J. Am. Ceram. Soc., 87 [7] 1280 –1286 (2004)
journal
Preparation of Micrometer- to Sub-micrometer-Sized Nanostructured
Silica Particles Using High-Energy Ball Milling
J. Eric Hampsey, Claudio L. De Castro, Byron McCaughey, Donghai Wang, Brian S. Mitchell,*
and Yunfeng Lu
Department of Chemical and Biomolecular Engineering, Tulane University, New Orleans, Louisiana
Nanostructured porous silica particles with sizes in the micrometer to sub-micrometer range are of great interest due to
their potential applications as catalyst supports and nanocomposite materials. However, if these particles are to be used in
industry, a process must be developed to affordably produce
them on a large scale. This paper reports on a high-energy
ball-milling process that has been used to create micrometerto sub-micrometer-sized mesoporous silica particles starting
from a silica xerogel prepared by a surfactant self-assembly
sol– gel process. We have studied various milling conditions
such as milling media (zirconia, stainless steel, or steelcentered nylon balls), milling time, and the presence of surfactants during milling and the resulting effect on particle size
and pore structure. Results from transmission electron microscopy, scanning electron microscopy, X-ray diffraction, light
scattering, and nitrogen adsorption demonstrate the feasibility
of producing large quantities of nanostructured particles by
this simple milling process.
I.
70118
also been produced using an aerosol-assisted self-assembly process.16,17 However, these processes usually require large amounts
of solvent and produce relatively small amounts of material.
High-energy ball milling has been a common physical technique used to produce sub-micrometer particles without the need
for excessive solvent. During the high-energy milling process, the
grinding media are accelerated to much larger velocities than those
achieved in traditional ball milling. These large velocities provide
a high transfer of kinetic energy from the balls to the sample
resulting in the production of very fine powders. High-energy ball
milling is usually used in three main applications: (1) mechanical
alloying, (2) reactive milling, and (3) particle deformation, the
latter of which will be the focus of this paper. Zoz and Ren used
yttria-stabilized ZrO2 balls to mill SiC powder with a starting
particle size of 850 ␮m.18 After 10 min of milling, the particle size
was reduced to ⬍15 ␮m, and after 60 min the particle size was
further reduced to ⬍10 ␮m with some particles in the 1–2-␮m
range. Quartz and Al2O3 powders were also milled and particle
sizes of ⬍1 and ⬍5 ␮m, respectively, were achieved after 2 h of
milling.
In this study, we use a high-energy ball-milling process and the
coassembly of silica and surfactant to produce mesoporous silica
particles. This simple process allows us to produce large
quantities of mesoporous silica particles without using excessive solvent, making this process suitable for scale-up and industrial applications.
Introduction
its discovery by Mobile researchers in 1992,1,2 MCM-41
and other ordered mesoporous materials have attracted great
interest. These mesoporous materials are usually prepared by the
cooperative self-assembly of surfactants and inorganic species
such as silicates. Removal of the surfactant through either calcination or solvent extraction results in mesoporous materials with
hexagonal, cubic, lamellar, disordered, or other pore channels. The
size and structure of these pores can be controlled by careful
choice of the surfactant system and reaction conditions. Among
these materials, mesoporous silica synthesized in the form of thin
films,3 aerogels,4 and particles are the most cited materials.
Synthesis of mesoporous silica particles is of particular interest
due to their possible use as catalysts, fillers, chromatographic
materials and other applications. Mesoporous silica particles are
typically synthesized in dilute solutions, and depending on the
reaction conditions such as pH,5 stirring rate,6 choice of surfactant,
or presence of cosolvents,7 various particle shapes and morphologies can be achieved. Mesoporous silica hollow rods,8 hard
spheres,9,10 hollow spheres,11,12 and other curved shapes such as
toroidal, disklike, spiral, and gyroid morphologies13–15 have been
reported in the literature. Mesostructured spherical particles have
S
INCE
II.
Experimental Procedure
(1) Chemicals
All commercial chemicals and surfactants were used asreceived. The surfactants chosen for this study were cetyltrimethylammonium
bromide
(CTAB;
Aldrich),
Brij-58
(C16H33(OCH2CH2)20OH; Aldrich), P123 (EO20PO70EO20;
BASF), and F127 (EO106PO70EO106; BASF). EO and PO are used
to designate ethylene oxide and propylene oxide, respectively.
