- Email: [email protected]
Structure characteristics and valence state study for La1−xNaxTiO3 synthesized under high-pressure and high-temperature conditions
January 2000
Materials Letters 42 Ž2000. 1–6 www.elsevier.comrlocatermatlet
Structure characteristics and valence state study for La 1yx Na xTiO 3 synthesized under high-pressure and high-temperature conditions Ji-Peng Miao a,b,) , Li-Ping Li a,b, Hong-Jian Liu a , Da-Peng Xu a , Zhe Lu a , Yan-Bin Song a , Wen-Hui Su a,b,c,d , Ying-Guang Zheng e b
a Department of Physics, Jilin UniÕersity, Changchun 130023, China Key Laboratory of Rare Earth Chemistry and Physics, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China c International Center for Materials Physics, Academia Sinica, Shenyang 110015, China d Center for Condensed Matter and Radiation Physics, CCAST (World Lab.), P.O. Box 8730, Beijing 100080, China e Analysis Measurement Center, Jilin UniÕersity, Changchun 130023, China
Received 29 January 1999; received in revised form 14 June 1999; accepted 16 June 1999
Abstract By using a novel high-pressure, high-temperature method, perovskite oxides of La 1yx Na xTiO 3 Ž x s 0.05, 0.1–0.8. with mixed valence state were synthesized. XRD analysis shows a cubic cell for the samples. Cell volumes of the samples with 0.1 F =F 0.5 decreases as x increases, and the cell volume for x s 0.05 is smaller than that for x s 0.1. XPS of surface and EPR measurements indicate that Ti ions are of mixed valence of q3 and q4 and that A-cations vacancies exist in the samples. As x increases, the amount of Ti 3q ions decreases and the amount of A-cations vacancies increases. The valence state of Ti ions can be altered by changing both pressure and temperature. q 2000 Elsevier Science B.V. All rights reserved. Keywords: High-pressure and high-temperature synthesis; XPS; EPR; Mixed valence oxides
1. Introduction Due to their many interesting properties, perovskite oxides, RTiO 3 compounds ŽR s rare earth., have attracted many researchers recently. RTiO 3 compounds can be viewed as distorted perovskites with orthorhombic structure ŽGdFeO 3-type.. Stoichiometric LaTiO 3 is a typical Mott insulator at room temperature. Substitution of La3q by divalent ions ŽBa2q, Sr 2q . results in rich structural character)
Corresponding author.
istics, transport and magnetic properties w1–4x, whereas substitution by Naq, i.e., La 0.5 Na 0.5TiO 3 produces paraelectricity w5x. Polycrystalline materials of La 1y x Na xTiO 3 Ž0 F x F 0.4. were prepared by solid state reaction in an Ar atmosphere by mixing La 0.5 Na 0.5TiO 3 and LaTiO 3 , both previously obtained by heating a mixture of La 2 O 3 , Na 2 CO 3 , TiO, Ti 2 O 3 and TiO 2 in an Ar atmosphere. Semiconductor to metal transitions has been found in La 1y x Na xTiO 3 Ž x G 0.1. as temperature decreased w6x. Because Ti 4q ions are considerably stable at ambient condition, the synthesis of R 1y x M xTiO 3
00167-577Xr00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 5 7 7 X Ž 9 9 . 0 0 1 4 9 - 4
2
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
compounds ŽR s La, Nd and Y, M s Ba2q, Sr 2q, Ca2q, Kq and Naq . required a reducing H 2 atmosphere when using TiO 2 as starting material or an inert Ar atmosphere when using Ti, TiO, Ti 2 O 3 and TiO 2 . It has been reported that the valence states can be altered under high-pressure and high-temperature conditions w7,8x. In this paper, we use a new synthesis route. By using a high-pressure, high-temperature method, perovskite oxides of La 1y x Na xTiO 3 were synthesized under 4.0 GPa and at 9308C by using TiO 2 as starting material. The structure and mixed valence characteristics of the samples were investigated by X-ray diffraction ŽXRD., XPS and EPR.
2. Experimental The samples were prepared by mixing TiO 2 , La 2 O 3 and NaHCO 3 according to the nominal ratio of La 1y x Na xTiO 3 . The mixtures were thoroughly ground and placed into a high-pressure chamber Žsee Fig. 1.. High-pressure, high-temperature synthesis was carried out using a belt-type high-pressure apparatus. The magnitude of the pressure generated inside the cell was calibrated by the electrical resistance change related to the phase transformation of Bi, Tl and Ba Ž2.55, 2.69 and 7.7 GPa.. The samples were heated electrically through a graphite heater, and the temperature was measured by inserting a Pt30%Rh–
Fig. 1. High-pressure chamber: Ž1. pyrophyllite, Ž2. molybdenum metal pellet, Ž3. graphite heater, Ž4. sample, Ž5. hexagonal borazon Žh-BN. tube, Ž6. pyrophyllite, Ž7. steel cylinder.
