- Email: [email protected]
Thermal stability and high temperature polymorphism of topochemically-prepared Dion–Jacobson triple-layered perovskites
Journal of Alloys and Compounds 647 (2015) 370e374
Contents lists available at ScienceDirect
Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom
Thermal stability and high temperature polymorphism of topochemically-prepared DioneJacobson triple-layered perovskites Stephen L. Guertin, Elisha A. Josepha, Dariush Montasserasadi, John B. Wiley* Department of Chemistry and Advanced Materials Research Institute, University of New Orleans, 2000 Lakeshore Dr., New Orleans, LA 70148, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 4 May 2015 Accepted 5 June 2015 Available online 10 June 2015
The thermal stability of six DioneJacobson-related triple layered perovskites, ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl), was explored to 1000 C. Each compound was produced topochemically by low-temperature (<500 C) ion exchange from RbCa2Nb3O10. The thermal behavior of the series was examined by variable temperature X-ray powder diffraction experiments in tandem with thermogravimetric analysis and differential scanning calorimetry. Five of the species were found to be low temperature/metastable phases, decomposing below 900 C, where the stability of the series decreased with decreasing interlayer cation size. The compounds, A ¼ Li, Na, K, exhibited high temperature polymorphism, with a completely reversible transition evident for KCa2Nb3O10. © 2015 Elsevier B.V. All rights reserved.
Keywords: Oxide materials Preferential site ordering Phase transitions Calorimetry X-ray diffraction
1. Introduction Layered perovskite compounds have been extensively studied due to their wide-ranging potential for applications, as well as for their susceptibility to soft chemical manipulation [1e4]. Many high temperature superconductors (e.g., YBa2Cu3O7 and Tl2Ba2Ca2Cu3O10) share perovskite-related structures [5e7], and layered perovskite materials have been variously utilized in photocatalysts [8e11], hydrogen sensors [12e14], and fatigue-free capacitors [15e18]. Such materials offer a great versatility with respect to solid state synthesis, allowing important electronic and ferroic properties to be tailored with convenience and ease [4]. Nanostructuring of related perovskites is another exciting new area of research [19]. Soft chemistry (chimie douce) involves the use of lowtemperature synthetic methods to form desired products that cannot be prepared by traditional high temperature ceramic techniques. For many compounds, one-step, direct synthesis from simple precursors is thermodynamically unfavored. In such cases, soft chemical topotactic routes such as ion exchange and intercalation may be used to retain resilient structural features of a precursor in the products. The strata in layered perovskite oxides can be utilized in this manner to create families of metastable materials
* Corresponding author. E-mail address: [email protected] (J.B. Wiley). http://dx.doi.org/10.1016/j.jallcom.2015.06.045 0925-8388/© 2015 Elsevier B.V. All rights reserved.
with variant interlayer chemistry [20]. These can, in turn, be used as secondary reagents in multistep synthesis [3]. Layered perovskites of the DioneJacobson (DJ) family have the general formula A[A0 n1BnO3nþ1], where typically A ¼ alkali metal, A' ¼ rare earth or alkaline-earth, B ¼ transition metal, and n ¼ number of perovskite layers. This family has been well characterized, including reports on their structures [21e25], electronics [26e28], and magnetics [23,29,30]. In the case of a doped potassium variant, superconductive properties have been observed [31]. The utility and development of topochemical reaction strategies in part relies on determining the effective window of thermal stability for the reagents and intermediates. By defining the limits of these systems and their modifications, strategies can be designed to more effectively access target materials. Recent studies have examined the stability of double-layered perovskites [32,33]. Of the triple-layered species, only HCa2Nb3O10 and NaCa2Nb3O10 have been explicitly probed above 500 C [23,34]. In the research reported herein, the thermal stability of the DioneJacobson perovskite series ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl) was investigated. These species can be categorized into three classes based on relative thermal stability and similarity of rearrangement/ decomposition pathways. Most interestingly, members of the class ACa2Nb3O10 (A ¼ Li, Na, K) were discovered to undergo endothermic rearrangement into high temperature polymorphs, only one of which has been previously identified [23, 38].
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
2. Experimental 2.1. Synthesis RbCa2Nb3O10 was synthesized by solid state reaction. Rb2CO3 (Alfa Aesar, 99.8%), CaCO3 (Alfa Aesar, 99.99%), and Nb2O5 (Alfa Aesar, 99.9985%) were finely ground together in the molar ratio of 1.25:1:1, respectively. The mixture was heated in an alumina crucible at 850 C for 24 h, re-ground, and then heated at 1050 C for 24 h. The product was washed thoroughly with distilled water followed by acetone and oven-dried at 110 C for 4 h. KCa2Nb3O10 was also synthesized by direct reaction for comparison to the ion exchanged species. K2CO3 (Alfa Aesar, 99.997%), CaCO3 (Alfa Aesar, 99.99%), and Nb2O5 (Alfa Aesar, 99.9985%) were ground together and pressed into 7 mm pellets by hand press (SigmaeAldrich). The pellets were heated at 850 C for 24 h and 1200 C for two periods of 4 h with intermittent grinding and hand-press pelletization between each step. RbCa2Nb3O10 was used for ion-exchange with various monovalent cationic species to prepare the series ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl) as detailed in Table 1. For the ammonium and alkali-metal species, the parent was finely ground with the corresponding nitrate reagent (NH4NO3, Alfa Aesar, 99.999%; LiNO3, Alfa Aesar, 99.99%; NaNO3, Alfa Aesar, 99.999%; KNO3, Alfa Aesar, 99.994%) and placed into an alumina crucible. The exchange reactions were carried out for 4 days in an ANO3 molten salt with the ratio 1:10 for ammonium and lithium and 1:20 for sodium and potassium, at temperatures of 200 C, 350 C, 400 C, and 400 C, respectively. To synthesize (CuCl)Ca2Nb3O10, RbCa2Nb3O10 and anhydrous CuCl2 (Alfa Aesar, 99.995%) were ground together in a drybox, hand-pressed into pellets, sealed under vacuum in a Pyrex tube and reacted at 325 C for 10 days. Each of this series of products was washed and dried in the same fashion as the starting material. HCa2Nb3O10 was formed by stirring RbCa2Nb3O10 in 6 M HNO3 (aq) at 60 C for 4 days, with replenishment of the acid solution after 3 days. This product required isolation by centrifugation for the washing step because much of the solid was fine enough to permeate standard filter paper. 2.2. Characterization Variable temperature X-ray powder diffraction (XRD) data were collected on a Philips X-Pert PW 3020 MPD diffractometer (Cu Ka radiation, l ¼ 1.5418 Å) equipped with a curved graphite monochromator and an Anton Paar HTK1600 variable temperature stage. To determine the structural evolution of the compounds with increasing temperature, XRD data were taken from 25 C to 1000 C under flowing nitrogen in 50 C heating increments. Kinetically hindered transitions were studied with longer periods of heat treatment, ranging from 6 to 24 h. To isolate endothermic and exothermic events associated with phase changes, the materials were examined by differential scanning calorimetry (DSC) on a Table 1 Summary of ion-exchange reactions with RbCa2Nb3O10 host. Ion-exchanged product HCa2Nb3O10$(H2O)n (NH4)Ca2Nb3O10 LiCa2Nb3O10 b-NaCa2Nb3O10∙(H2O)n KCa2Nb3O10 (CuCl)Ca2Nb3O10 a b c
