Open Access
Add to BibSonomy
Abstract
In this paper, a high sensitivity compact multi-slot sub-wavelength Bragg grating refractive index (RI) sensor was investigated. The structural parameters were optimized for higher sensitivity to RI change of the surrounding medium from viewpoints of a wavelength shift, an extinction ratio and a transmission loss, and a record-high sensitivity was experimentally demonstrated with a compact size. In this sensor, the first side-lobe at the Bragg grating (BG) stop-band end was focused as a sensing peak wavelength for moderate transmission loss and efficient sensing. To realize the compactness, a period count of the BG was kept as small as 20. By increasing the RI of the surrounding medium, the sensing peak shifts toward a longer wavelength side; thus due to the high sharpness and easy tracing of the first side-lobe, the device worked as an efficient RI sensor. The structural optimization was carried out by using 3D finite-difference time-domain (FDTD) simulation approach, and also influences of the structural parameters to sensitivities were discussed. Based on these optimized parameters, the devices were fabricated using the lift-off technique. By exposing the sensor to various liquid samples with different RIs such as pure water, sugar-dissolved water with various concentrations, acetone and isopropyl alcohol (IPA), a record-high sensitivity of 730 nm/RIU was attained for a sensor fabricated on SOI platforms with a length of as small as 9.5 µm and a transmission loss of 3 dB.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Silicon phonics is widely used in integrated circuit elements for various applications, and the advancement of fabrication precision for shape modulation in nano-meter sizes shows promise for device miniaturization [1,2]. One of the promising application fields of such silicon photonic devices is for compact sensors [3,4]. For this purpose most popular devices reported so far are Mach-Zehnder interferometers, but their footprints seem rather large [5]. For more compact size, micro ring resonators (MRRs) are investigated [6]. However, compact MRRs with small radiuses are liable to high transmission losses. To avoid this issue, while keeping compactness, a BG type configuration should be considered. BGs have widely been used in lasers, signal processing and filtering [7–10]. Furthermore, their use in sensors with lab-on chip has been studied for measuring multiple reagents at the same time [11].
The main purpose of these photonic sensors is sensing of RI of the surrounding region. One common parameter used for the evaluation in such sensors is sensitivity, which utilizes the change of information as a result of sensing peak wavelength shift due to RI change. The first reported BG RI sensor used rib waveguides (WGs) with top etched BG, and the evaluated sensitivity was 33.6 nm/RIU with a BG length of 172 µm [12]. A BG transmission spectrum with a large coupling coefficient exhibits a broad stop-band, where their steep side-lobes can be used as peaks of interest to measure the change of RI. To simplify the sensing and have a sharp wavelength peak inside the stop-band the introduction of a phase shift (PS) in the grating was proposed [13]. To reduce costs, a single step patterning in fabrication is effective. For this purpose, side BGs on a strip WG with a PS was adopted, and slightly increased sensitivity to 59 nm/RIU was reported [14]. To further enhance sensitivities, a slot inside the WGs was introduced, where the reported sensitivity value was much improved to 340 nm/RIU [15]. The sensors in high-sensitive configurations worked, however, for transverse electric (TE) mode. Another promising WG configuration was a sub-wavelength grating (SWG), which worked in both TE and transverse magnetic (TM) modes with sensitivities of about 400 nm/RIU [16–18].
Another type of a compact RI sensing device, which has been widely studied for nano-photonic devices, is photonic crystal (PhC) cavities in 1D or 2D configurations due to their higher Q-factor characteristics, and exhibited sensitivities of around 200 nm/RIU [19–21]. Further addition of slots has shown to improve sensitivity similarly to about 400 nm/RIU [22,23]. Higher values can be reached by suspending the PhC cavity off from the substrate, this might, however, require rather complex fabrication methods [24–27].
By employing single fabrication patterning a WG with multiple slots and sub-wavelength BG (MS-SWBG) configuration has shown currently the highest sensitivities. Using this type of the configuration with a PS, the reported sensitivity was as large as 579 nm/RIU at a length of 74 µm [28]. Another recent work with a similar WG configuration using PhC cavity resonance was done by modifying the grating periods to form a Gaussian shape profile, whereby side-lobe peaks were pushed into the stop-band, exhibited a high sensitivity of 586 nm/RIU at a shorter length of 22.3 µm [29]. Having the sensing peak, however, in the stop-band may induce larger transmission losses, which are especially common for SWG type structures at smaller duty ratios [30].
In this study a high-sensitivity MS-SWBG RI sensor with the first side-lobe at the longer-wavelength-side end of stop-band as a sensing wavelength peak was investigated for structural optimization in terms of a wavelength shift, an extinction ratio and a transmission loss with a thicker WG of 340 nm. The analysis was carried out using rigorous 3D FDTD simulation software, and then the devices were fabricated to demonstrate experimentally a record-high sensitivity among current non-suspended SOI sensors.
