MESA+ University of Twente
Research Business & Innovation About MESA+ Education

Cazimir Bostan (promotion date: 25 February 2005)

Design and fabrication of quasi-2D photonic crystal components based on silicon-on-insulator (SOI) technology

Promotion Date: 25 February 2005

Cazmir Bostan

The basic idea was to make a periodic structure (a regular structure) in a high refractive index material, which is silicon. The kind of periodicity is related to the light wavelength you want to propagate. The next step was to create intentionally a fault, a defect in the structure and to guide light through the structure by means of the defect, thus creating an optical circuit. The sample we made comprises three layers: the top layer of silicon, below the insulator (silicon oxide) and the silicon substrate. The role of the insulator is not, like in electronics, to insulate the current, but to create a buffer layer between substrate and the top layer to allow the propagation of light in the top layer only. We use silicon for top layer because of its high refractive index. The main challenge was to set-up the technology to create the structure.

What was your thesis about?

The basic idea was to make a periodic structure (a regular structure) in a high refractive index material, which is silicon. The kind of periodicity is related to the light wavelength you want to propagate. The next step was to create intentionally a fault, a defect in the structure and to guide light through the structure by means of the defect, thus creating an optical circuit. The sample we made comprises three layers: the top layer of silicon, below the insulator (silicon oxide) and the silicon substrate. The role of the insulator is not, like in electronics, to insulate the current, but to create a buffer layer between substrate and the top layer to allow the propagation of light in the top layer only. We use silicon for top layer because of its high refractive index. The main challenge was to set-up the technology to create the structure. It sounds easy, but the periodicity you want to create is very small, in the order of half of a micron. For this kind of size the conventional lithography does not work. We looked into the structuring technology using laser interference lithography, and after about 1.5 years of research we found that, although very good from the periodicity point of view, the shape of the holes was not very satisfactory. We aimed at circular holes and with laser interference lithography we ended up with slightly elliptical holes. So a better solution had to be found, and we used a Focused Ion Beam (FIB), which worked perfectly with nanometer accuracy, although is a serial technique, making large arrays very difficult to achieve. Still, for making photonic crystals precisely aligned with respect to coupling structures, FIB gives excellent results.

What is the purpose of all this?

In integrated optics so far light has been guided through total internal reflection, but in order to bend the light path you need quite a big surface, - this is of course relatively speaking. Our main goal was to realize optical circuits, which are very compact. With this new technique we can make very sharp bends and the light stays in. But right now the most spectacular applications that have already been demonstrated are in lasers and in photonic crystal fibres. The periodic structure gives a much better confinement of the light in the core. But of course there is still work to be done in the understanding of the physics and improving the technology. The etching of the structure for instance needs to be very precise, because jagged edges will scatter the light strongly, leading to large losses.

How far are we off from integrated circuits using photonics?

Not very far I would say, but this depends on the application. For example, replacing the electronic circuits in computers with photonic circuits is still a dream. The industrial application in telecom will take a while I suppose. I expect applications sooner in sensing technologies. The fabrication technique we have demonstrated works very well. Now we can make a photonic crystal sample in a couple of days in the Mesa+-lab. But the fabrication of planar photonic crystals is still expensive. And not all the physics of waveguiding in photonic crystals is completely understood yet.

Why did you do your PHD at Mesa?

I was working in Romania in the National Institute for Microtechnologies and then one day one of my colleagues brought the possibility for this PhD to my notice. I applied and was accepted.

What are you going to do next?

I am going to return to Romania, I have my promotion on the 25th of February and on the 28th I am going back. I still have my job there. I was on temporary leave for four years to do my PhD.

Did you enjoy your stay here?

Yes, I did, especially the international environment. I made friends here from all over the world, from Sudan to Indonesia and of course in the Netherlands as well. The level of independence was a bit difficult though: I was the only one working on this subject in the group. I sometimes missed the lack of interaction with my colleagues.

And on a personal level?

I had a good time, my wife stayed with me for three years and only the last year returned to Romania because she got a job. She will be here for the promotion of course.

For the summary of the thesis, click here.