Synthesis of nitrogen and lanthanum codoped barium titanate with a novel thermal ammonolysis reactor

Synthesis of nitrogen and lanthanum codoped barium titanate with a novel thermal ammonolysis reactor

Journal of the European Ceramic Society 36 (2016) 2719–2725 Contents lists available at www.sciencedirect.com Journal of the European Ceramic Societ...

2MB Sizes 2 Downloads 10 Views

Journal of the European Ceramic Society 36 (2016) 2719–2725

Contents lists available at www.sciencedirect.com

Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc

Synthesis of nitrogen and lanthanum codoped barium titanate with a novel thermal ammonolysis reactor Yusuke Otsuka a,∗ , Christian Pithan b , Jürgen Dornseiffer c , Takahiro Takada a , Takehiro Konoike a , Rainer Waser b,1 a

Murata Manufacturing Co., Ltd., Nagaokakyo-shi, Kyoto 617-8555, Japan Peter Grünberg Institute, Elektronische Materialien, Forschungszentrum Jülich GmbH, D-52425 Jülich, Germany c Institute für Energie- und Klimaforschung, Werkstoffsynthese und Herstellungsverfahren, Forschungszentrum Jülich GmbH, D-52425 Jülich, Germany b

a r t i c l e

i n f o

Article history: Received 21 December 2015 Received in revised form 25 March 2016 Accepted 29 March 2016 Available online 8 April 2016 Keywords: Barium titanate Oxynitride Thermal ammonolysis

a b s t r a c t Lanthanum doped barium titanate was nitridated by thermal ammonolysis with a newly designed reactor, which performs extended efficiency to supply ammonia with less decomposition at elevated temperatures above 600 ◦ C. The kinetics of nitrogen-uptake was studied systematically at elevated temperatures. Nitrogen-content, thermal stability and crystallographic phase of nitridated lanthanum doped barium titanate were characterized by several techniques. The final nitrogen-content of nitridated lanthanum doped barium titanate was much higher than the content of lanthanum, resulting in the formation of oxygen vacancies compensating the additional negative electrical charge of nitrogen anions occupying regular oxygen sites in the lattice. Through thermal ammonolysis, strongly reducing conditions because of the decomposition of ammonia generating hydrogen gas causes the formation of barium orthotitanate as a secondary phase. Ammonolysis temperature below 800 ◦ C and suppression of thermal decomposition of ammonia gas are essential to realize a nitridated lanthanum doped barium titanate as a single phase. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Tuning of the physical properties in electroceramics based on barium titanate BaTiO3 is generally achieved by compositional or microstructural design, e.g. acceptor- or donor-doping [1–4], the use of nano-powders [5]. Such compositional design is mainly realized by substitutional chemical modifications on the cationic sites of Ba2+ and Ti4+ , but not on the anionic site of the perovskite lattice. In recent years perovskite-related oxides with a partial substitution of oxygen by nitrogen, referred to as oxynitrides, are increasingly stimulating interest due to their unusual characteristics making them attractive for applications like non toxic pigments [6], photocatalysts for water splitting [7], giant permittivity [8] and giant magnetic resistance [9]. Although there are some examples of nitrogen-doped functional electroceramics reported in the lit-

∗ Corresponding author. E-mail addresses: [email protected] (Y. Otsuka), [email protected] (C. Pithan), [email protected] (J. Dornseiffer), t [email protected] (T. Takada), [email protected] (T. Konoike), [email protected] (R. Waser). 1 Institut für Werkstoffe der Electrotechnik, RWTH Aachen, D-52056 Aachen, Germany http://dx.doi.org/10.1016/j.jeurceramsoc.2016.03.033 0955-2219/© 2016 Elsevier Ltd. All rights reserved.

erature [10–12], the total number of published studies is rather small since the field is still quite new. Especially regarding the material synthesis of barium titanate which is massively applied on an industrial scale into passive electronic components especially as insulating multilayered co-fired ceramic capacitors (MLCC) [13] and semiconducting thermistors showing a positive temperature coefficient of resistance (PTCR) [14] less information is available [15,16]. Therefore the questions, to what extent oxygen may be exchanged by nitrogen in barium titanate and how this substitution affects the crystallography, conductivity and dielectric characteristics of barium titanate based materials is highly relevant not only to the fundamental but also to the technological point of view. Most widely ammonia is used for nitridation of perovskiterelated oxides at high temperatures (600 ◦ C–1200 ◦ C) [17]. In this temperature range the rapid decomposition of ammonia NH3 into hydrogen H2 and nitrogen N2 strongly limits the effectiveness of the nitridation reaction. At 600 ◦ C, for example more than 99.8% of ammonia decomposes under conditions of thermodynamical equilibrium at standard pressure. This is illustrated in Fig. 1, showing the temperature dependence of the concentration of undecomposed ammonia [18] according to the relation: 2NH3 ↔ N2 + 3H2

