We present a bi-layer lift-off fabrication approach to create low-loss amorphous titanium dioxide (TiO2) integrated optical waveguides and resonators for visible and near-infrared applications. This approach achieves single-mode waveguide losses as low as 7.5 dB/cm around 633 nm and 1.2 dB/cm around 1550 nm, a factor of 4 improvement over previous reports, without the need to optimize etching conditions. Depositing a secondary 260-nm TiO2 layer can reduce losses further, with the optimized process yielding micro-ring resonators with loaded quality factors as high as 1.5 × 105 around 1550 nm and 1.6×105 around 780 nm. These losses render our TiO2 devices suitable for visible and telecommunications applications; in addition, the simplicity of this lift-off approach is broadly applicable to other novel material platforms, particularly using near-visible wavelengths.
© 2015 Optical Society of America
Titanium dioxide (TiO2) exhibits a wide, indirect bandgap (3.1 eV), a high refractive index (>2.2), a large optical nonlinearity [1–5], and a negative thermo-optic coefficient, making it an attractive candidate for future integrated optical biosensor [6–9], quantum optical [10–12], and all-optical switching applications [13–15] as well as a key component for temperature-stable integrated optical devices [16–19]. Furthermore, TiO2’s visible transparency make it compatible with the low water-absorption window for biophotonic sensing applications [6,20]. These properties are in contrast to silicon, which must operate at wavelengths longer than 1.1 μm , and place TiO2 in a similar class as silicon nitride , with the advantages of an increased refractive index and reduced luminescence . Realizing these possibilities in integrated photonic circuits requires the ability to create low-loss TiO2 waveguides and high quality-factor (Q-factor) resonators . As an additional challenge for sensing and biophotonics, these devices should be uncladded to enable interaction with surrounding gases, liquids, or molecules , which poses a particular challenge at visible wavelengths due to Rayleigh scattering from surface roughness. Toward this end, TiO2 resonators have demonstrated loaded Q-factors as high as 2.2 × 104 at 635 nm using sputtered amorphous TiO2 , 6×104 at 660 nm, 1×105 at 980 nm, and 1×104 at 1550 nm using sol-gel TiO2  and 5.1 ×104 at 1574 nm in polycrystalline anatase TiO2 . As intrinsic material absorption is negligible at these wavelengths, thin-film quality (determined by deposition conditions), and scattering from roughness in channel waveguides (determined by structuring conditions) limit these Q-factors. By comparing the film and structured-waveguide losses, we can determine the structuring-associated loss. For example, using reactive ion etching (RIE) with fluorine-based chemistry, losses increase from 1.2 dB/cm in amorphous TiO2 thin films to 28 dB/cm in channel waveguides at 633 nm and from 0.4 to 4 dB/cm at 1550 nm . Scattering from sidewall roughness created during etching is clearly the dominant source of loss , particularly for visible wavelengths (as the loss scales as 1/λ2) . Therefore, developing a fabrication method to reduce structuring-associated losses is critical for future TiO2 photonic applications.
There are several approaches to reduce losses in conventional etched TiO2 waveguides, from engineering waveguide geometries to post-etch surface treatments. For example, multi-mode waveguides better isolate the fundamental modes from the sidewalls, enabling losses as low as 9.7 dB/cm at 633 nm  and 3.4 at 980 nm  (extrapolated from quality-factor measurements) in TiO2. Alternatively, applying an additional TiO2 layer post-etching has reduced the surface roughness in rib waveguides, leading to a loss reduction from 5 dB/cm to 2.4 dB/cm at 1550 nm  (pre and post deposition, respectively). Other post-structuring treatments, including laser annealing or wet chemical etching, have also been demonstrated in other materials [30–32]. Although optimizing etching and post-processing can decrease losses, we seek to develop a contactless structuring process to avoid etching-associated losses altogether.