6 M HCl. ANO3 (A ¼ NH4, Li, Na, K). Anhydrous CuCl2.
Molar ratio a
Excess 1:10b 1:10b 1:20b 1:20b 1:2c
Temperature ( C)
Time (days)
60 200 350 400 400 325
4 4 3 7 7 10
371
Netzsch 404S thermal analysis system. Samples were heated and cooled in alumina or platinum pans at 10 C/min to 1000 C under flowing argon (50 m[⁄min). Thermogravimetric analysis (TGA) in tandem with DSC was used to monitor weight changes during decomposition on a TA Instruments SDT Q600 V8.3 system. Samples were characterized by scanning electron microscopy on a JEOL JSM-5410 SEM and electron dispersive spectroscopy (EDS) EDAX DX-PRIME microanalytical system to examine morphology and chemical composition at various stages of heating. The peak positions and lattice parameters were refined by a least-squares method with the programs ChekCell and FULLPROF [39, 40]. 3. Results RbCa2Nb3O10 was indexed in the tetragonal system with space group P4/mmm (#123). The lattice parameters presented in Table 2 were refined using the method of least squares refinement and were found to be in good agreement with previous reports [35]. The lattice parameters of ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl) were also consistent with literature values (Table 2), and EDS analysis showed minimal amounts of residual rubidium ion (Rb:Nb < 0.05:1.00). Two of the species were found susceptible to hydration. At 25 C, HCa2Nb3O10∙(H2O)n and b-NaCa2Nb3O10∙(H2O)n intercalate variable amounts of water inside the interlayer, increasing the magnitude of the c lattice parameter by up to 1.5 Å. The degree of hydration in air is strongly dependent on ambient humidity [23]. For HCa2Nb3O10, endothermic dehydration is observed between 50 C and 125 C with a mass loss of 3.3% corresponding to ~1.9 waters of hydration. The event is correlated with the shifting of the [001] reflection to a higher angle and a dramatic drop in its intensity. Dehydration of the sodium species requires higher temperatures. A pronounced endothermic signal is observed centered on 150 C, accompanied by a mass loss of 3.5% (~1.1 waters of hydration) and a significant [001] reflection shift. b-NaCa2Nb3O10 continues to lose mass up to 400 C, but above 225 C the weight loss is minimal and is not accompanied by observable structural rearrangement. Each of the compounds ACa2Nb3O10 (A ¼ NH4, Li, Na, K, CuCl) also demonstrate small degrees of hydration, measuring less than 1% of the total sample weight. The volatilizations, however, are unaccompanied by crystallographic changes, indicating that such processes likely represent loss of surface water. ACa2Nb3O10 when A ¼ H, NH4 are the least thermally stable of the family. In the decomposition, the interlayer cations are removed from the perovskite lattice via formation of a corresponding oxide. In a concerted step, HCa2Nb3O10 evolves H2O (1.6% mass loss observed, theoretically 1.73%) and (NH4)Ca2Nb3O10 evolves NH3 and H2O (4.7% observed, theoretically 4.85%). These transitions are correlated with an exothermic DSC signal centered at 315 C for the hydrogen species, while the ammonium compound gives a sharper 350 C exotherm followed by an endotherm at 365 C. The residual salt in both cases is the poorly crystallized single phase Ca4Nb6O19, previously described by Fang and coworkers [31]. Major reflections of this tetragonal byproduct appear in the XRD data at about 300 C and 340 C, respectively. Upon further heating to 570 C, Ca4Nb6O19 begins to disproportionate to CaNb2O6 (JCPDS File No. 39-1392) and Ca2Nb2O7 (JCPDS File No. 42-2), accompanied by a DSC slope change in the exothermic direction. Members of the series ACa2Nb3O10 (A ¼ Li, Na) are found to be more stable than the A ¼ H and NH4 species, yet with sufficient heating undergo rearrangement and decomposition. In neither case is mass loss observed after dehydration. In the XRD of LiCa2Nb3O10, no change in crystal structure occurs up to 600 C. At 680 C DSC registers an endothermic event, correlated with a major shifting of reflections in the 650 C and 700 C diffractograms, representing the irreversible formation of an alternative polymorph of LiCa2Nb3O10.
372
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
Table 2 Lattice parameters of ion-exchanged species. Species
a (Å)
Exp.a RbCa2Nb3O10 Ref.b RbCa2Nb3O10 Exp. HCa2Nb3O10 Ref. HCa2Nb3O10 Exp. NH4Ca2Nb3O10 Ref. NH4Ca2Nb3O10 Exp. LiCa2Nb3O10 Ref. LiCa2Nb3O10 Exp. NaCa2Nb3O10 Ref. NaCa2Nb3O10 Exp. KCa2Nb3O10 Ref. KCa2Nb3O10 Exp. (CuCl)Ca2Nb3O10 Ref. (CuCl)Ca2Nb3O10
3.862(3) 3.85865(6) 3.853(3) 3.8536(6) 7.720(1) 7.724(2) 7.719(6) 7.720(7) 5.461(4) 5.4731(2) 7.741(5) 7.7418(6) 3.850(2) 3.84994(6)
a b
b (Å)
c (Å)
Layer spacing 14.93
e e e e e e e e e 7.7073(6) e e
14.9306(2) 14.9108(3) 14.381(1) 14.403(5) 14.993(1) 14.977(8) 28.305(6) 28.331(3) 28.935(1) 29.0138(9) 14.861(8) 14.859(1) 15.674(1) 15.6706(5)
Reference [35]
14.38 [21] 14.99 [38] 14.15 [38] 14.47 [23] 14.86 [24] 15.67 [37]
Exp. ¼ experimental values. Ref. ¼ literature values.
