Effect of milling on the formation of nanocrystalline x-Al 2O3 from gibbsite

Article abstract of DOI:10.1111/j.1551-2916.2010.03849.x

Gibbsite (FG) with mean particle diameter (d₅₀ = 3μm) was milled in an attrition mill for 12 and 24 h using alumina balls as grinding media and calcined at different temperatures in the range of 350°-600°C. The properties of the alumina obtained were determined by X-ray diffraction, N₂ physisorption, thermogravimetric/ differential thermal analyses, and transmission electron microscopy.Without milling, the alumina obtained normally contained the mixed phases between γ- and x-phase alumina. On the other hand, high purity of nanocrystalline vphase alumina (100 wt%) can be produced by calcination of the 24-h milled FG at 600°C. The isothermal kinetics measurements revealed that the rate constant (κ) for phase transformation increased as the particle size of gibbsite decreased and the calculated activation energy for transformation from FG to alumina decreased from 20.6 to 14.7 and 6.8 kJ/mol after milling for 12 and 24 h, respectively. The physical properties of nanocrystalline x-alumina obtained by the calcination of milled FG were comparable to those produced by the solvothermal method. The present results offer a simple way to prepare a large amount of pure x-phase alumina for particular industrial applications.

Full text of DOI:10.1111/j.1551-2916.2010.03849.x

J. Am. Ceram. Soc., 93 [10] 3377–3383 (2010)
DOI: 10.1111/j.1551-2916.2010.03849.x
r 2010 The American Ceramic Society
ournal
J
Effect of Milling on the Formation of Nanocrystalline v-Al2O3
from Gibbsite
Wasu Chaitree,z Sirithan Jiemsirilers,y Okorn Mekasuwandumrong,z Piyasan Praserthdam,z
Tawatchai Charinpanitkul,z and Joongjai Panpranotw,z
zDepartment of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok 10330, Thailand
yDepartment of Materials Science, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand
zDepartment of Chemical Engineering, Faculty of Engineering and Industrial Technology, Silpakorn University,
Nakorn Phathom 73000, Thailand
Gibbsite (FG) with mean particle diameter (d50 5 13 lm) was
milled in an attrition mill for 12 and 24 h using alumina balls as
grinding media and calcined at different temperatures in the
range of 3501–6001C. The properties of the alumina obtained
were determined by X-ray diffraction, N2 physisorption, the-
rmogravimetric/differential thermal analyses, and transmission
electron microscopy. Without milling, the alumina obtained nor-
mally contained the mixed phases between c- and v-phase
alumina. On the other hand, high purity of nanocrystalline v-
phase alumina (100 wt%) can be produced by calcination of the
24-h milled FG at 6001C. The isothermal kinetics measurements
revealed that the rate constant (k) for phase transformation in-
creased as the particle size of gibbsite decreased and the calcu-
lated activation energy for transformation from FG to alumina
decreased from 20.6 to 14.7 and 6.8 kJ/mol after milling for 12
and 24 h, respectively. The physical properties of nanocrystalline
v-alumina obtained by the calcination of milled FG were com-
parable to those produced by the solvothermal method. The
present results offer a simple way to prepare a large amount of
pure v-phase alumina for particular industrial applications.
line pure g-Al2O3, w-Al2O3, and mixed-phase g- and w-Al2O3 in
the dehydration reaction of methanol to dimethyl ether. Me-
ephoka et al.17 used both g- and w-Al2O3 as supports for the
preparation of Pt/Al2O3 for CO oxidation reaction. Most of the
recent studies show that highly stable nanocrystalline w-Al2O3
can be prepared by the solvothermal method using aluminum
iso-propoxide as the precursor. Such a technique, however, is
quite tedious and costly because it requires a high-pressure
reactor, high temperature, and long reaction time.
Gibbsite (a-Al(OH)3) is a cheap starting material widely used
in the transformation of alumina because it can be dehydrated
to various phases (w,g,y,k,a).18,19 Figure 1 shows the diagram of
phase transformation of alumina from gibbsite. The transfor-
mation route depended on temperature, heating environment,
particle size of staring gibbsite, and heating rate.20 Typically, for
small particles (o1 mm), the phase transformation process
occurs by route 1 and produces w-alumina.14 In addition, the
transformation of bayerite21 and diaspore22 to a-Al2O3 has been
studied. It was indicated that w-alumina was not formed from
these starting materials.
