Erasing-Modes Dependent Performance of A-Igzo TFT Memory with Atomic-Layer-Deposited Ni Nanocrystal Charge Storage Layer

Shi-Bing Qian,Yan Shao,Wen-Jun Liu,David Wei Zhang,Shi-Jin Ding
DOI: https://doi.org/10.1109/ted.2017.2702702
2017-01-01
Abstract:Amorphous indium–gallium–zinc–oxide (a-IGZO) thin-film transistor (TFT) memory devices are fabricated with plasma-assisted atomic-layer-deposition (ALD) charge storage medium of high-density Ni nanocrystals (NCs). The effect of Al2O3 tunneling layer thickness (i.e., 8 and 12 nm) on the memory characteristics is investigated. The results indicate that the 8-nm tunneling layer device shows a programming window as large as 4.7 V after 5-ms programming at 18 V, reflecting that the ALD Ni NCs have a high charge storage capacity. To achieve a high erasing efficiency, different erasing modes are explored, including electrical erasing, monochromatic light (ML) erasing, and ML-assisted electrical erasing. It is demonstrated that the ML-assisted electrical erasing can attain the highest erasing efficiency, e.g., an erasing window as large as 9.11 V is obtained after 100-s ML (300 nm)-assisted electrical erasing at −20 V for the 8-nm tunneling layer. Based on the above programming and erasing conditions, the ten-year memory window is extrapolated to be 4.7 V at room temperature. Though increasing the tunneling layer thickness may enhance the data retention, it also degrades the programming and erasing efficiencies slightly.
What problem does this paper attempt to address?