(1)

2720

Y. Otsuka et al. / Journal of the European Ceramic Society 36 (2016) 2719–2725

its poor scalability. Therefore we designed and constructed an ammonolysis reactor to suppress the decomposition of ammonia at elevated temperatures. In this study we investigated the influence of the nitridation parameters (temperature, ammonia flow rate and reaction time) on phase purity, crystallography and nitrogen-content of La-doped barium titanate powders. These ceramic compositions are possibly technically relevant to eventual applications as semiconducting donor doped PTC-thermistor components, those need acceptor compensation in the form of oxygen exchange by nitrogen balancing the loss of oxygen that results from firing in reducing conditions. In this case La3+ cations substituting Ba2+ cations on the regular A-sites of the perovskite lattice are electrically equalizing the negative surcharge caused by N3− that replace O2− ions. 2. Experimental procedure 2.1. Sample preparation Fig 1. Temperature dependence of concentration of ammonia under conditions of thermodynamical equilibrium at standard pressure.

Two strategies are known to improve the efficiency of the ammonolysis reaction. The first one is achieved by increasing the reactivity of ammonia, e.g. by plasma gasification, and the second one is realized by suppressing the decomposition of ammonia. Although it is known that microwave induced ammonia plasma improves the ammonolysis reaction [17], this method is at present not applicable for large scale ammonolysis processes because of

The nitridation process of La-doped barium titanate Ba0.997 La0.003 TiO3 was studied in the powdery state in order to achieve an as homogeneous nitrogen uptake as possible. For the preparation of barium titanate powder, the conventional mixed oxide route was used. Commercial grade of BaCO3 (≥99% pure, Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany), TiO2 (≥99.9% pure, Sigma-Aldrich) and La2 O3 (Alfa Aesar GmbH & Co., KG, Karlsruhe, Germany) as a dopant were milled with yttriastabilized zirconia (YSZ) balls and 2-propanol for 6 h on a roller bench. After milling and separation from the balls, the resulting slurry was dried in a rotary evaporator and calcined at 1150 ◦ C for

Fig. 2. Illustration of the thermal ammonolysis reactor, (a) overall image, magnified image of (b) the heating zone inside the tube furnace, and (c) the cooling zone outside of the furnace on the right side.

Y. Otsuka et al. / Journal of the European Ceramic Society 36 (2016) 2719–2725

2721

with nitrogen in order to eliminate excess ammonia gas, before locking out the nitridated sample. 2.3. Characterization techniques

Fig. 3. The relationship between specimen temperature and ammonia temperature. Ammonia temperature is measured on the outer surface of the quartz tube in which flowing ammonia is introduced with different flow rates of cooling air (in slm, standard liter per minute).

Thermochemical methods were employed to characterize the nitrogen content and thermal stability of the nitridated powders. The nitrogen uptake of nitridated La-doped barium titanate was analyzed by the hot gas extraction method using a LECO TC500 system (Michigan, U.S.A.). The principle of this technique is based on the extraction of nitrogen in the form of gaseous N2 in the strongly reducing environment of a graphite crucible in a flow of helium at elevated temperatures. The amount of extracted N2 is then detected over thermal conductivity. Thermogravimetric analysis (TG) was used (Netsch TG439, Selb, Germany) in order to clarify the extent of nitrogen uptake within nitridated barium titanate. Crystallographic phases were analyzed by X-ray powder diffraction (XRD) using a Huber G670 diffractometer (Rimsting, Germany) with monochromated Cu-K␣ radiation. Scanning electron micrographs of the powdery specimens morphology were taken on an SEM Hitachi High Technologies SU6000 instrument (Tokyo, Japan). 3. Results and discussion 3.1. Nitrogen content

2 h in air. The as calcined powders were then milled again with YSZ balls and 2-propanol for 6 h and dried. The average particle size d of the powders was 0.32 ␮m, which was estimated by specific surface area (SSA) S determined by N2 -adsorbtion (BET-method) using the equation d = 6/S (theoretical density of BaTiO3 was assumed to be  = 6.02 g/cm3 ).

