Mechanical and thermal carbonation of strontium ferrite SrFeOx

Article abstract of DOI:10.1016/S0025-5408(02)00898-X

Reaction of non-stoichiometric SrFeO_x (2.5 ≤ x ≤ 3.0) with carbon dioxide induced by high temperature and mechanical energy was investigated. The high-temperature reaction route was found to convert SrFeO_x to a mixture of SrFe₁₂O₁₉ and SrCO₃ at temperatures below 1128 K under 0.98 atm of CO₂ pressure. The mechanical reaction carried out at room temperature causes a complete decomposition of SrFeO_x to strontium carbonate and hematite via a metastable ferric carbonate phase. We examined the phase transformations, the reaction kinetics and morphology of carbonated SrFeO_x and compared the two processes.

Full text of DOI:10.1016/S0025-5408(02)00898-X

Materials Research Bulletin 37 (2002) 2093±2105
Mechanical and thermal carbonation of
strontium ferrite SrFeOx
*
M. Schmidt
ISIS Facility, Rutherford Appleton Laboratory, Chilton, Didcot OX11 OQX, Oxon., UK
(
Refereed)
Received 19 April 2002; accepted 19 July 2002
Abstract
Reaction of non-stoichiometric SrFeO (2:5 ꢀ x ꢀ 3:0) with carbon dioxide induced by
x
high temperature and mechanical energy was investigated. The high-temperature reaction
route was found to convert SrFeO to a mixture of SrFe O and SrCO at temperatures below
x
12 19
3
1
temperature causes a complete decomposition of SrFeO to strontium carbonate and hematite
128 K under 0.98 atm of CO pressure. The mechanical reaction carried out at room
2
x
via a metastable ferric carbonate phase. We examined the phase transformations, the reaction
kinetics and morphology of carbonated SrFeO and compared the two processes.
x
#
2002 Elsevier Science Ltd. All rights reserved.
Keywords: A. Ceramics; A. Oxides; C. Electron microscopy; C. Thermogravimetric analysis (TGA);
C. X-ray diffraction
1
. Introduction
Non-stoichiometric strontium iron ferrite SrFeO (2:5 ꢀ x ꢀ 3:0) is a mixed
x
oxygen conductor at elevated temperatures. Its properties are important not only
from scienti®c point of view but also because of its possible applications. The ferrite
or its derivatives can be used in pressure-driven oxygen generators, as electrodes in
solid oxide fuel cells (SOFCs) or partial oxidation generators [1,2].
The composition x and the crystal structure of SrFeO are functions of temperature
x
and oxygen partial pressure [3±9]. Unfortunately there is very little information about
the interaction of SrFeO with gaseous oxides [10]. In this paper we study the
x
*
Tel.: ‡44-1235-445581; fax: ‡44-1235-445720.
E-mail address: marek.schmidt@rl.ac.uk (M. Schmidt).
0025-5408/02/$ ± see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S 002 5 - 5 4 08 ( 02 ) 00 8 9 8 - X

