This is a long overdue goodbye… We left Paris early on July 31st. It is bittersweet - I’m happy to be home with family on one hand, but I miss wandering the streets of Paris. My research was also a wonderful experience and I am grateful to have been part of the team at Institut d’Optique.
I haven’t posted much in the last couple of days due to the fact that the program is now winding down.
Friday was my last day at Institut d’Optique and it was a wonderful and relaxed day. Benoit drove Charles and I back to Paris; it was nice to take a break from the bus and RER. The freeway there has such an amazing view of Paris!
Saturday I wandered Paris with some friends. We went to some lovely stores such as Merci Merci and Anaim. We finished the day with an organ recital at Notre Dame.
Yesterday (Sunday), my friends and I travelled to a lovely place called Rose Bakery where I tried some fabulous and expensive eggs benedict. Their carrot cake was also wonderful. We went to Luxembourg Garden to soak up the sun for an hour or so and then parted ways. I finished up my presentation and paper.
Today (Monday), I had an interview with Steve at a cafe. He filmed part of the interview so we can make a video to attract students to the program. After, I met up with Chelsea, Michael, and my friend Adam (whom I did research with last year at Fresno) at the Opera Garnier and we took a guided tour. The opera house is so beautiful! Much smaller than I expected. I also had a chance to go up to the top of the Eiffel Tower. I finished up the day with a great steak at Cafe Constant. Pictures to come!
My last week here at the lab has been a very very busy and stressful week, but I wouldn’t have it any other way! Friday is my last day here at the lab. Next Tuesday we will be giving presentations on our research at one of Ecole Polytechnique’s many amphitheaters.
Bad News: The code I thought was working the last couple of weeks or so actually wasn’t. I tried to run a simulation back at my apartment but an error popped up. So I tried to fix that error…but then a few other errors popped up. It was no good. The first three days of this week I have been fixing and modifying the code, so…
Good News: Simulation is back up and running perfectly! Optimization code is looking brilliant too. See?
The red is the optimized graph whereas the blue is the experimental data for the 41-layer aperiodic mirror that Benoit created a couple of weeks ago.
Bad News: Unfortunately, I’m not going to finish all my research up by Friday. I’m hoping that I can wrap up this project once I return back to the states since I have all the tools I need. It wouldn’t be right for me to just say, “see you guys later” and drop everything I have been working on. I would really like to finish this up in a week or two.
As I mentioned in my previous research post, I am now working with aperiodic multilayer mirrors. My simulation code is working quite well, and so now it is time to modify my optimization code.
I have mentioned before that we use a magnetron sputtering deposition machine to create our multilayer mirrors. Sputtering is when you use an inert gas, such as argon, to “collide” with a a target. When this happens, the target ions are ejected from its source and then deposited on a substrate. You are probably scratching your head, so here is a picture (thanks to Wikipedia):
The magnetron deposition machine contains four different material sources. The substrate spins over the four different materials at a speed v (measured in angles per second). Here is a terrible picture I drew in Paint (until I can draw a better one with Xfig):
We can control v, as well as how many times our substrate passes over the material. If we then take the number of times the substrate passes over the material multiplied by a constant and divided by v, we obtain a desired thickness. Note that this constant depends on the material of the layer. Thus, in equation form:
Since we are trying to obtain a good fit to our mirror, we define a new equation for our fitted thickness:
We can then substitute our desired thickness into the equation:
We are ultimately trying to optimize the thicknesses to each of our layers in the mirror. Before, the reflectivity of the mirror depended on energy, materials of the layers, their thickness and roughness, incidence angles, and polarization. In the program, we will be leaving the number of scans and the speed constant, but we will be optimizing the constants in order to optimize the thickness. So the reflectivity depends on the same parameters as before, but instead of it depending on the thickness of the layers it will depend on these new constants: