Hostname: page-component-74656f9884-j7cmm Total loading time: 0 Render date: 2026-11-07T14:52:00.069Z Has data issue: false hasContentIssue false
  • English
  • Français

Effects of high-temperature ambient on cyclic fatigue of La0.8Sr0.2MnO3+δ

Published online by Cambridge University Press:  15 August 2011

Makoto Tanaka ,
Tsuneaki Matsudaira ,
Daisuke Igimi  and
Satoshi Kitaoka
Makoto Tanaka*
Affiliation:
Japan Fine Ceramics Center, Atsuta-ku, Nagoya 456-8587, Japan
Tsuneaki Matsudaira
Affiliation:
Japan Fine Ceramics Center, Atsuta-ku, Nagoya 456-8587, Japan
Daisuke Igimi
Affiliation:
Japan Fine Ceramics Center, Atsuta-ku, Nagoya 456-8587, Japan
Satoshi Kitaoka
Affiliation:
Japan Fine Ceramics Center, Atsuta-ku, Nagoya 456-8587, Japan
*
a)Address all correspondence to this author. e-mail: m_tanaka@jfcc.or.jp

Abstract

The effects of water vapor and oxygen on the cyclic fatigue behavior of oxygen-excess La0.8Sr0.2MnO3+δ (LSM) were investigated under three-point bending at 1273 K. Because the fatigue life did not obviously depend on the number of cycles, which also represented the effective time of the applied stress, the fracture was presumed to not be significantly controlled by stress-corrosion cracking. Under a low oxygen partial pressure (${\rm{P}}_{{\rm{O}}_2 } $), however, wet exposure inhibited both fatigue fracture and permanent deformation, in which the LSM crystal lattice was distorted and the unit cell free volume was reduced. Under a high ${\rm{P}}_{{\rm{O}}_2 } $, on the contrary, the crystal symmetry was increased by the wet exposure. The inhibition of fatigue fracture and deformation at both high ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and low ${\rm{P}}_{{\rm{O}}_2 } $ was probably caused by retardation of lanthanum diffusion through its vacancies.

Keywords

CeramicFatigueCrystallographic structure

Information

Type
Articles
Information
Journal of Materials Research , Volume 26 , Issue 18 , 28 September 2011 , pp. 2450 - 2457
Copyright
Copyright © Materials Research Society 2011

I. INTRODUCTION

Superheated steam has attracted considerable attention in recent years for use in sustainable heat treatment, for applications such as food processing, dewaxing, and deodorization because of its higher heat transfer and superior drying and infiltration properties compared to air heating.Reference Schwartze and Bröcker1–Reference Kitaoka, Wada, Kawashima, Osa and Nagai3 Since the perovskite-type oxides of the series La1−xSrxMnO3 (LSM) have high electrical conductivity and excellent corrosion resistance at high temperatures, they are promising candidate heater materials for the generation of clean superheated steam at high temperatures by induction heating (IH), which enables exact control of treatment temperatures.Reference Kitaoka, Wada, Kawashima, Osa and Nagai3 There have been many reportsReference Minh4–Reference Nakamura and Ogawa16 of variations in crystallographic, electrical, and chemical characteristics induced by oxygen nonstoichiometry of LSM for application in solid oxide fuel cell (SOFC) cathodesReference Minh4 and magnetroresistive read heads of electronic storage devices.Reference Urushibara, Morimoto, Arima, Asamitsu, Kido and Tokura5 It is well known, for example, that LSM incorporates oxygen in oxidizing atmospheres and has the nominal composition of “apparent” oxygen-excess La1−xSrxMnO3+δ (δ > 0). The main ionic defects in oxygen-excess LSM are believed to be vacancies at cationic sites, which are charge-compensated by electron holes. On the contrary, LSM becomes oxygen deficient at very low oxygen partial pressures (${\rm{P}}_{{\rm{O}}_2 } $) below 10−4 Pa with x = 0.2 at 1273 K. The equilibrium ${\rm{P}}_{{\rm{O}}_2 } $ value in superheated steam is thermodynamically about 1 Pa at 1273 K under atmospheric pressure, so LSM in a superheated steam heater would certainly reside in an oxygen-excess environment.

There have been limited studies on the effects of oxygen nonstoichiometry and doped cations on the mechanical properties of LSM, primarily involving ${\rm{P}}_{{\rm{O}}_2 } $-controlled dry environments at high temperatures to investigate the use of LSM as a cathode in SOFCs. It was reported that LSM (x = 0.1–0.3) deforms at 1523 K by grain-boundary sliding because of lattice diffusion and possibly cavitation and/or interface reaction control.Reference Cook, Goretta, Wolfenstine, Nash and Routbort17, Reference Routbort, Goretta, Cook and Wolfenstine18 The creep rates of LSM with x = 0.3 in the range 10 ≤ ${\rm{P}}_{{\rm{O}}_2 } $ ≤ 105 Pa and with x ≤ 0.2 in the range 10−2 ≤ ${\rm{P}}_{{\rm{O}}_2 } $ ≤ 102 Pa were presumed to be controlled by cation diffusion through the cation vacancies, but that of x ≤ 0.2 at ${\rm{P}}_{{\rm{O}}_2 } $ > 102 Pa was governed by oxygen-ion diffusion through oxygen vacancies.Reference Cook, Goretta, Wolfenstine, Nash and Routbort17 Because the oxygen-excess region grew at lower ${\rm{P}}_{{\rm{O}}_2 } $ with decreasing annealing temperature,Reference Mizusaki, Mori, Takai, Yonemura, Minamiue, Tagawa, Dokiya, Inaba, Naraya, Sasamoto and Hashimoto13, Reference Mizusaki, Yonemura, Kamata, Ohyama, Mori, Takai, Tagawa, Dokiya, Naraya, Sasamoto, Inaba and Hashimoto14 the diffusion species governing the creep rate appeared to be strongly dependent on the temperature.