The paper first introduces the operation mechanism of the sensors investigated, followed by the structural optimization. In the experiment, the fabrication process was explained with the consideration of the fabricated structures affecting the resultant performances. Then, the measurement results of the fabricated MS-SWBG sensors were shown to demonstrate a record-high sensitivity with a compact size. Finally a short discussion was addressed regarding the measured results related to the simulated ones, and the conclusion was summarized.
2. Structure and operation principle
A schematic structure of the studied device is shown in Fig. 1(a). It is composed of a silicon WG formed on a silicon on insulator (SOI) platform with MS-SWBG configuration satisfying the BG condition of:
where ${\lambda _B}$ is a Bragg wavelength, $\Lambda $ a pitch of a BG period, ${n_{eff - MS - SWBG}}$ an effective refractive index of the MS-SWBG WG, and m a diffraction order. In this study, $m$ = 1 was adopted since it gives the largest diffraction efficiency. Due to the high-index BG structure, transmission spectra with a large and deep stop-band are displayed, and by changing RIs of the surrounding medium the BG wavelength and the associated stop-band shift, as is clear in Eq. (1). By measuring these spectral shifts the device can work as an RI sensor. The introduction of slots makes the shift of the BG wavelength and BG stop-band more efficient due to enhanced interaction between a guided optical field and a surrounding medium. To have easy tracing of the spectral shift with a high extinction ratio and low transmission loss during measurement, the first side-lobe at the longer-wavelength-side stop-band end was selected as the peak of interest. In order to be applicable to optical communication systems, the side-lobe was set around 1550 nm wavelength region. In the following simulation, the polarization of TE mode for an input light was assumed.
Fig. 1. (a) Schematic structure of the silicon MS-SWBG RI sensor with structural parameters. (b) Transmission spectra of the MS-SWBG RI sensor with RI change from 1.32 to 1.33 at duty ratio of 50%. Inset shows the whole stop-band at duty ratios of 50%, 60% and 70%. (c) Electric field profiles in logarithmic scale at the 10th pillar area cross-section (left) and the whole device horizontal cross-section (right) for the wavelengths in the stop-band $(\lambda = 1550\;nm).$ and at the first side-lobe $(\lambda = 1550\;nm)$.
An example of the transmission spectra shift with the noted parameters in Fig. 1(a) is shown in Fig. 1(b) for RIs of the surrounding medium nmed of 1.32 and 1.33 by solid lines and dashed ones, respectively. The inset of Fig. 1(b) shows the whole stop-band spectra at different silicon-pillar duty ratios defined by $a/\Lambda .$ The performances of the device as an RI sensor are evaluated by a wavelength shift $\Delta \lambda ,$ an extinction ratio $ER$ and a transmission loss $TL,$ as defined in the spectra zoomed in for the yellow-highlighted wavelength region of the inset for $a/\Lambda $ = 50%. To show performance potential of the device, the RI change in the surrounding media are ranging from 1.3164 to 1.3661 assuming liquids of pure-water, sugar-dissolved waters at various sugar concentrations, acetone and isopropanol used in the experiment, and 1.32 for pure-water and 1.33 for ethanol in the simulation [31,32].
In order to make clear the guided mode in this MS-SWBG waveguide, the electric field profiles at the 10th pillar area cross-section and the whole device horizontal cross-section for TE mode are shown in Fig. 1(c) for the wavelength in the stop-band of 1500 nm and at the peak of the first side-lobe at the longer wavelength side of 1550 nm, corresponding to the transmission spectra of Fig. 1(b). It clearly indicates that the fields are mostly well confined in the slots and guide along the edges of the boundary between silicon pillars and surrounding medium due to TE mode feature, which enhances the RI sensitivities.
In the next section, the structural parameters are optimized for higher sensitivities.
3. Simulation
3.1 Modeling consideration
Simulations were conducted using Lumerical 2D finite-difference eigenmode (FDE) and 3D finite-difference time-division (FDTD) methods. Under the simulation of structural optimization, an operation wavelength is preferable to be fixed in order to avoid its effect to the confinements in silicon pillars and surrounding medium. For this purpose, $\Lambda $ of the BG is adjusted to have the first side-lobe of longer-wavelength-side stop-band near 1550 nm for the various slot widths and silicon pillar widths in the considered MS-SWBG structure, and ${n_{eff - MS - SWBG}}$ is derived with the following equation, given by S. M. Ryotov [33]:
The transmission spectral responses of the MS-SWBG sensor for the change of RI of the surrounding media were then simulated to optimize the structure for higher performances from viewpoints of large $\Delta \lambda ,$ $ER$ and low $TL,$ as indicated in Fig. 1(b) under the criteria of a rather low loss device such as TL less than 3 dB. Structural parameters in the optimization were a total waveguide width W, a slot width s, a slot count, slot configuration, a duty ratio $a/\Lambda $ and a period count N.