Pt6%Rh thermocouple in the cell. The high-pressure and high-temperature synthesis procedure is aimed to increase the pressure to 4.0 GPa, then increase the temperature to 9308C. After being kept under highpressure and high-temperature for 30 min, the samples were quenched to room temperature under high-pressure, and finally, the pressure was released. Powder XRD data were collected at room temperature on a Rigaku 12 kW copper rotating anode X-ray diffractometer. The XRD data for index and cell parameter calculations were collected by a scanning mode with a step interval of 0.028 and a preset time of 4 s per step with silicon used as an internal standard. The XPS for the powder samples were measured on an ESCALAB MKII X-ray photoelectron spectrometer with MgK a radiation and the base pressure was 10y7 Pa. C1s s 284.6 eV was used to correct the charge effect. The EPR for the powder samples were measured on a BRUKER ER200D EPR spectrometer at room temperature with 9.79 GHz microwave frequency, 6.5 mW microwave power, 100 kHz modulator frequency, 0.32 mT field modulation intensity, 0.348 T middle range and 0.6 T scan range. During the measurements, both the amount Ž10y3 mol. and measurement conditions were kept the same for all the samples.
3. Results and discussion Fig. 2 shows the XRD patterns of La 1y x Na xTiO 3 Ž x s 0.05, 0.1–0.8. samples. The samples with x s 0.05, 0.1, 0.5 are single phases and belong to the cubic perovskite structures, however, for other samples, trace of TiO 2 exist as impurity, though the main phase still keep the cubic symmetry. A structural study on double perovskite oxide ŽAAX .ŽBBX .O 3 w5,9x showed that an ordered distribution of A and AX or B and BX cations produced a relatively strong superstructure peak Ž1r2 1r2 1r2. around 198. In our samples, the absence of this peak suggests that La3q and Naq ions are disorderly distributed at A-site. The weak superstructure peak Ž3r2 1r2 1r2. Žat ca. 388. observed for all the samples results from the tilt of TiO6 octahedron, though the symmetry is determined as cubic. As reported in Ref. w6x, the superstructure peak Ž3r2 1r2 1r2. is observed for x s 0.1 but does not appear for 0.2 F x F 0.4. It is
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
Fig. 2. XRD patterns of La 1y x Na xTiO 3 samples Ž x s 0.05, 0.1– 0.8..
clear that the different preparation routes by using different starting materials may produce materials with different microstructure characteristics. For the samples with 0.2 F x F 0.4 and 0.6 F x F 0.8, TiO 2 exists as the impurity. According to the formula La 1y x Na xTiO 3 , the ratio of ŽLa,Na.rTi should be equal to 1. When the amount of A-cations deviate from this ratio, the number of B-cations ŽTi ions. must decrease as to maintain the perovskite lattice. It is reasonable that the appearance of TiO 2 is due to loss of Na ions during the synthesizing process, which is similar to the report in Ref. w6x. This leads to A-cations vacancies as well as the observed excess of TiO 2 . The lattice parameters were calculated for the samples with x F 0.5. The relationship between the cell volume and x is shown in Fig. 3. In the range of 0.1 F x F 0.5, the cell volume decreases as x increases, but the volume of the x s 0.05 sample is smaller than that of the x s 0.1 sample. In La 0.5 Na 0.5 TiŽIV.O 3 , La3q and Naq ions have their
3
respective 50% occupancy at A-site. The average charge of A-site in La 1y x Na xTiO 3 increases as x decreases. In order to maintain electrical neutrality, the reduction from TiŽIV. to TiŽIII. or A-cation vacancies must be produced. These two factors have an important influence on the cell volume. The following analysis of XPS and EPR shows that both TiŽIII. ions and A-cation vacancies exist in the samples with 0.1 F x F 0.4. The decrease of cell volume with the increasing of x suggests that the effect of B-site substitution of TiŽIII.Ž0.067 nm . by TiŽIV.Ž0.0605 nm. is predominant over that of A-site substitution of LaŽIII.Ž0.136 nm. by NaqŽ 0.139 nm.. The smaller cell volume for the sample with x s 0.05 is due to the absence of A-cation vacancies as detected by EPR. The surface characteristics of the samples were studied by XPS. XPS spectra recorded at room temperature are shown in Fig. 4. The binding energies of core levels are listed in Table 1. It can be seen that the binding energies of La3d 5r2 are between 835.0 eV and 835.6 eV, suggesting the appearance of LaŽIII.. The satellite peak appearing at the higher binding energy side of the La3d 5r2 level is ca. 3.8 eV away from the main peak. The electronic structures of LaBO 3 ŽB s Ti, Cr, Mn, Fe and Co. has been investigated using XPS w10x. In the La3d spectra of LaBO 3 , the satellite peaks observed on the high binding energy side of the main peak by about 4 eV were interpreted in terms of the excitation of an electron from the anion valence band into the La 4f band. From these results, it is considered that the
Fig. 3. The change of cell volumes vs. dopant x.