2.2. Thermal ammonolysis The reactor designed and used in the present study consists of a “heating zone” and a “cooling zone” (Fig. 2). The heating zone for nitridation at elevated temperatures (typically 750–1000 ◦ C) inside a tube furnace and the cooling zone for quenching the nitridated specimen outside the furnace are enclosed in a gas tight quartz tube. For kinetic studies a specimen holder (driving plate) containing the material to be nitridated can be moved between both zones, in order to investigate the time dependent nitrogen uptake (15 min to 16 h). In the heating zone, ammonia is showered over the specimen through a quartz pipe that has been in turn coaxially embedded into another tube, in which flowing air is cooling down the ammonia gas stream in the inner tube. Due to the stream of cooling air flowing, the temperature of the quartz tube from which ammonia is showered onto the sample became considerably lower than the temperature of the specimen (Fig. 3). The NH3 flow rates were systematically varied in the range from 50 standard cubic centimeter per minute (sccm) up to 200 sccm. Eventual dilution of NH3 with N2 was realized by introducing mixtures of these two gases (100 sccm of NH3 with 200 or 300 sccm of N2 ). For the thermal ammonolysis of a specimen, sample powder was placed in an alumina crucible and settled on the specimen holder at the end of the connecting rod. After purging residual and moist air out of the reactor with dried nitrogen gas, the sample was quickly slid into the heating zone which was heated at a certain target temperature in advance. After the stabilization of the specimen temperature, flowing ammonia was introduced. The ammonolysis reaction was then—after a certain reaction time—quenched by sliding the driving rod from the heating zone into the cooling zone. Ammonia was kept to flow there too until the specimen temperature decreased below 250◦ C. Finally, the reactor tube was purged

As a representative composition of La-doped barium titanate, La0.003 Ba0.997 TiO3 –typically applied as thermistor material—was used for the following investigation (because it is expected that La3+ cations occupying regular sites of Ba2+ electrically compensate N3− anions on O2 −sites.) The incorporation of N3− anion into regular sites of O2− anions in La-doped barium titanate causes additional negative charge, and it needs to be compensated. Such charge compensation could be realized by following mechanisms expressed in a defect equation according to Kröger and Vink [19] notation: 1/2La2 O3 + TiO2 + NH3 ↔ LaBa • + TiTi x + 2Oo x + No  + 3/2H2 O (2)

3Oo x + 2NH3 ↔ 2No  + Vo •• + 3H2 O

(3)

These equations represent the additional charge from N3− anions are compensated by La3+ cations and oxygen vacancy, respectively. The nitrogen uptake in nitridated La-doped barium titanate was measured as a function of the reaction conditions of ammonolysis. The nitrogen-content in mol% relative to titanium in a nitridated La-doped barium titanate was estimated by assuming a chemical composition of La0.003 Ba0.997 TiO3+0.0015-1.5x Nx (x > 0.003), which was derived from the Eq. (2) and (3). The reaction kinetics of nitrogen uptake is very sensitive to the ammonolysis temperature (Fig. 4). At relatively high temperatures (900 ◦ C) a very abrupt increase of nitrogen uptake is observed and for longer processing times the nitrogen-content decreases again. The nitrogen content of specimens nitridated at 900◦ C showed a maximum value of 9.1 mol% after 0.5 h of reaction time. This value is comparable to 10 mol% reported for nitridated barium titanate in literature [15]. After reaching this maximum value, the nitrogen content gradually decreases again with ongoing reaction time at 900◦ C. Oxynitride perovskites are generally known to be meta-stable at elevated temperatures [20], thus this decrease is believed to originate from the decomposition of the oxynitride. At temperatures lower than 900◦ C (e.g. 750 ◦ C), nitridation develops more continuously and the nitrogen content saturates at a certain level depending on the reaction temperature. There occurs no significant decrease after the

2722

Y. Otsuka et al. / Journal of the European Ceramic Society 36 (2016) 2719–2725  

7* Z W

 E



D 

F

 