The formation of w-alumina from gibbsite has been observed
in previous works. Jang et al.12 studied the effect of grinding on
the phase transformation of gibbsite (starting particle size
64.2 mm). The w-phase alumina was observed at 4501C for the
sample grind for 5 h (particle size 9.7 mm) and completed at
around 9101C. Bhattacharya et al.20 investigated the thermal
decomposition of fine gibbsite (1.5 mm) and the formation of w-
alumina appeared at 5001C and its crystal remained until
around 8001C. Mercury et al.23 studied the decomposition of
synthetic gibbsite by neutron thermodiffractometry and found
that w-alumina was formed at the temperature above 5001C. The
grinding method is an effective method for the reduction of
particle size due to its simplicity, minimal environmental prob-
lems, convenient operation, and absence of wastes. Attrition
mills are widely used in order to reduce particle size; moreover,
they are ideally appropriate for the industrial process due to
their high efficiency and availability in large scale.24
I. Introduction
LUMINA (Al2O3) is one of the most common crystalline ma-
A
terials used as catalysts, catalyst supports, sorbent, coating,
and ceramics.1–4 Compared with the other oxides, alumina has
high surface area, good catalytic activity, high mechanical resis-
tance, good thermal stability, high strength, and toughness.5–7
There are many methods to synthesize alumina such as solvo-
thermal,5–7 molten salt synthesis,8 shock wave action,9 sol–gel,10
spray pyrolysis,11 and thermal decomposition of aluminum hy-
droxide (boehmite and gibbsite).12,13
w-alumina is a crystallographic form of series of alumina,
which is normally obtained by dehydration of gibbsite
(o200 nm).14 When it is fired, w-alumina will transform to k-
alumina at a temperature in the range of 6501–7501C and con-
sequently form a-alumina at a temperature around 10001C. The
w-phase alumina has been used as catalysts and catalyst supports
and interesting results were obtained.15–17 For examples, CO hy-
drogenation activities of Co/Al2O3 catalysts increased when the
catalysts were prepared on the mixed g- and w-phase Al2O3.15
Khom-in et al.16 studied the solvothermal-derived nanocrystal-
In the present study, the effect of milling on the phase trans-
formation behavior of gibbsite was extensively investigated. The
properties of the alumina obtained were investigated using
a
laser diffraction-based size analyzer, X-ray diffraction
(XRD), thermogravimetric/differential thermal analyses
(TGA/DTA), N2 physisorption, transmission electron micros-
copy (TEM), and isothermal kinetics measurement.
A. Krell—contributing editor
II. Experimental Procedure
(1) Sample Preparation
Manuscript No. 27288. Received December 23, 2009; approved April 8, 2010.
This work was financially supported by the Thailand Research Fund (TRF) and the
Office of Higher Education Commission, Ministry of Education, Thailand.
wAuthor to whom correspondence should be addressed. e-mail: joongjai.p@chula.ac.th
Gibbsite with d50 5 13.0 m (Merck, Darmstadt, Germany) was
used as a starting material. In order to reduce its particle size,
3377

3378
Journal of the American Ceramic Society—Chaitree et al.
lnðꢀ lnð1 ꢀ xÞÞ ¼ n ln k þ n ln t
Vol. 93, No. 10
(2)
Fig. 1. Phase transformation of alumina from gibbsite to a-alumina.
The value of k and n can be calculated from the slope and
interception of a linear plot of ln t and ln(-ln(1-x)). After the
value of k is obtained, the activation energy (Ea) can be esti-
mated by the Arrhenius equation (Eq. (3)).
the starting material was milled in an attrition mill for 12 and 24
h using alumina balls as grinding media and water as milling
fluid. Approximately 100 g of gibbsite and water 300 mL were
first mixed in a plastic pot. Alumina balls were weighed before
and after milling as well as filled to a half volume of the milling
pot. Cooling water was fed continuously in the attrition mill in
order to remove heat. Rotational speed of the mill was fixed at
500 rpm. The sample was collected every 2 h and dried at 1051C
overnight in an oven to remove water. The dried sample was
subsequently milled using a mortar to deagglomerate. The sam-
ples were calcined in a tube furnace in air (95 mL/min) by heat-
ing to a desired temperature at a rate of 101C/min and was held
at this temperature for 4 h. Then, the samples were cooled down
Ea
ln k ¼ ln A ꢀ
(3)
RT
where
A
is
a
constant,
R
is the gas constant (8.314
J ꢂ (Kꢂ mol)ꢀ1), and T is the working temperature. The value
of Ea can be calculated from the slope of a linear plot of ln k and
l/T.
