The hardness, crystalline structure and wide bandgap of gallium nitride (GaN) make it ideal for a variety of applications, including light-emitting diodes (LEDs), laser diodes that read blu-ray discs, transistors that operate at high temperatures, solar cell arrays for satellites, biochemical sensors and, because of GaN’s relative biocompatibility, electronic implants in humans. But Yujie Ding, professor of electrical and computer engineering at Lehigh University, sees another, potentially more revolutionary role for GaN.
The compound can be engineered so that light passing through GaN actually cools it instead of heating it, he says. This phenomenon of laser cooling (or laser refrigeration) would eliminate the need for costly heat-dispersion methods that are employed to prevent electronic devices from overheating.
“GaN can be used to make lasers, optoelectronic and electronic devices,” notes Ding, who is a fellow of both the Institute of Electrical and Electronics Engineers and the Optical Society of America. “What if we could also use GaN for cooling? This would be one-stop shopping. We could monolithically integrate everything — the laser, the laser-cooling device and the electronic devices — on the same substrate.”
Overturning the Stokes ratios
Ding’s group has reached the threshold for achieving laser refrigeration by utilizing the anti-Stokes photoluminescence (APSL) phenomenon, which refers to the small fraction of photons whose frequency increases after striking a material. Stokes scattering occurs when the frequency of scattered photons is lower than the frequency of incident photons.
The ratio of the occurrence of Stokes to anti-Stokes scattering is typically 35:1, says Ding. Scientists would like to reduce this to 1:1, at which point a material neither heats nor cools when struck by light, and even further when, with more anti-Stokes than Stokes scattering, a material imparts its energy, and thus its heat, to the light passing through it.
Two years ago, Ding and his students - working with Jacob B. Khurgin, professor of electrical and computer engineering at Johns Hopkins - succeeded in reducing the ratio of Stokes to anti-Stokes to 2:1 in GaN, in numerical simulations and in lab experiments. The ratio was the most favorable achieved to that point. Recently, the group improved upon their results and recorded a ratio of 1:4. “We have not yet demonstrated cooling,” says Ding. “That will require further work. But we have demonstrated that we are above the threshold for laser cooling.”
The cooling potential of phonons
Laser cooling was first demonstrated 20 years ago on glass doped with a rare earth element. This method is ineffective, says Ding, because only the relatively small portion of the material that is doped contributes to cooling.
By contrast, GaN’s crystalline structure makes it possible for a much larger portion of the compound to play a role in cooling. Of critical importance are the phonons (collective vibrations at a uniform frequency) of the GaN molecules in the compound’s crystalline lattice.
“Because of the nonlinear properties of the lattices, phonons vibrating at very high frequency break down to lower-frequency vibrations,” says Ding. “At this lower acoustical vibrational frequency, the phonons become heat.”
To prevent the breakdown of phononic vibrations, Ding’s group combines the higher-frequency-vibration phonons with incoming photons. The high-frequency phononic vibrations are hence removed before they break down, and the vibrations, instead of generating heat, are emitted as high-frequency photons.
“The advantage of GaN is that the collective vibration of all the GaN molecules in the lattice makes it possible for the entire lattice to potentially contribute to cooling by promoting the upconversion of high-frequency phonons,” says Ding.
“We have learned how to use ASPL to convert input photons with low energy to outgoing photons with higher energy. To do this, we remove phonons by using resonance enhancement of outgoing photons’ energy with energy states of GaN. Thus we enhance ASPL,” he adds. “This is the best way to achieve laser cooling, because once the breakdown of high-frequency phonons occurs and produces heat, the process is not reversible. You have to work to remove heat, and this is never effective,” Ding continues. “It shows that laser refrigeration can be obtained with a III-V semiconductor, that is, with the very materials from which the optoelectronic devices that require cooling are themselves made.”
The project is funded by the US Defense Advanced Research Projects Agency (DARPA). Ding’s other collaborators include Guan Sun (who received his Ph.D. from Lehigh in 2013) and Ruolin Chen (a Ph.D. candidate). Sun now works for optical communications component maker JDS Uniphase Corp of San Jose, CA.