2094
M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
interaction of SrFeO with carbon dioxide, since in real world applications SrFeO
x
x
would be exposed to this gas. The chemical reaction was induced by temperature and
additionally by mechanical energy. The investigations involve description of com-
positional and morphological changes in SrFeO exposed to CO and comparison of
x
2
the thermal and the mechanical reaction paths.
2
. Experimental procedure
SrFeO in its oxygen-de®cient form (Sr Fe O , x ˆ 2:5) was prepared from a
x
2
2 5
mixture of hematite (Fe O , 99.9% purity from Electronic Space Product Interna-
2
3
tional) and strontium carbonate (SrCO , 99.9‡% purity from Aldrich) combined in
3
the molar ratio 1:2, respectively. The mixture was ball milled for 12 h and then ®red
on an alumina combustion boat at 1373 K in air for total of 55 h. During the procedure
the sinter was ground and analysed for traces of transient phases using X-ray powder
diffraction. Annealing was ®nished upon the disappearance of impurities. SrFeOx
prepared this way is oxidised and was subsequently reduced to Sr Fe O by annealing
2
2 5
at 1473 K for 8 h in a ¯ow of ultra high purity argon (pO2 < 1 ppm).
Carbon dioxide used in experiments was 98% pure food-grade gas. The 2%
impurity was most likely air.
Thermogravimetric experiments were carried out using Shimadzu TGA-50analyser.
Different phases were identi®ed by X-ray powder diffraction using Philips PW1050
diffractometer equipped with a Co tube and a diffracted beam monochromator. The
morphology of samples was examined using a ®eld emission scanning electron
microscope (SEM, Hitachi S4500). SEM samples were prepared by dispersing powder
in alcohol using an ultrasonic probe and applying a droplet of slurry on a graphite tape.
Samples after drying were sputter coated with Pt to prevent charging in electron beam.
Ball milling experiments were carried out using a vertical stainless steel mill [11].
The mill was charged with four 1 in. ball bearing, 10g of powder and after evacuating
was ®lled with CO to a pressure of 300 kPa. The mill operated in a high-energy mode
2
with a speed of 200 rpm. During milling small samples of powder were taken for
analysis and the mill was re®lled to the initial pressure.
3
. Results and discussion
Since CO is an oxide apart from its carbonation capabilities the gas is also
2
expected to dissociate at high temperature. Calculated partial pressure of oxygen in
CO due to dissociation is presented in Fig. 1 as a function of temperature. It was
assumed that the dissociation reaction takes place according to the equation:
2
2
CO ˆ O ‡ 2CO;
2
2
and the calculations were made using thermodynamical data [12,13]. However, the
À4
oxygen partial pressure due to dissociation is rather low reaching 10 atm at 1473 K

M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
2095
2
Fig. 1. Calculated oxygen partial pressure due to dissociation of CO as a function of temperature.
and had negligible effect on the ferrite since the amount of oxygen in SrFeOx
decreases with temperature [3]. However, the 2% air impurity gives rise to a
0.004 atm partial pressure of oxygen, which was temperature independent and caused
partial oxidation of SrFeO on cooling.
x
3
.1. Thermal carbonation
À1
Sr Fe O was heated in TGA at a rate of 10K min
2
in a CO atmosphere, the
2
2
5
effects of which are presented in Fig. 2. The ferrite started gaining mass around 673 K
and because of relatively slow reaction rate the sample was kept at 1073 K for 1 h. The
compound gained almost 7% of mass, which is more than that can be accomplished by
oxidation alone so the mass gain must correspond to a carbonation reaction. The
shape of the TG curve (Fig. 2) indicates that the reaction was not completed (lack of
saturation). On further heating the carbonate suddenly decomposed around 1140K.
However, the sample gained some mass on the cooling cycle. X-ray diffraction
examination of the remains revealed traces of strontium carbonate, Sr Fe O and a
2
2 5
product of its oxidation.

2096
M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
À1
.
Fig. 2. Thermogravimetric trace of Sr
2
Fe
2
O
5
heated in CO
2
at a rate of 10K min
The peakinFig. 2 isa result offormationand subsequentdecomposition ofstrontium
carbonate. In order to establish the decomposition temperature more precisely, a ball
milledstoichiometricmixture ofhematiteand strontiumcarbonatewas heatedinCO at
2
À1
a rate of 10 K min . The resulting TGA scan presented in Fig. 3, shows that the
temperatureforthe onset ofmasslossis1128 Æ 5 K. Thistemperatureisapproximately
2
40K higher than the temperature required to initiate reaction in an argon atmosphere
see Fig. 3). Because of slow carbonation kinetics, further analysis was carried out on
(
samples annealed in a furnace. Sr Fe O samples were annealed in ¯owing CO at
2
2
5
2
1
073 K for up to 480 h. X-ray diffraction patterns taken after different carbonation
times are presented in Fig. 4. Analysis of these patterns revealed increasing amount of
SrCO , decreasing Sr Fe O content, signs of oxidation in form of Sr Fe O and
SrFe O as a strontium-de®cient ferrite. However, there were no traces of pure iron
3
2
2
5
4
4 11
1
2 19
oxides such as hematite. Although, Sr Fe O partially oxidised to Sr Fe O during the
4 11
2
2
5
4
carbonation and its strongest diffraction line is quite conspicuous in Fig. 4 (the 17 h
sample). Most of the oxidation occurred on cooling to room temperature [3].
The thermal carbonation reaction of Sr Fe O can be written as follows:
2
2 5
6
Sr Fe O ‡ 11CO ˆ SrFe O ‡ 11SrCO :
2
2
5
2
12 19
3

