ИСТИНА

Titania (TiO2) is a widely investigated semiconductor for photocatalytic decomposition of organic impurities in air and water and photoelectrochemical water splitting due to its chemical stability, photocorrosion resistance, low cost, and a large specific surface area with a high density of defects – own and adsorptive origin [1]. The type of defects and their prop-erties are strongly depend on way of sample preparation and therefore are determined by the morphology of nanostructures [2].Therefore, the aim of this work was to identify the defects in the TiO2 nanotubes (NTs) with different morphology and to study their peculiarities. Since the most of defects in the TiO2 has a nonzero spin, we used EPR spectroscopy. Multi-walled (MW) TiO2 NTs were formed by Ti foil anodization in a potentiostatic mode at 60V and temperature of 20°C for 60 minutes. Electrolyte solution consisted of 0.3 wt. % NH4F in ethylene glycol with the addition of 2 wt. % of deionized water. The duration of the anodization of pretreated Ti foil was 60 minutes. The obtained samples (NTs) were cleaned in ethyl alcohol and dried in an argon stream. Vacuum heat treatment of TiO2 NTs samples were carried out in a chamber of the vacuum-thermal evaporation unit at a tempera-ture of 450 0C for 60 min with the residual gas pressure was 3•10-5 Torr (NTs-1). Heat treat-ment in air of TiO2 NTs samples were carried out in a muffle furnace at a temperature of 450 0C for 60 min (NTs-2). Also we have made combined annealing: first in vacuum at 450 0C for 60 min, then in air at 300 0C for 60 min (NTs-3) and first in vacuum at 450 0C for 60 min, then in air at 450 0C for 60 min (NTs-4). The single-walled (SW) TiO2 nanotubes (NTs-5) were prepared using a technique similar to one described in [3]. Note that the multi-wall NTs means that ones have an outer layer of pure titania and an inner layer of carbon doped titania; single-wall NTs have only an outer layer. The EPR spectrum of NTs-1 and NTs-3 samples consists of a single line of high intensity, line width is approximately 4 G and the g-factor is 2.0027±0.0005. According to the literature data, such signals are due to carbon dangling bonds [4]. The presence of carbon in the samples under the investigation is due to use of a polyethylene glycol in the synthesis process and is confirmed by our data of ele-mental analysis. Notice, that in NTs-1 and NTs-3 samples we have detected a slightly asym-metric EPR signal. Such form of EPR signal can be due to superposition of different EPR line. Origin of several EPR lines with close g-factors can be explained by the presence of carbon in various forms, which is confirmed by our structural investigation and Raman data, which indicate the presence of carbon in the form of amorphous and crystalline inclusions. EPR spectrum of NTs-2 samples is a superposition of two signals originated from carbon dangling bond and Ti3+/oxygen vacancies centers. The concentration of spin centers (Ns) in the synthesized samples is in the range of 1015 ÷ 1017 spin/g, depending on the conditions of thermal treatment. The highest Ns value is observed in the vacuum sample NTs-1, which con-tains the largest amount of carbon. In the NTs-3 and NTs-4 samples the Ns value is signifi-cantly lower because of the amount of carbon is less. NTs-5 samples has only Ti3+/oxygen vacancies centers (Ns=1015 g-1). The energy position of the spin centers in the band gap of TiO2 NTs was determined using our original EPR method with illumination of the samples in situ [1]. It is established that the energy level of carbon dangling bond is 2.1 eV below con-duction band and the energy level of Ti3+/oxygen vacancy is 2.9 eV above valence band. An-alyzing the saturation curves (dependence of the EPR signal intensity vs the root of the mi-crowave power), we found that the inner layer of NTs has a mesoporous structure.