Standard silicon photonic strip waveguides offer a high intrinsic refractive index contrast; this permits strong light confinement, leading to compact bends, which in turn facilitates the fabrication of devices with small footprints. Sub-wavelength grating (SWG) based waveguides can allow the fabrication of low loss devices with specific, engineered optical properties. The combination of SWG waveguides with optical micro-resonators can offer the possibility of achieving resonators with properties different from the traditional SOI rings. One important property that SWG rings can offer is decreased light confinement in the waveguide core; this improves the resonator’s sensitivity to changes in the cladding refractive index, making the rings ideal for refractive index sensing applications. In this paper, we present the design and experimental characterization of SWG based rings realized on SOI chips without upper cladding (permitting their use as sensors). The fabricated rings offer quality factors in the range of ~1k-6k, depending on SWG parameters. Based on the comparison of experimental and simulated data we expect sensitivities exceeding 383 nm/RIU in water and 270 nm/RIU in air, showing excellent potential for use in sensing applications.
© 2015 Optical Society of America
The silicon photonics platform has recently made impressive progress, in particular several devices and systems for telecommunication have been demonstrated , and it is now attracting research interests also for other applications requiring compact devices and high fabrication volumes. One important application is sensing, which can be used for biological analysis  (e.g. molecule screening, biomarkers investigation, reactions monitoring, and drug development) as well as for environmental monitoring  (e.g. pollutant and toxin monitoring, food safety). In all those cases, it is not obvious that standard integrated silicon photonics devices and components (particularly the standard SOI strip waveguides based on a silicon core 500 nm wide and 220 nm high) are the optimized solution for the specific application. Having more degrees of freedom during the design steps to customize the optical properties could lead to the fabrication of devices better suited to a particular application.
A very promising way to tailor the optical properties of integrated devices is to use sub-wavelength grating (SWG) waveguides, which exploit a periodic, segmented waveguide core to properly guide the optical mode and offer a customized degree of optical confinement [4,5]. While traditional waveguides have a continuous high refractive index core, which allows for high field confinement, SWGs have a waveguide core made up of small silicon blocks, which repeat periodically and are interlaced with a lower refractive index material (i.e. the upper waveguide cladding, Fig. 1). The period of the blocks is upper limited by the Bragg condition; their aim is not to filter out or reflect some spectral components, but to propagate light in a low loss but application-tailored way.
SWGs waveguides were first proposed in the early 1990s to make more flexible implanted or doped core waveguides , but were not extensively investigated due to the lower demand for optical device integration at that time. Only very recently, they have been proposed again as a way to obtain tailored refractive index waveguides using the silicon on insulator platform [5,7]. Building upon these preliminary studies we have demonstrated several devices that can be fabricated using SWG waveguides, such as tapers, bends, and directional couplers [8,9]. Using this previous knowledge, we have designed and fabricated, for the first time, SOI ring resonators based on SWG with the use of no upper cladding, thus suitable for sensing. SWG based micro-rings add to the flexibility offered by SWG waveguides the many advantages of having a resonator configuration, which is a basic component for several optical devices, such as filters , modulators , detectors , and sensors . Very recently SWG filters on the SOI platform and covered with silicon dioxide have been demonstrated  mainly for telecommunication applications, however, this paper shows the possibility of using SWG resonators without any upper cladding, paving the way for their use as evanescent sensors in gaseous and fluidic environments. Furthermore, the described modeling techniques are not application-specific and thus will aid in the design of SOI SWG ring resonators with application-tailored properties for other fields as well (from sensing and real-time monitoring to telecommunications and optical interconnects).
In this manuscript we describe the modeling and optimization of SWG-based ring resonators, and we give some details about their fabrication and discuss the results of SWG rings’ experimental characterization. Our SWG-based resonators have no upper cladding, so we have measured their performance in air to demonstrate their potential use as gas sensors; however they can be used also in fluids, using microfluidic channels bonded on top of the chip . Finally, we compare the experimental performance to our simulation models, summing up where our work can be improved and paving the way for future work on SWG resonators’ modeling and sensing characterization.