III. Results and Discussion
(1) Particle Size Distribution
The size distributions of gibbsite agglomerates and particles
before and after milling are shown in Fig. 2. The median particle
size (d50) decreased from 13 to 3 mm and 0.6 mm after milling
for 12 and 24 h, respectively. Moreover, the unimodal pore
size distribution in the starting gibbsite changed to bimodal
distribution after milling for 24 h. During milling, the evolution
of the distribution of gibbsite is lower, and the mean size
decreased.24 Regarding the starting specific surface area of
25 m2/g and microscopic evidence (shown in Fig. 9), the 13
mm is referred to the agglomerate size of the raw powder,
whereas the particle size in the as-delivered state was ca. 2–3
mm. It is noted that particle size distribution of the samples did
not come from alumina balls because the weight loss of balls was
determined to be o1% (initial weight 761.9 g and after milling
759.9 g). Milling run of the alumina balls for 24 h in pure water
without gibbsite was carried out in order to determine the typ-
ical mean particle size of the wear debris (possible a seeds and
abrasives coming from the grinding media). The result is also
shown in Fig. 2 and the typical mean particle size was deter-
mined to be 24 m.
to room temperature in N2 flow (75 mL/min).
ÃÃ ÃÃÃ
The products are designed as FG
h
C where the first two
asterisks represent the milling time and the latter three asterisks
represent the temperature of calcination. The abbreviation of
FG is gibbsite (starting material).
(2) Characterization
Particle size distribution was measured using a laser diffraction-
based size analyzer (reflective index 1.57; Malvern Mastersizer,
Worcester Shire, England). All the samples were measured on
DI water as dispersant. Dispersion and deagglomeration of par-
ticles were ensured by ultrasonic treatment before measurement.
XRD analysis was carried out using a Siemens D5000 diffracto-
meter (Karlsruhe, Germany) with CuKa radiation with Ni filter.
The scans were recorded in the 2y range of 101–801 using a step
size of 0.041. The average crystallite size was estimated using the
Scherrer equation. The surface areas of samples were measured
by N2 physisorption using a micromeritics chemisorb 2750
(Norcross, GA). TGA/DTA of the ground and unground sam-
ples were performed (SDT Analyzer Model Q600 from TA In-
struments, New Castle, DE) at a heating rate of 101C/min in
flowing air (100 mL/min). The morphology of samples was in-
vestigated by TEM (JEOL JEM 2010, Tokyo, Japan), operating
at 200 kV.
The fraction of w-Al2O3 was determined by quantitative
XRD analysis, using CaF2 (Merck, Germany) as an internal
standard. A 0.200070.0001 g of sample was placed in a porce-
lain dish. One tenth of the sample weight of CaF2 was weighed
and mixed with the sample for 5 min. The analysis was calcu-
lated by the ratio of the integrated intensities of the 431 (CuKa
2y) of w-Al2O3 and integrated intensities of the 28.11 (CuKa 2y)
of CaF2. Their ratios were compared against a standard cali-
bration curve of w-Al2O3/CaF2. The calibration curve was
determined using a mixture of w-alumina and the same amount
of CaF2, spanning the range of 20–90 wt% w-alumina.
(2) XRD and TGA/DTA Results
Figures 3 and 4 show the results from the TGA and DTA plots,
of the gibbsite powder after milling for various times. Two
weight processes were detected corresponding to two endother-
mic and one exothermic process. The decrease in mass at around
1501–6001C, accompanied by the endothermic peaks in DTA
signal, is attributed to the dehydration of gibbsite to form
(3) Isothermal Kinetic Measurements
The samples were heated at three different temperatures in the
range of 4501–6001C with different holding times. The activation
energy for the transformation of gibbsite to w-Al2O3 was esti-
mated by isothermal experiments according to the Arrhenius
method.25 First of all, the value of the rate constant (k) was
calculated by the Johnsom–Mehl–Avrami equation (Eq. (1)).25
n
xðtÞ ¼ 1 ꢀ exp½ꢀðktÞ ꢁ
(1)
where x is the phase fraction of w-Al2O3 at time (t) and n is the
reaction exponent. For analyzing Eq. (1), it can be rewritten in
the form of the following Eq. (2).
Fig. 2. The particle size distribution of unmilled and milled gibbsite at
various milling times.

October 2010
Formation of w-Al2O3 from Gibbsite
3379
Fig. 3. The thermogravimetric curves of unmilled and milled gibbsite at
various milling times.
