Sub-solidus phase relations and a structure determination of new phases in the CaO–La2O3–TiO2 system

Available online at www.sciencedirect.com

ScienceDirect Journal of the European Ceramic Society 35 (2015) 2801–2814

Sub-solidus phase relations and a structure determination of new phases in the CaO–La2O3–TiO2 system b,∗ ˇ Maja Vidmar a , Amalija Golobiˇc a , Anton Meden a , Danilo Suvorov b , Sreˇco D. Skapin a

Faculty of Chemistry and Chemical Technology, University of Ljubljana, Veˇcna pot 113, 1000 Ljubljana, Slovenia b “Joˇzef Stefan” Institute, Jamova 39, 1000 Ljubljana, Slovenia Received 16 February 2015; received in revised form 26 March 2015; accepted 29 March 2015 Available online 19 April 2015

Abstract Sub-solidus phase relations in the ternary CaO–La2 O3 –TiO2 system at 1400 ◦ C in air were determined. The multi-phase samples were prepared by a solid-state reaction method, whereas the single-phase samples for the structure analysis of selected solid solutions were prepared by a wetprecipitation method in order to provide good homogeneity of the starting mixtures. The phases in the prepared samples were characterized by X-ray powder diffraction (XRD), scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS). The oxides form seven ternary compounds in the equilibrium state, many solid solutions (which extend across a broad concentration region), and a large, single-phase area based on the CaTiO3 solid solution. The structures of several new phases – solid solutions on the tie lines CaTiO3 –Ca3 La4 Ti3 O15 and La2 TiO5 –Ca3 La4 Ti3 O15 – were determined in detail. © 2015 Elsevier Ltd. All rights reserved. Keywords: Phase relations; CaO–La2 O3 –TiO2 ternary system; Structure determination; X-ray powder diffraction; Wet precipitation synthesis

1. Introduction The tremendous developments in wireless communications and information technology over the past two decades have relied on new ceramic materials that are used as the resonant elements. The most important requirements during the development of new, microwave dielectrics are a high permittivity (k), a high quality factor (Qxf) and a low temperature coefficient of the resonant frequency (τ f ). Thus, some compounds in the La2 O3 –TiO2 system, for example, La2 Ti2 O7 and La4 Ti9 O24 , have been examined as microwave dielectric ceramics [1,2], and their dielectric properties can be additionally tuned by the partial substitution of the La by Nd and/or Sm [3]. Furthermore, the addition of selected oxides to this binary system, for example, Al2 O3 [4], Ga2 O3 [5], CaO [6], causes the

∗

Corresponding author. Tel.: +386 1 477 37 08; fax: +386 1 477 38 75. ˇ E-mail address: [email protected] (S.D. Skapin).

http://dx.doi.org/10.1016/j.jeurceramsoc.2015.03.038 0955-2219/© 2015 Elsevier Ltd. All rights reserved.

formation of an A-site-deficient perovskite La2/3 TiO3 compound. Such a stabilized La2/3 TiO3 compound forms solid solution with the perovskites LaAlO3 and CaTiO3 , according to the formula (1 − x)La2/3 TiO3 − xLaAlO3 , which are stable over the range 0.04 ≤ x ≤ 1 [4], and (1 − x)La2/3 TiO3 − xCaTiO3 , which is stable over the range 0.04 ≤ x ≤ 1 [6,7]. Singlephase ceramics based on these solid solutions exhibit excellent microwave-dielectric properties, which can be easily tuned by changing the La2/3 TiO3 :(LaAlO3 or CaTiO3 ) ratio [4,8]. Moreover, in the ternary CaO–La2 O3 –TiO2 system, several compounds were identified, showing interesting microwave dielectric properties [9,10]. In our investigation we focused on a determination of the high-temperature sub-solidus phase relations in this ternary system. We also performed a structural characterization of some of the solid solutions that are formed. Until now, all the compounds in this ternary system were mostly prepared using a solid-state preparation technique. The advantages of wet-chemical techniques are a better homogeneity, the desired particle size and shape on the nanoscale, less

2802

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

power consumption and a lower cost of preparation. After calcination and sintering, the resulting product is well defined and suitable for a structural characterization. 1.1. Literature survey Phase relations in the La2 O3 –TiO2 binary system are well described [11–15]. The stable compounds are La4 Ti9 O24 , La2 Ti2 O7 , La4 Ti3 O12 and La2 TiO5 . Firing of the composition La2 O3 :3TiO2 in a reducing atmosphere [16], or the addition of a small quantity of 1+ ions, such as Li, Na, K [17], or 2+ or 3+ ions, such as Ca [6], Al [4], Ga [5] or Fe [18], leads to the stabilization of the A-site-deficient perovskite compound La2/3 TiO3 . It is also documented in the literature that a minute amount of impurities, the presence of a liquid phase during sintering, or even a small amount of Ti4+ reduced to Ti3+ , may enable the formation of an otherwise unstable phase in the La2 O3 –TiO2 system [2,19–21]. In the binary system CaO–La2 O3 no compounds that would be stable over 1000 ◦ C have been identified [22]. The phase relations in the system CaO–TiO2 have been studied experimentally and calculated by numerous investigators and their results are somewhat inconsistent [23–30]. According to these results, the existence of a perovskite-type CaTiO3 that is stable above 1900 ◦ C, and a compound with the composition Ca3 Ti2 O7 , was confirmed [23,29,30], while others have also reported a compound with the composition Ca4 Ti3 O10 [24–28,30]. Similarly controversial data were obtained in some investigations dealing with a sub-solidus phase-relations determination in the ternary CaO–TiO2 –M oxide systems. In the case where M = Al [31], M = Sr [32], M = Ba [32], only CaTiO3 was identified, whereas in the system where M = Sn [33,34] the compounds CaTiO3 and Ca3 Ti2 O7 were identified, and in the system where M = In [35] the compounds CaTiO3 and Ca4 Ti3 O10 were identified. In the systems M = Gd [36] and M = Y [37], all three compounds, CaTiO3 , Ca3 Ti2 O7 and Ca4 Ti3 O10 , were confirmed. Moreover, in the system CaO–MgO–TiO2 , Coughanour et al. [33] identified CaTiO3 and Ca3 Ti2 O7 , but Shultz [38] found CaTiO3 , Ca3 Ti2 O7 and Ca4 Ti3 O10 . In the ternary system CaO–La2 O3 –TiO2 Nanot et al. [39] reported a perovskite-related series of compounds with the general formula (CaLa)n Tin O3n+2 , where n = 4.5, 5, and 6 along the compositional line CaTiO3 –La2 Ti2 O7 . Thus, the stable compounds are: with n = 4.5 CaLa8 Ti9 O31 (149), with n = 5 CaLa4 Ti5 O17 (125) and with n = 6 Ca2 La4 Ti6 O20 (226). Additionally, the crystal structures of the compounds 149, 125, and 226 were determined by Demˇsar et al. [9] and Stare et al. [40]. The binary phase diagram of the system was reported by Pivovarova et al. [41], showing the stability of the three compounds 149, 125 and 226 above 1400 ◦ C. On the compositional line CaTiO3 –La4 Ti3 O12 two trigonal compounds with a known structure and the general formula Lan−x Cax Tin−1 O3n were identified [42–44], with n = 5 and x = 1 CaLa4 Ti4 O15 (124) and n = 6 and x = 2 Ca2 La4 Ti5 O18 (225). A phase diagram of the binary system was published by Pivovarova

