Avalanche-mode silicon LEDs, for monolithic optical coupling in CMOS technology

This thesis deals with the physics, design, and analysis of efficient light emitting diodes (LEDs) in silicon CMOS technology. ‘Within every modern appliances - for example in our smartphones - two separate worlds co-exist: the electric components and the optical functionality based on light based sensing,’ says Satadal Dutta. In car technology these worlds are also challengingly interconnected: high voltage components and low voltage communication.

Nowadays, many of such separate systems require data communication or signal transfer with galvanic isolation, providing fairly safe solutions. To make these systems smaller and integrated on the same substrate is a challenge, in which light can serve as a medium of communication. ‘The fundamental question underlying this thesis PhD work is to see how we can “do” light using silicon-based solutions, not needing far-off exotic materials, interfaces or devices,’ Satadal says.

Silicon LEDs conventionally emit infrared light (~1100 nm), which is not compatible with the spectral detection efficiency of silicon photo-detectors. This is because silicon can only efficiently detect light having wavelengths of less than ~1000 nm.

‘We worked on a special design solution for the silicon LED, which is promising in improving the detection efficiency of silicon photodetectors,’ Satadal says. ‘Our focus was on silicon LEDs, biased in "avalanche breakdown".’

In this situation, there exists a high electric field in the device which is responsible for light being emitted at shorter wavelengths (400 nm-900 nm). Such an emission - if properly guided laterally across a CMOS IC - can be detected by a silicon photodiode with a relatively high quantum efficiency.

In Satadal’s view, wide-spectrum Si LEDs are promising for integration of opto-electronics in CMOS. Therefore, in this thesis, the feasibility of realizing monolithic optical links is investigated as well, from a device physics viewpoint. Here silicon LEDs are used in silicon-on-insulators (SOI) using CMOS/DMOS (double diffusedmetal-oxide-semiconductor) technology.

Integration avalanche-mode LEDs

‘The optical coupling is treated as a conversion process from electrons to photons (in the LED) and back again to electrons (in the detector),’ Satadal explains.

Analysis is done from the viewpoint of coupling efficiency, where also the effect of heating across such a link, due to high power dissipation in the avalanche-mode LED, is taken into account. Optical propagation via built-in waveguides in SOI technologies is also studied using: finite difference time-domain simulations. The analysis of this link is aimed at integrating avalanche-mode LEDs, which have the potential to be driven at high speeds (~ GHz), with single-photon sensitive optical detectors.

Collaborations

Satadal stresses the importance of his cooperation with NXP Semiconductors in Nijmegen. ‘They supported us in fabricating the test devices we needed,’ he says. ‘Also the expertise and experimental support from Professor Lis Nanver (TU Delft) was decisive. She is a guest researcher at the Semiconductor Components Group in Twente. Simulations and modelling of the experimental materials was requisite for progressing in this PhD project. Here I got the fundamental insights needed, how light behaves in these key materials.’

Further, Satadal mentions the cooperation with the Integrated Circuit Design Group (University of Twente), where the project was led by Professor Anne-Johan Annema. ‘His personal connections with NXP proved very worthwhile,’ Satadal says. ‘Also the expertise on chip design and tape-out processes was very welcome to me, which was the focus of my project partner Vishal Agarwal. I succeeded in getting the devices integrated and fabricated using current wafer fabrication processes.’

On ‘sensitive detection’ using Single photon avalanche diodes (SPADs) Satadal profited from his collaboration with Professor Edoardo Charbon, from the Circuits and Systems Group at TU Delft. ‘The language used in these new forms of communication on chips, must be understood at the final detector side,’ he says. ‘Only then can we hope to handle light on chips in new ways in the future. The latest tape-out of our fully integrated optical link is still in progress.’

Publications

Satadal mentions the publication of his work in Journal of Applied Physics, was a milestone for his Group. Speaking at the CLEO conference on laser science to photonic applications, also was special, following his supervisor Professor Jurriaan Schmitz on this.

‘One of the important breakthroughs in my research was to show a direct connection between increasing the uniformity in the electric field in the LED, and consequently increasing the quantum efficiency of the avalanche-mode LED. This was shown using a rather well established concept of Superjunction RESURF, which is so far used only in power semiconductor devices for power electronics. We were the first to use it for improving intensity of optical emissions from silicon, a statement still quite surprising and unexpected for many people, both in academia and industry.’

Satadal mentions his communication and presentation skills improved massively. ‘Also my vision on research has increased considerably, thinking many more strategic steps ahead then I did four years ago. I fully enjoyed the collaboration with my colleagues. The open working atmosphere is a typically European - or perhaps even Dutch - characteristic, I believe.’

Future work

Regarding his upcoming work, the near future is still somewhat uncertain.

Satadal: ‘I’m looking for a post-doc position, and I am writing on a Veni research proposal as well. In the end I hope to find my way into academics. I like to feel free to go to the lab and do tests that I believe are important. Also the teaching part appeals to me. I really enjoyed giving tutorials for the Semiconductor Physics course. Helping the students to work neat, using the blackboard all the time, will help them to develop as all-round scientists. It is inspiring to see them develop, and to hear the feedback from them on my own work.'