Tetraethyl orthosilicate (TEOS; Aldrich), ethanol (Aldrich), HCl
(Aldrich), hexamethyldisilazane (HMDS; Aldrich), and deionized
water were also used to prepare the xerogels.
(2) Preparation of Xerogels
Ordered mesoporous silica xerogels were prepared using a
sol– gel technique. A typical sol was prepared using the molar
ratios TEOS:EtOH:water:HCl:CTAB ⫽ 1:3.8:0.15:5:0.01. The
mixtures were sonicated for 30 min and then placed in Petri dishes
in thin layers to dry in an ambient condition for several days. The
uncalcined xerogels were heated before the milling process under
nitrogen at 110°C for 2 h. Mesoporous silica was prepared by
removing the surfactant by either calcination in air at 450°C for 1 h
or by a solvent extraction technique. The solvent extraction
process was performed using a mixed solvent that contained 80%
ethanol and 20% HMDS. The mixtures were sonicated for 1 h and
then centrifuged to separate the solid. This procedure was repeated
five times to ensure complete removal of the surfactant.
C. Kurkjian—contributing editor
Manuscript No. 10019. Received March 6, 2003; approved January 21, 2004.
This work is financially supported by NASA (Grant No. NAG-1-02070 and
NCC-3-946), Office of Naval Research, Louisiana Board of Regents (Grant No.
LEQSF(2001-04)-RD-B-09) and National Science Foundation (Grant No. NSFDMR-0124765 and CAREER award), and the Advanced Materials Research Institute
at the University of New Orleans through DARPA Grant No. MDA972-97-1-0003
and through the Louisiana Board of Regents Contract No. NSF/LEQSF(2001-04)RII-03.
*Member, American Ceramic Society.
1280
July 2004
Preparation of Nanostructured Silica Particles
1281
(3) Milling of the Xerogels
All samples were milled with an 8000D SPEX CentriPrep Dual
Mixer/Mill. The unit was operated in a Fisher Scientific Isotemp
refrigeration unit maintained at 0°C. The grinding media used
were 13-mm (1⁄2-in.) diameter stainless steel, zirconia, or nyloncoated stainless steel balls. The vials used for milling were made
of either stainless steel or nylon, depending on the media. Milling
times varied from 10 to 100 min, a ball to sample weight ratio of
10:1 was used, and the xerogels were milled without adding a
dispersant such as ethanol.
(4) Analysis of the Milled Silica Particles
A JEOL JSM-5410 scanning electron microscope (SEM) was
used to examine the size and morphology of the particles.
Transmission electron microscope (TEM) images were taken using
a JEOL JSM-2010 to show the ordered pore structures. A Phillips
Xpert X-ray diffractometer was used to obtain low-angle X-ray
diffraction (XRD) patterns while the particle sizes were measured
using a Brookhaven 90 Plus particle size analyzer. Nitrogen
adsorption/desorption isotherms were obtained using a Micromeritics ASAP 2010, and the BET surface areas of the particles were
calculated from the isotherms using the ASAP 2010 v.5 software.
III.
Results and Discussion
Fig. 2. Small-angle X-ray diffraction patterns of CTAB-templated silica
xerogels milled using nylon-coated stainless steel media after the surfactant
was removed by calcination. The top curve is the starting xerogel before
milling, while the curves below it were milled at 10, 30, and 60 min,
respectively.
(1) Preserving the Pore Structure during Milling
Calcined silica xerogels containing ordered pore structures were
first milled using zirconia, stainless steel, and steel-centered nylon
balls at different time intervals. While sub-micrometer-sized particles were achieved using this process, many of the features of the
ordered xerogels were greatly reduced after milling. Figure 1
shows the nitrogen adsorption/desorption isotherms of CTABtemplated silica xerogels milled with nylon-coated stainless steel
media at different milling times. At time zero, the isotherm is a
type IV, consistent with mesoporous materials with high surface
areas. After 10 min of milling, the isotherm still has the characteristics of type IV, but the volume adsorbed has decreased. As the
milling time increases to 30 min and beyond, the isotherms no
longer exhibit type IV behavior and the volume adsorbed decreases rapidly. The low-angle XRD patterns of the samples are
also consistent with these findings. As shown in Fig. 2, at time zero
the xerogel exhibits a sharp (100) peak at 2␪ ⬃ 2.4° and a second
broader (200) peak at 2␪ ⬃ 4.5°. After 10 min of milling, the peaks
are still visible but less intense. However, after 30 min of milling,
the broader (200) peak has disappeared and the first (100) peak has
become less intense. Finally, when the sample is milled for 60 min,
both peaks virtually disappear.