3.2 Simulation results
Figure 2 shows the effective RI change $\Delta {n_{eff - a}}$ of the MS-SWBG WG for an RI change of the surrounding medium $\Delta {n_{med}}$ of from 1.32 to 1.33, giving the measure of sensitivity, with parameters of a slot width s in the vertical axis and a total WG width W in the horizontal one. These results were obtained under a slot count of 3 and a silicon WG height of 340 nm. The values in “Leaky mode region” values were ignored due to ${n_{eff - a}}$ being lower than bottom layer, whereby $TL$ would be high. Dark red regions indicate larger $\Delta {n_{eff -a}}$ in the higher range of 10−3. It was found that an optimal total WG width increases as the increase of a slot width, and the maximum $\Delta {n_{eff - a}}$ was derived to be 0.0098 for $s$ = 60 nm and $W$ = 600∼800 nm. Similar result was found by E. Luan et al. [28], but they utilized a phase shift and slot counts were 4 or 5 in the MS-SWBG WG as well as a smaller WG height of 220 nm.
Fig. 2. Dependence of ${\Delta n_{eff - a}}$, by ${\Delta n_{med}}\;=\;0.01$, on a slot and a total WG widths for a WG height of 360 nm and slot count of 3.
In optimizing the structure of the MS-SWBG sensor with 3D FDTD, the dependence of the sensor characteristics, that is, $\Delta \lambda ,$ $ER$ and $TL$ on a slot count was investigated, as shown in Fig. 3(a). The slot counts of 0 to 4 were considered with a slot width of s = 60 nm and a BG period count of N = 20 at a duty ratio of 50%. As a reference, the red lines are for a slot count 0, which corresponds to a SWG at BG condition. These results indicate that the maximum values of $\Delta \lambda $ and ER are almost the same independent of the slot count, as $\Delta \lambda $ ∼ 7 nm and $ER$ ∼ 14 dB, where each optimum total WG width increases as an increase of a slot count. For slot counts 1 and 0 the maximum $\varDelta \lambda$ value would be at total WG widths below the minimum range in the graph of 500 nm. In order to predict and discuss the sensing performances, it is convenient to introduce the factor, which indicates the ratio of the optical confined component in the surrounding medium relative to the whole mode-field area, which is defined as “an optical space-confinement factor” $(OSCF),$ and it is expressed in the following equation with an electric field of TE mode Ex, as:
Fig. 3. $\Delta \lambda ,$ $ER$ and $TL$ as a function of total WG width W of the MS-SWBG RI sensors dependent on (a) slot counts with slot widths s of 60 nm, and (b) slot configurations with a slot count of 3 at duty ratio of 50%, N of 20 and a WG height of 340 nm.
The next parameters for optimizations were a slot width s and a slot configuration under a slot count of 3, since the maximum $\varDelta \lambda$ was obtained for this case in the previous discussion, and the simulated results are shown in Fig. 3(b). Three-slot configurations with various slot-width combinations were taken such as 1) the same widths of 30nm-30nm-30nm, 45nm-45nm-45nm and 60nm-60nm-60nm, and 2) the different widths of 30nm-60nm-30-nm and 60nm-30nm-60nm. The characteristic tendencies look almost similar. From the results of the same slot-width configuration, it was noticed that the wider slots have the larger wavelength sensitivities $\Delta \lambda ,$ which was due to enhanced interaction between the optical field and the surrounding medium, namely the larger $OSCF,$ and smaller $ERs$ due to the larger $TL.$ For the different slot-width configurations the characteristics were almost the same for all as the same slot-width configuration of $s$ = 45 nm, which was a result of the almost same total slot widths in the range of 120∼150 nm. In general, the narrower slot widths lead to the better characteristics, but the slot widths of s = 30 nm and 45 nm may meet difficulty in fabrication.
The dependence on a duty ratio was considered, as shown in Fig. 4(a) under a slot count of 3 and a slot width of $s$ = 45 nm. Here, duty ratios of 40% to 80% with 10% step were studied. On the contrary to the previous cases discussed above, the smaller duty ratios, which had the larger $OSCF,$ lead to the larger $TL,$ but $ER$ increased. This may be due to a high-index contrast feature of silicon pillar grating, and required large Fourier components for the first-order Bragg condition correspond to a smaller duty ratio, trading larger $TL.$ While, the largest values of $\Delta \lambda $ were almost the same of about 7 nm.
Fig. 4. $\Delta \lambda ,$ $ER$ and $TL$ as a function of total WG width W of the MS-SWBG RI sensors dependent on (a) duty ratios with N of 20, a WG height of 340 nm, s of 45 nm and a slot count of 3 and (b) transmission spectra dependent on $N$ with a duty ratio of 50%, W of 1000 nm, a WG height of 340 nm, $\Lambda $ of 436 nm and s of 60 nm.