To the authors’ knowledge, this is the first time that integrated optical resonators are demonstrated using SWG waveguides with no upper cladding. They were fabricated on an SOI platform, with fabrication methods compatible with CMOS foundry standards. The combination of different components in SWG-based rings, such as the straight SWG bus, the directional coupler, the bends, and the tapers, needs to be carefully implemented in order to optimize the resonators and to achieve high quality factors.
2. Sub-wavelength gratings (SWGs): basic principles and optical properties
Sub-wavelength gratings are periodic structures, like Bragg gratings , but they differ from this widely used structures on SOI chips because they are not operated in their resonant condition. They are used to tailor the optical properties and effective refractive index of the optical waveguide, and they are achieved using a composite waveguide core based on the periodic alternation of two materials with different refractive indexes. On an SOI chip, the first material, with higher index of refraction, nH, is silicon. The second material, with lower index of refraction nL, can be the same material used as waveguide top cladding (e.g. SiO2, SU-8 , or air [8,9]). Even if the waveguide core is discontinuous, the optical field is confined and forced to propagate by total internal reflection, like in a traditional optical fiber or waveguide. The light propagation in the SWG waveguide can be modeled by treating the SWGs as a continuous strip waveguide with an equivalent refractive index, neq, which lies between nL and nH and can be calculated using the process described in  and . neq depends on the refractive indexes of the core materials, and also on the grating period Π, and duty cycle δ, which defines how much core volume is occupied by one material with respect to the other (δ = Πon/Π, where Πon is the length of the silicon blocks in the core). By varying the volume ratio of higher and lower refractive index materials, the SWG waveguide optical properties can be modulated. The shape of the propagating mode in these structures is very peculiar, since it has a periodic behavior along propagation axis , and a remarkable amount of the field propagates outside the higher refractive index material, i.e. silicon. In regular strip waveguides, however, the mode shape is constant along propagation axis. It is worth noting that the equivalent refractive index model is very useful to have a practical and easy to use model for the SWG waveguide, but it is not comprehensive, since it does not take into account radiation loss due to core discontinuities as well as scattering loss. Radiation loss is due to the nature of light propagation in the SWG structure, and can be minimized by changing SWG parameters ; scattering loss is related to the technological process, since depends on several factors such as Si surface roughness and etch precision.
To understand why in SWG waveguides we have more light propagation outside the silicon volume, we have to consider two different aspects: (1) the core is discontinuous and light also propagates in the voids between the silicon blocks; (2) using the equivalent refractive index model the SWG can be represented by a strip waveguide with the same cross-section but with lower refractive index with respect to a traditional silicon strip waveguide, thus light is clearly less confined and the evanescent field is broader for SWGs . This second effect is summarized in Table 1, reporting the equivalent refractive index step, Δneq, and effective index neff (calculated using the modeling process described in the next section), for the SWG waveguides used in the rings described in this paper. The equivalent strip waveguide has a refractive index that can be calculated as expressed in Eq. (1), where ncl is the refractive index of the upper cladding.