The flexural strength of LSM in air tended to increase with increasing testing temperature.Reference Atkinson and Selçuk19–Reference Meixner and Cutler21 This strengthening was speculated to result from the expansion of the unit cell with temperature, relaxation of residual stresses upon heating, etc., although the defect chemistry of LSM strongly depends on temperature. The LaMnO3+δ specimen in the oxygen-excess region expanded with increasing ${\rm{P}}_{{\rm{O}}_2 } $ because of lattice creation on the surfaces, even though the unit cell volume is known to decrease with increasing ${\rm{P}}_{{\rm{O}}_2 } $.Reference Miyoshi, Hong, Yashiro, Kaimai, Nigara, Kawamura, Kawada and Mizusaki15 Therefore, even if an LSM specimen is isothermally exposed during rapid ${\rm{P}}_{{\rm{O}}_2 } $ change, the local stress induced by the volume change in the LSM subsurface may lead to fracture of the LSM. However, there have been no fundamental investigations of the defect chemistry of LSM under wet atmospheres at high temperatures, and no studies have investigated the thermomechanical properties such as flexural strength and fatigue of LSM in such environments to evaluate the applicability and long-term durability of LSM as an IH heater for the production of superheated steam.

It is well known for glasses and oxide ceramics that subcritical crack growth, also called stress-corrosion cracking (SCC), occurs by dissociative adsorption of water molecules onto the strained crack tip bonds of the materials, resulting in a time-dependent reduction in strength called delayed failure.Reference Wiederhorn, Fuller and Thomson22, Reference Freiman, Barkatt, Verink and Smith23 Since SCC is a stress-aided chemical reaction, it is rapidly accelerated by increasing the temperature. Therefore, when LSM specimens are exposed to a rapid ${\rm{P}}_{{\rm{O}}_2 } $ change under isothermal superheated steam, the local stress that is chemically induced by the subsurface volume change may promote SCC.

In this study, the isothermal cyclic fatigue behavior of LSM was evaluated at 1273 K under varying partial pressures of both water vapor (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $) and oxygen (${\rm{P}}_{{\rm{O}}_2 } $). The fatigue mechanism was investigated on the basis of a vacancy excluding model,Reference Mizusaki, Mori, Takai, Yonemura, Minamiue, Tagawa, Dokiya, Inaba, Naraya, Sasamoto and Hashimoto13, Reference Mizusaki, Yonemura, Kamata, Ohyama, Mori, Takai, Tagawa, Dokiya, Naraya, Sasamoto, Inaba and Hashimoto14 which incorporated the interaction between La vacancies and oxygen ions around the vacancies.

II. EXPERIMENTAL PROCEDURES

Commercial LSM powder with doped Sr (Daiichi Kigenso Kagaku Kogyo Co., Ltd., Osaka, Japan) of x = 0.2 was used because this composition has a stable rhombohedral structure ($R\bar 3c$) over a wide ${\rm{P}}_{{\rm{O}}_2 } $ range at 1273 K,Reference Mitchell, Argyriou, Potter, Hinks, Jorgensen and Bader9 which are typical operating conditions for an IH heater. The powder was molded by a uniaxial press at 20 MPa and then subjected to a cold isostatic press at 245 MPa. The green compacts were pressureless sintered in O2 gas flow at 1773 K for 5 h. Specimens with dimensions 4 × 3 × 20 mm3 were cut from the sintered bodies, and a single V-notch with a tip radius, ρ, of about 20 μm was introduced at the center of the tensile surface (4 × 20 mm2) as shown in Fig. 1 so that the notch tip could act as a fracture origin under the influence of the surrounding environment. The relative density of the specimens was 98.3% of their theoretical density, and the crystalline phase was only identified as $R\bar 3c$ by x-ray diffraction analysis. As shown in Fig. 1, the specimens were placed in a three-point bending jig.

FIG. 1.

Schematic of cyclic bending fatigue tests (notch depth: 0.4 mm; notch width: 0.17 mm).

Gases such as Ar, O2, and 0.1% O2–Ar, in which ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ was controlled in the range of 2.7 × 102 and 4 × 103 Pa by bubbling temperature of distilled water through the gases, were blown against the V-notched surface. The specimens were subsequently heated to 1273 K at a heating rate of 10 K/min and maintained at that temperature for 15 min, after that fatigue tests were performed while blowing the gases. A gases flow rate was 0.3 L/min at 303 K; eventually, it was 1.3 L/min at 1273 K because of the expansion of the gases. Before bubbling, each gas contained water vapor as an impurity of the order of ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 10 Pa, which was measured at room temperature using an optical dew-point sensor. The Ar gas also included a tiny amount of O2 (${\rm{P}}_{{\rm{O}}_2 } $ = about 1 Pa), which was monitored by a zirconia oxygen sensor at 973 K. The ${\rm{P}}_{{\rm{O}}_2 } $ values in the gases at 1273 K, which were estimated thermodynamically from the values of ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and ${\rm{P}}_{{\rm{O}}_2 } $ measured at lower temperatures, were nearly the same as the measured values. Therefore, the ${\rm{P}}_{{\rm{O}}_2 } $ values used in this study are considered to be for the oxygen-excess nonstoichiometry for LSM.Reference Mizusaki, Mori, Takai, Yonemura, Minamiue, Tagawa, Dokiya, Inaba, Naraya, Sasamoto and Hashimoto13, Reference Mizusaki, Yonemura, Kamata, Ohyama, Mori, Takai, Tagawa, Dokiya, Naraya, Sasamoto, Inaba and Hashimoto14