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
4
Fig. 4. Core level spectra of La3d and Ti2p. The data in brackets are half height width of Ti2p 3r 2 .
presence of satellite of the La3d peak in our samples is due to the monopole excitation arising from a sudden change in the screening of the valence electrons upon the removal of a core electron. The half height widths of Ti2p 3r2 of the samples are given in Table 1 and Fig. 4. Rao and Sarma w11x studied the Ti2p spectra for Ti 4 O 7 and Ti 3 O5 . It was found that there was a small shoulder at low binding energy side apart from main peak about 1 eV, which is thought to be Ti 3q 2p 3r2 . As the Ti 3q ions are easily oxidized to Ti 4q ions in air and the effective electron escape depth is no more than 5 nm in the process of X-ray photoelectron emission, i.e., only the surface of the samples is detected, the Ti 3q 2p 3r2 peaks for our samples are not easily detected. However, because the positions of Ti 3q and Ti 4q ions are different, if Ti ions are in mixed valence state, Table 1 Binding energies of core level for La 1y x Na xTiO 3 ŽeV. x
O IS
La3d 5r2
Na IS
Ti2p 3r2 ŽU .
0.05 0.1 0.2 0.3 0.4 0.5
530.3 530.5 530.9 531.3 531.8 532.0
835.6 835.0 835.1 835.0 835.3 835.6
1071.2 1071.5 1071.6 1071.6 1071.5 1072.2
458.7 Ž2.26. 458.1 Ž2.20. 458.5 Ž2.18. 458.4 Ž2.17. 458.2 Ž2.15. 458.4 Ž1.96.
U
The data in parentheses are half height width of Ti2p 3r 2 .
the half height width of Ti ions in mixed valence state is wider than that of single Ti 4q ions. The measurements for core level of Ti2p show that the half height widths of Ti2p 3r2 for the samples with 0.05 F x F 0.4 are larger than that of TiO 2 Ž1.94 eV. and for x s 0.5 is almost same. Furthermore, the half height widths of Ti2p 3r2 for the samples decreases with the increasing x. This suggests that Ti ions on the surface of the samples are in mixed valence except for x s 0.5. The electronic configurations of Ti 4q and Ti 3q ions in the ground state are 3d 0 and 3d1 , respectively. There is no EPR signal for Ti 4q ions. If Ti 3q ions are in absolutely symmetric cubic crystal field, there is no EPR signal. If Ti 3q ions are in slightly asymmetric cubic crystal field, there is EPR signal. From the results of XRD, though the samples belong to cubic structures, the TiO6 octahedron tilts and La3q and Naq ions are disorderly distributed at A-site forming the positive electronic center around La3q and the negative electronic center around Naq, which lead to the distortion of lattice at a very small scale, so Ti 3q ions are in slightly asymmetric cubic crystal field and EPR signal can be observed. So TiŽIII. having one unpaired electron can be distinguished from TiŽIV. by EPR. The EPR spectra of the samples with 0.05 F x F 0.4 detected at room temperature are shown in Fig. 5. Two strong signals are
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
Fig. 5. EPR spectra of La 1y x Na xTiO 3 samples Ž x s 0.05, 0.1– 0.4..