3. SWG based rings modeling and optimization
SOI resonators have been deeply investigated in the last decade, since they offer the possibility to fabricate very compact optical devices (with footprints of few hundreds microns squared or even less) combined with unique properties and very high quality factors . These properties mean that the light can be effectively ‘trapped’ in the device itself, so selective filters or very sensitive sensors can be achieved. The performance of a racetrack resonator is strictly related to the properties of its constituting waveguides and components, and its resonance wavelength, λres, can be expressed using the resonance equation [Eq. (2)] .Eq. (3) .20]) and the equivalent refractive index method, as described in . As explained previously, the SWG is first simplified using an equivalent strip waveguide; the analyses are then performed on the waveguide transverse cross section. The optical properties of straight and bent SWG waveguides were calculated varying several waveguide and SWG parameters, as well as the properties of the directional coupler. Simulations allowed us to extract several optical properties, among them the effective and group indexes, neff and ng respectively, propagation loss due to material absorption, and mode mismatch loss. Those results were combined by a dedicated Matlab script, which evaluates the single bus ring properties (quality factor, Q, free spectral range, FSR, and extinction ratio, ER, of resonance peaks), using the resonance equation given in Eq. (2), as well as the quality factor described by Eq. (3). It is possible to use a similar method to evaluate the performance of double bus resonators. The rings were tuned to resonate at around 1550 nm, which is the typical wavelength used for telecommunications; however, the central wavelength can be changed in the simulation according to the application wavelengths, in this case the use of different central wavelengths would affect the choice of design parameters for an optimized SWG and ring. Using MODE for the simulations, the ring response was evaluated for several different structure parameters and ring radii (10-40 μm) in order to quantify these parameters’ effects on ring performance. Some of the SWG based structures were also simulated using the finite difference time domain method (Lumerical FDTD Solutions ,), and results were compared with MODE ones for completeness. Furthermore, to ensure validity of the simulation methods, full-ring 3D FDTD simulations were also used to simulate smaller rings (10- 20 μm radius) for comparison with the other simulation methods. Figure 2 shows some mode profiles (from 3D FDTD simulations) on SWG transverse cross-section, Fig. 2(a) and 2(b), and on the longitudinal axis, Fig. 2(c), for straight waveguide, 10 μm bend and directional coupler (a, b, c, respectively). Some relevant SWG waveguide parameters, along with monitor positions (red line for 2-D cross-sectional monitor and dotted box for longitudinal monitor), are schematically shown in the top-view depicted in Fig. 2(d).
It is worth noting that the SWG structures, which were simulated and fabricated as a part of this work, are compatible with standard SOI technology; they use a 220 nm thick silicon layer to form the core blocks, silicon oxide as the lower cladding, and no upper cladding. The SWG rings were simulated (and subsequently measured) with air as the top cladding. The refractive index discontinuity in the SWG waveguide core is due to the use of silicon (nH = 3.47 at λ = 1550 nm) and air (nL = 1 at λ = 1550 nm).
Using our simulation results, we designed and fabricated point-coupled SWG based ring resonators, coupled to a straight SWG waveguide bus. The SWG bus tapers out to a standard SOI strip waveguide connected to the sub wavelength grating couplers (GCs) to couple the light in and out of the chip . SEM images of some of our SWG ring resonators are presented in Fig. 3. Figure 3(a) shows the ring test structures, packed in order to optimize chip area usage; GCs connected to the same device are 127 μm apart to match the off-chip fiber array (more details on the measurements setup will be given in next paragraph). Rings with different radii and SWG parameters were fabricated. Figures 3(b)-3(c) show the ring coupling region, from lower to higher magnification, and Fig. 3(d) shows the tapers used to convert strip waveguides in SWG and vice-versa.
4. Fabrication and test
The devices presented in this paper were fabricated using SOI wafers (from Soitech, Peabody, MA), with a top silicon layer that is 220 nm thick. The fabrication was carried out at the University of Washington (UW)’s Microfabrication Facility (MFF) , using a JEOL JBX-6300FS Direct Write E-Beam Lithography System (EBL) (Peabody, MA). As presented in our previous publication on SWG components , EBL at UW has been optimized to produce silicon photonic components with performance compatible with foundry fabrication [23,24].