Our strategy is to avoid etching-related surface roughness by using a dielectric lift-off process. This process minimizes surface roughness by avoiding contact with the waveguide itself [Fig. 1(a)]. We expose a lift-off resist/deep-ultraviolet resist (LOR/DUV) bilayer using positive-tone lithographic techniques to create openings and undercut channels [Fig. 1(b)]. Next, we deposit TiO2 over the resist and through the opening onto the exposed substrate, and then liftoff the resist and excess TiO2 to form channel waveguides [Fig. 1(c)]. By directly depositing channel waveguides without contact with other materials (e.g. resists and hard masks such as chromium [1,2]), we significantly reduce sidewall roughness while avoiding several processing steps. This technique has successfully been applied to several chalcogenide glasses for operation at infrared wavelengths [33–35]. However, this dielectric lift-off approach is unexplored for visible photonics, which are more susceptible to surface scattering, and thus should greatly benefit from this technique.
In this communication, we report the first application of a dielectric lift-off fabrication process to TiO2 photonics at both visible and telecommunications wavelengths. We demonstrate single- and multi-mode TiO2 waveguides with four-times lower losses than previous reports. These structured-waveguide losses become equal to thin-film values at telecommunications wavelengths, thus demonstrating that we can minimize structure-related loss. To show the applicability of this approach, we demonstrate high quality-factor micro-ring resonators around 1550 and 780 nm, and discuss strategies to reduce losses further.
2. Fabrication and simulation
We fabricate TiO2 waveguides and micro-ring resonators using a dielectric lift-off process. We start with a thermal oxide (3 μm) silicon wafer with a 260-nm LOR-layer and a 600-nm DUV resist layer [Fig. 1(a)]. After exposing using a 4x-reduction DUV stepper (λ = 248 nm), developing produces openings in the photoresist as defined by our mask dimensions with a 750-nm undercut into the LOR layer [Fig. 1(b)]. Next, we deposit 250 nm of TiO2 using reactive direct-current (DC) sputtering of titanium metal with oxygen at room temperature. Ellipsometry on the bare film shows an index of refraction between 2.33–2.23 for 633–1550 nm. After sputtering, we lift-off using n-methyl pyrrolidinone to obtain channel waveguides, as shown in Fig. 1(c). As the geometry is non-polygonal, we identify waveguide dimensions based on the mask’s design width. For rib waveguide and micro-ring fabrication, we deposit a second 265-nm thick TiO2 slab layer post lift-off. The waveguides do not have an upper cladding. By using an undercut bilayer-resist geometry and dielectric lift-off, we form channel and rib waveguides that avoid contact with photoresist to minimize edge and surface roughness.
To identify waveguides that support a single, well-confined mode at different wavelengths, we extract dimensions of our waveguides using atomic force microscopy (AFM) and calculate the number, polarization, and confinement of the modes using an eigenmode solver. An exemplary AFM image of a TiO2 waveguide [Fig. 1(c)] reveals root-mean-squared surface roughnesses of 1.2 nm and 3.8 nm on the film and waveguide, respectively, calculated using commercial software and standard methods . For comparison, previous reports show surface roughness of 0.4 nm on the film using radio-frequency (RF) sputtering  and etched sidewall roughness of 10 nm . We approximate the waveguide profile, h(x), using a Gaussian expansion of the form:Fig. 2(a). Interestingly, the waveguide width profiles (full-width at half maximum, FWHM) are 210–280 nm wider than the masks design width, which likely stems from the angular spread of the sputtering deposition. In addition, we observe the waveguide height increases with wider openings due to an aperture-effect during deposition. We use these fitting-functions to simulate the fundamental transverse-electric-like (TE-like) modal distribution. We then determine dimensions that support singlemode operation (with no transverse-magnetic-like mode) with modal core-confinement between 40–50% for each wavelength [Figs. 2(b)–2(e)]. Based on these simulation results, we choose 250-, 500-, 750-, and 1000-nm wide waveguides for wavelengths of 633, 980, 1310, and 1550 nm, respectively. By maintaining similar confinement across different wavelengths, we can directly compare losses as a function of wavelength to understand the merits of this fabrication approach.
3. Loss measurements
We measure planar waveguide losses in an unstructured film and use this result as a basis for comparison to structured waveguides. Using the same deposition procedure, we prepare a 250-nm thick TiO2 film on an identical wafer. We couple into the fundamental TE mode using prism coupling and measure the losses using a camera . Comparing structured to planar waveguide loss values allows us to decouple the film versus structuring losses to provide insight into loss-reduction strategies.