R

7HPSHUDWXUH &

Fig. 6. Thermogravimetric measurement of (a) an original non nitridated La0.003 Ba0.997 TiO3 sample as reference under dried air, La0.003 Ba0.997 TiO3 nitridated at 900◦ C for 2 h with 100 sccm of ammonia under (b) dried air, and (c) a gas mixture of Ar and 4% of H2 . Fig. 4. Nitrogen content of nitridated La0.003 Ba0.997 TiO3 as a function of reaction time at different reaction temperatures with 100 sccm of ammonia flow.

saturation. Not only temperature but also the flow rate of ammonia affected the nitrogen uptake (Fig. 5). In all investigated cases the nitrogen-content of the nitridated powders first raises with increasing reaction time and then falls again. For short reaction times (below 30 min) higher NH3 -flow rates result in higher values of the nitrogen-uptake. On the other hand the opposite is true for longer nitridation times above 1 h. At 900 ◦ C and with 50 sccm of ammonia flow, the nitrogen content reached 19 mol% after 1 h, this was almost twice of the value of a specimen treated with 100 sccm of NH3 -supply. This increase could be attributed to the pronounced formation of secondary phases and will be discussed in chapter 3.4. No significant variations in nitrogen-content could be detected for different gassing condition at the top surface layer of the powder containing Al2 O3 crucible compared to deeper surface regions, indicating that the nitridation process of the powders was spatially homogenous.

It is worthy to note that some of the measured nitrogen contents were much higher than the La-content of 0.3 mol%, which means charge compensation of N3− anions is mainly realized by oxygen vacancies (refer to Eq. (3)). The reaction with nitrogen overcompensates the effect of donor doping with La. The influence of further processing parameters during ammonolysis on the nitrogen uptake investigated in the present study were the effect of diluting NH3 with nitrogen gas, which turned out to be beneficial on the exchange of lattice oxygen by N3− anions and the quenching conditions in the cooling zone of the reactor, which do not seem to strongly affect the nitrogen-content of the nitridated powders. Generally it can be concluded that the development of the nitrogen-concentration determined for the reaction products after ammonolysis is a result of two competing processes: uptake of nitrogen and release of nitrogen due to a decomposition reaction, that occurs apparently and preferentially at higher reaction temperatures and lower ammonia flow rate.

θ

Fig. 5. Nitrogen content in nitridated La0.003 Ba0.997 TiO3 as a function of reaction time with different ammonia flow rates at a fixed reaction temperature of 900◦ C.

Fig. 7. XRD patterns of La0.003 Ba0.997 TiO3 powders in their original condition before nitridation (black curves: 0 h) and after nitridation at 900◦ C with 100 sccm of ammonia flow for different reaction times (red curves: 0.25 h; green curve: 0.5 h; blue curve: 16 h). The inset is a magnified view on the (200) and (002) Bragg reflections of the tetragonal barium titanate phase.

Y. Otsuka et al. / Journal of the European Ceramic Society 36 (2016) 2719–2725

(c)

(b)

(a)

2723

but such change is difficult to be seen in XRD pattern because of the overlap of peaks. However, the main crystallographic phase was not changed by ammonolysis, slight contents of a secondary phase were detected in specimens exposed to air after nitridation but not in specimen not exposed to air after nitridation (Fig. 8). This secondary phase was attributed to barium carbonate BaCO3 . Barium carbonate is expected to arise from barium orthotitanate (Ba2 TiO4 ) because barium orthotitanate easily reacts with air humidity and forms barium hydroxide Ba(OH)2 ; then barium hydroxide reacts with carbon dioxide CO2 from the air and forms barium carbonate. Such barium hydroxide formation was indirectly

Fig. 8. XRD pattern of La0.003 Ba0.997 TiO3 powder, (a) non-nitridated, nitridated at 900◦ C with 100 sccm of ammonia flow for 1 h (b) without exposure to air, (c) after exposure to air.

