Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 43

VLHENMA HKI 18-19 BBKH 1.

Author’s Accepted Manuscript

Mechanical aspects of ceramic membrane materials

J. Malzbender

www.elsevier.com/locate/ceri

PII: S0272-8842(16)30074-8
DOI: http://dx.doi.org/10.1016/j.ceramint.2016.02.136
Reference: CERI12327
To appear in: Ceramics International

Received date: 29 January 2016


Accepted date: 23 February 2016
Cite this article as: J. Malzbender, Mechanical aspects of ceramic membran
m a t e r i a l s , Ceramics Internationa
http://dx.doi.org/10.1016/j.ceramint.2016.02.136
This is a PDF file of an unedited manuscript that has been accepted fo
publication. As a service to our customers we are providing this early version o
the manuscript. The manuscript will undergo copyediting, typesetting, an review
of the resulting galley proof before it is published in its final citable form Please
note that during the production process errors may be discovered whic could
affect the content, and all legal disclaimers that apply to the journal pertain

NGUYỄN ANH PHƯƠNG – 16146449 1


VLHENMA HKI 18-19 BBKH 1.5

Mechanical aspects of ceramic membrane materials

J. Malzbender

Forschungszentrum Jülich, IEK-2, 52425 Jülich, Germany

Abstract

Interest in ceramic transport membrane materials has increased significantly leading also

questions with respect to mechanical reliability and robustness, requiring knowledge of the

mechanical properties. The current review focuses on the mechanical properties of such

ceramics, emphasizing in particular relationships between mechanical properties, non-elastic

effects, phase changes and materials’ stability. Room and elevated temperature application is

considered with a main emphasis on elastic and creep deformation as well as fracture.

Consideration is given to dense membranes as well as porous substrate materials for advanced

asymmetric concepts. Properties are summarized for selected oxygen and proton conductors.

Furthermore, mechanical properties of some selected porous ceramic and metallic substrate

materials are given. In addition to the failure probability associated with the Weibull

distribution of fracture stresses, creep rupture of dense materials and enhanced creep

deformation of porous materials are aspects that need special consideration in the application

of these materials in gas separation systems.

Keywords: Ceramics, membranes, mechanical properties, stability, fracture

1. Introduction

Mixed ionic–electronic conducting (MIEC) materials operating at high temperatures have,

since they provide high counteractive mobility of ions and electrons, high potential for the use

in advanced electrochemical systems, in particular, as oxygen transport membranes for

oxygen separation from ambient air and potential use combustion in pure oxygen in Oxy-fuel

1 2 3 4 5
power plants [ ] and [ ] in catalytic membrane reactors [ ], [ ] and [ ], and as a cathode in

NGUYỄN ANH PHƯƠNG – 16146449 2


VLHENMA HKI 18-19 BBKH 1.5

6
solid oxide fuel cells (SOFCs) [ ]. Oxygen production using ceramic mixed ionic–electronic

conducting membranes can be a cost-effective alternative to cryogenic methods. Since H 2-

related technologies gain growing attention due to the increasing importance of H 2 as the
environmentally friendly energy source of the future with great social impact and economical
7
relevance [ ], selected proton conductors are also considered in the current work.

Among many available MIEC compounds mainly two groups of materials, the perovskite ABO 3-

type and the fluorite AO2-type structure [3], appear to have the most promising oxygen

permeation properties. In fact, cubic SrCo 0.8Fe0.2O3−δ was found to exhibit the highest oxygen

8
permeation flux values among the perovskite-structured materials [ ]. Since phase stability of this
compound appears to be an issue, Sr was partially replaced in later works. Currently, the most
9
promising membrane material due to its high oxygen permeation rates [ ] is

Ba0.5Sr0.5Co0.8Fe0.2O3-δ (BSCF). However, it appears not to be chemically stable under high

oxygen chemical gradients (e.g. air/methane) and against CO 2, hence, it can only be used in Oxy-
fuel process without flue gas recirculation, i.e. a three-end concept where no contact between flue
10
gas and membrane occurs without presence of CO 2 [ ]. La0.6Sr0.4Co0.2Fe0.8O3-δ and the variation

11
La0.58Sr0.4Co0.2Fe0.8O3−δ (LSCF) have lower permeation rates than BSCF [ ], but are expected

12
to be stable under CO2 containing atmospheres [ ]. Recent the potential of oxygen separation by
monolithic doped/multidoped ceria membranes has been shown, where in particular, gadolinium

doped ceria (Ce0.9Gd0.1O1.95−δ, CGO) was suggested as oxygen transport material for syngas

13 14
applications [ , ]. Among the fluorite-structured materials, 20% Pr-doped CeO 2−δ is reported to
possess one of the highest permeation fluxes

[9]. However, this material has a very large chemical expansion at elevated temperatures,

which might result in mechanical reliability issues under application conditions. Such

unfavorable chemical-expansion effects can be reduced by partial replacement of Pr-doping

15
with Gd [ ]. However, this modification decreases ionic and electronic conductivities.

NGUYỄN ANH PHƯƠNG – 16146449 3


VLHENMA HKI 18-19 BBKH 1.5

Lanthanum tungstate LaxWO3+1.5x ( , LaWO) is a promising membrane candidate for a

16
temperature range 600-800 °C [7, ]. LWO has an appreciable chemical and (hydro-)thermal

stability, as well as high values of mixed proton-electronic conductivity [7], which are
advantageous for its application, especially under highly reducing and sulfur-containing gas

17
environments. Nd5.5WO11.25−δ has been suggested as alternative material [ ].

Within the bulk-diffusion limited permeation range, the oxygen flux is inversely proportional

to the membrane thickness. Hence, advanced membrane designs are based on thin, defect-free

18
membrane layers mechanically supported by porous substrates [9, ]. The porous substrate

should provide mechanical stability of the membrane structure, since application-related

loads, related to thermal or chemical strains, pressure gradients between feed and permeate

side, cannot be sustained by a thin membrane layer. In order to reduce the materials’ costs for

such a solution, especially with respect to the amount of very expensive oxides, MgO or

19 20
Al2O3 are alternative substrate materials or ferritic steels [ , ]. Where obviously in addition

to the chemical strain of the dense membrane also the thermal expansion mismatch of layer

and substrate needs consideration. In order to limit the review chemical and thermal

expansion is not reviewed in the current work, data can be found for example in [9].

Micromechanical properties are usually assessed via indentation testing, yielding elastic moduli

21 22 23 24 25
[ , , , , ], an important parameter for analysis and simulation of stress states and from length

of indentation induced cracks, as measure of crack growth behavior, the fracture toughness of the

26 27
material [ , ]. Long-term reliability of the component is obviously an important aspect, which

depends on the mechanical properties of membrane materials and in the case of supported

28 29 30 31
structures, also on the properties of substrate materials [ , , , ]. With respect to failure, the

fracture strength of materials is typically characterized, requiring an

NGUYỄN ANH PHƯƠNG – 16146449 4


VLHENMA HKI 18-19 BBKH 1.5

assessment of fracture stresses using different methods (where mostly experimentally simpler

bending methods are preferred) and a subsequent statistical analysis via a Weibull approach

32 33 34
[ , , ]. Such data are representative for the material in a particular geometry and loading

state, hence to assess the reliability in application rescaling of the strength to the different

loading situation and geometry via effective volumes or areas is necessary (depending if

35 36
volume or area defects) [ , ]. Reliability does not only depend on the initial strength, but

37
also on subcritical crack growth (SCG) [ ]. The associated correlation between loading rate

and strength is important to assess effects associated with thermally induced stresses of start-

up and shut-down process of units and can be used as a basis for a strength–probability–time

(SPT) lifetime prediction, incorporate the time dependence of strength into failure statistics,

and as basis to assess failure relevance of transient and service stages. Another important

thermo-mechanical aspect with respect to structural stability at elevated temperatures is creep

38 39
resistance [ , ]. A strain of 1 % per year in a compressive mode has been suggested [15] as

a tolerable creep rate to warrant long-term structural stability in membrane application.

The aim of the current review is summarize elastic and fracture properties, with emphasis on

origins of anomalies, sensitivity to subcritical crack growth and creep behavior of fluorite-and

perovskite-based dense ceramic MIEC membranes and possible substrate materials. Porosity

is known to have a large effect on the mechanical properties and is therefore discussed where

data on dense and porous state are available.

2.1. BSCF and BSCFZ

The perovskite-structured barium strontium cobalt iron oxide (Ba–Sr–Co–Fe–O) has attracted a

lot of attention as a promising candidate material for oxygen separation membrane applications

and also as novel cathode material for intermediate-temperature solid oxide fuel cells [9], where

among various compositions, Ba0.5Sr0.5Co0.8Fe0.2O3−δ (BSCF) has emerged to

NGUYỄN ANH PHƯƠNG – 16146449 5


VLHENMA HKI 18-19 BBKH 1.5

be one of the best candidates for both applications [8], indicating the importance of oxygen

non-stoichiometry and B-site cation valence.

The observed high permeation rates are mainly attributed to oxygen vacancies in the cubic

40
perovskite structure [ ], whose amount significantly increases with temperature. However, in

order to be economically justified, oxygen transporting membranes (OTMs) and SOFCs


should operate at temperatures between 773 and 1123 K. Permeation rates in this temperature
range show a degradation which is attributed to a partial phase decomposition of the cubic

41
structure into a hexagonal phase [ ], where the change of radius and spin state of Co is the

main driving force for this phase decomposition. Many scientific studies were performed in

42
order to find a way to eliminate this effect. One promising solution [ ] seems to be the partial

substitution of B-site cations by fixed-valent cations which counteract the radius decrease

3+
during the oxidation to low-spin Co , where the optimal partial substitution to preserve long-

term stability of the cubic structure can be achieved by doping with 3 mol% of zirconium,

which results in chemical composition Ba0.5Sr0.5(Co0.8Fe0.2)0.97Zr0.03O3−δ (BSCFZ).