The new system exothermically decomposes at 730 C into Li2O, CaNb2O6, and Ca2Nb2O7. Reflections of lithium oxide were not observed in the diffraction pattern of the decomposed sample, but the other products were clearly present. Presumably, the invisibility of Li2O is due to a low degree of crystallinity and/or poor X-ray scattering capacity of second period elements. b-NaCa2Nb3O10 undergoes an irreversible phase change at high temperature into a-NaCa2Nb3O10, as previously reported [23,38]. Byeon and workers located this transition at 650 C. In the present study, DSC registers an endothermic event at 690 C, well correlated with changes in its XRD patterns. Like the low temperature species, the a-phase is found to form a hydrate in air at 25 C. At 800 C the polymorph begins to decompose into a two phase system consisting of NaNbO3 and NaCa4Nb5O17 [32,41]. The XRD data reveal the rapid appearance of major reflections of decomposition products near the transition temperature, but total decomposition appears kinetically limited and requires hours to reach completion. Similarly to the other alkali species, KCa2Nb3O10 transforms into an alternative polymorph at high temperatures. Notably, the high temperature structure lacks the very prominent (207) reflection (Fig. 1 of the low temperature species). The reflections collected from 15 to 40 2q strongly resemble those of tetragonal RbCa2Nb3O10, suggesting similarity of structure. As expected by analogy to the lithium and sodium variants, the rearrangement is accompanied by an endothermic trace. The transition temperature for KCa2Nb3O10 synthesized by direct reaction was found to be 1000 C, but this threshold is decreased by up to 200 C for the ionexchanged species, presumably due to the presence of residual rubidium ions in the interlayer. Independent batches of ion exchanged samples demonstrated divergent transition temperatures, while distinct batches of the direct product behaved identically. Most notably, when allowed to cool, the high temperature phase reverts to the low temperature structure, with this transition lagging 50 C lower than the heated transition. The sample was subjected to three consecutive heatingecooling cycles without change in transition temperatures or sample crystallinity. (CuCl)Ca2Nb3O10 exhibits a more complicated bonding scheme than the other compounds investigated, as the interlayer is occupied by an extended two-dimensional square planar copper (II) chloride framework rather than by a single cation. TGA revealed two major volatilization events. A steep 3.1% mass loss is centered at 550 C, and a second loss of 2.2% centered at 825 C. The first event is accompanied by a very gradual DSC slope change in the exothermic direction, while the latter registers a sharp exotherm. EDS analysis confirmed that the mass loss in both cases was due to removal of chloride from the structure (5.3% total mass loss
Fig. 1. Reversible phase transition of KCa2Nb3O10. a) Tetragonal RbCa2Nb3O10 at 25 C, b) K-exchanged RbCa2Nb3O10 at 950 C, c) 900 C, d) 850 C, e) 800 C, f) 750 C, g) 700 C, h) 650 C, and i) 600 C.
observed, theoretically 5.74%). When the sample was heated in an alumina pan, the alumina surface in contact with the sample was found to be scorched to a dark brown hue, possibly due to oxideechloride exchange. Experiments carried out with platinum containment showed no visual evidence of such interaction. XRD demonstrated major structural transitions in good agreement with DSC-TGA data. No previously reported diffraction patterns could be found for comparison with the intermediate species, which appears to be a poorly crystallized, complex material. Above 825 C, Ca2Nb2O7 is clearly present, but CaNb2O6 is absent, distinguishing this process from those of ACa2Nb3O10 (A ¼ H, NH4, Li). Intact (CuCl) Ca2Nb3O10 is light green in color. The anhydrous intermediate species (heating arrested at T ¼ 700 C) appears unevenly olive green and light grey, and the anhydrous terminal decomposition products are light gray.
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
373
Table 3 Summary of thermal behavior of the series ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl). Species
DSC event temperature
Reaction
HCa2Nb3O10
300 C 550 C 340 C 550 C 680 C 730 C 690 C 800 C (slow) 1000 C 800 C e 900 C 330 C 550 C
2 HCa2Nb3O10 / Ca4Nb6O19 þ H2O (g) Ca4Nb6O19 / 2 CaNb2O6 þ Ca2Nb2O7 2 NH4Ca2Nb3O10 / Ca4Nb6O19 þ 2NH3 (g) þ H2O (g) Ca4Nb6O19 / 2 CaNb2O6 þ Ca2Nb2O7 LiCa2Nb3O10 (LT)* 4 LiCa2Nb3O10 (HT)** 2 LiCa2Nb3O10 / 2 CaNb2O6 þ Ca2Nb2O7 þ Li2O b-NaCa2Nb3O10 4 a-NaCa2Nb3O10 (HT) 2 a-NaCa2Nb3O10 (HT) / NaNbO3 þ NaCa4Nb5O17 KCa2Nb3O10 (LT) 4 KCa2Nb3O10 (HT) KCa2Nb3O10 (LT) 4 KCa2Nb3O10 (HT) (CuCl)Ca2Nb3O10 / unidentified intermediate Unidentified intermediate / Ca2Nb2O7 þ unknown phase
(NH4)Ca2Nb3O10 LiCa2Nb3O10
b-NaCa2Nb3O10 KCa2Nb3O10 (Direct) KCa2Nb3O10 (Ion-exchanged) (CuCl)Ca2Nb3O10
*LT e low temperature phase; **HT e high temperature phase; 4 indicates a reversible process.
A reaction scheme for all examined compounds is summarized in Table 3. Fig. 2 displays an XRD stack plot of the series ACa2Nb3O10 (A ¼ H, Li, NH4, CuCl) after being maintained at 1000 C for five days compared to the JCPDS reference patterns of the terminal products. 4. Discussion Rationalization of the thermal stability of the series is largely dependent on the size of the interlayer cations. The utility of layered perovskites derives from the chemical resilience of the A'BO3 motif at extreme temperatures and pH values. The perovskite “slabs” can become chemical templates for affecting difficult, desirable transformations. The region of chemical and mechanical susceptibility, as in the ion exchange reactions of this study, is the interlayer. Larger, chemically softer cations, can be expected to coordinate greater numbers of oxide ions, thereby stabilizing anionic repulsion across the interlayer. Consistent with this model, the ammonium-substituted structure was found to be somewhat more stable than its protic counterpart, and the alkali series increased in stability down the group. The structure of LiCa2Nb3O10 was first studied in 1981 by Dion et al., who characterized the compound in the tetragonal system with a ¼ 7.720(7) Å and c ¼ 28.331(3) Å [38]. In 2001, Ishizawa and coworkers presented a synchrotron X-ray study of the material
which indicated triclinic P1 symmetry with a ¼ 5.4809(3) Å, b ¼ 5.4804(3) Å, c ¼ 26.5533(16) Å, a ¼ 89.999(4) , b ¼ 90.245(4) , g ¼ 89.999(5) [22]. The latter paper has since been cited as the standard model for LiCa2Nb3O10 structure. Interestingly, the present study identifies the diffraction pattern of the Ishizawa structure with that of the high temperature polymorph. Ion exchange synthesis at 350 C produces the low temperature phase, with calculated lattice parameters in line with the results of Dion et al. [38]. Ishizawa and coworkers, however, produced single crystals of the material in a molten flux with a much higher reaction temperature program. The present study implies that the synchrotron X-ray examination was performed on the high temperature phase. The reversible phase change in ceramically synthesized KCa2Nb3O10 correlates very well with a thermoelastic transition of the compound at 1000 C reported by Dion et al. [36], and presumably this structural rearrangement is responsible for the bulk phenomenon. In the same paper, Dion describes a thermoelastic transition for RbCa2Nb3O10 occurring at 620 C. Small occupancy of potassium sites by rubidium in KCa2Nb3O10 (Rb < 5%) greatly lowered the polymorphic transition temperature by as much as 200 C (Table 3). The potassium compound has been reported as 6coordinate [24], while the rubidium has been identified as 8coordinate [35]. Fig. 3 displays a proposed model for the low and high temperature polymorphs. If the high temperature phase
Fig. 2. Stack plot of terminal products versus CaNb2O6 and Ca2Nb2O7. a) HCa2Nb3O10, b) LiCa2Nb3O10, c) NH4Ca2Nb3O10, d) (CuCl)Ca2Nb3O10, e) CaNb2O6 and f) Ca2Nb2O7.