After fabrication, the chip was tested automatically, using our automated optical test setup, which has been shown previously . A dedicated Matlab code controls and optimizes the chip position in order to have the grating couplers aligned with the fiber array, to inject the light into the chip and collect the output from the devices. By specifying the devices’ coordinates, thousands of them can be measured in few hours, scanning their response on a 100 nm wavelength span. To the aim of this paper devices’ spectra were measured on a wavelength span around 1550 nm (1500 nm- 1580 nm), which is the typical wavelength used for third window telecommunication signals. On chip, grating couplers targeted to EBL single etch process were used to couple light in/out from/to the fiber array . The off chip optical fiber array (from PLC Connections, Columbus, OH) is four-port polarization maintaining optical fiber array; its ports can be used as inputs or outputs, depending on the specific configuration of devices on chip. The fiber array is used to couple light from the tunable laser to the chip and to collect the devices’ outputs, connecting them to the optical detectors. More information on the automated test setup we are using at UBC can be found in our previous publications , as well as information regarding the features and design of the on-chip grating couplers . In the case of fluidic measurements we have a controlled syringe pump as well, integrated with all the other setup devices by the Matlab code; more information of the optofluidic platform can be found in .
5.1 Ring experimental characterization
Each of the designed SWG based ring resonators was fabricated two times on the same SOI chip, spaced widely across the chip. The first instance was fabricated in one spatial quarter of the chip (from now on named Q1) and the other in a different quarter (from now on Q2). Thus, for each device (meaning for each set of design parameters, namely SWG Π, δ and ring radius, R) we collected two sets of measurements to include fabrication variability . Figure 4(a) depicts typical measured spectra from two identical SWG ring resonators fabricated in the two different chip quarters (wavelength sweep is done with a resolution of 10 pm). All spectra were normalized by subtracting the response of an optical link containing only the GCs . Both rings have 20 μm radius, with SWG Π of 250 nm and δ = 0.8. The spectra of the two resonators (from now on we will use RR1 to indicate the resonator on Q1 and RR2 to indicate the one on Q2) are overlaid and represented with different colors. As expected, the FSR of RR1 and RR2 are almost identical (4.92 nm versus 4.96 nm) but they present a shift of about 2 nm of their central wavelength. This result is completely in agreement with the findings of . On the other hand, the variability of fabrication parameters has a strong impact on the loss of the resonators (the differences are amplified by the nature of the resonator), thus the two rings present different mean values for the Q factor (1.2k for RR1 versus 1.7k for RR2) and ER (~20 dB for RR1 versus ~10 dB for RR2). Figure 4(b) depicts a magnified version of the spectrum from the RR2 around the central wavelength region of its spectrum; the resonance peaks and their linewidths are clearly shown.
As described in the previous section, the rings on the EBL SOI chip were measured using the automated test setup. All the results from working rings were analyzed using a dedicated Matlab script; measured quality factors, Q, and Extinction Ratio, ER (in dB) versus resonator radius are summarized in Figs. 5(a)-5(b), Q factor, and Figs. 5(c)-5(d), ER. The results from the devices in the two different chip quarters are reported using different markers (crosses for Q1 and squares for Q2). We can observe a trend in the results, with devices in Q2 behaving slightly better both in terms of higher quality factor and higher extinction ratio. Both of these metrics would be related to the resonator losses, as the ER is dependent upon the relationship between the coupling and loss in the device. In addition to these differences due to the position of the rings on the chip, the performance of the rings show a dependence on the SWG duty cycle. Rings made up of gratings with smaller duty cycle (δ = 0.7) present on average a higher Q and ER, in particular for larger rings (R≥30μm). This is reasonable, taking into account that 3D FDTD simulation of SWG waveguides showed that the loss of waveguides with δ = 0.8 are about one order of magnitude higher than the loss for δ = 0.7, as reported in . Rings with larger radii (R≥30μm) and higher duty cycles (δ>0.8) will be more affected by this length dependent loss, hindering the performance and lowering the Q of larger rings with higher duty cycles. As can be observed in Figs. 5(c)-5(d), in few cases the measured ER is lower than 10 dB, in these cases there is an error in the Q factor measured from the ring through port. Accordingly, in those rings the estimated value for the Q factor has been properly corrected depending on ER value [Figs. 5(a)-5(b)].
