Photonic-spintronic integrated circuit technology
Abstract
We propose a conceptually new photonic-spintronic memory technology, which we anticipate can have higher writing and reading speed and lower energy consumption than state-of-the-art electronic technologies. We build this idea on the recently discovered all-optical switching of magnetic layers by short optical pulses and the initial devices shown, a.o., by us. The next step is now to explore how we can envision such materials, devices and concepts in an actual memory, thus scalable to multiple bits and having a network to address all of these spintronic elements optically. This will be a technology paradigm shift, as for the first time we will integrated spintronics and photonics at the circuit level. To achieve this, in this work we will address three fundamental research challenges. The first challenge is to design novel spintronic devices that can be optically addressable at the lowers possible optical pulse energies. This can be achieved by maximizing optical light-matter interaction, for example using plasmonic or resonant light concentrators. To address for the mismatch in footprint of electronic and photonic elements, we will leverage our own unique magnetic racetrack technology, where a single optical concentrator will address a whole array of magnetic bits. The second challenge is to come up with an integration technology that supports such photonic and spintronic, and their interfaces, seamlessly, and at best-in-class level. We will combine the indium phosphide (InP) membrane technology, pioneered by us, with the spintronic elements, to create a unique platform, that can in principle support all required functionalities, such as lasers, photodetectors, switches and passives. Finally, the third challenge is to conceptually design architectures that would support such a memory. The unique features of our memory, i.e., optical pulses to write and read the bits, require a radical rethinking of existing (electronic) memory architectures. This technology platform will be foundational for many other applications too, including spintronic-photonic neuromorphic computing, and magneto-optic based sensors.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 22, 2024
- Source ID
- FA86552317066
Entities
People
- Martijn Heck
Organizations
- Air Force Office of Scientific Research
- United States Air Force