observed in the range of 0.048–0.648 T. The g-factors of the narrow and broad signals are 2.003 and 1.943, respectively. But for the x s 0.05 sample, only one broad signal with g s 1.943 is observed. Both signals are symmetric. Studies on BaTiO 3 and donored BaTiO 3 showed two EPR signals with g s 2.00 and 1.96, respectively w12x. The signal with g s 2.00 resulted from barium vacancies and the signal with g s 1.96 was due to the TiŽIII. ions. For our samples, the signal with g s 2.003 is ascribed to A-cation vacancies, whereas the signal with g s 1.943 to the TiŽIII. ions. A-cation vacancies are due to the loss of Naq ions. The existence of Ti 3q ions shows that the valence state of Ti 4q ions can be altered by changing both pressure and temperature. The relationship between relative EPR signal intensity and x is shown in Fig. 6. The relative signal intensity of g s 1.943 decreases with the increasing of x, indicating that the number of TiŽIII. ions in
5
La 1y x Na xTiO 3 decreases with the decreasing of the average effective charge of A-site cations. The reverse change of the g s 2.003 signal intensity with increasing of x indicates that the amount of A-cations vacancies increases with increasing of x. The extremely weak signal of g s 2.003 is detected for x s 0.1. Due to the very small amount of Naq ions in the x s 0.05 sample, the signal of g s 2.003 is not observed, indicating that the charge compensation can be achieved only by reduction of Ti ions under high-pressure and high-temperature. The high-pressure, high-temperature synthesis method has some advantages: Ž1. with the effect of air-tight sealing, high-pressure and high-temperature can provide a reducing condition, which is equivalent to the use of reducing and inert gas. This causes the reduction of Ti 4q to Ti 3q ions. Ž2. High-pressure can densify the starting materials, increase the cross-section of the reaction, accelerate the diffusion between the reaction materials powders and decrease the reaction activation energy, thus accelerate the reaction rate, shorten the synthesis time and decrease the synthesis temperature. In comparison with the other preparation method in Ref. w6x, temperature is lower and time is shorter in the present case. Ž3. Perovskite oxides of La 1y x Na xTiO 3 Ž x s 0.05–0.4. can be synthesized under high-pressure and hightemperature, but can not be synthesized by the solid
Fig. 6. Relative intensity of EPR signal vs. dopant x.
6
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
state reaction method in air at temperature below 13008C Žno higher temperatures were investigated..
4. Conclusions From the above discussion, it can be concluded that high-pressure, high-temperature has an important influence on crystal symmetry and valence state of Ti ions. Therefore, the high-pressure, high-temperature synthesis method is very effective in synthesizing the perovskite oxides of mixed valence, especially those containing Ti 3q ions. It is expected that the perovskite oxides of La 1y x Na xTiO 3 with mixed valence state induced by high-pressure and high-temperature might have some interesting properties which are to be studied.
Acknowledgements This work was supported by NSFC ŽNo. 19804005..
References w1x Y. Tokura, T. Taguchi, Y. Okada, Y. Fujishima, T. Arima, K. Kumagai, Y. Iye, Phys. Rev. Lett. 70 Ž14. Ž1993. 2126– 2129. w2x Y. Okaka, T. Arima, Y. Tokura, C. Murayama, N. Mori, Phys. Rev. B 48 Ž13. Ž1993. 9677–9683. w3x S.W. Robey, L.T. Hudson, C. Eylem, B. Eichorn, Phys. Rev. B 48 Ž1. Ž1993. 562–568. w4x J.E. Sunstrom IV, S.M. Kauzlarich, Chem. Mater. 5 Ž1993. 1539–1544. w5x Y. Inaguma, J.H. Sohn, I.S. Kim, M. Itoh, T. Nakamura, J. Phys. Soc. Jpn. 61 Ž10. Ž1992. 3831–3832. w6x H. Yamamoto, T. Tahara, Y. Sugahara, K. Kuroda, C. Kato, Phase Transitions 41 Ž1993. 137–141. w7x L. Li, Q. Wei, H. Liu, D. Zheng, W. Su, Z. Phys. B 96 Ž1995. 451–454. w8x W. Su, H. Liu, L. Li, G. Chen, Acta Sci. Nat. Univ. Jilin., Suppl. Ž1992. 188–201. w9x M.T. Anderson, K.B. Greenwood, G.A. Taylor, K.R. Poeppelmeier, Prog. Solid State Chem. 122 Ž1993. 197–233. w10x D.J. Lam, B.W. Veal, D.E. Ellis, Phys. Rev. B 22 Ž12. Ž1980. 5730–5739. w11x C.N.R. Rao, D.D. Sarma, J. Solid State Chem. 45 Ž1982. 14–39. w12x T.R.N. Kutty, P. Murugaraj, N.S. Gajbhiye, Mater. Res. Bull. 20 Ž1985. 565–574.
