Pranalytica’s novel design pushes up mid IR laser power

The team from Pranalytica, Harvard University, and the University of California, Los Angeles, has fabricated a 4.6 μm QCL with a non-resonant extraction design that delivered 3W at 293K. The laser, which incorporates a highly strained InGaAs/AlInAs active region, produced a peak wall-plug efficiency of 12.7 percent.

Funding from the Defense Advanced Research Projects Agency (DARPA) assisted the development of this laser, which could aid directed infrared countermeasures, such as deflecting heat-seeking missiles. In addition, this QCL could aid free space optical communications and provide a battery-powered source for infrared target illumination.

Corresponding author Kumar Patel, who is Pranalytica’s CEO, explains that the team’s design differs from the majority of today’s QCLs, which employ a two-phonon resonance approach. In this more common design, a radiative electron transition between upper and lower laser levels is followed by two consecutive, non-radiative transitions involving resonant interaction with longitudinal optical phonons.

Patel says that this two-resonance condition leads to very fast removal of electrons from the lower laser level, preventing electrons backscattering into it. Stopping this from happening is beneficial, because it prevents a decrease in population inversion that would degrade laser performance. However, the penalty paid by this design is its lack of flexibility, because layer thicknesses in the active region are dictated by the resonance condition and the desired laser transition energy.

Pranalytica overcomes this weakness with a non-resonant extraction approach that uses parallel non-resonant transitions to realize fast carrier extraction from the lower laser level. Non-resonant transitions are slower than the resonant one, but multiple paths mean that the total lifetime in the lower laser level is essentially the same.

Two major benefits result from the non-resonant design, according to Patel: it enables higher continuous-wave output powers and wall-plug efficiencies; and it allows greater freedom in QCL design, which can also lead to improved performance. The researchers have also been developing a low-reflectivity, Al2O3 coating for one of the facets, in order to increase wall plug efficiency. This adjustment, alongside the introduction of longer cavities, has led to the fabrication of a 1.1W QCL mounted on AlN that does not require cooling.

Micro-pixelated lamp combats current crowding

Nitek and Asif Khan’s group at the University of South Carolina have fabricated a 42 mW lamp emitting at 280 nm.

This ultraviolet lamp is a promising source for air and water purification and polymer curing. “42 mW is certainly enough to purify water at the tap at a rate of about a gallon per minute,” says Khan.

One of the challenges of making the chip was to overcome lateral current crowding, and this was addressed with a monolithic chip that contains 1600 micropixels, each with a diameter of 20 μm. Driven at an output power of 22 mW, this lamp has a lifetime of over 1500 hours.

Khan says that the team should be able to double the output power of the chip by optimizing the surface roughening of the chip.

Ammonothermal approach yields non-polar substrates

A Polish partnership has employed an ammonothermal growth method to produce non-polar GaN substrates with incredibly low threading dislocation densities (TDDs). Development of high-quality non-polar substrates is seen as an important goal in the nitride community, because it enables the fabrication of optoelectronic devices that are free from the large internal electric fields that hamper electron-hole recombination in conventional LEDs and lasers.

Led by the Warsaw firm Ammono, this Polish team has produced m-plane substrates up to 11 mm by 22 mm in size that have a TDD below 5 x 104 cm-2. These substrates have fewer defects and are slightly larger in size than typical pieces produced by the leading commercial supplier of m-plane GaN, Mitsubishi Chemical, which produces its material by HVPE.

Ammono’s substrate production begins by dissolving GaN-containing feedstock in ammonia in one zone of a high-pressure autoclave. A temperature gradient drives material to a second zone, leading to crystallization of GaN on native seeds, due to supersaturation of the solution. Measurements indicate that the quality of Ammono’s material is excellent: X-ray diffraction spectra have peaks with a full-width half maximum below 20 arc seconds; optical microscopy reveals that the TDD in the substrates and a 2μm thick GaN epilayer deposited on them is below 5 x 104 cm-2; optical excitation experiments produce strong emission at 3.4 eV, indicating good optical quality; and reflectance measurements reveal the strong hexagonal symmetry of the crystals.

This ammonothermal approach also has the edge over HVPE in terms of the growth process, according to Ammono’s Robert Dwili´nski: “ In HVPE, gases are flowing through the open reactor and only a small fraction of the raw materials is converted into the product. But with the ammonothermal method, thanks to recrystallization of polycrystalline GaN feedstock in a closed system, almost 100 percent of the raw material can be converted into the final product.”

Dwili´nski says that other strengths of the ammonothermal approach include lower growth temperatures that reduce energy consumption, easier reactor maintenance, and the scaleable nature of the process, which means that it is possible to grow hundreds of crystals in one run. Ammono aims to lead the world in the development of ever-larger high-quality, non-polar substrates. “We can easily keep this position because our growth method can be scaled up to any thickness,” says Dwili´nski.

Calculations confirm the benefit of semi-polar planes for green emitters

Calculations by a US researcher have revealed why semipolar planes of gallium nitride offer the best platform for growing green semiconductor lasers.

John Northrop from the Palo Alto Research Center performed first-principles calculations based on the chemical potential, which show that indium will incorporate in higher concentrations on the semi-polar (1122) surface than the non-polar (1010) surface. Northrop says that he considered the chemical potential - which is the free-energy per atom for an atomic species - because it allows comparisons of energy between surfaces that have differing numbers of atoms.

He reached his conclusions by employing “fairly complex” code developed at the Fritz Haber Institute, Berlin, and calculating the indium chemical potential of (1122) and (1010) layers with various degrees of indium incorporation. These calculations included a low hydrogen chemical potential, a condition that is typically found in an MOCVD growth chamber.

So far, the longest emission wavelength for a nitride laser has been realized on the (2021) plane.

Northrop is interested in these results that were produced by Sumitomo, but he has not performed calculations for that particular plane.