Intergrowth between the Oxynitride Perovskite SrTaO2N and the

Jul 16, 2018 - TDK Corporation, Technology & Intellectual Property HQ, 2-15-7 ... to recover its stoichiometric composition and electrical insulating ...
0 downloads 0 Views 10MB Size
Download PDF
Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX

pubs.acs.org/IC

Intergrowth between the Oxynitride Perovskite SrTaO2N and the Ruddlesden−Popper Phase Sr2TaO3N Yuya Suemoto,† Yuji Masubuchi,† Yuki Nagamine,‡ Atsuo Matsutani,‡ Takeshi Shibahara,‡ Kumiko Yamazaki,‡ and Shinichi Kikkawa*,† †

Faculty of Engineering, Hokkaido University, N13W8, Sapporo 060-8628, Japan TDK Corporation, Technology & Intellectual Property HQ, 2-15-7 Higashi-Ohwada, Ichikawa, Chiba 272-8558, Japan

‡

Downloaded via UNIV OF READING on July 21, 2018 at 00:38:38 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Strontium tantalum oxynitrides were prepared within the nominal composition range of 1.0 ≤ x ≤ 2.0, where x = Sr/Ta atomic ratio. A gradual structural transition was observed between the perovskite SrTaO2N and the Ruddlesden−Popper phase Sr2TaO3N with increasing SrO content. X-ray diffraction analyses showed that a single-phase perovskite was obtained up to x = 1.1, after which Sr2TaO3N gradually appeared at x ≥ 1.25. High-resolution scanning transmission electron microscopy observations identified the gradual intergrowth of a Ruddlesden−Popper Sr2TaO3N type planar structure interwoven with the perovskite crystal lattice upon increasing x. The crystal lattice at x = 1.4 was highly defective and consisted primarily of perovskite intergrown with a large amount of the Ruddlesden− Popper phase structure. This Ruddlesden−Popper phase layer intergrowth is a characteristic of an oxynitride perovskite rather than the Ruddlesden−Popper defects previously reported in oxide perovskites. Partial substitution of Ta with Sr was also evident in this perovskite lattice. Just below x = 2, a perovskite-type structure was intergrown as defects in the Ruddlesden− Popper Sr2TaO3N. Characterization of Sr2TaO3N in ambient air was challenging due to its moisture sensitivity. Thermal analysis demonstrated that this material was relatively stable up to approximately 1400 °C in comparison with SrTaO2N perovskite, especially under nitrogen. Sr2TaO3N could keep its structure in a sealed tube, and some amount of SrCO3 was observed in XRD after 10 days of exposure to 75% relative humidity under prior ambient conditions. A compact of this material had a relative density of 96% after sintering at 1400 °C under 0.2 MPa of nitrogen, even though a drastic loss of nitrogen was previously reported for a SrTaO2N perovskite under these same conditions. Postammonolysis of the Sr2TaO3N ceramics was not required prior to studying its dielectric behavior. This is in contrast to the SrTaO2N perovskite, which requires postammonolysis to recover its stoichiometric composition and electrical insulating properties.

■

0.402752(6) nm.11 In prior work, an almost fully dense ceramic was obtained by heating under a 0.2 MPa nitrogen atmosphere to prevent further decomposition at higher temperatures. Insulating, stoichiometric SrTaO2N was then completely recovered by ammonolysis of the SrTaO2N0.7 ceramics.12,13 Ferroelectric behavior was confirmed on the surface of a thin slice of this material.14 An addition of an excess of SrCO3 was found to be necessary to compensate for a slight loss of strontium during sintering.15 Ruddlesden−Popper (hereafter denoted as RP) type Sr2TaO3N (SrO·SrTaO2N) is reported to have a K2NiF4 type crystal structure with excess SrO layers interleaved between SrTaO2N perovskite units.16 Neutron diffraction was employed to refine the crystal structure, and the space group I4/mmm with a = 0.403899(4) nm and c = 1.26007(3) nm was determined. The diffraction data also showed that the axial 4e sites in this structure are occupied by oxygen atoms,5,17 suggesting that the nitrogen atoms in the TaO4N2 octahedra are positioned equatorially within each plane. A similar anionic coordination has been proposed for Sr2NbO3N on the basis of

INTRODUCTION Oxynitrides are attracting attention as promising white LED phosphors, photocatalysts, and dielectric materials.1−3 However, the use of oxynitride perovskites is very challenging because their properties vary upon substituting their anions or cations.4 SrTaO2N crystallizes in the I4/mcm space group with a = 0.569411(7) nm and c = 0.80658(2) nm.5 The nitride anions in this compound are in a cis configuration in slightly tilted TaO4N2 octahedra.6,7 The stability of cis-TaO4N2 octahedra was confirmed by using first-principles calculations.8 This configuration is typical for Ta5+ in the d0 state and plays an important role in establishing local polarity and relaxer-type dielectric properties.6−9 The possibility of dielectric properties has been reported for ceramics having porosity values of approximately 45% prepared in an ammonia flow at 1020 °C.10 Sintering to generate dense ceramics is very important for obtaining dielectric materials with superior performance. In connection with this process, partial nitrogen loss has been reported in the case of a SrTaO2N perovskite heated above 1000 °C. This process changes the chemical composition to nonstoichiometric SrTaO2N0.7, which is assumed to be a mixture of a tetragonal I4/mcm phase with a = 0.569796(4) nm and c = 0.806788(8) nm and a cubic Pm3̅m phase with a = © XXXX American Chemical Society

