Stanford
Caltech
I have been interested in photonics since the beginning of my undergrad years at Caltech, in the Caltech Nanofabrication Group under the direction of Axel Scherer. Optoelectronics holds many opportunities for improving on existing electronic devices and systems since light is made up of photons, which are bosons. This means photons don't mind being around each other, making them easier to herd and control. Ultimately this translates to faster raw signal processing from the increased bandwidth of optical devices. Optical phenomena such as diffraction are also capable of performing massively parallel computations (e.g. Fourier transforms) with little effort compared with electronic equivalents. Given all these advantages, it is unsurprising there is so much work currently devoted to the field of optoelectronics. I hope to see the reimplementation and expansion of the present electronic infrastructure with photonics, as well as a grand unification of optical, electrical, mechanical, and fluidic devices into truly complete miniature systems on a chip.
In electronics, if you need an electron source, you find a power supply or function generator to create a continuous stream of electrons or a signal. In the frequency domain, the ideal electronic source would be a sinusoidal oscillator or radiating antenna. The optical equivalent of this device is a laser, which (largely) emits coherent single-frequency optical radiation. As a source, lasers occupy the most important role in my conception of the field of photonics. (After all, if we can't generate signals in the first place, what's the point of manipulating them?)
The basic theory of lasers involves a gain medium to amplify light and confining light to a cavity to bounce around many times through the gain medium. In semiconductors, this is achieved by generating electron-hole pairs in a material with the proper bandgap. Confinement is achieved with quantum wells and the particular physical geometry of the device.
I have been working on InGaAlP/InGaP quantum well disk lasers [1]; tiny cylindrical disks on a post that look like mushrooms. These are probably the smallest lasers in the world, actually. They emit red light (670 nm) and actually have physical dimension smaller than the vacuum wavelength.
I am also interested in photonic crystal lasers, since they present more opportunities for miniaturization and integration into larger systems. Recently my mentor Zhaoyu Zhang succeeded in creating the first visible photonic crystal laser using the same material system as above. We are currently expanding on this work.
My first two summers at Caltech were spent exploring the use of surface plasmons in enhancing light emission from solid-state sources. Surface plasmons are a special type of wave that can exist at a metal-dielectric interface. Their useful property is that they can be used to store energy like a capacitor, except they work at optical frequencies. In practice, the deposition of tiny metal particles (like gold) on a dielectric surface (like glass) suffice to support surface plasmons (this is essentially stained glass). The electrons in the particles slosh back and forth forcefully if the frequency of the incident radiation matches the resonance frequency of the particle.
All this still doesn't sound as if you can get more light out of a device, since after all, no additional energy is being put into the system. The key here is that normally optical devices emit only a small fraction of light. So by enhancement of light emission, I mean increasing the percentage of generated light that gets radiated. In this way, we can imagine using surface plasmons as a channel to extract light; by first coupling energy from light in a material into surface plasmons, and then coupling surface plasmons into radiation. This process seems unnecessarily roundabout, but in practice it does work. We observed a three-fold enhancement of silicon quantum dot emission by using a solid gold layer nearby. The original plan entailed patterning the gold layer, but we ran out of time (there's only so much that can be done over summers).
The overall goal of the surface plasmon project was to enhance silicon light emission, and to perhaps attain the holy grail of silicon photonics: a silicon laser. Unfortunately silicon is such a poor light emitter than extensive light extraction techniques are required to observe any radiation output at all. This lofty goal will always have a place in my mind, and I hope to see a true silicon laser realized.
Publications
- Zhaoyu Zhang, Lan Yang, Victor Liu, Ting Hong, Kerry Vahala, and Axel Scherer, Visible submicron microdisk lasers, Applied Physics Letters 90, 111119 (2007).