et al. [19], where the existence and the thermal stability of the compounds 124 and 225 above 1400 ◦ C were confirmed. She reported that the compound 124 dissolves approximately 5 mol % of La4 Ti3 O12 in the temperature range 1200–1400 ◦ C. Saltykova et al. [20] and Pivovarova [45] reported perovskitetype solid solution on the tie line CaTiO3 –Ca3 La4 Ti3 O15 (323) with the formula Ca1−x/2 Lax Ti1−x/2 O3 (at 1200 ◦ C up to the melting point), where the compounds with x < 0.2 were supposed to be cubic and the other ones orthorhombic. The end member of this solid solution, 323, was identified, but the structure was not determined. Later on Vanderah et al. [46] denied the existence of the compound 323 and claimed that CaTiO3 dissolves only up to 38 mol % La2 O3 . The structure of only one composition from this part of the phase diagram, with the formula Ca2/3 La2/3 Ti2/3 O3 , is known [47]. It crystallizes in a monoclinic unit cell in the P21 /n space group. However, this is not in agreement with Pivovarova [45], where powder patterns were indexed in more symmetrical crystal systems. To clarify these disagreements, this paper reports the crystal structures of solid solution over the whole compositional range on the tie line CaTiO3 –323. Furthermore, the solid solubility of 323 in La2 TiO5 was identified and structurally characterized.

2. Experimental 2.1. Synthesis For the structure determination the wet-precipitation method for the synthesis of samples was used since this method enables a better homogeneity and a higher crystallinity compared to the solid-state method. In all the precipitation experiments weighed amounts of Ca(CH3 COO)2 ·H2 O (99%, Alfa Aesar) and La(NO3 )3 ·6H2 O (99.9%, Alfa Aesar) were dissolved in 10 mL of distilled water. Separately, the Ti[OCH(CH3 )2 ]4 (97%, Aldrich Chemical Company Inc.) was pre-diluted in 10 mL of distilled water, acidified with HNO3 (>65%, Aldrich Chemical Company Inc.), to obtain a stable solution. Both solutions were mixed and evaporated at approximately 100 ◦ C, while mixing with a magnetic stirrer. The white solids were additionally heated up to 210 ◦ C for 2 h. The product was crushed and well homogenized in an agate pestle and mortar and calcined at 750 ◦ C in air for 1 h. The so-prepared powder was uniaxially pressed into pellets and finally sintered at 1400 ◦ C from 20 h to 100 h. After the heat treatment the samples were cooled by quenching in air to room temperature. The majority of the samples used for the high-temperature phase-relation determination were prepared by a solid-state reaction method from the oxides TiO2 (Alfa Aesar, 99.8%), La2 O3 (Alfa Aesar, 99.99%) and CaCO3 (Alfa Aesar, 99.5%). Due to the reactivity of the La2 O3 with the environmental moisture and CO2 , the oxide was routinely analyzed before weighing using a firing test at 1300 ◦ C. The starting reagents were mixed in the proper molar ratio and homogenized in a planetary YTZ ball mill for 0.5 h in ethanol media. The dried powders were biaxially pressed into pellets and fired in a tube furnace in air at 1400 ◦ C for 20 h. The air-quenched samples were ground and

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

well homogenized in the ball mill and additionally fired and cooled under the same conditions as described previously. Polished cross-sections of the dense samples were analyzed by scanning electron microscopy (SEM) (Jeol JXA 840A, Japan) and a quantitative analysis of the identified phases was performed with energy-dispersive spectroscopy (EDS) using TRACOR software (Tracor Northern, Model NORAN Series II, USA). The compositions (mol %) and labels of the samples that were structurally characterized are collected in Table 1. 2.2. XRD data collection The phase assembly of the sintered samples was determined by powder X-ray diffractometry, using Cu K␣ radiation and an Endeavor D4, Bruker AXS diffractometer. For the structure determination additional powder patterns were collected by PANalytical X’Pert PRO MPD over a larger 2θ range with a longer dwell time. The data-collection parameters are given in Table 2. 2.3. Structure analysis The powder patterns of (2)–(3) are similar to that of the perovskite CaTiO3 , sample (1) (Fig. 1). They were indexed in the orthorhombic crystal system (space group Pbnm). They could not be indexed in cubic or (other higher symmetries) due to several superstructural reflections – the appearance of new, weak reflections or the splitting of some strong ones (the most visible reflections are marked in Fig. 1 with  and  for the new and split reflections, respectively). The powder patterns of (4)–(6) could not be indexed with an orthorhombic unit cell due to the additional splitting of the strong reflections and the appearance of weak peaks at low 2θ values (the most visible are marked in Fig. 1 with * and ♦ for the new and split reflections, respectively). All the reflections were indexed with a monoclinic unit cell (space group P21 /n, sub-group of Pbnm). The dimensions of the unit-cell edges are very similar to the orthorhombic ones with β slightly deviating from 90◦ (Table 2). Superstructural reflections, orthorhombic and monoclinic, were also present in the powder patterns in the paper of Saltykova [20], but no structure determination was performed. The appearance of the superstructural reflections was, in this case, attributed to a lowering of the cubic symmetry to orthorhombic. The patterns of (7) and (8) (Fig. 2) correspond to solidsolution compositions between La2 TiO5 and the compound 323. They were indexed with an orthorhombic unit cell similar to the unit cell of La2 TiO5 [48]. The structures of the samples were refined using the Rietveld method, incorporated into the TOPAS-Academic V4.1 programme [49]. The unit-cell parameters and the coordinates of the known structures, i.e., CaTiO3 [50] for samples (2)–(3), Ca4/3 Nb2/3 O3 [51] for samples (4)–(6), and La2 TiO5 [48] for samples (7)–(8), were used as the initial models with the La and Ca sharing the same crystallographic sites. In the case of samples (4)–(6) all the Nb sites were replaced by Ti and an appropriate proportion of the Ca on the A site was replaced by La. In the final

2803

refinements the occupancies were set corresponding to the synthetic ratio (details are given in Table 1 and in Section 3). Since all the samples were single phase, the molar ratio among the cations in the reaction mixtures corresponds to the molar ratio of the cations in the obtained compounds. During the structure analysis many trial refinements of the occupancy parameters for the different atoms were applied. In this way, we excluded the distribution of La over the B sites in samples (2)–(6), confirmed the equal distribution of Ca over both A sites in the samples (7) and (8) and also checked that the final occupancies are similar to the values obtained with the trial refinements. At first, the zero error, the scale factor and the background parameters were refined. The background was described with the 5th-order Chebyshev polynomial, and for the profile description the fundamental-parameters approach was used [52]. In the next step of the refinement, the unit-cell parameters and two parameters for the crystallite size and one for the strain were released. After that the atomic positions were refined – the coordinates of the atoms that share the sites were constrained to be equal. Afterwards, the atomic isotropic displacement parameters were turned on. The parameters of the final refinement for the samples (1)–(8) are collected in Table 2. The Rietveld plots of the final refinement for the full 2θ range for samples (1)–(8) are available as supplementary material (Appendix A, Figs. A.1, A.2, and A.3). All the structural details (atomic and other structural parameters) and also the parameters of the data collection and the refinement are also given in CIF files, which have been deposited with the FIZ Karlsruhe Crystal Structure Deposition (CSD) Centre as supplementary material with the deposition numbers 429094, 429095, 429096, 429097, 429098, 429099, 429100, and 429101 for the ceramics (1), (2), (3), (4), (5), (6), (7), and (8), respectively. Copies of the data can be obtained free of charge by contacting [email protected]. 3. Results and discussion In the ternary system CaO–La2 O3 –TiO2 , more than one hundred samples were prepared and analyzed. Based on the XRD and SEM analyses of these samples the sub-solidus phase relations in the system were determined at 1400 ◦ C, as shown in Fig. 3. Some compounds in the system form solid solutions, resulting in the extended two-phase areas and one single-phase area. 3.1. CaO–TiO2 system In order to avoid any possible influence that could result in the formation of some additional phases in the CaO–TiO2 system, the wet-precipitation method for sample preparation was applied in the present investigation along the line CaO–TiO2 . In such a case, due to complete homogenization at the atomic level of the mixed precursors, an additional homogenization/milling was not required – in contrast to the solid-state preparation method. It was assumed that the prolonged milling process and the insufficient purity of the starting chemicals were a possible reason for the variation of the phase relations obtained in the above-cited publications.