Fig. 5. The X-ray diffraction patterns of unmilled and milled gibbsite at
various milling times; (a) FG0h, (b) FG12h, and (c) FG24h.
alumina. The endothermic peak was shifted toward the lower
temperature with an increasing milling time, indicating an ac-
celeration of the dehydration of gibbsite by the milling process.21
The overall weight losses of all the samples were approximately
around 35%, which were in good agreement with the calculated
value of 34.6% for the dehydration reaction
terns of calcined FG0h, FG12h, and FG24h at various temper-
atures. For the calcined FG12h, the mixture between boehmite
and w-phases was found at 4001C and the w-phase was observed
at a higher calcination temperature. In the case of prolonged
milling (FG24h), the transformation of gibbsite to w-alumina
was completed at 3501C. This implies that the small particle sizes
are effective for reduction in the transformation temperature of
gibbsite to w-alumina.26
2AlðOHÞ3 ! Al2O3 þ 3H2O
(4)
When particle size of gibbsite is small, the dominant peak
(2y 5 14.41) of boehmite would be reduced (Figs. 6(d) and 7(d)).
As a result, w-alumina occurred from the small particle size of
gibbsite. It has been known that the dehydration sequence of
gibbsite in air is affected by its particle size. In small gibbsite
particles (o10 mm), boehmite is rarely formed, but in the case of
larger particles (4100 mm), boehmite is formed because the wa-
ter formed by the decomposition of the gibbsite cannot rapidly
escape from the larger particles.2,12,20,27 Consequently, transfor-
mation of gibbsite occurred through route 1 (Fig. 1). However,
the boehmite phase may be observed from the small particle size
of gibbsite, if the gibbsite particle is a crystal.20 Bhattacharya
et al.20 found that crystal gibbsite (0.25 mm) can produce boeh-
mite at 4001C. Nevertheless, Jang et al.12 used ground gibbsite
losing the crystalline structure, and boehmite was not observed
at low temperature. When FG0h was heated, boehmite was
observed at 4001C and w-alumina was formed at 4501C. How-
ever, w-alumina was not the only phase of alumina that occurred
The dehydration of FG0h (Fig. 4(a)) in first endothermic step
(2361C) was due to loss of water in the sample. The second en-
dothermic peak (3061C) is corresponding to a weight loss of
22.5%, XRD measurements of FG0h treated at 4001C also
show a peak at 2y 5 14.41 (Fig. 6(d)). It points out that boehmite
formed at this range of temperature. In addition, the endother-
mic peak at around 5261C indicated that boehmite transforms to
alumina. Figure 6(b) shows that diffraction peaks of mixed
phases of alumina were observed at 5001C. However, this en-
dothermic peak disappeared when milling time increased, indi-
cating that the milling conditions affected the transformation
sequence of gibbsite.12
XRD data were undertaken to study the transformation
phases from gibbsite at various temperatures. The XRD pat-
terns of FG0h, FG12h, and FG24h are shown in Fig. 5. The
intensities of XRD peaks decreased and became wider as the
milling time increased. It is due to decreasing of the crystallinity
and particle size of gibbsite. Figures 6–8 shows the XRD pat-
Fig. 6. The X-ray diffraction patterns of FG0h calcined at various
temperatures, b, boehmite; x, w-Al2O3; o, g-Al2O3 (a) 6001C, (b) 5001C,
(c) 4501C, and (d) 4001C.
Fig. 4. The differential thermal analyses curves of unmilled and milled
gibbsite at various milling times.

3380
Journal of the American Ceramic Society—Chaitree et al.
Vol. 93, No. 10
Table I. The BET Surface Area and Alumina Phase of
Unmilled and Milled Samples after Calcination at 6001C
Samples
Phase
Surface area (m2/g)
FG0h
—
g and w
w
w
25.1
101.8
110.5
122.6
FG0h600C
FG12h600C
FG24h600C
high-purity Al2O3 balls when the seed concentration was in-
creased to 410 wt%. Moreover, the average particle size of the
abrasives in this study was much larger than those reported in
the literature for reduction in the transformation temperature
of gibbsite powder (i.e., 24 mm (this work) as compared with
0.6 mm in Xie et al.30).
(3) BET Surface Area
Fig. 7. The X-ray diffraction patterns of FG12h calcined at various
temperatures, b, boehmite; x, w-Al2O3 (a) 6001C, (b) 5001C, (c) 4501C,
and (d) 4001C.
Based on BET surface analyses of calcined samples summarized
in Table I, the FG0h had significantly smaller surface area. After
calcinations, the structure of gibbsite was destroyed, resulting in
a drastic increase in surface area of the calcined samples. With
the same calcination temperature, FG24h600C had the highest
surface area of 122.6 m2/g. The increase in the surface area of
treated samples is due to the decrease in particle size. These
results are in good agreement with that of Ogata et al.,19 who used
gibbsite as the starting material and found that the specific surface
area increased when the calcination temperature increased.