Several studies concentrated onto mechanical properties of BSCF, in particular fracture and high-

temperature creep [15,24,29,37,38,39]. For real components the mechanical integrity of this brittle

perovskite will be challenged by the boundary conditions of transient and stationary temperature

exposure. In particular, long-term failure mechanisms such as static room temperature fatigue and

creep rupture at operation temperature may occur. Anomalous mechanical behavior observed at

43 44 45 46
rather low temperatures [ , , , ] was attributed to a spin-state transition of B-site ions.

Detailed results and implications are discussed in the following.

2.1.1. Elastic and fracture behavior

NGUYỄN ANH PHƯƠNG – 16146449 6


VLHENMA HKI 18-19 BBKH 1.5

Different mechanical testing methods have been used to assess the elastic behavior of the

47
mixed ion electron conducting perovskite BSCF in [ ]. Different experimental methods were

applied for disc shape and tubular geometries from RT to 900 °C on materials provided by

different suppliers. The preparation method, i.e. extrusion as compared to hot pressing did not

influence the materials properties. Although the stress field in terms of the combination of

compressive and tensile stresses as compared to pure compressive stress differed in the used

tests, there was overall agreement with respect to the determined elastic properties as well as

its temperature dependence, implying that the material does not experience a stress dependent

stiffness changes under the testing conditions.

Elastic modulus and internal friction of (BSCF) were determined from resonance frequencies

and damping behavior, respectively, using an impulse excitation method in [46]. An elastic

anomaly around 476 K, with a corresponding peak in the internal friction, was attributed to an

3+
experimentally confirmed spin transition of Co in agreement with finding reported in [44]

based on ring-on-ring tests and depth-sensitive micro-indentation. Measurements of the

thermo-mechanical properties with ring-on-ring test and depth-sensitive micro-indentation

testing revealed an anomaly between 200 °C and 400 °C also in fracture stress and toughness

(Fig. 1) [44,45,46]. At higher temperatures the properties increased.

The effect of subcritical (slow) crack growth at room temperature has been investigated using

fracture stresses obtained in biaxial bending under different loading rates in [37,29]. The

provided data permitted to assess the fracture stresses for different loading rates. Furthermore,

a strength–probability–time plot (SPT in Fig. 2) was derived that permits a prediction of the

lifetime under static loading conditions and hence the long-term reliability at room

temperature. The provided fracture data permit an assessment of the fracture stress in a fast

loading situation. For example in a rather fast loading situation (3 MPa/s) the fracture stress

NGUYỄN ANH PHƯƠNG – 16146449 7


VLHENMA HKI 18-19 BBKH 1.5

for a failure probability of 1 % will be ∼46 MPa implying that the material will sustain this
stress if loaded within 15 s. Slower loading decreases this stress, faster loading increases it. In
a loading situation under constant load the failure probability has been assessed using the
strength–probability–time analysis, for example for a failure probability of 1 % this yields a
tolerable stress limit of 25 MPa if a lifetime of 40 years is envisaged. The results imply a
rather low risk of failure for BSCF due to subcritical crack growth.

2.1.2. Porous BSCF

The mechanical stability of porous BSCF material was studied using depth-sensitive micro-
48 49
indentation and ring-on-ring biaxial bending in [ , ]. The porous BSCF was characterized as
potential substrate for the deposition of a dense membrane layer. Indentation tests yielded
values for hardness and fracture toughness up to a temperature of 400 °C, while bending tests
permitted an assessment of elastic modulus and fracture stress up to 800 °C. In addition, the
fracture toughness was evaluated up to 800 °C via bending tests of pre-indented specimens.
Compared to dense BSCF, fracture stress and elastic modulus of porous BSCF (porosity ∼38
%) are ∼50–60% lower for the entire temperature range. At 800 °C fracture toughness values
of porous and dense material are similar. The temperature dependencies of elastic modulus,
fracture strength and toughness were similar to those of dense BSCF (see Fig. 3).

Porous and dense BSCF ceramics were also characterized with respect to their mechanical

behavior and failure limits at room temperature in [29]. Porous materials seemed to be less

sensitive to subcritical crack growth, which might be associated with pores blunting the crack

tip. SPT analysis yielded for porous BSCF a tolerable static stress of 9 MPa (lower than dense

BSCF due to porosity) for a failure probability of 1 % if a lifetime of 40 years is envisaged.

2.1.3. Intermediate temperature anomalies

NGUYỄN ANH PHƯƠNG – 16146449 8


VLHENMA HKI 18-19 BBKH 1.5

Up to 200 °C all characteristic mechanical parameters decreased; at high temperatures they

3+
increased. This anomaly agreed with an experimentally confirmed Co spin transition

[43,46]. Detailed thermal expansion measurements along with magnetic as well as

50
conductivity behavior characterizations in [ ] indicated that the temperature dependence of

BSCF properties can be divided into six phases, i.e. five transitions: (I) ~200 K, (II) 200~350

K, (III) 350~500 K, (IV) 500~800 K (V) 800~1050 K, and (VI) 1050~1273 K. The latter three

phases (IV, V, and VI) have been the focus of the earlier studies and attributed to a spin-state

3+
transition of Co , whereas the former three phases (I, II, and III), especially I and II,

appeared to be related to valence and spin-state, since being represented in the susceptibility.

2.1.4. Creep

Although the material clearly exhibits an apparent creep deformation at room temperature

51
[ ], this low temperature creep rate is rather small and will most probably not cause serious

problems for the high-temperature use of BSCF. The influence of the partial cubic-to-

hexagonal phase transformation of BSCF on the elevated temperature creep behavior was

investigated in [38,39] in air in the temperature range 750–950 °C. Arrhenius plots of the

creep rate revealed a significant hysteresis effect. During heating, the creep rate changed

slightly up to ~850 °C and increased substantially above this temperature. Upon cooling from

950 °C it decreased in a nearly linear manner to temperature (see Fig. 4). Microstructural

investigations supported that the hysteresis is linked with the growth and transformation

kinetics of a hexagonal polymorph at the boundaries of cubic BSCF grains. Creep rates under

reduced oxygen partial pressures were significantly higher due to the absence of these

secondary phases. The temperature dependent creep rates demonstrated strong influence of

the hexagonal phase on the creep. It was suggested that, with appropriate thermal treatment,

the presence of the hexagonal phase, which deteriorates the permeation, might be avoided.

NGUYỄN ANH PHƯƠNG – 16146449 9


VLHENMA HKI 18-19 BBKH 1.5

2.1.4.1. Creep rupture

Creep rupture at typical operating temperatures was analyzed in [37] using three-point
bending tests at 850 °C and 900 °C, permitting a determination of the failure stress in this
combined tensile–compressive mode. The creep rupture data were described by a modified
Monkman–Grant relationship. The creep rupture duration and hence creep survival time
decreases considerably with temperature and applied stress (Fig. 5). A temperature increase of
50 K reduces the rupture time by a factor of ∼5–10. Under this combined tensile–compressive

creep mode, an apparent failure strain of ∼40 % was estimated. However to operate the
component for 40 years the applied stress at 850 °C should be less than 0.4 MPa. For a
temperature of 800 °C the respective stress limit was estimated to be approximately 4 MPa.

C-ring creep tests in [38] revealed the formation of cavities along the grain boundaries normal
to the tensile stresses that ultimately might lead to creep rupture. Note that, whereas under
compressive loading creep strains of ∼1 % per year might be sustained, a strain of ∼0.2 % can
already lead to damage under tensile loading. A tensile creep rupture test has been developed
52
in [ ], based on the application of tensile stresses at elevated temperatures, where only a
limited central length of the tubes was heated and cold ends clamped. Figure 3.c presents
creep rupture times for two tests on BSCF tubes. Based on the BSCF bending test failure,
significantly longer rupture times would be expected, i.e. failure at ~ 375 h should occur in
four-point bending for a bending stress of almost 25 MPa. The time to rupture will be affected
by the stress gradient in the case of four-point bending, hence leading in this case to a less
critical situation or in other words, in the case of a tensile test to a more critical situation and
rupture occurs for similar times under almost 50 % reduced stresses.

2.1.5. Zr doped BSCF

NGUYỄN ANH PHƯƠNG – 16146449 10


VLHENMA HKI 18-19 BBKH 1.5

BSCF suffers from partial decomposition of cubic to hexagonal polymorphs at ∼825 °C,
associated with a significant oxygen permeation drop [40,41]. Partial B-site substitution with 3
mol% zirconium has been suggested as one solution for transformation suppression [42].

Thermo-mechanical characteristics of Ba0.5Sr0.5(Co0.8Fe0.2)0.97Zr0.03O3−δ (BSCFZ) were

53
assessed and compared to BSCF data in [ ]. Elastic moduli and fracture strengths of BSCFZ

showed similar temperature dependencies with lower absolute values. The creep deformation

was found to be similar for both materials. Slight differences in mechanical properties of

BSCF and BSCFZ were most likely a result of different production procedures also being

reflected in the microstructures and not necessarily induced by the Zr doping.