2804

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

Table 1 Compositions (mol %) and labels of structurally characterized samples. Sample name

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

CaO La2 O3 TiO2

50.0 0 50.0

47.37 5.26 47.37

46.15 7.69 46.15

42.86 14.28 42.86

40.0 20.0 40.0

37.5 25.0 37.5

4.84 46.77 48.39

9.37 43.75 46.88

1.0 39.3 59.7

3.0 37.5 59.5

7.0 34.3 58.7

10.0 32.0 58.0

Fig. 1. Diffraction patterns of (1)–(6) in the range from 10◦ to 70◦ 2θ and a detailed view of the superstructural and split peaks. The most visible orthorhombic and monoclinic superstructural peaks are marked with  (new orthorhombic),  (split orthorhombic), * (new monoclinic), and ♦ (split monoclinic peaks), respectively.

Fig. 2. Diffraction patterns of La2 TiO5 and samples (7) and (8) from 10◦ to 70◦ 2θ and the detailed positions of the peaks with the corresponding indices in the ranges 40–42◦ and 68.5–70.5◦ 2θ.

Fig. 4 presents XRD powder patterns of the ceramics prepared along the CaO–TiO2 line. The powder pattern of the composition with the molar ratio CaO = 50.0 mol %, TiO2 = 50.0 mol % was a good match with the PDF card of the perovskite CaTiO3

(01-070-7337) and the pattern of the composition with the molar ratio CaO = 60.0 mol %, TiO2 = 40.0 mol % (corresponding to Ca3 Ti2 O7 ) agreed with the PDF card of Ca3 Ti2 O7 (00-0891384). Thus, we confirmed that the compound Ca3 Ti2 O7 is

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

2805

Table 2 Crystal data, data collection and refinement data for samples (1)–(8).

Crystal data Chemical formula Mr Crystal system, space group Temperature (K) ˚ a, b, c (A) ˚ 3) V (A ρ (g cm−3 ) Z Radiation type ˚ Wavelength of incident radiation (A) μ (mm−1 ) Specimen colour Data collection Diffractometer Specimen mounting Scan method Data-collection mode 2θ values (◦ ) Refinement Profile function Rp (Rp ), Rwp (Rwp ), Rexp (Rexp ), RBragg , χ2 =

Excluded regions No. of data points No. of structural parameters No. of profile parametersa No. of restraints

Crystal data Chemical formula Mr Crystal system, space group Temperature (K) ˚ a, b, c (A) β (◦ ) V (Å3 ) ρ (g cm−3 ) Z Radiation type ˚ Wavelength of incident radiation (A) μ (mm−1 ) Specimen colour Data collection Diffractometer Specimen mounting Scan method Data-collection mode 2θ values (◦ )

(1)

(2)

(3)

CaTiO3 135.95 Orthorhombic, Pbnm 298 5.38445(10), 5.44067(11), 7.64786(16) 224.044(8) 4.030 4 Cu K␣1 1.5406 50.190 White yellowish powder

Ca0.901 La0.198 Ti0.901 O3 154.76 Orthorhombic, Pbnm 298 5.45535(5), 5.51959(5), 7.75023(8) 233.369(4) 4.410 4 Cu K␣1 1.5406 71.049 White yellowish powder

Ca0.857 La0.286 Ti0.857 O3 163.07 Orthorhombic, Pbnm 298 5.48206(4), 5.55097(4), 7.79093(6) 237.084(3) 4.569 4 Cu K␣1 1.5406 79.614 White yellowish powder

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

Fundamental parameters approach 0.08705 (0.16136), 0.11543 (0.17537), 0.04622 (0.07022), 0.06499, 2.498 None 3333 13 11 0

Fundamental parameters approach 0.05413 (0.12670), 0.07189 (0.12463), 0.03647 (0.06323), 0.03139, 1.971 None 3333 13 11 0

Fundamental parameters approach 0.04619 (0.11656), 0.05993 (0.10873), 0.03618 (0.06564), 0.03016, 1.656 None 3333 13 11 0

(4)

(5)

(6)

Ca0.75 La0.5 Ti0.75 O3 183.41 Monoclinic, P21 /n 298 5.55914(5), 5.65097(5), 7.92215(8) 90.0785(10) 248.870(4) 4.895 4 Cu K␣1 1.5406 98.785 White yellowish powder

Ca0.667 La0.667 Ti0.667 O3 199.23 Monoclinic, P21 /n 298 5.60004(6), 5.72706(6), 7.99995(8) 90.1665(7) 256.572(5) 5.159 4 Cu K␣1 1.5406 113.186 White yellowish powder

Ca0.6 La0.8 Ti0.6 O3 211.89 Monoclinic, P21 /n 298 5.63262(5), 5.78498(5), 8.05366(7) 90.1749(6) 262.424(4) 5.363 4 Cu K␣1 1.5406 124.184 White yellowish powder

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

2806

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

Table 2 (Continued).

Refinement Profile function Rp (Rp ), Rwp (Rwp ), Rexp (Rexp ), RBragg , χ2 =

Excluded regions No. of data points No. of structural parameters No. of profile parametersa No. of restraints

Crystal data Chemical formula Mr Crystal system, space group Temperature (K) ˚ a, b, c (A) V (Å3 ) ρ (g cm−3 ) Z Radiation type ˚ Wavelength of incident radiation (A) μ (mm−1 ) Specimen colour Data collection Diffractometer Specimen mounting Scan method Data-collection mode 2θ values (◦ ) Refinement Profile function Rp (Rp ), Rwp (Rwp ), Rexp (Rexp ), RBragg , χ2 =

Excluded regions No. of data points No. of structural parameters No. of profile parametersa No. of restraints a

(4)

(5)

(6)

Fundamental parameters approach 0.04620 (0.12427), 0.06253 (0.12042), 0.03675 (0.07076), 0.02320, 1.702 None 3333 19 11 0

Fundamental parameters approach 0.05715 (0.14870), 0.07647 (0.14666), 0.03700 (0.07097), 0.03798, 2.067 None 3333 19 11 0

Fundamental parameters approach 0.06590 (0.17668), 0.09184 (0.17945), 0.04060 (0.07933), 0.07418, 2.262 None 3333 19 11 0

(7)

(8)

Ca0.0984 La1.9016 Ti0.9836 O4.918 393.86 Orthorhombic, Pnam 298 11.00896(9), 11.36347(9), 3.94606(3) 493.652(7) 5.301 4 Cu K␣1 1.5406 138.542 White yellowish powder