In contrast, the rings employing larger duty cycles have higher Q and ER for R = 10μm. This is in good agreement with what we have observed and described in , where TE SWG bends have been studied and analyzed. SWG bends with lower duty cycles and smaller radii presented higher mode mismatch loss (more than one order of magnitude larger than their counterparts with δ = 0.8) and thus higher total bend loss (about two times the loss per bend of bends with δ = 0.8). The lower field confinement of SWG waveguides with δ = 0.7 hinders the performance of smaller rings. When the mode mismatch decreases thanks to increased radius, these rings present Q factors of around 6k. This result is of remarkable importance in particular for applications in refractive index sensing, because the SWG rings with larger bend radii and lower duty cycles may have elevated Q factors combined with a substantial increase of the evanescent field penetration, with more than the 30% of the field intensity propagating in direct contact with the upper cladding (that is the medium under test in the case of sensing applications) due to the core discontinuities. The penetration depth of the evanescent field of the SWG waveguides presented in this paper is about 199 nm for δ = 0.7 and about 162 nm for δ = 0.8, which is 1.7 and 1.4 times higher than the penetration depth of a regular SOI strip waveguide with the same cross-section (i.e. 500 nm wide and 220 nm high).
5.2 Ring performance discussion
In Fig. 6(a) we present the simulation result for the electromagnetic field propagation in a 10 μm radius SOI SWG ring resonator coupled to an SWG waveguide bus, calculated via 3D FDTD simulation. Light is injected in the SWG bus (in the lower left side of the image the source is represented by a black arrow) and effectively coupled to the resonator; propagation is visualized using a 2D longitudinal monitor. FDTD was used to extract the Q and the FSR from the simulation results of 10 μm and 20 μm rings, for comparison with experimental results. Figure 6(b) reports the extracted quality factor from simulations (dotted lines) together with experimental results (continuous line) and related errors due to the measurements of double copies of the same device onto different chip quarters. As discussed, for smaller radii the SWG based resonators with smaller duty cycles present a higher Q, this is clearly shown in simulation and experimental results. On the other hand, the Q factors achieved by the simulations are higher than the measured ones, showing that there are some loss sources that are not included in our simulation model. This is in agreement with our study on SWG bends, reported in , where we noticed that the loss per 90° bend is underestimated by 3D FDTD simulations. One source of loss that is not included in the simulations is the wall roughness of the silicon blocks; this roughness introduces unwanted field scattering. The introduced scattering loss has a greater effect than in traditional strip waveguides, since the SWG field not only interacts with waveguide sidewalls but also with longitudinal walls of silicon blocks in the core. This effect is expected to be particularly prevalent in bends, where the mode is pushed towards the outer edge of the waveguide and causes radiation loss, which would add to the scattering losses. It is worth noting that SWG waveguides with loss comparable with traditional strip waveguides have been reported ; however, they have different geometries with respect to the SWG waveguides reported in this manuscript, and they are covered with silicon dioxide. The thermal deposition of an upper cladding reduces scattering loss, since it reduces the number of surface states at the edges of silicon blocks and it diminishes the roughness of the sidewalls .
Figures 7(a)-7(b) depict the SWG resonator FSR and group index, ng, comparing their value obtained by simulations and experimental measurements. The group index has been extracted from the data using its relation with FSR and total length of the resonator, L, that is ng = λ2/(FSR × L). Experimental and simulation results are in good agreement, with the simulated FSR (via 3D FDTD) slightly higher than the measured one. This effect is also shown by the results achieved for the group index; in fact, the simulated ng is slightly lower than the one extracted from measurements, with a bigger difference between model and measurements for the higher δ. This can be caused by imprecisions in the etches between the silicon blocks constituting the SWG waveguide core. This effect is more apparent when the features are smaller (for δ = 0.8 the spacing between silicon blocks in the core is only 50 nm; this is a challenging feature size for EBL writing with regular pitch, that is 6 nm). As shown in Fig. 3(c), SWGs with higher duty cycles yield spaces between the silicon blocks which are actually deformed, and not rectangular as designed. These spaces will impact the actual optical properties of the SWG waveguide.
