Journal of Applied Spectroscopy, Vol. 80, No. 4, September, 2013 (Russian Original Vol. 80, No. 4, July–August, 2013)
MORPHOLOGY AND STRUCTURE OF ZrO2/TiO2/SiO2 NANOCOMPOSITES OBTAINED BY SYNTHESIS ON CARBON MICROFIBER V. V. Zheleznov, Yu. V. Sushkov, E. I. Voit,* and V. G. Kuryavyi
UDC 539.2:(546.824-31+546.831.4-31)
We present novel nanomaterials based on Ti, Zr, and Si oxides, synthesized by the template method on a carbon fiber support. We have studied their structural and morphological properties. All the oxide fibers obtained have an extended surface formed from nanoparticle aggregates. We show that by selecting the components and varying their ratios, we can change the microorganization of the fiber surface and the structure of the composites. Due to microstress arising in the lattice, rearrangement of the structure of the composite is accompanied by a decrease in the fiber length and diameter. Increasing the amount of stabilizing additives of SiO2 in the titanium dioxide reduces the size of the nanoparticles and inhibits their growth, and makes it possible to stabilize the structure of the TiO2 nanoparticles in the anatase phase. At the same time, the presence of ZrO2 in the composition of the composites allows us to modify the particles at the reactive interface between the components and to obtain a fiber composite with a new crystal structure. For a low ZrO2 concentration (~10 wt.%), the structure of the oxide fibers is formed by TiO2 nanoparticles of the anatase modification, coated with an amorphous mixture of the oxides SiO2 and ZrO2. When the ZrO2 concentration is increased (~50.5 wt.%), we observe formation of nanoparticles with the structure of the orthorhombic phase of the compound ZrTiO4. Keywords: vibrational spectroscopy, scanning electron microscopy, titanium dioxide, nanostructured oxide fibers, nanocomposite, template synthesis method. Introduction. An important problem in modern materials science is controlling the properties of material by means of their structural organization at the nano level. However, today it is already insufficient to obtain material with small particle sizes and an extended surface; approaches are needed for creating organized nanostructures with the required functional properties [1–4]. Currently there is tremendous interest, from both a basic research and an applied standpoint, in nanocomposite materials whose properties depend on the morphology and interphase characteristics. Template sol-gel synthesis is one promising method for obtaining nanocomposites and materials characterized by high specific surface area and having pores of a certain size and shape. This paper focuses on using the sol-gel method to obtain composite oxide materials based on TiO2, ZrO2, and SiO2 on a carbon fiber template, followed by its annealing, and investigation of the structural and morphological properties of the fiber composites obtained. We have used scanning electron microscopy (SEM) and vibrational spectroscopy methods (IR, Raman). Each of the Ti, Zr, and Si oxides individually have valuable properties and have been extensively studied [2–5]. The surface characteristics of the pure oxides can be varied by forming new sections at the reactive interface between the components or by penetration of one oxide into the lattice of the other oxide. In the latter case, mixing of the oxides can lead to formation of new crystalline phases with properties that are different from those of the original oxides. Among twocomponent materials, the systems TiO2–SiO2, ZrO2–SiO2, TiO2–ZrO2 are most often studied [6–13]. Mixed oxides are widely used as thin-film coatings, selective sorbents, ion-exchange materials, and catalysts for many reactions. The methods for obtaining such systems are quite diverse, and the synthesis conditions are a determining factor for the acid-base properties of the surface, the structural features, and the possibility of their further use. Depending on _____________________ *
To whom correspondence should be addressed.