Ca0.1935 La1.8065 Ti0.9677 O4.8385 382.46 Orthorhombic, Pnam 298 11.00664(11), 11.33061(10), 3.94583(4) 492.092(8) 5.162

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

PANalytical X’Pert PRO MPD Flat plate Continuous Reflection, θ–2θ 2θ min = 10, 2θ max = 120, 2θ step = 0.033

Fundamental parameters approach 0.06387 (0.14159), 0.08903 (0.15994), 0.03632 (0.06525), 0.05095, 2.451 None 3333 22 11 0

Fundamental parameters approach 0.06183 (0.14082), 0.08458 (0.15621), 0.03497 (0.06459), 0.04163, 2.418 None 3333 22 11 0

Cu K␣1 1.5406 133.371 White yellowish powder

Profile parameters: 6 background (Chebyshev 5th order), 1 zero correction, 2 crystallite size L, G (nm), 1 strain L, 1 scale factor.

stable in the CaO–TiO2 system at 1400 ◦ C. On the other hand, the starting composition corresponding to Ca4 Ti3 O10 (molar ratio CaO = 57.15 mol %, TiO2 = 42.85 mol %) did not result in Ca4 Ti3 O10 – the powder pattern of that composition showed that the product was a mixture of CaTiO3 and Ca3 Ti2 O7 , regardless of the synthesis route, i.e., a solid-state or a wet-precipitation technique. Even after 100 h of firing at 1400 ◦ C only the mixture of CaTiO3 and Ca3 Ti2 O7 was identified (Fig. 4). From this it could be deduced that the compound Ca4 Ti3 O10 does not exist in the binary join CaO–TiO2 at 1400 ◦ C. Also, no solid solubility of the oxides CaO and TiO2 in CaTiO3 was detected at 1400 ◦ C. The composition with the molar ratio CaO:TiO2 = 0.530:0.470

resulted in a mixture of CaTiO3 and TiO2 , and that with the ratio 0.470:0.530 in a mixture of CaTiO3 and Ca3 Ti2 O7 , as shown in Fig. 4. 3.2. Stabilization of the La2/3 TiO3 phase The addition of 4 mol % of CaTiO3 was necessary for complete stabilization of the unstable stoichiometric La2/3 TiO3 compound, which is in agreement with the published results [6]. For a smaller CaTiO3 addition, the secondary phases La2 Ti2 O7 and La4 Ti9 O24 are present along with the stabilized La2/3 TiO3 (Fig. 7a). Such stabilized La2/3 TiO3 forms a perovskite-type

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

2807

in preparing it and determining its structure. The CaTiO3 and the compound 323 form a solid solution over the whole concentration range. A micrograph of a polished single-phase sample lying on the tie line is shown in Fig. 7c. In contrast, no solid solubility of La2 O3 in the compound 323 was detected.

Fig. 3. Sub-solidus phase relations in the ternary system CaO–La2 O3 –TiO2 in air, at 1400 ◦ C (L2 T9 : La4 Ti9 O24 , ‘LT3 ’: La2/3 TiO3(STAB) , LT2 : La2 Ti2 O7 , L2 T3 : La4 Ti3 O12 , LT: La2 TiO5 , 149: CaLa8 Ti9 O31 , 125: CaLa4 Ti5 O17 , 226: Ca2 La4 Ti6 O20 , 124: CaLa4 Ti4 O15 , 225: Ca2 La4 Ti5 O18 and 323: Ca3 La4 Ti3 O15 ).

solid solution with CaTiO3 up to 100 mol %. The solid solution CaTiO3 –La2/3 TiO3(STAB) in whole the concentration range is compatible with TiO2 ; both phases can be seen in the microstructure in Fig. 7b. 3.3. CaTiO3 –La2 O3 sub-system In spite of doubts in the literature [46] about the existence of the compound 323 on the CaTiO3 –La2 O3 line, we succeeded

3.3.1. Structures of solid solution on the tie line CaTiO3 –323 As was explained in Section 2, all the solid solution Ca1−x/2 Lax Ti1−x/2 O3 has a slightly distorted perovskite structure. Compositions with x < 0.500 are orthorhombic and those with x > 0.500 are monoclinic. The superstructural reflections, orthorhombic and monoclinic, were also present in the powder patterns in the paper of Saltykova [20], but the structure determination was not performed. The appearance of superstructural reflections in this case was attributed to the lowering of the cubic symmetry to the orthorhombic. The lattice parameters a, b, and c, and the unit-cell volume, of the ceramics (1)–(6) increase with an increasing amount of La2 O3 incorporated into the CaTiO3 , obeying Vegard’s rule, which is presented in Fig. 5. This is in accordance with the slightly larger ionic radius of La3+ (for the coordination ten 1.27 ˚ in comparison with Ca2+ (1.23 and 1.12 A, ˚ and for eight 1.16 A) respectively) [53]. In (2) and (3) the A site, lying on the mirror plane, is 20.0% and 28.6%, occupied with La3+ in (2) and (3), respectively. The rest of the A sites belong to Ca2+ . This means that all the

Fig. 4. X-ray powder-diffraction patterns of the compositions along the CaO–TiO2 line, with the ratio CaO:TiO2 = 0.470:0.530, CaO:TiO2 = 0.500:0.500 (corresponds to CaTiO3 , sample (1)), CaO:TiO2 = 0.530:0.470, CaO:TiO2 = 0.5715:0.4285 (corresponds to Ca4 Ti3 O10 ), and CaO:TiO2 = 0.600:0.400 (corresponds to Ca3 Ti2 O7 ), fired at 1400 ◦ C for 100 h.

2808

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

Fig. 5. The graph of lattice parameters a, b, and c, and the unit-cell volume, versus the molar ratio x (n(La2 O3 )/(3(CaTiO3 ) + n(La2 O3 )) of the ceramics (1)–(6). The red circles, magenta rhombuses, cyan squares and blue triangles represent the data points of the unit-cell parameters a, b, and c, and the volume, respectively. The solid black lines represent the linear trend lines of each data set. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Table 3 Occupancies (%) of the A and B sites of the ABO3 perovskite structures of samples (1)–(6). Site

(1) (2) (3) (4) (5) (6)

A

B (orthorhombic)

Ca

La

Ca

Ti

100 80.0 71.4 50.0 33.3 20.0

20.0 28.6 50.0 66.7 80.0

10.0 14.3

100 90.0 85.7

lanthanum is positioned on the A site, while the remaining calcium is randomly distributed over the six-coordinated B site, located at the centre of the inversion, together with Ti4+ . All the sites, including the oxygen atoms (O1 on a mirror plane and O2 at a general position), are fully occupied. The occupancies of the perovskite A and B sites are collected in Table 3. As shown in Fig. 6 the (Ca/Ti)O6 octahedra are

B (monoclinic, 2c)

B (monoclinic, 2d)

Ca

Ti

Ti

50.0 66.7 80.0

50.0 33.3 20.0

100 100 100

slightly distorted and tilted in comparison with the ideal BX6 octahedra in the cubic perovskite. The bond lengths Ti1/Ca2 O, presented in Table 4, increase with the increasing content of lanthanum in the sample and, consequently, with the increase in the occupancy of calcium on the B site, which is in agreement with the larger ionic radius of the Ca2+ in comparison with the Ti4+ ˚ respectively) [53]. (for the octahedral sites 1.00 and 0.605 A,

Table 4 ˚ in samples (1)–(6). Bonding and some contact distances between the metal ions and the oxygen (A) Sample

