Experimental study of the sensitivity of a porous silicon ring resonator sensor using continuous in-flow measurements

Abstract

A highly sensitive photonic sensor based on a porous silicon ring resonator was developed and experimentally characterized. The photonic sensing structure was fabricated by exploiting a porous silicon double layer, where the top layer of a low porosity was used to form photonic elements by e-beam lithography and the bottom layer of a high porosity was used to confine light in the vertical direction. The sensing performance of the ring resonator sensor based on porous silicon was compared for the different resonances within the analyzed wavelength range both for transverse-electric and transverse-magnetic polarizations. We determined that a sensitivity up to 439 nm/RIU for low refractive index changes can be achieved depending on the optical field distribution given by each resonance/polarization.

Full text

Experimental study of the sensitivity of a
porous silicon ring resonator sensor using
continuous in-flow measurements
RAFFAELE CAROSELLI,1 SALVADOR PONCE-ALCÁNTARA,1 FRANCISCO
PRATS QUILEZ,1 DAVID MARTÍN SÁNCHEZ,1 LUIS TORRIJOS MORÁN,1
AMADEU GRIOL BARRES,1 LAURENT BELLIERES,1 HANNA BANDARENKA,2
KSENIYA GIREL,2 VITALY BONDARENKO,2 AND JAIME GARCÍA-RUPÉREZ1,*
1Nanophotonics Technology Center, Universitat Politècnica de València, Camino de Vera s/n, 46022
Valencia, Spain
2Micro- and Nanoelectronics Department, Belarusian State University of Informatics and
Radioelectronics, P. Brovka str. 6, 220013 Minsk, Belarus
*jaigarru@ntc.upv.es
Abstract: A highly sensitive photonic sensor based on a porous silicon ring resonator was
developed and experimentally characterized. The photonic sensing structure was fabricated
by exploiting a porous silicon double layer, where the top layer of a low porosity was used to
form photonic elements by e-beam lithography and the bottom layer of a high porosity was
used to confine light in the vertical direction. The sensing performance of the ring resonator
sensor based on porous silicon was compared for the different resonances within the analyzed
wavelength range both for transverse-electric and transverse-magnetic polarizations. We
determined that a sensitivity up to 439 nm/RIU for low refractive index changes can be
achieved depending on the optical field distribution given by each resonance/polarization.
© 2017 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
OCIS codes: (130.3120) Integrated optics devices; (130.6010) Sensors; (230.5750) Resonators.
References and links
1. M. C. Estevez, M. Alvarez, and L. M. Lechuga, “Integrated optical devices for lab-on-a-chip biosensing
applications,” Laser Photonics Rev. 6(4), 463–487 (2012).
2. D.-X. Xu, A. Densmore, A. Delâge, P. Waldron, R. McKinnon, S. Janz, J. Lapointe, G. Lopinski, T. Mischki, E.
Post, P. Cheben, and J. H. Schmid, “Folded cavity SOI microring sensors for high sensitivity and real time
measurement of biomolecular binding,” Opt. Express 16(19), 15137–15148 (2008).
3. M. S. Luchansky and R. C. Bailey, “High-Q optical sensors for chemical and biological analysis,” Anal. Chem.
84(2), 793–821 (2012).
4. K. De Vos, I. Bartolozzi, E. Schacht, P. Bienstman, and R. Baets, “Silicon-on-Insulator microring resonator for
sensitive and label-free biosensing,” Opt. Express 15(12), 7610–7615 (2007).
5. A. L. Washburn, L. C. Gunn, and R. C. Bailey, “Label-free quantitation of a cancer biomarker in complex media
using silicon photonic microring resonators,” Anal. Chem. 81(22), 9499–9506 (2009).
6. Genalyte, Inc., http://www.genalyte.com/
7. M. Iqbal, M. A. Gleeson, B. Spaugh, F. Tybor, W. G. Gunn, M. Hochberg, T. Baehr-Jones, R. C. Bailey, and L.
C. Gunn, “Label-free biosensor arrays based on silicon ring resonators and high speed optical scanning
instrumentation,” IEEE J. Sel. Top. Quantum Electron. 16(3), 654–661 (2010).
8. T. Claes, J. G. Molera, K. De Vos, E. Schachtb, R. Baets, and P. Bienstman, “Label-free biosensing with a slot
waveguide-based ring resonator in silicon on insulator,” IEEE Photonics J. 1, 197–204 (2009).
9. S. T. Fard, V. Donzella, S. A. Schmidt, J. Flueckiger, S. M. Grist, P. Talebi Fard, Y. Wu, R. J. Bojko, E. Kwok,
N. A. F. Jaeger, D. M. Ratner, and L. Chrostowski, “Performance of ultra-thin SOI-based resonators for sensing
applications,” Opt. Express 22(12), 14166–14179 (2014).
10. J. Flueckiger, S. Schmidt, V. Donzella, A. Sherwali, D. M. Ratner, L. Chrostowski, and K. C. Cheung, “Sub-
wavelength grating for enhanced ring resonator biosensor,” Opt. Express 24(14), 15672–15686 (2016).
11. V. Yakovtseva, V. Bondarenko, M. Balucani, N. Kazuchits, and A. Ferrari, “Integrated optical waveguides based
on porous silicon: state-of-the-art and outlook for progress,” Physics, Chemistry and Application of
Nanostructures 99, 375–377 (1999).
12. S. Dhanekar and S. Jain, “Porous silicon biosensor: Current status,” Biosens. Bioelectron. 41(1), 54–64 (2013).
13. P. A. Snow, E. K. Squire, P. S. J. Russell, and L. T. Canham, “Vapor sensing using the optical properties of
porous silicon Bragg mirrors,” J. Appl. Phys. 86(4), 1781–1784 (1999).
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31651
#306445
https://doi.org/10.1364/OE.25.031651
Journal © 2017
Received 6 Sep 2017; revised 2 Nov 2017; accepted 4 Nov 2017; published 5 Dec 2017