Institute of Chemistry, Far-Eastern Branch, Russian Academy of Sciences, 159, 100 let Vladivostoku Ave., Vladivostok, 690022; e-mail:
[email protected]. Translated from Zhurnal Prikladnoi Spektroskopii, Vol. 80, No. 4, pp. 596– 603, July–August, 2013. Original article submitted February 12, 2013. 0021-0937/13/8004-0581 ©2013 Springer Science+Business Media New York
581
TABLE 1. Oxide Content (wt.%) in the Samples and Annealing Temperature for Carbon Template Sample
TiO2
ZrO2
SiО2
T, °С
1
60.0
0.0
40.0
520
2
99.2
0.0
0.8
520
3
53.5
18.4
28.1
600
4
38.4
61.0
0.6
640
the method and the synthesis conditions, when varying the ratio of the components in binary materials we can obtain novel structures with predictable properties that are not characteristic of the pure oxides. Experimental Section. Objects of investigation. Nanostructured fibers made from Ti, Zr, and Si oxides were obtained by stepwise hydrolysis of a solution of Ti2(SO4)3 or a mixture of this solution with a solution of ZrO(NO3)2, by precipitation of the oxides on the surface of a carbon fiber template. As the template, we used Busofit-TM-4 carbon fibers manufactured by the Svetlogorsk Chemical Fiber Plant (Belarus). Since the original carbon fiber contains ~0.01% silicon, the presence of SiO2 is detected in the nanostructures. Preliminary treatment of the carbon fiber with NH4HF2 can reduce the silicon content in the sample by a factor of 30. The synthesized samples were subjected to gentle annealing at a temperature of ~500°C–640°C in an air atmosphere for 15–20 minutes (preventing active combustion of the carbon) to remove the template. The selected synthesis method allows us, by varying the composition of the components in solution when precipitating the oxides, to vary the ratio of the components over a broad range and to ensure that we obtain homogeneous and uniform nanoparticles on the template surface. By selecting the annealing conditions and temperature, the method makes it possible to obtain nanostructured materials with different morphologies and different crystal structures. In our work, we selected conditions for which nanostructured oxide fibers are obtained. We examined four samples with significantly different compositions and studied the mutual effect of the components on their structure and the morphology of the surface. Investigation methods. Table 1 gives the chemical composition of the Ti, Zr, Si samples (selected according to x-ray fluorescence analysis (XFA) data) with different oxide ratios; their annealing temperatures are also indicated. We studied the morphology of the synthesized nanocomposites by SEM on a Hitachi S5500 high-resolution microscope. We used a Shimadzu EDX 700 energy-dispersive spectrometer. We used vibrational spectroscopy methods to refine the structure, confirm the chemical composition of the samples, and determine the structural role of the components. The Raman spectra were recorded on a Bruker RFS-100/S Fourier transform Raman spectrometer with germanium detector. As the excitation source, we used an Nd:YAG laser with λ = 1064 nm. The IR spectra of the samples in the 150–4000 cm–1 region were recorded at room temperature using a Vertex-70 Fourier transform spectrometer with resolution 4 cm–1. We recorded the spectra of the samples as liquid paraffin mulls. Results and Discussion. SEM images of the samples obtained are shown in Fig. 1. In the structures of all the samples, we can observe microfibers with diameter varying within the range 1–3 μm. Depending on the elemental composition of the sample, the microfibers have different surface organization. The surface of the microfibers consists of nanoparticles and has a branched porous structure. Sample 2 has a small amount of SiO2. It is made up of microfibers consisting of aggregates of nanoparticles of irregular shape; the average diameter of the nanoparticles is 10–15 nm. Pure nanostructured TiO2, obtained by the sulfate sol-gel method at a temperature of 500–520°C, has a similar grain size [3]. We see that the surface of the samples with significant SiO2 content is different from the surface of samples with low SiO2 content. For samples 1 and 3, a characteristic feature is a less extended fiber surface, coated on top with an amorphous phase. This causes difficulties in estimating the size of the primary particles forming them. Also for sample 1 we can note that the surface of the fibers consists of layers of agglomerated nanoparticles of size ~100–200 nm, which probably is caused by the presence of microstress in the lattice due to the presence of dopant SiO2 [6]. Taking into account the SiO2 content in composite 1 and data from many studies of TiO2–SiO2 systems [6–12], we can assume that in the structure of the annealed sample 1, crystalline nanoparticles of TiO2 are formed that are stabilized by SiO2 interlayers. This is confirmed by the results of many studies of xTiO2–(1–x) SiO2 composites obtained by the sol-gel method, and the fact that TiO2 is poorly soluble in SiO2 [9–12]. Earlier in a study of nanocrystalline mesoporous ZrO2–TiO2 photocatalysts, it was shown that incorporation of a small amount of ZrO2 (up to 20%) in a TiO2 lattice, as for SiO2, increases its microstress and prevents formation of large crystallites. Furthermore,
582
Fig. 1. SEM images of the surface of microfibers corresponding to samples 1–4, with magnification 500 (a), 1000 (b), 10,000 (c), and 20,000 (d). incorporation of additives plays an important role in formation of nanoparticles with an anatase structure and a branched surface, having increased acidity and increased thermal stability [13]. Judging from the micrographs (Fig. 1), the size of the nanoparticles in sample 3 is appreciably reduced compared with sample 2. In sample 3, the nanoparticle diameter is ~5–10 nm. In sample 4, as in 2, the SiO2 content is low, the surface of its component fibers has a mesoporous structure, it is uniform and homogeneous and consists of well defined, irregularly shaped nanoparticles of diameter ~7–10 nm.