We measure the optical propagation loss for both TiO2 channel and rib waveguides by measuring the relative insertion loss for devices of different lengths. We fabricate serpentine waveguides with 8 repeated units for total lengths from 12–28 mm in 4 mm increments . We show two of these units in Fig. 3(a). Dicing and polishing our chips ensures consistent facets. Using a series of lasers coupled to single-mode fibers, we quantify the losses of our waveguide for visible and NIR wavelengths (633, 780, 850, 980, 1310, and 1550 nm). Using fiber collimators followed by a half-wave plate and polarizer, we excite the fundamental TE mode of our waveguides using a 0.85 NA objective. An automated piezo stage ensures consistent coupling. An identical objective collimates the output and a large-area silicon or InGaAs detector measures the signal (visible or infrared wavelengths, respectively). We estimate a coupling efficiency of −7 and −4 dB/facet for channel and rib waveguides, respectively.
We measure the transmission of three identical, adjacent waveguides. To avoid spurious data points from scattering on our uncladded waveguides we analyze only the waveguide with maximum transmission for each length. We calculate the error bars based on the linear fit uncertainty and we find that repeated measurements of at least three times display variations that fall within this uncertainty. Plotting the maximum relative transmission data as a function of length allows us to fit and extract the propagation losses, as shown in Fig. 3(b).
We show the extracted loss values versus wavelength for single-mode channel, multi-mode rib, and planar waveguides in Fig. 3(c). We find that the planar waveguide losses decrease monotonically from 633 to 980 nm, are experimentally identical from 980 nm to 1310 nm, and display a slight increase around 1550 nm. The single-mode losses are higher than the planar-waveguide losses at shorter wavelengths, decreasing monotonically with increasing wavelength, and becoming equal to the planar losses around 1550 nm. Comparing single- and multi-mode waveguides, we find the multi-mode rib waveguide losses are roughly 1 dB/cm lower from 633–980 nm, become equal to single-mode losses at 1310 nm and abruptly increase from 1310 nm to 1550 nm. These reduced losses at visible wavelengths in high index-contrast multi-mode versus single-mode waveguides are consistent with other reports at telecommunications wavelengths .
To understand the wavelength-dependent losses of the structured waveguides, we fit a line to the loss-versus-wavelength data plotted using a log-log scale. This representation determines the exponential wavelength dependence of the loss (1/λx). For 633–1310 nm, we find x-values of 2.1 (R2 = 0.981) and 2.0 (R2 = 0.997) for the channel and ridge waveguides, respectively. This analysis approximately shows a 1/λ2-dependence. This scaling is consistent with Rayleigh surface scattering in planar waveguides [27, 39], which suggests that our lift-off waveguides are limited by surface scattering at visible and near-infrared wavelengths. The convergence of losses for planar and single-mode waveguides at telecommunications wavelengths suggests that the film quality determines the low-loss limit.
Comparing our losses to the literature shows that our lift-off approach can reduce losses significantly at visible wavelengths, which has been a major challenge for TiO2 visible integrated optics. For comparison, in our single-mode channel waveguides we observe visible losses of 7.5 dB/cm around 633 nm that are nearly a factor of 4 lower than previous single-mode reports (28 dB/cm, 637 nm) . Our multi-mode rib waveguides also exhibit lower losses (6.5 dB/cm) than other reported multi-mode waveguides (9.7 dB/cm) , also around 633 nm. The low losses we measure reflect the advantage of the bi-layer liftoff approach for single-mode waveguides at visible wavelengths.