Received: April 19, 2018

A

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

identification and crystal structural determination. Structural parameters were refined by using the RIETAN-FP software package.24 X-ray fluorescence spectroscopy (XRF, SEA6000VX, Hitachi) was employed to assess the chemical composition of cations, working in the FP mode and using SrCO3, Ta2O5, and their various mixtures as standards. STEM observations were performed with a high-resolution transmission electron microscope (Titan G2 60-300, FEI). Thin sample sections were prepared from powder specimens for which x = 1.2 and 2.0 were dispersed in epoxy resins, using a focused ion beam. In addition, a powder sample for which x = 1.4 was dispersed in ethanol and then applied to a copper grid in preparation for TEM observations. HR-STEM images were acquired after identifying the zone axis in the TEM mode at 300 kV, employing a selected-area aperture of 50 nm at the largest. Diffuse reflectance spectroscopy (DRS) data were acquired over the range of 400−800 nm using a UV−vis spectrometer (V-750, JASCO), with a white reference plate (Spectralon, Labsphere) to establish 100% reflectance. The band gap energy was estimated via the Kubelka−Munk technique of extrapolating the onset of absorption to the wavelength axis. Structural changes in response to humidity were studied by exposing the as-prepared Sr2TaO3N powder to 75% relative humidity over a supersaturated aqueous sodium chloride solution for 10 days. Synchrotron XRD experiments were performed using a one-dimensional semiconductor detector (MYTHEN) associated with the BL02B2 beamline in the SPring-8 facility at the Japan Synchrotron Radiation Research Institute, Hyogo, Japan. The X-ray wavelength was 0.0496 nm at 25 keV, and the powder samples were loaded into borosilicate glass capillaries. The thermal stability of each oxynitride sample was assessed by loading 10−20 mg into an alumina crucible and acquiring thermogravimetry−differential thermal analysis−mass spectrometry (TG-DTA-MS, STA2500-QMS403, Netzsch) data. During these trials, the sample temperature was raised to 1550 °C at 20 °C/min under either helium or nitrogen flow (200 mL/min) after evacuating the sample chamber to 10 Pa. The Sr2TaO3N was sintered on the basis of our previous investigations of the SrTaO2N perovskite.12−14,25 In preparation, each uniaxially pressed powder specimen (approximately 0.12 g, 6 mm in diameter) was isostatically cold pressed at 150 MPa. The green compacts were subsequently transferred to a boron nitride crucible and sintered at 1400 °C for 3 h under a 0.2 MPa nitrogen atmosphere in a carbon furnace (High Multi 5000, Fuji Dempa Kogyo). The heating rates were 20 °C/min and 10 °C/min below and above 1200 °C, respectively. After they were held at the desired temperature, the samples were cooled to room temperature in the furnace, after which the ceramic surface was polished with 1000 grit sandpaper. The electrical properties of a Sr2TaO3N ceramic sample (4.9 mm diameter, 0.61 mm thickness) were assessed using an LCR meter (4274A, Hewlett-Packard) at an applied voltage of 0.2 V and over the frequency range from 102 to 105 Hz. Silver paste (4922N, Du Pont) was used to form electrodes after drying under vacuum at 60 °C.

Pauling’s second crystal rule, although nitrogen atoms occupy one of the two axial positions of the octahedra in Nd2AlO3N.18 Recurrent intergrowth along the c axis in the case of other RP members (n = 3, 4, ∞) was reported for the (SrO)(SrNbO2N)n family (for which n = 1, 2) on the basis of high-resolution electron microscopy in the same study. Density functional theory calculations on Sr2TaO3N recently demonstrated that the most energetically favorable configuration consists of in-plane cis nitrogen atoms in conjunction with tilted octahedra, as observed in the SrTaO2N perovskite.19 This configuration suggests that Sr2TaO3N could be ferroelectric, as is the case with SrTaO2N.14,15 The proton-conducting perovskite SrCeO3 was found to be nonstoichiometric on the basis of X-ray diffraction (XRD) data.20 Nominally stoichiometric SrCeO3 was determined to be slightly Sr rich, and its structural defects, as characterized by high-resolution transmission electron microscopy (HR-TEM), include twin domain boundaries and SrO-rich RP type planar defects (that is, antiphase boundaries). Multiferroic BiMnO3 prepared under high pressure exhibits a high density of defects such as twin boundaries, RP antiphase boundaries, and superdislocations associated with small segments of RP defects.21 Similar structural defects, such as intergrowth and dislocation loops, were observed in the magnetoresistive material (Ln,A)3Mn2O7, where Ln = Nd, Pr and A = Ca, Sr, Ba.22 Dissociated superdislocations were also observed in connection with RP type planar defects at the SrRuO3/LaAlO3 interface.23 In the present study, changes in the crystal structure of a SrTaO2N−Sr2TaO3N pseudobinary material were assessed while the strontium oxide content was varied, using XRD and high-resolution scanning transmission electron microscopy (HR-STEM). The stability of the Sr2TaO3N specimens were also assessed in response to variations in both humidity and heating. The sintering behavior of this material under a nitrogen atmosphere was additionally studied so as to investigate its electrical properties.