La/Ca1 O

La/Ca1 Oa

Ti1/Ca2 O

Ti2 O

CaTiO3 [50] (1) CaTiO3 (2) (3) (4) (5) (6) 323

2.360(4)–2.660(3) 2.360(5)–2.686(5) 2.368(5)–2.749(4) 2.374(5)–2.763(4) 2.376(6)–2.826(14) 2.387(9)–2.804(12) 2.358(10)–2.906(14)

3.04(4)–3.05(5) 3.018(6)–3.046(7) 3.092(5)–3.120(5) 3.114(4)–3.142(5) 3.219(6)–3.229(6) 3.288(9)–3.351(8) 3.386(10)–3.501(11)

1.950(5)–1.961(4) 1.938(4)–1.977(4) 1.9846(11)–1.991(4) 1.992(3)–2.008(3) 2.058(14)–2.120(13) 2.166(12)–2.192(12) 2.172(13)–2.290(14)

1.976(13)–2.023(14) 1.949(12)–1.973(12) 1.935(14)–2.055(13)

a

˚ First two oxygen atoms with distance La/Ca1 O > 3 A.

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

2809

Fig. 6. Structure drawings of (3) – left – and (6) – right. The orange polyhedra represent the (Ca/Ti)O6 octahedra and magenta, the TiO6 octahedra, respectively; circles – O* represents O1 in (3) and O3 in (6), respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Due to the tilting of the (Ca/Ti)O6 octahedra, the coordination number of the A-site cations is lowered from 12 to 8+2 in the shape of an irregular polyhedron. The solid solution with compositions (4)–(6) has a lower, monoclinic, P21 /n symmetry. In this symmetry there are two B (octahedral) sites, i.e., Wyckoff position 2c (B ) and 2d (B ), respectively. With the rise of the lanthanum content on the A site (in this symmetry, at a general position), the amount of calcium also increases, which needs to be distributed over the octahedral B sites (see Table 3). Only bond lengths around the Wyckoff position 2c show an obvious trend of enlargement with an increase of the lanthanum content, indicating that this site is randomly occupied by both titanium and calcium ions, while the Wyckoff position 2d is only occupied by titanium ions. In both cases the BO6 octahedra are slightly distorted and centrosymmetric, and as in all perovskites they share all six vertices with the neighbouring octahedra. From Fig. 6 it is clear that the (Ca/Ti)O6 octahedra are larger in comparison with TiO6 and along the [1 1 1] direction there are alternating planes of TiO6 and (Ca/Ti)O6 octahedra, respectively. In the monoclinic phases the tilting angles are larger when compared to those in the orthorhombic phases, which is reflected in an increase of the La/Ca1 Oa distances (see Table 4). For this reason the coordination number of the La3+ and the Ca2+ on the A site is only 8 (the next two oxygen atoms belong to the second coordination sphere). The La/Ca1 O bond lengths from the first coordination sphere show a trend of increasing with an increasing content of La, which is in accordance with the slightly larger ionic radius ˚ (for of the La3+ in comparison with the Ca2+ (1.16 and 1.12 A coordination 8), respectively) [53].

Such a partial ordering of the calcium ions, reflected in a reduction of the symmetry from the orthorhombic to the monoclinic, was also observed by Aatiq in Ca2/3 La2/3 Ti2/3 O3 [47] and Levin in Ca4/3 Nb2/3 O3 [51]. It is true that the Ca2+ ions in most of the perovskite structures occupy the A site. However, there are also other known structures where it is distributed over the B site, which is usually reserved for smaller cations: Ca2 CaWO6 [54], Sr2 CaWO6 [55], Sr2 CaMoO6 [56], Ca2 CaUO6 [57] with a monoclinic P21 /n symmetry. Also in the orthorhombic modifications of Sr2 CaWO6 [58] and Sr2 CaMoO6 [59] the calcium is distributed on the B octahedral site. In all these cases – orthorhombicand monoclinic – the approximate  unit-cell parameters were 2ap , 2ap and 2ap (ap refers to the cubic perovskite) and the angles were equal or close to 90◦ . As was mentioned in Section 2, using trial refinements we also checked the possibility that some lanthanum was distributed over the B sites. However, due to the significantly different scattering factors of lanthanum in comparison to those of other atoms, the Rietveld refinement was sensitive to the positions and occupancies of the La atoms. Trial refinements for samples (2)–(6) showed that the refined occupancies of La on the B sites were very close to zero. Furthermore, if we fixed higher values for the occupancies of La on the B sites (similar to those of Ca) the agreement between the calculated and the observed diffraction patterns was significantly worse in comparison to the situation with no La on the B sites. It can be concluded that although the difference between the ionic radii of La3+ and Ca2+ is small, it is large enough so that only the smaller Ca2+ ions are distributed over the B sites (together with Ti4+ ).

2810

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

Fig. 7. SEM micrographs of the polished cross-sections of the selected samples prepared in the system La2 O3 –TiO2 –CaO: (a) the sample with the starting composition 24.3 mol % La2 O3 , 74.3 mol % TiO2 and 1.4 mol % CaO is composed of a matrix grey phase, stabilized La2/3 TiO3 (‘LT3 ’), dark phase La4 Ti9 O24 (L2 T9 ) and bright phase, La2 Ti2 O7 (LT2 ); (b) the sample with the starting composition 10.0 mol % La2 O3 , 85.0 mol % TiO2 and 5.0 mol % CaO is composed of a matrix bright phase, CaTiO3 –La2/3 TiO3(STAB) solid solution ((CT–‘LT3 ’)ss ) and dark phase TiO2 ; (c) the sample with the starting composition 7.7 mol % La2 O3 , 46.15 mol % TiO2 and 46.15 mol % CaO shows a single-phase microstructure of the CaTiO3 –323 solid solution; (d) the sample with the starting composition 30.0 mol % La2 O3 , 30.0 mol % TiO2 and 40.0 mol % CaO, contains a matrix phase, 323, dark isolated grains of CaO and a bright phase La2 O3 ; (e) the sample with the starting composition 20.0 mol % La2 O3 , 45.0 mol % TiO2 and 35.0 mol % CaO is composed of two solid solutions, the matrix grey phase, CaTiO3 solid solution (CTss ) and light phase La2 TiO5 solid solution (LTss ); (f) the sample with the starting composition 21.0 mol % La2 O3 , 66.0 mol % TiO2 and 13.0 mol % CaO has a two-phase microstructure, the dark phase, CaTiO3 –La2/3 TiO3(STAB) solid solution ((CT–‘LT3 ’)ss ) and bright phase 149 compound.