4. Ring resonators
To demonstrate the utility of the dielectric lift-off fabrication technique, we fabricate and measure micro-ring resonators around 1550 nm and 780 nm, which are key components for sensors as well as nonlinear and quantum optical applications. Using a similar fabrication method, we form rib waveguides with mask design widths of 750 nm and 500 nm into micro-ring resonators with radii (r) of 150 μm for operation around 1550 nm and 780 nm, respectively. Using a straight bus-waveguide, we couple into the micro-rings using a design-gap of 1.4 μm for 1550 nm and 1.2 μm for 780 nm. We show a 150-μm ring in Fig. 4(a). We characterize the rings at telecommunications wavelengths by sweeping with a tunable laser and observing the transmission, as shown in Fig. 4(b). We observe a free-spectral range (FSR) of 1.052 nm around 1550 nm. For 780-nm measurements, we current-tune a distributed feedback (DFB) laser around a single resonance. We find these resonators are undercoupled around 780 nm, prohibiting accurate transmission measurements. Consequently, we measure the light scattered from the ring using a top-view camera as a function of wavelength to determine the quality factor.
Around 1550 nm, we fit individual transmission resonances using a Lorentzian-lineshape and measure the FWHM, as shown in Fig. 5(a). We observe loaded quality factors (Qloaded = λ/Δλ) as high as 1.5 × 105. Using our measured FSR, we calculate the group index using :
Conservatively assuming that the resonator is undercoupled, we find that Q0 corresponds to 2 × 105. We use this intrinsic quality factor to estimate the loss using:
Around 780 nm, we perform a similar fit using integrated top-view images of the scattered light within the ring as shown in Fig. 5(b). We observe loaded Q-factors of 1.6×105. Assuming we are strongly undercoupled, this Q-factor approaches Q0 and serves as a lower-bound estimate. We see that while the loaded Q-factor is slightly higher than around 1550 nm, our estimated Q0 is lower than for 1550 nm, consistent with Rayleigh scattering. From simulation, we calculate a group index of 2.562. Using this value with our Q0 estimate we obtain a corresponding loss of 5.6 dB/cm. This value falls between similar measurements taken from 633–850 nm with a design-width of 750 nm [Fig. 3(c)].
Our measurements show the highest reported Q-factors for TiO2 micro-ring resonators at telecommunication wavelengths. While sol-gel micro-goblet resonators have reported Q0’s of 1:98×105 around 980 nm, these values fall precipitously for the 1550-nm band, down to 2×104, attributed to surface adsorbed water and hydroxyl (OH) groups in thin-films measurements . Meanwhile, anatase micro-ring resonators have demonstrated loaded Q-factors as high as 5×104 at telecommunications wavelengths . Comparing these results to our measurements, we see that our fabrication approach improves losses by a factor of 4 around 1550 nm. Although there are no previous reports of TiO2 resonators around 780 nm, our measured Q-factors are higher than measurements around 660 nm (6 × 104) and comparable to 980 nm .
We show that our dielectric lift-off process significantly reduces losses over etching methods without the need for etch chemistry optimization. At telecommunication wavelengths, our single-mode and planar waveguide losses are comparable, suggesting that we are not limited by structuring in this wavelength regime. We attribute the residual deposition-related loss to scattering within the film and possible sub-bandgap absorption from impurities. Therefore, optimizing the deposition process, such as using RF-sputtering , should further decrease losses by a factor of 2 or greater at telecommunications wavelengths. At visible wavelengths, surface Rayleigh scattering limits losses. Reducing the surface roughness via deposition conditions or post-treatments such as laser annealing should reduce losses [30–32]. Additionally, using a top cladding (such as SiO2) to decrease the index contrast can minimize the impact of the surface roughness. Our rib waveguide data supports this analysis, correlating decreased losses with reduced overlap between the fundamental mode and the waveguide surface.
We attribute the disagreement between losses in multimode waveguides and resonators around 1550 nm to the partial excitation of the first higher-order mode during endfire couplings. Our simulation shows a higher-order mode that is near cutoff, making it more susceptible to surface scattering and contributing to increased losses. This higher-order mode is suppressed in the ring data due to its decreased coupling efficiency in the evanescent-field coupler.