■

EXPERIMENTAL SECTION

Oxide precursors were obtained by firing mixtures of SrCO3 and Ta2O5 (99.9%, Wako Pure Chemical Industries, Ltd.) at 1.0 ≤ x ≤ 1.85 (where x is the nominal Sr/Ta compositional ratio) twice at 1200 °C for 10 h in alumina crucibles after sufficient grinding. The exception was the x = 2.0 specimen, which was fired only once. The products were mixtures of Sr2Ta2O7 (PDF 01-072-0921), Sr5Ta4O15 (PDF 01-054-1251), and Sr(OH)2(H2O)8 (PDF 01-073-3738) having 1 ≤ x ≤ 1.25. Sr1.4Ta0.6O2.7 (PDF 01-070-7299) also appeared at x = 1.3, as shown in Figure S1 in the Supporting Information. Above x = 1.4, the primary product was Sr5Ta4O15 contaminated with Sr1.4Ta0.6O2.7 (PDF 01-070-7299), with other possible contaminants, as evidenced by several weak unidentified diffraction peaks. The main phase was Sr1.4Ta0.6O2.7 contaminated with impurities such as Sr5Ta4O15 at x = 2.0. Approximately 0.3 g quantities of these oxide precursors were subsequently nitrided several times in alumina boats held within a silica glass tube furnace until there was no apparent change in XRD by the successive firing, applying a temperature of 1000 °C for 20 h under ammonia (99.9%, Sumitomo Seika Chemicals Co., Ltd.) at a flow of 100−150 mL/min. The nitrided products were stored in a nitrogen-filled glovebox in an atmosphere with a dew point below −80 °C and an oxygen concentration of 0.01 ppm (Miwa MFG Co., Ltd., NM3-P60S). XRD analyses were performed using monochromated Cu Kα radiation generated at 40 kV and 40 mA with a diffractometer (Ultima IV Protectus ADS, Rigaku) over the 2θ range of 5−90°. A sampling width of 0.02° and sampling speed of 10°/min were applied for phase

■

RESULTS AND DISCUSSION X-ray Diffraction Analysis of Oxynitrides. Ammonolysis products obtained from simply mixed SrCO3 and Ta2O5 were inhomogeneous mixtures of SrCN2, SrTaO2N, and Sr2TaO3N. Oxide precursors were then used as starting materials to improve the homogeneity in the ammonolysis products. The ammonolysis products having nominal x values of less than 1.1 were found to consist solely of the oxynitride perovskite SrTaO2N without any impurities, as can be seen in the XRD patterns in Figure 1. However, a Sr2TaO3N phase gradually appeared with further increases in x. No apparent change was observed on XRD in additional firing after grinding the oxynitride products. The XRD lines associated with the perovskite slowly shifted to lower diffraction angles, especially at 2θ > 40°, in conjunction with increases in x. These lines also became slightly broader and exhibited more asymmetric tailing B

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

at higher x) in x < 1.5. The amount of Sr2TaO3N impurity gradually increased with x. The SrTaO2N perovskite was not a line phase but exhibited an excess compositional range in terms of its strontium content. The amount of Sr2TaO3N increased further with x, and mixtures with the perovskite phase were obtained up to x = 1.85, as presented in Figure 2. Slightly smaller Sr/Ta actual

Figure 1. X-ray powder diffraction patterns for oxynitride products with nominal x values of 1.0−1.5. Red diamonds and blue triangles indicate diffraction peaks for the SrTaO2N perovskite and Sr2TaO3N Ruddlesden−Popper phases, respectively. Dotted lines are included as visual guides for comparison with the product at x = 1.0.

to lower diffraction angles. These line profiles suggest possible defect formation in the perovskite lattice. The microstructure was studied in detail by TEM observation, as shown in next section. It was difficult to accurately estimate the lattice parameters of the perovskite products with increasing x due to the nature of these diffraction lines. However, the lattice parameters were roughly estimated using Rietveld fitting assuming some structural defects seen in HR-STEM observations mentioned in HR-STEM Observations of Oxynitrides, and the results are summarized in Table 1. The actual Sr/Ta atomic ratios determined by XRF were 0.98(2), 1.03(2), 1.07(2), 1.16(2), 1.21(3), 1.26(3), 1.35(3), and 1.43(3) for x = 1.00, 1.05, 1.10, 1.20, 1.25, 1.30, 1.40, 1.50, respectively. They were evidently slightly less than the nominal values of x, likely due to strontium loss during firing, although some values were close, such as 0.98(2) at x = 1.00. The tetragonal lattice parameters were estimated to be a = 0.5700(3) nm and c = 0.8063(8) nm for SrTaO2N at x = 1.0, both of which are quite similar to those reported in the literature.5,7,10 These values gradually expanded with increasing strontium amounts (that is,

Figure 2. X-ray powder diffraction patterns for oxynitride products with nominal x values of 1.6−2.0. Red diamonds and blue triangles indicate diffraction peaks for the SrTaO2N perovskite and Sr2TaO3N Ruddlesden−Popper phases, respectively.

atomic ratios were observed similarly to the above perovskite products; 1.54(3), 1.65(3), 1.78(3), and 1.93(4) for x = 1.6, 1.7, 1.85, 2.0, respectively. An apparently pure product in XRD was obtained at x = 2.0 but can be analyzed as an about 10% perovskite mixture in Table 1. There were no apparent changes in the XRD line profile for Sr2TaO3N although many differences related to defects were observed in the case of the perovskites discussed above. The lattice parameters were determined to be a = 0.40440(1) nm and c = 1.25966(4) nm, in good agreement with literature values for Sr2TaO3N.5,16,17 The c value decreased slightly with decreasing strontium

Table 1. Structural Parameters Refined by Rietveld Fitting on the Oxynitride Products Having Various Nominal x Valuesa SrTaO2N perovskite lattice param

Sr2TaO3N RP phase

Ta site occupancy

x

a/nm

c/nm

Sr

Ta

1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.85 2.0

0.5700(3) 0.5708(2) 0.57239(2) 0.57165(3) 0.57203(3) 0.57327(5) 0.5732(1) 0.5732(3) 0.5732 (fixed) 0.5732 (fixed)

0.8063(8) 0.8070(6) 0.8067(5) 0.8110(1) 0.8115(1) 0.8086(1) 0.8092(3) 0.8101(9) 0.8101 (fixed) 0.8101 (fixed)