3.4. CaTiO3 -based single-phase region The solid solution along the CaTiO3 –323 and CaTiO3 –La2/3 TiO3(STAB) lines circumscribes a nearly triangular, single-phase region based on CaTiO3 , with the corner phases CaTiO3 , 323 and the CaTiO3 –La2/3 TiO3(STAB) solid solution with the composition La2/3 TiO3 :CaTiO3 = 40.0:60.0 (corresponding to: La2 O3 = 7.7 mol %, TiO2 = 57.7 mol %, and CaO = 34.6 mol %). The extension of the solid solubility of the lanthanum titanates into the CaTiO3 was determined by the systematic preparation of the samples along the lines CaTiO3 –La2 TiO5 , CaTiO3 –La4 Ti4 O12 and CaTiO3 –La2 Ti2 O7 , monitoring the appearance of the secondary phase using XRD and SEM (EDS). Thus, it was found that the CaTiO3 formed a solid solution with:

(1) La2 TiO5 ; up to the composition CaTiO3 :La2 TiO5 = 74.0:26.0 (corresponds to: La2 O3 = 13.0 mol %, TiO2 = 50.0 mol % and CaO = 37.0 mol %). (2) La4 Ti3 O12 ; up to the composition CaTiO3 :La4 Ti3 O12 = 85.0:15.0 (corresponds to: La2 O3 = 12.2 mol %, TiO2 = 53.1 mol % and CaO = 34.7 mol %). (3) La2 Ti2 O7 ; up to the composition CaTiO3 :La2 Ti2 O7 = 80.0:20.0 (corresponds to: La2 O3 = 9.1 mol %, TiO2 = 54.5 mol % and CaO = 36.4 mol %). 3.5. Phase relations in the ternary CaO–La2 O3 –TiO2 system The sub-solidus phase relations in the system are illustrated in Fig. 3. The compound Ca3 Ti2 O7 is introduced in the binary

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

2811

sub-system CaTiO3 –CaO; however, this compound was not identified in a single sample, prepared in the ternary sub-system CaTiO3 –CaO–323, even after 100 h of heat treatment at 1400 ◦ C with intermittent crushing and homogenization (neither by solidstate nor wet-precipitation synthesis). The compound 323 is in equilibrium with the La2 O3 and CaO, which is illustrated by the microstructure of the sample containing all three phases (Fig. 7d). (1) On the tie line 323–La2 TiO5 , the La2 TiO5 forms a solid solution with 323 in a limited compositional range up to the composition CaLa4/3 TiO5 :La2 TiO5 = 20:80 (corresponds to: La2 O3 = 43.7 mol %, TiO2 = 46.9 mol % and CaO = 9.4 mol %). This composition with a saturated solubility is additionally connected with the end points of the CaTiO3 - and La4 Ti3 O12 -based solid solutions, La2 O3 and 124. The SEM micrograph in Fig. 7e shows the two-phase microstructure of the sample that is composed of La2 TiO5ss and CaTiO3ss . 3.5.1. Structures of solid solution on the tie line 323–La2 TiO5 Structure determination using powder XRD showed that the solid solution with compositions CaLa4/3 TiO5 : La2 TiO5 = 10.0:90.0 (7) and 20.0:80.0 (8) is isostructural with La2 TiO5 ; it crystallizes in the orthorhombic Pnam space group [48]. In La2 TiO5 Ti4+ is coordinated by five O2− atoms. The coordination polyhedron can be described as a distorted trigonal bi-pyramid where the basal plane of the bi-pyramid, consisting of a Ti and three O atoms, coincides with the crystallographic mirror plane. Through the apical O atoms lying on the neighbouring mirror planes, the bi-pyramids are connected into chains running parallel with the c axis (see Fig. 8). The lanthanum is partitioned over two crystallographically independent sites, both located on mirror planes. The La3+ on both sites is 7 coordinated. In the solid solution the Ca atoms are equally distributed over both La sites in such a way that these two sites are fully occupied, i.e., 4.92% and 9.68% Ca and 95.08% and 90.32% La for the samples (7) and (8), respectively. The bond lengths Ti1 O, collected in Table 5, do not increase in the solid solution, which shows that the Ca atoms do not occupy the Ti sites, as observed in the solid solution corresponding to the samples (2)–(6). The Ti and O sites are occupied by 98.36% and 96.77% for the samples (7) and (8), respectively. With the increasing calcium content in the ceramics, the volume of the unit cell decreases, which is in agreement with the smaller ionic radius of the Ca2+ in comparison with the La3+ [53]. In solid solution only the b edge of the unit cell decreases, while the unit-cell parameters a and c remain almost constant. In accordance with this, it is clear from Fig. 2 that the 2θ angle of the peaks with the indices hkl, where k is large and h and l are small (for example 150, 162), decreases with an increasing amount of the compound 323 in the solid solution (7) and (8), while the

Fig. 8. Drawing of the structure of (8). The orange polyhedra represents the TiO5 bi-pyramids, the large circles are the La/Ca atoms and the small circles are the O1 atom, surrounded only by La.

position of the peaks hkl, where k is small (for example 411), remain almost constant. (2) On the tie line CaTiO3 –La4 Ti3 O12 the compounds 124 and 225 are stable. The solid solubility of the 124 in La4 Ti3 O12 was identified in the concentration range up to the composition CaLa4 Ti4 O15 :La4 Ti3 O12 = 37:63 (corresponds to: La2 O3 = 34.3 mol %, TiO2 = 58.7 mol % and CaO = 7 mol %). This composition with saturated solubility is additionally connected with the end point of the La2 TiO5 -based solid solution, La2 Ti2 O7 and 149. The compounds 225 and 124 are additionally compatible with: (i) 225 is in equilibrium with the CaTiO3 -based solid solution, thus forming a two-phase area, and with the compounds 226 and 125. (ii) 124 is connected with tie lines with solid solutions based on CaTiO3 and La2 TiO5 , with saturated solubility and the compounds 149. Several compositions (9)–(12) of solid solution between La4 Ti3 O12 and the compound 124 were prepared (see Table 1) and their X-ray diffraction patterns (Fig. 9) were collected in order to determine their structure. The powder patterns of the ceramics with an increasing amount of calcium, (9)–(11) are very similar (Fig. 9). On the basis of this similarity we were able to label the indices of the reflections of the solid solution with the corresponding indices of La4 Ti3 O12 [13]. Nevertheless, these indices are not suitable for the proper indexing of the solid solution. The reflections hk0, for example 110 and 300,

2812

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

Table 5 ˚ for samples (7) and (8). Bonding distances between the metal ions and the oxygen (A) Sample

La1/Ca1 O

La2/Ca2 O

Ti1 O

La2 TiO5 [48] (7) (8)

2.439(12)–2.589(13) 2.449(9)–2.579(15) 2.445(9)–2.616(15)

2.412(13)–2.549(12) 2.381(15)–2.717(15) 2.387(15)–2.678(15)

1.789(13)–2.033(3) 1.784(15)–2.014(3) 1.772(15)–2.013(3)

Fig. 9. XRD powder patterns of the solid solution (9)–(12) in comparison with the XRD patterns of the end member La4 Ti3 O12 [13] and the compound 124 [44]. The peaks, corresponding to the indices 101, 102, 104, 105 and 107 in La4 Ti3 O12 , are marked with ; the peaks 110, 330, which remain at almost the same angles, with ; and the diffraction peaks of the compound 124 are marked with 䊉.

have in (9)–(11) very similar diffraction angles and can also be indexed without problems in the solid solution, indicating that the lattice parameters a and b in the solid solution are similar to those of La4 Ti3 O12 . On the other hand, the 2θ angles of the appropriate reflections with l = / 0 in (9)–(11) differ and the difference becomes greater with the increase of the 2θ angle, indicating changes in the lattice parameter c. The problem is that the diffraction angles of some reflections, for example, 102, 104, and 107, increase, and some other, for example 101 and 105, decrease with an increasing amount of incorporated calcium. Consequently, it was not possible to find a reasonable unit cell, neither in the hexagonal nor in the lower symmetries. The reason is very probably that these solids are incommensurate modulated phases, as have already been observed for the compound La4 Ti3 O12 . [21]. For this reason the structure determination of the solid solution (9)–(11) is still under investigation and is beyond the scope of this paper. The XRD powder pattern of composition (12) includes reflections of (11) and the compound 124, thus it can be concluded that composition (11) is the

end member of the solid solution, with the highest content of incorporated calcium. (3) The ternary compounds 226, 125, and 149 are formed in the binary system CaTiO3 –La2 Ti2 O7 . They are compatible with the CaTiO3 –La2/3 TiO3(STAB) solid solution with a varying Ca/La ratio. The extension of the compatibility triangles was determined by a combination of XRD and a detailed microanalysis of the CaTiO3 –La2/3 TiO3(STAB) solid solution in the corresponding samples. Thus: (i) 226 is compatible with the CaTiO3 –La2/3 TiO3(STAB) solid solution within the concentration range 60:40 and 31:69 of the CaTiO3 :La2/3 TiO3 ratio. (ii) 125 is compatible with the CaTiO3 –La2/3 TiO3(STAB) solid solution at a single point for the concentration CaTiO3 :La2/3 TiO3 = 31:69.