374
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
Fig. 3. Proposed structural transition of KCa2Nb3O10. Left: Low temperature polymorph. Right: High temperature polymorph.
features 8-coordinate potassium, then residual rubidium would be expected to facilitate the crystallographic transition. Future studies on the solid solution KnRb1nCa2Nb3O10 across the range 0 n 1 would help in the exploration of trends in the range 500 C < T < 1000 C. 5. Conclusions The thermal stability of the triple-layered calcium niobate perovskite family depends fundamentally on its interlayer chemistry. Reaction schemes intended to incorporate the triple-layered perovskite structural framework at T > 500 C must remain cognizant of the specific integrities of precursor species to ensure that rearrangements do not impede synthetic goals. Alternatively, knowledge of decomposition and transition temperature thresholds could be exploited to access novel materials. Acknowledgments This study has been supported by the National Science Foundation (DMR1005856). References [1] A.S. Bhalla, Ruyan Guo, Rustum Roy, Mater. Res. Innov. 4 (2000) 3e26. [2] M. Johnsson, P. Lemmens, Handbook of Magnetism and Advanced Magnetic Materials, John Wiley and Sons, New York, 2007. [3] K.G. Ranmohotti, E. Josepha, J. Choi, J. Zhang, J.B. Wiley, Adv. Mater. 23 (4) (2011) 442e460.
[4] R.E. Schaak, T.E. Mallouk, Chem. Mater. 14 (4) (2002) 1455e1471. [5] Neeraj Khare, Handbook of High Temperature Superconductor Electronics, CRC Press, 2003. [6] C.N.R. Rao, P. Ganguly, K. Sreedhar, R.A. Mohan Ram, P.R. Sarode, Mater. Res. Bull. 22 (6) (1987) 849e855. [7] C.C. Torardi, M.A. Subramanian, J.C. Calabrese, J. Gopalakrishnan, K.J. Morrissey, T.R. Askew, R.B. Flippen, U. Chowdry, S.W. Sleight, Science 240 (4852) (1988) 631e634. [8] Jindo Kim, Hwang Dong, Hyun Kim, Sang Bae, Jae Lee Wei, Se Oh Li, Top. Catal. 35 (3e4) (2005) 295e303. [9] Masato Machida, Jun-Ichi Yabunaka, Tsuyoshi Kijima, Chem. Mater. 12 (3) (2000) 812e817. [10] Shunsuke Nishimoto, Yoshihiro Okazaki, Motohide Matsuda, Michihiro Miyake, J. Ceram. Soc. Jpn. 117 (2009) 1175e1179. [11] Tsuyoshi Takata, Yoko Furumi, Kiyoaki Shinohara, Akira Tanaka, Michikazu Hara, Junko N. Kondo, Kazunari Domen, Chem. Mater. 9 (5) (1997) 1063e1064. [12] M. Sakthivel, W. Weppner, Sens. Actuators B 125 (2) (2007) 435e440. [13] Peng Song, Jifan Hu, Hongwei Qin, Ling Zhang, Tianfeng Xing, Xue Liu, Shanxing Huang, Minhua Jiang, J. SoleGel Sci. Tech. 35 (1) (2005) 65e68. [14] Xiling Tang, Kurtis Remmel, Daniel Sandker, Zhi Xu, Junhang Dong, Xudong Fan, Hai Xiao, Anbo Wang (Eds.), Proceedings of the SPIE, 2010, p. 7682, 76820L-76820L-7. [15] C. A-Paz de Araujo, J.D. Cuchiaro, L.D. McMillan, M.C. Scott, J.F. Scott, Nature. 374 (1994) 627e629. [16] B.H. Park, B.S. Kang, S.D. Bu, T.W. Noh, J. Korean Phys. Soc. 1999 (1999) 35. S13067eS1309. [17] S.R. Shannigrahi, Hyun Jang, Appl. Phys. Lett. 79 (2001) 1051. ~ es, B.D. Stoyanovic, M.A. Ramirez, A.A. Cavalheiro, E. Longo, [18] A.Z. Simo J.A. Varela, Ceram. Int. 34 (2008) 257e261. [19] Debasish Mohanty, Girija S. Chaubey, Amin Yourdkhani, Shiva Adireddy, Gabriel Caruntu, John B. Wiley, RSC Adv. 2 (2012) 1913e1916. [20] Jacques Livage, New J. Chem. 25 (2001) 1. [21] Margaret J. Geselbracht, Helen K. White, Jeanette M. Blain, Miranda J. Dia, Jennifer L. Hubbs, Nicole Adelstein, Joshua A. Kurzman, Mater. Res. Bull. 46 (2011) 398e406. [22] Nobuo Ishizawa, Reiko Yamashita, Shuji Oishi, James R. Hester, Shunji Kishimoto, Acta Cryst. C57 (2001) 1006e1009. [23] Song-Ho Byeon, Jin Hee Kim, Dong-Kuk Kim, Nam Hwi Hur, Chem. Mater. 15 (2) (2003) 383e389. [24] Toshiaki Tokumitsu, Kenji Toda, Tsukuru Daiki Sakuraba, Kazuyoshi Uematsu, Mineo Sato, JCS-Jpn. 114 (9) (2006) 795e797. [25] Ramesh Chitrakar, Ind. Eng. Chem. Res. 47 (1) (2008) 176e179. [26] Venkataraman Thangadurai, Werner Weppner, J. Mater. Chem. 11 (2001) 636e639. [27] Venkataraman Thangadurai, Werner Weppner, Chem. Mater. 14 (2002) 1136e1143. [28] Yoshihiko Takano, Yoshihide Kimishima, Tokio Yamadaya, Shinji Ogawa, Nobuo Mori, Rev. High. Press. Sci. Technol. 7 (1998) 589e591. [29] X.L. Wang, E. Takayama-Muramaki, Phys. Rev. B: Condens. Matter Mater. Phys. 72 (2005) 064401. [30] Yoshihiko Takano, Shigeru Takayanagi, Shinji Ogawa, Tokio Yamadaya, ^ri, Solid State Commun. 103 (4) (1997) 215e217. Nobuo Mo [31] Mingming Fang, Chy Hyung Kim, Thomas E. Mallouk, Chem. Mater. 11 (1999). [32] F.J. Zuniga, J. Darriet, Acta Cryst. C59 (2003) i18ei20. [33] Andrew T. Hermann, John B. Wiley, Mater. Res. Bull. 44 (2009) 1046e1050. [34] E.A. Josepha, S. Farooq, C.M. Mitchell, J.B. Wiley, J. Solid State Chem. 216 (2014) 85e90. [35] Zhen-Hua Liang, Kai-Bin Tang, Qian-Wang Chen, Hua-Gui Zheng, Acta Cryst. E65 (2009) i44. [36] M. Dion, M. Ganne, M. Tournoux, Rev. Chim. Min. 23 (1986) 61, 61-? [37] T.A. Kodenkandath, A.S. Kumbhar, W.L. Zhou, J.B. Wiley, Inorg. Chem. 40 (4) (2001) 710e714. [38] M. Dion, M. Ganne, M. Tournoux, Mater. Res. Bull. 16 (1981) 1429e1435. [39] J. Laugier and B. Bochu, http://www.inpg.fr/LMGP, Domaine Univ. BP 46, res, France. 38402 Saint Martin d'He [40] J. Rodríguez-Carvajal, FullProf Suite. http://www.ill.eu/sites/fullprof/. [41] Luis Elcoro, Javier F. Zuniga, J. Manuel Perez-Mato, Acta Cryst. B60 (2004) 21e31.