A final investigated parameter was a period count $N,$ and it was taken from 20 to 60, as shown in Fig. 4(b) for a medium RI change by 0.01. Here, a duty ratio of 50%, a slot count of 3 and a slot width of 60 nm were used taking fabrication possibility into account. By an increase of $N,$ the first side-lobes became steeper and their peaks narrower due to the increase of total coupling strengths given by multiplication of a coupling coefficient and a BG length. The steeper first side-lobe of the BG stop-band enhanced an $ER,$ and a maximum $ER$ of about 26 dB was expected for $N$ = 40. Further increase of N such as 60 decreased the first side-lobe peak due to a larger total coupling strength, and to an increase of $TL,$ and in this case the maximum ER should be taken from the second side-lobe rather than the first one. It is noted that a wavelength shift $\Delta \lambda $ was almost the same independent of N since the effect of the $OSCF$ had no difference for N larger than 20.
To summarize the structural optimizations, the MS-SWBG waveguide was optimized considering its various parameters and fabrication point of view for $\varDelta \lambda ,$ $ER$ and $TL.$ If compactness and larger $\varDelta \lambda$ are pursued with a $TL$ smaller than 3 dB, the optimal structure has a slot count of 3 at a slot configuration of 60nm-60nm-60nm, a total WG width of about 900 nm, a duty ratio of 50% and a period count N of 20, corresponding to a length of about 9 µm. The expected performances would be $\varDelta \lambda$ of about 7 nm for $\Delta {n_{med}}$ = 0.01 corresponding to a high sensitivity $(S = \Delta \lambda /\Delta {n_{med}})$ of 700 nm/RIU, an ER of about 10 dB and a $TL$ of 2.9 dB. If a longer device size, a $TL$ at the first side-lobe of larger than 3 dB and even usage of the second side-lobe as sensing do not matter, a larger period count $N$ than 20 such as 40 or 60 would be selectable with the improvement of $ER$.
In the previous part the $OSCF$ was introduced to explain the sensing characteristics, especially the wavelength sensitivities, depending on various structural parameters such as a slot count, a slot size, a duty ratio and a total WG width. Calculated $OSCF$ as a function of a total WG width is shown in Fig. 5 with electric field profiles in TE mode for the structure with a slot count of 3 and a slot width of 60 nm. This trend seems to fit well with the wavelength shift $\varDelta \lambda$ of Fig. 3(b) and Fig. 4(a) with a maximum $OSCF$ at a total WG width of about 750 nm called as the peak width. The optical electric field profiles indicate that for wider total WG widths than the peak width optical fields tend to be more confined in larger RI silicon pillars and for narrower than that they are shifted toward a SiO2 under-cladding layer due to reduction of ${n_{eff - a}}$.
Fig. 5. $OSCF$ of the MS-SWBG WG as a function of a total WG width with a slot count of 3, a slot widths of 60 nm, a period count of 20 and a duty ratio of 50%. Insets are the electric field profiles in logarithmic scale at the 10th pillar area cross-section.
4. Experiment
4.1 Fabrication process and EBL conditions
The MS-SWBG sensors were fabricated on SOI substrates employing electron beam lithography (EBL) and inductively coupled plasma reactive ion etching (ICP-RIE) process. The SOI wafers used had 340 nm-thick silicon and 2 µm-thick buried oxide (BOX) layers. Fabrication process consisted of 8 steps, as shown in Fig. 6.
First, the SOI substrates were cleaned by submerging in acetone, isopropanol, NMD-3% for 5, 10 and 3 minutes, respectively, and finally dipping in pure water. Second, an EBL resist ZEP520A7 film diluted by the thinner ZEP-A with a concentration of 1:1 was spin-coated, and then samples were pre-baked for 3 minutes at 180 $^\circ C$. Third, EBL (ELS-7700W, Elionix) was carried out at an acceleration energy of 75 keV for a field size of 150 µm square. Fourth, the exposed samples were developed with ZEP-N50 and rinsed with ZMD-B for 30 and 60 seconds, respectively. Fifth, 20 nm-thick nickel (Ni) film was deposited with electron beam deposition (EBX-6B, ULVAC). Sixth, unwanted nickel film was removed by lift-off technique with NMD-3% at 130 $^\circ C$ for 30 minutes. Seventh, etching of 340 nm-thick silicon waveguide was done by inductively coupled plasma reactive ion etching (ICP-RIE, RIE-200iP, Samco) with gases of CF4+Ar in 8:2 mixture at a pressure of 0.1 Pa for 3 and half minutes. Finally, samples were submerged in chromium etchant at 130 $^\circ C$ for 20 minutes to remove Ni film, followed by final cleaning with pure water.
To find the minimum slot width which could be fabricated by our EBL technique, the EBL conditions were investigated. This was done using the fabrication steps 1 to 4 in Fig. 6, and the samples thus fabricated were observed by scanning electron microscope (SEM, SU8200 and SU4800, Hitachi). In the EBL exposure the MS-SWBG WGs with a slot width s of 10 to 60 nm were fabricated at dose values of 50 to 250 µC/cm2. The best dose value was 150 µC/cm2. A SEM top view of the fabricated pattern with the minimum slot width of 30 nm in which the narrow EBL resist remained for the following lift-off process as the formation of slots is shown in Fig. 7(a). For smaller slot widths than 30 nm the slot patterns were broken, as shown in Fig. 7(b).