3.2. Thermal stability During thermogravimetric analysis, a nitridated specimen previously treated at 900 ◦ C for 2 h under a flow of 100 sccm of ammonia started to increase its weight around 350◦ C under dried air (Fig. 6). Such weight increase was neither observed in the case of nonnitridated specimen measured under dried air nor in the case of a nitridated specimen measured under a dried reducing gas mixture of Ar-H2 (4%). From this experimental evidence, the weight increase can be attributed to the substitution of nitrogen by oxygen, due to its higher atomic weight than nitrogen: nitrogen is chemically desorbed and oxygen absorbed instead. The temperature at which this substitution starts is comparable to the value at which the exchange of nitrogen with oxygen in an analogous material, LaTiO2 N, is reported to initiate [21]. The nitridated specimen decreased more weight than non-nitridated specimen at temperature above 500◦ C under dried air. As will be discussed in the next chapter, nitridated specimen contains barium carbonate which was formed after ammonolysis. Barium carbonate in nitridated specimen was confirmed to decompose at temperatures between 400◦ C and 600◦ C under air by XRD measurement. Therefore it can be concluded that the additional weight decrease of nitridated specimen at temperature above 500◦ C under air was originated from the decomposition of barium carbonate. 3.3. Crystallography of nitridated La-doped barium titanate Nitridated La0.003 Ba0.997 TiO3 keeps its crystallographic structure as originally tetragonal phase. A typical example (nitridated at 900◦ C with an ammonia flow rate of 100 sccm) is shown in Fig. 7. In consequence, peak positions of X-ray reflections observed for specimens before and after nitridation were same within the experimental resolution, (for example, see Fig. 7 inset). This means that the lattice parameter was not changed by nitridation under these conditions, although the ionic radius of nitrogen is larger (1.5 Å) than that of oxygen (1.4 Å). The amount of nitrogen might be too low to affect on the lattice parameters. An another explanation for this result might be the possibility, that only surface regions of particles within the nitridated powders reacted upon ammonolysis and consequently are enriched in nitrogen. It is known that the surface of barium titanate particle is composed of pseudo-cubic phase [22]. If the nitridation occurred at only surface of La-doped barium titanate, lattice parameter of pseudo-cubic phase might change,

Fig. 9. SEM image of La0.003 Ba0.997 TiO3 powders (a) non nitridated, nitridated with (b) 50 sccm or (c) 200 sccm of ammonia flow for 2 h at 900◦ C. The Insets show macroscopic images of powder revealing a clear difference in color.

2724

Y. Otsuka et al. / Journal of the European Ceramic Society 36 (2016) 2719–2725

,QWHQVLW\ DX

%D&2

E D





 θ GHJUHH





Fig. 10. XRD pattern of La0.003 Ba0.997 TiO3 powder nitridated at 900◦ C for two hour with (a) 200 sccm of ammonia flow and (b) 50 sccm of ammonia flow. Both samples were exposed to air before the measurement.

confirmed by the alkaline character (pH > 7) of suspensions of nitridated powders in water. 3.4. Powder morphology Thermal ammonolysis strongly changed the morphology of the particles of the reacted powders (Fig. 9). Rectangularly shaped pores were occasionally formed after nitridation. They seem to be the result of a leaching mechanism by which “etching pits” are formed. In the initial phase of the nitridation reaction at elevated temperatures (e.g. 900 ◦ C) only very few of these pores are formed. After longer reaction times well above 30 min, the frequency of these pores, however, drastically increased resulting in a significant increase of the SSA, as determined by BETmethod. The typical SSA value of 3.1 m2 /g for a raw powder raised up to 5.8 m2 /g after 2 h of ammonolysis at 900 ◦ C. In the case of relatively low reaction temperatures (e.g. 800 ◦ C) pores can also be formed but it takes much more time. If the temperature is reduced to 750 ◦ C even at such a long gassing duration as 16 h porosity does not seem to appear. The frequency of the rectangular pores increased with decreasing ammonia flow rate for thermal ammonolysis. In other words, ammonolysis under stronger reducing conditions which resulted from the intense decomposition of ammonia because of longer retention periods of time of this nitridation agent gas in the hot reactor at lower flow rates created more pores. This effect manifests itself also in a distinct reduction of SSA. (Example: ammonolysis for 2 h at 900 ◦ C with a flow rate of 50 sccm resulted in a SSA of 6.6 m2 /g whereas a flow rate of 200 sccm caused a SSA of only 4.6 m2 /g). Diluting the ammonia flow with nitrogen gas N2 of increasing the La-content of the powder up to 2 at.-% did not seem to affect the phenomenon of pore formation. It was found that specimens with higher frequency of pores also showed higher peak values of barium carbonate Bragg reflections in XRD (Fig. 10). Directly after the ammonolysis reaction was completed (e.g. after reaction times ranging from 30 min up to 4 h at 900 ◦ C) no XRD reflections of BaCO3 could be detected. They only appeared after a subsequent longer exposure to air. Similar rectangular pores appeared in La-doped barium titanate annealed in reducing conditions. They were found to originate from leached out precipitates of barium orthotitnate [23]. Therefore it is