M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
2097
Fig. 3. Thermogravimetric trace of Fe2O3 ‡ 2SrCO3 mixtures heated in CO and Ar and a trace of
2
À1
6
SrCO3 ‡ Fe2O3 mixture heated in CO
2
at a rate of 10K min . The curves are shifted for clarity.
The oxidation of Sr Fe O was neglected in the reaction because of its small
2
2 5
contribution at high temperature (x ! 2:5 at elevated temperature [3]) and dimin-
ishing amount of the oxidised phase. Also the decreasing amount of Sr Fe O
4 11
4
4
‡
indicates that the presence of Fe does not inhibit the carbonation reaction. The
extent of the carbonation reaction was calculated from the mass gain of the samples
and is presented in Fig. 5 as a function of annealing time. There are two stages of the
carbonation reaction with different kinetics mechanisms. The first stage lasted for few
hours and was responsible for creation of 40% of the product (SrFe O ). It was
1
2 19
followed by a second stage characterised by slow mass gain. In order to estimate a
time required to complete conversion of Sr Fe O to SrFe O the experimental
2
2
5
12 19
points corresponding to the second stage were fitted assuming that we are dealing
with a first-order reaction. Calculation showed that to obtain 99% of completion, the
annealing would have to be carried out for 1500 h and therefore further heat treatment
was abandoned.
During the carbonation experiments there were no signs of SrFe O decom-
1
2 19
position and annealing of pure SrFe O in CO for few days did not cause any
19
12
2

2098
M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
Fig. 4. X-ray diffraction patterns of Sr
time. Figures denote the annealing time; strokes mark diffraction lines of Sr Fe O (a), SrCO (b),
2 2 5
Fe O samples carbonated at 1073 K for different periods of
2
2
5
3
SrFe O (c). The arrow marks the strongest line of Sr Fe O .
4 11
1
2
19
4
changes either. The interaction of SrFe O with carbon dioxide was examined by
annealing of a ball milled mixture of hematite and strontium carbonate in the molar
ratio of 6:1, respectively. The TGA scan obtained at a constant heating rate of
12 19
À1
1
However, in a CO -de®cient atmosphere SrFe O formation starts around 880K
0K min
revealed an onset of SrFe O formation at 1088 Æ 5 K (see Fig. 3).
1
2 19
2 12 19
[14]. This experiment indicated that strontium hexaferrite is affected by CO but its
decomposition in CO must be extremely slow.
2
2
The SEM images of pure Sr Fe O and the thermally carbonated samples are
2
2 5
presented in Fig. 6. The crystal size of the starting Sr Fe O , obtained from X-ray
2
2 5
diffraction line broadening, was around 440nm. Starting Sr Fe O forms agglom-
2
2 5
erates of several tens of micrometers in size. A representative particle is presented in
Fig. 6 (top row). The sintered fragments appear as solid and free of cracks. Sr Fe O
did not develop crystal facets. Annealing in CO did not change the size of particles
2
2 5
2
and the carbonated material still consists of sintered grains. The appearance of the
carbonated materials was found to be the same irrespective of the annealing time
(Fig. 6, middle rows). However, the carbonated grains are divided by obvious cracks
and images taken at higher magni®cations show clearly separated domains of material
approximately 1±2 mm in size not observed in pure Sr Fe O . Apart from the ferrite
2
2 5
there are elongated crystals visible on the surface of grains. Even though, the electron
microscope was not equipped with energy dispersive detector, the strong scattering
of electrons by the rod like crystals indicate that they contain a large fraction of
strontium. Elongated crystals are characteristic for SrCO and just for comparison
3

