- Email: [email protected]
Hydration thermodynamics of proton-conducting perovskite Ba4Ca2Nb2O11
Materials Letters 235 (2019) 97–100
Contents lists available at ScienceDirect
Materials Letters journal homepage: www.elsevier.com/locate/mlblue
Hydration thermodynamics of proton-conducting perovskite Ba4Ca2Nb2O11 Vladimir Sereda ⇑, Dmitry Malyshkin, Dmitry Tsvetkov, Andrey Zuev Institute of Natural Sciences and Mathematics, Ural Federal University, 620002, 19 Mira St., Yekaterinburg, Russia
a r t i c l e
i n f o
Article history: Received 6 September 2018 Received in revised form 26 September 2018 Accepted 30 September 2018 Available online 1 October 2018 Keywords: Perovskite Defects Thermal properties Proton conductors Hydration enthalpy
a b s t r a c t The defect structure model based on the single reaction of water incorporation, involving structural oxygen vacancies, was discussed and successfully verified using existing pH2 O T x data for cubic Ba4Ca2Nb2O11xH2O – a promising electrolyte material for proton-conducting solid oxide fuel cells. As a result, the enthalpy of hydration (DH0hydr ) was obtained. It was found to be close to the partial molar enthalpy of water in Ba4Ca2Nb2O11xH2O as well as to the calorimetrically measured value of DH0hydr = 107.9 ± 14.4 kJ/mol of H2O. Using DH0hydr and the value of low-temperature heat of hydration, the enthalpy of the hydration-induced cubic!monoclinic phase transition in Ba4Ca2Nb2O110.92H2O was calculated as 63.9 ± 14.2 kJ/mol of Ba4Ca2Nb2O11. Ó 2018 Elsevier B.V. All rights reserved.
1. Introduction The oxygen nonstoichiometry index d, i.e. the number of oxygen vacancies per formula unit, in perovskite-type BaCa(1+y)/3Nb(2-y)/3O3–d (BCNy) oxides can be tailored by varying the Ca–Nb ratio y, and equals d ¼ y=2. These oxygen vacancies can be hydrated under humid atmosphere, providing nonstoichiometric BCNy oxides with good proton conductivity [1–5]. It makes them promising materials for proton-conducting solid oxide fuel cell (SOFC) electrolytes and high-temperature humidity sensors. Among the BCNy compounds, Ba4Ca2Nb2O11 (BCN50) stands out clearly because of its large theoretical water uptake, up to x = 1.0 in Ba4Ca2Nb2O11xH2O, and its fully 1:1-ordered structure [1]. Besides, in contrast with the other BCNy, the crystal structure of highly hydrated BCN50 is known to differ from that of the ‘‘dry” sample. At lower temperature (T < 573 K) under humid atmospheres (water partial pressure pH2 O > 101:8 atm), when x is close to 1.0, BCN50 was found to be monoclinic with either C2=m [4,6] or P21 =n [7] space group assigned. Partially hydrated BCN50 (at higher temperatures
and/or in less humid environment) is cubic (Fm 3 m) [4,6–8]. Despite a number of works on the crystal structure [1], water content [2,3] and other properties of various BCNy compounds, including BCN50 [4–9], there is a lack of fundamental thermodynamic studies on these materials, and the defect chemistry of ⇑ Corresponding author. E-mail address: [email protected] (V. Sereda). https://doi.org/10.1016/j.matlet.2018.09.172 0167-577X/Ó 2018 Elsevier B.V. All rights reserved.
BCN50 was discussed only qualitatively [5,7,9]. Therefore, the present work aimed to partly address these issues by investigating the heat of low-temperature hydration-induced phase transition as well as the higher-temperature thermodynamics of hydration and related defect chemistry of BCN50 oxide.