583
Fig. 2. IR absorption spectra of samples 1–4 (1–4), polycrystalline TiO2 (5), and amorphous SiO2 (6). Thus from SEM analysis it follows that oxide fibers synthesized by the template method have different surface organization depending on their components (Fig. 1). Samples with SiO2 additive have a nonuniform branched and layered surface consisting of agglomerates or crystallites of fused TiO2 particles, coated with an amorphous mixture. The modifying additive SiO2 breaks down the fibers and leads to their closer packing (Fig. 1). When the content of both SiO2 and ZrO2 increases, the fiber diameter and the size of the nanoparticles on the fiber surface decrease, which is probably connected with the microstress arising in the TiO2 crystal lattice. For a more detailed determination of the phase composition and nanoparticle size and in order to clarify the role of the modifying additives in the studied samples, we examined their vibrational spectra. We know that vibrational spectroscopy allows us to observe changes occurring in amorphous or nanocrystalline material consisting of particles too small for observation by conventional diffraction methods. Raman spectroscopy can be successfully used to determine the lattice symmetry of the crystalline system for oxide materials. At the same time, in this case IR spectroscopy is less informative but may be useful for revealing changes in short-range order in the structure of oxides [3, 6, 14–18]. General analysis of the IR spectra shows that after annealing of the composites, they were practically dehydrated. Broad bands in the 3700–3300 cm–1 region disappeared when the sample was annealed at ≈300°C. Only bands of insignificant intensity remain in the spectra with maxima at 3437 and 1630 cm–1, which correspond to vibrations of molecularly adsorbed water on the surface of the sample and possibly residues of silanol groups [8, 19]. Annealing of the carbon template was monitored from the presence of the G or D lines characteristic for carbon structures in the Raman spectra [20]. On the whole, the IR spectra show a correspondence to the established content of the samples (Table 1, Fig. 2). From the spectra of samples 1 and 3, we detect the presence of SiO2 in their composition. In the 1200–400 cm–1 region, absorption bands appear that correspond to stretching and bending vibrations of the tetrahedral SiO 44 − group [21]: a broad band with maximum at 1100 cm–1 and a shoulder at 1150 cm–1 (TO and LO modes of the antisymmetric stretch of Si–O–Si), a shoulder in the 800 cm–1 region (symmetric stretch of Si–O–Si), and a medium-intensity band at 470 cm–1 (antisymmetric transverse bend of Si–O–Si). In the spectra of samples 2 and 4, only insignificant traces of SiO2 vibrations are noted, which corresponds to its residual content (Table 1, Fig. 2). We note that for the indicated synthesis method and annealing temperatures (500–640°C), from the vibrational spectra it is impossible to
584
TABLE 2 Frequencies (ν, cm–1) and Half-Widths (Γ, cm–1) of Peaks in Raman Spectra of Polycrystalline TiO2 and Samples 1, 2, and 3 Sample
ν8
Г
ν7
Г
TiO2 anatase
143.5
7.0
196.7
4
1
148.3
17.0
198.9
8
2
146.0
12.0
198.2
3
148.4
24.6
199.0
ν7
Г
ν4
Г
ν 1, ν 3
Г
ν6
Г
–
396.4
18.0
515.3
21
639.6
23.4
293.0
15
398.5
29,0
516.0
27
639.7
34.6
8
293.2
16
397.2
22.0
516.4
25
639.1
31.7
–
293.0
–
398.4
34.2
516.1
29
639.7
42.0
