Despite the considerable promise of self-assembling quantum dots (QDs) as optoelectronic building blocks for nanoscale devices and systems, basic issues still must be addressed. These include quantitatively measuring the interaction between a QD and its optical stimulus, and learning how to control QD growth processes to arbitrarily design spatial distributions to suit specific needs.
A group of researchers from the National Institute of Standards and Technology (NIST; Boulder, CO) and the National Renewable Energy Laboratory (NREL; Golden, CO) has begun to address the first of these issues by directly measuring the absorption coefficient of indium gallium arsenide/gallium arsenide (InGaAs/GaAs) QDs. Another research team in the NIST Materials Reliability Division has found a correlation between QD growth patterns and the elastic-energy release rate (EERR) that is generally used to describe the spreading of cracks in solid materials.
A waveguide contains quantum dots (QDs). Each pyramid-shaped InGaAs/GaAs quantum dot (QD) shown in this micrograph of the etched waveguide is about 20 nm wide and 8 nm high.
Directly measuring the absorption coefficient of QDs involved coupling light into a waveguide containing QDs and resolving the output in time (see figure).1 "The QDs were grown in a semiconductor waveguide structure and were etched to form ridges in the waveguide," said Kevin Silverman of NIST. "Once we had that structure, we cleaved the facets of the waveguide, which provided the mirrors on either side of the waveguide with the reflection between gallium arsenide and air as end mirrors for the cavity."
Pulses from a synchronously pumped optical parametric oscillator were coupled into the waveguide; multiple round-trips within the cavity provided increased interaction and observation time, while also removing uncertainties in coupling efficiencies for both input and output signals. Light coupled out of the waveguide was time-resolved by mixing it with a gating pulse in a nonlinear beta barium borate crystal and scanning the delay of the gating pulse, resulting in a lower uncertainty on the value of the dipole moment.
Also unlike QD dipole measurements based on analysis of threshold currents, this method did not rely upon an assumed laser model with the Fermi level pinned above threshold. The results obtained—26 to 30 Debye for the dipole moment of single-layer InGaAs/GaAs QDs—were consistent with previously reported results obtained by threshold-current analysis.
Energy release rate
Bo Yang's research team in the NIST Materials Reliability Division performed a mathematical simulation to calculate how the growth of new indium arsenide QDs on the thin-film surface of a gallium arsenide substrate might be affected by the relative position of established neighboring QDs.2 To do so, they applied the concept of an EERR previously conceived to describe the formation of quantum islands (such as dots and wires) in which new islands nucleate and grow in the elastic field of an established neighboring quantum island.
"When you grow QDs, you have conservation of mass, but you also have a change of geometry, so the total elastic energy is going to change," Yang said. "And the change of the elastic energy for a unit volume of QD growth is the elastic-energy release rate."
Yang's group observed a significant preference for surface QD formation close to established subsurface QDs, as opposed to near established QDs on the same surface. An observed effect of edge placement on the new dot formation shows that the shape of the dots also has an effect.
The researchers will next attempt to manipulate the driving force for QD formation electrically—in piezoelectric semiconductor materials perhaps, rather than mechanically.
REFERENCES
- K. L. Silverman et al., Appl. Physics Lett. 82(25) 4552 (June 23, 2003).
- B. Yang, V. K. Tewary, Phys. Review B 68, 035301-1 (July 1, 2003).