54
In another study the dopant level was varied [ 54], i.e. BSCF-Z100·x (x = 0, 0.01, 0.03, 0.05

and 0.1). Diffusional creep was confirmed as the predominant creep mechanism for T ≥ 850

°C. This was supported by the high activation energies and low stress exponents observed for

steady-state creep, and by the fact that the grain-size-normalized creep rates of the different

compositions coincided within the limits of experimental uncertainty (see Fig. 6).

Creep characteristics of dense BSCFZ and its dependence on the secondary phase have been

55
extensively analyzed in [ ] indicating that creep behavior of both BSCF and BSCFZ at

temperatures below 850 °C depends on the thermal history of the tested specimen (see also

Fig. 4). The analyses indicated possible existence of secondary phase even at 950 °C and

proved a nonlinear creep rate – temperature dependence in both heating and cooling

sequences, which are associated with a change of the secondary phase amount.

The creep deformation of porous BSCFZ was found to be larger than for the dense material,

where the difference between the results for porous and dense material increased with

NGUYỄN ANH PHƯƠNG – 16146449 11


VLHENMA HKI 18-19 BBKH 1.5

increasing temperature (see Fig. 5 [55]), indicating that an operation of dense membrane layer

on a porous substrate might be difficult above the dissolution limit of the hexagonal phase.

2.1.6. Post-operational inspection


56
On the contrary to the annealing studies [ ], that revealed the formation of hexagonal phase

only in air, the studies of operated tubes confirmed hexagonal phase also on the vacuum side,

which appears to be related to oxygen transport. Fractographic investigations suggest that the

thermal exposure of BSCF under application relevant conditions can lead to crack initiation

(see Fig. 8) and eventually failure (i.e. proofing the importance of mechanical

characterization), which could be avoided by a modification of the operation procedure.

2.2. LSCF

La0.6Sr0.4Co0.2Fe0.8O3−δ and La0.58Sr0.4Co0.2Fe0.8O3–δ (due to its similar properties abbreviated

LSCF) perovskite-type oxide are well known for their high elevated temperature electronic and

57
good oxygen ionic conductivity [ ], which qualifies the material for application as cathode in

58 59
solid oxide fuel cells [ ] and oxygen transport membrane (OTM) [ ]. Especially in the case of

SOFCs the material lead to large success in the reduction of degradation effects and long term

60
operation [ ]. Although yielding lower permeation than BSCF [11], for membrane application it

is expected to be more thermo-chemically stable with respect to CO 2 exposure and does not show

formation of secondary phases at elevated temperatures [59].

A number of reports exist on its mechanical properties, in particular fracture and elevated-

61 62 63 64 65 66
temperature creep [29,45,46,50, , , , , , ]. Again mechanical integrity will be challenged by

transient and stationary temperature exposure and long-term failure mechanisms such as static

room temperature fatigue and creep rupture. Due to its rhombohedral room temperature

NGUYỄN ANH PHƯƠNG – 16146449 12


VLHENMA HKI 18-19 BBKH 1.5

67
structure, ferroelastic deformation and domain switching occurs at low temperatures [ ] and a

phase change to a cubic structure at operation relevant temperatures [61,62,50,46].

2.2.1. Elastic and fracture behavior

Polycrystalline LSCF revealed in bending tests a non-linear deformation behavior at room

temperature and a linear one at elevated temperatures (≥ 800 °C) [61,62]. The non-linearity

originates from ferroelastic domain switching in grains with rhombohedral symmetry (see Fig.

9). Such domains cannot form in the cubic structure that exists at elevated temperatures.

The stiffness as measured via bending tests showed an increase of about 50 % above 600 °C
and 700 °C in vacuum and air [61,62], respectively. Elastic moduli determined from
resonance frequencies exhibited an anomaly between 473 and 1113 K [46], which was
68
attributed to the transition from rhombohedral to cubic symmetry, see also [ ]. The
mechanical deformation of lanthanum strontium cobalt ferrite under uniaxial compression was
investigated at various temperatures in [64]. The material revealed a rather complex
mechanical behavior related to its ferroelasticity and stress–strain curves obtained in the 1st
and 2nd loading cycles were completely different. Susceptibility measurements indicated that
69
LSCF is in a ferromagnetic state below ∼220 K and paramagnetic above ∼250 K [ ]. The
fracture stress decreases about 10 % from RT to 200 °C, and increases again from 200 °C to
700 °C to values slightly higher than at RT [61,62]. Temperature dependency of elastic
modulus, fracture strength and fracture toughness were in good agreement [46]. With respect
to subcritical crack growth, SPT predictions indicated that dense LSCF might tolerate a static
stress of 35 MPa for 40 years, with a failure probability of 1 %, being similar to BSCF [29].

Whereas it was explicitly shown in [65] that data and temperature dependence of LSCF60

agrees with reported results for LSCF58, LSCF38 behaves differently, especially in terms of

NGUYỄN ANH PHƯƠNG – 16146449 13


VLHENMA HKI 18-19 BBKH 1.5

the temperature dependence of the elastic modulus (see Table 1). Lower strength and Weibull

moduli determined for LSCF38 suggest that improvement of production and processing are

necessary. Weibull moduli appeared to decrease at elevated temperatures for dense materials.

2.2.2. Porous LSCF

The porosity effect on elastic modulus and strength appears to be similar for LSCF as reported

for BSCF [65], very low characteristic strengths (see Table 1) led to low tolerable stresses for

long term exposure to static loads for porous LSCF [37], where the overall elevated

temperature behavior agreed with that observed for the respective dense variant [65]. A rather

low Weibull modulus for the porous material was associated with the used tape casting [65].

2.2.3. Ferroelastic domains and ferromagnetic–paramagnetic transition

Reports indicated LSCF comprises at RT of rhombohedral and cubic phase, with a cooling

rate and atmosphere dependent ratio [62]. Purely rhombohedral material was obtained with

low cooling rates; purely cubic by fast cooling in air or slow cooling under vacuum. SEM and

TEM verified a domain structure in rhombohedral grains [62,67]. Direct SEM observations

verified the influence of the stress on domain switching (see Fig. 9). In-situ TEM observations

on a heating stage verified the limited thermal stability of domains. Mechanical stresses cause

domain switching being reflected macroscopically in non-linear deformation behavior [64].

Associated with the ferromagnetic–paramagnetic transition around 250 K, the thermal

expansion coefficient possesses its minimum and increases by almost a factor of two from 250

K to 350 K (below to above RT). Thermal expansion, susceptibility and conductivity display

monotonic changes with increasing temperature from 350 K to 800 K, whilst the conductivity

started to decrease and α demonstrates a drastic increase around 800 K due to the oxygen loss

accompanied by the reduction of the transition metals [19].

NGUYỄN ANH PHƯƠNG – 16146449 14


VLHENMA HKI 18-19 BBKH 1.5

2.2.4. Creep

A distinctive ferroelastic creep that due to its low magnitude should not endanger the

application of the material was observed at 293 K in [64], whilst typical ferroelastic stress–

strain curves were obtained in the temperature range from 473 K to 873 K.

The creep behavior of dense La0.58Sr0.4Co0.2Fe0.8O3−δ (LSCF) perovskite was studied in the
temperature range 750–950 °C in air and vacuum (PO2 ≈ 4 mbar) in [63]. A transition in the
apparent activation energy was found between 800 and 850 °C for both oxygen partial
−1
pressures. The apparent activation energy is ∼250 kJ mol for the temperature range 700–
−1
800 °C under vacuum (PO2 ≈ 4 mbar) and ∼480 kJ mol for 850–950 °C in both
atmospheres. Above 850 °C, the creep rate of LSCF is higher in vacuum than in air although
the same cubic structure exists. The stress exponent of the creep law is in the range 1.9–2.5
for all temperatures, which excludes a transition of creep mechanism. It is suggested that,
below 800 °C, cation vacancies originate from the necessary balance with the substituted
cations in LSCF, and the determined activation energy reflects the energy barrier for cation
migration via these vacancies. Above 850 °C, additional vacancies appear to be formed
intrinsically, and the activation energy represents the sum of the thermally activated formation
energy of cation vacancies and migration energy of cations. A single creep rupture test on
dense LSCF in [37] indicated that this material is less prone to creep rupture than BSCF.

2.2.4.1. Creep of porous LSCF

Porous LSCF creep was studied in [66] in the temperature range 750 to 950 °C. The creep

exponent was between 3.0 and 3.9 in the stress range 15–30 MPa, which is higher than that

reported for dense LSCF (1.9–2.5 [63]). The activation energy from 800 to 950 °C was ~530

kJ/mol (dense LSCF ~480 kJ/mol [63]). The ratio of creep of porous to dense material

NGUYỄN ANH PHƯƠNG – 16146449 15


VLHENMA HKI 18-19 BBKH 1.5

increased progressively with increasing temperature (see Fig. 7) that might be related to

surface diffusional effects. Additional tests at room temperature revealed non-linear stress-

strain curves and a ferroelastic creep similar as for dense LSCF due to domain switching.

Since the creep deformation appeared to exceed an acceptable level [15], a strategy for

reduction of creep deformation could be a decrease of operation temperature or pressure ratio.

Both will reduce the oxygen flux through the membrane. In order to compensate the creep, a

smart design of the support might be an option as reported for a new generation supports with

70
straight pore channels perpendicular to load direction, freeze drying manufactured [ ].