Materials Letters 42 Ž2000. 1–6 www.elsevier.comrlocatermatlet
Structure characteristics and valence state study for La 1yx Na xTiO 3 synthesized under high-pressure and high-temperature conditions Ji-Peng Miao a,b,) , Li-Ping Li a,b, Hong-Jian Liu a , Da-Peng Xu a , Zhe Lu a , Yan-Bin Song a , Wen-Hui Su a,b,c,d , Ying-Guang Zheng e b
a Department of Physics, Jilin UniÕersity, Changchun 130023, China Key Laboratory of Rare Earth Chemistry and Physics, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022, China c International Center for Materials Physics, Academia Sinica, Shenyang 110015, China d Center for Condensed Matter and Radiation Physics, CCAST (World Lab.), P.O. Box 8730, Beijing 100080, China e Analysis Measurement Center, Jilin UniÕersity, Changchun 130023, China
Received 29 January 1999; received in revised form 14 June 1999; accepted 16 June 1999
Abstract By using a novel high-pressure, high-temperature method, perovskite oxides of La 1yx Na xTiO 3 Ž x s 0.05, 0.1–0.8. with mixed valence state were synthesized. XRD analysis shows a cubic cell for the samples. Cell volumes of the samples with 0.1 F =F 0.5 decreases as x increases, and the cell volume for x s 0.05 is smaller than that for x s 0.1. XPS of surface and EPR measurements indicate that Ti ions are of mixed valence of q3 and q4 and that A-cations vacancies exist in the samples. As x increases, the amount of Ti 3q ions decreases and the amount of A-cations vacancies increases. The valence state of Ti ions can be altered by changing both pressure and temperature. q 2000 Elsevier Science B.V. All rights reserved. Keywords: High-pressure and high-temperature synthesis; XPS; EPR; Mixed valence oxides
1. Introduction Due to their many interesting properties, perovskite oxides, RTiO 3 compounds ŽR s rare earth., have attracted many researchers recently. RTiO 3 compounds can be viewed as distorted perovskites with orthorhombic structure ŽGdFeO 3-type.. Stoichiometric LaTiO 3 is a typical Mott insulator at room temperature. Substitution of La3q by divalent ions ŽBa2q, Sr 2q . results in rich structural character)
Corresponding author.
istics, transport and magnetic properties w1–4x, whereas substitution by Naq, i.e., La 0.5 Na 0.5TiO 3 produces paraelectricity w5x. Polycrystalline materials of La 1y x Na xTiO 3 Ž0 F x F 0.4. were prepared by solid state reaction in an Ar atmosphere by mixing La 0.5 Na 0.5TiO 3 and LaTiO 3 , both previously obtained by heating a mixture of La 2 O 3 , Na 2 CO 3 , TiO, Ti 2 O 3 and TiO 2 in an Ar atmosphere. Semiconductor to metal transitions has been found in La 1y x Na xTiO 3 Ž x G 0.1. as temperature decreased w6x. Because Ti 4q ions are considerably stable at ambient condition, the synthesis of R 1y x M xTiO 3
00167-577Xr00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 5 7 7 X Ž 9 9 . 0 0 1 4 9 - 4
2
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
compounds ŽR s La, Nd and Y, M s Ba2q, Sr 2q, Ca2q, Kq and Naq . required a reducing H 2 atmosphere when using TiO 2 as starting material or an inert Ar atmosphere when using Ti, TiO, Ti 2 O 3 and TiO 2 . It has been reported that the valence states can be altered under high-pressure and high-temperature conditions w7,8x. In this paper, we use a new synthesis route. By using a high-pressure, high-temperature method, perovskite oxides of La 1y x Na xTiO 3 were synthesized under 4.0 GPa and at 9308C by using TiO 2 as starting material. The structure and mixed valence characteristics of the samples were investigated by X-ray diffraction ŽXRD., XPS and EPR.
2. Experimental The samples were prepared by mixing TiO 2 , La 2 O 3 and NaHCO 3 according to the nominal ratio of La 1y x Na xTiO 3 . The mixtures were thoroughly ground and placed into a high-pressure chamber Žsee Fig. 1.. High-pressure, high-temperature synthesis was carried out using a belt-type high-pressure apparatus. The magnitude of the pressure generated inside the cell was calibrated by the electrical resistance change related to the phase transformation of Bi, Tl and Ba Ž2.55, 2.69 and 7.7 GPa.. The samples were heated electrically through a graphite heater, and the temperature was measured by inserting a Pt30%Rh–
Fig. 1. High-pressure chamber: Ž1. pyrophyllite, Ž2. molybdenum metal pellet, Ž3. graphite heater, Ž4. sample, Ž5. hexagonal borazon Žh-BN. tube, Ž6. pyrophyllite, Ž7. steel cylinder.
