The objective of the proposal is to investigate novel compact geometries that overcome the traditional temperature sensitivity of Si nanophotonics enabling future computing systems that with high bandwidth and low power.
Intellectual Merit: We propose to control the thermal drift of photonic structures by tailoring the degree of confinement in silicon waveguides. We aim to achieve structures that self compensate for temperature effects based on such waveguides with dissimilar thermal drifts. Our preliminary work shows the possibility to tailor the thermo-optic drift of an interferometric structure. This technique of thermal stabilization is completely passive and fully compatible with current microelectronics fabrication processing. We propose to use this approach for demonstrating a 40 GHz WDM modulator system stable with temperature.
Broader Impacts: The fundamental knowledge gained in this research on temperature insensitive devices and in particular temperature insensitive modulators is also applicable to other novel photonic structures such as optical routers, switches and filters as well as other optical materials system such as glass-based compounds. We propose to develop an educational program for exposing students to photonics as an interdisciplinary area and enhancing the undergraduate and graduate Electrical and Computer Engineering curriculum in the area of photonics. An interdisciplinary seminar series will be developed for undergraduate and graduate students and will cover research areas related to photonics such as: photonic materials, biomedical applications, and optical computing. We also propose to introduce novel subjects of photonics in courses of the electrical and computer engineering in Cornell university to include active research areas.
We investigated and demonstrated novel compact geometries and schemes for silicon photonic structures that self-stabilize when temperature variations are encountered. We demonstrated the control of the thermal drift of photonic structures by tailoring the degree of confinement in silicon waveguides, and showed self-compensation (passive) of temperature effects based on such waveguides with dissimilar thermal drifts. The first approach showed a device that works on the principle that the guided mode encounters different effective mode index changes with temperature in the two arms of Mach-Zehnder Interferometer (MZI), and by choosing the arm lengths carefully the overall temperature sensitivity can be set to any desired value or even brought down to zero. A schematic of the device is shown in Fig. 1. The device was fabricated in the Cornell CNF and tested in our labs. Fig.3 shows a microscope image of the MZI designed to give zero spectral shift with temperature along with insets showing SEM of the wide and narrow waveguides. The temperature insensitive MZI was found to operate over a very wide range of temperature (greater than 50 degrees) with no significant degradation in performance, as shown in Fig. 3. Fundamentally the device operates as expected as long as the dielectric refractive indices change linearly with temperature. Hence it provides an attractive platform for designing devices to work at extremely high temperatures. We also demonstrated athermal-like behavior over a wide temperature range of a ring resonator. By using the temperature insensitive approach, it is possible to include a ring resonator to the device and achieve a temperature insensitive modulator. We demonstrated this as shown in Fig. 4, where a ring resonator is coupled to a MZI, where the latter is used for thermal compensation as described above. The upper right spectrum in Fig. 4 shows the optical transmission response of the device (without any temperature effects, i.e., the desired output of the device). The lower right section of Fig. 4 shows how the tailored MZI can compensate temperature changes. The fabricated device is shown in Fig. 5a. The transmission characteristics of the device during modulation for different temperatures are shown in Fig. 5b. A finite extinction (> 3 dB) between the ON and OFF states is maintained at every temperature. To test the modulation properties of the device, a 2 Gbps square wave signal was applied to the diode. The resulting eye diagrams are shown in Fig. 5c, for various temperatures (15-50 C). The eye remains open showing that modulation is possible over a wide temperature range of 40 degrees. This is approximately 40 times the operating temperature range of 1°C for a standard resonator modulator with Q = 10,000. Lastly, using a bi-material cantilever, we demonstrate a temperature athermal operation over 14 degrees of Si ring resonators. Figure 6(a) shows a schematic of the device. The bi-material cantilever is designed such that it bends upwards as temperature increases, thereby decreasing the effective index of the coupled mode and counteracting the thermo-optic effect. This is achieved with a bottom layer of the cantilever that has a higher thermal expansion coefficient (α) than the top layer. We choose Al2O3 (α = 8e-6 K-1) for the bottom layer and Si (α = 3e-6 K-1) for the top layer. When light circulates in the resonator, a small fraction of the optical field is coupled to the cantilever above, as shown in Fig 6(a). For athermal operation we design the coupling gap so that slight changes in the position of the cantilever due to temperature fluctuations. Figure 6(d) shows the change in resonance wavelength of the TM mode (neff ~ 2.15) as a function of temperature. When the gap is optimally chosen, the resonance wavelength changes minimally over a certain temperature range yielding an athermal operating zone. We demonstrate that the resonance sensitivity changes from conventional red shift to zero to strong blue shift with temperature as the cantilever interaction with the optical mode increases, when it is brought closer to the waveguide. Fig. 6(e) shows the device transmission spectrum of the TM mode as a function of temperature over a range of 15 degrees. At temperatures of 10 C and above, the thermo-optic effect dominates and the resonance redshifts with increase in temperature. As the temperature is reduced, the cantilever moves closer to the waveguide such that any change in effective index of the mode due to thermo-optic effect is compensated by coupling to the mechanical deflection of the cantilever. Further reduction in temperature brings the cantilever even closer to the waveguide and leads to overall increase in cavity size, hence the blue shift of resonance wavelength with temperature.