2. Experimental BCN50 oxide was prepared via the standard ceramic technique from the high-purity BaCO3, CaCO3 and Nb2O5. Stoichiometric quantities of precursors were calcined in air at 1073–1473 K with 100 K step for 8 h at each temperature with intermediate regrindings. Finally, the powder was uniaxially pressed into pellets, annealed in air at 1773 K for 8 h and subsequently slowly (100 K/h) cooled to room temperature. The pellets prepared in this way were then crushed in an agate mortar. Phase-purity of the as-obtained BCN50 powder was confirmed by means of X-ray diffraction (XRD) with 7000S diffractometer (Shimadzu, Japan) using Cu Ka radiation (see Fig. 1). Calorimetric measurements were performed with an original heat-flux differential scanning calorimeter (DSC). The calorimetric cell consisted of two identical alumina crucibles with a thermopile in the form of 16 (8 on each crucible) K-type thermocouple junctions rigidly mounted on crucibles’ sides. Two aluminum crucibles, one of which empty and another one containing around 0.5 g of BCN50 powder, were inserted into the alumina ones. The calorimetric cell was then placed in a furnace and held at a given
98
V. Sereda et al. / Materials Letters 235 (2019) 97–100
equilibrium constant of reaction (1), mass and charge balances is expressed as
8 00 2 OV ½OHO > DH0 D S0 > > K ¼ ¼ exp RT1 þ R1 1 > 2 > V O p ½ O ½ O H2 O > > > h 00 i > > > < OHO ¼ 2 OV h 00 i > > OO þ OHO þ OV ¼ 11 þ x > > > > > > 1x > > VO ¼ > : OHO ¼ 2x
ð2Þ
where x is the amount of water in hydrated BCN50 (Ba4Ca2Nb2O11xH2O). The analytical solution of the set (2) with respect to pH2 O yields
pH2 O
temperature in 50 ml/min flow of dry (log pH2 O =atm 3:5) air for at least 48 h to achieve equilibrium water content in the sample. The inlet air was dried by passing it through the column with pre-annealed zeolites. After that, the air flow was redirected abruptly so as to bubble it through the flask with a saturated solu tion of KBr (log pH2 O =atm ¼ 1:67), and the corresponding heat effects resulting from the sample hydration were recorded. The DSC hydration experiment was repeated, at least, 3 times at each temperature. The heat sensitivity of DSC setup was calibrated using standard metals’ heats of fusion in scanning mode with various heating rates, and the sensitivity coefficient extrapolated to zero heating rate was used to calculate the heat of hydration. Temperature-dependent equilibrium water content in BCN50 samples in dry (log pH2 O =atm 3:5) and wet (log pH2 O =atm ¼ 1:67) air was measured by thermogravimetry (TG) using CI Precision (UK) microbalances.
3. Results and discussion Let us discuss first the defect reactions accompanying the
hydration of cubic (Fm 3 m) BCN50 at higher temperatures (T = 623–773 K). It was shown previously [1,7,9] that BCN50 is a perovskite-type oxide with 1:1 ordered structure, where Ca and Nb ions are placed on alternate (1 1 1) planes. The intrinsic oxygen vacancy concentration in dehydrated BCN50 is determined by the Ca:Nb ratio, i.e. the electroneutrality conditions, and equals 1 mol of structural vacancies per mole of the Ba4Ca2Nb2O11. Taking into account that the oxygen vacancies in BCN50 are of a structural nature [7,9], the water uptake by the sample can be written in KrögerVink notation as 00
!
ð3Þ
According to Eq. (3), the slope of log x ¼ f log pH2 O ; T depen-
Fig. 1. XRD pattern for as-sintered BCN50.