To reinforce and preserve the pore structures of the xerogels
during milling, samples were milled with the surfactant still inside
the pore framework of the xerogel. The effect on the pore structure
was quite remarkable. Figure 3 shows the nitrogen adsorption/
desorption isotherms of calcined CTAB-templated silica xerogels
milled with nylon-coated stainless steel balls with the surfactant
Fig. 1. Nitrogen adsorption/desorption isotherms of CTAB-templated
silica xerogels milled using nylon-coated stainless steel media after the
surfactant was removed by calcination. The top curve is the starting xerogel
before milling, while the curves below it were milled at 10, 30, and 60 min,
respectively.
Fig. 3. Nitrogen adsorption/desorption isotherms of CTAB-templated
silica xerogels milled using nylon-coated stainless steel media before the
surfactant was removed by calcination. The top curve is the starting xerogel
before milling, while the curves below it were milled at 10, 30, 60, and 100
min, respectively.
1282
Journal of the American Ceramic Society—Hampsey et al.
inside the pores during milling. As before, at time zero the xerogel
exhibits type IV behavior similar to that of mesoporous materials.
However, unlike the precalcined xerogels (without surfactant in
the pore framework during the milling process), the type IV
nitrogen adsorption behavior is maintained throughout the milling
process and the volume adsorbed decreases only slightly up to 100
min of milling time. Another important feature of the isotherms is
the increase in the volume adsorbed near a relative pressure of 1.
This corresponds to interparticle nitrogen condensation and indicates that smaller particles are being produced. The low-angle
XRD data are also consistent with this result. As shown in Fig. 4,
a sharp peak at 2␪ ⬃ 2.4° and a broader secondary peak at 2␪ ⬃
4.5° are visible even after milling for 60 min. Only after a milling
time of 100 min are these features absent.
To further illustrate this marked difference between the two
types of milled xerogels, a comparison of the BET surface area as
a function of milling time is displayed in Fig. 5. As shown in the
figure, the surface areas of the xerogels milled after the surfactant
was removed decreases rapidly as the milling time increases.
However, the xerogels milled with the surfactant still inside the
pore framework show a much smaller decrease in surface area.
This result suggests that the surfactant reinforces the pore walls of
the xerogel and prevents the collapsing of the pores during the
milling process. Another possible factor is that the silica framework of the precalcined milled samples is more brittle than the
uncalcined milled samples due to further silica condensation and
cross-linking during the calcination process. Whatever the case
may be, it was determined that all samples should be milled with
the surfactant still incorporated in the pore network to preserve the
mesopore structure.
(2) Characterization of the Milled Silica Particles
TEM micrographs were taken of the milled samples to confirm
that the ordered pore structures were maintained throughout the
milling process. Figure 6 shows representative TEM images of
CTAB-templated silica xerogels (a) before milling and after (b) 10,
(c) 30, and (d) 60 min of milling with nylon-coated stainless steel
balls. As shown in the figure, the ordered pore structures are
maintained in the silica even after milling for 30 min. However,
Fig. 4. Small-angle X-ray diffraction patterns of CTAB-templated silica
xerogels milled using nylon-coated stainless steel media before the
surfactant was removed by calcination. The top curve is the starting xerogel
before milling, while the curves below it were milled at 10, 30, 60, and 100
min, respectively.
Vol. 87, No. 7
Fig. 5. BET surface areas of calcined CTAB-templated silica xerogels as
a function of milling time. The top curve was milled with the surfactant still
inside the framework of the xerogel, while the bottom curve was milled
after the surfactant was removed by calcination.
after 60 min of milling, the ordered pore structures become less
apparent.
After analyzing the milled particles under a SEM microscope,
very few differences in particle size and agglomeration could be
determined between different milling media. Figure 7 shows
representative SEM images of CTAB-templated silica xerogels
after milling with nylon-coated stainless steel balls for (a) 0, (b)
10, (c) 30, (d) 60, and (e) 100 min. After 10 min of milling, the
particle sizes are greatly reduced from ⬎100 to ⬍5 ␮m. As the
milling time increases to 30 min and above, it appears that greater
amounts of sub-micrometer particles are being produced. However, it is difficult to quantify the particle sizes due to the large amount
of agglomeration of the particles during the milling process.
