Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://semiconductors.phys.msu.ru/publ/JAdvDielectrics_3_1350031.pdf Äàòà èçìåíåíèÿ: Sat Apr 5 13:55:50 2014 Äàòà èíäåêñèðîâàíèÿ: Sat Apr 9 23:20:27 2016 Êîäèðîâêà:

JOURNAL OF ADVANCED DIELECTRICS Vol. 3, No. 4 (2013) 1350031 (7 pages) © World Scientific Publishing Company DOI: 10.1142/S2010135X13500318

Structural position and oxidation state of nickel in SrTiO
Irina A. Sluchinskaya , Alexander I. Lebedev* and Alexei *Physics Department, Moscow State University
*,

3

Erko



Leninskie gory, 119991 Moscow, Russia Helmholtz-Zentrum Berlin Albert-Einstein-Str. 15, 12489 Berlin, Germany irinasluch@nm.ru
Received 19 July 2013; Revised 2 October 2013; Accepted 14 November 2013; Published 13 December 2013 The properties of Ni-doped strontium titanate are studied using X-ray diffraction and XAFS spectroscopy. It is shown that regardless of the preparation conditions, the SrTi1þx NixO3 solid solution and the NiTiO3 phase are the most stable phases which can coexist. According to the EXAFS data, in the single-phase sample of SrTi0:97 Ni0:03 O3, the Ni atoms substitute for the Ti ones and are oncenter. The distortion of the oxygen octahedra is not observed. The XANES spectra analysis shows that the oxidation state of nickel in NiTiO3 is 2×, and in the SrTi1þx NixO3 solid solution it is close to 4×. It is shown that the strongest light absorption in doped samples is associated with the presence of tetravalent nickel in the SrTi1þx NixO3 solid solution. This doping seems the most promising one for solar energy converters that exploit the bulk photovoltaic effect. Keywords : XAFS spectroscopy; impurities; strontium titanate; solar energy conversion.

J. Adv. Dielect. 2013.03. Downloaded from www.worldscientific.com by 95.143.213.243 on 02/16/14. For personal use only.

1. Introduction The bulk photovoltaic effect consists in the occurrence of a photocurrent and very large photovoltages when illuminating homogeneous crystals having no inversion symmetry.1 The idea of the practical application of this effect in ferroelectrics for solar energy conversion was discussed back in the 70s.2 However, because of the short lifetime of photo-excited carriers, the quantum yield of this effect is generally small, and the idea was decided to be unproductive. In recent years, the interest to the ferroelectric oxides with the perovskite structure has revived because new ideas on how to increase the efficiency of solar energy converters based on the bulk photovoltaic effect have been proposed.3,4 The main disadvantage of the ferroelectric oxides is their relatively large bandgap, causing them to absorb only a small fraction of the solar radiation. Recent theoretical studies have shown that the substitution of Ti atoms at the B sites of the perovskite structure with divalent impurities having the d 8 electron configuration (Ni, Pd, Pt), compensated by the oxygen vacancy, decreases the bandgap and the obtained perovskites are polar semiconductor oxides.5,6 An additional interest to the study of the Ni impurity is associated with the results obtained from the recent experimental and theoretical studies of new materials -- a recently synthesized PbNiO3 which has a very high calculated spontaneous polarization7­9 and BiNiO3 with unexpected oxidation states of nickel and bismuth atoms.10 In addition, the search for new magnetic off-center impurities in incipient ferroelectrics is still important because they can result in simultaneous emergence of the ferroelectricity and magnetic ordering and give rise to the magnetoelectric

interaction. Materials with these properties belong to multiferroics. SrTiO3 doped with Mn at the A site is an example of such a material, in which a new type of magnetoelectric interaction was recently discovered.11,12 Ni-doped samples could be an another example. Since the doping impurity can enter several different sites in the perovskite structure and stay in them in different oxidation states, the aim of this work was to study the structural position and the oxidation state of the Ni impurity in SrTiO3 prepared under different conditions using XAFS spectroscopy. We planned to check the possibility of preparing samples doped with divalent Ni at the B site, to evaluate the possibility of incorporating the Ni impurity into the A site, and to establish a correlation between the optical properties of the samples, on the one hand, and the structural position and the oxidation state of the Ni impurity, on the other hand. The choice of SrTiO3 was dictated by the fact that earlier we have studied the structural position and the oxidation state of a number of 3d elements in SrTiO3 (Mn,13,14 Co,15 Fe16), and their combined analysis can allow to find new promising impurities for solar energy converters.

