Make Semiconductors Like 'Lego'... Swap Atoms to Create Insulating Layers and Electrodes All at Once [Reading Science]

IBS selectively transforms atomic layers in 0.7-nm increments, offering an alternative to complex material-stacking processes
Current-control performance nears theoretical limit... Findings published in leading journal Science

What if, instead of painstakingly stacking different materials to make a semiconductor, we could create the components we need simply by changing atoms? A Korean research team has succeeded in changing the properties of individual atomic layers to integrate a semiconductor, an insulating layer and an electrode contact layer into a single thin film. The team has created a next-generation transistor whose current-control performance approaches the theoretical limit while reducing the complexity of the manufacturing process.

The Ministry of Science and ICT and the Institute for Basic Science (IBS) announced on Oct. 16 that a research team led by Jo Moonho, director of the IBS Center for Van der Waals Quantum Materials and a distinguished professor at POSTECH, had developed a "modular transistor fabrication technology" that selectively substitutes atoms in a 2D semiconductor to create materials with different electrical properties. The findings were published that day in the international journal Science.

Schematic of modular 2D transistor fabrication using atomic-layer-selective substitution. It shows how atoms in a 2D semiconductor thin film are selectively replaced at desired layers and positions to form an insulating layer and an electrode contact layer, thereby integrating the transistor's main components into a single thin film. Provided by the research team.

As artificial intelligence (AI) becomes more widespread, technologies that increase semiconductor integration density and power efficiency are growing in importance. In particular, 2D semiconductors, which can control current even when only a few atomic layers thick, are attracting attention as next-generation materials that could overcome the limits of continued miniaturization in conventional silicon semiconductors.

The challenge lies in the manufacturing process. A transistor requires a semiconductor through which current flows, an insulating layer to control it, and electrodes that allow current to flow in and out. However, it is difficult to coat the surface of a 2D semiconductor uniformly with an insulating layer, and directly attaching metal electrodes can increase contact resistance. Another option is to transfer and stack different thin films, but the process is complex and difficult to apply to mass production.

The research team took a different approach. Instead of bringing in and stacking other materials, it devised a way to directly create the required materials by changing the atoms in the existing semiconductor.

The starting material was tungsten diselenide (WSe2), in which a layer of tungsten (W) atoms is sandwiched between layers of selenium (Se) atoms. The team left the tungsten atoms unchanged and selectively replaced the selenium atoms with oxygen (O), sulfur (S) or tellurium (Te). This enabled them to create tungsten oxide (WO?), an insulator; tungsten disulfide (WS2), an n-type semiconductor; and tungsten ditelluride (WTe2), a highly conductive semimetal.

Changing atomic layers in 0.7-nm increments brings current-control performance close to theoretical limit

The key is the ability to precisely control the depth and location of the atomic substitutions. The team took advantage of the tiny gaps between layers in the thin film, which block the movement of reactive atoms. By adjusting the reaction temperature and gas pressure, they transformed the material one atomic layer at a time, in increments of approximately 0.7 nm (one-billionth of a meter). They also successfully used a semiconductor fabrication mask to change only the desired locations.

Using this technology, the team fabricated a transistor with a semiconductor, an insulating layer and a semimetal electrode contact layer integrated into a single thin film. With the semimetal contact layer, the energy barrier that charge carriers must overcome to move from the metal electrode into the semiconductor fell to 35.5 millielectronvolts (meV). This is about one-third of the level when a metal electrode is attached directly.

Current switching performance also improved. The change in gate voltage required to vary the current by a factor of 10 was 61.3 millivolts (mV), approaching the theoretical minimum of about 60 mV achievable by conventional transistors at room temperature. The ratio between the current in the on and off states was also more than 10 million to one.

Key members of the research team. From left: Jo Hyunjae, student researcher at the IBS Center for Van der Waals Quantum Materials and a PhD candidate in the Department of Materials Science and Engineering at POSTECH; Center Director Jo Moonho; and postdoctoral researcher Min Seokyoung. Photo courtesy of IBS.

The equivalent oxide thickness (EOT) of the insulating layer was 0.3-0.4 nm, meeting the goal of 1 nm or less set by the International Roadmap for Devices and Systems (IRDS). This means that the thin insulating layer can effectively control current.

The team also fabricated a complementary metal-oxide-semiconductor (CMOS) inverter that operates at a low voltage of 1 volt (V) by vertically stacking p-type and n-type transistors. An inverter is a basic component of digital circuits that reverses an input electrical signal. The result demonstrates the potential to move beyond individual devices and extend the technology to stacked integrated circuits.

However, applying the technology to commercial semiconductor production will require uniform and reproducible atomic substitution across large, wafer-scale areas. The team plans to expand its work to a wider range of 2D semiconductor materials and conduct follow-up research on applications in 3D stacked integrated circuits.

"This study is significant because it opens a path to a qualitative leap in future semiconductor technology by applying a fundamental materials science technique - the manipulation of materials at the atomic level - to modern semiconductor technology," said Jo Moonho, director of the IBS Center for Van der Waals Quantum Materials. "The technology could be extended to ultralow-power and ultraminiature semiconductors, as well as next-generation stacked semiconductor devices."

This content was produced with the assistance of AI translation services.