To determine more accurately the particle sizes of the milled
xerogels, light scattering measurements were performed. Figure 8
displays the effective diameter and polydispersity of the CTABtemplated silica xerogels after milling with nylon-coated stainless
steel for 10, 30, 60, and 100 min. As shown in the figure, the
effective diameter decreases from ⬃900 nm after 10 min of
milling to less than 650 nm after 100 min of milling. It is important
to note that the particle size distribution was quite large with
particle sizes ranging from 100 nm to over 5 ␮m. Also, the
polydispersity of the particles is observed to increase over milling
times of 10, 30, and 60 min but decrease after milling for 100 min.
(3) Effect of Different Milling Media
Employing different types of milling media is known to affect
the outcome of a milled material. While stainless steel and zirconia
balls are commonly used in the milling of ceramic materials, the
use of polymeric media has also been studied. Stainless steel balls
coated with nylon have recently been demonstrated to be an
effective milling medium for contamination reduction in a highenergy ball-milling process.19 To eliminate contamination from
the nylon media, the nylon wear particles in the sample can simply
be removed by thermal treatment or solvation after milling,
resulting in a high-purity product.
While examining the effect of different milling media on the
mesoporous silica particles, significant differences in the BET
surface area were observed for samples milled with different types
of balls. Figure 9 shows the BET surface area of CTAB-templated
silica xerogels milled at different times with zirconia, stainless
steel, and nylon-coated stainless steel balls. At a milling time of 10
min, all three samples display similar surface areas. After 30 min,
there is a sharp decrease in the stainless-steel-milled samples, and
July 2004
Preparation of Nanostructured Silica Particles
1283
Fig. 6. Representative TEM images of CTAB silica xerogels (a) before milling and after milling with nylon-coated stainless steel balls for (b) 10, (c) 30,
and (d) 60 min.
after 60 min both the zirconia- and stainless-steel-milled samples
show significant decreases in surface area. The nylon-coated
stainless steel balls, however, show the smallest decrease in
surface area while producing the same micrometer-sized agglomerated particles.
(4) Effect of Different Surfactant Systems
The effect of using different surfactant systems such as CTAB,
Brij-58, F127, and P123 was also investigated. Different surfactants
are used to synthesize different pore structures, but the properties of
each surfactant system also have large influences in the high-energy
ball-milling process. Table I lists the BET surface area and the total
volume adsorbed for silica xerogels with different surfactants milled
for 100 min. After milling, the CTAB system exhibits the largest
decrease in surface area (33%) followed by F127 (26%) and Brij-58
(8%). However, unlike the F127 and Brij-58 systems, the CTAB
system has a substantial increase in the pore volume. Since the total
volume of nitrogen adsorbed may be comprised of the pore volume
and the interparticle volume, this indicates a significant contribution
from the interparticle volume. The P123 system is not shown in the
data due to the poor condition of the xerogel after milling. Instead of
having the consistency of a fine powder like the other systems, the
P123 xerogel formed a dense coating on the interior of the milling
vial. Sub-micrometer- or micrometer-sized silica particles were also
not evident during SEM investigation. Instead, large continuous silica
agglomerates were observed (data not shown).
The SEM micrographs of the CTAB, F127, and Brij-58
systems are shown in Fig. 10. In these images, the F127 (Fig.
10(a)) and CTAB (Fig. 10(c)) systems appear to be comprised
of small agglomerations of sub-micrometer-sized silica particles. However, in the Brij-58 system (Fig. 10(b)), the individual
particles are not apparent due to very high agglomeration. In
fact, the Brij-58 system agglomerated into particles as large as
100 ␮m (data not shown). It is believed that the texture of the
1284
Journal of the American Ceramic Society—Hampsey et al.
Vol. 87, No. 7
Fig. 7. Representative SEM micrographs of calcined silica xerogels milled with CTAB surfactant inside the pores. The milling times varied from (a) 0, (b)
10, (c) 30, (d) 60, and (e) 100 min.
surfactant plays a large part in the agglomeration of the
particles. At room temperature, P123 has a sticky and resinlike
texture, while Brij-58 has a waxy texture. Both of these systems
exhibited large amounts of agglomeration. However, the CTAB
and F127 systems both have a texture more like a dry powder,
and both of these systems resulted in the production of
sub-micrometer-sized particles.