Pt6%Rh thermocouple in the cell. The high-pressure and high-temperature synthesis procedure is aimed to increase the pressure to 4.0 GPa, then increase the temperature to 9308C. After being kept under highpressure and high-temperature for 30 min, the samples were quenched to room temperature under high-pressure, and finally, the pressure was released. Powder XRD data were collected at room temperature on a Rigaku 12 kW copper rotating anode X-ray diffractometer. The XRD data for index and cell parameter calculations were collected by a scanning mode with a step interval of 0.028 and a preset time of 4 s per step with silicon used as an internal standard. The XPS for the powder samples were measured on an ESCALAB MKII X-ray photoelectron spectrometer with MgK a radiation and the base pressure was 10y7 Pa. C1s s 284.6 eV was used to correct the charge effect. The EPR for the powder samples were measured on a BRUKER ER200D EPR spectrometer at room temperature with 9.79 GHz microwave frequency, 6.5 mW microwave power, 100 kHz modulator frequency, 0.32 mT field modulation intensity, 0.348 T middle range and 0.6 T scan range. During the measurements, both the amount Ž10y3 mol. and measurement conditions were kept the same for all the samples.
3. Results and discussion Fig. 2 shows the XRD patterns of La 1y x Na xTiO 3 Ž x s 0.05, 0.1–0.8. samples. The samples with x s 0.05, 0.1, 0.5 are single phases and belong to the cubic perovskite structures, however, for other samples, trace of TiO 2 exist as impurity, though the main phase still keep the cubic symmetry. A structural study on double perovskite oxide ŽAAX .ŽBBX .O 3 w5,9x showed that an ordered distribution of A and AX or B and BX cations produced a relatively strong superstructure peak Ž1r2 1r2 1r2. around 198. In our samples, the absence of this peak suggests that La3q and Naq ions are disorderly distributed at A-site. The weak superstructure peak Ž3r2 1r2 1r2. Žat ca. 388. observed for all the samples results from the tilt of TiO6 octahedron, though the symmetry is determined as cubic. As reported in Ref. w6x, the superstructure peak Ž3r2 1r2 1r2. is observed for x s 0.1 but does not appear for 0.2 F x F 0.4. It is
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
Fig. 2. XRD patterns of La 1y x Na xTiO 3 samples Ž x s 0.05, 0.1– 0.8..
clear that the different preparation routes by using different starting materials may produce materials with different microstructure characteristics. For the samples with 0.2 F x F 0.4 and 0.6 F x F 0.8, TiO 2 exists as the impurity. According to the formula La 1y x Na xTiO 3 , the ratio of ŽLa,Na.rTi should be equal to 1. When the amount of A-cations deviate from this ratio, the number of B-cations ŽTi ions. must decrease as to maintain the perovskite lattice. It is reasonable that the appearance of TiO 2 is due to loss of Na ions during the synthesizing process, which is similar to the report in Ref. w6x. This leads to A-cations vacancies as well as the observed excess of TiO 2 . The lattice parameters were calculated for the samples with x F 0.5. The relationship between the cell volume and x is shown in Fig. 3. In the range of 0.1 F x F 0.5, the cell volume decreases as x increases, but the volume of the x s 0.05 sample is smaller than that of the x s 0.1 sample. In La 0.5 Na 0.5 TiŽIV.O 3 , La3q and Naq ions have their
3
respective 50% occupancy at A-site. The average charge of A-site in La 1y x Na xTiO 3 increases as x decreases. In order to maintain electrical neutrality, the reduction from TiŽIV. to TiŽIII. or A-cation vacancies must be produced. These two factors have an important influence on the cell volume. The following analysis of XPS and EPR shows that both TiŽIII. ions and A-cation vacancies exist in the samples with 0.1 F x F 0.4. The decrease of cell volume with the increasing of x suggests that the effect of B-site substitution of TiŽIII.Ž0.067 nm . by TiŽIV.Ž0.0605 nm. is predominant over that of A-site substitution of LaŽIII.Ž0.136 nm. by NaqŽ 0.139 nm.. The smaller cell volume for the sample with x s 0.05 is due to the absence of A-cation vacancies as detected by EPR. The surface characteristics of the samples were studied by XPS. XPS spectra recorded at room temperature are shown in Fig. 4. The binding energies of core levels are listed in Table 1. It can be seen that the binding energies of La3d 5r2 are between 835.0 eV and 835.6 eV, suggesting the appearance of LaŽIII.. The satellite peak appearing at the higher binding energy side of the La3d 5r2 level is ca. 3.8 eV away from the main peak. The electronic structures of LaBO 3 ŽB s Ti, Cr, Mn, Fe and Co. has been investigated using XPS w10x. In the La3d spectra of LaBO 3 , the satellite peaks observed on the high binding energy side of the main peak by about 4 eV were interpreted in terms of the excitation of an electron from the anion valence band into the La 4f band. From these results, it is considered that the
Fig. 3. The change of cell volumes vs. dopant x.
