Stable, high-mobility transistors for next-gen displays


Wednesday, 22 December, 2021

Stable, high-mobility transistors for next-gen displays

The trade-off between carrier mobility and stability in amorphous oxide semiconductor-based thin-film transistors (TFTs) has been overcome by researchers at the Tokyo Institute of Technology (Tokyo Tech), in a breakthrough that could pave the way for display technologies that are cheaper than current silicon-based technologies.

Amorphous oxide semiconductors (AOSs) are a promising option for the next generation of display technologies due to their low costs and high electron (charge carrier) mobility. High mobility, in particular, is essential for high-speed images. But AOSs also have a drawback that is hampering their commercialisation: the mobility–stability trade-off, as indicated by the negative-bias temperature stress (NBTS) stability test.

Two AOS TFTs of interest are indium gallium zinc oxide (IGZO) and indium tin zinc oxide (ITZO). IGZO TFTs have high NBTS stability but poor mobility, while ITZO TFTs have the opposite characteristics. Tokyo Tech scientists have now reported a solution to this trade-off, as published in the journal Nature Electronics.

“In our study we focused on NBTS stability, which is conventionally explained using ‘charge trapping’,” explained Assistant Professor Junghwan Kim, who headed the study. “This describes the loss of accumulated charge into the underlying substrate. However, we doubted if this could explain the differences we see in IGZO and ITZO TFTs, so instead we focused on the possibility of a change in carrier density or Fermi level shift in the AOS itself.”

To investigate the NBTS stability, the team used a bottom-gate TFT with a bilayer active-channel structure comprising an NBTS-stable AOS (IGZO) layer and an NBTS-unstable AOS (ITZO) layer. They then characterised the TFT and compared the results with device simulations performed using the charge-trapping and the Fermi-level shift models. They found that the experimental data agreed with the Fermi-level shift model.

“Once we had this information, the next question was, ‘What is the major factor controlling mobility in AOSs?’” Prof Kim said.

The fabrication of AOS TFTs introduces impurities, including carbon monoxide (CO), into the TFT, especially in the ITZO case. The team found that charge transfer was occurring between the AOSs and the unintended impurities. In this case, the CO impurities were donating electrons into the active layer of the TFT, which caused the Fermi level shift and NBTS instability.

“The mechanism of this CO-based electron donation is dependent on the location of the conduction band minimum, which is why you see it in high-mobility TFTs such as ITZO but not in IGZO,” Prof Kim explained.

The researchers developed an ITZO TFT without CO impurities by treating the TFT at 400°C and found that it was NBTS stable, with an electron mobility as high as 70 cm2 (Vs)-1 — far above the 40 cm2 (Vs)-1 required by super-high vision technologies. However, CO impurities alone do not cause instability.

“Any impurity that induces a charge transfer with AOSs can cause gate-bias instability,” Prof Kim said. “To achieve high-mobility oxide TFTs, we need contributions from the industrial side to clarify all possible origins for impurities.”

The results could pave the way for fabrication of other similar AOS TFTs for use in display technologies, as well as chip input/output devices, image sensors and power systems. Moreover, given their low cost, they might even replace more expensive silicon-based technologies.

Image credit: ©stock.adobe.com/au/Сake78 (3D & photo)

Please follow us and share on Twitter and Facebook. You can also subscribe for FREE to our weekly newsletter and bimonthly magazine.

Related News

Researchers develop a fully 3D-printed, flexible OLED display

Researchers combined two different modes of printing to print the six device layers that resulted...

Samsung demonstrates MRAM-based in-memory computing

It has so far been difficult to use MRAM for in-memory computing, despite MRAM's merits such...

Improving energy storage with better supercapacitors

More supercapacitor development is needed to enable them to effectively store enough electricity.


  • All content Copyright © 2022 Westwick-Farrow Pty Ltd