The fatigue tests were performed with a maximum load P max of 60 N and a minimum load P min of 6 N and sinusoidal loading cycles of constant amplitude at a frequency of 20 Hz for N = 2 × 106 cycles, which corresponded to 105 s. The fracture load, P f, in the three-point bending test under dry Ar at 1273 K with no cyclic load was ∼70 N so that P max/P f ≈ 0.85. The initial stress intensity factor at the notch tip estimated by an equation proposed for single-edged-cracked specimen under three-point bendingReference Srawley24 is $K_{\rm{I}}^{\max } $ = 1.10 MPam1/2, which induced a maximum load P max of 60 N. The corresponding maximum notch-tip stress, $\sigma _{{\rm{tip}}}^{\max } $, was estimated using the equation σ = 2K I / (πρ)1/2 to be 278 MPa.Reference Wiederhorn and Bolz25 Ten specimens were tested for each condition. The specimens that survived the fatigue tests at P max = 60 N were permanently deformed, whether or not a crack propagated from the notch-tip, where crack propagation was observed with an optical microscopy of both side surfaces of the specimen.

When three specimens were tested at the same temperature at the lower P max of 58 N ($K_{\rm{I}}^{\max } $ = 1.06 MPam1/2, $\sigma _{{\rm{tip}}}^{\max } $ = 268 MPa), which all samples survived and crack propagation did not occur, the permanent deformation of each specimen was measured at room temperature from the profiles of the tensile notched surfaces. Figure 2 shows a typical permanent deformation profile of the tensile surface of a V-notched specimen that survived fatigue testing under a P max of 58 N at 1273 K while the vicinity of the notch tip was exposed to a wet O2 atmosphere (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 4 × 103 Pa; ${\rm{P}}_{{\rm{O}}_2 } $ = 105 Pa).

FIG. 2.

Typical permanent deformation profile on the tensile surface of a V-notched specimen that survived fatigue testing under P max = 58 N at 1273 K while the vicinity of the notch tip was exposed to a wet O2 atmosphere (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 4 × 103 Pa; ${\rm{P}}_{{\rm{O}}_2 } $ = 105 Pa).

The deformation ratio D is defined as

(1)
$$D = {{h_1 + h_2 } \over {2L}}\quad,$$

where L is the lower span of 16 mm in the three-point bending jig, and h 1 and h 2 are the deformations at a distance of L/2 from the notch. In addition, three-point bending tests for specimens that survived fatigue tests at P max = 58 N and 1273 K were performed at a crosshead speed of 0.5 mm/min in the same environment as that used in the fatigue tests, and the corresponding fracture surfaces were observed by scanning electron microscopy (SEM).

To elucidate the effects of ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and ${\rm{P}}_{{\rm{O}}_2 } $ on the oxygen-excess nonstoichiometry around the notch tip, which was the fracture origin in the LSM specimens during fatigue testing, LSM powders were exposed to the same environmental conditions for the same time (105 s) as the fatigue test specimens, and then quenched. The concentrations of La, Sr, and Mn cations in the LSM powders were quantified by inductively coupled plasma spectroscopy. The average Mn valences of the powders were determined by an iodometry titration method described in the literature.Reference Andersen, Andersen, Norby and Skou26 According to the defect model of LSM in an oxygen-excess nonstoichiometry,Reference Mizusaki, Mori, Takai, Yonemura, Minamiue, Tagawa, Dokiya, Inaba, Naraya, Sasamoto and Hashimoto13 the La/Mn site populations for La, Sr, Mn (Mn2+, Mn3+, and Mn4+), and La vacancies were calculated from the corresponding average Mn valences and cation concentrations. Furthermore, room-temperature structural parameters such as the Mn–O bond length, the Mn–O–Mn bond angle, the lattice parameters, and the corresponding unit cell volumes were determined from x-ray Rietveld analysis of the LSM powders, taking into account the La/Mn site populations.

III. RESULTS AND DISCUSSION

A. Cyclic fatigue behavior

The specimens either fractured during the initial cycles (N = 1–67 cycles) or did not fail before the upper cycle limit. Because the fatigue life did not obviously depend on the number of cycles, which also represented the effective time of the applied stress, the fracture was presumed to not be significantly controlled by SCC because of the dissociative adsorption of water molecules at the crack tip. However, the fraction of specimens surviving the fatigue tests depended significantly on the environment surrounding the notch tip. Figure 3 shows the ratio of surviving specimens under P max = 60 N at 1273 K as a function of ${\rm{P}}_{{\rm{O}}_2 } $ around the notch tip. All of the specimens fractured under the high ${\rm{P}}_{{\rm{O}}_2 } $, independent of ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $. On the contrary, under both a lower ${\rm{P}}_{{\rm{O}}_2 } $ and a higher ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $, the ratio of surviving specimens sharply increased.

FIG. 3.

Specimen survival ratio during fatigue testing under P max = 60 N at 1273 K as a function of ${\rm{P}}_{{\rm{O}}_2 }$ around the notch tip; n is the number of specimens tested under each condition.