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
4
Fig. 4. Core level spectra of La3d and Ti2p. The data in brackets are half height width of Ti2p 3r 2 .
presence of satellite of the La3d peak in our samples is due to the monopole excitation arising from a sudden change in the screening of the valence electrons upon the removal of a core electron. The half height widths of Ti2p 3r2 of the samples are given in Table 1 and Fig. 4. Rao and Sarma w11x studied the Ti2p spectra for Ti 4 O 7 and Ti 3 O5 . It was found that there was a small shoulder at low binding energy side apart from main peak about 1 eV, which is thought to be Ti 3q 2p 3r2 . As the Ti 3q ions are easily oxidized to Ti 4q ions in air and the effective electron escape depth is no more than 5 nm in the process of X-ray photoelectron emission, i.e., only the surface of the samples is detected, the Ti 3q 2p 3r2 peaks for our samples are not easily detected. However, because the positions of Ti 3q and Ti 4q ions are different, if Ti ions are in mixed valence state, Table 1 Binding energies of core level for La 1y x Na xTiO 3 ŽeV. x
O IS
La3d 5r2
Na IS
Ti2p 3r2 ŽU .
0.05 0.1 0.2 0.3 0.4 0.5
530.3 530.5 530.9 531.3 531.8 532.0
835.6 835.0 835.1 835.0 835.3 835.6
1071.2 1071.5 1071.6 1071.6 1071.5 1072.2
458.7 Ž2.26. 458.1 Ž2.20. 458.5 Ž2.18. 458.4 Ž2.17. 458.2 Ž2.15. 458.4 Ž1.96.
U
The data in parentheses are half height width of Ti2p 3r 2 .
the half height width of Ti ions in mixed valence state is wider than that of single Ti 4q ions. The measurements for core level of Ti2p show that the half height widths of Ti2p 3r2 for the samples with 0.05 F x F 0.4 are larger than that of TiO 2 Ž1.94 eV. and for x s 0.5 is almost same. Furthermore, the half height widths of Ti2p 3r2 for the samples decreases with the increasing x. This suggests that Ti ions on the surface of the samples are in mixed valence except for x s 0.5. The electronic configurations of Ti 4q and Ti 3q ions in the ground state are 3d 0 and 3d1 , respectively. There is no EPR signal for Ti 4q ions. If Ti 3q ions are in absolutely symmetric cubic crystal field, there is no EPR signal. If Ti 3q ions are in slightly asymmetric cubic crystal field, there is EPR signal. From the results of XRD, though the samples belong to cubic structures, the TiO6 octahedron tilts and La3q and Naq ions are disorderly distributed at A-site forming the positive electronic center around La3q and the negative electronic center around Naq, which lead to the distortion of lattice at a very small scale, so Ti 3q ions are in slightly asymmetric cubic crystal field and EPR signal can be observed. So TiŽIII. having one unpaired electron can be distinguished from TiŽIV. by EPR. The EPR spectra of the samples with 0.05 F x F 0.4 detected at room temperature are shown in Fig. 5. Two strong signals are
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
Fig. 5. EPR spectra of La 1y x Na xTiO 3 samples Ž x s 0.05, 0.1– 0.4..