2.2.5. Post-operational inspection

Interesting with respect to mechanical behavior is also a study on strontium surface

71
segregation for LSCF annealed with and without uniaxial compressive loads [ ]. In both

cases, surface segregation of Sr-rich phase was clearly observed on grain boundaries. The size

of segregated particles increased with increasing annealing time. The number of particle was

fewer, whereas the particle size was larger for a compressed sample, which was attributed to

grain-boundary sliding as well as Sr diffusion aiding high-temperature creep deformation.

The effect of the partial pressure difference on the mechanical behavior of LSCF is reported

72
in [ ] for samples annealed under an oxygen gradient. The mechanical properties of cross-

sections were characterized using indentation testing. Chemical strains for BSCF and LSCF

were too small to detect them after cooling to RT by XRD; however, the results suggest that

the indentation crack length is affected by chemical strains for LSCF, but not for BSCF. An

anisotropy of the indentation crack length and corresponding apparent fracture toughness was

related to interaction of domain switching and residual strain that was also associated with the

−5
observation that vacuum (10 mbar) annealed LSCF showed surface cracking on heating in

NGUYỄN ANH PHƯƠNG – 16146449 16


VLHENMA HKI 18-19 BBKH 1.5

air (see Fig. 10), whereas for BSCF such fracture features were not observed. In this light,

components made form LSCF may fail during start-up or shut due to internal stresses.

2.3. STF

Recently, the cubic perovskite SrTi 1-xFexO3-δ (STF) has been identified as a promising stable

73
membrane material [ ], compositions, especially with high iron content, exhibit high ambipolar

conductivities and oxygen surface exchange kinetics comparable to LSCF. Compressive creep

74
tests on STF (x = 0.3, 0.5 and 0.7) in [ ] yielded activation energies (341– 477 kJ mol−1) and

stress exponents in the temperature range 800–1000 °C and stress range 10–100 MPa. Results

indicated that steady-state creep rate is mainly limited by cation lattice diffusion. For STF50 and

STF70, the stress exponents are ~1.1–1.5. For STF30, however, they are in the range 2.0–2.4,

suggesting a contribution of dislocation creep. Creep rates (Fig. 11) were similar to LSCF and

LSC; but significantly lower than for BSCF and BSCFZ.

2.4. CGO

With respect to chemical stability, cerium oxide and doped cerium oxides are promising

materials for membrane applications. Ce0.8Gd0.2−yPryO2−δ membrane materials and the

75
influence of an addition 2 mol% of CoO was assessed in [ ]. CGO material showed

significantly higher elastic modulus and fracture strength compared to LSCF and BSCF as

reported in [29]. Materials preparation and compositional variations were compared with

respect to elastic modulus, hardness and fracture toughness [75]. Overall, the elastic modulus

of ceria-based ceramics was only slightly affected by the Gd- and/or Pr-doping and the

differences in elastic properties resulted mainly from the variations in material porosity.

Fracture toughness of the Gd- and Pr-doped materials, measured locally with the aid of

indentation, appeared to reduce with increasing temperature. This temperature-dependent

NGUYỄN ANH PHƯƠNG – 16146449 17


VLHENMA HKI 18-19 BBKH 1.5

behavior was not reflected, however, in fracture toughness measurements carried out using an

indentation-strength method, which are more representative for the global properties.

Critical strains of CGO estimated fracture strength and elastic modulus in [19] increases

slightly with increasing porosity (see Fig. 12), which might be a result of crack tip blunting by

pores. Although both dense and porous CGO indicated a higher sensitivity to subcritical crack

growth [29], the allowable stress for a 40 years static load at room temperature conditions was

comparable to the data obtained for porous BSCF and LSCF.

The creep behavior of doped and undoped cerium oxides was assessed in [75]. Values of the

activation energies ranged between 50 and 160 kJ/mol, being very close to the activation

energy of oxygen self-diffusion in Ce1−xGdxO2−0.5x. The stress exponents suggested diffusion-

or a combination of diffusion- and dislocation-related mechanisms. Creep properties of the

cerium oxides without Co-doping appeared to be superior to Co-doped material.

2.5. CTF

Creep of Ca0.5Sr0.5Ti0.6Fe0.15Mn0.25O3-δ and CaTi0.6Fe0.15Mn0.25O3-δ have been characterized in

[14]. Fig. 13 shows the creep rate along with literature data on other relevant membrane

materials. CTFMn25 exhibits a high creep resistance. LSCF and BSCF/BSCFZ exhibit higher

creep rates with similar activation energies [63,55]. Evaluation of the creep data on CTFMn25

revealed a stress exponent of 1.5, implying that creep was dominated by a diffusional

mechanism. The determined activation energy of 5.1 ± 0.4 eV excluded oxide ions as the rate

controlling species [16] and implied that cation diffusion is rate controlling for creep in these

materials. The creep data on CTF10 exhibited a stress exponent of 1.6 and activation energy

of 4.4 ± 1.1 eV, which implies a similar creep mechanism as for CTFMn25. Considering the

NGUYỄN ANH PHƯƠNG – 16146449 18


VLHENMA HKI 18-19 BBKH 1.5

significant porosity of the CTF10 specimen and rather similar grain sizes, the lower creep rate

of CTF10 indicated a decrease in creep resistance with Mn-doping for CTFMn25.

A three-dimensional defect study was carried out on porous CaTi 0.9Fe0.1O3−d ceramic

membrane material as a model study for the use of computed tomography in microstructural

76
analysis and mechanical characterization in [ ]. Exemplified for CaTi0.9Fe0.1O3−δ ceramics,

the study showed that three-dimensional visualization, data extraction and statistical analysis

make the computed tomography clearly advantageous in defect analysis over the commonly

used fractography. Although strength data implied that materials production and processing

required further improvement, it was verified that CT data can be used for numerical

simulation and hence for a direct assessment of the material behavior under realistic loading

configurations. The study also showed that information on microstructures and defects can

lead to misleading conclusions if the loading via FEM is not considered, since geometrical

imperfections can become more important than the defects. It was concluded that, in general,

if a material’s requirement is to possess pores and if they are the failure-relevant defects, finer

pores, compensated by their higher amount to reach the same porosity, are more desirable.

2.6. LNO
77
La2NiO4+δ (LNO) has been suggested as a rather chemical stable membrane material in [ ].

A strong change of elastic modulus and internal friction observed for dense LNO between RT

and 100 °C was attributed to an orthorhombic-tetragonal transition. No further indication of

phase transition could be observed up to 950 °C. An increase of fracture stress of porous

La2NiO4+δ from 700 to 900 °C was attributed to the primary creep. Oxygen was continuously

released from ~400 °C up to 1000 °C with a stable thermal expansion coefficient.

2.7. LWO

NGUYỄN ANH PHƯƠNG – 16146449 19


VLHENMA HKI 18-19 BBKH 1.5

La5.4WO12-δ (LWO) was studied in [78]. The room temperature Young’s modulus was ~ 130

GPa, independent of grain size, with a decrease by ~ 20 % up to 1000 °C. Ring-on-ring

bending test yielded a characteristic strength of 59 ± 3 MPa. It was shown that a factor ~ two

higher strength value could be obtained for more homogeneous microstructure. Large

agglomerates, irregular pores and pore zone areas were identified as strength limiting flaws by

fractography analysis. The strength was found to decrease by ~ 30 % up to 1000 °C. An

unexpected strength increase by ~ 4 % at 400 °C was observed. Complementary annealing

studies indicated a relationship to the onset of chemical / thermal expansion, where longer

annealing led to lower strength values. The strength was independent of the loading rate;

hence, LWO does not seem to be prone to subcritical crack growth effects. Creep appeared to

be governed by cation diffusion as the predominant mechanism. Creep parameters were n =

1.2 ± 0.3, p = 1.3 ± 0.4 and Qa = 370 ± 120 kJ / mol in the temperature and stress range 1050

– 1450 °C and 20 – 63 MPa, respectively. Creep experiments did not reveal any significant

change in the creep rates under different atmospheres.

2.7.1. Porous LWO

The mechanical properties were determined for La 5.4WO12-δ for an application relevant porosity

range 30 – 40 % [78]. It was shown that elastic modulus, hardness and strength decrease with

increasing porosity. Strength values decreased almost linear with increasing porosity. The

Weibull modulus was in the range 8-16 indicating a homogeneous flaw distribution. For

specimens with low fracture stresses rather large pores were identified as failure origin. The

secondary phase La6W2O15 in the surface resulted in lower strength values.

2.8. NWO

Creep results of single phase cubic Nd 5.5WO11.25−δ (NWO) with ~ 8% porosity and fine grains

(grain size 1 ± 0.5 μm, elastic modulus 125 ± 5 GPa) are reported in [78]. Deformation rates

NGUYỄN ANH PHƯƠNG – 16146449 20


VLHENMA HKI 18-19 BBKH 1.5

were obtained at T ≥ 1050 °C. An average stress exponent of n = 1.3 ± 0.2 and an activation

energy = 540 ± 30 kJ/mol suggested a diffusional creep mechanism. The stress exponent

increases above 1150 °C. Overall, NWO55 shows higher creep rates and higher activation

energy than LWO, former being associated with a smaller grain size.

2.9. Alternative substrate materials

Although it is often preferred to use identical materials for membrane and porous substrate to

reduce thermal expansion mismatch, due to materials cost alternative materials might be used.

2.9.1. MgO

Advanced design concepts for the application of oxygen transport ceramic membranes are

based on thin layers supported by porous substrates, where porous MgO has been suggest.