When P max was decreased slightly from 60 to 58 N, all of the specimens survived and underwent no crack growth from the notch tip, but were obviously deformed. Therefore, catastrophic failure of the specimens would ultimately occur through creep deformation with increasing N, that is, after a longer effective time. Figure 4 shows the deformation ratio D of the surviving specimens under P max = 58 N at 1273 K as a function of ${\rm{P}}_{{\rm{O}}_2 } $ around the notch tip. The D values increased with increasing ${\rm{P}}_{{\rm{O}}_2 } $ under both the low and the high ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ environments. Cook et al.Reference Cook, Goretta, Wolfenstine, Nash and Routbort17 reported that the compressive steady-state creep rate of LSM specimens with x = 0.2 increased with increasing ${\rm{P}}_{{\rm{O}}_2 } $ above 103 Pa, a dry environment at 1523 K. Although temperature dependence of the creep rate is unclear, the reported tendency of the specimens to deform easily under a higher ${\rm{P}}_{{\rm{O}}_2 } $Reference Cook, Goretta, Wolfenstine, Nash and Routbort17 is consistent with our results, as shown in Fig. 4. It was reported that a LaMnO3+δ specimen in the oxygen-excess region contracted with decreasing ${\rm{P}}_{{\rm{O}}_2 } $ because the crystal lattice was dissipated on the surfaces.Reference Miyoshi, Hong, Yashiro, Kaimai, Nigara, Kawamura, Kawada and Mizusaki15 If these contractions occur locally nearby the notch tip, where the specimen is not in chemical equilibrium, a stress will be induced in the crack-closing direction around the tip for the lower ${\rm{P}}_{{\rm{O}}_2 } $. Therefore, it may be harder to deform the specimen tested under the lower ${\rm{P}}_{{\rm{O}}_2 } $ than that tested under the higher ${\rm{P}}_{{\rm{O}}_2 } $. The average D value was decreased by exposure to a wet environment under a low ${\rm{P}}_{{\rm{O}}_2 } $, whereas it was relatively unaffected by exposure to a wet environment under a high ${\rm{P}}_{{\rm{O}}_2 } $.

FIG. 4.

Deformation ratio, D, of the specimens that survived fatigue tests under P max = 58 N at 1273 K as a function of ${\rm{P}}_{{\rm{O}}_2 } $ around the notch tip; n is the number of specimens tested under each condition.

Figure 5 shows SEM micrographs of fracture surfaces after bending tests were conducted on specimens that survived fatigue tests with P max = 58 N at 1273 K. The bending tests were performed in the same environments as those used in the fatigue tests. The broken lines and arrows in the micrographs indicate the edges of the V-notch and the crack propagation direction, respectively. The specimen tested in dry Ar mainly underwent intergranular fracture [Fig. 5(a)], while the specimen tested in wet Ar with a ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 4 × 103 Pa obviously underwent transgranular fracture [Fig. 5(b)]. The fracture surfaces tested in dry and wet O2 had similar morphologies to the intergranular fracture shown in Fig. 5(a). The morphology of the fracture surfaces in Fig. 5 was similar to that of the specimens that were merely exposed without loading in the same environments as those of the fatigue test. Therefore, only wet exposure under a low ${\rm{P}}_{{\rm{O}}_2 } $ seemed to enhance the grain boundary strength relative to the cleavage strength.

FIG. 5.

Scanning electron microscopy micrographs of fracture surfaces after bending tests on specimens that survived fatigue tests under P max = 58 N at 1273 K. The bending tests were carried out in the same environments as the fatigue tests: (a) dry Ar (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 10 Pa; ${\rm{P}}_{{\rm{O}}_2 } $ = 1.5 Pa), (b) wet Ar (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 4 × 103 Pa; ${\rm{P}}_{{\rm{O}}_2 } $ = 1.5 Pa). The broken lines and arrows in the micrographs indicate the edges of the V-notch and the crack propagation directions, respectively.

B. Crystal structural change

Various defect structure models for LSM have been proposed to elucidate the defect chemistry of LSM with an oxygen-excess nonstoichiometry. However, no conclusive defect model for LSM in oxygen-excess nonstoichiometry has been established because of the discrepancy in interpretations on neutron powder diffraction results for LSM.Reference van Roosmalen, Cordfunke, Helmholdt and Zandbergen6–Reference Tofield and Scott8 However, Mizusaki and coworkersReference Kamata, Yonemura, Mizusaki, Tagawa, Naraya and Sasamoto12–Reference Miyoshi, Hong, Yashiro, Kaimai, Nigara, Kawamura, Kawada and Mizusaki15 proposed that the oxygen nonstoichiometry of LSM was caused by the formation of only La vacancies, which were created by site-mixing from La to Mn sites. They introduced the vacancy exclusion model to explain the upper limit for the excess oxygen δ, in which vacancy excluding spaces around the La vacancy and Sr substituted at the La site consisted of nine unit cells and three unit cells of LSM, respectively. The maximum oxygen excess was predicted by the maximum number of vacancy excluding spaces available in the crystal lattice. For LSM at x > 0.4, there was no place for the vacancy excluding space around the La vacancy, resulting in the disappearance of the oxygen-excess nonstoichiometry. The vacancy exclusion model fits its respective experimental results well, and only this modelReference Mizusaki, Mori, Takai, Yonemura, Minamiue, Tagawa, Dokiya, Inaba, Naraya, Sasamoto and Hashimoto13, Reference Mizusaki, Yonemura, Kamata, Ohyama, Mori, Takai, Tagawa, Dokiya, Naraya, Sasamoto, Inaba and Hashimoto14 accounts for the participation of oxygen ions around the cation vacancies. Thus, in this study, the crystal structure of LSM was investigated on the basis of the vacancy excluding model.