Table II summarizes the crystallite size (dXRD) and the frac-
tion of w-phase of the calcined samples. It can be seen that the
increase in calcination temperature resulted in a larger crystallite
size. Meanwhile, the fraction of w-alumina increased with the
increasing calcination temperature. These results indicated that
high-purity w-alumina can form at the higher temperature. Du
et al.21 also reported that the fraction of a-phase in a-alumina
synthesized from bayerite would increase with increasing calci-
at this temperature; the peak of g-alumina was also observed.
For the micrometer sizes, both transition routes (Fig. 1) may
possibly occur, which depends on the heat treatment condi-
tions.28 The decomposition of boehmite would occur at 3001C at
atmospheric pressure. The typical temperature reported for the
decomposition of boehmite to g-Al2O3 is 4501C.29
In the past, a-seeding has been shown to influence phase
development and transformation kinetics of alumina.26,30 It may
be possible that the g peaks in Figs. 6–8 could have been en-
hanced by such a seeding effect. However, in the present study,
the contamination from abrasives appeared to have very little
impact on the reduction of transformation temperature of gibbs-
ite to w-alumina due to the very small amount presented. The
amounts of wear debris obtained by milling the alumina balls in
water for 24 h with and without gibbsite were only 2 and 3 wt%
of the gibbsite weight, respectively. According to the literature,
the effect of a-seeding was dependent not only on the size of the
seeds but also on the amount of the seeds presented. For exam-
ples, Kano et al.26 showed that the presence of 50 wt% a-
alumina seeding reduced the transformation temperature of
gibbsite to a-alumina by 1201C (from 10301 to 9101C), while
for 5 wt% a-alumina seeding, the transformation temperature
decreased only about 301C. Similarly, Xie et al.30 reported that
Al(OH)3 can be transformed to a-Al2O3 at a relatively lower
temperature (1100oC) with the help of seeds by wet grinding of
nation temperature. Maceˆ
do et al.25 found that the high fraction
of a-alumina can be produced from g-alumina when the calci-
nation temperature increased from 7501 to 9001C. At low tem-
perature (4501C), the fraction of w-alumina rapidly increased
from 0.58 to 0.89. It is implied that mechanical activation
affected the formation of w-alumina. Under isothermal condi-
tion, the fraction of w-alumina increased when the particle size
decreased. It is confirmed that the small particle size of gibbsite
can produce high purity of w-alumina. In comparison with pure
w-alumina prepared by the solvothermal method as reported
previously by our group (crystallite size 6 nm and BET surface
area 168 m2/g),17 the physical properties of w-alumina obtained
by the calcination of FG24h were comparable to those of the
solvothermally derived ones.
(4) Kinetics Measurements
Table III summarizes the transformation rate constants (k) of
unmilled and milled gibbsite, which was calcined at different
Table II. The Particle Size (dXRD) of the Samples after
Calcination and the Fraction of w-Phase
Samples
T (1C)
dXRD (nm)w
Fraction of w-phasez
FG0h
450
500
600
450
500
600
450
500
600
3.8
4.1
4.4
3.7
3.8
4.2
3.3
3.7
3.8
0.58
0.72
0.88
0.68
0.94
0.95
0.90
0.99
1
FG12h
FG24h
Fig. 8. The X-ray diffraction patterns of FG24h calcined at various
temperatures, b, boehmite; x, w-Al2O3 (a) 6001C, (b) 5001C, (c) 4001C,
(d) 3501C, and (e) 3001C.
wCalculated by the Scherrer equation. zCalculated by quantitative XRD.