2OO þ VO þ H2 O¢2OHO þ OV
4x3 DH01 DS01 ¼ 3 exp 4x 48x2 þ 165x 121 RT R
ð1Þ
provided that Ba4Ca2Nb2O11 with 1 structural vacancy, V O , per formula unit is chosen as a reference crystal. The oxygen from the water molecule fills the structural vacancy resulting in negatively 00
charged defect OV . For simplicity, we assume that the defect chemistry of BCN50 at T = 623–773 K and log pH2 O =atm < 1:6 is governed solely by the reaction (1). Corresponding set of equations comprising the
dencies should be close to 1/3 at low level of hydration in agreement with the qualitative assessment of the defect chemistry of BCN50 made by Animitsa et al. [5,7,9]. Standard enthalpy, DH01 , and entropy, DS01 , of reaction (1) were obtained as a result of nonlinear fitting of Eq. (3) to the equilibrium pH2 O T x data for BCN50 reported by Animitsa and Kochetova [7]. The results of such fitting are presented in Fig. 2a and Table 1. A good agreement between the calculated surface and the experimental data, as seen in Fig. 2a, is confirmed by close to unity coefficient of determination R2 = 0.98. In principle, the other defect equilibria, such as those expressed by Eq. (4) or Eq. (5): 00
OO þ VO ¢VO þ OV 0
ð4Þ 00
OO þ OHV ¢OHO þ OV
ð5Þ
could influence the defect chemistry of BCN50. However, it is implied in Eqs. (4) and (5) that there are different sites occupied 0
00
by protons (OHV and OHO ) and oxide ions (O O and OV ). In order to affect the hydration thermodynamics of BCN50, these sites should be regarded as crystallographically and energetically nonequivalent. Otherwise, the said reactions would describe the defect exchange between the equal positions, being essentially meaningless. However, nonequivalent crystallographic positions of oxygen were found only in fully hydrated monoclinic (P21 =n) BCN50 at lower (less than 573 K) temperatures [4,7] but not in the cubic
(Fm 3 m) BCN50 that is the focus of the present discussion. In addition, the reaction (4) is unlikely to proceed at T 773 K, because of the insufficient mobility of oxygen at such a low temperature. Besides, the introduction of either Eq. (4) or Eq. (5) into the defect structure model does not improve the results of the model verification (fitting). Consequently, one can assume that the equilibria (4) and (5) affect the defect chemistry of BCN50 insignificantly, at least, under the conditions discussed in the present work (T = 623–773 K and log pH2 O =atm < 1:6). Partial molar enthalpy, Dh, and entropy, Ds, of water, calculated using pH2 O T x data [7], are shown in Fig. 2b and c, respectively. It is seen that, while Ds increases slightly with water content increase, Dh in BCN50 is almost composition-independent. The average value of Dh, Dhavg (see Table 1), is in good agreement with
DH01 , reinforcing our assumption that the defect structure model based on reaction (1) describes the uptake of water by BCN50 quite well.
99
V. Sereda et al. / Materials Letters 235 (2019) 97–100
Fig. 2. The results of model analysis: points – experimental data [7], surface – fitted model (a) along with the partial molar enthalpy (b) and entropy (c) of water in BCN50.
Table 1 Thermodynamic parameters of hydration of BCN50 (see explanations in the text). Equilibrium pH2 O T x data [7]
DH01 ,
kJmol
1
93.5 ± 9.6 *
DS01 ,
Jmol
Calorimetry 1
K
154.3 ± 15.4
1
1
Dhavg , kJmol
DH0hydr , 692 K, kJmol1
DH0total , 473 K, kJmol1*
DH0tr , kJmol1*
95.9 ± 8.0
107.9 ± 14.4
35.4 ± 3.2
63.9 ± 14.2
kJ/mol of BCN50.
The calorimetric measurements, the results of which are presented in Fig. 3, were performed at two temperatures, 473 and 692 K. At higher temperature, the total heat effect, corresponding to the enthalpy of hydration, DH0hydr , was found to be – 29.13 ± 0.27 kJ/mol of BCN50, or 107.9 ± 14.4 kJ/mol of H2O, because, according to TG, 1 mol of BCN50 absorbs 0.27 ± 0.04 mol of water under the experimental conditions. Larger error of
DH0hydr per mole of H2O is due to the water content measurement uncertainties. It should be emphasized that directly measured DH0hydr and the hydration enthalpies obtained from the equilibrium pH2 O T x data, DH01 and Dhavg , coincide with each other considering their respective uncertainties (see Table 1). It is noteworthy that the DH0hydr of BCN50 is distinctively more exothermal than
DH0hydr reported for the other BCNy (y 0.18) [2,3]. As seen in Fig. 3, the reproducibility of the DSC is somehow lower at 473 K than at 692 K. It is not surprising, considering the sluggish hydration kinetics and the complex nature of hydration processes in BCN50 at 473 K. In the wet atmosphere at T = 473 K,
Fig. 3. DSC measurement results.