Contents lists available at ScienceDirect
Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom
Thermal stability and high temperature polymorphism of topochemically-prepared DioneJacobson triple-layered perovskites Stephen L. Guertin, Elisha A. Josepha, Dariush Montasserasadi, John B. Wiley* Department of Chemistry and Advanced Materials Research Institute, University of New Orleans, 2000 Lakeshore Dr., New Orleans, LA 70148, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 4 May 2015 Accepted 5 June 2015 Available online 10 June 2015
The thermal stability of six DioneJacobson-related triple layered perovskites, ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl), was explored to 1000 C. Each compound was produced topochemically by low-temperature (<500 C) ion exchange from RbCa2Nb3O10. The thermal behavior of the series was examined by variable temperature X-ray powder diffraction experiments in tandem with thermogravimetric analysis and differential scanning calorimetry. Five of the species were found to be low temperature/metastable phases, decomposing below 900 C, where the stability of the series decreased with decreasing interlayer cation size. The compounds, A ¼ Li, Na, K, exhibited high temperature polymorphism, with a completely reversible transition evident for KCa2Nb3O10. © 2015 Elsevier B.V. All rights reserved.
Keywords: Oxide materials Preferential site ordering Phase transitions Calorimetry X-ray diffraction
1. Introduction Layered perovskite compounds have been extensively studied due to their wide-ranging potential for applications, as well as for their susceptibility to soft chemical manipulation [1e4]. Many high temperature superconductors (e.g., YBa2Cu3O7 and Tl2Ba2Ca2Cu3O10) share perovskite-related structures [5e7], and layered perovskite materials have been variously utilized in photocatalysts [8e11], hydrogen sensors [12e14], and fatigue-free capacitors [15e18]. Such materials offer a great versatility with respect to solid state synthesis, allowing important electronic and ferroic properties to be tailored with convenience and ease [4]. Nanostructuring of related perovskites is another exciting new area of research [19]. Soft chemistry (chimie douce) involves the use of lowtemperature synthetic methods to form desired products that cannot be prepared by traditional high temperature ceramic techniques. For many compounds, one-step, direct synthesis from simple precursors is thermodynamically unfavored. In such cases, soft chemical topotactic routes such as ion exchange and intercalation may be used to retain resilient structural features of a precursor in the products. The strata in layered perovskite oxides can be utilized in this manner to create families of metastable materials
* Corresponding author. E-mail address: [email protected] (J.B. Wiley). http://dx.doi.org/10.1016/j.jallcom.2015.06.045 0925-8388/© 2015 Elsevier B.V. All rights reserved.
with variant interlayer chemistry [20]. These can, in turn, be used as secondary reagents in multistep synthesis [3]. Layered perovskites of the DioneJacobson (DJ) family have the general formula A[A0 n1BnO3nþ1], where typically A ¼ alkali metal, A' ¼ rare earth or alkaline-earth, B ¼ transition metal, and n ¼ number of perovskite layers. This family has been well characterized, including reports on their structures [21e25], electronics [26e28], and magnetics [23,29,30]. In the case of a doped potassium variant, superconductive properties have been observed [31]. The utility and development of topochemical reaction strategies in part relies on determining the effective window of thermal stability for the reagents and intermediates. By defining the limits of these systems and their modifications, strategies can be designed to more effectively access target materials. Recent studies have examined the stability of double-layered perovskites [32,33]. Of the triple-layered species, only HCa2Nb3O10 and NaCa2Nb3O10 have been explicitly probed above 500 C [23,34]. In the research reported herein, the thermal stability of the DioneJacobson perovskite series ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl) was investigated. These species can be categorized into three classes based on relative thermal stability and similarity of rearrangement/ decomposition pathways. Most interestingly, members of the class ACa2Nb3O10 (A ¼ Li, Na, K) were discovered to undergo endothermic rearrangement into high temperature polymorphs, only one of which has been previously identified [23, 38].
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
2. Experimental 2.1. Synthesis RbCa2Nb3O10 was synthesized by solid state reaction. Rb2CO3 (Alfa Aesar, 99.8%), CaCO3 (Alfa Aesar, 99.99%), and Nb2O5 (Alfa Aesar, 99.9985%) were finely ground together in the molar ratio of 1.25:1:1, respectively. The mixture was heated in an alumina crucible at 850 C for 24 h, re-ground, and then heated at 1050 C for 24 h. The product was washed thoroughly with distilled water followed by acetone and oven-dried at 110 C for 4 h. KCa2Nb3O10 was also synthesized by direct reaction for comparison to the ion exchanged species. K2CO3 (Alfa Aesar, 99.997%), CaCO3 (Alfa Aesar, 99.99%), and Nb2O5 (Alfa Aesar, 99.9985%) were ground together and pressed into 7 mm pellets by hand press (SigmaeAldrich). The pellets were heated at 850 C for 24 h and 1200 C for two periods of 4 h with intermittent grinding and hand-press pelletization between each step. RbCa2Nb3O10 was used for ion-exchange with various monovalent cationic species to prepare the series ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl) as detailed in Table 1. For the ammonium and alkali-metal species, the parent was finely ground with the corresponding nitrate reagent (NH4NO3, Alfa Aesar, 99.999%; LiNO3, Alfa Aesar, 99.99%; NaNO3, Alfa Aesar, 99.999%; KNO3, Alfa Aesar, 99.994%) and placed into an alumina crucible. The exchange reactions were carried out for 4 days in an ANO3 molten salt with the ratio 1:10 for ammonium and lithium and 1:20 for sodium and potassium, at temperatures of 200 C, 350 C, 400 C, and 400 C, respectively. To synthesize (CuCl)Ca2Nb3O10, RbCa2Nb3O10 and anhydrous CuCl2 (Alfa Aesar, 99.995%) were ground together in a drybox, hand-pressed into pellets, sealed under vacuum in a Pyrex tube and reacted at 325 C for 10 days. Each of this series of products was washed and dried in the same fashion as the starting material. HCa2Nb3O10 was formed by stirring RbCa2Nb3O10 in 6 M HNO3 (aq) at 60 C for 4 days, with replenishment of the acid solution after 3 days. This product required isolation by centrifugation for the washing step because much of the solid was fine enough to permeate standard filter paper. 2.2. Characterization Variable temperature X-ray powder diffraction (XRD) data were collected on a Philips X-Pert PW 3020 MPD diffractometer (Cu Ka radiation, l ¼ 1.5418 Å) equipped with a curved graphite monochromator and an Anton Paar HTK1600 variable temperature stage. To determine the structural evolution of the compounds with increasing temperature, XRD data were taken from 25 C to 1000 C under flowing nitrogen in 50 C heating increments. Kinetically hindered transitions were studied with longer periods of heat treatment, ranging from 6 to 24 h. To isolate endothermic and exothermic events associated with phase changes, the materials were examined by differential scanning calorimetry (DSC) on a Table 1 Summary of ion-exchange reactions with RbCa2Nb3O10 host. Ion-exchanged product HCa2Nb3O10$(H2O)n (NH4)Ca2Nb3O10 LiCa2Nb3O10 b-NaCa2Nb3O10∙(H2O)n KCa2Nb3O10 (CuCl)Ca2Nb3O10 a b c