October 2010
Formation of w-Al2O3 from Gibbsite
3381
Table III. Rate Constant (k) for Milled and Unmilled Gibbsite
at Different Temperature
Table IV. Activation Energy of Milled and Unmilled Gibbsite
Sample
Ea (kJ/mol)
Milling time (h)
k (minꢀ1
)
FG0h
FG12h
FG24h
20.6
14.7
6.8
T 5 6001C
0
12
24
T 5 5001C
0.0142
0.0178
0.0359
0
12
0.0110
0.0153
0.0248
Table IV shows the activation energy for phase transforma-
tion of unmilled and milled gibbsite to w-alumina. The activa-
tion energy decreased from 20.6 to 14.7 and 6.8 kJ/mol after
milling for 12 and 24 h, respectively. This result indicated that
the activation energy decreased with the decreasing particle size
of the starting gibbsite. In general, reducing the particle size
would lead to an increasing surface energy of the particle,
thereby resulting in the decreasing of activation energy for phase
transformation and transformation temperature as shown in the
DTA profile.28 The XRD patterns also confirmed that the trans-
formation temperature of gibbsite to w-alumina reduced from
4501 to 3501C after milling for 24 h. It may be attributed to the
fact that gibbsite grows to the critical size of phase transforma-
tion and then transform to w-alumina at a lower tempera-
ture.28,31 Moreover, the milling of gibbsite for 12 and 24 h can
be up to 28% and 67% of the activation energy compared with
that of the unmilled gibbsite. Several researchers observed the
24
T 5 4501C
0
12
24
0.0033
0.0054
0.0188
temperatures. For any constant temperature, k increased with
the increasing milling time. At a low temperature, the milling
can increase k more significantly. For example, at 4501C acti-
vation of the sample, milling for 24 h increased the reaction rate
constant 5.7 times (compared with the sample for 0 h), but at
6001C it increased only 2.5 times. Panchula and Ying31 synthe-
sizing a-alumina from milled and unmilled g-alumina reported
that k increased with the increase in calcination temperature and
milling time from 30 to 120 min.
kinetics of transition alumina. Maceˆ
do et al.25 studied the ki-
Fig. 9. The transmission electron micrographs of FG0h (a, b), FG12h (c), and FG24h (d).

3382
Journal of the American Ceramic Society—Chaitree et al.
Vol. 93, No. 10
Fig. 10. The transmission electron micrographs of FG0h400C (a), FG0h600C (b), FG12h400C (c), FG12h600C (d), and FG24h600C (e).
netics of g- to a-alumina. They found that the activation energy
of this phase transformation was 20174 kJ/mol. Chang et al.28
studied the size effect of w-alumina to a-alumina. The activation
energy reduced from 506 to 321 kJ/mol when d50 of w-alumina
decreased from 155 to 40, respectively. Yang et al.32 presented
the formation during y-Al2O3 to a-Al2O3 transformation. They
used three kinds of y-powder (as-received, homogenized,
homogenized, and additionally unaxial-pressed compact) as
the starting material. The activation energy was 299, 189, and
148 kJ/mol, respectively. Candela and Perlmutter33 studied the
kinetics of boehmite formation by the thermal decomposition
of gibbsite under water vapor pressure from 100 to 3200 Pa. The

October 2010
Formation of w-Al2O3 from Gibbsite
3383
7O. Mekasuwandumrong, P. Tantichuwet, C. Chaisuk, and P. Praserthdam,
‘‘Impact of Concentration and Si Doping on the Properties and Phase Transfor-
mation Behavior of Nocrystalline Alumina Prepared Via Solvothermal Synthesis,’’
Mater. Chem. Phys, 107, 208–14 (2008).
activation energy was 142710 kJ/mol. In the present work,
the value of activation energy was lower than those reported in
the above-mentioned works due to the lower transformation
temperature.
8J. He, W. Liu, L.-H. Zhu, and Q.-W. Huang, ‘‘Phase Transformation Behav-
iors of Aluminum Hydroxides to Alpha Alumina in Air and Molten Salt,’’
J. Mater. Sci., 40, 3259–60 (2005).
9A. N. Tsvigunov, V. G. Khotin, A. S. Krasikov, A. S. Valasov, and B. S.
Svetlov, ‘‘Synthesis of New Modification of Aluminium Oxide with the
Spinel Structure Under Shock-Wave Action on Gibbsite,’’ Glass Ceram., 56, 7–8
(1999).
(5) TEM Observation
The effect of grinding on the morphology of gibbsite and
w-Al2O3 were studied by TEM observation. Figure 9 shows
the TEM micrographs of unmilled (a and b) and milled (c and d)
gibbsite powder. The unmilled fine gibbsite clearly consisted
of pseudohexagonal plates, while irregular and flaky particles
were observed for milled samples. It revealed that the milling
effectively changed the morphology of gibbsite and reduced
particle size. Figures 10(a) and (b) show the TEM micrographs
of FG0h400C and FG0h600C, respectively. Figure 10(a) shows
that the structure is a strip. It was due to the formation of
boehmite. However, this structure disappeared when gibbsite
was milled for 12 h and calcined at the same temperature
(Fig. 10(c)). It was probably due to the decreasing of boehmite
phase in calcined samples. Figures 10(b), (d), and (e) exhibited
the TEM micrographs of FG0h600C, FG12h600C, and
FG24h600C, respectively. Figures 10(d) and (e) indicated that
the dispersed w-alumina with narrow size distribution was ob-
tained from milled gibbsite, whereas the larger particle of mixed
phase (w, g) was obtained from unmilled one (Fig. 10(b)).
