J. Am. Ceram. Soc.

, 83 [8] 2036 – 40 (2000)

journal
Reaction Mechanism in Alumina/Chromia (Al2O3–Cr2O3) Solid
Solutions Obtained by Coprecipitation

Federica Bondioli, Anna Maria Ferrari, Cristina Leonelli,* and Tiziano Manfredini*
Department of Chemistry, Faculty of Engineering, University of Modena, 41100 Modena, Italy

Laura Linati
Centro Grandi Strumenti, University of Pavia, 27100 Pavia, Italy

Piercarlo Mustarelli
Department of Physical Chemistry and CSTE-CNR, 27100 Pavia, Italy

The aim of this work is to study the structural characteristics of 945°C and a critical composition of 45 mol% at a pressure of
and properties of the solid solution (Al,Cr)2O3. XRD analysis, 1000 bar. At 1 bar pressure the miscibility gap was observed
27
Al MAS-NMR measurements, and microstructural charac- previously by Sitte3 on sintered pellets (1400°C, 72 h) by emf
terization were used to determine the relationship between measurements; in this work the upper critical temperature is about
color and crystallochemical properties of the compounds 300°C higher than that observed in earlier work and it was
formed. In particular, to determine more accurately the mech- associated with an increase in the molar volume compared with the
anism of solid solution formation above the miscibility gap of pure compounds.
the system, the marker technique was used. In order to define The high temperatures used for its synthesis limit the prepara-
the behavior of the system for temperatures below the misci- tion, the study, and the use of this solid solution. Therefore, there
bility gap at 1 bar pressure, the composition Al2O3:Cr2O3 1:1 is a growing interest in “nonconventional” methods of synthesis
was studied with high-temperature XRD. that require lower temperatures and shorter times. Of the chemical
methods, the colloidal sol– gel method gives rise to a molecular
homogeneity4 which in turn enables shorter average ionic diffu-
I. Introduction sion pathways and higher reaction rates. This synthesis method
generally produces amorphous and homogeneous materials which,
T HE increasing interest in solid solutions of ceramic oxide
materials is mostly due to their structural characteristics and
their thermal and chemical stability. Their properties (e.g., con-
given their high entropy, can form metastable crystalline phases.
The aim of this work is to study the structural characteristics
and the properties of the substitutional solid solution chromia–
ductivity, ferromagnetism) can be easily designed and modified in alumina between 1000° and 1500°C at 1 bar pressure, obtained by
a continuous range simply by changing the ratio of the two starting solid-state reaction starting from coprecipitated hydroxides.
materials. We proceeded as follows:
Concerning isovalent solid solutions, in which an atom or ion (a) Solid solutions were prepared by coprecipitation of hy-
replaces an atom or ion of the same charge in the parent structure, droxide precursors (starting from a solution of cation nitrates) and
the study of the (Al22xCrx)O3 (0 # x # 2) system is of great subsequently calcined.
interest because of its high refractoriness and chemical stability.
Al2O3 and Cr2O3 are sesquioxides having the same corundum
crystal structure (approximately hexagonal close-packed oxide
ions with the Al31 and Cr31 ions occupying two thirds of the
available octahedral interstitial sites), and by reaction at high
temperatures (T . 1000°C) complete ranges of substitutional solid
solutions are obtained. Below 950°C a miscibility dome is present,
inside which two crystalline phases (alumina-rich and chromia-
rich solid solutions) are present (Fig. 1). The actual decomposition
of a homogeneous solid solution into two phases takes place very
slowly, but it may be speeded up using hydrothermal or high-
pressure treatments. Neuhaus et al.1 investigated this system
hydrothermally at pressures between 1600 and 2000 bar and
reported the presence of a miscibility gap within the limits of ;10
and 60 mol% Cr2O3 and which narrowed with increasing temper-
ature to finally close at 800°C with a composition of 40 mol%
Cr2O3. In contrast, Roy and Barks2 reported a critical temperature

R. S. Roth—contributing editor

Manuscript No. 189736. Received November 23, 1998; approved February 15,
2000.
*Member, American Ceramic Society. Fig. 1. Phase diagram of the alumina– chromia system.