The effect of sintering conditions on porosity and mechanical properties, i.e. elastic modulus

and fracture stress as function of temperature are summarized in [19] for planar and tubular

geometry. O-ring and ring-on-ring bending test yielded similar characteristic fracture

strengths (~ 40 ± 7 MPa), whereas the elastic modulus decreases slightly with increasing

testing temperature. Data were in agreement with literature values for porous ceramics. The

addition of boehmite appeared to enhance the materials strength. An estimate of the fracture

toughness was also given. A linear increase of the elastic modulus with increasing sintering

temperature was observed, associated with a porosity reduction.

2.9.2. Al2O3

Aiming at the use in pre-combustion systems, alumina substrate with different pore sizes were

characterized with respect to their mechanical properties in [20]. Although porosity of the

micro- and nano-porous materials (see Fig. 14) were equal, bending test assessed elastic

moduli and fracture stress values revealed significant differences, indicating that based on

NGUYỄN ANH PHƯƠNG – 16146449 21


VLHENMA HKI 18-19 BBKH 1.5

mechanical considerations a nano-porous material is far superior. Microstructural

investigations supported that the higher strength of the nano-porous material is a result of a

larger contact area between grains. The nano-porous material appears also to be superior to

other membrane substrate suggested in literature (see Table 2), indicating the potential of

nano-porous materials for the use as porous substrates. However, initial permeation tests

indicated that the nano-porous substrates lead to an approximately one decade lower flow rate

that the micro-porous substrate material [20]. This will lead to a higher flow resistance and a

higher risk for concentration polarization. Hence for the application as membrane substrate an

optimization that considers both mechanical aspects and flow rate is required.

2.9.3. Metallic substrates

In an application as porous substrates for oxygen transport membranes, the alloy will be

exposed to temperatures of 700 to 900 °C for long durations, raising high demands in

particular with respect to corrosion resistance and creep. The necessary porosity will leads to

a reduction in elastic modulus and enhanced creep. A higher elastic modulus of the substrate

compared to the ceramic layers can be favorable as crack propagation in the ceramic layer

79 80
becomes less likely (see also [ , ]). The typically rather large creep deformation of steels at

elevated temperature might be problematic for the structural stability of the component,

although creep can relax stresses that might result from differences in thermal expansion.

From creep rates measured for dense FeCrAl alloy, the highest applied stress to stay below

this limit suggested in [79] of 1% strain per year is 11 MPa at 700 °C, 5 MPa at 800 °C, and 3

MPa at 900 °C. A membrane during operation can be exposed to an applied stress of 30 MPa

81
[ ] leading to very challenging conditions for long-term operation of a metal-supported

component at high temperature.

NGUYỄN ANH PHƯƠNG – 16146449 22


VLHENMA HKI 18-19 BBKH 1.5

ITM has verified its excellent performance as interconnector material for SOFCs in the so far

longest ever operated SOFC stack (> 70,000 hours) that is still under operation [60]. Hence it

is obviously also a promising materials as porous mechanical support for membrane materials

82
[ ]. The temperature dependence of mechanical properties, i.e. Young´s modulus, ultimate

tensile strength (UTS), Poisson´s ratio, coefficient of thermal expansion and creep behavior

have been compared for dense and porous ITM in [82]. Elastic modulus increases by a factor

of 4 for the porous material due to the smaller effective area. The UTS is reduced by a factor

of ~ 8 for the porous material. The Poisson’s ratio is approximately reduced by a factor of 1.5.

The coefficient of thermal expansion is independent of porosity. The creep mechanism

appears to depend on the temperature and on the applied stress.

2.10. Sealant materials

Tubes or plates have to be joined to larger components to enhance the flux. Depending on the

design joining technologies might be used. Options are also glass-ceramic and metallic

83 84 85 86
sealants that find already frequent use in solid oxide fuel cells [ , , , ]. Although a number

of studies concentrated on the properties of such materials for the SOFC application

[84,85,86], revealing in particular limitations due to atmospheric conditions for metallic

87
sealants [ ] and creep and reduced elevated temperature strength values for glass-ceramic

88 89 90
sealants [ , , ], studies on sealants for membrane application focused more on the materials

91 92 93 94 95
interactions [ , ] and reports on the mechanical properties are limited [ , , ]. In particular

in [95] it was stated that for BSCF, reactive brazing in air atmosphere (RAB) is the only

technical applicable solution for a heat resistant and gas tight sealing to metals, however,

brazing BSCF to heat resistant austenitic steel AISI 314 led to a change in BSCF

microstructure at the interface due to reactions with copper oxide (CuO) from the used AgCu

braze alloy. After aging for 500 h at 850 °C in air atmosphere, a further change in the

microstructure appeared, which leading to an additional interfacial porosity increase and

NGUYỄN ANH PHƯƠNG – 16146449 23


VLHENMA HKI 18-19 BBKH 1.5

needle shaped phase at ceramic grain boundaries that might be characteristic of the hexagonal

phase formation discussed in 2.1.1. Mechanical tests with the BSCF/AISI 314-joints showed a

decrease of the room temperature strength after aging by a factor of around two.

3. Conclusions

Mechanical properties were summarized for selected oxygen conductors un-doped and Zr doped

Ba0.5Sr0.5Co0.8Fe0.2O3−δ, La0.6/0.58Sr0.4Co0.2Fe0.8O3−δ, SrTi1-xFexO3-δ, Ce0.8Gd0.2−yPryO2−δ,

Ca0.5Sr0.5Ti0.6Fe0.15Mn0.25O3-δ, CaTi0.6Fe0.15Mn0.25O3-δ, La2NiO4+δ, and for the proton

conductors La5.4WO12-δ and Nd5.5WO11.25−δ. Furthermore, properties of some selected porous

ceramic and metallic substrate materials were given. The importance of failure probability

associated with the fracture stress Weibull distribution; creep rupture of dense and enhanced creep

deformation of porous materials have been emphasized. In particular, a study on micro-porous and

nano-porous alumina substrates indicated that the strength can be enhanced by the use of a nano-

porous material, although being on the expense of a reduction of the flow rate. Particular

advantage of the STF compositions is their elevated temperature chemical and single phase

stability under different atmospheres. Fractographic investigations suggest that the thermal

exposure of BSCF under application relevant conditions can lead to crack initiation and eventually

failure, which could be avoided by a modification of the operation procedure. On the other side a

study of the fracture toughness of LSCF after exposure to an oxygen gradient indicated, similar as

fracture features during air exposure after a vacuum annealing study, that gradient conditions

exposure can also be dangerous for this perovskite.

Advanced oxygen separation units are based on thin membrane layers that are mechanically

supported by a porous substrate. One of the most important aspects for the long-term stability

at elevated temperatures is the creep behavior, which was assessed in the current study for

various potential membrane and substrate materials. Fluorite-structured ceria membranes and

NGUYỄN ANH PHƯƠNG – 16146449 24


VLHENMA HKI 18-19 BBKH 1.5

magnesia substrates showed the lowest creep rates of the investigated materials and satisfy the

“1% creep strain per year” criterion up to 900 °C. For example LSCF38, MgO and all

investigated cerium oxides (CeO2−δ and its three Gd-/Pr-doped variations:

Ce0.8Gd0.2−xPrxO2−0.5(0.2−x)−δ, with x = 0, 0.1 and 0.2) satisfied at around 800 °C the creep

limit criterion, suggested to be necessary to warrant reliable long-term stability for membrane

systems, but only ceria materials and MgO fulfilled this requirement also at higher

temperatures. As might be expected, the metalls revealed stronger creep than the ceramics,

with creep rates significantly higher than 1 % per year. The creep deformation of porous

LSCF and BSCFZ was found to be larger than for the dense material, where the difference

between the results for porous and dense material increased with increasing temperature,

indicating that an operation of dense membrane layer on a porous substrate might be difficult.

The comparison with respect to a tolerable creep rate aims to avoid failure of the oxygen

generation system that might occur due to deformation of its functional elements. Such high

creep rates can result in large deformations and structural instability and in the case of tensile

strains even in creep rupture. Since the creep deformation appeared to exceed an acceptable

level for some materials, a strategy for reduction of creep deformation could be a decrease of

operation temperature or pressure ratio. Both will reduce the oxygen flux through the

membrane. In order to compensate the creep, a smart design of the support might be an option

as reported for a new generation supports with straight pore channels perpendicular to load

direction, freeze drying manufactured. However, the creep can also aid the relaxation of

stresses that might result from differences in thermal expansion and a detailed consideration

of the effects in an actual design requires detailed finite element simulations.

Acknowledgement

The author would like to thank the coworkers, financially supported by national and European

projects that supported the study of membrane materials and substrates in the last decade.

NGUYỄN ANH PHƯƠNG – 16146449 25


VLHENMA HKI 18-19 BBKH 1.5

Captions

Fig. 1: Normalized fracture toughness, strength and elastic modulus of BSCF after [45].

Fig. 2: Strength – probability – time plot of BSCF, LSCF and CGO after [52,37].

Fig. 3: Elastic modulus (a) and average fracture stress (b) of dense and porous BSCF after

[65].

Fig. 4: Creep rated of BSCF and BSCFZ after [55].

Fig. 5: a) Creep rupture times for BSCF in three-point bending along with a typical image of

creep rupture cracks in b) [38,37]. In c) tensile creep test of dense BSCF tubes at 850 °C (D =

10 mm, d = 8 mm), applied stresses 12.5 MPa (blue curve) and 16 MPa (red curve),

respectively.