2. Samples and Experimental Techniques Ni-doped samples of SrTiO3 with the impurity concentration of 2­3% and different deviation from stoichiometry were prepared by the solid-state reaction method. The starting materials were SrCO3, nanocrystalline TiO2, and Ni (CH3COO)2à4H2O. The components were weighed in proper proportions, mixed, ground in acetone, and calcined in air at 1100 C for 8 h. The calcined powders were ground once

1350031-1


I. A. Sluchinskaya, A. I. Lebedev & A. Erko

J. Adv. Dielect. 3, 1350031 (2013)

more and annealed again under the same conditions. Some of the samples were additionally calcined in air at 1500 C for 2 h. The composition of the samples was deliberately deviated from stoichiometry (excess titanium or excess strontium) in order to incorporate the impurity into the A or B site of the perovskite structure. The reference compounds NiO, NiTiO3, and BaNiO3þ used for determination of the oxidation state of Ni in SrTiO3 were prepared as follows. NiO was obtained by thermal decomposition of Ni(CH3COO)2à4H2O. Two other samples were prepared by the solid-state reaction method in air: the NiTiO3 sample was obtained from Ni(CH3COO)2à4H2O and TiO2 at 1100 C, the BaNiO3þ sample was prepared from BaO2 and NiO at 650 C. The phase composition of the samples was controlled by X-ray diffraction. Extended X-ray absorption fine structure (EXAFS) and X-ray absorption near-edge structure (XANES) spectra were obtained at KMC-2 station of the BESSY synchrotron radiation source (the beam energy 1.7 GeV; the beam current up to 290 mA) at the Ni K-edge (8340 eV) at 300 K. The radiation was monochromatized by a double-crystal Si1þx Gex(111) monochromator. Spectra were collected in fluorescence mode. The radiation intensity incident on the sample (I0 ) was measured by an ionization chamber; the fluorescence intensity (If ) was measured by a silicon energydispersive RæNTEC X-flash detector with 10 mm2 active area. Isolation of the oscillating EXAFS function Ïk÷ from the fluorescence excitation spectra ÏE ÷ ¼ If =I0 (where E is the X-ray photon energy) was performed in the traditional way.17,18 After subtracting the pre-edge background, splines were used to extract the monotonic atomic part of the spectrum 0 ÏE ÷ and then the dependence of ¼ Ï þ 0 ÷=0 was calculated as a function of the photoelectron wave vector 2 k ¼ Ï2mÏE þ E0 ÷=} ÷ 1=2 . The energy origin, E0 , was taken to be the position of the inflection point on the absorption edge. For each sample three spectra were recorded, they were then independently processed, and the resulting Ïk÷ curves finally averaged. Direct and inverse Fourier transforms with modified Hanning windows were used to extract the information about the first three shells from the obtained Ïk ÷ curves. The distances Rj and Debye­Waller factors 2 for jth shell (j ¼ 1­ j 3) as well as the energy origin correction dE0 were simultaneously varied to obtain the minimum root-mean-square deviation between the experimental and calculated k 2 Ïk ÷ curves. The coordination numbers were considered fixed for a given structural model. The number of adjustable parameters (8) was usually about a half of the number of independent data points Nind ¼ 2àk àR= % 16. The single- and multiple-scattering amplitudes and phase shifts, the central atom phase shift, and the photoelectron mean free path as a function of k, needed to calculate the theoretical curves Ïk÷, were computed using the FEFF6 software.19

EXAFS spectra were also analyzed with a widely used IFEFFIT software package.20 Isolation of the experimental EXAFS function was carried out using the ATHENA program and its fitting to the theoretical curve calculated for a given structural model was performed using the ARTEMIS program. In this approach, the amplitudes and phase shifts for all single- and multiple-scattering paths were also calculated using the FEFF6 software. The results obtained by two different data analysis approaches agreed well.

