Rutile TiO2 is well known for its ability to “trap” photoinduced electrons at Ti4+ ions and form Ti3+ ions with an unpaired d1 electron. This has been shown experimentally to result in a large family of similar, yet slightly different, Ti3+-related centers that include both intrinsic small polarons and donor-bound small polarons. In these latter centers, the Ti3+ ion is located next to an oxygen vacancy or an impurity such as fluorine, lithium, or hydrogen. These small polarons are easily formed in commercially available bulk single crystals of rutile TiO2 by illuminating oxidized (and nominally undoped) samples at temperatures between 5 and 30 K with sub-band-gap laser light (e.g., 442 nm) or by slight reducing treatments (in the case of hydrogen). Once formed, the ground states of the defects are readily studied at low temperature with magnetic resonance (EPR and ENDOR). Single crystals of rutile TiO2 provide complete sets of angular dependence data, and thus allow detailed information about the ground-state models of the electron traps to be extracted in the form of g matrices and hyperfine matrices. In this review, the differences and similarities of the various Ti3+-related trapped electron centers are described.

We have achieved a drastic improvement of the performance as thin film transistor (TFT) for solution-processed IGZO thin film by controlling drying temperature of solvents containing the precursor solution. The IGZO-precursor solution was prepared by mixing of metal nitrates and two kinds of organic solvents, 2-methoxyethanol (2ME) and 2,2,2-trifluoroethanol (TFE). 2ME was used for dissolving metal nitrates. TFE was added as a solvent for reducing surface tension as small as possible, leading to improvement of the wettability of the precursor solution on the surface of the substrate. In order to discuss the relationship between morphology and drying process, the spin-coated IGZO-precursor films were dried at room temperature and 140 °C on a hotplate, respectively. Annealing of the both films was carried out at 300 °C in an electric oven for 60 min after each drying process. Drying at room temperature provides a discontinuous film, resulting in a large variation of the TFT performance. On the other hand, drying at 140 °C provides a continuous film, resulting in the higher TFT performance and a minor variation. The difference in surface morphologies would be derived from the evaporation rate of the organic solvents. The rapid evaporation at 140 °C brings about rapid pinning of the spin-coated precursor layer on the substrate. Preparation process via the drying at 140 °C gave ∼ 1 cm2 V-1 s-1 of the saturated mobility, quite small hysteresis, and 107∼ 108 of the on-off ratio.

Tin oxide aerogels were synthesized using the epoxide-assisted technique and characterized with X-ray diffraction, diffusive reflectance spectroscopy, particle-induced X-ray emission and scanning electron microscopy. Our results indicate that the material is electrically semi-insulating as the result of oxygen vacancies that appear as fixed charges at the bottom of the conduction band. A modification of the technique with the addition of hydrogen peroxide is proposed to reduce the levels of defects and enhance the optical transparency of the material.

ZnO thin films were synthesized by radio-frequency (RF) magnetron sputtering of high purity ZnO solid targets on sapphire substrates. Depositions were carried out at selected temperatures between 293 K and 1173 K, and post-deposition annealing was performed at 1173 K for 3 min. in an O2 atmosphere. Samples for electron microscopy investigations were prepared by lift-out technique in a multi-beam FIB/SEM instrument. The ZnO thin films show generally uniform thickness (about 1µm), determined by transmission electron microscopy (TEM) imaging. Irrespective of the deposition temperature, the ZnO thin films are polycrystalline, with individual grains exhibiting columnar morphology with the long axis oriented perpendicular to the ZnO/sapphire interface. The grain size varies with the deposition temperature, and a direct correlation between grain size and photoluminescence has been observed. Analyses performed using low-temperature photoluminescence spectroscopy measurements at 12 K revealed luminescence peaks at 3.361, 3.317, 3.218 and 3.115 eV. The intensity of the luminescence peak at 3.317 eV decreased with increasing deposition temperature. The films deposited at lower temperatures also exhibited a higher density of stacking faults as observed from the atomic resolution TEM. The crystallographic imperfections/photoluminescence relationship is not clear. The purpose of this study is to quantify the observed crystallographic imperfections and understand their effect on the photoluminescence of undoped ZnO thin films deposited on sapphire substrates.