2036
August 2000 Reaction Mechanism in Alumina/Chromia Solid Solutions 2037

(b) A systematic color study of the solid solutions was the top of this reactant. The other reactant covers the wires. The
conducted to ascertain the relationship between the red-green color crucible is then heated at the reaction temperature. Information
change and the phases present. concerning ionic motion during the reaction is obtained from the
(c) The mechanism of solid solution formation was identified location of the product after firing. In the present study, the marker
and compared with that used to produce powders by the conven- technique was slightly modified. Instead of using a crucible, a
tional process based on mechanical mixing of commercial hydrox- pellet of precipitated aluminum and chromium hydroxides, distin-
ides and their subsequent calcination. guished by Pt wires, was prepared at 400 kg/cm2 pressure. This
pellet was heated in an electric furnace at 1200°C for 2 h (heating
rate 10°C/min). The samples were carefully removed from the
II. Experimental Procedure furnace and mounted for SEM examination.
(1) Sample Preparation
Amorphous hydroxides containing chromium oxide in the III. Results and Discussion
0 –100 wt% range were prepared by coprecipitation. These gels of
Al31 and Cr31 hydroxides were coprecipitated with ammonia (1) Powder Characterization
(NH4OH, RPE, Carlo Erba, Milan, Italy) from aluminum nitrate To gain insight into the thermal reactivity of the coprecipitated
(Al(NO3)3, RPE, Carlo Erba) and chromium nitrate (Cr(NO3)3, particles, the gels were first characterized by scanning electron
RPE, Carlo Erba) solution at pH 9. The amorphous coprecipitated microscopy. SEM (Fig. 2) demonstrated that the dried unconven-
hydroxides, carefully washed with distilled water, were dried in tional gels were homogeneous, spherical, and had a small particle
conventional furnaces at 120°C and then ground in an agate size distribution, with an average size of about 0.1 mm. The
mortar. The powders obtained were calcined in air, in an electric specific surface area was ;48 m2/g.
furnace, for different times (1– 6 h) and at different temperatures The formation of Cr–O–Al bonds was shown by NMR spectra
(1200 –1500°C). performed on dried powders. The formation of Cr–O–Al bonds in
the gel powders was investigated by 27Al MAS-NMR. Figure 3
(2) Samples Characterization shows the spectra of aluminum hydroxide (a), of the mechanically
mixed sample with 10 mol% Cr2O3 (b), and of the dried copre-
Sample morphology was examined by scanning electron mi- cipitated sample with 10 mol% Cr2O3 (c). In all cases, a single
croscopy (SEM, Philips XL 40, The Netherlands) and energy peak was found at ;8 ppm, which is a typical chemical shift
dispersion X-ray spectroscopy (EDS, Philips PV9900, The Neth- position for aluminum in octahedral coordination.8 These features
erlands). Particle morphology was maintained using a critical- represent the spin central transition 21⁄2 7 1⁄2, which in a 27Al
point drying (CPD) process5 (Balzers instrument, CPD 010, MAS experiment is generally dominated by second-order quadru-
Switzerland). The specific surface area of dried coprecipitated polar interaction.9
hydroxides was measured by the BET method6 (Micromeritics The peaks of the spectra (a) and (b) are very similar (full width
Instruments, Gemini 2360, GA USA), using nitrogen as the at half-height, FWHH > 650 Hz), the only difference being related
adsorbate. The 27Al NMR data were collected with an to the intensity of the spinning sidebands (marked with stars),
AMX400WB spectrometer (Bruker, Karlsruhe, Germany) based which are stronger in the sample containing Cr2O3. This is due to
on a wide-bore superconducting magnet (B0 5 9.4 T), at a Larmor the existence of anisotropic magnetic interactions between the Al
frequency of 104.3 MHz. Magic angle spinning (MAS) experi- nuclei and the unpaired electrons of Cr,10 which is substantially
ments were performed using a 4 mm probehead (Bruker, removed by MAS rotation. On the other hand, we do not observe
Karlsruhe, Germany) equipped with cylindrical zirconia rotors and isotropic contributions to either the MAS line width or position,
a boron nitride stator. The samples were spun at 13.5 kHz and the which means that the second-shell chemical bonds between Al and
spectra were acquired using a single-pulse sequence, with a 30° Cr are dissimilar. We would stress that the structure that we
pulse of 0.4 ms and a recycle time of 20 ms. The spectra were observe in peaks (a) and (b) is likely to be due to an imperfect
referenced to Al(H2O)31 6 . averaging of the quadrupolar interaction.
The thermal behavior of the samples after drying was studied by By contrast, the peak of the coprecipitated sample (c) does not
thermogravimetry and differential thermal analysis in air at a show any structure, and its line (FWHH > 1.7 kHz) is nearly 3
heating rate of 20°C/min, using a simultaneous TGA/DTA appa- times larger than that of the reference sample (a).
ratus (Netzsch STA409, Selb, Germany). Crystalline phase iden- We attribute the line enlargement to a distribution of isotropic
tification was performed using a computer-assisted X-ray (CuKa) chemical shifts which cannot be averaged under MAS conditions.
powder diffractometer (XRD, Philips PW 3710, The Netherlands) This points to the formation of Cr–O–Al bonds in the sample
hot-stage equipped. The bulk-heated samples were ground to less obtained by coprecipitation. In addition, peak (c) displays a tail
than 20 mm and the XRD patterns were collected in the 20 –70° 2u asymmetry on the low-frequency side, as generally observed in
range at room temperature. The scanning rate was 0.004°/s, step
size 0.02°. The lattice parameters were determined by a least-
squares refinement.
To determine the reaction mechanism of powders with compo-
sition 50 wt% Cr2O3 in the temperature range inside the miscibility
gap, XRD analyses with a hot stage were performed. The powders
were obtained either by coprecipitation or by mechanical mixing
of a synthetic mixture of hydroxides (Al(OH)3 and Cr(OH)3, RPE,
Carlo Erba) and the diffraction patterns were collected in the
24 –27° 2u range at temperatures varying from 600° to 1200°C in
200°C increments. The scanning rate was 0.004°/s, step size 0.02°,
the heating rate 20°C/min, and the dwell time at each temperature
15 min.
To investigate the color and the optical properties of the
powders, the samples were analyzed using a UV–vis spectropho-
tometer with analytical software for color measurements (Model
Lambda 19, Perkin-Elmer, Norwalk, CT).
To study the mechanism of solid solution formation above the
miscibility gap, the marker technique was applied.7 In the classical
method, one reactant is placed in a crucible; markers, such as Pt
wires, which are inert to the reaction to be studied, are placed on Fig. 2. SEM micrograph of the dried gel.
2038 Journal of the American Ceramic Society—Bondioli et al. Vol. 83, No. 8