100
V. Sereda et al. / Materials Letters 235 (2019) 97–100
BCN50 is almost fully hydrated and possesses monoclinic structure [4,7]. Thus, the hydration at 473 K is accompanied by the phase transition, and the total enthalpy of water uptake, DH0total , the value of which is presented in Table 1, can be expressed as
DH0total ¼ DH0hydr Dx þ DH0tr
ð6Þ
where Dx = 0.92 ± 0.04 is the amount of water absorbed by 1 mol of BCN50 at 473 K as measured by TG, and DH0tr and DH0hydr are the phase transformation (from cubic to monoclinic Ba4Ca2Nb2O110.92H2O) and hydration enthalpies, respectively. Supposedly, the distortions of the crystal lattice of BCN50 caused by the hydration should not lead to significant change in the hydration energetics. In other words, DH0hydr for the cubic and monoclinic BCN50 should be quite close to each other. Neglecting their difference and assuming that DH0hydr is temperature-independent, at least between 473 and 692 K, DH0tr can be calculated using Eq. (6) (see the value of
DH0tr in Table 1). 4. Conclusions The defect structure model for BCN50 based on the single reaction of water uptake was discussed and successfully verified using the proton content dependencies,x T; pH2 O , at 623–773 K [7]. The
values of the hydration enthalpy of cubic (Fm 3 m) BCN50, either measured directly or evaluated using the equilibrium pH2 O T x data [7], were shown to be close to each other. The enthalpy of the cubic!monoclinic phase transition for Ba4Ca2Nb2O110.92H2O was calculated using calorimetrically measured low-temperature heat of hydration. As this transition introduces a degree of disorder, lowering the crystal lattice symmetry, its entropy,
DS0tr ¼ DH0tr =T tr , where T tr is the phase transition temperature, should be positive, making the phase transformation enthalpy DH0tr endothermic. Acknowledgements Sereda V.V. acknowledges the project SP-3103.2018.1 funded by the The Council for grants of the President of Russian Federation. References [1] Y. Du, A.S. Nowick, Structural transitions and proton conduction in nonstoichiometric A3B’B”O9 perovskite-type oxides, J. Am. Ceram. Soc. 78 (11) (1995) 3033–3039. [2] F. Krug, T. Schober, The high-temperature proton conductor Ba3(Ca1. 18Nb1.82)O9-d: thermogravimetry of the water uptake, Solid State Ionics 92 (3) (1996) 297–302. [3] T. Schober, J. Friedrich, The mixed perovskites BaCa(1+x)/3Nb(2–x)/3O3x/2 (x= 0. . .0.18): proton uptake, Solid State Ionics 136–137 (2000) 161–165. [4] I. Animitsa, A. Neiman, N. Kochetova, B. Melekh, A. Sharafutdinov, Proton and oxygen-ion conductivity of Ba4Ca2Nb2O11, Solid State Ionics 162–163 (2003) 63–71. [5] D.V. Korona, A.Y. Neiman, I.E. Animitsa, A.R. Sharafutdinov, Effect of humidity on conductivity of Ba4Ca2Nb2O11 phase and solid solutions based on this phase, Russ. J. Electrochem. 45 (5) (2009) 586–592. [6] N.A. Kochetova, I.E. Animitsa, A.Y. Neiman, The synthesis and properties of solid solutions based on Ba4Ca2Nb2O11, Russ. J. Phys. Chem. A 83 (2) (2009) 203–208. [7] I.E. Animitsa, N.A. Kochetova, Crystal structure and imperfection of the perovskite-like proton conductor Ba4Ca2Nb2O11, Chim. Techno Acta 3 (1) (2016) 5–13. [8] I. Animitsa, T. Denisova, A. Neiman, A. Nepryahin, N. Kochetova, N. Zhuravlev, P. Colomban, States of H+-containing species and proton migration forms in hydrated niobates and tantalates of alkaline-earth metals with a perovskiterelated structure, Solid State Ionics 162–163 (2003) 73–81. [9] I.E. Animitsa, High-temperature proton conductors with structure-disordered oxygen sublattice, Russ. J. Electrochem. 45 (6) (2009) 668–676.