3. Results 3.1. X-ray data The diffraction patterns of all investigated samples are shown in Fig. 1. It is seen that the SrTi0:97 Ni0:03 O3 sample annealed at 1500 C is the only single-phase sample which has a cubic perovskite structure; for other samples, additional reflections were observed on the diffraction patterns. For the SrTi0:97 Ni0:03 O3 and Sr0:98 Ni0:02 TiO3 samples annealed at 1100 C, along with the reflections characteristic of the perovskite phase, additional reflections indicating a small amount of TiO2 and, presumably, NiTiO3 were observed. The identification of a possible NiO phase was complicated by the closeness of its reflections to the position of NiTiO3 reflections. In addition, stronger lines of the Ruddlesden­Popper phases Sr3Ti2O7 and Sr4Ti3O10 were observed in the SrTi0:97 Ni0:03 O3 sample annealed at 1100 C. In the Sr0:98 Ni0:02 TiO3 sample annealed at 1500 C the only additional phase, NiTiO3, was found. Since barium nickelate BaNiO3þ is a defective phase with the value depending on the preparation conditions, its lattice parameters dependence on the value21 was used to determine the actual oxygen content in our BaNiO3þ

J. Adv. Dielect. 2013.03. Downloaded from www.worldscientific.com by 95.143.213.243 on 02/16/14. For personal use only.

Fig. 1. (Color online) Diffraction patterns of samples: (1) NiTiO3, (2) SrTi0:97 Ni0:03 O3 sample annealed at 1500 C, (3) Sr0:98 Ni0:02 TiO3 sample annealed at 1100 C, (4) Sr0:98 Ni0:02 TiO3 sample annealed at 1500 C, and (5) SrTi0:97 Ni0:03 O3 sample annealed at 1100 C. Reflections of TiO2 and Ruddlesden­Popper (RP) phases are also indicated.

1350031-2


I. A. Sluchinskaya, A. I. Lebedev & A. Erko

J. Adv. Dielect. 3, 1350031 (2013)

J. Adv. Dielect. 2013.03. Downloaded from www.worldscientific.com by 95.143.213.243 on 02/16/14. For personal use only.

Fig. 2. (Color online) XANES spectra for SrTiO3(Ni) samples and nickel reference compounds: (1) SrTi0:97 Ni0:03 O3 sample annealed at 1500 C, (2) SrTi0:97 Ni0:03 O3 sample annealed at 1100 C, (3) Sr0:98 Ni0:02 TiO3 sample annealed at 1500 C, (4) Sr0:98 Ni0:02 TiO3 sample annealed at 1100 C, (5) BaNiO3þ , (6) NiTiO3 and (7) NiO.

Fig. 3. (Color online) EXAFS spectrum recorded at the Ni K-edge at 300 K for SrTi0:97 Ni0:03 O3 sample annealed at 1500 C (thin black line) and its best theoretical fit (thick green line).

sample. The hexagonal lattice parameters of a ¼ 5:568 ô 0:001 å, c ¼ 4:838 ô 0:001 å in our sample correspond to % 0:4. 3.2. XANES data analysis To determine the oxidation state of the Ni impurity in SrTiO3, the position of the absorption edge in XANES spectra of the samples was compared with the edge positions in the reference compounds. XANES spectra of all studied samples and three reference compounds are shown in Fig. 2. Comparison of the spectra of the Sr0:98 Ni0:02 TiO3 sample annealed at 1500 C with the spectra of cubic NiO and rhombohedral NiTiO3 (with the ilmenite structure) shows that the absorption edges in these samples are very close. So, we conclude that the Ni impurity in the sample under consideration is in the 2× oxidation state. The absorption edges of the SrTi0:97 Ni0:03 O3 and Sr0:98 Ni0:02 TiO3 samples annealed at 1100 C are close to the absorption edge of the BaNiO3þ reference compound. In the single-phase SrTi0:97 Ni0:03 O3 sample annealed at 1500 C, the absorption edge is shifted to even higher energies (by 2.5 eV as compared to NiO, by 2.9 eV as compared to NiTiO3, and by 1.3 eV as compared to BaNiO3þ ). 3.3. EXAFS data analysis To determine the structural position of the Ni impurity, EXAFS spectra were analyzed. A typical EXAFS spectrum k 2 Ïk ÷ for the single-phase SrTi0:97 Ni0:03 O3 sample annealed at 1500 C and its best theoretical fit (which takes into account the multiple-scattering effects) are shown in Fig. 3.