M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
2099
Fig. 5. Extent of Sr
discussed in the text.
2 2 5
Fe O carbonation reaction as a function of annealing time. Solid lines are
images of pure SrCO and SrFe O are presented in Fig. 6 (bottom row). The
3
12 19
appearance of SrCO and SrFe O is suf®ciently different to allow identi®cation of
3
12 19
two main compounds from morphology itself.
The cracked surface of the ferrite grains leads to a likely explanation of the two
different kinetics mechanisms of SrFeO conversion to SrFe O . Fig. 5 shows that
x
12 19
the ®rst stage of reaction was very rapid and produced approximately 40% of the
product. The initial stages of solid±gas reaction are rapid since the surface material is
consumed. It appears that carbon dioxide can diffuse along the grain boundaries,
effectively increasing the surface area, which explains the initial rate of reaction. The
cracks in the material appeared as a result of molar volume changes between the
starting material (Sr Fe O ) and SrFe O . Once the material from the surface
2
2
5
12 19
converts to the hexaferrite it shrinks exposing the gaps. At the end of the ®rst stage of
reaction the ferrite consist of SrFe O grains with a SrFeO core. Since the
12
19
x
hexaferrite is resistant to carbonation further reaction is limited by diffusion of Sr
through the layer of SrFe O slowing down the process and changing the rate-
12 19
determining step from gaseous diffusion to solid-state diffusion, which is typically
much slower.

2100
M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
Fig. 6. SEM images of: pure Sr Fe O (top row), a sample carbonated at 1073 K for 45 h (2nd and 3rd
2 2 5
rows), pure SrFe12
magni®cations.
3
O19 (bottom row, left) and pure SrCO (bottom row, right) taken at different

M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
2101
3
.2. Mechanical carbonation
Milling of Sr Fe O was carried out without any liquid although it is possible to use
2
2 5
water which leads to a very rapid formation of the carbonate due to the action of
carbonic acid [15,16]. However, this line of experiment was not pursued for two
reasons. First of all, the acid forms a corrosive environment which attacks the mill
[17] and the products of corrosion would spoil the analysis. Secondly, water is not
inert towards iron oxides and it would most certainly form hydrated forms of the
oxides [16]. The route for the hydration reaction depends on the amount of added
water, which would greatly complicate the analysis since there would be several
chemical reactions occurring simultaneously.
The milled samples were examined using X-ray diffraction and selected patterns
are presented in Fig. 7. As expected the main product of the carbonation reaction was
strontium carbonate; it appeared after just 4 h of milling and its presence is marked by
the strongest line at 2y ˆ 29:38 (the observed line is actually a doublet). The fraction
of carbonate and thus the intensity of line increased with time as the intensity of
Sr Fe O lines decreased. At the same time, the lines broadened due to decreasing
2
2 5
crystal size and accumulated strain. An amorphous phase appeared after 32 h of
milling and persisted to the end of the process. Its presence was indicated by a broad
peak around 2y ˆ 358.
Formation of strontium carbonate should lead to a simultaneous appearance of a
strontium-de®cient ferrite such as SrFe O in the case of thermally carbonated
12 19
samples. However, up to 256 h of milling there were no signs of any crystalline form
of iron compound (apart from Sr Fe O of course). On further milling hematite lines
2
2 5
2 2 5 2
Fig. 7. X-ray diffraction patterns of Sr Fe O milled in CO for different periods of time. Figures
denote the milling time and the arrows mark the positions of two strongest hematite lines.