completely rule out the presence of SO 24 − groups in the composition of the samples, since the bands in the IR spectrum that are characteristic for them overlap the SiO2 bands. However, we can say that they are present only in a small amount, at the defect level [15, 17], since x-ray fluorescence analysis did not detect the presence of sulfur either on the surface or within the interior of the sample. The broad band in the 900–400 cm–1 region in the IR spectrum allows us to conclude that Ti/Zr oxide fibers are formed [14–16, 22]. From the SEM and XFA and vibrational spectroscopy results, it follows that the oxide fibers obtained can be grouped according to composition as follows: those not containing ZrO2 (samples 1 and 2, TiO2/SiO2 composites); and those containing ZrO2 (samples 3 and 4, ZrO2/TiO2/SiO2 composites). TiO2/SiO2 composites. In nature, titanium oxide is encountered in three polymorphic modifications: orthorhombic 15 19 14 brookite ( D2h ) and tetragonal anatase ( D4h ) and rutile ( D4h ) [23]. We know that for TiO2, only the rutile modification is thermodynamically stable. It has also been experimentally established that as the particle size decreases, we observe an inverse change in the relative stability of the polymorphic modifications. The Raman spectrum of anatase has six first-order active modes (Table 2), while for rutile there are four such modes: B2g (826 cm–1, ν5), A1g (612 cm–1, ν1), Eg (447 cm–1, ν8), B1g (143 cm–1, ν4) [14]. For samples 1 and 2 at a temperature of 520°C, we obtained Raman spectra (Fig. 3) corresponding to an incipient crystal lattice. We see that the type of nanocrystallite lattice in samples 1 and 2 involves anatase particles. For sample 1, the intensity of the lines is reduced compared with 2, which may be connected with the presence of an amorphous SiO2 layer on the surface of the nanoparticles. Comparison of the Raman spectra of samples 1, 2 and a TiO2 polycrystalline material of the anatase modification shows that their frequencies and half-widths are different (Table 2). The difference may be due both to size effects in the sample and to incorporation of “guest” atoms in the TiO2 crystal lattice [24]. As we see, in the spectrum of sample 1 the half-widths of the characteristic peaks for the ν4 and ν6 vibrations are different, which is connected with the small contribution to their contour from vibrations ν1 and ν8 of TiO2 with the rutile structure. We have not assigned the low intensity peak in the 290 cm–1 region, but it was observed previously in the spectrum of TiO2 with an anatase structure [25]. For peaks ν7 and ν8, in addition to their shift toward shorter wavelengths, we see asymmetric broadening Table 2). Such changes and broadening of the main peak of the anatase phase is mainly related to the effect of a change in the size of the nanoparticles [24] or to effects of compression of the particle by the external amorphous layer [6]. Assuming that the observed spectra belong to three-dimensionally confined nanoparticles, from the position and half-width of the ν8 peak we calculated the average size of the nanoparticles in the samples using the phonon spatial confinement model. In the calculations, we used the MAPLE program. In order to calculate the dependence of the nanoparticle size on the line shift, we used the Bersani dispersion [26]; for the dependence of the nanoparticle size on the line half-width, we used the Ivanda dispersion [27]. Calculation of the nanoparticle size by both methods showed practically identical results within the error limits: ~7 nm for sample 1, 13 nm for sample 2. We know that for the IR spectra of single-crystal and nanosized TiO2 samples with the anatase structure, a broad intense absorption band is characteristic in the 900–400 cm–1 region, with the corresponding stretch of the Ti–O–Ti bonds [15–17]. It was shown earlier that the shape of this band depends on the phase composition of the sample (the purity) and on the crystallite size [14]. When the TiO2 particle size decreases, a small absorption maximum close to 440 cm–1 appears on the broad band, and this small maximum is more pronounced for fine particles. In the IR spectrum of sample 2, a maximum is formed at ~445 cm–1 on the broad anatase band, connected with a decrease in the particle size [14]. In the spectrum of sample 1, this peak is hidden under the broad band for the antisymmetric bend of Si–O–Si with maximum at 470 cm–1. The major