Fig. 7. SEM images of different slot configuration designs: (a) 30nm-60nm-30nm and (b) 25nm-50nm-25nm after EB-lithography development.
4.2 Device fabrication
For the device fabrication two kinds of the MS-SWBG samples with different duty ratios of 60% and 70% in the EBL process were fabricated under a total WG width W of 1000 nm and 1150 nm at a slot count of 3. The slot configurations were in 60nm-60nm-60nm, 60nm-30nm-60nm, 30nm-60nm-30nm and 45nm-45nm-45nm with period counts N of 20 and 50. Adoption of the duty ratios in the EBL process mentioned above, which were larger than the designed one of 50%, was due to the predicted reduction of silicon-pillar widths in the following fabrication process such as the proximity effect in EBL [34] as well as possible side etching by ICP-RIE process [35].
SEM top views of the fabricated devices are shown in Fig. 8. Figures 8(a) and (b) are for the slot-width configuration of 75nm-75nm-75nm with period counts of 20 and 50, respectively, and Fig. 8(c) is their zoom in top view. It can be confirmed that silicon pillars were well isolated uniformly. These well-isolated features were not realized for the slot-width configurations including narrow slot widths of 30 nm, as shown in Fig. 8(e) where the 30 nm-width slot is located at the outside rows in this case. The realizable minimum slot width was 60 nm shown in Fig. 8(f), which may be the technical limit for our fabrication conditions. It is noted that the actual fabricated sizes were liable to be a little deviated from the designed ones, as mentioned above, but the structures with the uniformly isolated silicon pillars seem to be suitable ones with duty ratios of 52∼58%, evaluated by the zoom in images of Figs. 8(c) and (f) as well as an angled view of the silicon BG of Fig. 8(d), leading to expected performances discussed in the previous section.
Fig. 8. SEM images of the MS-SWBG WG sensors for various slot-width configurations with slot widths s and period counts $N:$ (a-d) $s$ = 75nm-75nm-75nm with (a) $N$=20 and (b) 50 and their (c) zoom in top view and (d) angled view. (e) Top view with $s$ = 30nm-60nm-30nm and (f) zoom in view with $s$ = 60 nm-60nm-60nm at $N$ = 50.
4.3 Measurements
In the experimental setup, tunable semiconductor lasers (TSL-210, Santec) with bandwidths of from 1500 to 1580 nm and from 1530 to 1610 nm were used as light sources to measure the spectral response. The input and output lights in TE mode were connected to the MS-SWBG sensors, fabricated as in Figs. 8(a)–(d) and (f), by cleaved fibers with 5.5 µm core diameters through grating couplers (GCs), formed on taper silicon waveguides with a period of 640 nm and a duty ratio of 80%. Fibers were set at angles of 10 degrees inclined from the normal to the subsutrate. The ouput fiber was connected with an optical power meter (AQ2202 with AQ2200-211 sensor module, Yokogawa). For measuring the sensitivities of the sensors for different RIs of the surrounding medium ${n_{med}},$ sample liquids were dropped on the sensing part. In order to swiftly obtain the measurement data before fast liquid-drying, especially for acetone, the automatic measurement system using LabVIEW was employed. The measurement setup is shown in Fig. 9(a). When the measurement took longer time than the liquid drying the GCs were contaminated, as shown in Fig. 9(b), and a piranha etch cleaning was carried out to confirm the reprocucibility of measurement. The sample liquids selected to measure were sugar-dissolved waters at 5%, 10% and 15% weight concentrations, acetone and IPA where estimated RI changes $(\Delta {n_{med}})$ compared to that of pure water were 0.0094, 0.0182, 0.0274, 0.0319 and 0.0497, respectively [31–32].
Fig. 9. (a) Experimental setup for measuring MS-SWBG RI sensors and (b) zoomed photographs of the sensor areas for various sample media and the contaminated case.
4.4 Experimental results
Three kinds of the MS-SWBG sensors labeled by I, II and III were measured. While measuring spurious Fabry–Pérot (FP) resonances were observed which may be caused due to the short distance between the input and output GCs being only 550 μm [36]. To remove this effect the spectra were averaged. The device I had a slot-width configuration of 75nmn-75nm-75nm, a duty ratio of the silicon-pillar of about 58%, a period count of 20 and a total waveguide width W of 1145 nm, as shown in Fig. 8(a). The device II had the same parameters as those of the device I except a period count being 50 shown in Fig. 8(b), and the device III had a slot-width configuraion of 60nm-60nm-60nm, a duty ratio of about 52%, a period count N of 50 and a total waveguide of 1000 nm.