6 M HCl. ANO3 (A ¼ NH4, Li, Na, K). Anhydrous CuCl2.
Molar ratio a
Excess 1:10b 1:10b 1:20b 1:20b 1:2c
Temperature ( C)
Time (days)
60 200 350 400 400 325
4 4 3 7 7 10
371
Netzsch 404S thermal analysis system. Samples were heated and cooled in alumina or platinum pans at 10 C/min to 1000 C under flowing argon (50 m[⁄min). Thermogravimetric analysis (TGA) in tandem with DSC was used to monitor weight changes during decomposition on a TA Instruments SDT Q600 V8.3 system. Samples were characterized by scanning electron microscopy on a JEOL JSM-5410 SEM and electron dispersive spectroscopy (EDS) EDAX DX-PRIME microanalytical system to examine morphology and chemical composition at various stages of heating. The peak positions and lattice parameters were refined by a least-squares method with the programs ChekCell and FULLPROF [39, 40]. 3. Results RbCa2Nb3O10 was indexed in the tetragonal system with space group P4/mmm (#123). The lattice parameters presented in Table 2 were refined using the method of least squares refinement and were found to be in good agreement with previous reports [35]. The lattice parameters of ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl) were also consistent with literature values (Table 2), and EDS analysis showed minimal amounts of residual rubidium ion (Rb:Nb < 0.05:1.00). Two of the species were found susceptible to hydration. At 25 C, HCa2Nb3O10∙(H2O)n and b-NaCa2Nb3O10∙(H2O)n intercalate variable amounts of water inside the interlayer, increasing the magnitude of the c lattice parameter by up to 1.5 Å. The degree of hydration in air is strongly dependent on ambient humidity [23]. For HCa2Nb3O10, endothermic dehydration is observed between 50 C and 125 C with a mass loss of 3.3% corresponding to ~1.9 waters of hydration. The event is correlated with the shifting of the [001] reflection to a higher angle and a dramatic drop in its intensity. Dehydration of the sodium species requires higher temperatures. A pronounced endothermic signal is observed centered on 150 C, accompanied by a mass loss of 3.5% (~1.1 waters of hydration) and a significant [001] reflection shift. b-NaCa2Nb3O10 continues to lose mass up to 400 C, but above 225 C the weight loss is minimal and is not accompanied by observable structural rearrangement. Each of the compounds ACa2Nb3O10 (A ¼ NH4, Li, Na, K, CuCl) also demonstrate small degrees of hydration, measuring less than 1% of the total sample weight. The volatilizations, however, are unaccompanied by crystallographic changes, indicating that such processes likely represent loss of surface water. ACa2Nb3O10 when A ¼ H, NH4 are the least thermally stable of the family. In the decomposition, the interlayer cations are removed from the perovskite lattice via formation of a corresponding oxide. In a concerted step, HCa2Nb3O10 evolves H2O (1.6% mass loss observed, theoretically 1.73%) and (NH4)Ca2Nb3O10 evolves NH3 and H2O (4.7% observed, theoretically 4.85%). These transitions are correlated with an exothermic DSC signal centered at 315 C for the hydrogen species, while the ammonium compound gives a sharper 350 C exotherm followed by an endotherm at 365 C. The residual salt in both cases is the poorly crystallized single phase Ca4Nb6O19, previously described by Fang and coworkers [31]. Major reflections of this tetragonal byproduct appear in the XRD data at about 300 C and 340 C, respectively. Upon further heating to 570 C, Ca4Nb6O19 begins to disproportionate to CaNb2O6 (JCPDS File No. 39-1392) and Ca2Nb2O7 (JCPDS File No. 42-2), accompanied by a DSC slope change in the exothermic direction. Members of the series ACa2Nb3O10 (A ¼ Li, Na) are found to be more stable than the A ¼ H and NH4 species, yet with sufficient heating undergo rearrangement and decomposition. In neither case is mass loss observed after dehydration. In the XRD of LiCa2Nb3O10, no change in crystal structure occurs up to 600 C. At 680 C DSC registers an endothermic event, correlated with a major shifting of reflections in the 650 C and 700 C diffractograms, representing the irreversible formation of an alternative polymorph of LiCa2Nb3O10.
372
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
Table 2 Lattice parameters of ion-exchanged species. Species
a (Å)
Exp.a RbCa2Nb3O10 Ref.b RbCa2Nb3O10 Exp. HCa2Nb3O10 Ref. HCa2Nb3O10 Exp. NH4Ca2Nb3O10 Ref. NH4Ca2Nb3O10 Exp. LiCa2Nb3O10 Ref. LiCa2Nb3O10 Exp. NaCa2Nb3O10 Ref. NaCa2Nb3O10 Exp. KCa2Nb3O10 Ref. KCa2Nb3O10 Exp. (CuCl)Ca2Nb3O10 Ref. (CuCl)Ca2Nb3O10
3.862(3) 3.85865(6) 3.853(3) 3.8536(6) 7.720(1) 7.724(2) 7.719(6) 7.720(7) 5.461(4) 5.4731(2) 7.741(5) 7.7418(6) 3.850(2) 3.84994(6)
a b
b (Å)
c (Å)
Layer spacing 14.93
e e e e e e e e e 7.7073(6) e e
14.9306(2) 14.9108(3) 14.381(1) 14.403(5) 14.993(1) 14.977(8) 28.305(6) 28.331(3) 28.935(1) 29.0138(9) 14.861(8) 14.859(1) 15.674(1) 15.6706(5)
Reference [35]
14.38 [21] 14.99 [38] 14.15 [38] 14.47 [23] 14.86 [24] 15.67 [37]
Exp. ¼ experimental values. Ref. ¼ literature values.
The new system exothermically decomposes at 730 C into Li2O, CaNb2O6, and Ca2Nb2O7. Reflections of lithium oxide were not observed in the diffraction pattern of the decomposed sample, but the other products were clearly present. Presumably, the invisibility of Li2O is due to a low degree of crystallinity and/or poor X-ray scattering capacity of second period elements. b-NaCa2Nb3O10 undergoes an irreversible phase change at high temperature into a-NaCa2Nb3O10, as previously reported [23,38]. Byeon and workers located this transition at 650 C. In the present study, DSC registers an endothermic event at 690 C, well correlated with changes in its XRD patterns. Like the low temperature species, the a-phase is found to form a hydrate in air at 25 C. At 800 C the polymorph begins to decompose into a two phase system consisting of NaNbO3 and NaCa4Nb5O17 [32,41]. The XRD data reveal the rapid appearance of major reflections of decomposition products near the transition temperature, but total decomposition appears kinetically limited and requires hours to reach completion. Similarly to the other alkali species, KCa2Nb3O10 transforms into an alternative polymorph at high temperatures. Notably, the high temperature structure lacks the very prominent (207) reflection (Fig. 1 of the low temperature species). The reflections collected from 15 to 40 2q strongly resemble those of tetragonal RbCa2Nb3O10, suggesting similarity of structure. As expected by analogy to the lithium and sodium variants, the rearrangement is accompanied by an endothermic trace. The transition temperature for KCa2Nb3O10 synthesized by direct reaction was found to be 1000 C, but this threshold is decreased by up to 200 C for the ionexchanged species, presumably due to the presence of residual rubidium ions in the interlayer. Independent batches of ion exchanged samples demonstrated divergent transition temperatures, while distinct batches of the direct product behaved identically. Most notably, when allowed to cool, the high temperature phase reverts to the low temperature structure, with this transition lagging 50 C lower than the heated transition. The sample was subjected to three consecutive heatingecooling cycles without change in transition temperatures or sample crystallinity. (CuCl)Ca2Nb3O10 exhibits a more complicated bonding scheme than the other compounds investigated, as the interlayer is occupied by an extended two-dimensional square planar copper (II) chloride framework rather than by a single cation. TGA revealed two major volatilization events. A steep 3.1% mass loss is centered at 550 C, and a second loss of 2.2% centered at 825 C. The first event is accompanied by a very gradual DSC slope change in the exothermic direction, while the latter registers a sharp exotherm. EDS analysis confirmed that the mass loss in both cases was due to removal of chloride from the structure (5.3% total mass loss
Fig. 1. Reversible phase transition of KCa2Nb3O10. a) Tetragonal RbCa2Nb3O10 at 25 C, b) K-exchanged RbCa2Nb3O10 at 950 C, c) 900 C, d) 850 C, e) 800 C, f) 750 C, g) 700 C, h) 650 C, and i) 600 C.