The best agreement between the calculated and experimental data was obtained in the model in which Ni atoms substitute for Ti atoms in SrTiO3. The interatomic distances and Debye­Waller factors for three nearest shells are given in Table 1. Small values of the Debye­Waller factor for the first and second shells, which are typical for the thermal vibrations in perovskites at 300 K, enable to draw two conclusions: (1) the off-centering of Ni atoms at the B sites can be excluded and (2) there is no distortion of the oxygen octahedra around the impurity atoms.a The analysis of the EXAFS spectra helped us to determine the composition of the second phase precipitating in the Sr0:98 Ni0:02 TiO3 sample annealed at 1500 C. Although the distances and the coordination numbers for the first shell of Ni are close in NiO and NiTiO3, the coordination numbers for the second shell in these compounds differ by three times. A comparison of the Fourier transforms of the EXAFS spectra for the sample under consideration and two reference compounds, NiTiO3 and NiO, shows a better agreement of its spectrum with that of NiTiO3 (Fig. 4). This means that among two possible phases, NiTiO3 and NiO, the second phase in our samples is the NiTiO3 one. As concerns to two-phase SrTi0:97 Ni0:03 O3 and Sr0:98 Ni0:02 TiO3 samples annealed at 1100 C, the comparison of their EXAFS spectra with the EXAFS spectrum of the single-phase solid solution and the EXAFS spectrum of NiTiO3 shows that the spectra of the samples under discussion may be represented as a superposition of the spectra of NiTiO3 and of the solid solution in a ratio close to 1:1. The optical properties of the samples are consistent with the data obtained above. The Sr0:98 Ni0:02 TiO3 sample

aThe sensitivity of the Debye­Waller factor to small distortions of the oxygen octahedra can be illustrated by the EXAFS data for NiTiO . In this compound, 3 2 there are two Ni­O distances in the first shell which differ by $ 0:07 å; such a small distortion increases the Debye­Waller factor to 0.0073 ô 0.0012 å .

1350031-3


I. A. Sluchinskaya, A. I. Lebedev & A. Erko
Table 1. Structural parameters obtained from the data analysis of the EXAFS spectra for SrTi0:97 Ni0:03 O3 sample annealed at 1500 C (Ri is the distance to the ith shell, 2 is i the Debye­Waller factor for this shell). Shell 1 2 3 Ri (å) 1.914 ô 0.004 3.342 ô 0.006 3.877 ô 0.004 2 (å2) i 0.0035 ô 0.0006 0.0084 ô 0.0007 0.0053 ô 0.0005 Atom O Sr Ti

J. Adv. Dielect. 3, 1350031 (2013)

Fig. 4. (Color online) Comparison of Fourier transforms of the EXAFS k 2 Ïk ÷ spectra obtained for (1) Sr0:98 Ni0:02 TiO3 sample annealed at 1500 C and for (2) NiO and (3) NiTiO3 reference compounds.

annealed at 1500 C had a light brown color; the SrTi0:97 Ni0:03 O3 and Sr0:98 Ni0:02 TiO3 samples annealed at 1100 C were dark brown; the SrTi0:97 Ni0:03 O3 sample annealed at 1500 C had an almost black color. Thus, the color of the samples reflects the relative amounts of black SrTi1þx NixO3 and yellow NiTiO3 phases in the samples.

4. Discussion Combined analysis of X-ray and EXAFS data shows that the SrTi0:97 Ni0:03 O3 sample annealed at 1500 C is the singlephase solid solution in which the Ni atoms substitute for the Ti atoms at the B sites and are on-center. This means that the solubility of nickel at the B sites in SrTiO3 exceeds 3% at 1500 C. In the sample with a nominal composition of SrTi0:97 Ni0:03 O3 annealed at 1100 C, the appearance of reflections of NiTiO3 and of the Ruddlesden­Popper phases indicates that a part of Ni atoms is spent on the NiTiO3 formation, while the other part remains at the B sites. As a result, some untapped Sr atoms form the SrO planes which embed into the perovskite structure and form the Sr3Ti2O7 and Sr4Ti3O10 phases. When trying to incorporate Ni atoms into the A sites of SrTiO3 at 1500 C (the samples with a nominal