2102
M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
become visible but after 656 h of milling the X-ray diffraction line intensity ratio
of hematite to carbonate was still smaller than expected for a mixture of 2SrCO ‡
3
Fe O . There was no evidence of SrFe O in the diffraction patterns, although the
2
3
12 19
samples were slightly magnetic implying its presence. However, their magnetic
nature was not nearly as strong as in the thermally carbonated ferrite. The lack of an
iron compound and the small intensity of hematite lines indicate that Fe was bound in
some metastable compound, possibly a form of iron carbonate.
The unknown phase decomposed slowly during milling with hematite as a product.
The decomposition to hematite had started before the starting phase (Sr Fe O )
2
2 5
vanished (the 256 h pattern in Fig. 7 contains traces of hematite). However, this
overlap may be caused by the randomness of the milling reaction. Grains of powder
interact with milling media randomly and in the case of multistage reactions a fraction
of powder may undergo the next step of reaction while other particles are still in the
previous stage. This causes the overlap of the reaction stages. The carbonation
reaction appears to follow the scheme:
Sr Fe O ! unknown phaseꢁs† ‡ nSrCO ! Fe O ‡ 2SrCO ;
2
2
5
3
2
3
3
that ultimately produces SrCO and Fe O . The mechanically carbonated samples
3
2 3
À1
were analysed using TGA by heating them at a rate of 10K min up to 1473 K in an
argon atmosphere. The results are presented in Fig. 8. The figure also shows the
reaction trace of a hematite-strontium carbonate mixture (dashed line in Fig. 8). This
mixture shows only one stage of reaction, which corresponds to decomposition of
SrCO and formation of Sr Fe O . However, the mechanically carbonated samples
3
2
2 5
exhibit an additional low temperature reaction, which commences around 473 K and
overlaps with the decomposition of SrCO . This indicates the decomposition of an
3
additional phase. An attempt to separate the stages by decreasing the heating rate was
unsuccessful. Also extended annealing of milled powders in argon and CO below
2
8
73 K did not cause crystallisation of the unknown phase.
The mass loss due to the low temperature stage increased signi®cantly up to 256 h
of milling with only a slight increase to 656 h of milling (compare the curves for 256
and 656 h in Fig. 8). This corresponds to the increasing fraction of the metastable
phase and is consistent with the X-ray diffraction ®ndings. Samples milled for longer
than 256 h exhibited greater total mass loss than expected from a reaction of
2
SrCO ‡ Fe O .
3
2
3
The unknown carbonate is metastable and decomposes upon annealing and
milling. Since hematite appeared after 256 h of mechanical treatment the compound
must contain a signi®cant amount of iron and the presence of some strontium in the
compound cannot be excluded. X-ray diffraction did not show any unexplained lines
that might correspond to the unknown carbonate phase proving, that the carbonate is
either amorphous or its re¯ections overlap with broad lines of the other phases.
However, intensities of lines characteristic to the particular compounds are not visibly
changed so the former supposition is likely to be true. There are two known iron
carbonates, ferrous carbonate FeCO (siderite) and ferric carbonate Fe O CO
3
2
2
3
described by Erd oÈ s and Altorfer [18]. The starting material contained only ferric

M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
2103
Fig. 8. Thermogravimetric traces of mechanically carbonated Sr Fe O heated in argon at a rate of
2
2 5
0K min À1. The dashed line represents a trace of 2SrCO3 ‡ Fe2O3 mixture as discussed in text. The
1
®
gures denote milling time.
iron so to form siderite; the iron would have to be reduced. Such a reduction is feasible
in the mill used for experiments due to possible contamination with elemental iron.
2‡
The milled samples were tested for the presence of Fe using 0.1% solution of
o-phenanthroline. The test did not detect any ferrous iron and the test concentration
limit is 1 in 1,500,000 [19]. The test excluded siderite as a potential iron carbonate.
Additionally the materials were examined for excess iron by X-ray examination of the
TGA products. The examination revealed only Sr Fe O , which indicates lack of iron
2
2 5
contamination. So it can be safely concluded that the unidenti®ed carbonate does not
contain ferrous iron and the milling did not change the overall iron to strontium ratio
in the compound.
Even though the gas used for experiments contained a small amount of oxygen
there was no visible signs of oxidation [20]. Therefore, the fraction of tetravalent iron
has to be negligibly small and the unknown carbonate has to contain the trivalent iron.
Lack of any structural information does not allow positive identi®cation of the
carbonate as Fe O CO since its strongest diffraction line at 2y ˆ 20:248 was not
2
2
3
observed in Fig. 7.

2104
M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
Fig. 9. SEM images of Sr
magni®cations.
2 2 5
Fe O sample mechanically carbonated for 256 h taken at different
Sr Fe O carbonated by milling was also examined using SEM. These images are
2
2 5
presented in Fig. 9. However, due to the destructive nature of milling the carbonation
products are featureless compared with the thermally carbonated samples and looked
very similar irrespective of the milling time. The powder consisted of grains up to
several micrometers across formed by a coalescence of smaller (submicron) crystals.
The crystal size of SrCO , estimated from X-ray diffraction line breadth, was found to
3
be 13 nm and hematite has a crystal size of around 40nm. Therefore even the smallest
particles visible under the microscope are polycrystalline.
The surface area of the 656 h sample was measured using the BET method using
nitrogen as an adsorbant. The sample was degassed under vacuum at 473 K prior to
2
À1
the analysis. The measurement gave a speci®c surface area of only 2.66 m g that
precludes any porosity of the milled material or a large fraction of loose submicron
crystals. It also shows that surface adsorption was not responsible for the mass loss
observed during the thermogravimetric analysis.
4
. Conclusions
Temperature induced carbonation of SrFeO starts around 673 K and produces
x
SrCO and SrFe O as products. SrFe O is resilient towards carbonation. The
12 19
3
12 19
carbonation reaction has a critical temperature above, which the products of
carbonation convert back to SrFeO . This temperature depends on partial pressure
x
of CO (1128 K at 0.98 atm of CO ) and decreases with decreasing partial pressure of
2
2
carbon dioxide. The thermal carbonation reaction has two kinetic stages, the ®rst
determined by surface reaction and the second, which is limited by diffusion of Sr
through SrFe O . The carbonation produces microcracks on the surface of SrFeO ,
1
2
19
x
which may be potentially very deleterious to performance if SrFeO is used in oxygen
x
conducting devices exposed to CO2.
Carbonation of SrFeO induced by mechanical energy causes the decomposition of
x
the ferrite to SrCO and hematite. However, milling produces an intermediate
3
metastable carbonate, which decomposes on prolonged milling forming hematite.