observed, theoretically 5.74%). When the sample was heated in an alumina pan, the alumina surface in contact with the sample was found to be scorched to a dark brown hue, possibly due to oxideechloride exchange. Experiments carried out with platinum containment showed no visual evidence of such interaction. XRD demonstrated major structural transitions in good agreement with DSC-TGA data. No previously reported diffraction patterns could be found for comparison with the intermediate species, which appears to be a poorly crystallized, complex material. Above 825 C, Ca2Nb2O7 is clearly present, but CaNb2O6 is absent, distinguishing this process from those of ACa2Nb3O10 (A ¼ H, NH4, Li). Intact (CuCl) Ca2Nb3O10 is light green in color. The anhydrous intermediate species (heating arrested at T ¼ 700 C) appears unevenly olive green and light grey, and the anhydrous terminal decomposition products are light gray.
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
373
Table 3 Summary of thermal behavior of the series ACa2Nb3O10 (A ¼ H, NH4, Li, Na, K, CuCl). Species
DSC event temperature
Reaction
HCa2Nb3O10
300 C 550 C 340 C 550 C 680 C 730 C 690 C 800 C (slow) 1000 C 800 C e 900 C 330 C 550 C
2 HCa2Nb3O10 / Ca4Nb6O19 þ H2O (g) Ca4Nb6O19 / 2 CaNb2O6 þ Ca2Nb2O7 2 NH4Ca2Nb3O10 / Ca4Nb6O19 þ 2NH3 (g) þ H2O (g) Ca4Nb6O19 / 2 CaNb2O6 þ Ca2Nb2O7 LiCa2Nb3O10 (LT)* 4 LiCa2Nb3O10 (HT)** 2 LiCa2Nb3O10 / 2 CaNb2O6 þ Ca2Nb2O7 þ Li2O b-NaCa2Nb3O10 4 a-NaCa2Nb3O10 (HT) 2 a-NaCa2Nb3O10 (HT) / NaNbO3 þ NaCa4Nb5O17 KCa2Nb3O10 (LT) 4 KCa2Nb3O10 (HT) KCa2Nb3O10 (LT) 4 KCa2Nb3O10 (HT) (CuCl)Ca2Nb3O10 / unidentified intermediate Unidentified intermediate / Ca2Nb2O7 þ unknown phase
(NH4)Ca2Nb3O10 LiCa2Nb3O10
b-NaCa2Nb3O10 KCa2Nb3O10 (Direct) KCa2Nb3O10 (Ion-exchanged) (CuCl)Ca2Nb3O10
*LT e low temperature phase; **HT e high temperature phase; 4 indicates a reversible process.
A reaction scheme for all examined compounds is summarized in Table 3. Fig. 2 displays an XRD stack plot of the series ACa2Nb3O10 (A ¼ H, Li, NH4, CuCl) after being maintained at 1000 C for five days compared to the JCPDS reference patterns of the terminal products. 4. Discussion Rationalization of the thermal stability of the series is largely dependent on the size of the interlayer cations. The utility of layered perovskites derives from the chemical resilience of the A'BO3 motif at extreme temperatures and pH values. The perovskite “slabs” can become chemical templates for affecting difficult, desirable transformations. The region of chemical and mechanical susceptibility, as in the ion exchange reactions of this study, is the interlayer. Larger, chemically softer cations, can be expected to coordinate greater numbers of oxide ions, thereby stabilizing anionic repulsion across the interlayer. Consistent with this model, the ammonium-substituted structure was found to be somewhat more stable than its protic counterpart, and the alkali series increased in stability down the group. The structure of LiCa2Nb3O10 was first studied in 1981 by Dion et al., who characterized the compound in the tetragonal system with a ¼ 7.720(7) Å and c ¼ 28.331(3) Å [38]. In 2001, Ishizawa and coworkers presented a synchrotron X-ray study of the material
which indicated triclinic P1 symmetry with a ¼ 5.4809(3) Å, b ¼ 5.4804(3) Å, c ¼ 26.5533(16) Å, a ¼ 89.999(4) , b ¼ 90.245(4) , g ¼ 89.999(5) [22]. The latter paper has since been cited as the standard model for LiCa2Nb3O10 structure. Interestingly, the present study identifies the diffraction pattern of the Ishizawa structure with that of the high temperature polymorph. Ion exchange synthesis at 350 C produces the low temperature phase, with calculated lattice parameters in line with the results of Dion et al. [38]. Ishizawa and coworkers, however, produced single crystals of the material in a molten flux with a much higher reaction temperature program. The present study implies that the synchrotron X-ray examination was performed on the high temperature phase. The reversible phase change in ceramically synthesized KCa2Nb3O10 correlates very well with a thermoelastic transition of the compound at 1000 C reported by Dion et al. [36], and presumably this structural rearrangement is responsible for the bulk phenomenon. In the same paper, Dion describes a thermoelastic transition for RbCa2Nb3O10 occurring at 620 C. Small occupancy of potassium sites by rubidium in KCa2Nb3O10 (Rb < 5%) greatly lowered the polymorphic transition temperature by as much as 200 C (Table 3). The potassium compound has been reported as 6coordinate [24], while the rubidium has been identified as 8coordinate [35]. Fig. 3 displays a proposed model for the low and high temperature polymorphs. If the high temperature phase
Fig. 2. Stack plot of terminal products versus CaNb2O6 and Ca2Nb2O7. a) HCa2Nb3O10, b) LiCa2Nb3O10, c) NH4Ca2Nb3O10, d) (CuCl)Ca2Nb3O10, e) CaNb2O6 and f) Ca2Nb2O7.