composition of Sr0:98 Ni0:02 TiO3), the precipitation of the NiTiO3 second phase occurs. The concentration of nickel in the solid solution phase is small, as it follows from the EXAFS data and the sample color. In contrast, when annealing the Sr0:98 Ni0:02 TiO3 sample at a lower temperature (1100 C), nickel in the sample is in a mixture of NiTiO3 and SrTi1þx NixO3 solid solution as follows from the XANES data, EXAFS data, and the sample color. The precipitation of a small portion of TiO2 confirms the incorporation of Ni into SrTiO3 (the appearance of TiO2 is a consequence of removing of some Ti atoms from the B sites when doping strontium titanate with nickel). As concerns to the appearance of the TiO2 phase in the SrTi0:97 Ni0:03 O3 sample annealed at 1100 C, we suppose that its appearance is possible for kinetic reasons. The formation of the solid solution occurs as a result of a chain of chemical reactions in which the NiTiO3 phase initially formed at $ 750 C from NiO and TiO2 reacts with excess SrO resulted from the thermal decomposition of SrCO3 at $ 1000 C to produce SrNiO2:5 ×TiO2 as intermediate phases. This reaction is quite slow (the preparation of SrNiO2:5 usually takes 48 h at 1000 C, Ref. 22). As a result, in the samples annealed at 1100 C for 16 h, a mixture of the solid solution, NiTiO3, and TiO2 is observed. At higher annealing temperature, the kinetic processes are faster, the reaction completes, and the composition of the samples is fully controlled by the deviation from stoichiometry. Therefore, the stable phases in the samples are the singlephase SrTi1þx NixO3 solid solution and NiTiO3; their ratio depends on the deviation from stoichiometry and the annealing temperature. Nickel cannot be incorporated into the A sites of strontium titanate. The low stability of Ni at the A site of SrTiO3 is apparently related to the large difference in the ionic radii of Ni 2× and Sr 2×. The 12-fold coordination is not typical for nickel; for the coordination number of 6, the ionic radius of Ni 2× (0.69 å) is much less than that of Sr 2× (1.18 å).23 The XANES data, which were used to determine the oxidation state of nickel, are fully consistent with X-ray and EXAFS data. In the sample with a nominal composition of Sr0:98 Ni0:02 TiO3 annealed at 1500 C, in which nickel is in the NiTiO3 phase, the oxidation state of Ni coincides with its oxidation state in NiO and NiTiO3 and is 2×. In the singlephase SrTi0:97 Ni0:03 O3 sample annealed at 1500 C, the shift of the absorption edge with respect to NiTiO3 is maximum and is about twice the shift between the reference compounds NiTiO3 and BaNiO3þ . If, following Ref. 21, we start from the number of ions, their nominal charge, and the % 0:4 value determined from the lattice parameters, the average oxidation state of nickel in BaNiO3þ is Ï4 þ 2÷ % 3:2. Then, according to the shift of the absorption edge in SrTi0:97 Ni0:03 O3 sample annealed at 1500 C, the oxidation state of nickel in it should be close to 4×. For the SrTi0:97 Ni0:03 O3 and Sr0:98 Ni0:02 TiO3 samples annealed at 1100 C, which are the mixture of two nickel-containing

J. Adv. Dielect. 2013.03. Downloaded from www.worldscientific.com by 95.143.213.243 on 02/16/14. For personal use only.

1350031-4


I. A. Sluchinskaya, A. I. Lebedev & A. Erko

J. Adv. Dielect. 3, 1350031 (2013)

phases in proportion close to 1:1, the position of the absorption edge is intermediate between those in two stable phases and is close to the position of the absorption edge in BaNiO3þ (about 3×). It should be noted that the question about the oxidation state of Ni in SrTi0:97 Ni0:03 O3 is not so simple. The discussion about the oxidation state of Ni in BaNiO3þ is still going on, in particular, the doubts were expressed24 about the validity of its formal determination based on the number of ions and their nominal charges. In the cited work, the data of M ossbauer spectroscopy for BaNiO3 indicated the Ni oxidation state close to 4×, whereas the photoelectron spectroscopy gave a value close to 3+. The authors of Ref. 24 suggested a model in which Ni is trivalent and its charge is compensated by a hole bound by one or two negative oxygen ions. Within this model, the Ni oxidation state determined from the absorption edge shift in our SrTi0:97 Ni0:03 O3 solid solution should be closer to 3×. This explanation, however, contradicts the fact that our EXAFS data analysis revealed no distortion of the oxygen octahedra, whereas the localization of a hole at one or two oxygen ions should cause its distortion. At the same time, the experimental study of XANES spectra of LixNiO2 compound used in lithium batteries have shown that a variation in the degree of intercalation x changes the Ni oxidation state from 2× to 4× and shifts the absorption edge in XANES spectra by $ 3.5 eV.25 This shift is close to the shift of 2.9 eV observed in our spectra between NiTiO3 and the SrTi0:97 Ni0:03 O3 sample annealed at 1500 C. Moreover, the interatomic Ni­O distance (1.914 å) obtained from our EXAFS measurements is less than the sum of the ionic radii of Ni 3× and O 2þ (0:56 × 1:4 ¼ 1:96 å) and is closer to the sum of the ionic radii of Ni 4× and O 2þ (0:48 × 1:4 ¼ 1:88 å). Yet another argument in favor of tetravalent nickel can be the small value of the Debye­Waller factor 2, which 1 indicates the absence of the Ni displacement from the B site in the SrTi0:97 Ni0:03 O3 solid solution. If the nickel atom in this phase was in the Ni 3× oxidation state, it required to be charge-compensated by the oxygen vacancy located nearby the impurity atom (such Ni 3× ­VO axial centers with the Ni displacement from the center of the octahedron up to $ 0.3 å have been observed in electron spin resonance (ESR) spectra).26,27 However, our EXAFS measurements did not reveal noticeable distortion of the oxygen octahedra surrounding the Ni atom. It should be noted that the EXAFS data analysis of the SrTi0:97 Ni0:03 O3 sample annealed at 1500 C revealed a reduced value of 4.90 ô 0.34 for the coordination number for the first shell of Ni. There are two possible explanation of this fact. First, it can be an evidence for the existence of a small amount of Ni 3× ­VO complexes. The coordination number for these complexes is effectively reduced to 4 because one of the oxygen atoms is missing and the other one is located at a different distance compared to four remaining O atoms and does not contribute much to the EXAFS signal. This explanation agrees with the reduced shift of the absorption edge in our sample with respect to the Ni 4× state in LixNiO2. Another