585
and most pronounced change in the spectrum of sample 1 involves the fact that the position of all the bands corresponding to SiO2 and TiO2 are preserved practically unchanged in shape compared with the reference standards. The spectrum of sample 1 looks like the superposition of the bands in the SiO2 and TiO2 spectra, but in the 940–950 cm–1 region a small shoulder appears, the intensity of which is reduced with annealing. This band belongs to the stretching mode of Si–O– or Si–O–Ti (the coordination number of the Ti atom is equal to 4) [6, 7, 9, 12]. This fact may indicate both phase separation during annealing of the composite and a small number of Si–O–Ti bonds formed only at the phase interfaces, as well as dehydration of the sample. As a result of the template synthesis carried out, oxide fibers of composition TiO2/SiO2 were formed (samples 1 and 2) consisting of agglomerates of nanosized TiO2 particles with anatase structure. An amorphous mixture of SiO2, probably with a small amount of TiO2, coats the nanosized particles with anatase structure. Consequently, modification of the TiO2 nanocrystals occurs mainly on the surface. The particle size decreases as the SiO2 content increases in the samples. ZrO2/TiO2/SiO2 composites. We know that zirconium dioxide at atmospheric pressure can crystallize in three 5 15 modifications: monoclinic (m-ZrO2, C2h ), tetragonal (t-ZrO2, D4h ), and cubic (c-ZrO2, Td1 ). At room temperature, the monoclinic modification is stable, and as the temperature rises it reversibly undergoes a transition first to the tetragonal modification and then to the cubic modification. It has been experimentally established that at room temperature, metastable t-ZrO2, c-ZrO2 modifications may exist, which are characteristic for nanocrystalline ZrO2. Methods have been developed and the conditions have been studied for stabilization of the metastable phases [17, 28, 29]. The t-ZrO2 modification is characterized by high specific surface area, and along with anatase is of interest as a catalyst and as a support for different types of catalysts. The Raman spectra of the different modifications of ZrO2 are well known [17, 18, 28, 29]; they differ markedly from the spectra of TiO2 with a rutile and anatase structure. It has been previously noted many times that mixing TiO2 with a sufficiently large amount (up to 50%) of ZrO2 does not have an effect on the change in structure of titanium dioxide, but rather x-ray diffraction does not detect any effect of the dopant ZrO2 and, according to the data in [13, 22], the lattice remains the anatase structure. The frequencies of the peaks in the Raman spectrum of sample 3 match the frequencies for sample 1 (Fig. 3). At the same time, the peaks in the spectrum of 3 are more symmetric, and are characterized by a larger half-width (Table 2), which may be due to the change in the environment of the TiO2 particles. Furthermore, based on the spectrum in the 100–900 cm–1 region, we observe scattering typical of amorphous ZrO2. The effect of the presence of dopant atoms in the lattice of the composites is also quite noticeable in the IR spectrum. A question arises concerning the structural role that might be played by the components SiO2 and ZrO2. In the spectrum, we see a broad band at 870–370 cm–1 corresponding to anatase, but a maximum was formed on it in the 480–450 cm–1 region which, despite