Because the simulations results give a good estimation of the SWG rings’ group index, we can use them to calculate the expected sensitivity for SWG based ring resonators. Equations (4)-(5) show the expressions we used to evaluate respectively the sensitivity of the ring resonators, Sring, and the bulk sensitivity of a waveguide to cladding variations, Swg (details can be found in ).
The simulated SWG waveguide sensitivities (evaluated for λ = 1550 nm) and the related SWG ring sensitivities versus SWG duty cycle are reported in Figs. 8(a)-8(b); they were calculated using MODE and equivalent refractive index method. SWG waveguides were simulated with air, as well as water, as upper cladding. In particular, the results in water for the SWG waveguide bulk sensitivity (Swg) are in good agreement with values that have been recently reported in , using a 3D model and FDTD, further validating our modeling method. SWG sensitivity results have been plotted with the simulation results of the same quantities for strip and slot waveguides, showing the beneficial effect of the use of SWG based resonators as sensors. In particular, SWGs present a remarkably higher sensitivity with respect to traditional strip waveguides, both in terms of Swg and Sring. Indeed, for the ring resonators the use of SWGs gives up to more than 7-times enhancement of the sensitivity, and in the worst case (i.e. δ = 0.9) there is a 2-times sensitivity enhancement. With respect to slot-based resonators, SWGs present room for improvement, in particular when smaller duty cycles are used (δ≤0.7). Devices with air as the upper cladding and δ<0.7 are nonfunctional due to insufficient confinement of the field in SWG waveguide core; these cases are not plotted.
Furthermore, we have recently presented  some preliminary characterization of SWG ring resonators as fluidic sensors, using them to measure sodium chloride salt concentrations in water. A more comprehensive characterization of SWG rings is in progress and is outside the aim of this paper, but it is worth mentioning the potentiality held by SWG ring resonators for sensing applications. We have summarized the measured performance of ring resonators based on strip, slot and SWG waveguides, as reported in  and , in Table 2. Simulation results reported in Fig. 8(b) are in good agreement with the data presented in the Table.
In this manuscript we presented the modeling, design, and test of sub-wavelength gratings based ring resonators fabricated in an SOI platform. This is the first time that SWG based ring resonators with no upper cladding were demonstrated. As a consequence of this work, SWG based resonators can now be used on silicon photonics chips for the fabrication of tailored resonator-based devices, such as sensors and filters. SWG-based resonators offer the flexibility and versatility of SWG devices combined with the advantages of resonators, such as small footprint and compact design; for these reasons they can be of particular importance in sensing applications. We designed our devices using MODE and the equivalent refractive index method, and we used 3D FDTD to simulate some full-ring structures; simulation results are in good agreement with the measurements from fabricated SWG rings (with radii in the range of 10-40 μm). In particular, we compared the two simulations methods for the smaller SWG rings (R = 10-20μm); both methods accurately modeled SWG structures with smaller duty cycles. This effect may be due to fabrication errors that have a greater effect on structures with larger duty cycles due to challenging etch dimensions (50 nm or less). Fabricated devices present a lower quality factor with respect to simulated values, due to the presence of scattering loss (potentially accentuated by the SWG core discontinues) that was not included in our model. The Q factors of the SWG rings measured in air are in the range of ~1000-6000, with ER in the range of ~5-35 dB. Very recently, we have also reported the first measurements of these SWG rings in fluidic environment. In this case we measured a Q factor of ~4000, showing that the performance of SWG rings is not limited by water absorption but more likely by scattering loss due to the discontinuous nature of the SWG core . In air, the best device performance is achieved for rings with δ = 0.7 and larger radii (R≥30μm); this design configuration presents moderate propagation loss and reduced mode mismatch loss. The SWG model, validated by the measurements, allows us to estimate the sensitivity of SWG-based rings in water; when smaller duty cycles are used, SWG-based rings can offer a ~7-times improvement of the sensitivity with respect to traditional SOI strip waveguide based rings and a ~1.6-times improvement with respect to rings based on slot SOI waveguides (in the case of δ = 0.6). The reported preliminary measurements in water are very promising , and future work will perform further characterization and validation of SWG resonators as biological and chemical sensors. SWG rings improve upon the performance of slot and strip based rings in terms of limit of detection and sensitivity, as reported in Table 2. One important aspect for future performance enhancement of SWG based ring sensors will be the reduction of bend loss in order to improve resonator Q factor. The results presented in this manuscript show that SWG rings can pave the way for ultra-sensitive integrated optical biosensors.