Fig. 6: Normalized creep rates for BSCF-Z100·x (x = 0, 0.01, 0.03, 0.05 and 0.1) [54].

Fig. 7: Ratio of creep rate of porous to dense BSCF and LSCF after [66].

Fig. 8: Crack extending from elongated pores after the operation in a demonstrator unit [56].

Fig. 9: Change of ferroelastic domain direction due to the application of an in-plane load [67].
−5
Fig. 10: Vacuum (10 mbar) annealed LSCF showed surface cracking on heating in air [72].

Fig. 11: Comparison of creep rates of STF with data for BSCF, BSCFZ, LSCF and LSC [74].

Fig. 12: Critical strain of CGO (diamond) and SYT (round symbol) as function of porosity

[65].

Figure 13: Creep rate of dense CTFMn25 and porous CTF10 under a compressive load of 30

MPa along with literature data for dense LSCF, BSCF and GGO [14].

Fig. 14: a) Micro- and b) nano-porous Al2O3 [20].

Fig. 15. Comparison of the creep rates of different membrane and substrate materials [15].

NGUYỄN ANH PHƯƠNG – 16146449 26


VLHENMA HKI 18-19 BBKH 1.5

TABLES

Table 1. Characteristic strengths and Weibull moduli of LSCF material [65].

Material LSCF58-dense LSCF60_6% LSCF60_46% LSCF38_6%

σ0(RT) [MPa] 73 ± 2 75 ± 3 18 ± 2 77 ± 3

m (RT) 10.6 ± 0.4 (15) 11.4 ± 1.2 (7) 3.4 ± 0.4 (5) 6.2 ± 0.4 (14)

σ
0(800 °C) - 170 ± 30 21 ± 3 125 ± 33
[MPa]

m (800 °C) - 3 ± 1 (3) 4 ± 1 (4) 1.7 ± 0.3 (5)

Table 2. Comparison of mechanical properties of different substrate materials [20].

Consignment Al2O3 Al2O3 MgO [19] CGO [29] BSCF [29] LSCF [29]

Atech IEK-1

Porosity 42 43.5 36 43 34 46

E [GPa ] 48 ± 10 102 ± 13 55 ± 3 57 ± 14 35 ± 3 12 ± 4

σ0 [MPa ] 42 ± 1 157 ± 11 41 ± 2 55 ± 3 31 ± 1 18 ± 2

Fracture strain 0.088 0.15 0.075 0.096 0.089 0.15

m 7±1 3±1 4±2 5±2 4±2

NGUYỄN ANH PHƯƠNG – 16146449 27


VLHENMA HKI 18-19 BBKH 1.5

FIGURES

Fig. 1.

Fig. 2.

NGUYỄN ANH PHƯƠNG – 16146449 28


VLHENMA HKI 18-19 BBKH 1.5

Fig. 3.

Fig. 4.

Fig. 5.

NGUYỄN ANH PHƯƠNG – 16146449 29


VLHENMA HKI 18-19 BBKH 1.5

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

NGUYỄN ANH PHƯƠNG – 16146449 30


VLHENMA HKI 18-19 BBKH 1.5

150 µm

Fig. 10.

Fig. 11.

Fig. 12.

NGUYỄN ANH PHƯƠNG – 16146449 31


VLHENMA HKI 18-19 BBKH 1.5

Fig. 13.

Fig. 14.

Fig. 15.

NGUYỄN ANH PHƯƠNG – 16146449 32


VLHENMA HKI 18-19 BBKH 1.5

References

1
J.F. Vente, W.G. Haije, Z.S. Rak, Performance of functional perovskite membranes for
oxygen production, J. Membr. Sci., 276 (2006), pp. 178–184
2
I. Pfaff, A. Kather, Comparative thermodynamic analysis and integration issues of CCS
steam power plants based on oxy-combustion with cryogenic or membrane based air
separation, Energy Proc., 1 (2009), pp. 495–502

3 H.J.M. Bouwmeester, Dense ceramic membranes for methane conversion, Catal.


Today, 82 (2003), pp. 141–150

4 L. Olivier, S. Haag, C. Mirodatos, A.C. van Veen, Oxidative coupling of methane using
catalyst modified dense perovskite membrane reactors, Catal. Today, 142 (2009), pp. 34–41

5 F.T. Akin, Y.S. Lin, Controlled oxidative coupling of methane by ionic conducting
ceramic membrane, Catal. Lett., 78 (2002), pp. 239–242

6 Z. Shao, S.M. Haile, A high-performance cathode for the next generation of solid-
oxide fuel cells, Nature, 431 (2004), pp. 170–173

7 D. van Holt, E. Forster, M.E. Ivanova, W.A. Meulenberg, M. Müller, S. Baumann, R.


Vaßen, Ceramic materials for H2 transport membranes applicable for gas separation under
coal-gasification-related conditions, J. Eur. Ceram. Soc., 34 (2014), pp. 2381–2389
8 H. Wang, Y. Cong, W. Yang, Oxygen permeation study in a tubular

Ba0.5Sr0.5Co0.8Fe0.2O3−δ oxygen permeable membrane, J. Membr. Sci., 210 (2002), pp. 259–271

9 J. Sunarso, S. Baumann, J.M. Serra, W.A. Meulenberg, S. Liu, Y.S. Lin, et al. Mixed
ionic– electronic conducting (MIEC) ceramic-based membranes for oxygen separation, J.
Membr. Sci., 320 (2008), pp. 13–41

10 R. Kneer, D. Toporov, M. Forster, M. Förster, D. Christ, C. Broeckmann, E. Pfaff, M.


Zwick, S. Engels, M. Modigell, OXYCOAL-AC: towards an integrated coal-fired power plant
process with ion transport membrane-based oxygen supply, Energy Environ. Sci., 3 (2010),
pp. 198–207

11 J.M. Serra, J. Garcia-Fayos, S. Baumann, F. Schulze-Küppers, W.A. Meulenberg,


Oxygen permeation through tape-cast asymmetric all – La 0.6Sr0.4Co0.2Fe0.8O3−δ membranes,
J. Membr. Sci., 447 (2013), pp. 297–305

12 A.J. Ellett, Oxygen permeation and thermo-chemical stability of oxygen separation


membrane materials for the oxyfuel process, Ph.D. Thesis, RWTH Aachen, Germany, (2009)
13 D.P. Fagg, A.L. Shaula, V.V. Kharton, J.R. Frade, High oxygen permeability in
fluorite-type Ce0.8Pr0.2O2−δ via the use of sintering aids, J. Membr. Sci., 299 (2007), pp. 1–7

NGUYỄN ANH PHƯƠNG – 16146449 33


VLHENMA HKI 18-19 BBKH 1.5

14J.M. Polfus, W. Xing, G. Pećanac, A. Fossdal, S.M. Hanetho, Y. Larring, J. Malzbender,


M.-L. Fontaine, R. Bredesen, Oxygen permeation and creep behavior of Ca 1-
xSrxTi0.6Fe0.15Mn0.25O3-δ (x=0,0.5) membrane materials, J.Membr.Sci. 499 (2016), pp.172–178

15 M. Lipińska-Chwałek, G. Pećanac, J. Malzbender, Creep behaviour of membrane and


substrate materials for oxygen separation units, J. Eur. Ceram. Soc., 33 (2013), pp. 1841–1848
16 S. Escolástico, C. Solís, T. Scherb, G. Schumacher, J.M. Serra, Hydrogen separation in
La5.5WO11.25−δ membranes, J. Membr. Sci., 444 (2013), pp. 276–284
17 M. Ruf, C. Solís, S. Escolástico, R. Dittmeyer, J.M. Serra, Transport properties
and
oxidation and hydration kinetics of the proton conductor Mo doped Nd 5.5WO11.25−δ, J. Mater.
Chem. A., 2 (2014) pp. 18539–18546
18
S. Baumann, W.A. Meulenberg, H.P. Buchkremer, Manufacturing strategies for
asymmetric ceramic membranes for efficient separation of oxygen from air, J. Eur. Ceram.
Soc., 33 (2013), pp. 1251–1261

19 M. Lipińska-Chwałek, L. Kiesel, J. Malzbender, Mechanical properties of porous MgO


substrates for membrane applications, J. Eur. Ceram. Soc., 34 (2014), pp. 2519–2524

20 J. Zhang, J. Malzbender, Mechanical characterization of micro- and nano-porous


alumina, Ceram. Int., 41 (2015) pp. 10725–10729
21 Z. Chen, X. Wang, V. Bhakhri, F. Giuliani, A. Atkinson, Nanoindentation of porous
bulk and thin films of La0.6Sr0.4Co0.2Fe0.8O3−δ, Acta Mater., 61 (2013), pp. 5720–5734

22 J. Malzbender, G. de With, The use of the loading curve to assess soft coatings, Surf.
Coat. Technol., 127 (2000), pp. 265–272
23 J. Malzbender, G. de With, The use of the indentation loading curve to assess soft
coatings, Surf. Coat. Technol., 135 (2001), pp. 202–207
24 A. Chanda, B.X. Huang, J. Malzbender, R.W. Steinbrech, Micro- and macro-indentation
behaviour of Ba0.5Sr0.5Co0.8Fe0.2O3−δ perovskite, J. Eur. Ceram. Soc., 31 (2011), pp. 401–408

25 J. Malzbender, G. de With, J.M.J. den Toonder, Determination of the elastic modulus


and hardness of sol–gel coatings on glass: influence of indenter geometry, Thin Solid Films,
372 (2000), 134–143