(5) Prevention of Agglomeration
Finally, in an attempt to reduce the agglomeration of the
particles, the surfactant was removed from the xerogels after
milling by a solvent-extraction technique that involves ultrasonication instead of calcination. The results did show some
improvement in the agglomeration of the particles. Figure 11
shows a representative SEM image of CTAB-templated silica
July 2004
Preparation of Nanostructured Silica Particles
1285
Fig. 8. Effective particle diameter and polydispersity determined by light
scattering of the CTAB-templated silica xerogels as a function of milling
time.
Fig. 9. BET surface area for CTAB-templated silica xerogels milled with
nylon-coated stainless steel, stainless steel, and zirconia balls as a function
of milling time.
Table I. Surface Area and Pore Volume of
Silica Xerogels before and after Milling†
Surfactant
system
CTAB
F127
Brij-58
†
Milling
time (min)
BET surface
area (m2/g)
Total pore
volume (cm3/g)
0
100
0
100
0
100
937.3
630.5
384.9
285.3
708.0
651.6
0.570
0.716
0.342
0.237
0.365
0.372
Fig. 10. Representative SEM micrographs of silica xerogels milled with
zirconia balls. The surfactant used in each sample was (a) F127, (b)
Brij-58, and (c) CTAB.
Milled with zirconia balls.
xerogels milled with nylon-coated stainless steel balls and the
surfactant removed by solvent extraction. As seen in the figure,
individual particles in the micrometer to sub-micrometer range
with almost no agglomeration were observed. It is believed that
the ultrasonication of the particles in solution helps break up the
agglomerates. However, when examining the nitrogen adsorption/desorption isotherms of the solvent-extracted samples (Fig.
12), marked differences from that of the calcined particles are
observed. The volume adsorbed decreases more rapidly with
increased milling time and has less interparticle adsorption in
the solvent-extracted samples than in the calcined samples. It is
believed that the solvent-extraction and cetrifugation cycles
1286
Journal of the American Ceramic Society—Hampsey et al.
Vol. 87, No. 7
milling process. In addition, nylon-coated stainless steel balls were
found to preserve the surface area of the silica better than the
zirconia and stainless steel balls. Milling times of 10 –30 min were
sufficient for producing sub-micrometer particles, and milling
times greater than 30 min just produced more excessive agglomeration. When examining different surfactant systems, CTAB and
F127 were found to be suitable surfactants in this process, while
Brij-58 and P123 surfactants promoted excessive agglomeration. A
solvent-extraction technique was found to reduce the agglomeration of the particles but also removes many of the sub-micrometer
particles. Future directions of this project include surface modification of the silica and incorporation of metal nanoparticles into
the pore framework for use at catalysts.
Acknowledgments
The authors thank Dr. Wayne Reed and Alina Alb for their assistance with the
light scattering measurements.
Fig. 11. Representative SEM micrograph of a nylon-milled silica xerogel
with CTAB surfactant inside the pores. The surfactant was removed after
milling by solvent extraction with ethanol and HMDS.
Fig. 12. Nitrogen adsorption/desorption isotherms for the CTABtemplated silica xerogels milled with nylon-coated stainless steel and the
surfactant removed by solvent extraction.
remove many of the very small particles that have agglomerated
on the surface of the larger particles.
IV.
Conclusions
From these findings, it is concluded that mesoporous silica
particles in the sub-micrometer to micrometer range can be
produced from silica xerogels using a high-energy ball-milling
process. It was discovered that the pore structure of the silica is
reasonably maintained throughout the milling process if the
surfactant is not removed from the xerogel before milling. The
surfactants help to stabilize and reinforce the pore walls during the
References
1
C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli, and J. S. Beck, “Ordered
Mesoporous Molecular Sieves Synthesized by a Liquid-Crystal Template Mechanism,” Nature (London), 359 [6397] 710–12 (1992).
2
J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge, K. D.
Schmitt, C. T. W. Chu, D. H. Olson, E. W. Sheppard, S. B. McCullen, J. B. Higgins,
and J. L. Schlenker, “A New Family of Mesoporous Molecular Sieves Prepared with
Liquid Crystal Templates,” J. Am. Chem. Soc., 114 [27] 10834 – 43 (1992).