Contents lists available at ScienceDirect
Materials Letters journal homepage: www.elsevier.com/locate/mlblue
Hydration thermodynamics of proton-conducting perovskite Ba4Ca2Nb2O11 Vladimir Sereda ⇑, Dmitry Malyshkin, Dmitry Tsvetkov, Andrey Zuev Institute of Natural Sciences and Mathematics, Ural Federal University, 620002, 19 Mira St., Yekaterinburg, Russia
a r t i c l e
i n f o
Article history: Received 6 September 2018 Received in revised form 26 September 2018 Accepted 30 September 2018 Available online 1 October 2018 Keywords: Perovskite Defects Thermal properties Proton conductors Hydration enthalpy
a b s t r a c t The defect structure model based on the single reaction of water incorporation, involving structural oxygen vacancies, was discussed and successfully verified using existing pH2 O T x data for cubic Ba4Ca2Nb2O11xH2O – a promising electrolyte material for proton-conducting solid oxide fuel cells. As a result, the enthalpy of hydration (DH0hydr ) was obtained. It was found to be close to the partial molar enthalpy of water in Ba4Ca2Nb2O11xH2O as well as to the calorimetrically measured value of DH0hydr = 107.9 ± 14.4 kJ/mol of H2O. Using DH0hydr and the value of low-temperature heat of hydration, the enthalpy of the hydration-induced cubic!monoclinic phase transition in Ba4Ca2Nb2O110.92H2O was calculated as 63.9 ± 14.2 kJ/mol of Ba4Ca2Nb2O11. Ó 2018 Elsevier B.V. All rights reserved.
1. Introduction The oxygen nonstoichiometry index d, i.e. the number of oxygen vacancies per formula unit, in perovskite-type BaCa(1+y)/3Nb(2-y)/3O3–d (BCNy) oxides can be tailored by varying the Ca–Nb ratio y, and equals d ¼ y=2. These oxygen vacancies can be hydrated under humid atmosphere, providing nonstoichiometric BCNy oxides with good proton conductivity [1–5]. It makes them promising materials for proton-conducting solid oxide fuel cell (SOFC) electrolytes and high-temperature humidity sensors. Among the BCNy compounds, Ba4Ca2Nb2O11 (BCN50) stands out clearly because of its large theoretical water uptake, up to x = 1.0 in Ba4Ca2Nb2O11xH2O, and its fully 1:1-ordered structure [1]. Besides, in contrast with the other BCNy, the crystal structure of highly hydrated BCN50 is known to differ from that of the ‘‘dry” sample. At lower temperature (T < 573 K) under humid atmospheres (water partial pressure pH2 O > 101:8 atm), when x is close to 1.0, BCN50 was found to be monoclinic with either C2=m [4,6] or P21 =n [7] space group assigned. Partially hydrated BCN50 (at higher temperatures
and/or in less humid environment) is cubic (Fm 3 m) [4,6–8]. Despite a number of works on the crystal structure [1], water content [2,3] and other properties of various BCNy compounds, including BCN50 [4–9], there is a lack of fundamental thermodynamic studies on these materials, and the defect chemistry of ⇑ Corresponding author. E-mail address: [email protected] (V. Sereda). https://doi.org/10.1016/j.matlet.2018.09.172 0167-577X/Ó 2018 Elsevier B.V. All rights reserved.
BCN50 was discussed only qualitatively [5,7,9]. Therefore, the present work aimed to partly address these issues by investigating the heat of low-temperature hydration-induced phase transition as well as the higher-temperature thermodynamics of hydration and related defect chemistry of BCN50 oxide.