26 G.R. Anstis, P. Chantikul, B.R. Lawn, D.B. Marshall, A critical-evaluation of


indentation techniques for measuring fracture-;toughness. 1. Direct crack measurements, J.
Am. Ceram. Soc., 64 (1981), pp. 533–38

27 J. Malzbender, G. de With, Cracking and residual stress in hybrid coatings on float


glass, Thin Solid Films, 359 (2000), pp. 210–214

28 P.V. Hendriksen, J.R. Høgsberg, A.M. Kjeldsen, B.F. Sørensen, H.G. Pedersen,
Failure modes of thin supported membranes, Ceram. Eng. Sci. Proc., 27 (2008), pp. 347–360

NGUYỄN ANH PHƯƠNG – 16146449 34


VLHENMA HKI 18-19 BBKH 1.5

29
G. Pećanac, S. Foghmoes, M. Lipińska-Chwałek, S. Baumann, T. Beck, J. Malzbender,
Strength degradation and failure limits of dense and porous ceramic membrane materials, J.
Eur. Ceram. Soc., 33 (2013), pp. 2689–2698

30 J. Wei, G. Pećanac, J. Malzbender, Review of mechanical characterization methods for


ceramics used in energy technologies, Ceram. Int., 40 (2014), pp. 15371–15380
31 J. Malzbender, Mechanical and thermal stresses in multilayered materials, J. Appl.
Phys.,
95 (2004), pp. 1780–1782

32A. Atkinson, A. Selçuk, Mechanical behaviour of ceramic oxygen ion-conducting


membranes, Solid State Ionics, 134 (2000), pp. 59–66
33 M. Lipińska-Chwałek, F. Schulze-Küppers, J. Malzbender, Strength and elastic
modulus of lanthanum strontium cobalt ferrite membrane materials, Ceram. Int., 41 (2015),
pp. 1355– 1360

34 J. Malzbender, R.W. Steinbrech, Fracture test of thin sheet electrolytes for solid oxide
fuel cells, J. Eur. Ceram. Soc., 27 (2007), pp. 2597–2603

35 K. Kwok, L. Kiesel, H.L. Frandsen, M. Søgaard, P.V. Hendriksen, Strength


characterization of tubular ceramic materials by flexure of semi-cylindrical specimens, J. Eur.
Ceram. Soc., 34 (2014), pp. 1423–1432

36 K. Kwok, H.L. Frandsen, M. Søgaard, P.V. Hendriksen, Stress analysis and fail-safe
design of bilayered tubular supported ceramic membranes, J. Membr. Sci., 453 (2014), pp.
253–262
37 G. Pećanac, S. Baumann, J. Malzbender, Mechanical properties and lifetime
predictions for
Ba0.5Sr0.5Co0.8Fe0.2O3−δ membrane material, J. Membr. Sci., 385–386 (2011), pp. 263–268
38 B. Rutkowski, J. Malzbender, T. Beck, R.W. Steinbrech, L. Singheiser, Creep
behaviour of
tubular Ba0.5Sr0.5Co0.8Fe0.2O3−δ gas separation membranes, J. Eur. Ceram. Soc., 31 (2011),
pp. 493–499
39 B. Rutkowski, J. Malzbender, R.W. Steinbrech, T. Beck, H.J.M. Bouwmeester,
Influence of thermal history on the cubic-to-hexagonal phase transformation and creep
behaviour of
Ba0.5Sr0.5Co0.8Fe0.2O3−δ ceramics, J. Membr. Sci., 381 (2011), pp. 221–225
40
S. Baumann, J.M. Serra, M.P. Lobera, S. Escolástico, F. Schulze-Küppers, W.A.
Meulenberg, Ultrahigh oxygen permeation flux through supported Ba0.5Sr0.5Co0.8Fe0.2O3−δ
membranes, J. Membr. Sci., 377 (2011), pp. 198–205
41
K. Efimov, Q. Xu, A. Feldhoff, Transmission electron microscopy study of
Ba0.5Sr0.5Co0.8Fe0.2O3−δ perovskite decomposition at intermediate temperatures, Chem.
Mater., 22 (2010), pp. 5866–5875
NGUYỄN ANH PHƯƠNG – 16146449 35
VLHENMA HKI 18-19 BBKH 1.5

42 S. Švarcová, K. Wiik, J. Tolchard, H.J.M. Bouwmeester, T. Grande, Structural


instability of cubic perovskite BaxSr1−xCo1−yFeyO3−d, Solid State Ionics, 178 (2008), pp.
1787–1791

NGUYỄN ANH PHƯƠNG – 16146449 36


VLHENMA HKI 18-19 BBKH 1.5

43 B.X. Huang, J. Malzbender, R.W. Steinbrech, P. Grychtol, C.M. Schneider, L.


Singheiser, Anomalies in the thermomechanical behavior of Ba0.5Sr0.5Co0.8Fe0.2O3−δ ceramic
oxygen conductive membranes at intermediate temperatures, Appl. Phys. Lett., 95 (2009),
051901
44 B.X. Huang, J. Malzbender, R.W. Steinbrech, L. Singheiser, Discussion of the complex
thermo-mechanical behavior of Ba0.5Sr0.5Co0.8Fe0.2O3−δ, J. Membr. Sci., 359 (2010), pp.80–85

45 B.X. Huang, A. Chanda, R.W. Steinbrech, J. Malzbender, Indentation strength method


to determine the fracture toughness of La0.58r0.4Co0.2Fe0.8O3−δ and Ba0.5Sr0.5Co0.8Fe0.2O3−δ,
J. Mater. Sci., 47 (2012), pp. 2695–2699
46 B. Huang, J. Malzbender, R.W. Steinbrech, Elastic anomaly and internal friction of
Ba0.5Sr0.5Co0.8Fe0.2O3−δ and La0.58Sr0.4Co0.2Fe0.8O3−δ, J. Mater.Res., 26 (2011), pp. 1388–1391
47 J. Malzbender, B. Huang, J. Moench, R.W. Steinbrech, A comparison of results
obtained
using different methods to assess the elastic properties of ceramic materials exemplified for
Ba0.5Sr0.5Co0.8Fe0.2O3−δ, J. Mater. Sci., 45 (2010), pp. 1227–1230
48 M. Lipińska-Chwałek, F. Schulze-Küppers, J. Malzbender, Stability aspects of porous

49 M. Lipińska-Chwałek, J. Malzbender, A. Chanda, S. Baumann, R.W. Steinbrech,


Mechanical characterization of porous Ba0.5Sr0.5Co0.8Fe0.2O3−δ, J. Eur. Ceram. Soc., 31
(2011), pp. 2997–3002
50 W. Araki, Y. Arai, J. Malzbender, Transitions of Ba0.5Sr0.5Co0.8Fe0.2O3−δ and
La0.58Sr0.4Co0.2Fe0.8O3−δ, Mater. Lett., 132 (2014), pp. 295–297
51 W. Araki, J. Malzbender, Mechanical behaviour of Ba0.5Sr0.5Co0.8Fe0.2O3−δ under
uniaxial compression, Scripta Mater., 69 (2013), pp. 278–281

52 G. Pećanac, J. Malzbender, Strength and creep behaviour of ceramic materials,


HETMOC project report, 2015.
53 G. Pećanac, L. Kiesel, R. Kriegel, J. Malzbender Comparison of thermo-
mechanical
characteristics of non-doped and 3 mol% B-site Zr-doped Ba 0.5Sr0.5Co0.8Fe0.2O3−δ, Ceram.
Int., 40 (2014), pp. 1843–1850
54
V. Stournari, S.F.P. ten Donkelaar, J. Malzbender, T. Beck, L. Singheiser, H.J.M.
Bouwmeester, Creep behavior of perovskite-type oxides Ba0.5Sr0.5(Co0.8Fe0.2)1–xZrxO3−δ, J.
Eur. Ceram. Soc., 35 (2015), pp. 1841–1846
55
G. Pećanac, L. Kiesel, J. Malzbender, Steady-state creep of porous and an extended
analysis on the creep of dense BSCFZ perovskite, J. Membr. Sci., 456 (2014), pp. 134–138

NGUYỄN ANH PHƯƠNG – 16146449 37


VLHENMA HKI 18-19 BBKH 1.5

56 B. Rutkowski, R. Kriegel, J. Malzbender, Ex-service analysis of membrane tubes after


the operation in a demonstrator unit, J. Membr. Sci., 462 (2014) pp. 69–74

NGUYỄN ANH PHƯƠNG – 16146449 38


VLHENMA HKI 18-19 BBKH 1.5

57 Y. Teraoka, H.M. Zhang, K. Okamoto, N. Yamazoe, Mixed ionic-electronic


conductivity of
La1−xSrxCo1−yFeyO3−δ perovskite-type oxides, Mater. Res. Bull., 23 (1988), pp. 51–58