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

(iii) 149 is compatible with the CaTiO3 –La2/3 TiO3(STAB) solid solution within the concentration range 31:69 and 8:92 of the CaTiO3 :La2/3 TiO3 ratio. Fig. 7f presents the twophase microstructure of the sample containing 149 and the CaTiO3 –La2/3 TiO3(STAB) solid solution. 4. Conclusions Sub-solidus compatibility relations in the ternary CaO–La2 O3 –TiO2 system at 1400 ◦ C were determined and a phase diagram was constructed. In the binary CaO–TiO2 system the compound Ca3 Ti2 O7 is stable along with CaTiO3 ; however, the proposed Ca4 Ti3 O10 compound does not form up to 1400 ◦ C. In contrast, the Ca3 Ti2 O7 phase was not identified in the ternary system, even after a prolonged firing time up to 100 h. In the system, a remarkable single-phase region exists, due to an extensive solid solubility of lanthanum titanates and the 323 compound in the perovskite CaTiO3 . The powder XRD structure analysis revealed that 323 and the compositions Ca1−x/2 Lax Ti1−x/2 O3 with x > 0.5 have a monoclinic perovskite structure and those with x < 0.5 an orthorhombic perovskite structure. The solid solution of La2 TiO5 and 323 are not perovskites and are isostructural with La2 TiO5 . Acknowledgements The financial support of the Ministry of Education, Science and Sport of the Republic of Slovenia is gratefully acknowledged (grants P2-0091, P1-0175, MR-33158). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jeurceramsoc. 2015.03.038. References [1] Fuierer PA, Newnham RE. La2 Ti2 O7 ceramics. J Am Ceram Soc 1991;74:2876–81. [2] Kimura M, Nanamatsu S, Doi K, Matsushita S, Takahasi M. Electrooptic and piezoelectric properties of La2 Ti2 O7 single crystal. J Appl Phys 1972;11:904. [3] Takahashi J, Kageyama K, Kodaira K. Microwave dielectric properties of lanthanide titanate ceramics. Jpn J Appl Phys 1993;32:4327–31. ˇ [4] Skapin SD, Kolar D, Suvorov D. X-ray diffraction and microstructural investigation of the Al2 O3 –La2 O3 –TiO2 system. J Am Ceram Soc 1993;76:2359–62. ˇ [5] Kolar D, Skapin SD, Suvorov D, Valant M. Phase equilibria and dielectric properties in the La2 O3 –Ga2 O3 –TiO2 system. High Temperature Mat Chem 1997;IX:109–15. [6] Kim IS, Jung WH, Inaguma Y, Nakamura T, Itoh M. Dielectric properties of A-site deficient perovskite-type lanthanum–calcium–titanium oxide solid solution system [(1 − x)La2/3 TiO3 − xCaTiO3 (0.1 < x < 0.96)]. Mater Res Bull 1995;30:307–16. [7] Vasylechko L, Vashook V, Guth U. (CaTiO3 )–(LaCrO3 )–(CaCrO3 ) and (CaTiO3 )–(LaCrO3 )–(La2/3 TiO3 ) quasi-ternary systems. In: Sammes N, Smirnova A, Vasylyev O, editors. Full Cell Technologies: State and Perspectives. Dordrecht: Springer; 2005. p. 373–80.

2813

ˇ [8] Suvorov D, Valant M, Skapin SD, Kolar D. Microwave dielectric properties of ceramics with compositions along the La2/3 TiO3(STAB) –LaAlO3 tie line. J Mater Sci 1998;33:85–9. ˇ [9] Demˇsar K, Skapin SD, Meden A, Suvorov D. Rietveld refinement and dielectric properties of CaLa4 Ti4 O17 and SrLa4 Ti5 O17 ceramics. Acta Chim Slov 2008;55:966–72. [10] Iqbal Y, Manan A, Reaney IM. Low loss Sr1−x Cax La4 Ti5 O17 microwave dielectric ceramics. Mater Res Bull 2011;46:1092–6. [11] MacChesney JB, Sauer HA. The system La2 O3 –TiO2 : phase equilibria and electrical properties. J Am Ceram Soc 1962;45:416–22. ˇ [12] Skapin SD, Kolar D, Suvorov D. Phase stability and equilibria in the La2 O3 –TiO2 system. J Eur Ceram Soc 2000;20:1179–85. [13] Konstantinov P, Krezhov K, Svab E, Meszaros G, Török G. Neutron powder diffraction refinement of the crystal structure of La4 Ti3 O12 . Phys B 2000;276–278:260–1. [14] Bontchev R, Darriet B, Darriet J, Weill F, Van Tendeloo G, Amelinckx S. New cation deficient perovskite-like oxides in the system La4 Ti3 O12 –LaTiO3 . Eur J Sol State Inorg Chem 1993;30:521–37. [15] Teneze N, Mercurio D, Trolliard G, Frit B. Cation-deficient perovskiterelated compounds (Ba,La)n Tin−1 O3n (n = 4, 5, and 6): a Rietveld refinement from neutron powder diffraction data. Mater Res Bull 2000;35:1603–14. [16] Abe M, Uchino K. X-ray study of the deficient perovskite La2/3 TiO3 . Mater Res Bull 1974;9:147–56. [17] Belous AG, Gavrilova LG, Makarova ZR, Caljic VP. Stabilization of lanthanum titanate perovskite structure. Ukr Khim Zh 1984;50: 460–1. ˇ ˇ [18] Skapin SD, Sever Skapin A, Suvorov D, Gaberˇscˇ ek M. A stabilization mechanism for the perovskite La2/3 TiO3 compound with Fe2 O3 : a structural and electrical investigation. J Eur Ceram Soc 2008;28: 2025–32. [19] Pivovarova AP, Strakhov VI, Saltykova VA. Phase relations in the CaTiO3 –La4 Ti3 O12 system. Inorg Mater 1999;35:1288–90. [20] Saltykova VA, Mel’nikova OV, Fedorov NF, Polevaya IA. Phase-equilibria in the CaTiO3 –La2 O3 system. Zh Neorg Khim 1989;34:2659–63. [21] Elcoro L, Perez-Mato JM, Withers R. Intergrowth polytypoids as modulated structures: the example of the cation deficient oxides LaTi1−x O3 . Z Kristallogr 2000;215:727–39. [22] Lopato LM, Lugin LI, Shevchenko AV, Gerasimyuk GI. Phase ratios in systems La2 O3 –CaO and Ce2 O3 –CaO. Ukr Khim Zh 1972;38: 143–6. [23] DeVries RC, Roy R, Osborn EF. Phase equilibria in the system CaO–TiO2 . J Phys Chem 1954;58:1069–73. [24] Roth RS. Revision of the phase equilibrium diagram of the binary system calcia–titania showing the compound Ca4 Ti3 O10 . J Res Natl Bur Stand (US) 1958;61:437–40. [25] Jongejan A, Wilkins AL. A re-examination of the system CaO–TiO2 at liquidus temperatures. J Less-Common Met 1970;20:273–9. [26] Tulgar HE. Solid state relationships in the system CaO–TiO2 . Bull Tech Univ Istanb 1976;29:111–29. [27] Kaufman L. Calculation of multicomponent ceramic phase diagrams. Phys B 1988;150:99–114. [28] DeCapitani C, Kirschen M. A generalized multicomponent excess function with application to immiscible liquids in the system CaO–SiO2 –TiO2 . Geochim Cosmochim Acta 1998;62:3753–63. [29] Danek V, Nerad I. Phase diagram and structure of melts of the system CaO–TiO2 –SiO2 . Chem Pap 2002;58:241–6. [30] Elcombe MM, Kisi EH, Hawkins KD, White TJ, Goodman P, Matheson S. Structure determinations for Ca3 Ti2 O7 , Ca4 Ti3 O10 , Ca3.6 Sr0.4 Ti3 O10 and a refinement of Sr3 Ti2 O7 . Acta Crystallogr B 1991;47:305–14. [31] Morgan PED, Koutsoutis MS. Phase relations in the Ca–Ti–Al–O system; further studies on members of the CTA family. J Mater Sci Lett 1985;4:321–3. [32] Kwestroo W, Paping HA. The systems BaO–SrO–TiO2 , BaO–CaO–TiO2 , and SrO–CaO–TiO2 . J Am Ceram Soc 1959;42:292–9. [33] Coughanour LW, Roth RS, Marzulo S, Sennett FE. Solid-state reactions and dielectric properties in the system magnesia–lime–tin oxide–titania. J Res Natl Bur Stand (US) 1955;54:149–62.