2. Experimental BCN50 oxide was prepared via the standard ceramic technique from the high-purity BaCO3, CaCO3 and Nb2O5. Stoichiometric quantities of precursors were calcined in air at 1073–1473 K with 100 K step for 8 h at each temperature with intermediate regrindings. Finally, the powder was uniaxially pressed into pellets, annealed in air at 1773 K for 8 h and subsequently slowly (100 K/h) cooled to room temperature. The pellets prepared in this way were then crushed in an agate mortar. Phase-purity of the as-obtained BCN50 powder was confirmed by means of X-ray diffraction (XRD) with 7000S diffractometer (Shimadzu, Japan) using Cu Ka radiation (see Fig. 1). Calorimetric measurements were performed with an original heat-flux differential scanning calorimeter (DSC). The calorimetric cell consisted of two identical alumina crucibles with a thermopile in the form of 16 (8 on each crucible) K-type thermocouple junctions rigidly mounted on crucibles’ sides. Two aluminum crucibles, one of which empty and another one containing around 0.5 g of BCN50 powder, were inserted into the alumina ones. The calorimetric cell was then placed in a furnace and held at a given
98
V. Sereda et al. / Materials Letters 235 (2019) 97–100
equilibrium constant of reaction (1), mass and charge balances is expressed as
8 00 2 OV ½OHO > DH0 D S0 > > K ¼ ¼ exp RT1 þ R1 1 > 2 > V O p ½ O ½ O H2 O > > > h 00 i > > > < OHO ¼ 2 OV h 00 i > > OO þ OHO þ OV ¼ 11 þ x > > > > > > 1x > > VO ¼ > : OHO ¼ 2x
ð2Þ
where x is the amount of water in hydrated BCN50 (Ba4Ca2Nb2O11xH2O). The analytical solution of the set (2) with respect to pH2 O yields
pH2 O
temperature in 50 ml/min flow of dry (log pH2 O =atm 3:5) air for at least 48 h to achieve equilibrium water content in the sample. The inlet air was dried by passing it through the column with pre-annealed zeolites. After that, the air flow was redirected abruptly so as to bubble it through the flask with a saturated solu tion of KBr (log pH2 O =atm ¼ 1:67), and the corresponding heat effects resulting from the sample hydration were recorded. The DSC hydration experiment was repeated, at least, 3 times at each temperature. The heat sensitivity of DSC setup was calibrated using standard metals’ heats of fusion in scanning mode with various heating rates, and the sensitivity coefficient extrapolated to zero heating rate was used to calculate the heat of hydration. Temperature-dependent equilibrium water content in BCN50 samples in dry (log pH2 O =atm 3:5) and wet (log pH2 O =atm ¼ 1:67) air was measured by thermogravimetry (TG) using CI Precision (UK) microbalances.
3. Results and discussion Let us discuss first the defect reactions accompanying the
hydration of cubic (Fm 3 m) BCN50 at higher temperatures (T = 623–773 K). It was shown previously [1,7,9] that BCN50 is a perovskite-type oxide with 1:1 ordered structure, where Ca and Nb ions are placed on alternate (1 1 1) planes. The intrinsic oxygen vacancy concentration in dehydrated BCN50 is determined by the Ca:Nb ratio, i.e. the electroneutrality conditions, and equals 1 mol of structural vacancies per mole of the Ba4Ca2Nb2O11. Taking into account that the oxygen vacancies in BCN50 are of a structural nature [7,9], the water uptake by the sample can be written in KrögerVink notation as 00
!
ð3Þ
According to Eq. (3), the slope of log x ¼ f log pH2 O ; T depen-
Fig. 1. XRD pattern for as-sintered BCN50.