58J. Malzbender, P. Batfalsky, R. Vaßen, V. Shemet, F. Tietz, Component interactions after


long-term operation of an SOFC stack with LSM cathode, J.PowerSources,201(2012)196–203
59 F. Schulze-Küppers, S. Baumann, F. Tietz, H.J.M. Bouwmeester, W.A.
Meulenberg,
Towards the fabrication of La0.98−xSrxCo0.2Fe0.8O3−δ perovskite-type oxygen transport
membranes, J. Eur. Ceram. Soc., 34 (2014), pp. 3741–3748
60 L Blum, P Batfalsky, Q Fang, LGJ de Haart, J Malzbender, N Margaritis, SOFC Stack
and System Development at Forschungszentrum Jülich, J. Electrochem. Soc., 162 (2015)
F1199– F1205
61 B.X. Huang, J. Malzbender, R.W. Steinbrech, L. Singheiser, Mechanical properties of
La0.58Sr0.4Co0.2Fe0.8O3−δ membranes, Solid State Ionics, 180 (2009), pp. 241–245
62
B.X. Huang, J. Malzbender, R.W. Steinbrech, E. Wessel, H.J. Penkalla, L. Singheiser,
Mechanical aspects of ferro-elastic behavior and phase composition of
La0.58Sr0.4Co0.2Fe0.8O3−δ, J. Membr. Sci., 349 (2010), pp. 183–188
63
B.X. Huang, R.W. Steinbrech, S. Baumann, J. Malzbender, Creep behavior and its
correlation with defect chemistry of La0.58Sr0.4Co0.2Fe0.8O3−δ, Acta Mater., 60 (2012), pp.
2479–2484
64 W. Araki, J. Malzbender, Ferroelastic deformation of La0.58Sr0.4Co0.2Fe0.8O3−δ under
uniaxial compressive loading, J. Eur. Ceram. Soc., 33 (2013), pp. 805–812
65 M. Lipińska-Chwałek, F. Schulze-Küppers, J. Malzbender, Strength and elastic modulus
of lanthanum strontium cobalt ferrite membrane materials, Ceram. Int., 41 (2015), pp.1355–1360

66 Y. Zou, F. Schulze-Küppers, J. Malzbender, Creep behavior of porous


La0.6Sr0.4Co0.2Fe0.8O3−δ oxygen transport membrane supports, Ceram. Int., 41 (2015), pp.
4064–4069

67 B.X. Huang, R.W. Steinbrech, J. Malzbender, Direct observation of ferroelastic


domain effects in LSCF perovskites, Solid State Ionics, 228 (2012) pp. 32–36
68 Y. Kimura, T. Kushi, S. Hashimoto, K. Amezawa, T. Kawada, Influences of
temperature
and oxygen partial pressure on mechanical properties of La0.6Sr0.4Co1-yFeyO3-δ, J. Am.
Ceram. Soc., 95 (2012), pp. 2608–2613

NGUYỄN ANH PHƯƠNG – 16146449 39


VLHENMA HKI 18-19 BBKH 1.5

69 W Araki, T Abe, Y Arai, Anomalous variation of the mechanical behaviour of


ferroelastic La0.6Sr0.4Co0.2Fe0.8O3−δ during the ferro-to-paramagnetic transition, Scripta
Mater., 99 (2015), pp. 9–12

NGUYỄN ANH PHƯƠNG – 16146449 40


VLHENMA HKI 18-19 BBKH 1.5

70
C. Gaudillere, J. Garcia-Fayos, M. Balaguer, J.M. Serra, Enhanced oxygen separation
through robust freeze-cast bilayered dual-phase membranes, Chem. Sus. Chem., 7 (2014), pp.
2554–2561
71
W. Araki, T. Yamaguchi, Y. Arai, J. Malzbender, Strontium surface segregation in
La0.58Sr0.4Co0.2Fe0.8O3−δ annealed under compression, Solid State Ionics, 268 (2014), pp. 1–
72
6 B.X. Huang, J. Malzbender, The effect of an oxygen partial pressure gradient on the
mechanical behavior of perovskite membrane materials, J. Eur. Ceram. Soc., 34 (2014), pp.
1777–1782
73 C.-Y. Yoo, H.J.M. Bouwmeester, Oxygen surface exchange kinetics of SrTi1-xFexO3-δ
mixed conducting oxides, Phys. Chem. Chem. Phys., 11 (2012), pp. 11759–11765

74 S.F.P. ten Donkelaar, V. Stournari, J. Malzbender, A. Nijmeijer, H.J.M. Bouwmeester,


High-temperature compressive creep behaviour of perovskite-type oxides SrTi 1-xFexO3-δ, J.
Eur. Ceram. Soc., 35 (2015), pp. 4203–4209

75 M. Lipińska-Chwałek, F. Schulze-Küppers, J. Malzbender, Mechanical properties of


pure and doped cerium oxide, J. Eur. Ceram. Soc., 35 (2015) pp. 1539–1547
76 G. Pećanac, J. Malzbender, F. Pauly, M.L. Fontaine, P. Niehoff, S. Baumann, T. Beck,
L.
Singheiser, Mechanical characterization of ceramics by means of a 3D defect analysis,
Ceram. Int., 41 (2015), pp. 2411–2417
77 B.X. Huang, J. Malzbender, R.W. Steinbrech, Thermo-mechanical properties of
La2NiO4+δ, J. Mater. Sci., 46 (2011), pp. 4937–4941

78 V.K. Stournari, Thermo-mechanical properties of mixed ionic-electronic conducting


membranes for energy application, Ph.D. Thesis, RWTH Aachen, Germany, 2015.
79 J.A. Glasscock, L. Mikkelsen, Å.H. Persson, G. Pećanac, J. Malzbender, P. Blennow,
F.
Bozza, P.V Hendriksen, Porous Fe21Cr7Al1Mo0.5Y metal supports for oxygen transport
membranes: Thermo-mechanical properties, sintering and corrosion behavior, Solid State
Ionics, 242 (2013), pp. 33–44

80 J. Malzbender, G. de With, Cracking and residual stress in hybrid coatings on float


glass, Thin Solid Films, 359 (2000), pp. 210–214
81 M. Schulz, R. Kriegel, W. Burckhardt, Modeling of oxygen flux and stress distribution
for
Ba0.5Sr0.5Co0.8Fe0.2O3−δ membranes at application conditions, in: Proceedings of the 10th
International Conference on Inorganic Membranes, Tokyo, Japan, 2008.
82 V. Vasechko, G. Pećanac, B. Kuhn, J. Malzbender, Mechanical properties of porous
ITM alloy, Int. J. Hydrogen Energy, 41 (2016), pp. 562–569

NGUYỄN ANH PHƯƠNG – 16146449 41


VLHENMA HKI 18-19 BBKH 1.5

83 J. Malzbender, R. W. Steinbrech, L. Singheiser, A review of advanced techniques for


characterising SOFC behavior, Fuel Cells, 9 (2009), pp. 785–793

NGUYỄN ANH PHƯƠNG – 16146449 42


VLHENMA HKI 18-19 BBKH 1.5

84 J. Malzbender , J. Mönch, R.W. Steinbrech, T. Koppitz, S.M. Gross, J. Remmel,


Symmetric shear test of glass-ceramic sealants at SOFC operation temperature, J. Mater. Sci.,
42 (2007), pp. 6297–6301

85 B. Kuhn, F.J. Wetzel, J. Malzbender, R.W. Steinbrech, L. Singheiser, Mechanical


performance of reactive-air-brazed (RAB) ceramic/metal joints for solid oxide fuel cells at
ambient temperature, J. Power Sources, 193 (2009), pp. 199–202.

86 B. Kuhn, E. Wessel, J. Malzbender, R.W. Steinbrech, L. Singheiser, Effect of


isothermal aging on the mechanical performance of brazed ceramic/metal joints for planar
SOFC-stacks, Int. J. Hydrogen Energy, 35 (2010), pp. 9158–9165.

87 T. Bause, J. Malzbender, M. Pausch, T. Beck, L. Singheiser, Damage and Failure of


Silver Based Ceramic/Metal Joints for SOFC Stacks, Fuel Cells, 13 (2013), pp. 578–583.
88 J. Malzbender, Y. Zhao, T. Beck, Fracture and creep of glass-ceramic solid oxide fuel
cell sealant materials, J. Power Sources, 246 (2014), pp. 574–580

89 Y. Zhao, J. Malzbender, Elevated temperature effects on the mechanical properties of


solid oxide fuel cell sealing materials, J. Power Sources, 239 (2013), pp. 500–504
90 J. Malzbender, Y. Zhao, Flexural Strength and Viscosity of Glass Ceramic Sealants for
Solid Oxide Fuel Cell Stacks, Fuel Cells, 12 (2012), pp. 47–53

91 K. Bobzin, N. Kopp, S. Wiesner, Influence of Filler and Base Material on the Pore
Development during Reactive Air Brazing , Adv. Eng. Mater., 16 (2014), pp. 1456–1461

92 S. Dabbarh, E. Pfaff, A. Ziombra, A. Bezold, Brazing of MIEC ceramics to high


temperature metals, Ceram. Trans., 215 (2010), pp.213–223
93 A. Kaletsch, A. Bezold, E.M. Pfaff, C. Broeckmann, Effects of Copper Oxide Content
in AgCuO Braze Alloy on Microstructure and Mechanical Properties of Reactive-Air-Brazed
Ba0.5Sr0.5Co0.8Fe0.2O3−δ (BSCF), J. Ceram. Sci. Technol., 3 (2012) pp. 95–103
94
A. Kaletsch, E.M. Pfaff, C. Broeckmann, Effect of Aging on Microstructure and
Mechanical Strength of Reactive Air Brazed BSCF/AISI 314-Joints, Adv. Eng. Mater., 16
(2014), pp. 1430–1436
95 K. Bobzin, M. Ote, S. Wiesner, A. Kaletsch, C. Broeckmann, Characterization of Reactive
Air Brazed Ceramic/Metal Joints with Unadapted Thermal Expansion Behavior, Adv. Eng.
Mater., 16 (2014), pp. 1490–1497

NGUYỄN ANH PHƯƠNG – 16146449 43

You might also like