14. C. Baratto, G. Faglia, E. Comini, G. Sberveglieri, A. Taroni, V. La Ferrara, L. Quercia, and G. Di Francia, “A
novel porous silicon sensor for detection of sub-ppm NO2 concentrations,” Sens. Actuators B Chem. 77(1–2),
62–66 (2001).
15. L. De Stefano, L. Rotiroti, I. Rea, L. Moretti, G. Di Francia, E. Massera, A. Lamberti, P. Arcari, C. Sanges, and
I. Rendina, “Porous silicon-based optical biochips,” J. Opt. A, Pure Appl. Opt. 8(7), S540–S544 (2006).
16. H. Zhang, Z. Jia, X. Lv, J. Zhou, L. Chen, R. Liu, and J. Ma, “Porous silicon optical microcavity biosensor on
silicon-on-insulator wafer for sensitive DNA detection,” Biosens. Bioelectron. 44, 89–94 (2013).
17. K. Kim and T. E. Murphy, “Porous silicon integrated Mach-Zehnder interferometer waveguide for biological
and chemical sensing,” Opt. Express 21(17), 19488–19497 (2013).
18. G. A. Rodriguez, S. Hu, and S. M. Weiss, “Porous silicon ring resonator for compact, high sensitivity biosensing
applications,” Opt. Express 23(6), 7111–7119 (2015).
19. F. A. Harraz, “Porous silicon chemical sensors and biosensors: A review,” Sens. Actuators B Chem. 202, 897–
912 (2014).
20. O. Bisi, S. Ossicini, and L. Pavesi, “Porous silicon: A quantum sponge structure for silicon based
optoelectronics,” Surf. Sci. Rep. 38(1), 1–126 (2000).
21. D. A. G. Bruggeman, “Dielectric constant and conductivity of mixtures of isotropic materials,” Ann. Phys. 24,
636–679 (1935).
22. R. B. Balili, “Transfer matrix method in nanophotonics,” Int. J. Mod. Phys. Conf. Ser. 17, 159–168 (2012).
23. L. Pavesi, “Porous silicon dielectric multilayers and microcavities,” Riv. Nuovo Cim. 20(10), 1–76 (1997).
24. W. S. Rasband, “ImageJ”, https://imagej.nih.gov/ij/.
1. Introduction
Optical ring resonators (RRs) are probably the planar photonic structures mostly employed
for the development of sensing devices due to their structural simplicity, compact size and
high sensitivity. Many research groups and commercial companies have demonstrated strong
activity in the development of RR arrays for the highly sensitive and multiplexed detection of
analytes for different applications [1–6]. Nowadays, a traditional Silicon-on-Insulator (SOI)
RRs can provide sensitivities up to around 160 nm/RIU (Refractive Index Units) [7].
However, the performance of traditional planar photonic sensing structures based on a
configuration with a high refractive index (RI) contrast is limited by the fact that only the
evanescent field propagating outside of the photonic structure is used for sensing purposes,
while the majority of the optical field distribution associated with the guided mode propagates
within the structure itself. Novel RR configurations have been proposed to overcome this
limitation, as those based on slot waveguides [8], on ultra-thin SOI substrates [9] or on sub-
wavelength waveguides [10]. Another option to increase the sensitivity, which is considered
in this work, is the use of porous silicon (PS) for the development of the sensing structures.
The high potential of porous silicon for sensing purposes, what has been demonstrated in
several works [11–18], comes from the possibility of infiltrating the target analytes directly
into the pores in order to obtain a significantly improved sensitivity [19]. Additionally, PS is
characterized by a very high surface-to-volume ratio [20], what allows immobilizing a
significantly larger amount of bioreceptors over the inner walls of the pores for the better
detection of biorecognition events. Moreover, PS can be formed simply, quickly and
inexpensively since it is the result of the electrochemical etching of a silicon substrate.
In this work, we study the sensing performance of a porous silicon ring resonator (PSRR)
sensor for transverse-electric (TE) and transverse-magnetic (TM) polarizations. In case of
each polarization, we measured the sensitivity for every resonance within the analyzed
wavelength range, which is determined by the optical field distribution of the PSRR resonant
mode and its interaction with the medium filling the pores. Experimental results obtained by
flowing several ethanol-water dilutions over the porous structure provided a sensitivity up to
439 nm/RIU for the low ethanol concentrations when working with TE-polarized light.
2. Fabrication of the PSRR
First, we selected the RI profile of the double-layered PS to be used, aiming at having a
strong light confinement in the top layer, where the photonic sensing structures will be
patterned. To reach this goal, the RI contrast between the top and bottom layers of PS needs
to be as high as possible when water is infiltrated into the pores. However, there is a
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31652

