Jared Keeley and Dr. Aaron Hawkins, Electrial and Computer Engineering
Main Text
In the field of micro-electromechanical systems (MEMS) and micro-opto-electromechanical systems (MOEMS) fabrication, photoresists are used to form 3D structures on a scale of micro- or even nanometers. This is necessary to create the microelectronics used in many devices. One structure that photoresist can be used to create is a hollow-core optical waveguide. To complete the waveguide something must be used to remove the photoresist from the core. Piranha is a mixture of hydrogen peroxide (H2O2) and sulfuric acid (H2SO4) that can be used to remove photoresist. One goal of my ORCA project was to analyze piranha’s effectiveness in fabricating hollow-core optical waveguides. It had been unofficially observed that piranha was faster at fabricating these waveguides than alternative methods. My project needed to confirm this observation. My project’s final goal was to find the optimal piranha mixture and conditions for fabricating hollow-core optical waveguides.
There are a few basic steps to creating a hollow-core optical waveguide. The first step is to form the core of the waveguide by patterning photoresist into the desired shape on top of a planar surface such as a silicon wafer. SU8 (MicroChem Corp.) is a photoresist used in many MEMS and MOEMS devices, and it was the photoresist used in my project to create the sacrificial core of the waveguides. After the core has been deposited, a material with appropriate optical properties must be deposited over the core to form the actual structure of the waveguide. For my project SiO2 was deposited over the SU8 core. The next step is to expose the ends of the waveguide so that the core can be removed. The equipment necessary to complete all of these steps is available in the BYU cleanroom. After the core is removed the waveguide will be hollow, and thus the hollow-core optical waveguide will be complete. The major focus of my project was analyzing the effectiveness of piranha in removing the core. It is important to have a basic knowledge of these fabrication steps to understand the results of my project.
One challenge of my project was finding a piranha mixture that could quickly remove SU8 without damaging the hollow-core optical waveguides. Piranha can be created with different ratios of H2O2 to H2SO4. To determine which ratio would remove SU8 the fastest, waveguides were immersed in piranha mixtures with different ratios of H2O2 to H2SO4 for one hour at a constant temperature. Each waveguide was examined to see how much SU8 had been removed. It was determined that a piranha mixture with a ratio of (1:20) (H2O2: H2SO4) produces the highest etch rate of SU8 [1]. However, many of the faster mixture ratios caused damage to the waveguides such as cracks or breaks. Waveguides with damage such as this are no longer usable. Thus, it needed to be determined what piranha mixture ratio would produce a high or 100% yield of usable waveguides. Further experiments revealed that a piranha mixture of (1:1) (H2O2: H2SO4) had the highest etch rate while still achieving 100% yield at 100 °C [1]. Therefore, it was determined that a piranha mixture of (1:1) (H2O2: H2SO4) was the best to use with the geometry of the SU8 waveguides used in my project.
The temperature of the piranha was another factor that had to be considered in my project. Once again a balance needed to be found between speed and yield. At higher temperatures piranha will etch SU8 faster, but it may also cause damage to the waveguides. Waveguides were immersed in piranha mixtures of (1:1) (H2O2: H2SO4) at different temperatures. The results showed that a temperature of 100 °C had the fastest etch rate while still achieving 100% yield [1]. I learned that there are many factors to consider in experiments like those done for my project. You must keep the others constant and only change one factor at a time to properly study the effects of that factor. In my project, keeping temperature the same in the first experiments while varying mixture ratio, and then keeping the ratio the same while varying temperature is an example of this principle.
There were several other factors we had to study. For example, it had to be determined how often the piranha mixture should be replaced with a fresh mixture because H2O2 self-decomposes into H2O and O2. It was determined that if cost is not an issue then the piranha mixture should be replaced every 60 to 90 minutes, but since cost usually is important it is satisfactory to only replace the mixture every 24 hours [1]. It also needed to be determined if piranha had negative effects on the materials that commonly form the waveguide structure such as SiN, SiO2, and Si. It was determined that piranha did not alter or only negligibly altered the roughness, index, and thickness of those materials [1]. I saw the importance of team work on these parts of the project as I worked with my mentor and a graduate student to complete them.
After the analysis of piranha was complete, we wanted to have a practical example of the use of piranha in fabrication. For the example we created ARROWs (anti-resonant reflecting optical waveguides) which are a more complicated form of hollow-core optical waveguides than those used in the analysis. ARROWs can be used to perform optical fluorescence sensing and detecting of single particles. Previously, it typically took 6-8 weeks to remove the core of an ARROW using Nanostrip. Our optimized piranha mixture removed the ARROW core in about 5 days [1]. This example showed that the optimized piranha mixture significantly improved the fabrication speed of ARROWs while still achieving high yields.
The results of my ORCA project were published in an article in the Journal of Micromechanics and Microengineering. In this report I have summarized much of that article, and referenced it many times. Those who desire more information may use the reference below. This project has given me valuable undergraduate research experience. In the future, we will continue to find ways to improve ARROWs and other MEMS and MOEMS devices.