Here, we show results on the deposition of ZnO:Al thin films by the successive ionic layer adsorption and reaction method. The growing of the films was performed by sequentially immersing glass and SiO2/Si substrates in water at temperatures close to the boiling point, a precursor reaction solution, water at room temperature and ultrasonic water bath. The resulting ZnO:Al films were transparent and well adhered to the substrates. From X-ray diffraction analysis was determined that the ZnO:Al films had hexagonal wurtzite structure with preferential orientation along the c-axis. Changes in the morphology of the films were obtained from ellipsoidal-shaped aggregates for the undoped ZnO films to spherical-shaped aggregates for the ZnO:Al films. The optical transparency and bandgap of the ZnO:Al films was about 85% and 3.28 eV, respectively. Thin film transistors were fabricated with ZnO:Al films as active layers. The characterized device had a saturation mobility of 0.048 cm2/V-s, threshold voltage of approximately 16.1 V and a drain current on-to-off ratio (Ion/Ioff) in the order of 103.

In this work, the annealing effects at 180°C in Aluminum-ZnO contacts as function of time were studied. Also, the application in TFTs of ZnO films obtained at low-temperature (200°C) are presented. The ZnO films obtained by ultrasonic Spray Pyrolysis at 200 °C were deposited over Aluminum contacts on SiO2/Si wafers to demonstrate the use of active layer in thin-film transistors. The results show that an improvement can be obtained in metal-ZnO interfaces by low-temperature annealing treatments. However, long annealing time degrade the metal-ZnO interface and may affect the electrical performance of the device.

We investigated the electrical properties of the rf-sputtered HfxZn1-xO/ZnO heterostructures. The thermal annealing on ZnO prior to the HfxZn1-xO deposition greatly influences the properties of the heterostructures. A highly conductive interface formed at the interface between HfxZn1-xO and ZnO thin films as the ZnO annealing temperature exceeded 500°C, leading to the apparent decrease of the electrical resistance. The resistance decreased with an increase of either thickness or Hf content of the HfxZn1-xO capping layer. The Hf0.05Zn0.95O/ZnO heterostructure with a 200-nm-thick 600°C-annealed ZnO exhibits a carrier mobility of 14.3 cm2V-1s-1 and a sheet carrier concentration of 1.93×1013 cm-2; the corresponding values for the bare ZnO thin film are 0.47 cm2V-1s-1 and 2.27×1012 cm-2, respectively. Rf-sputtered HfZnO/ZnO heterostructures can potentially be used to increase the carrier mobility of thin-film transistors in large-area electronics.

This work investigates the quality of back-channel passivation applied to sputter-deposited IGZO bottom-gate TFTs. Passivation materials investigated were alumina, silicon dioxide, and B-staged bisbenzocyclobutene-based (BCB) resins. Sputtered quartz and PECVD (TEOS) SiO2 rendered the IGZO material highly conductive (ρ < 0.01 Ω·cm), with subsequent annealing in oxidizing ambient unable to restore a high-resistivity state. Appropriate channel resistivity was restored on devices passivated with electron-beam evaporated alumina and spin-coated BCB when followed by annealing in air. Alumina passivated devices demonstrated improved stability; however slight distortions in measured I-V and C-V characteristics were observed. TCAD simulation was used to develop an IGZO material/device model, with results indicating the significant presence of oxygen-vacancy (OV) interface traps and negative fixed charge remaining at the back-channel.

Thin film Ti doped ZnO (Ti-ZnO) film were grown on sapphire (0001) substrate by RF and DC magnetron sputtering. Films were grown at a substrate temperature of 250 °C with different Ti/Zn concentration. Surface chemical study of the samples was performed by X-ray photoelectron spectroscopy to determine the stoichiometry and Ti/Zn ratio for all samples. Surface morphology of the samples were studied by atomic force microscopy. X-ray diffraction was carried out to determine the crystallinity of the film. No secondary phases of TixOy was observed. We observed a slight increase in the lattice constant with the increase in Ti concentration in ZnO. No ferromagnetic signal was observed for any of the samples. However, some samples showed super-paramagnetic phase.

Amorphous oxide semiconductors (AOS) are promising candidate materials for thin film transistors in display devices, but one major challenge for mass application is their instability under illumination. In this work, a theoretical method for analyzing transient photoconductivity response in such AOS thin films is reviewed, namely the continuous multi-exponential model. This model can deduce a continuous distribution of decay time constants representing activation energy levels in an AOS, and is shown to reliably reproduce a model of density of states (DOS) of mid-gap traps assumed to be responsible for the transient photoconductivity. Provided the data collection time is sufficiently long, the continuous multi-exponential model was verified to reconstruct the modeled continuous DOS spectrum, thus providing a powerful tool to analyze photoresponse in AOS. This method has the advantage that no prior assumptions about the form of the density of states are needed, but the drawback that long data collection times are required for the transient to be fully relaxed.