Figure 6 shows the mean valence of Mn ions in LSM powders exposed at 1273 K for 105 s as a function of ${\rm{P}}_{{\rm{O}}_2 } $. In the low ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ environments, the mean valence increased with increasing ${\rm{P}}_{{\rm{O}}_2 } $. At a ${\rm{P}}_{{\rm{O}}_2 } $ of 105 Pa, it was nearly consistent with the dotted line in Fig. 6 corresponding to the maximum oxygen excess predicted by the vacancy excluding model.Reference Mizusaki, Mori, Takai, Yonemura, Minamiue, Tagawa, Dokiya, Inaba, Naraya, Sasamoto and Hashimoto13 Therefore, the powders exposed to dry conditions and a high ${\rm{P}}_{{\rm{O}}_2 } $ were considered to be in the equilibrium state. On the other hand, in the high ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ environment, the mean valences under a low ${\rm{P}}_{{\rm{O}}_2 } $ were slightly larger than those in the low ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ environment, but it is worth noting that the valence under the high ${\rm{P}}_{{\rm{O}}_2 } $ of 105 Pa greatly exceeded the predicted maximum value, shown as a dotted line. The molar fractions of the cations and vacancies occupying the La/Mn sites in LSM powders exposed at 1273 K for 105 s are listed in Table I. The La vacancy concentration increased with increasing ${\rm{P}}_{{\rm{O}}_2 } $ and was accelerated by wet exposure.

FIG. 6.

Mean valence of Mn ions in La0.8Sr0.2MnO3+δ (LSM) powders exposed at 1273 K for 105 s as a function of ${\rm{P}}_{{\rm{O}}_2 } $.

Incorporation of oxygen molecules into the LSM lattice in the oxygen-excess nonstoichiometry was presumed to proceed as followsReference Mizusaki, Mori, Takai, Yonemura, Minamiue, Tagawa, Dokiya, Inaba, Naraya, Sasamoto and Hashimoto13, Reference Mizusaki, Yonemura, Kamata, Ohyama, Mori, Takai, Tagawa, Dokiya, Naraya, Sasamoto, Inaba and Hashimoto14:

(2)
$${3 \over 2}{\rm{O}}_2 + {\rm{La}}_{{\rm{La}}}^ \times \to 3{\rm{O}}_{\rm{O}}^ \times + {\rm{La}}_{{\rm{Mn}}}^ \times + 2{\rm{V}}\prime\!\prime\!\prime_{{\rm{\!\!\!\!\!\!\!La}}} + 6{\rm{h}}^ \bullet \quad.$$

The produced holes are mainly expended in oxidation from Mn2+ to Mn3+, while a vacancy excluding space consisting of nine unit cells (nine La ions) is formed around the La vacancy. On the other hand, incorporation of water molecules into the LSM lattice should be accompanied by the formation of H+ as follows:

(3)
$$3{\rm{H}}_2 {\rm{O}} + {\rm{La}}_{{\rm{La}}}^ \times \to 3{\rm{O}}_{\rm{O}}^ \times + {\rm{La}}_{{\rm{Mn}}}^ \times + 2{\rm{V}}\prime\!\prime\!\prime_{{\rm{\!\!\!\!\!\!La}}} + 6{\rm{H}}^ \bullet\quad .$$

The vacancy excluding spaces would likely be reduced by the formation of complex defects such as$\left[ {\rm{V}}′\!′\!′_{{{\!\!\!\!\!\!\rm La}}} \bullet {\rm H}^ \bullet } \right]^{\prime \!\prime } $, $\left[ {\rm{V}}′\!′\!′_{{{\!\!\!\!\!\!\rm La}}} \bullet 2{\rm H}^ \bullet } \right]^\prime $, and $\left[ {\rm{V}}′\!′\!′_{{{\!\!\!\!\!\!\rm La}}} \bullet 3{\rm H}^ \bullet } \right]^ \times $, resulting in still more incorporation of the coexisting oxygen molecules according to Eq. (2). In other words, water molecules are incorporated into the lattice without the direct oxidation of Mn ions because of the absence of holes. Thus, the high mean Mn valence that exceeded the predicted maximum value at both high ${\rm{P}}_{{\rm{O}}_2 } $ and high ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ was probably caused by an accelerated incorporation of coexisting oxygen molecules by reduction of the vacancy excluding spaces because of complex defect formation. At a low ${\rm{P}}_{{\rm{O}}_2 } $, on the other hand, a slight increase in the mean Mn valence by the wet treatment was probably related to the small number of La vacancies, as listed in Table I, and the surrounding oxygen molecules to incorporate to the lattice, even though the vacancy excluding spaces were diminished by the wet exposure.

TABLE I.

Molar fractions of cations and vacancies occupying the A and B sites in La0.8Sr0.2MnO3+δ powders exposed at 1273 K for 105 s.

The room-temperature x-ray diffraction patterns of all LSM powders that had been exposed at 1273 K for 105 s were identified as $R\bar 3c$, and there were no other crystalline phases present. Figure 7 shows the structural parameters determined from x-ray Rietveld analysis of the LSM powders based on the La/Mn site populations listed in Table I. Table II summarizes the corresponding lattice parameters and the fitness agreement index (R wp/R e). R wp/R e ≤ 1.3 is generally considered to be quite satisfactory, while R wp/R e = 1.0 indicates an ideal fit. In the low ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ atmospheres, both the Mn–O bond length and the Mn–O–Mn bond angle were nearly constant with increasing ${\rm{P}}_{{\rm{O}}_2 } $ up to 102 Pa, whereas the bond length and bond angle at ${\rm{P}}_{{\rm{O}}_2 } $ = 105 Pa were clearly shorter and larger than those below ${\rm{P}}_{{\rm{O}}_2 } $ = 102 Pa, respectively. The bond lengths in the high ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ atmospheres under a low ${\rm{P}}_{{\rm{O}}_2 } $ were longer than those in the dry environments, but they were shorter under a high ${\rm{P}}_{{\rm{O}}_2 } $. The decrease in the bond angle indicates an increase in the crystallographic distortion for the perovskite-type oxides of the LSM series. Therefore, the LSM crystal was evidently distorted by wet exposure at a low ${\rm{P}}_{{\rm{O}}_2 } $, but the crystal symmetry was increased under a high ${\rm{P}}_{{\rm{O}}_2 } $. It is more remarkable that the effect of water vapor on the structural parameters under a low ${\rm{P}}_{{\rm{O}}_2 } $ was the opposite of its effect under a high ${\rm{P}}_{{\rm{O}}_2 } $. These structural changes depend on whether Mn2+ ions exist in the LSM crystal (see Table I).

