the spectrum
or the refractive index of the biosolution, we can obtain
(c)
m
sin  cos  þ
@kz  m
¼ k0 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2 ffi ;
@ m 
n2B  sin  cos  þ  ðsin  sin Þ2

ð2Þ
@kz nB
¼ k0 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
  ffi : ð3Þ Fig. 4. Variation in S calculated using Eq. (4) for the CWG biosensor
@nB m 2 versus (a) Λ, (b) θ, and (c) φ.
2 2
nB  sin  cos  þ  ðsin  sin Þ

(a)
From Eqs. (2) and (3), the derivative related to S can be
simplified as
@  nB
S¼ ¼ : ð4Þ
@nB m m (b)
sin  cos  þ

To assess this interpretation, we computed S as a function of
(d)
the parameters in Eq. (4), i.e., Λ, θ, and φ. Figure 4(a) shows
(c)
the evolution of the S of the CWG biosensor as a function of
the related grating constant (Λ, in unit of nm). In this figure, a
linear dependence between S and Λ is observed. This means
that a higher S can be obtained by designing CWG biosensors
with larger grating constants. A similar phenomenon has been
indicated in Ref. 15. However, the grating constant Λ can be
Fig. 5. Spectrum shifts when there are small variations in the refractive
varied only in a limited range when designing GBR sensors. index of the biosolution for CWG biosensors with different schemes of
The basic limitation is imposed by the fact that an appropriate (a) Λ = 980 nm, θ = 0°, (b) Λ = 862 nm, θ = 70°, and (c) Λ = 980 nm,
Λ is necessary for the coupling between the zeroth-order and θ = 70°. (d) The peak wavelength changes against the refractive index of the
high-order diffraction modes. However, if we choose an biosolution. The slopes of the curves show that the refractive index
sensitivities corresponding to (a)–(c) were 160, 300, and 345 nm=RIU,
oblique configuration (i.e., θ ≠ 0° or φ ≠ 0°), S can also be
respectively.
increased efficiently as shown in Figs. 4(b) and 4(c), even if
the grating constant remains unchanged. It should be noted
that different illumination configurations show different trends simultaneously increase Λ and θ. From this configuration, we
in S. For example, in the same case (m = −1), S increases with find that the optimized S is close to 345 nm=RIU in com-
a positive θ but decreases with φ always. parison with the S of the normal case shown in Fig. 2(c),
In accordance with the guidelines analyzed above, we now evaluated to be 140 nm=RIU. As a result, the combined
move forward with the optimization of the S of the former optimization of these parameters allows a twofold enhance-
example mentioned in Fig. 2(c) through wavevector modu- ment in the magnitude of S. It should be noted that S is
lation. We start with a direct transfer of the enlarged Λ to the dependent on the grating period, incident angle, and the zero-
normal incidence case as depicted in Fig. 5(a). Compared order effective index of the grating layer. If different kinds of
with S = 140 nm=RIU corresponding to Fig. 2(c), the S of this materials are used for device fabrication, the zero-order
longer-period case increases slightly (S =160 nm=RIU) as effective index of the grating layer would be slightly modified.
shown in Fig. 5(d). If we keep the grating constant unchanged However, the zero-order effective index of the grating layer
(Λ = 862 nm), by increasing the angle, S can be effectively can also be tuned by different filling factors, which means that
improved (S = 300 nm=RIU) as shown in Figs. 5(b) and 5(d). a similar sensitivity can be obtained by adjusting the structural
Here, the −1-order diffraction is chosen to excite the Fano-like parameters of the compound waveguide grating. Moreover, it
resonance in the CWG biosensors. Here again, to obtain a is noteworthy that the simulated sensitivities are lower than
higher improvement of S, we keep φ equal to zero and the theoretically predicted ones, in all of the three cases. The
082202-3 © 2018 The Japan Society of Applied Physics
Appl. Phys. Express 11, 082202 (2018) H. Lu et al.