3
Y. Lu, R. Ganguli, C. A. Drewien, M. T. Anderson, C. J. Brinker, W. Gong, Y.
Guo, H. Soyez, B. Dunn, M. H. Huang, and J. I. Zink, “Continuous Formation of
Supported Cubic and Hexagonal Mesoporous Films by Sol–Gel Dip-Coating,” Nature
(London), 389 [6649] 364 – 68 (1997).
4
M. T. Anderson, P. S. Sawyer, and T. Rieker, “Surfactant-Templated Silica
Aerogels,” Microporous Mesoporous Mater., 20 [1–3] 53– 65 (1998).
5
H. Yang, G. Vovk, N. Coombs, I. Sokolov, and G. A. Ozin, “Synthesis of
Mesoporous Silica Spheres under Quiescent Aqueous Acidic Conditions,” J. Mater.
Chem., 8 [3] 743–50 (1998).
6
L. Qi, J. Ma, H. Cheng, and Z. Zhao, “Micrometer-Sized Mesoporous Silica
Spheres Grown under Static Conditions,” Chem. Mater., 10 [6] 1623–26 (1998).
7
M. T. Anderson, J. E. Martin, J. G. Odinek, and P. P. Newcomer, “SurfactantTemplated Silica Mesophases Formed in Water:Cosolvent Mixtures,” Chem. Mater.,
10 [1] 311–21 (1998).
8
H.-P. Lin and C.-Y. Mou, “Tubules-within-a-Tubule Hierarchical Order of
Mesoporous Molecular Sieves in MCM-41,” Science (Washington, D.C.), 273 [5276]
765– 68 (1996).
9
R. I. Nooney, D. Thirunavukkarasu, Y. Chen, R. Josephs, and A. E. Ostafin,
“Synthesis of Nanoscale Mesoporous Silica Spheres with Controlled Particle Size,”
Chem. Mater., 14 [11] 4721–28 (2002).
10
C. Boissiere, A. van der Lee, A. E. Mansouri, A. Larbot, and E. Prouzet, “A
Double Step Synthesis of Mesoporous Micrometric Spherical MSU-X Silica Particles,” Chem. Commun. (Cambridge), 2047– 48 (1999).
11
P. S. Singh, and K. Kosuge, “The Synthesis of Mesoporous Silica Spheres by
Octylamine Templating,” Chem. Lett., 101–102 (1998).
12
H.-P. Lin, Y.-R. Cheng, and C.-Y. Mou, “Hierarchical Order in Hollow Spheres
of Mesoporous Silicates,” Chem. Mater., 10 [12] 3772–76 (1998).
13
H. Yang, N. Coombs, and G. A. Ozin, “Morphogenesis of Shapes and Surface
Patterns in Mesoporous Silica,” Nature (London), 386 [6626] 692–95 (1997).
14
D. Zhao, J. Sun, Q. Li, and G. D. Stucky, “Morphological Control of Highly
Ordered Mesoporous Silica SBA-15,” Chem. Mater., 12 [2] 275–79 (2000).
15
H. B. S. Chan, P. M. Budd, and T. deV. Naylor, “Control of Mesostructured
Silica Particle Morphology,” J. Mater. Chem., 11 [3] 951–57 (2001).
16
Y. Lu, H. Fan, A. Stump, T. L. Ward, T. Rieker, and C. J. Brinker, “AerosolAssisted Self-Assembly of Mesostructured Spherical Nanoparticles,” Nature (London), 398 [6724] 223–26 (1999).
17
G. V. R. Rao, G. P. Lopez, J. Bravo, H. Pham, A. K. Datye, H. Xu, and T. L.
Ward, “Monodisperse Mesoporous Silica Microspheres Formed by EvaporationInduced Self Assembly of Surfactant Templates in Aerosols,” Adv. Mater., 14 [18]
1301–304 (2002).
18
H. Zoz and H. Ren, “Processing of Ceramic Powders Using High-Energy
Milling,” Interceram, 49 [1] 24 –34 (2000).
19
C. L. De Castro and B. S. Mitchell, “The Use of Polymeric Milling Media in the
Reduction of Contamination during Mechanical Attrition,” J. Mater. Res., 17 [12]
2997–99 (2002).
䡺