limitation in the maximum reduction of the RI of the bottom layer since it implies increasing
the porosity of this PS layer, what compromises the stability of the photonic structure. Taking
these considerations into account, we selected porosities of the top and bottom PS layers of
35% and 65%, respectively. By using the Bruggeman equation [21], we determined the RI of
the PS layers for these porosities, being ntop-water = 2.67 and nlc-water = 1.95 when considering
water-filled pores, and ntop-air = 2.56 and nlc-air = 1.7 when considering air-filled pores.
The designed double-layered PS was fabricated by an electrochemical etching of a single-
side polished boron doped (resistivity of 0.01 – 0.02 Ω⋅cm) <100> Si wafer, in a solution of
48% hydrofluoric acid and 99% ethanol mixed in a 1:2 volume ration. Just before the etching
process, the silicon wafer was cleaned from possible organic residues using a piranha solution
(volume ration H2SO4:H2O2 = 3:1) and then immersed into aqueous solution of HF (5%) to
remove SiO2 from the surface. After each cleaning step, the wafer was rinsed in deionized
water (DIW). The top layer with high RI was formed by applying a current density of 5
mA/cm2 for 100 seconds and the bottom layer of 45 mA/cm2 for 110 seconds. The cleaning
procedures and the electrochemical etching were performed at room temperature. The
fabricated PS sample was washed with acetone, isopropyl alcohol and DIW to remove
residual surface impurities. Finally, an O2 plasma was applied for 6 minutes, in order to
improve the hydrophilicity of the PS surface.
The fabricated PS sample was cleaved into two parts: one for the physical characterization
of the PS layers and the other one for the fabrication of the photonic structures. Field
emission scanning electron microscope (FE-SEM) was used to study the thickness and the
uniformity of both PS layers as well as the average pore diameter of the top PS layer. Figure
1(a) shows a FE-SEM image of the cross section of the double-layered PS, where the
different porosities of both layers are clearly distinguished. The measured thicknesses for the
top and bottom PS layers were around 800 nm and 4350 nm, respectively. Refractive indices
of the double-layer were calculated from Fresnel coefficients under the Transfer Matrix
Method approach using reflection measurements of the porous structure [22, 23]. In this
respect, the refractive indices obtained were ntop-air = 2.48 and nlc-air = 1.75 for the air-filled
pores, what corresponds to porosities of 38% and 63%, respectively. These porosities are
practically identical to those previously selected in the design phase, thus indicating a good
control of the PS fabrication procedure. Additionally, the ImageJ software [24] was used to
analyse a top-view FE-SEM image, shown in Fig. 1(b), in order to obtain the average pore
diameter of the sensing structure, which was determined to be around 12 nm.
Fig. 1. FE-SEM images of (a) cross section and (b) top-view of the PS structure.
Finally, a lithographic process was used to create the planar photonic structures in the top
PS layer. Designs were patterned into a PMMA positive resist layer using e-beam lithography
with an acceleration voltage of 10 keV and a dose of ~87 μC/cm2. Then, after developing the
resist, the pattern was transferred to the top PS layer by inductively coupled plasma (ICP)
etching. The fabricated design consisted of a 20µm-radius RR coupled to a coupling
waveguide. Several sets of this structure were fabricated using different widths of the
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31653

