RWTH: 2D materials for next-generation computing
In Nature Communications, RWTH professor Max Lemme and colleagues outline the most promising areas of application for two-dimensional (2D) materials.
“More Moore” and “More than Moore”: These are two of the most important research directions in the semiconductor industry. More Moore (More Moore) is an expression of the effort to extend “Moore’s Law”, that is, the continuous striving to shrink transistors and to integrate more, smaller and faster transistors on each chip of the next production node. More than Moore instead suggests the combination of digital and non-digital capabilities on the same chip, a trend also known as “CMOS+X” that has accelerated with the rise of 5G connectivity and applications like the Internet of Things and autonomous driving is becoming increasingly important.
For these two research directions, 2D materials are an extremely promising platform. For example, their ultimate thinness makes them prime candidates to replace silicon as the channel material for nanosheet transistors in future technology nodes, which would allow for continued dimensional scaling. In addition, devices based on 2D materials can in principle be easily integrated into standard CMOS technology and can therefore be used to extend the capabilities of silicon chips with additional functions, such as sensors, photonics or memristive devices for neuromorphic computing. The RWTH scientists Max C. Lemme and Christoph Stampfer with Deji Akinwande (University of Texas, Austin, USA) and Cedric Huyghebaert (IMEC,
Big potential
“2D materials have the potential to become the X-factor in future integrated electronics,” says Professor Max Lemme, head of the Chair of Electronic Components at RWTH Aachen University and spokesman for the Aachen Graphene & 2D Materials Center. “I expect they will first come to market in niche applications for specific sensors as the manufacturing technology requirements could be lower. But I am also convinced that 2D materials will play an important role in photonic integrated circuits and in future neuromorphic computing applications. We are still in the early stages here, but the first results are already very promising.”
In fact, more than a dozen 2D materials have already been discovered that exhibit programmable switching resistance – the fundamental property for building devices (memristors) – that can be used to mimic the behavior of synapses and neurons. While many fundamental aspects remain to be understood, the first memristors based on 2D materials have demonstrated competitive performance and a wide range of other desirable features, such as non-clonability and high-frequency switching for communication systems. In fact, such memristors are being studied in detail in the Cluster4Future project “NeuroSys”, which started in January 2022.
Another future field in which 2D materials can play an important role is quantum technology. “There is consistent evidence that 2D materials have great potential for quantum computing as well as for quantum communication and for novel quantum sensors,” says Professor Christoph Stampfer, head of the “2D Materials and Quantum Devices Group” at RWTH Aachen University and co-author of the comments. “Speaking of quantum computing, 2D materials are now eight to 12 years ahead of other platforms like silicon – spin qubits based on 2D materials, for example, are within reach but not yet demonstrated. However, the flexibility offered by the 2D platform could offer great advantages in the medium to long term and make it possible to overcome some of the obstacles encountered by other platforms.