Growing crystals for lasers, rather than sculpting them, could be the key to a new generation of nanoscopic lasers, that will be embedded in communication and processing technology of the future.

Dr Wei Wen Wong said the new technique can streamline manufacturing of lasers a tenth of the diameter of a human hair, as well as open the door for further miniaturisation, without sacrificing efficiency.

“We have achieved a record high – 80 percent of the light emission was converted into laser light,” said Dr Wong, from the Department of Electronic Materials Engineering, and the ARC Centre of Excellence for Transformative Meta-Optical Systems (TMOS).

“We think this is because of the elegance in our fabrication approach, which allowed us to combine a high-quality cavity and active material into one.”

Dr Wong is the first author of the paper reporting the research in Science Advances.

The team were exploring making lasers from metasurfaces, which use sub-wavelength structures to manipulate light in novel ways not observed in everyday materials. Metasurfaces can be used to make devices such as lasers or sensors compact enough to be used in novel technologies – for example biochemical sensors that detect protein shapes without the need to alter their behaviour with the use of fluorescent biomarkers.

To date, nanoscale manufacturing has relied on sculpting structures through lithography and etching processes. These processes selectively carve into material surfaces via chemical reactions and physical bombardment, and are capable of the precision required to create nanostructures. These carving processes, however, have a major drawback: The resulting structures have rough surfaces. This is bad for optical uses as this roughness reduces lasing efficiency by causing detrimental light scattering.

This roughness is of a fixed size, and doesn’t scale down with the size of nano structures being manufactured. So, for smaller structures, surface roughness becomes a proportionately bigger issue, which limits how much smaller these kinds of devices can be made.

However, many researchers believed there was no alternative, arguing that that metasurface lasers could not be produced using bottom-up crystal growth methods, or at least would be very inefficient. 

Dr Wong was undeterred: Instead of carving into the surface, he tried techniques to selectively grow crystals with the desired sidewall facets, through a process called selective area epitaxy. This not only simplified the manufacturing process, but the team were able to grow nanostructures with atomically smooth sidewalls (surface roughness around 1.6 Angstrom).

The crystal design was based around bound-states in the continuum (BIC) to produce laser light, using structures that create resonances and trap light at a specific frequency. To do this the team created pairs of subwavelength nano sheets. These sheets each established a pair of magnetic dipole BICs whose symmetry prevents light leakage. By varying the size of one of the nano sheets in each pair they introduced an asymmetry that allowed light to leak out and produce laser radiation.

Optically pumping the laser they achieved a lasing efficiency of 0.8 – on-par with the current best etched room-temperature metasurface lasers (many of which need cryogenic temperatures).

The next step on the road to proving commercial viability of this crystal growth technique for these devices is to have the lasers powered electrically, to enable integration with semi-conductor electronics. However, attaching electrodes to the crystals is tricky, because it can create new areas where light can leak out.

Dr Wong hopes the work will benefit the wider industry, especially for applications that use LEDs, for example as pixels in augmented reality goggles.

“A lot of people had doubts, but we have proved that bottom-up nanolasers really work,” he said.

“Etching damage issue is much more critical for microLEDs, so people are paying serious attention to bottom-up techniques now.”