The authors would like to thank CMC Microsystems, Menthor Graphics, and Lumerical Solutions Inc. for providing the simulation software. We gratefully acknowledge the Centre for High-Throughput Phenogenomics at the University of British Columbia (UBC) for assistance with SEM imaging. We are also grateful to Prof. NAF Jaeger at UBC for his insights and support and to SA Schmidt for fruitful discussions. We wish to thank Prof. DM Ratner and the University of Washington (UW) Royalty Research Fund, NSF CBET (Award nos. 0930411 and 1264174), and the Washington Research Foundation. We gratefully acknowledge NSERC CREATE Silicon Electronic Photonic Integrated Circuits (SiEPIC) training program. This work was made possible by a National Priorities Research Program grant from the Qatar National Research Fund. Devices were fabricated by Richard Bojko, at the University of Washington Nanofabrication Facility (WNF), a member of the NSF National Nanotechnology Infrastructure Network.
References and links
1. M. Hochberg and T. Baehr-Jones, “Towards fabless silicon photonics,” Nat. Photonics 4(8), 492–494 (2010). [CrossRef]
5. P. J. Bock, P. Cheben, J. H. Schmid, J. Lapointe, A. Delâge, S. Janz, G. C. Aers, D. X. Xu, A. Densmore, and T. J. Hall, “Subwavelength grating periodic structures in silicon-on-insulator: a new type of microphotonic waveguide,” Opt. Express 18(19), 20251–20262 (2010). [CrossRef] [PubMed]
6. D. Ortega, J. M. Aldariz, J. M. Arnold, and J. S. Aitchison, “Analysis of “quasi-modes” in periodic segmented waveguides,” J. Lightwave Technol. 17(2), 369–375 (1999). [CrossRef]
7. P. Cheben, P. J. Bock, J. H. Schmid, J. Lapointe, S. Janz, D. X. Xu, A. Densmore, A. Delâge, B. Lamontagne, and T. J. Hall, “Refractive index engineering with subwavelength gratings for efficient microphotonic couplers and planar waveguide multiplexers,” Opt. Lett. 35(15), 2526–2528 (2010). [CrossRef] [PubMed]
8. V. Donzella, A. Sherwali, J. Flueckiger, S. Talebi Fard, S. M. Grist, and L. Chrostowski, “Sub-wavelength grating components for integrated optics applications on SOI chips,” Opt. Express 22(17), 21037–21050 (2014). [CrossRef] [PubMed]
9. V. Donzella, J. Flueckiger, A. Sherwali, S. Talebi Fard, S. M. Grist, and L. Chrostowski, “Compact and broad band directional coupler for sub-wavelength grating SOI components,” IEEE Photonics Conference (2014).