explanation is the existence of a small amount of the NiTiO3 phase located at the grain boundaries; the experiment shows that the contamination with this phase strongly decreases the coordination number because of out-of-phase EXAFS oscillations in two nickel-containing phases. Neverthless, we think that the oxidation state of Ni at the B site in SrTiO3 is close to 4×; this disagrees with the suggestion made earlier5,6 that Ni is in the 2× oxidation state in the related ferroelectric PbTiO3. Although the question about the actual oxidation state of Ni in the latter material should be tested experimentally, our findings indicate that the strong light absorption in Ni-doped SrTiO3 is associated with the tetravalent nickel. From the viewpoint of possible application of doped perovskites in solar energy conversion it is interesting to compare the properties of SrTiO3 doped with nickel and with other 3d elements.13­16 The absorption spectra of doped samples are systematically shifted to the infrared region with increasing atomic number from Mn to Ni: manganese-doped samples are greenish brown, iron-doped samples are brown, cobalt-doped samples are dark brown, and Ni gives an almost black color to the samples. Thus, for creating samples that strongly absorb light in the whole visible region, the nickel doping seems the most promising. Intense absorption in the samples suggests that it is associated with charge-transfer transitions. Interestingly, the small value of the Debye­Waller factor for the first shell ( 2 % 0:0035 å2) also excludes the possi1 bility of Jahn­Teller instability of the Ni 4× ion in SrTiO3, which is possible for the octahedral d 6 configuration. In LixNiO2, the Jahn­Teller instability of Ni 4× manifests itself in the distorted oxygen octahedra with Ni­O bond lengths of 1.88 and 2.08 å. In the EXAFS spectra, this distorted environment should be observed as one O shell with an average interatomic distance of 1.947 å and a static Debye­Waller factor of 0.009 å2. The experimental value of 2 in 1 SrTiO3(Ni) is much lower than the value estimated for Jahn­ Teller-distorted environment. The interpretation of our data on the oxidation state of nickel in SrTiO3 differs much from the results of earlier studies. The comparison with the data of Ref. 28, in which SrTiO3(Ni) samples were also studied by XAFS technique, shows that the XANES and EXAFS spectra obtained in our work and in Ref. 28 are qualitatively different. For example, in Ref. 28 the shift of the absorption edge in SrTiO3(Ni) with respect to NiO was only 1.1 eV (in our samples it is 2.5 eV). Moreover, even the color of the samples is different (beige in Ref. 28 and almost black in our work). We believe that these differences in the properties of the samples result from the different methods of their preparation (solid-state reaction method in our case and hydrothermal synthesis at 150 C in Ref. 28). In Refs. 26 and 27, single crystals of SrTiO3(Ni) were investigated by electron spin resonance. In these works, the ESR spectra associated with axial Ni 3× ­VO complexes were systematically observed along with the ESR spectra which were attributed to the \cubic" Ni 2× and Ni 3×