The measured transmission spectral responses around the side-lobe at the longer wavelength-side of the stop-band are shown for the device I in Fig. 10(a) at various liquid solutions, as denoted in the inset. The side-lobes clearly shifted as a change of the RIs of the liquid samples $\Delta {n_{med}},$ from which a record-high wavelength-shift sensitivity of 730 nm/RIU was achieved for the non-suspended silicon waveguide structure with a low $TL.$ of 3 dB, though an $ER$ was about 9 dB for $\Delta {n_{med}}$ of 0.01. Figure 10(b) shows the relation between a wavelength shift $\Delta \lambda $ of the first side-lobe and a medium refractive index change $\Delta {n_{med}}$. For the device II, where a period count N increased to 50, a wavelength-shift sensitivity decreased to 705 nm/RIU with a $TL$ increased to 12 dB and an $ER$ increased to 15 dB. This characteristic may be attributed to that the larger N was liable to the steeper side-lobe and the larger $TL$ for the first side-lobe. For the device III, a wavelength-shift sensitivity was a little decreased to 680 nm/RIU and a $TL$ increased to 13.5 dB with an $ER$ of 14 dB. This may be due to a smaller duty ratio, total waveguide width and slot size compared to the previous case, decreasing the $OSCF.$ Considering that the mimal wavelength resolusion $(\varDelta {\lambda _{TSL210}})$ of the conventional optical spectrum analyzer as small as 0.001 nm, the detection limit $(\varDelta {\lambda _{TSL210}}/S)$ is expected up to 1.37 × 10−6 RIU.
Fig. 10. (a) Measured transmission spectral responses of the MS-SWBG of Device I with various liquid samples. (b) Relation between side-lobe wavelength shifts $\Delta \lambda$ and. $\Delta {n_{med}}$ for the devices I, II and III. Error bars take into account the FP resonance averaging.
5. Discussion
The record-high wavelength-shift sensitivity of 730 nm/RIU experimentally obtained was a result of the structural optimization including the adoption of a thicker silicon WG of 340 nm compared to 220 nm for the previous results because of larger interaction area between a guided TE mode and a surrounding medium to be measured. The device I had a slot width of 75 nm, which was different from the optimized one of 60 nm discussed in Sec.3, and this difference may be attributed to the cross-sectional deformation of silicon pillars in height from rectangular shape to rectangular frustum, as shown in Fig. 8(d) as an example, formed by side etching during ICP-RIE. In addition, the improved sensitivity from the simulated one may be responsible for another deformation of the silicon pillars, at the corners to become round, as shown in Fig. 8(c), leading to a slightly larger $OSCF.$ The effect of these shape deformations to the wavelength sensitivities $\Delta \lambda $ for $\Delta {n_{med}}$ = 0.01 was simulated, as shown in Fig. 11, for the possible shapes denoted in the inset. The red line is for the rectangular pillar which is the same case discussion in Sec.2, where the optimum total WG width is about 75 nm, but in the experiment that of 1100 nm was adopted from a viewpoint of fabrication process. The optimum total WG widths shifted to wider values due to effective slot width increase and for the deformed pillar structures from rectangular to rectangular frustum or ellipsoid. This may be responsible for the improved $\Delta \lambda $ obtained in the experiments. It is noted that the tendency of $\Delta \lambda $ decrease and then increase for narrowing the total WG width seems attributed to the mode profile change from shifting to the under-cladding layer to moving back around the MS-SWBG WG in weakly guiding with the stop-band not being clearly observable.
Fig. 11. Pillar shape and slot width dependence of $\Delta \lambda $ under $\Delta {\lambda _{med}}$ = 0.01 of the MS-SWBG sensor with a slot count of 3, duty ratio 50% and period count of 20 with a top-bottom width difference for rectangular frustum of 20 nm.
6. Conclusions
In conclusion, compact and high sensitive MS-SWBG RI sensors were developed. This sensor is a combination of multi-slot sub-wavelength Bragg grating WG at the first-order Bragg condition. Different from the previous works with a similar WG configuration, the wavelength peak as a sensing marker is that of the first side-lobe at the long wavelength side of the stop-band whereby a transmission loss is much lower. After the designing of the structural parameters, the devices were fabricated through the fabrication steps in high-precision EB lithography and RIE. As a result, record-high sensitivity of 730 nm/RIU and very compact size of 9.5 µm were demonstrated for the non-suspended waveguide structure. In comparison to the fully SOI-based BG sensors firstly reported so far, [12] the improvements in sensitivity and length were by 22 times higher and by 18 times smaller, respectively. Comparing with another previous result exhibiting the best performances with an MS-SWBG WG configuration where BG periods were form in a Gaussian-shape profile by P. Xu et al. [30] the sensing performances were improved by 1.4 times in higher sensitivity and 2.4 times in smaller length with a transmission loss lower by about 5 times. From these mentioned above, the MS-SWBG WG on SOI platforms with well-designed structures should be expected as a high-sensitivity and compact RI sensor, which can be applied to a wide variety of chemical, environment, and bio sensors.