11. G. T. Reed, G. Mashanovich, F. Y. Gardes, and D. J. Thomson, “Silicon optical modulators,” Nat. Photonics 4(8), 518–526 (2010). [CrossRef]
12. D. F. Logan, P. Velha, M. Sorel, R. De La Rue, A. P. Knights, and P. E. Jessop, “Defect-Enhanced Silicon-on-Insulator Waveguide Resonant Photodetector With High Sensitivity at 1.55 m,” IEEE Photon. Technol. Lett. 22(20), 1530–1532 (2010). [CrossRef]
13. V. Donzella, S. Talebi Fard, and L. Chrostowski, “Fabrication and experimental characterization of cascaded SOI micro-rings for high-throughput label-free molecular sensing,” IEEE Photonics Conference (IPC) 614–615 (2013).
15. L. Chrostowski and M. Hochberg, Silicon Photonics Design (Cambridge University Press, 2014)
17. W. Bogaerts, P. De Heyn, T. Van Vaerenbergh, K. De Vos, S. Kumar Selvaraja, T. Claes, P. Dumon, P. Bienstman, D. Van Thourhout, and R. Baets, “Silicon microring resonators,” Laser Photonics Rev. 6(1), 47–73 (2012). [CrossRef]
18. 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] [PubMed]
21. Y. Wang, X. Wang, J. Flueckiger, H. Yun, W. Shi, R. Bojko, N. A. F. Jaeger, and L. Chrostowski, “Focusing sub-wavelength grating couplers with low back reflections for rapid prototyping of silicon photonic circuits,” Opt. Express 22(17), 20652–20662 (2014). [CrossRef] [PubMed]
23. M. Gnan, D. S. Macintyre, M. Sorel, R. M. De La Rue, and S. Thoms, “Enhanced stitching for the fabrication of photonic structures by electron beam lithography,” J. Vac. Sci. Technol. B 25(6), 2034 (2007). [CrossRef]
24. Y. Chen, J. Feng, Z. Zhou, C. J. Summers, D. S. Citrin, and J. Yu, “Simple technique to fabricate microscale and nanoscale silicon waveguide devices,” Front. Optoelectron. China 2(3), 308–311 (2009). [CrossRef]
25. S. M. Grist, S. A. Schmidt, J. Flueckiger, V. Donzella, W. Shi, S. Talebi Fard, J. T. Kirk, D. M. Ratner, K. C. Cheung, and L. Chrostowski, “Silicon photonic micro-disk resonators for label-free biosensing,” Opt. Express 21(7), 7994–8006 (2013). [CrossRef] [PubMed]
26. L. Chrostowski, X. Wang, J. Flueckiger, Y. Wu, Y. Wang, and S. Talebi Fard, “Impact of fabrication non-uniformity on chip-scale silicon photonic integrated circuits,” Optical Fiber Communication Conference, Th2A–37 (2014). [CrossRef]
27. 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). [CrossRef] [PubMed]
29. J. Gonzalo Wangüemert-Pérez, P. Cheben, A. Ortega-Moñux, C. Alonso-Ramos, D. Pérez-Galacho, R. Halir, I. Molina-Fernández, D.-X. Xu, and J. H. Schmid, “Evanescent field waveguide sensing with sub-wavelength grating structures in silicon-on-insulator,” Opt. Lett. 39(15), 4442–4445 (2014).
30. S. A. Schmidt, J. Flueckiger, W. Wu, S. M. Grist, S. Talebi Fard, V. Donzella, P. Khumwan, E. R. Thompson, Q. Wang, P. Kulik, J. T. Kirk, K. C. Cheung, L. Chrostowski, and D. M. Ratner, “Improving the performance of silicon photonic rings, disks, and Bragg gratings for use in label-free biosensing,” SPIE Optics + Photonics, 9166 (2014).
31. S Talebi Fard, S.M. Grist, V Donzella, S.A. Schmidt, J. Flueckiger, X Wang, W Shi, A Millspaugh, M Webb, DM Ratner, KC Cheung, and L Chrostowski, “Label-free silicon photonic biosensors for use in clinical diagnostics,” SPIE OPTO 862909 (2013)