In the field of electronics engineering, one of the main goals is to make transistors and other electronic parts that are smaller and more efficient with materials and processes that are easy to find. Ferroelectric field effect transistors (FE-FETs), which are similar to conventional FETs but also contain ferroelectric materials, are among the transistors that have been found to be particularly promising.
FE-FETs contain entryway separators made of ferroelectric materials that can both switch and store electrical charge. These ferroelectric-based transistors could also be used as memory devices in addition to controlling the flow of current in electronic devices like conventional FETs.
Because it could enable devices to better support their operation without using too much power, this dual function could be extremely beneficial for computationally demanding applications like running AI models. Despite their potential, FE-FETs have not yet been used on a large scale because it has been difficult to reliably fabricate them using existing processes.
Scientists at the College of Pennsylvania, Penn State College, and different colleges overall have as of late acquainted themselves with a methodology for making FE-FETs utilizing comparative cycles to those currently utilized to deliver FETs. The publication of their paper in Nature Nanotechnology may open the door to the widespread use of these dual-function transistors.
A MoS2/AlScN FE-FET schematic Credit: Kim et al.
“The primary motivation behind our study was to demonstrate that both 2D semiconductor materials such as MoS2 and nitride ferroelectrics such as AlScN (i.e., aluminum scandium nitride) are very attractive for realizing compact, low-power, and fast non-volatile memory devices that can be directly integrated on Si CMOS technology in a back end of line (BEOL) process,” said Deep Jariwala, one of the researchers. For this purpose, we have been investigating AlScN and 2D materials for some time. This paper is a masterful demonstration of devices and materials scaled down to very small dimensions and operating voltages, as well as of materials scaled up to large areas.”
FE-FETs have the potential to serve as non-volatile memory devices due to their capacity to store and switch electrical charge almost indefinitely, even when a voltage is applied at their gate electron. The primary objective of the study conducted by Jariwala and his colleagues was to demonstrate that FE-FETs can be manufactured to meet the high-bandwidth memory requirements of big data applications by successfully integrating them with silicon semiconductor materials.
According to Jariwala’s explanation, “The charge stored by FE-FETs also modulates the conductivity of the 2D semiconductor into a high or low resistance state, which actually represents the information stored in the memory device.” The combination of AlScN ferroelectric material, which has exceptional ferroelectric properties and can be deposited in BEOL-compatible processes, and 2D semiconductors, which, due to their thinness and van der Waals structure, can allow for strong conductivity modulation and can also be integrated with relative ease, is the main advantage of our strategy.”
The researchers’ FE-FETs combine AlScN, a ferroelectric material, with channels made of a 2D semiconductor. Both materials were grown using conventional wafer-scalable processes. In a series of tests, the team tested a large number of their FE-FETs and found that they performed exceptionally well, with an ON-current density greater than 250 A um-1 and ON/OFF ratios greater than 107 at a channel length of less than 80 nm.
The exchange qualities of the FE-FETs are recorded at a 10 Hz rate with 0.2 V door voltage dissipation. Credit: Kim et al.
“Our exhibit demonstrates that 2D semiconductor/AlScN FE-FET gadgets are prepared for reconciliation with Si CMOS to bring large information figures requiring high transfer speed memory with processors to people in the future of PC equipment,” Jariwala said.
“Our work creates a bridge for making this leap from the lab to the foundry of these materials and memory devices, as both these classes of materials are becoming mature.”
The recent efforts of Jariwala and his coworkers may soon contribute to the widespread use of FE-FETs. The prototypes they have developed thus far can switch voltage between 3 and 4 volts, store data effectively, and integrate easily with some existing silicon CMOS processors. The researchers hope to further reduce their size in their subsequent research, which could make it easier for them to be integrated into consumer electronics.
“We will need to shrink these devices further to truly see their advantages and performance gains in big data computing,” Jariwala added. Ferroelectric AlScN can be produced and switched reliably at a thickness of 5 nm, as demonstrated in another recent paper of ours. We are currently working toward this goal. To truly realize devices and achieve operation voltages compatible with cutting-edge Si CMOS processors, our next step will be to integrate 2D materials and fabricate FE-FETs from AlScN films 5 nm thick. We also need to work on making p-type FE-FET devices and improving metal/2D semiconductor contact resistance values at the FE-FET device level.”
More information: Kwan-Ho Kim et al, Scalable CMOS back-end-of-line-compatible AlScN/two-dimensional channel ferroelectric field-effect transistors, Nature Nanotechnology (2023). DOI: 10.1038/s41565-023-01399-y