Besides, the morphology of samples is similar to that of gibbs-
ite treated by mechanical milling.
10A. L. Ahmad and N. N. N. Muatafa, ‘‘Sol-Gel Synthesized of Nanocomposite
Palladium–Alumina Ceramic Membrane for H2 Permeability: Preparation and
Characterization,’’ Int. J. Hydrogen Energy, 32, 2010–21 (2007).
11J. H. Kim, K. Y. Jung, K. Y. Park, and S. B. Cho, ‘‘Characterization of
Mesoporous Alumina Particle Prepared by Spray Pyrolysis of Al(NO3)2.9H2O
Precursor: Effect of CTAB and Urea,’’ Micropor. Mesopor. Mater., 128, 85–90
(2010).
12S.-W. Jang, H.-Y. Lee, S.-M. Lee, S. W. Lee, and K.-B. Shim, ‘‘Mechanical
Activation Effect on the Transition of Gibbsite to a-Alumina,’’ J. Mater. Sci.
Lett., 19, 507–10 (2000).
13A. Boumaza, L. Favaro, J. Le
A. M. Huntz, P. Roy, and R. Te
´
tot, ‘‘Transition Alumina Phases Induced by Heat
dion, G. Attonnay, J. B. Brubach, P. Berthet,
´
Treatment of Boehmite: An X-Ray Diffraction and Infrared Spectroscopy Study,’’
J. Solid State Chem., 182, 1171–6 (2009).
14F. G. W. Brindley and J. O. Choe, ‘‘The Reaction Series, Gibbsite-chi
Alumina-kappa Alumina-Corundum,’’ J. Mineral. Soc. Am., 46, 771–85
(1961).
15K. Pansanga, J. Panpranot, O. Mekasuwandumrong, C. Satayaprasert, J. G.
Goodwin, and P. Praserthdam, ‘‘Effect of Mixed g- and w-Crystalline Phases in
Nanocrystalline Al2O3 on the Dispersion of Cobalt on Al2O3,’’ Catal. Commun., 9,
207–12 (2008).
16J. Khom-in, P. Praserthdam, J. Panpranot, and O. Mekasuwandumrong,
‘‘Dehydration of Methanol to Dimethyl Ether over Nocrystalline Al2O3 with
Mixed w- and g- Crystallite Phases,’’ Catal. Commun., 9, 1955–8 (2008).
17C. Meephoka, C. Chaisuk, P. Samparnpiboon, and P. Praserthdam, ‘‘Effect of
Phase Composition Between Nano g- and w-Al2O3 on Pd/Al2O3 Catalyst in CO
Oxidation,’’ Catal. Commun., 9, 546–50 (2008).
IV. Conclusions
18P. Souza Santos, H. Souza Santos, and S. P. Toledo, ‘‘Standard Transition
Aluminas. Electron Microscopy Studies,’’ Mater. Res., 3, 104–14 (2000).
19F. Ogata, N. Kawasaki, M. Kabayama, T. Nakamura, and S. Tanada,
‘‘Structure Transformation of Gibbsite by Calcinations,’’ J. Surf. Sci. Nanotech.,
4, 267–9 (2006).
High-purity nanocrystalline w-alumina can be produced from
milled gibbsite (FG12h and FG24h). The transformation tem-
perature of gibbsite to w-alumina decreased from 4501 to 3501C
when the milling time of gibbsite was increased from 12 to 24 h,
respectively. For the unmilled gibbsite (FG0h), the mixed w and
g-phase Al2O3 were formed at 4501C. The fraction of w-alumina
increased with the increase in the milling time and calcination
temperature. The activation energy for phase transformation of
gibbsite to w-Al2O3 also decreased with the reducing particle size
of the starting gibbsite.
20I. N. Bhattacharya, S. C. Das, P. S. Mukherjee, S. Paul, and P. K. Mitra,
‘‘Thermal Decomposition of Precipitated Fine Aluminium Trihydroxide,’’ Scand.
J. Metall., 33, 211–9 (2004).
21X. Du, X. Su, Y. Wang, and J. Li, ‘‘Thermal Decomposition of Grinding
Activated Bayerite,’’ Mater. Res. Bull., 44, 660–5 (2009).
22L. Lo
¨
ffler and W. Mader, ‘‘Transformation Mechanism of the Dehydration of
Diaspore,’’ J. Am. Ceram. Soc., 86 [4] 534–40 (2003).
23J. M. R. Mercury, P. Pena, and A. H. de Aza, ‘‘On the Decomposition of
SyntheticGibbsite Studied by Neutron Thermodiffractometry,’’ J. Am. Ceram.
