Instability and Bistability in Oxides (c-ZnO and a-InGaZnO4)
연사: 남호현 박사 (서울대학교, 물리학과)
<?xml:namespace prefix = "o" ns = "urn:schemas-microsoft-com:office:office" />
Semiconducting oxides, i.e., ZnO  and a-InGaZnO4 , have attracted great attention as promising materials to replace amorphous Si (a-Si) in thin-film transistors (TFT), due to the high electron mobility. Specially, because of high uniformity in amorphous phase amorphous oxide semiconductors (AOS) TFTs have been adopted in next generation displays. However, the facing bottleneck of the oxide applications is to have the instability of electrical conductivity under electric field and light illumination [3-5]. For examples, the negative bias and illumination stress (NBIS) and positive bias stress (PBS) negatively and positively shifts the threshold voltage (Vth) of the AOS TFT, respectively, and the shifted Vths can be very slowly recovered (it takes a few days at room temperature). Unless the instability problem is resolved, the AOS application will not be successfully formed.
It is well-known that persistent photoconductivity (PPC) is strongly-correlated with the TFT instabilities. Generally, as an origin of the PPC of c-ZnO and a-InGaZnO4, oxygen vacancy has been believed, but the recovery energy barrier of oxygen vacancy is (just predicted) unresolved  or is calculated to be very small [7, 8]. Here, two new microscopic origins of PPC (that may be regarded as an intrinsic origin of the TFT instabilities) are suggested in c-ZnO  and a-InGaZnO4 . Firstly, in ZnO, substitutional hydrogen at oxygen site (HO) is known to be a robust source of n-type conductivity in ZnO. But, a puzzling aspect is that the doping limit by hydrogen is only about 1018 cm-3, even if solubility limit is much higher. Another puzzling aspect of ZnO is persistent photoconductivity, which prevents the wide applications of the ZnO-based thin film transistor. Up to now, there is no satisfactory theory about two puzzles. We report the bistability of HO in ZnO through first-principles calculations. We identify the bi-stability of the HO . As Fermi level is close to conduction bands, the HO can undergo a large lattice relaxation, leading to a deep level. The bistability of HO (shallow Ho+ and deep H-DX- ) can give explanations to two puzzling aspects. Secondly, in order to explain the NBIS instability of AOS TFT, we provide the bistability of disorder in valance bands as the microscopic origin of the NBIS (and also the illumination stress) instability of AOS . The peroxide (O22-) is mediated by holes generated by the stress, and it becomes a meta-stable donor defect giving the negative-shift of Vth and persisting for a long time after the stress. It is found that the peroxide defect is a donor, raising the Fermi level (decreasing the threshold voltage), and meta-stable (persists for a long time after the stress). The calculated activation energy from peroxide into normal state of 0.97 eV is very close to the experimentally obtained values of 0.9-1.0 eV . Conclusively, our identified bi-stable defects are essential to understood and/or overcome the instabilities of oxide-based TFTs.
일시: 2014년 8월 1일 (금요일) 3:30 PM ~ 5:30 PM
장소: 신소재공학관 702호 세미나실