already at 1200°C after being treated for 2 h (Fig. 5), whatever the
level of chromium oxide content.
The compositions studied form a continuous range of solid
solutions as shown by XRD patterns (Fig. 6), where only the peak
(012) is represented. Increasing the amount of Cr2O3 in the
powders, the distances between a-Al2O3 planes increase and there
is a shift of the peaks corresponding to the pure aluminum oxide
which indicates formation of the solid solution.
According to Vegard’s law, the change in the unit cell param-
eters, a and c, is linear with composition (Table I), owing to the
substitution of the Cr31 ion for Al31 in the corundum crystal
structure.13 The agreement with published findings is very
good.14,15
A transition-metal ion forms an absorption spectra which has
absorption peaks shifted toward areas which are more or less
energetic, depending on the intensity of the crystalline field
surrounding the dopant ion.
Therefore, in a strong field, a shift occurs toward greater
energies (shorter l) of the absorption bands characteristic of pure
Cr2O3, where the Cr31 ion has octahedral coordination, associated
with the transitions
4
A2g 3 4 T2g
4
A2g 3 4 T1g

The evolution of the diffuse reflectance spectra for powders

with increasing Cr2O3 content is given in Fig. 7 with respect to the
pure Cr2O3 spectra.
Fig. 3. 27Al MAS-NMR spectra of aluminum hydroxide (a), of the The two bands, at 445 and 590 nm in the spectra of pure Cr2O3,
mechanically mixed sample with 10 mol% Cr2O3 (b), and of the dried fall as a consequence of the strong crystal field expected in the
coprecipitated sample with 10 mol% Cr2O3 (c). solid solutions. The intensity of the absorption edge depends critically
on the chromium content, increasing with increasing Cr31. Since the
crystal field splitting, D0, is directly correlated to distance between the
27 transition metal and surrounding ligands by the formula
Al spectra of disordered materials.9 After drying, the samples
exhibit a characteristic XRD amorphous pattern. Z Le 2 4 ^r 4 &
For the samples obtained by chemical synthesis TGA and DTA D 0 5 10Dq 5 5 ^r & 5 Q (1)
6R R5
analysis indicated an endothermic weight loss corresponding to the
elimination of humidity water at temperatures of about 120°C. The where Q is a costant, ZLe is the charge of the ligands separated by
measurements for the dried samples (Fig. 4) exhibit low- a distance R from the cation, and r is the polar coordinate of the
temperature endothermic peaks at 240°, 325°, and 420°C, corre- electron, the red shifts of the absorption bands with increasing
sponding to the conversion of gibbsite to boehmite, gibbsite chromium concentrations due to expansion of the octahedral sites
dehydration, and boehmite dehydration, respectively.11 The sharp by larger Cr31 (octahedral ionic radius, roct 5 61.5 pm) replacing
exothermic peak at about 345°C appears to be characteristic of the Al31 ions (roct 5 53.5 pm) in the corundum structure (Table I).16
alumina amorphous phase12 but does not correspond to crystalli- As regarding the color of the powders, in the case of a low
zation. An exothermic peak, not related to weight loss but due to chromium content (,20 wt%), the samples exhibit pink tonalities
the formation of the (Cr,Al)2O3 solid solution was observed in the and there are Cr31 ions isolated or interacting in pairs. On
DTA curve at a temperature of about 1300°C. A different amount increasing the chromium oxide content, the color becomes gray,
of chromium oxide does not lead to any qualitative differences in because there is an increase in the spin–spin interactions between
the DTA/TGA curves. the Cr31 ions lying along the crystal axis. Finally, in the case of
higher chromium oxide content (.80 wt%), the color becomes
(2) Temperature Range above the Miscibility Gap green, owing to Cr31–O22–Cr31 interactions, until the powder
To investigate the nature of the samples after calcination, all the consists of pure Cr2O3.17
heat-treated samples were subjected to powder X-ray diffraction.
The powders formed a highly pure and crystalline solid solution

Fig. 5. XRD patterns of powders (10 wt% Cr2O3) calcined at 1200°C for
Fig. 4. DTA/TGA curves of the dried samples. 2 (a), 4 (b), and 6 h (c).
August 2000 Reaction Mechanism in Alumina/Chromia Solid Solutions 2039

Fig. 6. XRD patterns of (012) peak of solid solutions with 10 (a), 30 (b), 50 (c), and 90 wt% (d) of chromium oxide. E 5 eskolaite, C 5 corundum.

Table I. Unit Cell Parameters, a and c Underneath the interface there is an off-white interior which is
unreacted alumina. The fact that the reaction product is confined to
Composition (wt%) a (Å) c (Å)
the alumina layer suggests that Cr is the only mobile cation.
Al2O3 †
4.758 12.991 Furthermore, a green stain appears around the entire periphery of
10% Cr2O3 4.77400(37) 13.0300(14) the alumina layer, not just at the interface with the Cr2O3 layer,
20% Cr2O3 4.79100(48) 13.0760(19) indicating that Cr is transported through the vapor state.
30% Cr2O3 4.80900(64) 13.1240(24) To verify this optical evidence, the pellet was sectioned and
50% Cr2O3 4.85000(35) 13.2360(13) studied by SEM/EDS analysis to determine the quantity of Cr
70% Cr2O3 4.88900(36) 13.3560(13) (mol%) in the alumina layer with respect to the distance from the
80% Cr2O3 4.91000(26) 13.42400(94) interface (Fig. 8). Figure 8 shows that the chromium transported
90% Cr2O3 4.93400(13) 13.5020(4) through the vapor state at the alumina surface of the pellet is more
Cr2O3‡ 4.95876(14) 13.5942(7) than that transported by solid-state diffusion at the interface.
†
a-Al2O3, PDF File No. 10-173. ‡Cr2O3, PDF File No. 38-1479.
The diffusion coefficient of chromium in the alumina matrix
was determined by Fick’s second law. Assuming that the diffusion
constant (D) is concentration-independent, we have
]c ] 2c
5D 2 (2)
]t ]x
where c is the concentration of the diffusing species, t the diffusion
time, and x the direction of diffusion. To obtain a solution to Eq.
(2), it is necessary to establish the initial and boundary conditions.
For a very thin deposit on a samples of “infinite” thickness, as
we could assume to be the situation at the Al2O3 free surface, after
diffusion for a time t the concentration is described by
1 2 /4Dt
c~ x,t! 5 exp2x (3)
ÎpDt

Fig. 7. Diffuse reflectance spectra obtained for Cr2O3 (a) and various
compositions of the solid solutions Al2O3/Cr2O3 ((b) 20%, (c) 15%, (d)
10%, and (e) 5 wt% Cr2O3).