coupling and the RR waveguides (800, 1000 and 1200 nm) and different coupling gaps
between the coupling waveguide and the RR (150, 200, 250, 300 and 350 nm). In order to
ease the butt-coupling characterization of the photonic structures, the coupling waveguide
was tapered to a width of 10 µm at the edges of the chip. Figure 2 shows several FE-SEM
images of the fabricated structures.
Fig. 2. FE-SEM images of the fabricated PSRR chip. (a) 60°-sectional image of the 10µm-
wide access waveguide. Thanks to the vertical cut of this sample edge, the vertical orientation
of the pores can be observed. (b) Cross-sectional image of the 10µm-wide access waveguide.
Thanks to the non-vertical cut of this sample edge, the sponge-like morphology of the PS and
the boundary between the two PS layers can be clearly observed. (c) Top-view image of the
top and bottom PS layers in the access waveguide region, where the porosity difference
between them can be clearly observed. (d) Top-view image of the RR and the coupling
waveguide.
3. Experimental setup and static characterization of the PSRR
An opto-fluidic setup, which is shown in Fig. 3, was developed to carry out the optical
characterization of the PSRR and the RI detection experiments. The optical part of the setup
consisted in a horizontal coupling interrogation platform where the light from a continuous
sweep tunable laser (Keysight 81980A) was coupled to the access waveguides in the PSRR
chip using a lensed fiber. The polarization of the input light was adjusted using a polarization
controller (Thorlabs FPC562) before injecting it into the chip. The light coming out from the
PSRR chip was collected with an objective (20X Olympus Plan Achromat, 0.4 NA) and
passed through a polarizer (Newport RM25A) in order to measure the signal level for the
selected polarization using an infrared (IR) camera (Xenics Xeva-1.7-320). The interrogation
platform was controlled using a software programmed in LabVIEW able to synchronize the
continuous sweep of the laser with the image acquisition of the IR camera via a trigger signal
in order to obtain the spectrum of the photonic structure with the desired spectral resolution.
Regarding the fluidic part of the setup, a microfluidic delivery system made in
polydimethylsiloxane (PDMS) was placed on top of the chip in order to flow all the solutions
over the sensing structure. A Finetech flip-chip tool was used to properly align the fluidic
microchannel with the PSRR structures. When the experiments were carried out, the target
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31654