0.005(4) 0.037(6) 0.136(7) 0.215(8) 0.293(9) 0.265(9) 0.53(1) 0.36(2) 0.36 (fixed) 0.36 (fixed)

0.995(4) 0.963(6) 0.864(7) 0.785(8) 0.707(9) 0.735(9) 0.47(1) 0.64(2) 0.64 (fixed) 0.64 (fixed)

lattice param a/nm

0.4037 (fixed) 0.4037(1) 0.40394(8) 0.40382(7) 0.40420(3) 0.40438(3) 0.40447(1) 0.40440(1)

Ta site occupancy c/nm

1.2514 (fixed) 1.2514(3) 1.2524(3) 1.2574(2) 1.2549(1) 1.25779(9) 1.25765(6) 1.25966(4)

Sr

0.46(2) 0.14(1) 0.28(1) 0.00(1)

phase ratio (wt %)

Ta

perovskite

1 1 1 1 0.54(2) 0.86(1) 0.72(1) 1.00(1)

100 100 97.7 91.2 90.1 89.8 54.8 37.8 27.6 9.6

RP phase

fitting value S

2.3 8.8 9.9 10.2 45.2 62.2 72.4 90.4

3.97 3.74 4.25 3.91 4.83 4.13 3.20 5.32 6.99 4.71

a The values on their minor impurities were fixed to those in x = 1.7 for the perovskite products in x = 1.85 and 2.0 and to those in x = 1.3 for the RP phase in x = 1.2.

C

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 3. TEM images and diffraction patterns for oxynitride products for which (a, c) x = 1.2 and (b, d) x = 2.0.

Figure 4. HR-STEM images of the oxynitride for which x = 2.0. Images (b) and (c) are magnified regions from image (a). The atomic image shown in the lower right of (b) was obtained from an inverse fast Fourier transform of data for Sr2TaO3N. Sr and Ta atoms are shown in green and yellow, respectively. Layer stacking is clearly evident along the c axis in the Sr2TaO3N crystal lattice in (b). Perovskite type stacking faults are observed in the Sr2TaO3N crystal lattice in the more magnified image in (c).

aggregates with diameters of 0.5−5 μm in the case of the perovskite having x = 1.2 (Figure 3a). The primary particles in Sr2TaO3N were slightly larger than those in the perovskite and exhibited necking with one another to form elongated secondary particles (Figure 3b). The electron diffraction patterns for these materials showed that the former generated

content in x ≥ 1.6, as shown in Table 1, but these variations were not as severe as those observed for the perovskites. HR-STEM Observations of Oxynitrides. The powder morphologies observed by TEM were different for the Sr1.2TaO2N perovskite and the Sr2TaO3N RP phase. Small particles approximately 100 nm in size formed round D

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 5. HR-STEM images of the oxynitride for which x = 1.2. The region in the yellow square in image (a) is magnified in (b). Dark stacking faults are evident, as indicated by yellow arrows in the SrTaO2N perovskite lattice. These represent intergrown RP phase Sr2TaO3N layers in the perovskite lattice, as shown in the more magnified image in (c).

Figure 6. (a) TEM image and (b) HR-STEM image of the oxynitride perovskite for which x = 1.4. The diffractogram in (c) was obtained by a Fourier transform of the region indicated by the blue square in (b).

a [100] zone axis pattern associated with an I4/mcm perovskite, while the latter produced a [100] zone axis pattern for an I4/mmm RP phase, as represented in Figure 3c,d, respectively. Several bright stripes indicating stacking faults were present in the HR-STEM image of the Sr2TaO3N specimen for which x = 2.0, associated with an electron beam incident along the [110] direction in the I4/mmm phase, as shown in Figure 4a. This well-ordered region is magnified in Figure 4b together with an atomic image calculated by a Fourier transform of the Sr2TaO3N crystal structure. The bright defect area is also enlarged in Figure 4c. Elastically scattered electron intensity by each atom is observed in Z-contrast mode in HR-STEM. It is proportional to atomic number Z of the observed atom. Sr and Ta can be distinguished by the technique because their Z numbers are very different. This region in Figure 4c is

attributed to a Ta-rich stacking fault, as suggested by the slightly lower Sr/Ta ratio of 1.93(4) observed in XRF. The value of less than 2 suggests the presence of an either Sr-poor or Ta-rich region in pure Sr2TaO3N crystals. The zone represents a perovskite type defect layer with more Ta than the regular Sr2TaO3N lattice. The bright stripe suggests a higher electron intensity due to the presence of an excess Ta layer. Thus, some perovskite layers were interleaved in the Sr2TaO3N RP type crystal structure as stacking faults. Other stacking faults were observed in the HR-STEM image of the x = 1.2 Sr1.2TaO2N perovskite, as indicated by the yellow arrows in Figure 5a. These show up as dark stripes and are attributed to intergrown Sr-rich layers in the SrTaO2N perovskite crystal lattice. This assignment is supported by the slight excess of SrO in the Sr1.2TaO2N perovskite, on the basis of a Sr/Ta ratio of 1.16(2). The region indicated by the yellow E

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 7. Three different perovskite areas as observed by HR-STEM at x = 1.4. The areas A−C are shown at higher magnifications in images for which the scattered electron intensities are defined by the color bar to the right of C (in the order of yellow, red, green, and blue).