believed that the rectangular pores in nitridated barium titanate are also an evidence for the formation of barium orthotitanate. Powders with higher amounts or volumes of pores were more bluish in color. Such bluish color may come from TiOx Ny [24], because the formation of barium orthotitanate coincides with the Ti-enriched phases like TiO2 . TiO2 can react with ammonia, forming TiOx Ny [25], according to the following formalism: 2BaTiO3 ↔ Ba2 TiO4 + TiO2

(4)

TiO2 + xNH3 ↔ TiO2−1.5x Nx + 1.5 × H2 O

(5)

The formation of Ba2 TiO4 and TiO2 from BaTiO3 [23] arises from the reducing effect of the gas during amonolysis. Increased nitrogen contents of specimens nitridated with 50 sccm of ammonia flow rate in Fig. 5 can be explained by the pronounced formation of barium orthotitanate because nitridated barium orthotitanate can contain more higher amounts of nitrogen than barium titanate [15]. Based on the results it can be concluded that relative low ammonolysis temperature (below 800 ◦ C) and suppression of thermal decomposition of ammonia gas are essential to realize a nitridated La-doped barium titanate as a single phase by avoiding the formation of secondary phase and the decomposition of the nitridated La-doped barium titanate. The effect of nitridation on the electrical conductivity and dielectric properties of La-doped barium titanate are under investigation. 4. Summary A new concept for nitridation of BaTiO3 -based materials via thermal ammonolysis was studied systematically in this work. Nitrogen uptake of La-doped barium titanate by thermal ammonolysis was confirmed by chemical and thermogravimetric methods. The amount of nitrogen in nitridated La-doped barium titanate was higher than the expected theoretical value deduced from the assumption of pure and only charge compensation of nitrogen anion No  by only lanthanum cation LaBa · . Thermogravimetric studies suggested that nitrogen exchanged by oxygen at annealing temperatures above 350◦ C. Nitrogen uptake didn’t affect the crystallographic tetragonal structure of BaTiO3 . Nitridated Ladoped barium titanate kept its tetragonal phase and the tetragonal

Y. Otsuka et al. / Journal of the European Ceramic Society 36 (2016) 2719–2725

distortion of the unit cell was not changed. Barium orthotitanate as a secondary phase formed eventually by thermal ammonolysis and the amount of this impurity increased with stronger reducing condition upon ammonolysis. References [1] Y. Sakabe, Y. Hamaji, T. Nishiyama, New barium titanate based material for MLCs with Ni Electrode, Ferroelectrics 133 (1992) 133–138. [2] Y. Sakabe, N. Wada, T. Hiramatsu, T. Tonogaki, Fine-grained BaTiO3 ceramics doped with CaO, Jpn. J. Appl. Phys. 41 (2002) 6922–6925. [3] Y. Sakabe, H. Takagi, Nonreducible mechanism of {(Ba1-x Cax )O}m TiO2 (m > 1) ceramics, Jpn. J. Appl. Phys. 41 (2002) 6461–6465. [4] S. Suzuki, T. Takeda, A. Ando, H. Takagi, Ferroelectric phase transition in Sn2+ ions doped (Ba, Ca)TiO3 ceramics, Appl. Phys. Lett. 96 (2010) 132903. [5] C. Pithan, Y. Shiratori, R. Waser, J. Dornseiffer, F.-H. Haegel, Preparation, processing, and characterization of nano-crystalline BaTiO3 powders and ceramics derived from microemulsion-mediated synthesis, J. Am. Ceram. Soc. 89 (2006) 2908–2916. [6] M. Jansen, H.P. Letschert, Inorganic yellow-red pigments without toxic metals, Nature 404 (2000) 980–982. [7] M. Higashi, R. Abe, T. Takata, K. Domen, Photocatalytic overall water splitting under visible light using ATaO2 N (A = Ca, Sr, Ba) and WO3 in a IO3− /I− shuttle redox mediated system, Chem. Mater. 21 (2009) 1543–1549. [8] Y.-R. Zhang, T. Motohashi, Y. Masubuchi, S. Kikkawa, Sintering and dielectric properties of perovskite SrTaO2 N ceramics, J. Eur. Ceram. Soc. 32 (2012) 1269–1274. [9] M. Yang, J. Oro-Sole, A. Kusmartseva, A. Fuertets, J.P. Attfield, Electronic tuning of two metals and colossal magnetoresistances in EuWO1+x N2-x perovskites, J. Am. Chem. Soc. 132 (2010) 4822–4829. [10] Y.-I. Kim, P.M. Woodward, K.Z. Baba-Kishi, C.W. Tai, Characterization of the structural, optical, and dielectric properties of oxynitride perovskites AMO2 N (A = Ba, Sr, Ca; M = Ta Nb), Chem. Mater. 16 (2004) 1267–1276. [11] Y.-I. Kim, P.M. Woodward, Syntheses and characterizations of complex perovskite oxynitrides LaMg1/3Ta2/3O2N, LaMg1/2 Ta1/2 O5/2 N1/2 LaMg1/3 Ta2/3 O2 N, and BaSc0.05 Ta0.95 O2.1 N0.9 , J. Sol. State Chem. 180 (2007) 3224–3233.