Our study reveals several fabrication considerations for this dielectric lift-off approach. First, the coupling region is limited by the minimum gap imposed by the undercutting geometry, which must be made smaller for shorter wavelength operation. Optimization of the resist and processing conditions should alleviate this problem. Alternatively, we can employ coupler configurations that require larger gaps for critical coupling . Second, to reduce the device’s insertion loss, we observe that overlap-based power-coupling calculations account for roughly −2 dB/facet in all cases. This analysis suggests that improved polishing or cleaving may improve coupling. In future devices, we can also achieve even higher coupling efficiency using tapered waveguides [43, 44].
Several applications could benefit from this fabrication approach as well as from low-loss integrated TiO2 waveguides directly. Applications such as Mach-Zehnder interferometers  and directional couplers  are two such examples if paired with a similarly deposited material. In addition to advancing applications at telecommunications wavelengths, TiO2 is compatible with the abundance of single and pair-photon sources that operate at visible wavelengths for quantum information applications [11,45]. Low-loss, high-index materials, such as TiO2, could enable high-density integrated photonic chips for quantum optics . Additionally, TiO2 possesses an indirect bandgap that suppresses fluorescence, which is often a source of background noise for high sensitivity applications. In addition to quantum devices, TiO2’s visible compatibility has advantages over NIR for spectroscopy applications. These advantages include compatibility with the low-absorption window of water, increased Raman signal (as Raman scattering scales as 1/λ4), and compatibility with inexpensive, high-performance silicon detectors. Using these advantages to develop TiO2 devices for spectroscopy is a subject that we will discuss in a future communication.
We have demonstrated a bi-layer lift-off fabrication approach to structure TiO2 waveguides and resonators. We have shown that this approach, which does not etch the core material, can reduce losses of TiO2 single-mode waveguides by nearly a factor of 4 compared to previous reports. These losses are as low as 7.5 dB/cm around 633 nm and 1.2 dB/cm around 1550 nm. While we find that surface scattering limits the guiding loss in the visible band, in the telecommunication band the devices become limited to losses we measure in unstructured-films. These low-losses in TiO2 waveguides and high-Q resonators will be critical building blocks for future applications in telecommunications, quantum optics, and biosensors. In addition to achieving the lowest reported losses in TiO2 waveguides using a dielectric lift-off technique, our work shows how this approach can be applied to new materials beyond TiO2. Lastly, by circumventing etching-recipe development, this approach reveals an avenue to explore and develop new integrated optical materials quickly and easily.
CCE and JS conceived of the basic idea for this work. CCE, and CL designed and carried out the experiments, and analyzed the results. All authors (CCE, CL, and JS) contributed to the analysis of the data and the writing of this paper. We would like to thank Chris Phare for his assistance measuring resonators, as well as Katia Shtyrkova and Chuhyon John Eom for feedback on this manuscript. Support for this work is provided by the Cornell Center for Materials Research under (National Science Foundation) Grant DMR-1120296, part of the NSF MRSEC Program. CCE acknowledges support from the Kavli Institute at Cornell for Nanoscience Science. This work was performed at the Cornell NanoScale Facility, a member of the National Nanotechnology Infrastructure Network (NSF ECCS-0335765) and at the Cornell Center for Materials Research (National Science Foundation, DMR-1120296).