Figure 8. Schematic drawings of (a) an antiphase boundary in the pseudo-SrTaO2N perovskite, (b) the crystal structure of Sr2TaO3N, and (c) Ta atoms at B sites in the SrTaO2N perovskite partially substituted with Sr.

square in this figure is magnified in Figure 5b. This region was primarily composed of a regular perovskite crystal lattice with some amount of intergrown RP type Sr2TaO3N layers, as illustrated in Figure 5c. However, the stacking faults are different from a SrO-rich antiphase boundary reported on oxide perovskites because crystal lattices are in phase in both sides across the boundary.20−22 Round secondary particles observed on Sr1.2TaO2N were also seen on the Sr1.4TaO2N product during TEM assessments. Sr1.4TaO2N was a mixture of particles of the perovskite and a small amount of Sr2TaO3N in XRD. Figure 6a presents a TEM image of a representative perovskite particle, while an HRSTEM image is provided in Figure 6b. The latter shows that

the crystal lattice was less clear and more defective in comparison to the perovskite having x = 1.2 and that obtained at x = 2.0 for a Sr2TaO3N RP phase. Clear streaks are apparent on the reflections along the c axis in the diffractogram obtained from the area in the blue rectangle in Figure 6b, as shown in Figure 6c. They suggest that the layer stacking along the c axis was highly defective. Additionally the appearance of a 011 diffraction spot suggests that the crystal symmetry was changed somewhat. The crystal lattice was primarily composed of the perovskite with some RP type defects, similar to the case for the x = 1.2 specimen, but with a much more complicated structure in this particle. The defects in the blue squared area in Figure 6b were studied further in detail. F

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

About 30% of Ta in B sites was substituted with Sr in the perovskite at x = 1.4, as depicted in Table 1. Most of the residues in the fittings were ascribed to the intergrowth between the two phases as their stacking faults after the fitting calculations with an optimized exchange amount. UV−Visible Diffuse Reflectance Spectroscopy of Oxynitrides. The color of the oxynitrides changed from orange to reddish orange with an increasing x, and the diffuse reflectance spectra of these specimens are provided in Figure 9.

An ordered perovskite lattice was observed in area A in Figure 7A, while RP type Sr2TaO3N fault layers were intergrown within the perovskite lattice at the middle in area B (Figure 7B). This in-phase defect differs from the antiphase boundary of the oxide perovskite lattice shown in Figure 8a. This type of defect had been previously reported to occur in proton conductive SrCeO3 and multiferroic BiMnO3 perovskites in the compositional range having a slight excess of A site cations.20,21 The fault layers in oxide perovskites are typically formed by oxide anions coordinated to A and B site cations. However, it is difficult to form similar stacking fault layers solely from the oxide anions in the oxynitride perovskite SrTaO2N. The octahedra around Ta atoms must have a cisTaO4N2 configuration in electronic stabilization on the basis of the presence of Ta5+ in the 5d0 state, as suggested by firstprinciples calculation.8 The nitride anions cannot participate in the formation of an antiphase boundary in this case, and this formation is less favorable than interleaving Sr2TaO3N RP phase layers. Excess SrO layers are interleaved in the perovskite lattice, forming a RP type Sr2TaO3N layer fault as a result of the electronic requirement of the Ta5+, as depicted in Figure 8b. The recurrent intergrowth of other RP members (n = 3, 4) has been reported in the defect structures of (SrO)(SrNbO2N)n oxynitrides for which n = 1, 2.18 However, the present study is the first instance of the intergrowth of a Sr2TaO3N RP phase layer in an oxynitride perovskite. Another type of defect is evident in area C, and Figure 7C shows that yellow-red dots and green-blue dots are present alternatively. This result suggests that a portion of the Ta at the B sites was substituted with Sr atoms having weaker electron scattering intensity in comparison to Ta atoms to generate a rock salt type crystal lattice (Fm3̅m) in this area. Similar structural modulation had been reported on the perovskite-like oxide Sr1.4Ta0.6O2.9, where a great amount of Sr substituted Ta in the B sites of the oxygen-deficient perovskite.26 The extra diffraction spots in Figure 6c are attributed to this change in the space group. As an example, an extra 011 diffraction in the I4/mcm pattern corresponds to 111 in the Fm3̅m space group. The defect structure is shown schematically in Figure 8c. Rock salt type ordering is most common in association with B site cations in double-perovskite oxides such as Sr2FeMoO6, Ba2MgWO6, and Sr2NaOsO6.4,27−29 A similar structural transition to the double-perovskite oxides appears to have occurred to some extent in the defective SrTaO2N lattice in the area C. These intergrowths described in this section can be determined easily in the HR-STEM observations on specimens with the respective x values. We did not take any special care to observe these intergrowths. They were observed frequently, e.g. in 10−25% possibility in STEM observations. Rietveld fitting was performed by taking into account the Sr substitution of Ta observed in HR-STEM on all XRD patterns in Figures 1 and 2. Phase amounts were estimated on both the SrTaO2N perovskite and Sr2TaO3N RP phases in all products, as summarized in Table 1. The perovskite amount continuously decreased and the RP phase amount increased with x value. Fittings of the calculation were much worse in the cases of both x = 1.2 and 1.4 in comparison to x = 1.0 because of the very defective structure observed by HR-STEM. Much larger residues were observed between the observation and calculations of the products with x = 1.2, 1.4, and 2.0 in comparison to the x = 1.0 product, as shown in Figures S2−S5. It is not easy to deal with the stacking faults in Rietveld fitting, but the Ta site substitution with Sr is much easier to calculate.

Figure 9. UV−vis reflectance spectra of oxynitrides prepared at x = 1.0, 1.2, 1.4, 1.7, 2.0.

