Nanocrystals Enable Silicon Field-Effect LED
While several research teams around the globe attempt to make silicon emit light and others try to make transistors lase, one has combined the two technologies and made a light-emitting transistor based entirely on silicon.
Robert J. Walters and his colleagues from Harry A. Atwater's group at California Institute of Technology in Pasadena incorporated silicon nanocrystals into a silicon transistor and observed light emission at approximately 750 nm with a spectral width of ±80 nm. This relatively broad range is the result of a distribution in size of the nanocrystals and the indirect nature of the electronic transition.
In the silicon field-effect LED, a tunneling process sequentially charges the nanocrystals embedded in the gate oxide with electrons and then with holes. The electron-hole pairs radiatively recombine to yield light at approximately 750 nm. ©Nature Publishing Group.
The device, called a field-effect LED, uses a different mechanism to generate light than the other silicon-based light-emitting structures recently described. The researchers used the gate oxide of a metal-oxide semiconductor field-effect transistor as a host for the silicon nanocrystals, which are smaller than the approximately 5-nm nanocrystals used by other research groups. When an alternating current is applied to the gate of the metal-oxide semiconductor field-effect transistor, electrons and holes tunnel into it, thus charging the nanocrystals and resulting in field-effect electroluminescence.
"Unlike electroluminescence in bulk light-emitting diodes or previous nanocrystal structures, there is no DC current flow in field-effect electroluminescence," Walters explained. "The field-effect electroluminescence phenomenon, in which one carrier is stored prior to injection of the subsequent carrier, is a powerful scientific venue for studying time-resolved carrier injection and exciton emission in nanocrystals."
In previously reported silicon nanocrystal LEDs, electrical excitation was accomplished by applying a large DC voltage, and light emission occurred during the flow of a DC electrical current in devices that resemble conventional PN-diode LEDs. This process is known to degrade the insulating properties of the oxide matrix over time, leading to eventual device failure. The Caltech scientists have reported their test device to be stable over more than ~5 × 109 cycles so far.
Lorenzo Pavesi, whose research group at the University of Trento in Italy has been developing silicon nanocrystal-based lasers, is impressed. He said that this is an important piece of work because Walters and his colleagues "were able to solve the problem of how to achieve bipolar injection into silicon nanocrystals. Up to now, most of the silicon nanocrystal-based LEDs worked by ... unipolar -- one-particle -- injection. These authors provided a scheme to achieve bipolar -- electrons and holes -- injection, which is usually more effective and more stable than unipolar excitation. [And they have achieved] this by using a very simple device geometry."
Although Pavesi thinks the work is a further significant step toward the realization of an injection silicon laser, he believes that the problem remaining to be solved involves the integration of this bipolar injection scheme into a system of silicon nanocrystals where gain is reachable, therefore making a laser.
Walters said that his team is interested in making a silicon laser using the injection mechanism, but that the Caltech device is not a "magic bullet" that will enable the easy production of silicon lasers.
Even if an electrically pumped silicon laser could be produced using the design, many industry observers doubt whether a silicon laser could be competitive. One problem with silicon nanocrystals is their relatively long luminescence lifetime -- on the order of tens of microseconds -- compared with their direct-bandgap counterparts, which are some four orders of magnitude faster. This means that the brightness and modulation speeds of any nanocrystal silicon device are unlikely to compete with direct-bandgap structures.
"However, owing to the inherent simplicity of our design and the fact that the color of the light emission can be controlled simply by changing the size of the nanocrystals used, our devices could be particularly useful in display applications," Walters said. "While we feel our research is worthwhile aside from technological application as a scientific inquiry, we also believe that there is a commercial future for silicon optoelectronics and that our [field-effect LED] device may play a part in that future."
MORE FROM PHOTONICS MEDIA