Acknowledgements
The authors would like to thank Dr. Isamu Yuito of Nanotechnology Research Center of Waseda University for his support in EBL fabrication process.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
References
1. J. E. Bowers, T. Komljenovic, M. Davenport, J. Hulme, A. Y. Liu, C. T. Santis, A. Spott, S. Srinivasan, E. J. Stanton, and C. Zhang, “Recent Advances in Silicon Photonic Integrated Circuits,” Proc. SPIE 9774, 977402 (2016). [CrossRef]
2. M. Teng, A. Honardoost, Y. Alahmadi, S. S. Polkoo, K. Kojima, H. Wen, C. K. Renshaw, P. LiKamWa, G. Li, S. Fathpour, R. Safian, and L. Zhuang, “Miniaturized Silicon Photonics Devices for Integrated Optical Signal Processors,” J. Lightwave Technol. 38(1), 6–17 (2020). [CrossRef]
3. E. Luan, H. Shoman, D. M. Ratner, K. C. Cheung, and L. Christowski, “Silicon Photonic Biosensors Using Label-Free Detection,” Sensors 18(10), 3519 (2018). [CrossRef]
4. 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). [CrossRef]
5. D. Yuan, Y. Dong, Y. Liu, and T. Li, “Mach-Zehnder Interferometer Biochemical Sensor Based on Silicon-on-Insulator Rib Waveguide with Large Cross Section,” Sensors 15(9), 21500–21517 (2015). [CrossRef]
6. 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). [CrossRef]
7. C. Zhang, S. Srinivasan, Y. Tang, M. J. R. Heck, M. L. Davenport, and J. E. Bowers, “Low threshold and high speed short cavity distributed feedback hybrid silicon lasers,” Opt. Express 22(9), 10202–10209 (2014). [CrossRef]
8. X. Wang, W. Shi, R. Vafaei, N. A. F. Jaeger, and L. Chrostowski, “Uniform and sampled Bragg gratings in SOI strip waveguides with sidewall corrugations,” IEEE Photonics Technol. Lett. 23(5), 290–292 (2011). [CrossRef]
9. M. Burla, L. R. Cortés, M. Li, X. Wang, L. Chrostowski, and J. Azaña, “Integrated waveguide Bragg gratings for microwave photonics signal processing,” Opt. Express 21(21), 25120–25147 (2013). [CrossRef]
10. H.-C. Kim, K. Ikeda, and Y. Fainman, “Tunable transmission resonant filter and modulator with vertical gratings,” J. Lightwave Technol. 25(5), 1147–1151 (2007). [CrossRef]
11. A. S. Jugessur, J. Dou, J. S. Aitchison, R. M. De La Rue, and M. Gnam, “A photonic nano-Bragg grating device integrated with microfluidic channels for bio-sensing applications,” Microelectron. Eng. 86(4-6), 1488–1490 (2009). [CrossRef]
12. V. M. N. Passaro, R. Loiacono, G. D’Amico, and F. De Leonardis, “Design of Bragg Grating Sensors Based on Submicrometer Optical Rib Waveguides in SOI,” IEEE Sens. J. 8(9), 1603–1611 (2008). [CrossRef]
13. P. Prabhathan, V. Murukeshan, Z. Jing, and P. V. Ramana, “Compact SOI nanowire refractive index sensor using phase shifted Bragg grating,” Opt. Express 17(17), 15330–15341 (2009). [CrossRef]
14. S. T. Fard, S. M. Grist, V. Donzella, S. A. Schmidt, J. Flueckiger, X. Wang, W. Shi, A. Millspaugh, M. Webb, D. M. Ratner, K. C. Cheung, and L. Chrostowski, “Label-free silicon photonic biosensors for use in clinical diagnostics,” Proc. SPIE 8629, 862909 (2013). [CrossRef]
15. X. Wang, J. Flueckiger, S. Schmidt, S. Grist, S. T. Fard, J. Kirk, M. Doerfler, K. C. Cheung, D. M. Ratner, and L. Chrostowski, “A novel silicon photonic biosensor using phase-shifted bragg gratings in slot waveguide,” J. Biophotonics 6(10), 821–828 (2013). [CrossRef]
16. R. Halir, P. J. Bock, P. Cheben, A. Ortega-Moñux, C. Alonso-Ramos, J. H. Schmid, J. Lapointe, D.-X. Xu, J. G. Wangüemert-Pérez, Í. Molina-Fernández, and S. Janz, “Waveguide sub-wavelength structures: A review of principles and applications,” Laser Photonics Rev. 9(1), 25–49 (2015). [CrossRef]
17. H. Yan, L. Huang, X. Xu, S. Chakravarty, N. Tang, H. Tian, and R. T. Chen, “Unique surface sensing property and enhanced sensitivity in microring resonator biosensors based on subwavelength grating waveguides,” Opt. Express 24(26), 29724–29733 (2016). [CrossRef]