Soc., 89 [12] 3728–33 (2006).
24T. C. Alex, R. Kumar, S. K. Roy, and S. P. Mehrotra, ‘‘Stirred Bead Mill
Grinding of Gibbsite: Surface and Morphological Change,’’ Adv. Powder Technol.,
19, 483–391 (2008).
Acknowledgments
The authors also would like to thank the Research Unit of Advanced Ceramic
and Polymeric Materials, National Center of Excellence for Petroleum, Petro-
chemicals and Advanced Materials, Chulalongkorn University, Bangkok, Thai-
land for providing the attrition mill.
25M. I. F. Macedo, C. A. Bertran, and C. C. Osawa, ‘‘Kinetic of the gsa-Al-
ˆ
umina Phase Transformation by Quantitative X-Ray Diffraction,’’ J. Mater. Sci.,
42, 2830–6 (2007).
26J. Kano, S. Saeki, F. Saito, M. Tanjo, and S. Yamazaki, ‘‘Application of Dry
Grinding to Reduction in Transformation Temperature of Aluminum Hydrox-
ides,’’ Int. J. Miner. Process., 60, 91–100 (2000).
References
27C. Misra, Industrial Alumina Chemicals. American Chemical Society, Wash-
ington, DC, 1986 75pp..
1B. P. Zolotovskii, O. P. Krivoruchko, R. A. Buyanov, T. A. Kriger, and G. N.
Krjukova, ‘‘Study of the Effect of the Size of Hydrargillite Crystals on the Pecu-
liarities of its Thermal Decomposition and Formation of the Phase Composition
of Al (III) Oxide,’’ React. Kinet. Catal. Lett., 22, 153–7 (1983).
28P.-L. Chang, Y.-C. Wu, S.-J Lai, and F. S. Yen, ‘‘Size Effects on the w-to
a-Al2O3 Phase Transformation,’’ J. Eur. Ceram. Soc., 29, 3341–8 (2009).
29S. Serena, M. A. Raso, M. A. Rodrı
´
guez, A. Caballero, and T. J. Leo, ‘‘Ther-
2W. H. Gitzen, Alumina as a ceramic material, pp. 17–8. The American Ceramic
Society, Westerville, OH, 1970.
modynamic Evaluation of the Al2O3–H2O Binary System at Pressures Up to
30 MPa,’’ Ceram. Int., 35, 3081–90 (2009).
30Z.-P. Xie, J.-W. Lu, Y. Huang, and Y.-B. Cheng, ‘‘Influence of a-Alumina
Seed on Morphology of Grain Growth in Alumina Ceramics from Bayer Alumi-
num Hydroxide,’’ Mater. Lett., 57, 2501–8 (2003).
3M. Lacroix, G. M. Pajonk, and S. J. Teichner, ‘‘Silica and Alumina Catalysts,
Activated by Hydrogen Spillover, for the Conversion of n-Heptane at 2701C,’’
J. Catal., 101, 314–22 (1986).
31M. L. Panchula and J. Y. Ying, ‘‘Mechanical Synthesis of Nanocrytalline
a-Al2O3 Seeds for Enhanced Transformation Kinetics,’’ Nanostruct. Mater., 9,
161–4 (1997).
4J. S. Church, N. W. Cant, and D. L. Trimm, ‘‘Stabilisation of Aluminas by
Rare Earth and Alkaline Earth Ions,’’ Appl. Catal., 101, 105–16 (1993).
5O. Mekasuwandumrong, H. Kominami, P. Praserthdam, and M. Inoue, ‘‘Syn-
thesis of Thermally Stable w-Alumina by Thermal Decomposition of Aluminum
Isopropoxide in Toluene,’’ J. Am. Ceram. Soc, 87, 1543–9 (2004).
6O. Mekasuwandumrong, V. Pavarajarn, M. Inoue, and P. Praserthdam, ‘‘Prep-
aration and Phase Transformation Behavior of w-Alumina Via Solvothermal Syn-
thesis,’’ Mater. Chem. Phys, 100, 445–50 (2006).
32R.-J. Yang, F. S. Yen, S.-M. Lin, and C. C. Chen, ‘‘Microstructure-Controlled
Effects on Temperature Reduction of a-Al2O3 Crystallite Formation,’’ J. Cryst.
Growth, 299, 429–45 (2007).
33L. Candela and D. D. Perlmutter, ‘‘Kinetics of Boehmite Formation by Ther-
mal Decomposition of Gibbsite,’’ Ind. Eng. Chem. Res., 31, 694–700 (1992).
&