J. Adv. Dielect. 2013.03. Downloaded from www.worldscientific.com by 95.143.213.243 on 02/16/14. For personal use only.

1350031-5


I. A. Sluchinskaya, A. I. Lebedev & A. Erko

J. Adv. Dielect. 3, 1350031 (2013) A. M. Glass, D. von der Linde and T. J. Negran, High-voltage bulk photovoltaic effect and the photorefractive process in LiNbO3, Appl. Phys. Lett. 25, 233 (1974). 3 M. Qin, K. Yao and Y. C. Liang, High efficient photovoltaics in nanoscaled ferroelectric thin films, Appl. Phys. Lett. 93, 122904 (2008). 4 M. Alexe and D. Hesse, Tip-enhanced photovoltaic effects in bismuth ferrite, Nat. Commun. 2, 256 (2011). 5 J. W. Bennett, I. Grinberg and A. M. Rappe, New highly polar semiconductor ferroelectrics through d8 cation-o vacancy substitution into PbTiO3: A theoretical study, J. Am. Chem. Soc. 130, 17409 (2008). 6 G. Y. Gou, J. W. Bennett, H. Takenaka and A. M. Rappe, Post density functional theoretical studies of highly polar semiconductive Pb(Ti1­xNix)O3­x solid solutions: Effects of cation arrangement on band gap, Phys. Rev. B 83, 205115 (2011). 7 Y. Inaguma, M. Yoshida, T. Tsuchiya, A. Aimi, K. Tanaka, T. Katsumata and D. Mori, High-pressure synthesis of novel lithium niobate-type oxides, J. Phys.: Conf. Ser. 215, 012131 (2010). 8 Y. Inaguma, K. Tanaka, T. Tsuchiya, D. Mori, T. Katsumata, T. Ohba, K.-I. Hiraki, T. Takahashi and H. Saitoh, Synthesis, structural transformation, thermal stability, valence state, and magnetic and electronic properties of PbNiO3 with perovskite- and LiNbO3type structures, J. Am. Chem. Soc. 133, 16920 (2011). 9 X. F. Hao, A. Stroppa, S. Picozzi, A. Filippetti and C. Franchini, Exceptionally large room-temperature ferroelectric polarization in the PbNiO3 multiferroic nickelate: First-principles study, Phys. Rev. B 86, 014116 (2012). 10 M. Azuma, S. Carlsson, J. Rodgers, M. G. Tucker, M. Tsujimoto, S. Ishiwata, S. Isoda, Y. Shimakawa, M. Takano and J. P. Attfield, Pressure-induced intermetallic valence transition in BiNiO3, J. Am. Chem. Soc. 129, 14433 (2007). 11 V. V. Shvartsman, S. Bedanta, P. Borisov, W. Kleemann, A. Tkach and P. M. Vilarinho, (Sr, Mn)TiO3: A magnetoelectric multiglass, Phys. Rev. Lett. 101, 165704 (2008). 12 W. Kleemann, S. Bedanta, P. Borisov, V. V. Shvartsman, S. Miga, J. Dec, A. Tkach and P. M. Vilarinho, Multiglass order and magnetoelectricity in Mn 2× doped incipient ferroelectrics, Eur. Phys. J. B 71, 407 (2009). 13 A. I. Lebedev, I. A. Sluchinskaya, A. Erko and V. F. Kozlovskii, Direct evidence for off-centering of Mn impurity in SrTiO3, JETP Lett. 89, 457 (2009). 14 I. A. Sluchinskaya, A. I. Lebedev and A. Erko, Local environment and oxidation state of a Mn impurity in SrTiO3 determined from XAFS data, Bull. Russ. Acad. Sci.: Phys. 74, 1235 (2010). 15 I. A. Sluchinskaya, A. I. Lebedev and A. Erko. The 19th AllRussian Conf. Physics of Ferroelectrics, Moscow (2011), Abstract book (2011), p. 116. 16 I. A. Sluchinskaya, A. I. Lebedev, V. F. Kozlovskii and A. Erko, The 8th National Conf. \X-ray, synchrotron radiation, neutrons, and electrons for studying nanosystems and materials", Moscow (2011), Abstract book (2011), p. 347. 17 A. I. Lebedev, I. A. Sluchinskaya, V. N. Demin and I. Munro, EXAFS studies of of the influence of impurities on the phase transition in GeTe, Izv. Akad. Nauk SSSR, Ser. Fiz. 60, 46 (1996). 18 A. I. Lebedev, I. A. Sluchinskaya, V. N. Demin and I. H. Munro, Off-centering of Pb and Sn impurities in GeTe, Phys. Rev. B 55, 14770 (1997).
2

centers.26 Unfortunately, no arguments on the basis of which the oxidation state of the Ni impurity was identified were presented in Ref. 26, and a possible interpretation of the ESR spectra as the spectra of Ni 4× was not discussed.