374
S.L. Guertin et al. / Journal of Alloys and Compounds 647 (2015) 370e374
Fig. 3. Proposed structural transition of KCa2Nb3O10. Left: Low temperature polymorph. Right: High temperature polymorph.
features 8-coordinate potassium, then residual rubidium would be expected to facilitate the crystallographic transition. Future studies on the solid solution KnRb1nCa2Nb3O10 across the range 0 n 1 would help in the exploration of trends in the range 500 C < T < 1000 C. 5. Conclusions The thermal stability of the triple-layered calcium niobate perovskite family depends fundamentally on its interlayer chemistry. Reaction schemes intended to incorporate the triple-layered perovskite structural framework at T > 500 C must remain cognizant of the specific integrities of precursor species to ensure that rearrangements do not impede synthetic goals. Alternatively, knowledge of decomposition and transition temperature thresholds could be exploited to access novel materials. Acknowledgments This study has been supported by the National Science Foundation (DMR1005856). References [1] A.S. Bhalla, Ruyan Guo, Rustum Roy, Mater. Res. Innov. 4 (2000) 3e26. [2] M. Johnsson, P. Lemmens, Handbook of Magnetism and Advanced Magnetic Materials, John Wiley and Sons, New York, 2007. [3] K.G. Ranmohotti, E. Josepha, J. Choi, J. Zhang, J.B. Wiley, Adv. Mater. 23 (4) (2011) 442e460.
[4] R.E. Schaak, T.E. Mallouk, Chem. Mater. 14 (4) (2002) 1455e1471. [5] Neeraj Khare, Handbook of High Temperature Superconductor Electronics, CRC Press, 2003. [6] C.N.R. Rao, P. Ganguly, K. Sreedhar, R.A. Mohan Ram, P.R. Sarode, Mater. Res. Bull. 22 (6) (1987) 849e855. [7] C.C. Torardi, M.A. Subramanian, J.C. Calabrese, J. Gopalakrishnan, K.J. Morrissey, T.R. Askew, R.B. Flippen, U. Chowdry, S.W. Sleight, Science 240 (4852) (1988) 631e634. [8] Jindo Kim, Hwang Dong, Hyun Kim, Sang Bae, Jae Lee Wei, Se Oh Li, Top. Catal. 35 (3e4) (2005) 295e303. [9] Masato Machida, Jun-Ichi Yabunaka, Tsuyoshi Kijima, Chem. Mater. 12 (3) (2000) 812e817. [10] Shunsuke Nishimoto, Yoshihiro Okazaki, Motohide Matsuda, Michihiro Miyake, J. Ceram. Soc. Jpn. 117 (2009) 1175e1179. [11] Tsuyoshi Takata, Yoko Furumi, Kiyoaki Shinohara, Akira Tanaka, Michikazu Hara, Junko N. Kondo, Kazunari Domen, Chem. Mater. 9 (5) (1997) 1063e1064. [12] M. Sakthivel, W. Weppner, Sens. Actuators B 125 (2) (2007) 435e440. [13] Peng Song, Jifan Hu, Hongwei Qin, Ling Zhang, Tianfeng Xing, Xue Liu, Shanxing Huang, Minhua Jiang, J. SoleGel Sci. Tech. 35 (1) (2005) 65e68. [14] Xiling Tang, Kurtis Remmel, Daniel Sandker, Zhi Xu, Junhang Dong, Xudong Fan, Hai Xiao, Anbo Wang (Eds.), Proceedings of the SPIE, 2010, p. 7682, 76820L-76820L-7. [15] C. A-Paz de Araujo, J.D. Cuchiaro, L.D. McMillan, M.C. Scott, J.F. Scott, Nature. 374 (1994) 627e629. [16] B.H. Park, B.S. Kang, S.D. Bu, T.W. Noh, J. Korean Phys. Soc. 1999 (1999) 35. S13067eS1309. [17] S.R. Shannigrahi, Hyun Jang, Appl. Phys. Lett. 79 (2001) 1051. ~ es, B.D. Stoyanovic, M.A. Ramirez, A.A. Cavalheiro, E. Longo, [18] A.Z. Simo J.A. Varela, Ceram. Int. 34 (2008) 257e261. [19] Debasish Mohanty, Girija S. Chaubey, Amin Yourdkhani, Shiva Adireddy, Gabriel Caruntu, John B. Wiley, RSC Adv. 2 (2012) 1913e1916. [20] Jacques Livage, New J. Chem. 25 (2001) 1. [21] Margaret J. Geselbracht, Helen K. White, Jeanette M. Blain, Miranda J. Dia, Jennifer L. Hubbs, Nicole Adelstein, Joshua A. Kurzman, Mater. Res. Bull. 46 (2011) 398e406. [22] Nobuo Ishizawa, Reiko Yamashita, Shuji Oishi, James R. Hester, Shunji Kishimoto, Acta Cryst. C57 (2001) 1006e1009. [23] Song-Ho Byeon, Jin Hee Kim, Dong-Kuk Kim, Nam Hwi Hur, Chem. Mater. 15 (2) (2003) 383e389. [24] Toshiaki Tokumitsu, Kenji Toda, Tsukuru Daiki Sakuraba, Kazuyoshi Uematsu, Mineo Sato, JCS-Jpn. 114 (9) (2006) 795e797. [25] Ramesh Chitrakar, Ind. Eng. Chem. Res. 47 (1) (2008) 176e179. [26] Venkataraman Thangadurai, Werner Weppner, J. Mater. Chem. 11 (2001) 636e639. [27] Venkataraman Thangadurai, Werner Weppner, Chem. Mater. 14 (2002) 1136e1143. [28] Yoshihiko Takano, Yoshihide Kimishima, Tokio Yamadaya, Shinji Ogawa, Nobuo Mori, Rev. High. Press. Sci. Technol. 7 (1998) 589e591. [29] X.L. Wang, E. Takayama-Muramaki, Phys. Rev. B: Condens. Matter Mater. Phys. 72 (2005) 064401. [30] Yoshihiko Takano, Shigeru Takayanagi, Shinji Ogawa, Tokio Yamadaya, ^ri, Solid State Commun. 103 (4) (1997) 215e217. Nobuo Mo [31] Mingming Fang, Chy Hyung Kim, Thomas E. Mallouk, Chem. Mater. 11 (1999). [32] F.J. Zuniga, J. Darriet, Acta Cryst. C59 (2003) i18ei20. [33] Andrew T. Hermann, John B. Wiley, Mater. Res. Bull. 44 (2009) 1046e1050. [34] E.A. Josepha, S. Farooq, C.M. Mitchell, J.B. Wiley, J. Solid State Chem. 216 (2014) 85e90. [35] Zhen-Hua Liang, Kai-Bin Tang, Qian-Wang Chen, Hua-Gui Zheng, Acta Cryst. E65 (2009) i44. [36] M. Dion, M. Ganne, M. Tournoux, Rev. Chim. Min. 23 (1986) 61, 61-? [37] T.A. Kodenkandath, A.S. Kumbhar, W.L. Zhou, J.B. Wiley, Inorg. Chem. 40 (4) (2001) 710e714. [38] M. Dion, M. Ganne, M. Tournoux, Mater. Res. Bull. 16 (1981) 1429e1435. [39] J. Laugier and B. Bochu, http://www.inpg.fr/LMGP, Domaine Univ. BP 46, res, France. 38402 Saint Martin d'He [40] J. Rodríguez-Carvajal, FullProf Suite. http://www.ill.eu/sites/fullprof/. [41] Luis Elcoro, Javier F. Zuniga, J. Manuel Perez-Mato, Acta Cryst. B60 (2004) 21e31.