solutions were flowed using a syringe pump working in withdraw mode and set to a constant
flow rate of 20 μl/min. By working in withdraw mode, the liquid flowed through the PDMS
channel can be easily changed by simply changing the vial at the end of the tubing.
Fig. 3. Pictures of the fiber-to-camera horizontal interrogation setup. The PDMS microfluidic
flow cell on top of the photonic chip can be also observed.
After characterizing in water environment the TE and TM spectra for all the
configurations included in the photonic chip (i.e., different waveguide widths and coupling
gaps), we determined that the best spectral responses in terms of resonances quality and
optical losses were obtained for the configuration having a coupling waveguide width of 1000
nm and a coupling gap of 200 nm. Figure 4 shows the measured TE and TM spectra, in water
environment, for that PSRR configuration. The average quality factor of the measured
resonances is quite reduced (453 and 512 for TE and TM polarizations, respectively) due to
the higher losses that are produced in the PS-based waveguides compared to typical solid core
ones. The inset images in Fig. 4 represent the light spots acquired with the IR camera at the
10µm-wide output waveguide for each polarization. It is possible to appreciate that a better
confined mode is obtained for TE polarization, while a higher amount of light is going into
the lower cladding for TM polarization.
Fig. 4. Spectra of the PSRR for (a) TE and (b) TM polarization. The amplitude is represented
in terms of the analog-to-digital units (ADU) measured by the camera (in logarithmic scale).
The inset in each graph shows the measured optical profile at the output waveguide for each
polarization.
The selected waveguide configuration was simulated using the software FemSIM in order
to determine the modes existing on it for a water environment. Two fully propagating modes
were obtained for that waveguide configuration for each polarization, while a third mode also
begins to propagate for both polarizations. The field profiles of those modes are shown in Fig.
5. From the optical profiles measured in the experiments, which were depicted in Fig. 4 we
can see that only the fundamental mode is properly excited and propagated for TE
polarization, while higher order modes are excited and propagated for TM polarization,
leading to a higher delocalization of the optical field.
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31655

Fig. 5. Simulations of the optical modes propagating in the selected PS waveguide
configuration (height = 800 nm and width = 1000 nm) for (a) TE and (b) TM polarization. The
modes are ordered from left to right.
4. PSRR sensitivity characterization
Several RI sensing experiments were performed to determine the sensitivity of the photonic
structure. To this aim, different ethanol-water dilutions were flowed over the sensor while the
transmission spectrum was continuously acquired in order to determine the shift of the PSRR
resonances. The ethanol (EtOH) concentrations in DIW used in the experiments were 10%,
5% and 1%, which correspond to RI changes of 6.6·10−3, 3.3·10−3 and 6.6·10−4 RIU respect
DIW, respectively. For the spectra acquisition, the tunable laser was swept from 1520 to 1620
nm with a sweeping speed of 10 nm/s (the total time for each sweep is 10 seconds) and the
synchronization with the IR camera provided a spectral resolution of 20 pm. A Lorentzian
fitting of each resonance was then performed in Matlab to determine its position with a higher
accuracy. All the experiments were carried out for TE and TM polarizations.
Figures 6(a) and 6(b) show the time evolution for the most and least sensitive PSRR
resonances within the sweeping range for TE and TM polarizations, respectively. As it can be
observed in these figures, the resonances presented a similar behavior for both polarizations
and only a slightly higher sensitivity is obtained for TE polarization resonances. This is
related to the fact that the sensing occurs mainly inside the guiding structures, where the field
distribution is similar in both cases. However, since a better light confinement is obtained for
TE polarization, as it was shown in the spot images presented in Fig. 4, a higher sensitivity is
provided by this polarization. This is the opposite to what typically happens for evanescent
wave sensors, where TM modes are generally more sensitive because of the presence of a
higher amount of evanescent field over the top surface of the structure. The average
sensitivity of the PSRR sensor (i.e., for all the resonances within the measured range) is
around 350 nm/RIU for TE and 320 nm/RIU for TM when considering the whole range of
refractive index variations that has been measured. The maximum sensitivity value, around
380 nm/RIU, was provided by the resonance located at ~1607.5 nm for TE polarization,
whose quality factor is 513. Such values are more than twice higher than the sensitivity
achieved for a traditional SOI RR. The noise level measured for all the resonances is below 1
pm, reaching values even in the range of 0.1 pm for some resonances as it is depicted in the
inset of Fig. 6(a). Finally, note that the time required by the resonances to reach the plateau
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31656

