Christina Thantrakul applied for and received a Undergraduate Research Fellowship to do research with Dr. Connie Chang-Hasnain at the University of California, Berkeley for the Spring and Summer semesters in 2015. Christina completed her degree at UC Berkeley and graduated with a BS in engineering physics in December 2015. She plans to further her education through a graduate program to continue building her skills for a technical or scientific career.
Refractive Index Sensing with High Contrast Grating Resonators
Optical resonators such as ring resonators, photonic crystal resonators, and microsphere resonators, are used in many applications, including on-chip filters, lasers, and optical sensing experiments. During my Undergraduate Research Fellowship with the Center for Integrated Access Networks (CIAN), Professor Connie Chang-Hasnain, and Graduate Student Researcher Tianbo Sun, I studied High Contrast Grating (HCG) resonators designed for refractive index sensing. These resonators have high quality factors, large physical sizes, are easily coupled, and are very sensitive to refractive index changes.
HCGs are formed by a thin layer of semiconductor gratings surrounded by low index material such as air or SiO2 . The structure used for this experiment consists of a silicon HCG on a layer of SiO2. The HCG structure has a high index contrast at both the entrance and exit planes and can be designed to have a high quality factor resonance at surface normal or oblique incidence angle. The HCG’s quality factor and resonance wavelength depends on the dimensions of the grating: grating thickness tg, period Λ and duty cycle .
The setup used to characterize the HCG resonators is described as follows: Light from a continuous wavelength laser was focused on to the HCG resonator by a 5X objective lens. The reflected spectrum was collected by the same objective lens and directed to the detector by a beam splitter. A linear free-space polarizer placed between the laser source and the objective lens ensured the proper polarization of the EM wave incident on the resonator (transverse electric, TE). A second linear polarizer placed in front of the detector allowed the detection of the resonance mode by filtering all other reflections.
In air, I characterized several samples designed for different purposes, refractive index sensing of liquids among them. To facilitate repeatability, I used Matlab to synchronize the laser, the detector, and two Thorlabs steppers. The highest Q found in air was approximately 7000.
After characterizing many HCG resonators in air, the HCG resonators were coated with standardized refractive index liquids from Cargille Labs. The same process used to coat a silicon wafer with photoresist was used to coat the resonators with a very thin film of refractive index liquid.
The spectrum of the HCG resonators was recorded using the Matlab synchronized laser, detector, and steppers. The resonators were then cleaned and the process was repeated four more times for a total of five different standard refractive index liquids. We observed a linear relationship between the measured redshift of the resonance wavelength and the increase in the refractive index of the liquid coating the resonator, resulting in a sensitivity of a 530 nm shift per refractive index unit.
We then replaced the standardized liquids with deionized water, using HCG resonators designed for liquids with refractive indices around 1.3 to 1.4. Due to the water’s low viscosity, we were able to pipet small amounts of water onto the resonators. We measured a significant redshift (23.6 nm) in the resonance wavelength of the resonator immersed in water.
We also performed qualitative measurements to determine the HCG resonators’ wavelength change due to the change in concentration of aqueous D-glucose by dissolving glucose in the deionized water covering the resonators. As we predicted, the increase in the glucose concentration (and hence the refractive index of the solution) resulted in a corresponding increase in the HCG resonator’s resonance wavelength.
Achieving a sufficiently consistent, thin, and still layer of liquid on the resonator was challenging. To solve these challenges, we designed a small microfluidic package of Polydimethylsiloxane (PDMS), shown the figure below, to hold both the resonator and liquid. This packaging maintains a consistent and shallow liquid depth, is small and transportable, is sealed to prevent evaporation, and is easily and quickly constructed. These characteristics may be advantageous when using these resonators with volatile liquids such as ethanol or very thin layers of water.
During July and August of 2015, I continued my work on the HCG resonator sensor in Professor Connie Chang-Hasnain’s lab. I worked on recording the spectrum of HCG resonators using the PDMS microfluidic packaging shown in figure 1. Using two such PDMS packages, I investigated the spectrums of over two hundred HCG resonators. I used the free-space setup described above to measure the reflection of the HCG resonators with the PDMS chamber filled with air. Approximately one third of the resonators I measured had a visible resonance in their spectrums. Then, I filled the chamber with distilled water and repeated the measurements for all the HCG resonators. None of the HCG resonators had a measureable resonance peak in their spectrums.
There are a couple of possible reasons we do not see the expected peak at the resonant wavelength: the layer of water may still be thick enough to bend the incident light far enough from normal incidence to prevent excitation of the fundamental mode, which gives rise to the resonance I measured; the water may absorb enough power to prevent detection of the resonance peak in the spectrum.
We have considered a few solutions to the aforementioned problems. One possibility is to direct the light close to the resonator via an optical fiber inserted into to the liquid layer, which is used to excite and collect the reflection spectrum of the HCG resonator. This allows the light to bypass the complicated optical setup and strike the resonator through only a thin layer of water. Another option would be to shine the laser from below the HCG resonator, eliminating the need to pass through any liquid at all. This would also reduce the amount of optical components needed to manipulate the laser, further reducing reflections and optical power loss. The amount of power lost would depend on the material through which the light travels to reach the resonator and its thickness.
- C. J. Chang-Hasnain and W. Yang, “High-contrast gratings for integrated optoelectronics,” Adv. Opt. Photon. 4(3) 379-440 (2012)