The band gaps were determined to be 2.03, 1.97, 1.94, 1.90, and 1.90 eV at x = 1.0, 1.2, 1.4, 1.7, 2.0, respectively. These values therefore gradually decreased with increasing x. The end members were SrTaO2N and Sr2TaO3N according to the XRD data but there was some amount of stacking faults intergrown in one another, as noted above during the discussion of the HR-STEM observations. The band gap values reported for ATaO2N perovskite are 1.8, 2.1, and 2.4 eV for A = Ba, Sr, Ca, respectively.10 This change in band gap is related to modifications in the Ta−(O,N)−Ta bond angle as the size of the A site cations varies.30 The bond angles for these respective compounds are 180, 169.9, and 153.3°. This in turn changes the conduction bandwidth via overlap between the t2g orbitals of Ta5+ and anionic 2p orbitals. The band gap for BaTaO2N is the smallest because the bottom of its conduction band is the lowest among the ATaO2N perovskites. All of the Ta−(O,N)−Ta bond angles were found to be 180° in a previous neutron structural refinement of Sr2TaO3N, similar to the results obtained for BaTaO2N.5 The bond angle in Sr2TaO3N is in agreement with its slightly smaller band gap in comparison to that for the SrTaO2N perovskite. Most of the present oxynitride products were mixtures of SrTaO2N and Sr2TaO3N, with mutual intergrowth. Their relative amounts changed with increasing x, and this effect also contributed to a gradual decrease in band gap energy. Stability of Sr2TaO3N Oxynitride against Humidity and High Temperatures. Neutron diffraction studies of the crystal structure of Ba2TaO3N have been hampered in the past because of the extreme moisture sensitivity of this compound.5 The humidity sensitivity of a Sr2TaO3N powder sample was examined at the SPring-8 synchrotron radiation facility at ambient temperature. The sample was sealed in a borosilicate G

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

sample powder was mostly Sr2TaO3N even after its surface was hydrolyzed. Another nitrogen loss observed in the range of 700−800 °C under helium (Figure S6) might be related to the formation of Sr1.4Ta0.6O2.7 from the reaction of Sr2TaO3N with its hydrolysis products such as Sr(OH)2 and SrCO3. Further nitrogen loss at 1300 °C to form Sr1.4Ta0.6O2.7 under helium was not as clearly evident as that under a nitrogen atmosphere, as can be seen from Figure 11, on the basis of the additional mass loss above 1400 °C. Some amount of Sr2TaO3N apparently changed to the SrTaO2N perovskite along with a small amount of Sr1.4Ta0.6O2.7 as an impurity following thermal analysis up to 1300 °C under nitrogen (Figure 12). The

glass capillary to remove additional humidity effects during the measurement. A small amount of SrCO3 appeared upon exposure to 75% relative humidity for 10 days in advance, although there was no change in the nonexposed sample, as shown in Figure 10. The diffraction intensity of Sr2TaO3N was also diminished because its crystallinity was reduced due to its moisture sensitivity.

Figure 10. Synchrotron X-ray diffraction patterns for Sr2TaO3N before (green) and after (red) exposure to 75% RH for 10 days. Patterns were acquired from powder samples sealed in a borosilicate capillary to avoid any effect from the ambient atmosphere, in work at the SPring-8 facility. Red triangles are diffraction peaks assigned to SrCO3.

Figure 12. Sample color and powder X-ray diffraction patterns for Sr2TaO3N (a) before and (b) after thermal analysis up to 1300 °C under nitrogen. Orange triangles, blue diamonds, and black circles indicate diffraction peaks for Sr2TaO3N, SrTaO2N, and Sr1.4Ta0.6O2.7, respectively.

Thermal stability was studied on as-prepared fresh samples on the basis of TG-DTA-MS analyses under nitrogen up to 1550 °C with care taken to minimize the effect of humidity and CO2 adsorption. Both a mass loss of 0.3 wt % and gas evolution (m/z 18) were observed below 300 °C, as shown in Figure 11. Adsorbed water on the powder surface in advance

sample color also changed slightly, from reddish orange to a more characteristic orange associated with the SrTaO2N perovskite. The heated product remained mostly Sr2TaO3N. The lattice parameters of Sr2TaO3N were a = 0.40411(1) nm and c = 1.2595(2) nm after the thermal analysis and thus were essentially unchanged. Sintering and Electrical Properties of Sr2TaO3N Oxynitride. The data shows that Sr2TaO3N might be thermally stable below 1400 °C under a nitrogen atmosphere, although it is moisture sensitive, as detailed in Stability of Sr2TaO3N Oxynitride against Humidity and High Temperatures. This is different from the behavior of SrTaO2N, because the latter perovskite lost about 15% of its nitrogen at approximately 950 °C even under nitrogen. Postammonolysis of the SrTaO2N ceramics was necessary to recover its stoichiometric composition and electrically insulating behavior after sintering at 1400 °C under a 0.2 MPa nitrogen atmosphere.12,13,15 Sr2TaO3N ceramics sintered under the same conditions were densified to a 96% relative density in 6.5 g/cm3. XRD studies confirmed that these materials were composed mainly of Sr2TaO3N with smaller concentrations of the SrTaO2N perovskite and other impurities such as Sr1.4Ta0.6O2.7, Sr(OH)2, and SrCO3. This result is in agreement with the data obtained from TG-DTA-MS analysis of the powder product heated to 1300 °C. The resulting material was black on its exterior, but a significant number of purple grains were observed on the interior of the ceramic pellet and an orange coloration was evident after crushing. The electrical resistivity of the specimen was 2.4 MΩ m, as determined by ac measurements. The relative permittivity εr and dielectric loss tan δ varied from 7 × 103 to 5 × 103 and from 3 to 6 × 10−2, respectively, over the frequency range of 102−105 Hz, as shown

Figure 11. TG-DTA-MS data obtained from Sr2TaO3N under a nitrogen atmosphere.