References and links
1. J. T. Choy, J. D. B. Bradley, P. B. Deotare, I. B. Burgess, C. C. Evans, E. Mazur, and M. Lončar, “Integrated TiO2 resonators for visible photonics,” Opt. Lett. 37, 539–541 (2012). [CrossRef] [PubMed]
2. J. D. B. Bradley, C. C. Evans, J. T. Choy, O. Reshef, P. B. Deotare, F. Parsy, K. C. Phillips, M. Lončar, and E. Mazur, “Submicrometer-wide amorphous and polycrystalline anatase TiO2 waveguides for microphotonic devices,” Opt. Express 20, 23821–23831 (2012). [CrossRef] [PubMed]
3. C. C. Evans, K. Shtyrkova, J. D. B. Bradley, O. Reshef, E. Ippen, and E. Mazur, “Spectral broadening in anatase titanium dioxide waveguides at telecommunication and near-visible wavelengths,” Opt. Express 21, 18582–18591 (2013). [CrossRef] [PubMed]
4. C. C. Evans, J. D. B. Bradley, E. A. Martí-Panameño, and E. Mazur, “Mixed two- and three-photon absorption in bulk rutile (TiO2) around 800 nm,” Opt. Express 20, 3118–3128 (2012). [CrossRef] [PubMed]
5. C. C. Evans, K. Shtyrkova, O. Reshef, M. Moebius, J. D. B. Bradley, S. Griesse-Nascimento, E. Ippen, and E. Mazur, “Multimode phase-matched third-harmonic generation in sub-micrometer-wide anatase TiO2 waveguides,” Opt. Express 23, 7832 (2015). [CrossRef] [PubMed]
7. K. Schmitt, K. Oehse, G. Sulz, and C. Hoffmann, “Evanescent field sensors based on tantalum pentoxide waveguides –a review,” Sensors 8, 711–738 (2008). [CrossRef]
9. B. Adhikari and S. Majumdar, “Polymers in sensor applications,” Prog. Polym. Sci. 29, 699–766 (2004). [CrossRef]
10. J. L. O’Brien, A. Furusawa, and J. Vuckovic, “Photonic quantum technologies,” Nat. Photonics 3, 687–695 (2009). [CrossRef]
12. J. Wang, A. Santamato, P. Jiang, D. Bonneau, E. Engin, J. W. Silverstone, M. Lermer, J. Beetz, M. Kamp, S. Höfling, M. G. Tanner, C. M. Natarajan, R. H. Hadfield, S. N. Dorenbos, V. Zwiller, J. L. OBrien, and M. G. Thompson, “Gallium arsenide (GaAs) quantum photonic waveguide circuits,” Opt. Commun. 327, 49–55 (2014). [CrossRef]
13. D. J. Moss, R. Morandotti, A. L. Gaeta, and M. Lipson, “New CMOS-compatible platforms based on silicon nitride and Hydex for nonlinear optics,” Nat. Photonics 7, 597–607 (2013). [CrossRef]
14. B. J. Eggleton, B. Luther-Davies, and K. Richardson, “Chalcogenide photonics,” Nat. Photonics 5, 141–148 (2011).
15. J. S. Aitchison, D. C. Hutchings, J. U. Kang, G. I. Stegeman, and A. Villeneuve, “The nonlinear optical properties of AlGaAs at the half band gap,” IEEE J. Quantum Electron. 33, 341–348 (1997). [CrossRef]
17. S. S. Djordjevic, K. Shang, B. Guan, S. T. S. Cheung, L. Liao, J. Basak, H.-F. Liu, and S. J. B. Yoo, “CMOS-compatible, athermal silicon ring modulators clad with titanium dioxide,” Opt. Express 21, 13958–68 (2013). [CrossRef] [PubMed]
19. O. Reshef, K. Shtyrkova, M. G. Moebius, S. Griesse-Nascimento, S. Spector, C. C. Evans, E. Ippen, and E. Mazur, are preparing a manuscript to be called “Polycrystalline anatase titanium dioxide micro-ring resonators with negative thermo-optic coefficient.”