discrepancy arises from the fact that our theoretical analysis Key Project of Science and Technology of Henan Province (152102210079), and
the Natural Science Foundation of Henan Province of China (19B140003).
does not take the real refractive index of the grating layer into
consideration. [In a calculation using Eq. (4), nB was set
to 1.33.] However, the dependences of the simulated S on
these parameters show that the guidelines based on Eq. (4)
1) Y. H. Zheng, C. C. Zhang, and Q. M. Wang, Sens. Actuators B 245, 622
still enable sensitivity improvement. In addition, many (2017).
kinds of optical biosensors, such as photonic crystals and 2) J. Martínez, A. Ródenas, M. Aguiló, T. Fernandez, J. Solis, and F. Díaz,
optical fibers, have been developed.20–23) However, as seen in Opt. Lett. 41, 2493 (2016).
3) M. S. Diware, H. M. Cho, W. Chegal, Y. J. Cho, D. S. Kim, S. W. O, K. S.
Fig. 5(d), such a sensitivity (S = 345 nm=RIU) of the CWG
Kim, and S. H. Paek, Biosens. Bioelectron. 87, 242 (2017).
biosensor is four times higher than the theoretical upper limits 4) J. Homola, Anal. Bioanal. Chem. 377, 528 (2003).
(S = 88 nm=RIU) of recent photonic crystal biosensors.20) 5) F. Bahrami, J. S. Aitchison, and M. Mojahedi, Opt. Lett. 39, 3946 (2014).
In particular, compared with a similar low concentration solu- 6) T. J. Wang and W. S. Lin, Appl. Phys. Lett. 89, 173903 (2006).
7) T. J. Wang and C. W. Hsieh, Appl. Phys. Lett. 91, 113903 (2007).
tion case in near-infrared wavelength domains (∼1.5 µm), the 8) S. Foland, K. H. Choi, and J. B. Lee, Opt. Lett. 35, 3871 (2010).
sensitivity (S = 345 nm=RIU) of the CWG biosensor is eight 9) P. G. Hermannsson, K. T. Strensen, C. Vannahme, C. L. Smith, J. J. Klein,
times higher than that (S = 46.5 nm=RIU) of a microfiber M.-M. Russew, G. Grützner, and A. Kristensen, Opt. Express 23, 16529
biosensor.23) (2015).
10) P. G. Hermannsson, C. Vannahme, C. L. C. Smith, and A. Kristensen, Appl.
In summary, from the results of RCWA numerical Phys. Lett. 105, 071103 (2014).
investigations, we determined two primary obstacles to the 11) A. A. Yanik, A. E. Cetin, M. Huang, A. Artar, S. H. Mousavi, A.
optical sensing applications of a CWG, i.e., the constant zero- Khanikaev, J. H. Connor, G. Shvets, and H. Altug, Proc. Natl. Acad. Sci.
order effective relative permittivity of the grating and the U.S.A. 108, 11784 (2011).
12) W. Li, X. Jiang, J. Xue, Z. Zhou, and J. Zhou, Biosens. Bioelectron. 68, 468
high-order evanescent field trapped in the groove. Then, we (2015).
showed that the S of CWG biosensors through wavevector 13) M. Piliarik and J. Homola, Opt. Express 17, 16505 (2009).
modulation can be strongly enhanced. We addressed this 14) W. X. Liu, Y. H. Li, H. T. Jiang, Z. Q. Lai, and H. Chen, Opt. Lett. 38, 163
issue in two steps: first, by considering the enlarged Λ of a (2013).
15) J. Hu, X. Liu, J. Zhao, and J. Zou, Chin. Opt. Lett. 15, 030502 (2017).
CWG; and second, by rotating the angle of the incident 16) Y. F. Ku, H. Y. Li, W. H. Hsieh, L. K. Chau, and G. E. Chang, Opt. Express
light or the grating. A numerical S up to 345 nm=RIU was 23, 14850 (2015).
obtained; this value is two times higher in magnitude than the 17) M. G. Moharam and T. K. Gaylord, J. Opt. Soc. Am. 71, 811 (1981).
18) S. H. Fan and J. D. Joannopoulos, Phys. Rev. B 65, 235112 (2002).
normal case. The results reported here may allow for practical
19) G. D’Aguanno, D. de Ceglia, N. Mattiucci, and M. J. Bloemer, Opt. Lett.
implementations of optical biosensors showing high FOM 36, 1984 (2011).
and S. They also provide some general rules for improving 20) X. B. Kang, L. J. Liu, H. Lu, H. D. Li, and Z. G. Wang, J. Opt. Soc. Am. A
the S of a Fano-resonance-based sensor and, as such, could 33, 997 (2016).
21) X. B. Kang, H. Lu, and Z. G. Wang, Opt. Express 26, 12769 (2018).
enable the future development of CWG devices with 22) J. L. Chen, J. L. Gan, Z. S. Zhang, T. Yang, H. Q. Deng, and Z. M. Yang,
unexpected electromagnetic properties. Appl. Phys. Express 7, 012501 (2014).
Acknowledgments The research was supported by the National Natural 23) H. Q. Yu, L. B. Xiong, Z. H. Chen, Q. G. Li, X. N. Yi, Y. Ding, F. Wang,
Science Foundation of China (NSFC) (11404102, 11304078, and 11604136), the H. Lv, and Y. M. Ding, Appl. Phys. Express 7, 022501 (2014).

082202-4 © 2018 The Japan Society of Applied Physics