was evidently removed during this process, after which a small amount of hydroxide groups was lost from the hydrolyzed surface at approximately 400 °C. Additional gas evolution appeared in the two temperature ranges of 500−800 and 1200−1450 °C, in conjunction with m/z 44. These steps represented the desorption of CO2 from the amorphous hydroxyl carbonate and from the crystalline SrCO3 found to be formed during the hydrolysis of the Sr2TaO3N surface on the basis of XRD data. Partial nitrogen desorption from SrTaO2N has been reported at 950−1100 °C.11 This evolution might be associated with a mass loss (in the vicinity of 1000 °C) from the oxynitride perovskite formed during the hydrolysis of Sr2TaO3N in ambient air prior to the measurements. The H

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

■

in Figure 13. These values are much larger than those obtained for the annealed SrTaO2N (for example, εr = 450 and tan δ =

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.8b01079. XRD for oxide precursors up to x = 1.3, Rietveld fittings for the products with x = 1.0, 1.2, 1.4, 2.0, respectively, and TG-DTA-MS for Sr2TaO3N in helium. (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail for S.K.: [email protected]. ORCID

Yuji Masubuchi: 0000-0003-3601-7077 Yuki Nagamine: 0000-0003-1938-8079 Takeshi Shibahara: 0000-0003-0911-6387 Shinichi Kikkawa: 0000-0002-3498-0735 Figure 13. Dielectric properties of Sr2TaO3N sintered at 1400 °C under a 0.2 MPa nitrogen atmosphere. The sample was contaminated with Sr1.4Ta0.6O2.7, Sr(OH)2, and SrCO3 impurities.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors give their appreciation to Professor M. Higuchi at Hokkaido University for his valuable discussion. This research was financially supported by a KAKENHI Grant-in-Aid for Scientific Research (A) (No. 24245039) and by Scientific Research on Innovative Areas “Mixed Anion” (No. JP16H06439) from the Japan Society for the Promotion of Science (JSPS).

0.1 at 10 Hz) and BaTaO2N (εr = 620 and tan δ = 0.4 at 10 Hz).12,25 The larger permittivity can be expected from the crystal structure of Sr2TaO3N similarly to the discussion on its optical band gap in a comparison between ATaO2N perovskites (A = Ca, Sr, Ba).10 The Ta−(O,N)−Ta bond angles are 180° in both BaTaO2N and Sr2TaO3N, as discussed above in UV−Visible Diffuse Reflectance Spectroscopy of Oxynitrides. In future work, it will be necessary to remove impurities such as Sr1.4Ta0.6O2.7 from the ceramics to obtain the actual values for dielectric constants and losses by improving the sintering process, especially through reducing exposure to humidity. The Sr-rich perovskite has an advantage in obtaining SrTaO2N ceramics because the addition of a small amount of SrCO3 can be avoided in the high-temperature sintering process of stoichiometric SrTaO2N. We need not care about the CO2 gas evolution from the SrCO3 additive, which will hamper the densification of perovskite. 2

2

■

REFERENCES

(1) Takeda, T.; Xie, R.-J.; Suehiro, T.; Hirosaki, N. Nitride and oxynitride phosphors for white LEDs: Synthesis, new phosphor discovery, crystal structure. Prog. Solid State Chem. 2017, DOI: 10.1016/j.progsolidstchem.2017.11.002. (2) Maeda, K. CO2 reduction using oxynitrides and nitrides under visible light. Prog. Solid State Chem. 2017, DOI: 10.1016/ j.progsolidstchem.2017.11.003. (3) Kikkawa, S.; Hosono, A.; Masubuchi, Y. Remarkable effects of local structure in tantalum and niobium oxynitrides. Prog. Solid State Chem. 2017, DOI: 10.1016/j.progsolidstchem.2017.08.001. (4) Shimakawa, Y. Perovskite structure compounds. In Handbook of Solid State Chemistry; Dronskowski, R., Kikkawa, S., Stein, A., Eds.; Wiley-VCH: Weinheim, 2017; Vol. 1 (Materials and Structure of Solids), Chapter 6, pp 221−250. (5) Clarke, S. J.; Hardstone, K. A.; Michie, C. W.; Rosseinsky, M. J. High-temperature synthesis and structures of perovskite and n = 1 Ruddlesden−Popper tantalum oxynitrides. Chem. Mater. 2002, 14, 2664−2669. (6) Yang, M.; Oro-Sole, J.; Rogers, J.; Jorge, A.; Fuertes, A.; Attfield, J. P. Anion order in perovskite oxynitrides. Nat. Chem. 2011, 3, 47− 52. (7) Zhang, Y.; Motohashi, T.; Masubuchi, Y.; Kikkawa, S. Local anionic ordering and anisotropic displacement in dielectric perovskite SrTaO2N. J. Ceram. Soc. Jpn. 2011, 119, 581−586. (8) Hinuma, Y.; Moriwake, H.; Zhang, Y.; Motohashi, T.; Kikkawa, S.; Tanaka, I. First principles study on relaxor-type ferroelectric behavior without chemical inhomogeneity in BaTaO 2N and SrTaO2N. Chem. Mater. 2012, 24, 4343−4349. (9) Fuertes, A. Chemistry and applications of oxynitride perovskites. J. Mater. Chem. 2012, 22, 3293−3299. (10) Kim, Y.-I.; Woodward, P.; Baba-Kishi, K.; Tai, C. Characterization of the structural, optical, and dielectric properties of oxynitride perovskites AMO2N (A = Ba, Sr, Ca; M = Ta, Nb). Chem. Mater. 2004, 16, 1267−1276.