As previously stated, to study the mechanism of solid solution

formation we used a modified marker technique that enabled us to
obtain information concerning ionic motion during the reaction
from the location of the product after firing. If the reaction
proceeds by cationic counterdiffusion in a rigid oxygen matrix, the
product layer will appear on both sides of the inert markers;18,19 if
both oxygen and one of the cations diffuse, the product will appear
only on the side of the nondiffusing cation.20 In many oxide
systems, transport of oxygen ions occurs through the vapor state.21
Transport of a cation through the vapor state is indicated by a
characteristic product formation on the surfaces of the immobile
reactant, which have not been in contact with the other reactant.22
Optical investigation of the pellet sample leads us to postulate
the presence of two different reaction mechanisms. The Cr2O3
layer is green and the Al2O3 has a pink interface, which indicates Fig. 8. Chromium content of the alumina layer versus the distance from
the formation of a (Al,Cr)2O3 solid solution as reaction product. the interface.
2040 Journal of the American Ceramic Society—Bondioli et al. Vol. 83, No. 8

In a logarithmic plot, there is a linear trend, whose slope gives

the diffusion coefficient of the diffusion species.
On the other hand, at the Al2O3/Cr2O3 interface we could
assume to be in the situation of a juxtaposition of two “infinite”
samples one of which has a uniform concentration, c0, and the
other a concentration c1 of the diffusing species. After a time t the
solution of the Eq. (2) is

c~ x,t! 2 c 0 1
c1 2 c0
5
2 F
1 2 erf
x
2 H Î JGD
t
(4)

where erf is the Gaussian error function.