21. B. Jalali and S. Fathpour, “Silicon Photonics,” J. Lightwave Technol. 24, 4600–4615 (2006). [CrossRef]
22. L.-W. Yin, Y. Bando, Y.-C. Zhu, and Y.-B. Li, “Synthesis, structure, and photoluminescence of very thin and wide alpha silicon nitride (α-Si3N4) single-crystalline nanobelts,” Appl. Phys. Lett. 83, 3584 (2003). [CrossRef]
25. J. Park, S. K. Ozdemir, F. Monifi, T. Chadha, S. H. Huang, P. Biswas, and L. Yang, “Titanium dioxide whispering gallery microcavities,” Adv. Opt. Mater. 2, 711–717 (2014). [CrossRef]
28. M. Furuhashi, M. Fujiwara, T. Ohshiro, M. Tsutsui, K. Matsubara, M. Taniguchi, S. Takeuchi, and T. Kawai, “Development of microfabricated TiO2 channel waveguides,” AIP Adv. 1, 32102–32105 (2011). [CrossRef]
29. M. Häyrinen, M. Roussey, V. Gandhi, P. Stenberg, A. Säynätjoki, L. Karvonen, M. Kuittinen, and S. Honkanen, “Low-loss titanium dioxide strip waveguides fabricated by atomic layer deposition,” J. Lightwave Technol. 32, 208–212 (2013). [CrossRef]
30. J. Hu, N.-N. Feng, N. Carlie, L. Petit, A. Agarwal, K. Richardson, and L. Kimerling, “Optical loss reduction in high-index-contrast chalcogenide glass waveguides via thermal reflow,” Opt. Express 18, 1469–1478 (2010). [CrossRef] [PubMed]
31. K. K. Lee, D. R. Lim, L. C. Kimerling, J. Shin, and F. Cerrina, “Fabrication of ultralow-loss Si/SiO2 waveguides by roughness reduction,” Opt. Lett. 26, 1888 (2001). [CrossRef]
32. S. Dutta, H. Jackson, J. Boyd, R. Davis, and F. Hickernell, “CO2 laser annealing of Si3N4, Nb2O5, and Ta2O5 thin-film optical waveguides to achieve scattering loss reduction,” IEEE J. Quantum Electron. 18, 800–806 (1982). [CrossRef]
33. J. A. Frantz, L. B. Shaw, J. S. Sanghera, and I. D. Aggarwal, “Waveguide amplifiers in sputtered films of Er3+-doped gallium lanthanum sulfide glass,” Opt. Express 14, 1797 (2006). [CrossRef] [PubMed]
34. J. Hu, V. Tarasov, N. Carlie, N.-N. Feng, L. Petit, A. Agarwal, K. Richardson, and L. Kimerling, “Si-CMOS-compatible lift-off fabrication of low-loss planar chalcogenide waveguides,” Opt. Express 15, 11798 (2007). [CrossRef] [PubMed]
35. J. Hu, N. Carlie, N.-N. Feng, L. Petit, A. Agarwal, K. Richardson, and L. Kimerling, “Planar waveguide-coupled, high-index-contrast, high-Q resonators in chalcogenide glass for sensing,” Opt. Lett. 33, 2500 (2008). [CrossRef] [PubMed]
36. ISO 4287/1997 Geometrical product specifications (GPS)–Surface texture: Profile method–Terms, definitions and surface texture parameters.
38. A. Harke, M. Krause, and J. Mueller, “Low-loss singlemode amorphous silicon waveguides,” Electron. Lett. 41, 1377 (2005). [CrossRef]
39. J. S. Foresi, M. R. Black, A. M. Agarwal, and L. C. Kimerling, “Losses in polycrystalline silicon waveguides,” Appl. Phys. Lett. 68, 2052 (1996). [CrossRef]
40. P. Rabiei, W. H. Steier, C. Zhang, and L. R. Dalton, “Polymer micro-ring filters and modulators,” J. Lightwave Technol. 20, 1968–1975 (2002). [CrossRef]
41. P. E. Barclay, K. Srinivasan, and O. Painter, “Nonlinear response of silicon photonic crystal microresonators excited via an integrated waveguide and fiber taper,” Opt. Express 13, 801 (2005). [CrossRef] [PubMed]
44. T. G. Tiecke, K. P. Nayak, J. D. Thompson, T. Peyronel, N. P. de Leon, V. Vuletić, and M. D. Lukin, “Efficient fiber-optical interface for nanophotonic devices,” Optica 2, 70 (2015). [CrossRef]
46. N.-N. Feng, S. Liao, D. Feng, P. Dong, D. Zheng, H. Liang, R. Shafiiha, G. Li, J. E. Cunninham, A. V. Krishnamoorthy, and M. Asghari, “High speed carrier-depletion modulators with 1.4V-cm Vπ L integrated on 0.25μm silicon-on-insulator waveguides,” Opt. Express 18(8), 7994–7999 (2010). [CrossRef] [PubMed]
47. R. Soref, “Mid-infrared 2 × 2 electro-optical switching by silicon and germanium three-waveguide and four-waveguide directional couplers using free-carrier injection,” Photon. Res. 2(5), 102–110 (2014). [CrossRef]