■

CONCLUSION A strontium-rich compositional range was demonstrated for the SrTaO2N oxynitride perovskite. The Sr2TaO3N RP phase was found to interleave as stacking faults along with increases in the amount of SrO in the SrTaO2N perovskite. The domain size and amount of this RP phase were both increased in the highly defective perovskite for which x = 1.4. This was accompanied by the growth of a rock salt type defect domain in which Ta atoms at B sites were partially substituted by Sr. A small quantity of perovskite type stacking faults were also identified in Sr2TaO3N. The possibility that Sr2TaO3N ceramics could exhibit relaxer-type dielectric characteristics was confirmed similarly to ATaO2N perovskites where A = Sr, Ba.8,14 The sintering process should be improved to optimize these properties because this compound is evidently moisture sensitive, although it may be thermally stable up to 1400 °C under nitrogen. The present intergrowth of perovskite and RP phases directly depends on the Sr/Ta atomic ratio in the starting composition. This intergrowth can change the properties such as band gap and relaxer behavior between the end member products similarly to their solid solution formation. It will be useful to adjust the properties as we wish. I

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (11) Chen, D.; Habu, D.; Masubuchi, Y.; Torii, S.; Kamiyama, T.; Kikkawa, S. Partial nitrogen loss in SrTaO2N and LaTiO2N oxynitride perovskites. Solid State Sci. 2016, 54, 2−6. (12) Sun, S.-K.; Zhang, Y.-R.; Masubuchi, Y.; Motohashi, T.; Kikkawa, S. Additive sintering, post annealing, and dielectric properties of SrTaO2N. J. Am. Ceram. Soc. 2014, 97, 1023−1027. (13) Zhang, Y.-R.; Motohashi, T.; Masubuchi, Y.; Kikkawa, S. Sintering and dielectric properties of perovskite SrTaO2N ceramics. J. Eur. Ceram. Soc. 2012, 32, 1269−1274. (14) Kikkawa, S.; Sun, S.; Masubuchi, Y.; Nagamine, Y.; Shibahara, T. Ferroelectric response induced in cis-type anion ordered SrTaO2N oxynitride perovskite. Chem. Mater. 2016, 28, 1312−1317. (15) Masubuchi, Y.; Sun, S.-K.; Kikkawa, S. Processing of dielectric oxynitride perovskites for powders, ceramics, compacts and thin films. Dalton Trans. 2015, 44, 10570−10581. (16) Pors, F.; Marchand, R.; Laurent, Y. Nouveaux oxynitrures A2TaO3N (A = alcalinoterreux) de type structural K2NiF4. Ann. Chim. Fr. 1991, 16, 547−551. (17) Diot, N.; Marchand, R.; Haines, J.; Leger, J. M.; Macaudiere, P.; Hull, S. Crystal structure determination of oxynitride Sr2TO3N. J. Solid State Chem. 1999, 146, 390−393. (18) Tobias, G.; Beltran-Porter, D.; Lebedev, O.; Tendeloo, G.; Rodrigues-Carvajal, J.; Fuertes, A. Anion ordering and defect structure in Ruddlesden-Popper strontium niobium oxynitrides. Inorg. Chem. 2004, 43, 8010−8017. (19) Bouri, M.; Aschauer, U. Bulk and surface properties of the Ruddlesden-Popper oxynitride Sr2TaO3N. Phys. Chem. Chem. Phys. 2018, 20, 2771−2776. (20) Mather, G. C.; Figueiredo, F. M.; Paz, J. R.; Garcia-Martin, S. Strontium superstoichiometry and defect structure of SrCeO3 perovskite. Inorg. Chem. 2008, 47, 921−929. (21) Yang, H.; Chi, Z. H.; Yao, L. D.; Zhang, W.; Li, F. Y.; Jin, C. Q.; Yu, R. C. Structural defects in multiferroic BiMnO3 studied by transmission electron microscopy and energy-loss spectroscopy. J. Appl. Phys. 2006, 100, 044105. (22) Laffez, P.; Tendeloo, G.; Seshadri, R.; Hervieu, M.; Martin, C.; Maignan, A.; Raveau, B. Microstructural and physical properties of layered manganites oxides related to magnetoresistive perovskites. J. Appl. Phys. 1996, 80, 5850−5856. (23) Wu, J. S.; Jia, C. L.; Urban, K.; Hao, J. H.; Xi, X. X. A new mechanism for misfit dislocation generation: superdislocations associated with Ruddlesden-Popper planar defects. J. Cryst. Growth 2002, 234, 603−609. (24) Izumi, F.; Momma, K. Three-dimensional visualization in powder diffraction. Solid State Phenom. 2007, 130, 15. (25) Hosono, A.; Sun, S.; Masubuchi, Y.; Kikkawa, S. Additive sintering and post-annealing of dielectric BaTaO2N oxynitride perovskite. J. Eur. Ceram. Soc. 2016, 36, 3341−3345. (26) Deniard, P.; Caldes, M. T.; Zou, X. D.; Diot, N.; Marchand, R.; Brec, R. Structural modulation in Sr1.4Ta0.6O2.9: non-harmonicity on ADPs during Rietveld refinement. Int. J. Inorg. Mater. 2001, 3, 1121− 1123. (27) Anderson, M. T.; Greenwood, K. B.; Poppelmeier, K. R. Bcation arrangements in double perovskites. Prog. Solid State Chem. 1993, 22, 197−233. (28) King, G.; Woodward, P. M. Cation ordering in perovskites. J. Mater. Chem. 2010, 20, 5785−5796. (29) Vasala, S.; Karppinen, M. A2B′B″O6 perovskites: A review. Prog. Solid State Chem. 2015, 43, 1−36. (30) Aguiar, R.; Logvinovich, D.; Weidenkaff, A.; Rachel, A.; Reller, A.; Ebbinghaus, S. The vast colour spectrum of ternary metal oxynitride pigments. Dyes Pigm. 2008, 76, 70−75.

J

DOI: 10.1021/acs.inorgchem.8b01079 Inorg. Chem. XXXX, XXX, XXX−XXX