5. Conclusion The study of Ni-doped SrTiO3 using X-ray diffraction and XAFS spectroscopy enabled to draw the following conclusions: 1. The preparation conditions of single-phase Ni-doped SrTiO3 samples with the impurity concentration up to at least 3% have been found. When incorporating the impurity into the A sites, the NiTiO3 second phase was precipitated. Regardless of the preparation conditions, the SrTi1þx NixO3 solid solution and NiTiO3 phase are the most stable phases which can coexist. 2. EXAFS analysis showed that in the single-phase SrTi0:97 Ni0:03 O3 sample, the Ni atoms substitute for the Ti atoms and are on-center. The distortion of the oxygen octahedra, which could occur because of the Ni off-centering or the existence of the oxygen vacancies in the Ni environment, was not detected. The reduced value of the coordination number for the first shell was explained by the presence of a small amount of Ni 3× ­VO complexes and the NiTiO3 phase at the grain boundaries. 3. XANES analysis showed that in NiTiO3 the oxidation state of Ni is 2× and in the SrTi1þx NixO3 solid solution it is close to 4×. 4. The strongest light absorption in doped samples is associated with the presence of tetravalent nickel in the SrTi1þx NixO3 solid solution. This doping seems to be the most promising for solar energy converters that exploit the bulk photovoltaic effect. In the future, we plan to perform similar experiments on the ferroelectric BaTiO3 doped with 3d elements in order to test the feasibility of obtaining the polar material with similar optical and physical properties, in which the bulk photovoltaic effect can be observed for photons in the whole spectrum of solar radiation.

J. Adv. Dielect. 2013.03. Downloaded from www.worldscientific.com by 95.143.213.243 on 02/16/14. For personal use only.

Acknowledgments This work was supported by Russian Foundation for Basic Research (Grant No. 13-02-00724). I. A. S. and A. I. L. are grateful to Russian-German laboratory for hospitality and financial support during their stay at BESSY.

References
1

B. I. Sturman and V. M. Fridkin, Photogalvanic Effect in Media without the Inversion Center and Related Phenomena (Nauka, Moscow, 1992).

1350031-6


I. A. Sluchinskaya, A. I. Lebedev & A. Erko
19

J. Adv. Dielect. 3, 1350031 (2013)
25

FEFF project home page, http://leonardo.phys.washington.edu/ feff/. URL http://leonardo.phys.washington.edu/feff/. 20 IFEFFIT project home page, http://cars9.uchicago.edu/ifeffit/. URL http://cars9.uchicago.edu/ifeffit/. 21 M. Arjomand and D. J. Machin, Oxide chemistry. Part I. Ternary oxides containing nickel in oxidation states II, III, and IV, J. Chem. Soc., Dalton Trans., 1055 (1975). 22 Y. Takeda, T. Hashino, H. Miyamoto, F. Kanamaru, S. Kume and M. Koizumi, Synthesis of SrNiO3 and related compound, Sr2Ni2O5, J. Inorg. Nucl. Chem. 34, 1599 (1972). 23 R. D. Shannon, Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides, Acta Cryst. A 32, 751 (1976). 24 R. Gottschall, R. Sch ollhorn, M. Muhler, N. Jansen, D. Walcher and P. GÝtlich, Electronic state of nickel in barium nickel oxide, BaNiO3, Inorg. Chem. 37, 1513 (1998).

26

27

28

A. N. Mansour and C. A. Melendres, Analysis of X-ray absorption spectra of some nickel oxycompounds using theoretical standards, J. Phys. Chem. A 102, 65 (1998). K. A. MÝller, W. Berlinger and R. S. Rubins, Observation of two charged states of a nickel-oxygen vacancy pair in SrTiO3 by paramagnetic resonance, Phys. Rev. 186, 361 (1969). S.-Y. Wu, J.-Z. Lin, Q. Fu and H.-M. Zhang, Investigations on the impurity displacements and the g factors of the two tetragonal Ni 3× centres in SrTiO3, Phys. Scripta. 75, 147 (2007). A. M. Beale, M. Paul, G. Sankar, R. J. Oldman, C. R. A. Catlow, S. French and M. Fowles, Combined experimental and computational modelling studies of the solubility of nickel in strontium titanate, J. Mater. Chem. 19, 4391 (2009).

J. Adv. Dielect. 2013.03. Downloaded from www.worldscientific.com by 95.143.213.243 on 02/16/14. For personal use only.

1350031-7