My first exposure to nanofabrication was during my masters, working with carbon nanotubes and 2D materials like boron nitride and NbSe₂. What a pain it was to get bilayers or thin layers and then pattern them with electrodes! At the time, I had no idea what I was really getting into—no one explained the bigger picture of nanofabrication or cleanroom work to me. But despite how niche and tedious that work was, it didn’t push me away from the field. Fast forward to my PhD, I got exposed to a wide array of fabrication techniques. I was fortunate enough to work at DTU - Technical University of Denmark’s cleanroom—the largest in Denmark—and with each step, I gained more insight into the broader context of nanofabrication. Yet, even in such advanced facilities, the cleanroom processes differ significantly from those used in foundries, despite many similar techniques being employed. This brings me to superconducting qubit devices. Why aren't these made in large-scale foundries? Although foundry methods are similar, it’s not as straightforward as you’d think. A key method used in qubit fabrication today is still angled shadow deposition, which requires tilting the sample to create Josephson Junctions (JJs). Foundries, however, rely heavily on bottom-up or top-down processes, making this approach impractical. While it’s possible to create JJs using these methods, they often don’t achieve state-of-the-art coherence times—or perhaps I’ve missed some recent breakthroughs? Interestingly, despite all the advances in material science, I’ve yet to come across a comprehensive review on the various qubit fabrication techniques. Did I miss it? If you know of any, I’d love to hear about them! During my PhD, I combed through a lot of literature, but I’m still searching for that perfect review on fabrication techniques. Volunteers, anyone? 📸 Credits: Pishchimova et al. (2023)
Author, you gathered sizable processing expertise and instrumentation skills through this Ph.D. journey at highly ranked DTU, Denmark.
Técnico de Engenharia Sênior
2moHave you ever thought about encapsulating the qubit with ceramic?