Towards All-Silicon Opto-coupling Systems (ASOs)

Contact:

Maarten Swanenberg (NXP Nijmegen): maarten.swanenberg@nxp.com
Ray Hueting (Power Electronics group): r.j.e.hueting@utwente.nl
Anne-Johan Annema (IC Design group): a.j.annema@utwente.nl

 

Objective:

For today’s move towards electrifying the world, we need even more research, development and implementation of break-through electronics systems. This particularly holds, for example, for electric cars, solar energy conversion and other types of upcoming smart power systems. In the vast majority of these systems, some parts operate at high voltages (up to several hundreds of volts) while any control and interface electronics operates at low voltages (a few volts). For safety and immunity reasons, the interface between the high power domain and the control/interface domain must be galvanically isolated while data-communication between those domains is crucial.

To make a next step in fully galvanically isolated data transfer in a single silicon die, research was done at the UT (sponsored by NXP) that yielded the first fully silicon integrated opto-coupler. Such an opto-coupler basically makes use of light for data-communication via a light emitting diode and photodiode which are mutually galvanically isolated. See also e.g. https://www.utwente.nl/en/news/2019/12/287774/ut-invention-in-laser-focus-worlds-top-20.

We defined assignments that focus to a next step to enable smart power systems. In this, we aim at high-voltage (several kV’s) galvanic isolation through opto-coupling in an advanced electronic system formed by multiple smart-power ICs. Boundary conditions include fabrication in state-of-the-art technologies used for actual power ICs, satisfying the most stringent performance requirements (automotive grade 0), pursuing break-through concepts. This assignment is part of the “All Silicon Opto-coupling” project which is a joint project by NXP and the UT.

We defined a number of assignments:

BSc assignments:

  1. To reach multiple kV of isolation, novel single-package multi-Si-die solutions are probably required. In this assignment, a literature survey is done on various reported waveguiding and packaging methodologies to allow in-package opto-coupling between multiple dies. The main focus will be on power applications, but other types of applications such as RF and digital form also part of this study. Theoretical aspects of isolation, creep, breakdown and other package and IC technology related issues are key in this. (BSc)

 

  1. In this “All Silicon Optocoupler” project, the focus is on data transfer via light. The aim is to obtain a breakthrough performance for which it is necessary to also benchmark against competing approaches. In this assignment, a literature survey on various coupling approaches is carried out. These alternatives include capacitive and inductive coupling methods. In the assignment, the galvanic isolation performance and used silicon area should be estimated as a function of performance and cost (in terms of power and area). The focus is on fully integrated (i.e. monolithic, system-on-chip, SoC) solutions, but also multi-die single-package (i.e. system in package, SiP) are part of this study. Depending on the available time, circuit simulations will be performed for either validating or even improving the alternative approaches. We want to benchmark against the best possible alternative, so we challenge you to find the best alternative and even to improve that one!  (BSc)

 

  1. For the predecessor of the “All Silicon Optocoupler” research project (see e.g. https://www.utwente.nl/en/news/2019/12/287774/ut-invention-in-laser-focus-worlds-top-20) we made building blocks in the previously available technology. A number of building blocks have not been measured and characterized yet. They can however provide valuable information on many feasibility aspects. In this BSc-assignment, a mainly experimental study is done into various existing on-chip (hence lateral) optocouplers in our previously available chip technology. This chip includes various novel LED designs such as a “superjunction” LED and a “horse shoe shaped” LED. The basic idea is to determine the opto-coupling efficiency for all opto-couplers, benchmark them and explain the differences theoretically. (BSc)

 

  1. For the OiC project that predates the current research project (see e.g. https://www.utwente.nl/en/news/2019/12/287774/ut-invention-in-laser-focus-worlds-top-20) we made a number of building blocks in a specific industrial power IC technology technology (ABCD9). Particularly an experimental study to investigate die-to-die opto-coupling and galvanic isolation in existing ABCD9 material has not been carried out yet. Preferably those dies should be stacked vertically for efficient optical coupling, but laterally would also be interesting for electrical isolation reasons. In this BSc assignment, experimental characterization of die-to-die opto-coupler is targeted, augmented with theoretical analyses. The realization of the die-to-die stacking will be performed by NXP. For lateral side-by-side coupling a type of transparent droplet may be investigated. (BSc)

 

  1. For any sensible system, performance over lifetime is essential. This is already true for consumer products such as handheld phones, that only need to perform for about 5 years before they are out of fashion. For industrial and automotive (smart) power applications lifetime demands are very high (15 years) while failure rates should stay below single digit ppm levels. In this BSc assignment, lifetime issues and degradation phenomena in silicon (Si) avalanche-mode and forward-mode light emitting diodes, abbreviated as AM-LEDs respectively FM-LEDs, are studied. This assignment contains a large chunk of experimental work to measure degradation: study the light emission spectrum, efficiency and electrical characteristics before and after stressing, and as a function of time/temperature. Then correlate this with theoretical phenomena to model/explain the behavior.

 

Detecting hydrogen from the sputtered ion signal in low-energy ion scattering

Low-energy ion scattering (LEIS) probes the outermost atomic layer of a surface by backscattering of noble gas ions. Under certain conditions, also depth profiles up to 5–10 nm depth can be obtained. LEIS spectra of hydrogen containing samples have a background signal at the low energy side of the spectrum due to hydrogen atoms that are sputtered away and ionised by the noble gas ions used for probing the sample. Time-of-Flight (TOF) measurements can separate the hydrogen ion signal from other sputtered elements and backscattered ions. A previous bachelor student project has shown that the thus obtained hydrogen signal is indeed proportional to the hydrogen content of the sample. In this assignment, the focus will be on a better understanding of the detection depth of hydrogen by LEIS, the possibilities for quantification and the detection of hydrogen in sputter depth profiles.


LEIS spectra of clean and hydrogen exposed Pd. The red and blue line show the signal of backscattered He+ ions for Pd cleaned by Ar ion sputtering and after exposure to H atoms, respectively. The green line shows the signal due to sputtered H ions.

Contact: Dr. Marko Sturm (j.m.sturm@utwente.nl), or contact prof.dr.Marcelo Ackermann.

Spectral Unraveling (SUN) of EUV light sources

A broadband transmission grating spectrometer approach

The XUV Optics group is undertaking an exciting project based on a newly developed optical component: a nanoscale patterned transmission grating that enables high resolution spectroscopy. The project goal is to do a comprehensive spectral characterization of EUV sources, through measurement of source spectrum in a broad wavelength range, from EUV to visible. Industry has shown interest and a recent grant was received for further optical and technological development of spectrometers. Currently a wealth of spectroscopic data is being gathered from several EUV sources including the most powerful industrial sources in the world.

The goal in this assignment (BSc or MSc) can be designing of new gratings or developing methods for retrieval of the emitted spectra from measurements. A particular challenge in the retrieval is the overlap of several diffraction orders of the grating and contribution of the noise. Candidate solutions include a first method to stitch spectra retrieved from 1st order measurements with several bandpass filters, and a second method to remove all higher order diffraction contributions from measured spectrum. In this project, you can gain experience in principles of spectrometers and spectroscopic analysis.

Contact: Dr. Muharrem Bayraktar (m.bayraktar@utwente.nl) , or contact prof.dr.Marcelo Ackermann.