In the alumina matrix the diffusion speed of chromium is, in the
first case, 2.27 3 10213 cm2/s and, in the second case, 3.68 3 Fig. 10. XRD patterns of mechanical mixed powders (50 wt% Cr2O3) at
10213 cm2/s, as obtained by the bulk diffusion coefficient evalu- (a) 600°, (b) 800°, (c) 1000°, and (d) 1200°C; (*) Cr2O3, (E) Al2O3, (#) s.s.
ated in the two diffusing directions. The D values obtained at
1200°C are in agreement with those previously found by other
these bonds, the thermodynamic equilibrium, characterized by the
authors.23 Furthermore, the diffusion coefficient of Cr31 in Cr2O3
presence of two solid solutions, has been obtained in a shorter time
varies from 10212 to 1028 cm2/s (Ref. 24) in the temperature range
and at ambient pressure.
of 1200 –1600°C, while Al31 in polycrystalline Al2O3 presents in
a remarkably higher temperature interval (1670 –1905°C) a diffu-
References
sion coefficient ranging from 10212 to 10210 cm2/s.25
1
A. Neuhaus; p. 237 in Physics and Chemistry of High Pressure. Edited by the
Society of Chemical Industry. Gordon and Breach Science Publishers, New York,
(3) Temperature Range below the Miscibility Gap 1963.
2
Figures 9 and 10 report the XRD patterns of mixtures with a 50 D. M. Roy and R. E. Barks, “Subsolidus Phase Equilibria in Al2O3–Cr2O3,”
wt% of chromium oxide obtained by coprecipitation and by Nature Phys. Sci., 235, 118 (1972).
3
W. Sitte, “Investigation of the Miscibility Gap of the System Chromia–Alumina
mechanical mixing of synthetic hydroxides, collected in the below 1300°C,” Mater. Sci. Monogr., 28A, 451 (1985).
600 –1100°C temperature range. The reaction mechanism for the 4
B. Durand, Ceramic Powders; pp. 413–20. Elsevier Science Publishers, Amster-
coprecipitate powders involves the direct formation of a dam, Netherlands, 1983.
5
chromium-rich solid solution at a temperature of about 600°C and A. L. Cohen, “Critical Point Drying”; pp. 373–79 in Principles and Techniques for
SEM: Biological Applications, Vol. 1. Edited by M. A. Hayat. Van Nostrand
then, at a temperature of 1000°C, the separation of the aluminum- Reinhold, New York, 1974.
rich phase (Fig. 9). The same analysis performed on mechanically 6
L. Svarovsky, Powder Testing Guide: Methods of Measuring the Physical
mixed powders demonstrated that this mixture develops first Properties of Bulk Powders. Published on behalf of the British Materials Handling
(500°C) as chromium oxide, then (800°C) a-aluminum oxide, Board by Elsevier Applied Science, London, U.K., 1987.
7
R. A. Eppler, “Mechanism of Formation of Zircon Stains,” J. Am. Ceram. Soc., 53
ending at a temperature of 1200°C with the formation of a single [8] 457– 62 (1970).
solid solution (Fig. 10) that, in the range of temperature and 8
G. Engelhardt and D. Michel, High-Resolution Solid-State NMR of Silicates and
diffusion time that we studied, does not lead to thermodynamic Zeolites. Wiley, Chichester, U.K., 1987.
9
equilibrium, characterized by the presence of two solid solutions. See, for example: C. Jager; pp. 133–70 in Solid State NMR II: Inorganic Matter,
