We studied the photoconductivity responses in amorphous In-Ga-Zn-O (a-IGZO) films using a time-resolved microwave photoconductivity decay (μ-PCD) technique. The a-IGZO film characteristics are correlated with three components in the photoconductivity response: the peak value and two decay constants. The peak value originated from the density of the photo-generated free carriers through carrier generation and recombination processes during laser pulse irradiation. Power law characteristics indicated that the peak values are attributed to recombination process related to the exponential distribution of the conduction band tail states. After the laser pulse was turned off, the reflectivity signal decreased rapidly, indicating fast recombination of the photo-generated carriers. This fast decay component is suggested to be related to the recombination processes through the deep level states. Following the fast decay, a slow decay with a decay constant on the order of microseconds appeared. This slow decay was attributed to the reemission of trapped carriers with an activation energy of ∼0.2 eV. In addition, both the fast and slow decays for the wet annealed a-IGZO film were longer than those of the as-deposited a-IGZO film. The decay constants are considered to reflect the density of the subgap states that act as trapping or recombination centers. The μ-PCD method provides a useful estimation of the film quality, such as the density of the defect states, and the physical properties of electronic devices using a-IGZO films.

No abstract

Photoinduced transient spectroscopy (PITS) was applied to study the effects of thermal annealing in the thin-film transistor (TFT) fabrication process on the variations of the electron traps in the channel region of amorphous In-Ga-Zn-O (a-IGZO). A dominant peak with a maximum of around 130 K was observed in the PITS spectra, but the detailed features were varied depending on the annealing conditions. The six particular temperatures corresponding to the trap states were extracted at about 100, 140, 150, 210, 320, and 390 K from the differential PITS spectra, showing good correlation with the trap states observed in ZnO. The results of thermal desorption spectrometry suggested that the variation of electron traps in the a-IGZO thin films has its origin in the decomposition of O and Zn during the annealing process. The annealing after the etch-stop layer deposition was also examined. The peak at about 150 K extracted from the differential PITS spectra before and after the annealing was markedly decreased. The activation energy of the corresponding trap states was estimated to be around 0.3 eV, which was close to those known as the E3 center in ZnO. Secondary ion mass spectroscopy analysis suggested that the reduction of trap density was mainly due to a decrease in the number of defects which involve hydrogen atoms in their configuration. Considering these results, the variations in the electron traps in the a-IGZO thin films during the TFT fabrication process should be attributed to the introduction of Zn, O, and/or H-related defects into tetrahedra consisting of Zn-O bonds.

The film quality of amorphous In–Ga–Zn–O (a-IGZO), an amorphous oxide semiconductor (AOS), was studied by the microwave photoconductivity decay (μ-PCD) method. Also, μ-PCD mappings over a 6 in. wafer were undertaken. It was found that the peak signal of the decay curve had a strong correlation with the a-IGZO transistor performance and hence the film quality. The film annealed under a wet condition showed the highest mobility and had the highest peak signal. The μ-PCD method was found to be a very useful tool to evaluate the film quality and predict the performance of AOS transistors fabricated under different process conditions.

We present the design, fabrication, and evaluation of a large total-reflection mirror for focusing x-ray free-electron laser beams to nanometer dimensions. We used an elliptical focusing mirror made of silicon that was 400 mm long and had a focal length of 550 mm. Electrolytic in-process dressing grinding was used for initial-step figuring and elastic emission machining was employed for final figuring and surface smoothing. A figure accuracy with a peak-to-valley height of 2 nm was achieved across the entire area. Characterization of the focused beam was performed at BL29XUL of SPring-8. The focused beam size was 75 nm at 15 keV, which is almost equal to the theoretical size.