2725

[12] A. Ziani, C.L. Paven-Thivert, L.L. Gendre, D. Fasquelle, J.C. Carru, F. Tessier, J. Pinel, Structural and dielectric properties of oxynitride perovskite LaTiOx Ny thin films, Thin Sol. Films 517 (2008) 544–549. [13] Y. Sakabe, K. Minai, K. Wakino, High-dielectric constant ceramics for base metal monolithic capacitors, Jpn. J. Appl. Phys. 20 (Suppl. 20–4) (1981) 147. [14] O. Saburi, Properties of semiconductive barium titanates, J. Phys. Soc. Japn. 14 (1959) 1159–1174. [15] T. Bräuniger, T. Müller, A. Pampel, H.-P. Abicht, Study of oxygen-nitrogen replacement in BaTiO3 by 14 N solid-state nuclear magnetic resonance, Chem. Mater. 17 (2005) 4114–4117. [16] T. Müller, T. Großmann, H.-P. Abicht, Nitrogen containing barium titanate: preparation and characterisation, J. Phys. Chem. Solids 70 (2009) 1093–1097. [17] S.G. Ebbinghaus, H.-P. Abicht, R. Dronskowski, T. Müller, A. Reller, A. Weidenkaff, Perovskite-related oxynitrides—recent developments in synthesis, characterisation and investigations of physical properties, Prog. Solid State Chem. 37 (2009) 173–205. [18] W. Chase, NIST-jANAF themochemical tables, Fourth Edition, American Institute of Physics, 1998. [19] F.A. Kröger, H.J. Vink, Solid State Physics, vol. 3, Academic Press, New York, 1956. [20] R. Aguilar, D. Logvinovich, A. Weidenkaff, A. Reller, S.G. Ebbinghaus, Thermal oxidation of oxynitride perovskites in different atmospheres, Thermochim. Acta 471 (2008) 55–60. [21] D. Logvinovich, L. Bocher, D. Shptyakov, R. Figi, S.G. Ebbinghaus, R. Aguiar, M.H. Aguirre, A. Reller, A. Weidenkaff, Microstructure, surface composition and chemical stability of partly ordered LaTiO2 N, Solid State Sci. 11 (2009) 1513–1519. [22] T. Takeuchi, K. Ado, T. Asai, H. Kageyama, Y. Saito, C. Masquelier, O. Nakamura, Thickness of cubic surface phase on barium titanate single-crystalline grains, J. Am. Ceram. Soc. 77 (1994) 1665–1668. [23] D. Makovec, M. Dorofenik, Microstructural changes during the Reduction/Reoxidation process in donor-doped BaTiO3 ceramics, J. Am. Ceram. Soc. 83 (2000) 2593–2599. [24] F. Cheviré, F. Tessier, R. Marchand, Optical properties of the perovskite solid solution LaTiO2 N?ATiO3 (A = Sr, Ba), Eur. J. Inorg. Chem. (2006) 1223–1230. [25] R. Asahi, T. Morikawa, K. Ohwaki, K. Aoki, Y. Taga, Visible-light photocatalysis in nitrogen-doped titanium oxides, Science 293 (2001) 269–271.