2OO þ VO þ H2 O¢2OHO þ OV
4x3 DH01 DS01 ¼ 3 exp 4x 48x2 þ 165x 121 RT R
ð1Þ
provided that Ba4Ca2Nb2O11 with 1 structural vacancy, V O , per formula unit is chosen as a reference crystal. The oxygen from the water molecule fills the structural vacancy resulting in negatively 00
charged defect OV . For simplicity, we assume that the defect chemistry of BCN50 at T = 623–773 K and log pH2 O =atm < 1:6 is governed solely by the reaction (1). Corresponding set of equations comprising the
dencies should be close to 1/3 at low level of hydration in agreement with the qualitative assessment of the defect chemistry of BCN50 made by Animitsa et al. [5,7,9]. Standard enthalpy, DH01 , and entropy, DS01 , of reaction (1) were obtained as a result of nonlinear fitting of Eq. (3) to the equilibrium pH2 O T x data for BCN50 reported by Animitsa and Kochetova [7]. The results of such fitting are presented in Fig. 2a and Table 1. A good agreement between the calculated surface and the experimental data, as seen in Fig. 2a, is confirmed by close to unity coefficient of determination R2 = 0.98. In principle, the other defect equilibria, such as those expressed by Eq. (4) or Eq. (5): 00
OO þ VO ¢VO þ OV 0
ð4Þ 00
OO þ OHV ¢OHO þ OV
ð5Þ
could influence the defect chemistry of BCN50. However, it is implied in Eqs. (4) and (5) that there are different sites occupied 0
00
by protons (OHV and OHO ) and oxide ions (O O and OV ). In order to affect the hydration thermodynamics of BCN50, these sites should be regarded as crystallographically and energetically nonequivalent. Otherwise, the said reactions would describe the defect exchange between the equal positions, being essentially meaningless. However, nonequivalent crystallographic positions of oxygen were found only in fully hydrated monoclinic (P21 =n) BCN50 at lower (less than 573 K) temperatures [4,7] but not in the cubic
(Fm 3 m) BCN50 that is the focus of the present discussion. In addition, the reaction (4) is unlikely to proceed at T 773 K, because of the insufficient mobility of oxygen at such a low temperature. Besides, the introduction of either Eq. (4) or Eq. (5) into the defect structure model does not improve the results of the model verification (fitting). Consequently, one can assume that the equilibria (4) and (5) affect the defect chemistry of BCN50 insignificantly, at least, under the conditions discussed in the present work (T = 623–773 K and log pH2 O =atm < 1:6). Partial molar enthalpy, Dh, and entropy, Ds, of water, calculated using pH2 O T x data [7], are shown in Fig. 2b and c, respectively. It is seen that, while Ds increases slightly with water content increase, Dh in BCN50 is almost composition-independent. The average value of Dh, Dhavg (see Table 1), is in good agreement with
DH01 , reinforcing our assumption that the defect structure model based on reaction (1) describes the uptake of water by BCN50 quite well.
99
V. Sereda et al. / Materials Letters 235 (2019) 97–100
Fig. 2. The results of model analysis: points – experimental data [7], surface – fitted model (a) along with the partial molar enthalpy (b) and entropy (c) of water in BCN50.
Table 1 Thermodynamic parameters of hydration of BCN50 (see explanations in the text). Equilibrium pH2 O T x data [7]
DH01 ,
kJmol
1
93.5 ± 9.6 *
DS01 ,
Jmol
Calorimetry 1
K
154.3 ± 15.4
1
1
Dhavg , kJmol
DH0hydr , 692 K, kJmol1
DH0total , 473 K, kJmol1*
DH0tr , kJmol1*
95.9 ± 8.0
107.9 ± 14.4
35.4 ± 3.2
63.9 ± 14.2
kJ/mol of BCN50.
The calorimetric measurements, the results of which are presented in Fig. 3, were performed at two temperatures, 473 and 692 K. At higher temperature, the total heat effect, corresponding to the enthalpy of hydration, DH0hydr , was found to be – 29.13 ± 0.27 kJ/mol of BCN50, or 107.9 ± 14.4 kJ/mol of H2O, because, according to TG, 1 mol of BCN50 absorbs 0.27 ± 0.04 mol of water under the experimental conditions. Larger error of
DH0hydr per mole of H2O is due to the water content measurement uncertainties. It should be emphasized that directly measured DH0hydr and the hydration enthalpies obtained from the equilibrium pH2 O T x data, DH01 and Dhavg , coincide with each other considering their respective uncertainties (see Table 1). It is noteworthy that the DH0hydr of BCN50 is distinctively more exothermal than
DH0hydr reported for the other BCNy (y 0.18) [2,3]. As seen in Fig. 3, the reproducibility of the DSC is somehow lower at 473 K than at 692 K. It is not surprising, considering the sluggish hydration kinetics and the complex nature of hydration processes in BCN50 at 473 K. In the wet atmosphere at T = 473 K,
Fig. 3. DSC measurement results.
