While lasers integrated directly with silicon chips offer several advantages in photonics, their efficiency may be impacted by material mismatch and coupling loss. Researchers have now achieved the direct fabrication of quantum dot laser on silicon with broad and scalable photonics application. The integrated lasers were thermally stable and capable of efficient lasing at O-band frequency. This novel integration technique shows strong potential for broader adoption and commercialization.
Lasers that are fabricated directly onto silicon photonic chips offer several advantages over external laser sources, such as greater scalability. Furthermore, photonic chips with these “monolithically” integrated lasers can be commercially viable if they can be manufactured in standard semiconductor foundries.
III-V semiconductor lasers can be monolithically integrated with photonic chips by directly growing a crystalline layer of laser material, such as indium arsenide, on gallium arsenide, aluminum gallium arsenide, and gallium phosphide on a silicon substrate. However, photonic chips with such integrated laser source are challenging to manufacture due to mismatch between structures or properties of III-V semiconductor material and silicon. ‘Coupling loss’ or the loss of optical power during transfer from laser source to silicon waveguides in the photonic chip is yet another concern when manufacturing photonic chips with monolithically integrated lasers.
In a study that was recently published in the IEEE Journal of Lightwave Technology, Ms. Rosalyn Koscica from the University of California Santa Barbara, United States, and her team successfully integrated indium arsenide quantum dot (QD) lasers monolithically on silicon photonics chiplets. According to Ms. Koscica, “Photonic integrated circuit (PIC) applications call for on-chip light sources with a small device footprint to permit denser component integration.”
To achieve this monolithic integration, the authors combined three key concepts: the pocket laser strategy for monolithic integration, a two-step material growth scheme that includes both metalorganic chemical vapor deposition and molecular beam epitaxy for a smaller initial gap size, and a polymer gap-fill approach to minimize optical beam divergence in the gap, to develop monolithically integrated QD lasers on silicon photonics chiplets.
On testing, the chiplets with monolithically integrated lasers demonstrated sufficiently low coupling loss to allow optical feedback. As a result, the QD lasers operate efficiently on a single O-band wavelength within chiplets. The O-band wavelength is desirable because it is frequently used in data communications (for example, data center links). Single-mode lasing is achieved using ring resonators made from silicon or distributed Bragg reflectors made from silicon nitride.
“Our integrated QD lasers demonstrated a high temperature lasing up to 105 °C and a life span of 6.2 years while operating at a temperature of 35 °C,” says Ms. Koscica.
The laser integration technique has the potential to be adopted widely due to two reasons. Firstly, the photonics chips can be manufactured in standard semiconductor foundries. Secondly, the QD laser integration technique can work for a range of photonic integrated chip design without needing extensive or complex modifications.
The proposed integration technique can be applied to a variety of photonic integrated circuit designs by modifying the silicon photonics components, paving the way for a scalable, cost-effective monolithic integration of on-chip light sources for practical applications.