2814

M. Vidmar et al. / Journal of the European Ceramic Society 35 (2015) 2801–2814

[34] Dietzel A, Posch W. Über das Verhalten von SnO2 bei hohen Temperaturen und über einige System mit SnO2 . Radex Rundsch 1960;1:52–61. [35] Li ZF, Sun JL, You LP, Jiao H, Li GB, Liao XP, et al. Phase equilibrium of the In2 O3 –TiO2 –MO (M = Ca, Sr) systems and the structure of In6 Ti6 CaO22 . Chem Mater 2005;17:2186–92. [36] Kovba LM, Herman M. The possible types of defects in solid-solutions on the basis of perovskites. Vestn Mosk Univ Ser 2 Khim 1986;27:190–5. [37] German M, Kovba LM. X-ray crystallographic investigation of the CaO–Y2 O3 –TiO2 system. Inorg Mater 1983;19:1712–3. [38] Shultz RL. Effects of titanium oxide on equilibria among refractory phases in system CaO–MgO–iron oxide. J Am Ceram Soc 1973;56:33–6. [39] Nanot M, Queyroux F, Gilles JC, Carpy A, Galy J. Multiple phases in systems Ca2 Nb2 O7 –NaNbO3 and La2 Ti2 O7 –CaTiO3 : homologous series into formula An Bn O3n+2 . J Sol State Chem 1974;11:272–84. ˇ y R, Skapin ˇ SD, Suvorov D, Meden A. Crystal structures of [40] Stare K, Cern´ CaLa8 Ti9 O31 and Ca2 La4 Ti6 O20 determined from powder diffraction data. Acta Chim Slov 2011;58:465–70. [41] Pivovarova AP, Strakhov VI, Saltykova VA. Phase equilibria in the CaTiO3 –La2 Ti2 O7 system. Russ J Appl Chem 1998;71:1317–9. [42] German M, Kovba LM. Hexagonal perovskite phases in systems of La2 O3 –TiO2 –MO (M = Mg, Ca, Sr, Ba). Zh Neorg Khim 1983;28:2377–9. [43] German M, Kovba LM. Structure of hexagonal AN B(N−1) O3N phases. Zh Neorg Khim 1985;30:317–22. [44] Tohdo Y, Kakimoto K, Ohsato H, Yamada H, Okawa T. Microwave dielectric properties and crystal structure of homologous compounds ALa4 Ti4 O15 (A = Ba, Sr, and Ca) for base station applications. J Eur Ceram Soc 2006;26:2039–43. [45] Pivovarova AP. Synthesis of complex lanthanum and calcium titanates with a perovskite structure and their electric properties studied at high temperatures. Refract Ind Ceram 2002;43:329–31. [46] Vanderah TA, Miller VL, Levin I, Bell SM, Negas T. Phase relations, crystal chemistry, and dielectric properties in sections of the La2 O3 –CaO–MgO–TiO2 system. J Sol State Chem 2004;177: 2023–38.

[47] Aatiq A. Synthesis and crystal structure of the new perovskite CaLa2 CaTi2 O9 (=(Ca1/3 La2/3 )A (Ca1/3 Ti2/3 )B O3 ). Sol State Sci 2003;5:745–9. ˇ [48] Kasuniˇc M, Meden A, Skapin SD, Suvorov D, Golobiˇc A. Order-disorder of oxygen anions and vacancies in solid solutions of La2 TiO5 and La4 Ga2 O9 . Acta Crystallogr B 2009;65:558–66. [49] Coelho AA. TOPAS-Academic, V4.1. Brisbane, Australia: Coelho Software; 2007. [50] Dunyushkina LA, Gorbunov YA. Effect of crystal structure on the electrical properties of CaTi1−x Fex O3−δ . Inorg Mater 2001;37:1165–9. [51] Levin I, Chan JY, Geyer RG, Maslar JE, Vanderah TA. Cation ordering types and dielectric properties in the complex perovskite Ca(Ca1/3 Nb2/3 )O3 . J Sol State Chem 2001;156:122–34. [52] Cheary RW, Coelho A. A fundamental parameters approach to X-ray lineprofile fitting. J Appl Crystallogr 1992;25:109–21. [53] Shannon RD. Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallogr A 1976;32:751–67. [54] Day BE, Bley ND, Jones HR, McCullough RM, Eng HW, Porter SH, et al. Structures of ordered tungsten- or molybdenum-containing quaternary perovskite oxides. J Sol State Sci 2012;185:107–16. [55] Gateshki M, Igartua JM. Crystal structures and phase transitions of the double-perovskite oxides Sr2 CaWO6 and Sr2 MgWO6 . J Phys Condens Matter 2004;16:6639–49. [56] Prior TJ, Couper VJ, Battle PD. Structural chemistry of the cationordered perovskites Sr2 CaMo1−x Tex O6 (0 ≤ x ≤ 1). J Sol State Sci 2005;178:153–7. [57] van Duivenboden HC, Ijdo DJW. Redetermination of tricalcium uranate (VI). A Rietveld refinement of neutron powder diffraction data. Acta Crystallogr C 1986;42:523–5. [58] Fu ZM, Li WX, Liang DC. The phase-transition and crystal structures of a new compound – Sr2 CaWO6 . Sci Sin A 1983;26:835–47. [59] Fu ZM, Li ZW, Li WX. Phase transition and crystal structure of a new compound Sr2 CaMoO6 . Acta Phys Sin 1992;41:937–47.