FIG. 7.

Structural parameters for LSM powders exposed at 1273 K for 105 s as a function of ${\rm{P}}_{{\rm{O}}_2 } $.

TABLE II.

Lattice parameters and the fitness agreement index (R wp/R e) for LSM powders exposed at 1273 K for 105 s.

De Souza et al.Reference De Souza, Islam and Ivers-Tiffée10 investigated the defect chemistry of LaMnO3 by the static lattice calculations and reported that the migration energy of the La vacancy tended to increase when the crystal lattice became more distorted. It was also reported that the LSM deformed by grain-boundary sliding accommodated by lattice diffusion, with some possible cavitation and/or interface reaction control.Reference Cook, Goretta, Wolfenstine, Nash and Routbort17, Reference Routbort, Goretta, Cook and Wolfenstine18 Therefore, the increase in the specimen survival ratio and the deformation resistance under higher ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and lower ${\rm{P}}_{{\rm{O}}_2 } $ shown in Figs. 3 and 4 was probably caused by an inhibition of grain-boundary sliding, retarding La migration through La vacancies because of an increase in crystal distortion, as shown in Fig. 7. The transgranular fracture mainly observed in specimens tested under a wet environment at a low ${\rm{P}}_{{\rm{O}}_2 } $, as shown in Fig. 5, may be related to the suppression of grain-boundary sliding.

C. Unit cell free volume

In general, the ionic conductivity of oxides with structural defects strongly depends on the corresponding unit cell free volume V f.Reference Sammells, Cook, White, Osborne and MacDuff27, Reference Nomura and Tanase28 As mentioned above, the high-temperature deformation of LSM in an oxygen-excess nonstoichiometry was considered to be controlled by migration of the cations through their vacancies.Reference Cook, Goretta, Wolfenstine, Nash and Routbort17, Reference Routbort, Goretta, Cook and Wolfenstine18 Therefore, the fatigue degradation mechanisms in this study will be discussed with regard to variations of V f in exposed LSM powders. V f was defined here as the difference between the unit cell volume and the sum of volumes occupied by the individual ions in the unit cell, as

(4)
$$V_{\rm{f}} = {V \over z}-{{4\pi } \over 3}\left({\Sigma M_{\rm{i}} r_{\rm{i}}^3 + 3r_{\rm{o}}^3 } \right)\quad,$$

where V is the unit cell volume, z is the number of chemical formula units per unit cell, M i is the cation molar fraction at each site, and r i and r o are ionic radiiReference Shannon29 of cation and oxygen, respectively. M i and V are listed in Tables I and II, respectively. Figure 8 shows the fraction specimens that survived the fatigue tests under P max = 60 N at 1273 K as a function of V f. In the low ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ environments, V f at a low ${\rm{P}}_{{\rm{O}}_2 } $ was smaller than that at a high ${\rm{P}}_{{\rm{O}}_2 } $. Wet exposure (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 4 × 103 Pa) decreased V f at a low ${\rm{P}}_{{\rm{O}}_2 } $, but increased it at a high ${\rm{P}}_{{\rm{O}}_2 } $. The influence of water vapor on V f at low and high ${\rm{P}}_{{\rm{O}}_2 } $ was opposite, which was similar to the influence of water vapor on the structural parameters shown in Fig. 7. The reduction of V f may be related to an increase in crystal distortion in addition to a depletion of La vacancies. The specimen survival rate increased sharply with decreasing V f, as shown in Fig. 8.

FIG. 8.

Specimen survival ratio in fatigue tests under P max = 60 N at 1273 K as a function of V f; n is the number of specimens tested under each condition.

Figure 9 shows the deformation ratio D of the surviving specimens after fatigue tests under P max = 58 N at 1273 K as a function of V f. The smaller V f was, the smaller D became. Although D varied widely in wet exposure at both low and high ${\rm{P}}_{{\rm{O}}_2 } $, the reason for this variation remains unclear. The retardation of La diffusion by the reduction of V f is probably responsible for the suppression of deformation. Therefore, because LSM heaters are a promising candidate for use in superheated steam in which ${\rm{P}}_{{\rm{O}}_2 } $ is thermodynamically expected to be approximately 1 Pa at 1273 K under atmospheric pressure (similar to the low ${\rm{P}}_{{\rm{O}}_2 } $ condition examined in this study), they should be more difficult to deform and/or fracture than in high-temperature air.

FIG. 9.

Deformation ratio D of specimens that survived fatigue tests under P max = 58 N at 1273 K as a function of V f; n is the number of specimens tested under each condition.

IV. CONCLUSIONS

The effects of ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and ${\rm{P}}_{{\rm{O}}_2 } $ on the cyclic fatigue behavior of V-notched LSM specimens in an oxygen-excess nonstoichiometry were investigated at 1273 K. Fatigue tests were carried out under three-point bending at a frequency of 20 Hz with sinusoidal loading cycles of constant amplitude until the number of cycles N reached 2 × 106. Although LSM did not exhibit delayed failure caused by SCC, its fracture and deformation behavior depended strongly on ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and ${\rm{P}}_{{\rm{O}}_2 } $ near the V-notched tensile surface. The ratio of surviving specimens increased at higher ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and lower ${\rm{P}}_{{\rm{O}}_2 } $. The permanent deformation ratio was decreased by the wet exposure at low ${\rm{P}}_{{\rm{O}}_2 } $, but was relatively unaffected at high ${\rm{P}}_{{\rm{O}}_2 } $. The wet exposure resulted in additional distortion of the crystal lattice at low ${\rm{P}}_{{\rm{O}}_2 } $, but increased the crystal symmetry at high ${\rm{P}}_{{\rm{O}}_2 } $. The inhibition of fatigue fracture and deformation at both high ${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ and low ${\rm{P}}_{{\rm{O}}_2 } $ was probably caused by retardation of lanthanum diffusion through its vacancies because of increased distortion of the LSM crystal lattice and the reduction of the unit cell free volume.