M. Schmidt / Materials Research Bulletin 37 (2002) 2093±2105
2105
This metastable phase is most likely amorphous and contains signi®cant amount of
ferric iron, however the presence of some strontium in the compound cannot be
excluded. The metastable phase decomposes on annealing at relatively low tem-
peratures.
Acknowledgments
This work was performed under the sponsorship of The Board of Trustees of
Michigan State University. This work not necessarily represents the views of the
University. Reproduction of this article, with the customary credit to the source, is
permitted.
References
[
[
1] B.C.H. Steele, Curr. Opin. Solid State Mater. Sci. 1 (1996) 684.
2] B.C.H. Steele, Ceramic electrochemical reactors incorporating solid oxide electrolytes, in:
B.C.H. Steele, D.P. Thompson (Eds.), British Ceramic Proceedings. Advanced Ceramics in
Chemical Process Engineering, number 43, Institute of Ceramics, Stoke-on-Trent, England,
December 1988, pp. 163±171.
[
[
3] Y. Takeda, K. Kanno, T. Takada, O. Yamamoto, J. Solid State Chem. 63 (1986) 237.
4] M. Takano, T. Okita, N. Nakayama, Y. Bando, Y. Takeda, O. Yamamoto, J.B. Goodenough, J.
Solid State Chem. 73 (1988) 140.
[
[
[
[
[
5] J.-C. Grenier, N. Ea, M. Pouchard, P. Hagenmuller, J. Solid State Chem. 58 (1985) 243.
6] S. Shin, M. Yonemura, H. Ikawa, Mater. Res. Bull. 13 (1978) 1017.
7] J. Mizusaki, M. Okayasu, S. Yamaguchi, K. Fueki, J. Solid State Chem. 99 (1992) 166.
8] M. Schmidt, S.J. Campbell, J. Solid State Chem. 156 (2001) 292.
9] M. Schmidt, S.J. Campbell, J. Phys. Chem. Solids 63 (2002) 2085.
[
[
[
10] S. Shin, Y. Hatakeyama, K. Ogawa, K. Shinomura, Mater. Res. Bull. 14 (1979) 133.
11] A. Calka, A.P. Radlinski, Mater. Sci. Eng. A 134 (1991) 1350.
12] I. Barin, Thermodynamical Data of Pure Substances, VCH Verlags Gesellschaft, Weinheim,
1
989.
13] I. Barin, Thermodynamical Data of Pure Substances, VCH Verlags Gesellschaft, Weinheim,
993.
[
1
[
[
14] F. Habery, A. Kockel, IEEE Trans. Magn. 12 (1976) 983.
15] R.J. Reeder (Ed.), Carbonates: Mineralogy and Chemistry. Reviews in Mineralogy, vol. 11,
Mineralogical Society of America, 1990.
[
[
16] W. Stumm, Chemistry of the Solid±Water Interface. Processes at the Mineral±Water Particle±
Water Systems, Willey-Interscience, New York, 1992.
17] R.J. McKay, R. Worthington, Corrosion Resistance of Metals Alloys, Reinhold Publishing
Corporation, New York, 1936.
[
18] E. Erd oÈ s, H. Altorfer, Werkstoffe und Korrosion 27 (1976) 304.
19] A.I. Vogel, A Text-book of Macro Semimicro Qualitative Inorganic Analysis, 4th ed., Longmans,
[
1954.
[
20] M. Schmidt, Mater. Res. Bull. 35 (2000) 169.