for each flowed substance (around 1 minute) is determined by the diffusion that is produced
in the input tubing at the interface between the liquids.
Figure 7 represents the spectral shift measured for each resonance for both polarizations,
where an oscillatory-like behavior can be observed as we move towards higher order
resonances. This sensitivity variation indicates that a different interaction between the
resonance optical modes and the target substance is taking place, which might be determined
by the local variations of the pores properties and distribution of the fabricated PS layer
where the RR is created. Besides that oscillatory-like behavior, the average shift also
increases as a function of the wavelength, due to the larger evanescent field of the modes as
wavelength is increased.
Fig. 6. Time evolution of the resonance shift for (a) TE polarization and (b) TM polarization
when flowing three cycles of EtOH in DIW with concentrations 10%, 5% and 1% (meaning RI
changes of 6.6·10−3, 3.3·10−3 and 6.6·10−4 RIU respect DIW, respectively) with cycles of DIW
flow between them. The different colors indicate the different resonances within the measured
wavelength range being tracked.
Fig. 7. Wavelength shift of each resonance for each RI variation for (a) TE and (b) TM
polarizations.
In Fig. 8, the sensitivity curves of each resonance for TE and TM polarizations are shown.
As it can be observed, the sensitivity behavior of the PSRR sensing structure was not totally
linear. In fact, the sensor exhibited a higher sensitivity for lower RI variations, specifically,
for RI variation of 6.6·10−4 RIU corresponding to the EtOH 1% concentration. This is related
with the non-linear variation of the RI of the top porous layer that is produced when changing
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31657

the RI of the solution infiltrated into the pores, as it is shown in Fig. 9. We can see that a
higher variation of the effective refractive index of the top layer is obtained for the EtOH 1%
concentration than for the EtOH 5% and 10% concentrations. For this reason, the PS structure
showed a higher sensitivity for such EtOH concentration. Thanks to such behavior, the
highest sensitivity value provided by the PSRR sensor was 439 nm/RIU for the TE resonance
located at ~1607.5 nm when considering a RI variation of 6.6·10−4 RIU.
Fig. 8. Sensitivity curve of each resonance for (a) TE and (b) TM polarizations.
Fig. 9. Top PS layer RI variation as a function of the EtOH solution RI.
5. Conclusions
A highly sensitive PSRR was developed thanks to the formation of a PS double layer with a
high RI contrast and its sensing performance was characterized both for TE and TM
polarizations. The characterization was performed by monitoring in continuum the evolution
of the sensing structure spectrum. The experimental results indicate that the sensitivity of the
PSRR was slightly better for the TE polarization than for the TM polarization. This is because
the sensing occurs in the core of the structure, where the TE polarization plays a major
influence due to the better confinement of the light in the photonic structure. Working with
such polarization, a sensitivity of 439 nm/RIU was achieved for the detection of low RI
variations. The higher light-matter interaction that is produced due to the fact that the sensing
occurs directly inside the structure, together with the possibility of immobilizing the
bioreceptors on the inner surface of the pores, make PS a suitable platform for the
development of new biosensing devices exhibiting a high sensitivity. Finally, note that the
combination of PS substrates with some of the configurations that have been proposed to
increase the sensitivity of photonic structures as slot or sub-wavelength waveguides might
provide even higher sensitivities.
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31658

Funding
European Commission through the project H2020-644242 SAPHELY; Spanish government
through the projects TEC2013-49987-EXP BIOGATE and TEC2015-63838-C3-1-R-
OPTONANOSENS; Generalitat Valenciana through the Doctoral Scholarship
GRISOLIAP/2014/109.
Vol. 25, No. 25 | 11 Dec 2017 | OPTICS EXPRESS 31659