ACKNOWLEDGMENTS

This work was supported by the Ministry of Education, Culture, Sports, Science and Technology, Regional Innovation Cluster Program (City-area type) for Western Tono area in Gifu prefecture. The authors are grateful to Professor J. Mizusaki, Tohoku University, Japan, for valuable discussions and advice during the course of this research.

References

REFERENCES

1.Schwartze, J.P. and Bröcker, S.: The evaporation of water into air of different humidities and the inversion temperature phenomenon. Int. J. Heat Mass Transfer 43, 1791 (2000).CrossRefGoogle Scholar
2.Iyota, H., Nishimura, N., Onuma, T., and Nomura, T.: Drying of sliced raw potatoes in superheated steam and hot air. Drying Tech. 19, 1411 (2001).CrossRefGoogle Scholar
3.Kitaoka, S., Wada, M., Kawashima, N., Osa, N., and Nagai, T.: Development of ceramic induction heater. FC Report. Jpn. Fine Ceram. Assoc. 28, 145 (2010).Google Scholar
4.Minh, N.Q.: Ceramic fuel cells. J. Am. Ceram. Soc. 76, 563 (1993).CrossRefGoogle Scholar
5.Urushibara, A., Morimoto, Y., Arima, T., Asamitsu, A., Kido, G., and Tokura, Y.: Insulator-metal transition and giant magnetoresistance in La1-xSrxMnO3. Phys. Rev. B 51, 14103 (1995).CrossRefGoogle ScholarPubMed
6.van Roosmalen, J.A.M., Cordfunke, E.H.P., Helmholdt, R.B., and Zandbergen, H.W.: The defect chemistry of LaMnO3±δ: 2. Structural aspects of LaMnO3±δ. J. Solid State Chem. 110, 100 (1994).CrossRefGoogle Scholar
7.Alonso, J.A., Martínez-Lope, M.J., Casais, M.T., MacManus-Driscoll, J.L., de Silva, P.S.I.P.N., Cohen, L.F., and Fernández-Díaz, M.T.: Non-stoichiometry, structural defects and properties of LaMnO3+δ with high δ values (0.11≤d≤0.29). J. Mater. Chem. 7, 2139 (1997).CrossRefGoogle Scholar
8.Tofield, B.C. and Scott, W.R.: Oxidative nonstoichiometry in perovskites, an experimental survey; the defect structure of an oxidized lanthanum manganite by powder neutron diffraction. J. Solid State Chem. 10, 183 (1974).CrossRefGoogle Scholar
9.Mitchell, J.F., Argyriou, D.N., Potter, C.D., Hinks, D.G., Jorgensen, J.D., and Bader, S.D.: Structural phase diagram of La1-xSrxMnO3+δ: Relationship to magnetic and transport properties. Phys. Rev. B 54, 6172 (1996).CrossRefGoogle Scholar
10.De Souza, R.A., Islam, M.S., and Ivers-Tiffée, E.: Formation and migration of cation defects in the perovskite oxide LaMnO3. J. Mater. Chem. 9, 1621 (1999).CrossRefGoogle Scholar
11.van Roosmalen, J.A.M., and Cordfunke, E.H.P.: The defect chemistry of LaMnO3±δ: 4. Defect model for LaMnO3±δ. J. Solid State Chem. 110, 109 (1994).CrossRefGoogle Scholar
12.Kamata, H., Yonemura, Y., Mizusaki, J., Tagawa, H., Naraya, K., and Sasamoto, T.: High temperature electrical properties of the perovskite-type oxide La1-xSrxMnO3-d. J. Phys. Chem. Solids 56, 943 (1995).CrossRefGoogle Scholar
13.Mizusaki, J., Mori, N., Takai, H., Yonemura, Y., Minamiue, H., Tagawa, H., Dokiya, M., Inaba, H., Naraya, K., Sasamoto, T., and Hashimoto, T.: Oxygen nonstoichiometry and defect equilibrium in the perovskite-type oxides La1-xSrxMnO3+d. Solid State Ionics 129, 163 (2000).CrossRefGoogle Scholar
14.Mizusaki, J., Yonemura, Y., Kamata, H., Ohyama, K., Mori, N., Takai, H., Tagawa, H., Dokiya, M., Naraya, K., Sasamoto, T., Inaba, H., and Hashimoto, T.: Electronic conductivity, Seebeck coefficient, defect and electronic structure of nonstoichiometric La1-xSrxMnO3. Solid State Ionics 132, 167 (2000).CrossRefGoogle Scholar
15.Miyoshi, S., Hong, J.-O., Yashiro, K., Kaimai, A., Nigara, Y., Kawamura, K., Kawada, T., and Mizusaki, J.: Lattice creation and annihilation of LaMnO3+δ caused by nonstoichiometry change. Solid State Ionics 154–155, 257 (2002).CrossRefGoogle Scholar
16.Nakamura, K. and Ogawa, K.: Excess oxygen in LaMnO3+δ. J. Solid State Chem. 163, 65 (2002).CrossRefGoogle Scholar
17.Cook, R.E., Goretta, K.C., Wolfenstine, J., Nash, P., and Routbort, J.L.: High-temperature deformation and defect chemistry of (La1-xSrx)1-yMnO3+δ. Acta Mater. 47, 2969 (1999).CrossRefGoogle Scholar
18.Routbort, J.L., Goretta, K.C., Cook, R.E., and Wolfenstine, J.: Deformation of perovskite electronic ceramics—A review. Solid State Ionics 129, 53 (2000).CrossRefGoogle Scholar
19.Atkinson, A. and Selçuk, A.: Mechanical behavior of ceramic oxygen ion-conducting membranes. Solid State Ionics 134, 59 (2000).CrossRefGoogle Scholar
20.D’Souza, C.M. and Sammes, N.M.: Mechanical properties of strontium-doped lanthanum manganite. J. Am. Ceram. Soc. 83, 47 (2000).CrossRefGoogle Scholar
21.Meixner, D.L. and Cutler, R.A.: Sintering and mechanical characteristics of lanthanum strontium manganite. Solid State Ionics 146, 273 (2002).CrossRefGoogle Scholar
22.Wiederhorn, S.M., Fuller, E.R., and Thomson, R.: Micromechanisms of crack growth in ceramics and glasses in corrosive environment. Meat Sci. 14, 450 (1980).CrossRefGoogle Scholar
23.Freiman, S.W.: Environmentally enhanced fracture of ceramics, in Materials Stability and Environmental Degradation, edited by Barkatt, A., Verink, E.D. Jr., and Smith, L.R. (Mater. Res. Soc. Symp. Proc. 125, Pittsburgh, PA, 1988), p. 205.Google Scholar
24.Srawley, J.E.: Wide range stress intensity factor expressions for ASTM E 399 standard fracture toughness specimens. Int. J. Fract. 12, 475 (1976).CrossRefGoogle Scholar
25.Wiederhorn, S.M. and Bolz, L.H.: Stress corrosion and static fatigue of glass. J. Am. Ceram. Soc. 53, 543 (1970).CrossRefGoogle Scholar
26.Andersen, I.G.K., Andersen, E.K., Norby, P., and Skou, E.: Determination of stoichiometry in lanthanum strontium manganates(III)(IV) by wet chemical methods. J. Solid State Chem. 113, 320 (1994).CrossRefGoogle Scholar
27.Sammells, A.F., Cook, R.L., White, J.H., Osborne, J.J., and MacDuff, R.C.: Relational selection of advanced solid electrolytes for intermediate temperature fuel cells. Solid State Ionics 52, 111 (1992).CrossRefGoogle Scholar
28.Nomura, K., and Tanase, S.: Electrical conduction behavior in (La0.9Sr0.1)MIIIO3-δ(MIII=Al, Ga, Sc, In, and Lu) perovskites. Solid State Ionics 98, 229 (1997).CrossRefGoogle Scholar
29.Shannon, R.D.: Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallogr. A 32, 751 (1976).CrossRefGoogle Scholar
Figure 0

