Defect-Free Structures May Pave the Way to GaAs Lasers

Lynn M. Savage

If you are an optoelectrical engineer trying to create semiconductor devices that emit light, here is some advice: Don’t let zinc blende get into your wurtzite.

In recent years, gallium arsenide has gained traction as the basis for optoelectronic devices in lieu of silicon, which does not emit light well. Growing nanoscale crystals of GaAs, however, is problematic because most techniques allow the formation of zinc blende (cubic) and wurtzite (hexagonal) crystalline in the same structure. Defects at the boundaries between the two forms reduce the efficacy of the resulting GaAs nanoscale crystals.

Now Connie Chang-Hasnain and her group at the University of California, Berkeley, have developed a way to grow tiny needlelike GaAs structures that are defect free. The objects could be the first practical step toward GaAs lasers, but also could represent a new way to mass-produce tips for atomic force microscopes and to improve techniques for tip-enhanced Raman spectroscopy.

The researchers formed the so-called nanoneedles via metallorganic chemical vapor deposition, passing gallium and arsenic precursors over either silicon or GaAs substrates in the presence of a hydrogen carrier gas. They performed the technique at about 400 °C rather than at 600 °C, the temperature typically used to form III-V semiconductor layers. According to Chang-Hasnain, the combination of precursor and lower temperature caused the needles to grow spontaneously. Although typically gold is used as a catalyst to grow larger nanowires, no catalyst was required to make the needles.

Scanning electron microscopy images of a GaAs “nanoneedle” grown sans crystalline defects on a silicon substrate are shown, both at 30° from vertical (left) and from the top, down (center). A transmission electron microscope image shows the same type of GaAs needles grown on a GaAs substrate (right). Courtesy of Michael Moewe, University of California, Berkeley.

“The shape and size [of the needles] are just a result of the growth kinetics, given the parameters of our growth conditions,” said team member Michael Moewe. “The crystal quality is quite uniform, and most of the needles on the substrates have very similar, reproducible photoluminescence properties.”

The needles generally were up to 1 μm in diameter at their base and 3 to 4 μm long, tapering at 6° to 9° down to tips that were about 2 to 5 nm across.

Using a Coherent 532-nm diode-pumped solid-state laser focused onto a small spot on the nanoneedles, the investigators induced photoemissions with high — but unquantified — brightness.

According to Moewe, the group is concentrating on the possible lasing ability of the nanostructures, looking at passivation as a way to further improve photoluminescence and exploring ways to integrate the needles into lasing devices.

Moewe also noted that the needles would require no further processing to be used in atomic force microscopy or for tip-enhanced Raman spectroscopy, although the group would like to develop a way to grow the structures precisely on specific locations, such as on atomic force microscope cantilevers.

The team described its work May 6, 2008, at CLEO in San Jose, Calif., and is preparing its results for publication in an upcoming issue of Applied Physics Letters.