Optical Fiber coupler for Concentrated Solar Photovoltaics

This is, by far, one of the most exciting projects I worked on with Ernest Demaray, a researcher studying graded refractive index thin films. We created a thin film of Aluminum-Titanium Oxide that has a very low “extinction coefficient”, which allows light to travel through the medium with fewer losses, and a tunable refractive index. The idea behind it is to absorb as much optical energy as possible from reflective dishes and guide the light into a coupler that would heat up salt to create energy for a turbine. We were able to make a beautiful film, but Ernest still has to do some modeling to make the “black hole” material come to reality.

How I contributed:

Ran individual depositions and found sputtering transition point using historesis.

Set-up and solved minor issues with the tool to get the films out.

Analyzed refractive index and k-value of film using filmetrics spectroscopy tool.

Looking up at a sputter target when chamber lid is open, can see oxidation marks from sputtering oxide.

The red surface it where the tuned refractive index material would be to absorb the light and move it.

Via filling

We attempted to fill in Via’s ~1:1 dimension ratio with Aluminum. We were not able to fill the Vias in all the way, which is what the customer wanted.

How I contributed:

Tuned process conditions and recipe steps to get best results.

Hand cleaved coupons which were sputtered on, located imaging locations for SEM in optical microscope, then imaged cross-sections of the Vias (as seen on the left).

Worked with team to think of and develop a setup and then set it up on “Sapphire” tool.

Collected and documented the results from depositions done on “Sapphire” and “Bumper” tools (Bumper is a 3-chamber Axcela™ tool).

This shows one of the early attempts to fill the via before tuning process parameters.

High Aspect Ratio Self-Ionization Sputtering

Tango Systems Inc. has successfully deposited a continuous film of Titanium + Copper in Vias with 15:1 aspect ratios. This was done on our “Sapphire” tool by directing high DC power to the copper target, tuning the bias power, and stopping the flow of argon gas during deposition steps. This is the highest TSV step coverage known in the industry.

How I contibuted:

Ran process with team and helped to image via cross-section using SEM microscopy.

Documented process conditions.

Attained and tabulated SEM step coverage results.

Pictured above is the analysis of wall thickness down the length of the via.

3-Dimensional Reactive Plasma Vapor Deposition

This project consisted of getting a relatively uniform coating of a non-reflective thin film of Aluminum Oxide on the inside of a pyramid shaped object using reactive sputtering. We were not able to get the gas ratios correct on the underside and top side at the same time.

How I contributed:

Ran each individual deposition and prepared the setup on our “Sapphire” tool (single-chamber Axcela™ PVD tool).

Shared ideas with team to improve results and helped execute new design of experiments

Found front and backside gas ratios to develop anti-reflective (dark colored) coating.

Made a jig to run Argon gas directly from the backside cooling to beneath our substrate.

Reported process conditions and results to keep a record for further analysis.

Opened PVD Chamber with single 6" plate with thick quartz plate and open shutter.

Closer Image of the plate and 3D Object

Close up of hole that object was placed in.

Ball-Grid-Array Sputter Protection

The purpose of this project was to deposit Copper, for electromagnetic interference shielding , on dies of various sizes without depositing on the backside of the chip. We used a pocket design in conjunction with high-vacuum adhesive to solve this. By making the pocket sizes slightly smaller than the chips and then spraying an adhesive on the surface around the pocket, we were able to protect the backside of the chip from metal deposition.

How I contributed:

Made and sprayed adhesive formula developed by Daetek and documented process.

Ran the pocket-etched substrates and analyzed results by checking for backspill.

Communicated observations and suggested improvements and theories on the process to improve results.

Observing back-spill on a Ball-Grid-Array chip.

Prototype steel etched carrier wafer with adhesive spray.

Multilayered Gas Sensor of Doped SnO2

In order to detect various gases (such as SOx, NOx, CO, CH3 etc.) in the atmosphere one can use a Tin Oxide resin with various dopants like nickel, chromium, antimony etc. After the fabrication of the alumina substrate with the interdigitated gold electrodes the goal of this research was to deposit multi-walled carbon nanotubes (MWCNTs) onto the gold electrodes and then apply Tin Oxide nanoparticles with the different dopants and see if there was an increase in the response speed or magnitude of the sensor. Unfortunately, this goal remained unaccomplished due to the MWCNTs inability to respond to the electrophoretic deposition process. We also made the resins with the dopants but never applied them to the substrate/electrodes to test for any results. Overall we found issues in the electrophoresis deposition of the MWCNTs on the substrate as well as issues with preparing the Tin Oxide resin using Tin citrate.

Materials and Methods

Firstly, an alumina substrate with gold interdigitated electrodes was fabricated using photolithography and electron beam deposition. This can be seen in figure 1 as the surface upon which the carbon nanotubes and doped tin oxide will be placed.

In order to get the carbon nanotubes to stick to the gold which is positively charged, carboxyl groups need to be placed on the CNTs to charge or functionalize them. This was done by mixing nitric acid with the carbon nanotubes and heating up the mixture for several hours. The nanotubes were then centrifuged and diluted with water and then KOH until the solution’s pH became ~7.

After the carbon nanotubes were functionalized, we attempted to use electrophoresis deposition in order to attach the carbon nanotubes to the interdigitated gold electrodes. However, multiple attempts at this resulted in no significant deposition of carbon nanotubes on the electrodes. This may be due to the fact that the functionalization method is usually used for single-walled carbon nanotubes. Unfortunately, multi-walled carbon nanotubes are more variable depending on the conditions in which they are grown and therefore may not have responded to the functionalization or electrophoresis as would be expected. After a few attempts at this, a plan was made to spray the nanotubes on using a simple spray bottle technique. But we then moved on to the synthesis of Tin Oxide using the Pechini method.

This part of the experiment proved tricky purely on the instructions we had to work with. I learned to value clear and well-made instructions more after this. In this process we first synthesized tin citrate from tin chloride and citric acid. This was then turned into a powder by adding NaOH and then vacuum drying. We then mixed citric acid and ethylene glycol with water and heated the mixture. After this we added the tin citrate and nitric acid and heated the clear mixture to 140°C and let the NOx gas go into the fume hood. The mixture always ran off, it was quite hard to prevent this, we were more prepared the second time though. Then we did the high temperature long term bakes, the result was a slightly yellow powder as seen in figure 2.

After we eventually managed to synthesize the tin oxide powder, we moved on to doping the powder. This was done by adding 1% by weight of each dopant to each sample of tin oxide. Unfortunately we did not get beyond this point in our work. We also did not manage to sinter the resin on top of the alumina plate.

Analysis and Discussion

Despite the fact that we were not able to test the functionality of the gas sensor, Patricia and I managed to complete several steps involved in the creation of nanoparticles and functionalized nanostructures. I found it very interesting as well as challenging project due to the road blocks we found in CNT deposition and Tin Oxide synthesis. Time was also an issue towards the end of the quarter between projects being due and studying for finals.

In conclusion, the fabrication of an incomplete gas sensor has taught me a lot about communication, time management, and basic chemical processes. I am very grateful for the opportunity this project presented and hope that it will find its way to completion.