observed in the range of 0.048–0.648 T. The g-factors of the narrow and broad signals are 2.003 and 1.943, respectively. But for the x s 0.05 sample, only one broad signal with g s 1.943 is observed. Both signals are symmetric. Studies on BaTiO 3 and donored BaTiO 3 showed two EPR signals with g s 2.00 and 1.96, respectively w12x. The signal with g s 2.00 resulted from barium vacancies and the signal with g s 1.96 was due to the TiŽIII. ions. For our samples, the signal with g s 2.003 is ascribed to A-cation vacancies, whereas the signal with g s 1.943 to the TiŽIII. ions. A-cation vacancies are due to the loss of Naq ions. The existence of Ti 3q ions shows that the valence state of Ti 4q ions can be altered by changing both pressure and temperature. The relationship between relative EPR signal intensity and x is shown in Fig. 6. The relative signal intensity of g s 1.943 decreases with the increasing of x, indicating that the number of TiŽIII. ions in
5
La 1y x Na xTiO 3 decreases with the decreasing of the average effective charge of A-site cations. The reverse change of the g s 2.003 signal intensity with increasing of x indicates that the amount of A-cations vacancies increases with increasing of x. The extremely weak signal of g s 2.003 is detected for x s 0.1. Due to the very small amount of Naq ions in the x s 0.05 sample, the signal of g s 2.003 is not observed, indicating that the charge compensation can be achieved only by reduction of Ti ions under high-pressure and high-temperature. The high-pressure, high-temperature synthesis method has some advantages: Ž1. with the effect of air-tight sealing, high-pressure and high-temperature can provide a reducing condition, which is equivalent to the use of reducing and inert gas. This causes the reduction of Ti 4q to Ti 3q ions. Ž2. High-pressure can densify the starting materials, increase the cross-section of the reaction, accelerate the diffusion between the reaction materials powders and decrease the reaction activation energy, thus accelerate the reaction rate, shorten the synthesis time and decrease the synthesis temperature. In comparison with the other preparation method in Ref. w6x, temperature is lower and time is shorter in the present case. Ž3. Perovskite oxides of La 1y x Na xTiO 3 Ž x s 0.05–0.4. can be synthesized under high-pressure and hightemperature, but can not be synthesized by the solid
Fig. 6. Relative intensity of EPR signal vs. dopant x.
6
J.-P. Miao et al.r Materials Letters 42 (2000) 1–6
state reaction method in air at temperature below 13008C Žno higher temperatures were investigated..
4. Conclusions From the above discussion, it can be concluded that high-pressure, high-temperature has an important influence on crystal symmetry and valence state of Ti ions. Therefore, the high-pressure, high-temperature synthesis method is very effective in synthesizing the perovskite oxides of mixed valence, especially those containing Ti 3q ions. It is expected that the perovskite oxides of La 1y x Na xTiO 3 with mixed valence state induced by high-pressure and high-temperature might have some interesting properties which are to be studied.
Acknowledgements This work was supported by NSFC ŽNo. 19804005..
References w1x Y. Tokura, T. Taguchi, Y. Okada, Y. Fujishima, T. Arima, K. Kumagai, Y. Iye, Phys. Rev. Lett. 70 Ž14. Ž1993. 2126– 2129. w2x Y. Okaka, T. Arima, Y. Tokura, C. Murayama, N. Mori, Phys. Rev. B 48 Ž13. Ž1993. 9677–9683. w3x S.W. Robey, L.T. Hudson, C. Eylem, B. Eichorn, Phys. Rev. B 48 Ž1. Ž1993. 562–568. w4x J.E. Sunstrom IV, S.M. Kauzlarich, Chem. Mater. 5 Ž1993. 1539–1544. w5x Y. Inaguma, J.H. Sohn, I.S. Kim, M. Itoh, T. Nakamura, J. Phys. Soc. Jpn. 61 Ž10. Ž1992. 3831–3832. w6x H. Yamamoto, T. Tahara, Y. Sugahara, K. Kuroda, C. Kato, Phase Transitions 41 Ž1993. 137–141. w7x L. Li, Q. Wei, H. Liu, D. Zheng, W. Su, Z. Phys. B 96 Ž1995. 451–454. w8x W. Su, H. Liu, L. Li, G. Chen, Acta Sci. Nat. Univ. Jilin., Suppl. Ž1992. 188–201. w9x M.T. Anderson, K.B. Greenwood, G.A. Taylor, K.R. Poeppelmeier, Prog. Solid State Chem. 122 Ž1993. 197–233. w10x D.J. Lam, B.W. Veal, D.E. Ellis, Phys. Rev. B 22 Ž12. Ž1980. 5730–5739. w11x C.N.R. Rao, D.D. Sarma, J. Solid State Chem. 45 Ž1982. 14–39. w12x T.R.N. Kutty, P. Murugaraj, N.S. Gajbhiye, Mater. Res. Bull. 20 Ž1985. 565–574.