NMR Basic Principles and Progress, Vol. 31. Edited by B. Blumich. Springer-Verlag,
This is a demonstration that the coprecipitated powders present Berlin, Germany, 1994.
a faster reaction kinetic, not only because the particle size of the 10
M. Bose; pp. 335– 44 in Progress in Nuclear Magnetic Resonance Spectroscopy.
powders obtained is smaller, and hence much more reactive, but Edited by J. W. Emsley, J. Feeney, and L. H. Sutcliffe. Pergamon Press, London,
also because of the different reaction mechanisms involved. U.K., 1969.
11
A. L. Dragoo and J. J. Diamond, “Transitions in Vapor-Deposited Alumina from
300° to 1200°C,” J. Am. Ceram. Soc., 50 [11] 568 (1967).
12
S. G. Samdani, J. Laxminarasimhaiah, and A. Abde, “Preparation and Phase
IV. Conclusion Changes of Amorphous Alumina,” Trans. Indian Ceram. Soc., 31 [2] 57– 60 (1972).
13
R. M. Spriggs and S. L. Bender, “Vegard’s Law and the System Alumina–
We have demonstrated that in this system the coprecipitated Chromia,” J. Am. Ceram. Soc., 45 [10] 506 (1962).
14
L. R. Rossi and W. G. Lawrence, “Elastic Properties of Oxide Solid Solutions:
powders present faster reaction kinetics, not only because of the The System Al2O3–Cr2O3,” J. Am. Ceram. Soc., 53, 604 (1970).
smaller particle size, and consequently greater reactivity, but also 15
J. Graham, “Lattice Spacings and Color in the System Alumina–Chromic Oxide,”
because of Cr–O–Al bonds due to the coprecipitation process. In J. Phys. Chem. Solids, 17, 18 –25 (1960).
16
particular, for compositions inside the miscibility gap, because of D. Reinen, “Ligand-Field Spectroscopy and Chemical Bonding in Cr31-
Containing Oxidix Solids,” Struct. Bonding, 6, 30 –51 (1969).
17
G. Munuera, P. Valerga, and V. Rives, “Color y Transferencia de Carga en
a-Cr2O3 y sus Soluciones Solidas en d- y a-Al2O3 (Color and Charge Transfer in
a-Cr2O3 and Its Solid Solutions in d- and a-Al2O3,” An. Univ. Hisp., Ser.: Cienc., 21,
71– 84 (1977).
18
H. Schmalzried, “Solid State Reactions,” Angew. Chem., Int. Ed. Engl., 2 [5]
251–54 (1963).
19
E. Koch and C. Wagner, “Formation of Ag2HgI4 from AgI and HgI2 by Reaction
in the Solid State,” Z. Phys. Chem., B34 [3– 4] 317–21 (1936).
20
R. E. Carter, “Mechanism of Solid-State Reaction between Magnesium Oxide
and Aluminum Oxide and between Magnesium Oxide and Ferric Oxide,” J. Am.
Ceram. Soc., 44 [3] 116 –20 (1961).
21
C. Wagner, “Diffusion and High Temperature Oxidation of Metals in Atom
Movements,” Z. Phys. Chem., B34, 309 (1963).
22
B. Bengtson and R. Jagitsch, “Kinetic Studies on Formation of Spinel from Zinc
Oxide and Alumina,” Ark. Kemi, Mineral. Geol., 24A [18] 1–16 (1947).
23
E. Croquevielle Puelma and R. Cordero M., “Diffusion in the Solid State: The
Al2O3–Cr2O3 System,” Ceramica, 29 [168] 341– 45 (1983).
24
W. C. Hagel and A. V. Seybolt, Rept. No. 61-RL-2757M, General Electric
Research Laboratory, New York, 1961.
Fig. 9. HT-XRD patterns of coprecipitate powders (50 wt% Cr2O3) at (a) 25
A. E. Paladino and W. D. Kingery, “Al-Ion Diffusion in Al2O3,” J. Chem. Phys,
600°, (b) 800°, (c) 1000°, and (d) 1200°C; (*) ss-Cr2O3, (E) ss-Al2O3. 37 [5] 289 –93 (1962). M