the very noisy signal, is obviously split into two components: one at ~450 cm–1 (stretching vibration of Ti–O–Ti) and one at ~470–480 cm–1 (asymmetric bending of Si–O–Si or stretching vibration of Zr–O). Furthermore, in the IR spectrum, the SiO2 ν3 band becomes broad and is shifted toward lower frequencies. An appreciable shoulder appears on it close to 1040 cm–1, which suggests a large degree of disorder in the SiO2 matrix and indicates that the silicate lattice is perturbed by introduction of other atoms, since the greatest changes occur in the bands associated with a high contribution from the motion of O atoms [21]. Additional confirmation of the change in the silicate lattice may come from the maximum at ~616 cm–1 appearing on the anatase band which is characteristic for t-ZrO2 [17] but at the same time corresponds to the tetragonal (cristobalite-like) form of SiO2 [21]. Furthermore, in the spectrum of sample 3, a shoulder is noted close to 950 cm–1, which characterizes reaction at the phase contact interface. It is assigned to Si–O–Ti/Zr vibrations [30]. Thus based on the vibrational spectra, we can suggest formation in the structure of fibers in sample 3 of crystallites of anatase nanoparticles, located in an amorphous mixture of Si/ZrO2. This is consistent with the different crystallization kinetics of the pure oxides [22]. Comparative analysis of the Raman spectrum of sample 4 and literature data allowed us to assign the observed bands to modes of the incipient orthorhombic phase of the compound ZrTiO4(D2h), in which two different Zr4+ and Ti4+ cations are distributed in a disordered fashion over crystallographically equivalent positions. For the ZrTiO4(D2h) crystal, a characteristic feature is the presence of 18 active modes in the Raman spectrum [31]. In the observed spectrum, there are much fewer of them, but the broad band with maximum at 406 cm–1 allows us to assign the compound only to the lattice of the orthorhombic phase of ZrTiO4. The shape of the band suggests site-positional disorder due to a random distribution of the Zr4+ and Ti4+ cations between equivalent regions in the lattice. In [32], the salt ZrTiO4 was also obtained with Zr/Ti ratio of 1:1, and the presence of anatase was noted below 600°C while the appearance of ZrTiO4 was detected with a rise in temperature. For sample 4 at a temperature up to 600°C, we observed the Raman spectrum of the completely amorphous phase, while when the temperature was raised up to 640°C, we obtained a well-resolved spectrum (Fig. 3). In the IR spectrum of sample 4, there are three broad bands with maxima at ~760, 500, and 440 cm–1 (Fig. 2). We can conclude that in sample 4, fibers are formed from
586
Fig. 3. Raman spectra of samples 1–4.
homogeneous nanocrystalline particles of diameter ~5–7 nm, fused into grains (crystallites). The structure of the particles is close to the structure of ZrTiO4. The surface of the fiber is uniform and has an extended pore system (see Fig. 1). Conclusions. We propose a template method for synthesis of nanostructured oxide fibers containing TiO2, SiO2, and ZrO2. We have used vibrational spectroscopy and scanning electron microscopy to study their structural features. We show that by selecting the components and varying the ratio of the components, we can change the microorganization of the fiber surface and the structure of the composites. Increasing the amount of the stabilizing additive SiO2 in the titanium dioxide reduces the size of the nanoparticles and inhibits their growth, and lets us stabilize the structure of the TiO2 nanoparticles in the anatase phase. At the same time, the presence of ZrO2 in the composition of composites based on titanium dioxide allows us both to modify the particles at the reactive interface between components and to obtain a fiber composite with a new crystal structure.
