Atomic-thin chip breakthrough promises faster, thinner, and more power-efficient devices

A team of researchers — led by Chunsen Liu at Fudan University in Shanghai — has created the first-ever fully functional 2D-CMOS chip that combines atomically thin 2D memory cells with a traditional silicon chip. This achievement — detailed in the journal Nature — bridges the gap between the conceptual potential of 2D materials and their real-world application.

For decades, scientists have worked hard to shrink the circuits on silicon chips, but the technology is now approaching its physical limits. 2D materials, which are just a single layer of atoms thick, promise a solution to the problem, but integrating them with conventional processors has been a major engineering hurdle, given their fragility and instability.

To overcome these challenges, the team came up with a new technology they call Atom2Chip. This technology involves several key innovations, including a full-stack on-chip process that allows the 2D material (monolayer molybdenum disulfide) to be integrated onto the rough surface of the CMOS chip, and a special packaging to protect the fragile atomic layer. The researchers also designed a new cross-platform system to ensure the novel 2D circuits could communicate seamlessly with the mature CMOS platform.

The result isn't just a simple lab prototype, but a full-featured 1-Kb 2D NOR flash memory chip capable of complex, instruction-driven operations. In tests, this chip managed a clock speed of 5 MHz and demonstrated fast programming and erasing speeds of 20 nanoseconds with low energy consumption.

This 2D flash chip serves as a blueprint for next-generation memory devices, promising higher density and high energy efficiency. While the demonstration focuses on data storage, the same approach may one day be applied to processors. Such advancements could lead to faster, thinner, and more power-efficient devices.