100
V. Sereda et al. / Materials Letters 235 (2019) 97–100
BCN50 is almost fully hydrated and possesses monoclinic structure [4,7]. Thus, the hydration at 473 K is accompanied by the phase transition, and the total enthalpy of water uptake, DH0total , the value of which is presented in Table 1, can be expressed as
DH0total ¼ DH0hydr Dx þ DH0tr
ð6Þ
where Dx = 0.92 ± 0.04 is the amount of water absorbed by 1 mol of BCN50 at 473 K as measured by TG, and DH0tr and DH0hydr are the phase transformation (from cubic to monoclinic Ba4Ca2Nb2O110.92H2O) and hydration enthalpies, respectively. Supposedly, the distortions of the crystal lattice of BCN50 caused by the hydration should not lead to significant change in the hydration energetics. In other words, DH0hydr for the cubic and monoclinic BCN50 should be quite close to each other. Neglecting their difference and assuming that DH0hydr is temperature-independent, at least between 473 and 692 K, DH0tr can be calculated using Eq. (6) (see the value of
DH0tr in Table 1). 4. Conclusions The defect structure model for BCN50 based on the single reaction of water uptake was discussed and successfully verified using the proton content dependencies,x T; pH2 O , at 623–773 K [7]. The
values of the hydration enthalpy of cubic (Fm 3 m) BCN50, either measured directly or evaluated using the equilibrium pH2 O T x data [7], were shown to be close to each other. The enthalpy of the cubic!monoclinic phase transition for Ba4Ca2Nb2O110.92H2O was calculated using calorimetrically measured low-temperature heat of hydration. As this transition introduces a degree of disorder, lowering the crystal lattice symmetry, its entropy,
DS0tr ¼ DH0tr =T tr , where T tr is the phase transition temperature, should be positive, making the phase transformation enthalpy DH0tr endothermic. Acknowledgements Sereda V.V. acknowledges the project SP-3103.2018.1 funded by the The Council for grants of the President of Russian Federation. References [1] Y. Du, A.S. Nowick, Structural transitions and proton conduction in nonstoichiometric A3B’B”O9 perovskite-type oxides, J. Am. Ceram. Soc. 78 (11) (1995) 3033–3039. [2] F. Krug, T. Schober, The high-temperature proton conductor Ba3(Ca1. 18Nb1.82)O9-d: thermogravimetry of the water uptake, Solid State Ionics 92 (3) (1996) 297–302. [3] T. Schober, J. Friedrich, The mixed perovskites BaCa(1+x)/3Nb(2–x)/3O3x/2 (x= 0. . .0.18): proton uptake, Solid State Ionics 136–137 (2000) 161–165. [4] I. Animitsa, A. Neiman, N. Kochetova, B. Melekh, A. Sharafutdinov, Proton and oxygen-ion conductivity of Ba4Ca2Nb2O11, Solid State Ionics 162–163 (2003) 63–71. [5] D.V. Korona, A.Y. Neiman, I.E. Animitsa, A.R. Sharafutdinov, Effect of humidity on conductivity of Ba4Ca2Nb2O11 phase and solid solutions based on this phase, Russ. J. Electrochem. 45 (5) (2009) 586–592. [6] N.A. Kochetova, I.E. Animitsa, A.Y. Neiman, The synthesis and properties of solid solutions based on Ba4Ca2Nb2O11, Russ. J. Phys. Chem. A 83 (2) (2009) 203–208. [7] I.E. Animitsa, N.A. Kochetova, Crystal structure and imperfection of the perovskite-like proton conductor Ba4Ca2Nb2O11, Chim. Techno Acta 3 (1) (2016) 5–13. [8] I. Animitsa, T. Denisova, A. Neiman, A. Nepryahin, N. Kochetova, N. Zhuravlev, P. Colomban, States of H+-containing species and proton migration forms in hydrated niobates and tantalates of alkaline-earth metals with a perovskiterelated structure, Solid State Ionics 162–163 (2003) 73–81. [9] I.E. Animitsa, High-temperature proton conductors with structure-disordered oxygen sublattice, Russ. J. Electrochem. 45 (6) (2009) 668–676.