REFERENCES 1. A. Kubacka, M. Fernández-García, and G. Colón, Chem. Rev., 122, No. 3, 1555–1614 (2012). 2. X. Chen and S. S. Mao, Chem. Rev., 107, No. 7, 2891–2959 (2007). 3. E. Ortiz-Islas, R. Gomez, T. Lopez, J. Navarrete, D. H. Aguilar, and P. Quintana, Chem. Rev., 107, No. 7, 807–812 (2007). 4. M. Bocanegra-Bernal and S. Díaz De La Torre, J. Mater. Sci., 37, No. 23, 4947–4971 (2002). 5. A. Colder, F. Huisken, E. Trave, G. Ledoux, O. Guillois, C. Reynaud, H. Hofmeister, and E. Pippel, Nanotechnology, 15, No. 3, 1–4 (2004). 6. A. Amlouk, L. El Mir, S. Kraiem, and S. Alaya, J. Phys. Chem. Sol., 67, No. 23, 1464–1468 (2006). 7. S. Permpoon, M. Houmard, D. Riassetto, L. Rapenne, G. Berthome, B. Baroux, J. C. Joud, and M. Langlet, Thin Solid Films, 516, No. 6, 957–966 (2008). 8. Ch. F. Song, M. K. Lu, P. Yang, D. Xu, and D. R. Yuan, Thin Solid Films, 413, 155–159 (2002). 9. Ch. M. Whang and S. S. Lim, Bull. Korean Chem. Soc., 21, No. 12, 1181–1186 (2000). 10. J. B. Miller, J. L. Mathers, and E. I. Ko, J. Mater. Chem., 5, No. 10, 1759–1760 (1995). 587
11. G. Mountjoy, D. M. Pickup, G. W. Wallidge, R. Anderson, J. M. Cole, R. J. Newport, and M. E. Smith, Chem. Mater., 11, No. 10, 1253–1258 (1999). 12. A. Hodroj, O. Chaix-Pluchery, M. Audier, U. Gottlieb, and Jean-Luc Deschanvres, J. Mater. Res., 23, No. 3, 755–759 (2008). 13. B. Neppolian, Q. Wang, H. Yamashita, and H. Choi, Appl. Catal. A: General, 333, No. 2, 264–271 (2007). 14. G. Busca, G. Ramis, J. M. Gallardo Amore, V. S. Escriba, and P. Piaggio, J. Chem. Soc. Faraday Trans., 90, No. 20, 3181–3190 (1994). 15. X. Bokhimi, A. Morales, E. Ortiz, T. Lopez, R. Gomez, and J. Navarrete, J. Sol-Gel Sci. Technol., 29, No. 1, 31–40 (2004). 16. R. J. Gonzalez and R. Zallen, Phys. Rev. B, 55, No. 11, 7414–7417 (1997). 17. X. Bokhimi, A. Morales, O. Novaro, M. Portilla, T. Lopez, F. Tzompantzi, and R. Gomez, J. Sol. State Chem., 135, No. 1, 28–35 (1998). 18. E. F. Lopez, V. S. Escribano, M. Panizza, M. M. Carnasciali, and G. Busca, J. Mater. Chem., 11, No. 7, 1891–1897 (2001). 19. S. Sakthivel, M. C. Hidalgo, D. W. Bahnemann, S. U. Geissen, V. Murugesan, and A. Vogelpohl, Appl. Catal. B: Environ., 63, Nos. 1–2, 31–40 (2006). 20. R. E. Shroder, R. J. Nemanich, and J. T. Glass, Phys. Rev. B, 41, No. 6, 3738–3745 (1990). 21. M. Ocana, V. Fornes, J. V. Garcia-Ramos, and C. J. Serna, Phys. Chem. Minerals, 14, No. 6, 527–532 (1987). 22. S. Chang and R. Doong, J. Phys. Chem. B, 110, No. 42, 20808–20814 (2006). 23. H. Remy, Treatise on Inorganic Chemistry [Russian translation], Mir, Moscow (1966), Vol. 2, p. 68. 24. W. F. Zhang, Y. L. he, M. S. Zhang, Z. Yin, and Q. Chen, J. Phys. D: Appl. Phys., 33, No. 8, 912–916 (2000). 25. G. M. Kuz'micheva, E. V. Savinkina, L. N. Obolenskaya, L. I. Belogorokhova, B. N. Mavrin, M. G. Chernobrovkin, and A. I. Belogorokhov, Kristallografiya, 55, No. 5, 919–924 (2010). 26. D. Bersani, P. P. Lottici, and X.-Z. Ding, Appl. Phys. Lett., 72, No. 1, 73–74 (1998). 27. M. Ivanda, S. Music, and M. Gotic, J. Mol. Struct., 480–481, Nos. 1–2, 641–644 (1999). 28. L. K. Noda, N. S. Gonçalves, S. M. de Borba, and J. A. Silveira, Vibrat. Spectrosc., 44, No. 1, 101–107 (2007). 29. M. E. Manriquez, M. Picquart, X. Bokhimi, T. Lopez, P. Quintana, and J. M. Coronado, J. Nanosci. Nanotechnol., 8, No. 12, 6623–6629 (2008). 30. M. Andrianainarivelo, R. Corriu, D. Leclercq, P. H. Mutin, and A. Vioux, J. Mater. Chem., 6, No. 10, 1665–1671 (1996). 31. R. Dwivedi, A. Verma, R. Prasad, and K. S. Bartwal, Opt. Mater., 35, No. 10, 33-37 (2012). 32. L. G. Karakchiev, T. M. Zima, and N. Z. Lyakhov, Neorg. Mater., 37, No. 4, 469–473 (2001).
588