FIG. 1. Schematic of cyclic bending fatigue tests (notch depth: 0.4 mm; notch width: 0.17 mm).

Figure 1

FIG. 2. Typical permanent deformation profile on the tensile surface of a V-notched specimen that survived fatigue testing under Pmax = 58 N at 1273 K while the vicinity of the notch tip was exposed to a wet O2 atmosphere (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 4 × 103 Pa; ${\rm{P}}_{{\rm{O}}_2 } $ = 105 Pa).

Figure 2

FIG. 3. Specimen survival ratio during fatigue testing under Pmax = 60 N at 1273 K as a function of ${\rm{P}}_{{\rm{O}}_2 }$ around the notch tip; n is the number of specimens tested under each condition.

Figure 3

FIG. 4. Deformation ratio, D, of the specimens that survived fatigue tests under Pmax = 58 N at 1273 K as a function of ${\rm{P}}_{{\rm{O}}_2 } $ around the notch tip; n is the number of specimens tested under each condition.

Figure 4

FIG. 5. Scanning electron microscopy micrographs of fracture surfaces after bending tests on specimens that survived fatigue tests under Pmax = 58 N at 1273 K. The bending tests were carried out in the same environments as the fatigue tests: (a) dry Ar (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 10 Pa; ${\rm{P}}_{{\rm{O}}_2 } $ = 1.5 Pa), (b) wet Ar (${\rm{P}}_{{\rm{H}}_2 {\rm{O}}} $ = 4 × 103 Pa; ${\rm{P}}_{{\rm{O}}_2 } $ = 1.5 Pa). The broken lines and arrows in the micrographs indicate the edges of the V-notch and the crack propagation directions, respectively.

Figure 5

FIG. 6. Mean valence of Mn ions in La0.8Sr0.2MnO3+δ (LSM) powders exposed at 1273 K for 105 s as a function of ${\rm{P}}_{{\rm{O}}_2 } $.

Figure 6

TABLE I. Molar fractions of cations and vacancies occupying the A and B sites in La0.8Sr0.2MnO3+δ powders exposed at 1273 K for 105 s.

Figure 7

FIG. 7. Structural parameters for LSM powders exposed at 1273 K for 105 s as a function of ${\rm{P}}_{{\rm{O}}_2 } $.

Figure 8

TABLE II. Lattice parameters and the fitness agreement index (Rwp/Re) for LSM powders exposed at 1273 K for 105 s.

Figure 9

FIG. 8. Specimen survival ratio in fatigue tests under Pmax = 60 N at 1273 K as a function of Vf; n is the number of specimens tested under each condition.

Figure 10

FIG. 9. Deformation ratio D of specimens that survived fatigue tests under Pmax = 58 N at 1273 K as a function of Vf; n is the number of specimens tested under each condition.