18. L. Huang, H. Yan, X. Xu, S. Chakravarty, N. Tang, H. Tian, and R. T. Chen, “Improving the detection limit for on-chip photonic sensors based on subwavelength grating racetrack resonators,” Opt. Express 25(9), 10527–10535 (2017). [CrossRef]
19. E. Chow, A. Grot, L. Mirkarimi, M. Sigalas, and G. Girolami, “Ultracompact biochemical sensor built with two-dimensional photonic crystal microcavity,” Opt. Lett. 29(10), 1093–1095 (2004). [CrossRef]
20. M. G. A. Rahman, P. Velha, M. Richard, and N. P. Johnson, “Silicon-on-insulator (SOI) nanobeam optical cavities for refractive index based sensing. In Optical Sensing and Detection II,” Proc. SPIE 8439, 84391Q (2012). [CrossRef]
21. K. Yao and Y. Shi, “High-Q width modulated photonic crystal stack mode-gap cavity and its application to refractive index sensing,” Opt. Express 20(24), 27039–27044 (2012). [CrossRef]
22. P. Xu, K. Yao, J. Zheng, X. Guan, and Y. Shi, “Slotted photonic crystal nanobeam cavity with parabolic modulated width stack for refractive index sensing,” Opt. Express 21(22), 26908–26913 (2013). [CrossRef]
23. D. Yang, S. Kita, F. Liang, C. Wang, H. Tian, Y. Ji, M. Lončar, and Q. Quan, “High sensitivity and high Q-factor nanoslotted parallel quadrabeam photonic crystal cavity for real-time and label-free sensing,” Appl. Phys. Lett. 105(6), 063118 (2014). [CrossRef]
24. S. Kim, H. M. Kim, and Y. H. Lee, “Single nanobeam optical sensor with a high Q-factor and high sensitivity,” Opt. Lett. 40(22), 5351–5354 (2015). [CrossRef]
25. J. Jágerská, H. Zhang, Z. Diao, N. Le Thomas, and R. Houdré, “Refractive index sensing with an air-slot photonic crystal nanocavity,” Opt. Lett. 35(15), 2523–2525 (2010). [CrossRef]
26. D. Yang, H. Tian, and Y. Ji, “High-Q and high-sensitivity width-modulated photonic crystal single nanobeam air-mode cavity for refractive index sensing,” Appl. Opt. 54(1), 1–5 (2015). [CrossRef]
27. F. Sun, J. Zhou, L. Huang, Z. Fu, and H. Tian, “High quality factor and high sensitivity photonic crystal rectangular holes slot nanobeam cavity with parabolic modulated lattice constant for refractive index sensing,” Opt. Commun. 399(15), 56–61 (2017). [CrossRef]
28. E. Luan, H. Yun, M. Ma, D. M. Ratner, K. C. Cheung, and L. Chrostowski, “Label-free biosensing with a multi-box sub-wavelength phase-shifted Bragg grating waveguide,” Biomed. Opt. Express 10(9), 4825–4838 (2019). [CrossRef]
29. P. Xu, J. Zheng, J. Zhou, Y. Chen, C. Zou, and A. Majumdar, “Multi-slot photonic crystal cavities for high-sensitivity refractive index sensing,” Opt. Express 27(3), 3609–3616 (2019). [CrossRef]
30. J. D. Sarmiento-Merenguel, A. Ortega-Moñux, J.-M. Fédéli, J. G. Wangüemert-Pérez, C. Alonso-Ramos, E. Durán-Valdeiglesias, P. Cheben, Í. Molina-Fernández, and Halir Robert, “Controlling leakage losses in subwavelength grating silicon metamaterial waveguides,” Opt. Lett. 41(15), 3443–3446 (2016). [CrossRef]
31. A. Belay and G. Assefa, “Concentration, Wavelength and Temperature Dependent Refractive Index of Sugar Solutions and Methods of Determination Contents of Sugar in Soft Drink Beverages using Laser Lights,” J. Laser Opt. Photonics 05(02), 187 (2018). [CrossRef]
32. J. E. Saunders, C. Sanders, H. Chen, and H. Loock, “Refractive indices of common solvents and solutions at 1550 nm,” Appl. Opt. 55(4), 947–953 (2016). [CrossRef]
33. S. M. Rytov, “Electromagnetic Properties of a Finely Stratified Medium,” JETP 2(3), 466–474 (1956).
34. R. Liming and C. Baoqin, “Proximity effect in electron beam lithography,” J. Vac. Sci. Technol. 12(1), 579–582 (2004). [CrossRef]
35. M. D. Henry, “ICP Etching of Silicon for Micro and Nanoscale Devices,” Ph.D. Dissertation, California Institute of Technology (2010).
36. Q. Zhong, V. Veerasubramanian, Y. Wang, W. Shi, D. Patel, S. Ghosh, A. Samani, L. Chrostowski, R. Bojko, and D. V. Plant, “Focusing-curved subwavelength grating couplers for ultra-broadband silicon photonics optical interfaces,” Opt. Express 22(15), 18224–18231 (2014). [CrossRef]
