Ball milling of copper powder under dry and surfactant-assisted conditions—on the way towards Cu/Cu2O nanocatalyst

Article abstract of DOI:10.1166/jnn.2019.15784

Samples of copper powder was milled with varied grinding frequencies in the presence of various organic agents (oleylamine, ethylene glycol or dimethyl sulfoxide) or without additives. The effects of experimental conditions were investigated by X-ray diff

Full text of DOI:10.1166/jnn.2019.15784

Article
Journal of
Nanoscience and Nanotechnology
Vol. 19, 389–394, 2019
Copyright © 2019 American Scientific Publishers
All rights reserved
Printed in the United States of America
www.aspbs.com/jnn
Ball Milling of Copper Powder Under Dry and
Surfactant-Assisted Conditions—On the Way
Towards Cu/Cu O Nanocatalyst
2
1
ꢀ2
1ꢀ2
1ꢀ2
3ꢀ4
3ꢀ5
Katalin Musza , Márton Szabados , Adél Anna Ádám , Zoltán Kónya , Ákos Kukovecz ,
2ꢀ6
1ꢀ2ꢀ∗
Pál Sipos , and István Pálinkó
1Department of Organic Chemistry, University of Szeged, Szeged, H-6720, Hungary
2
Material and Solution Structure Research Group, University of Szeged, Szeged, H-6720, Hungary
Department of Applied and Environmental Chemistry, University of Szeged, Szeged, H-6720, Hungary
MTA-SZTE Reaction Kinetics and Surface Chemistry Research Group, Szeged, H-6720, Hungary
MTA-SZTE “Lendület” Porous Nanocomposites Research Group, Szeged, H-6720, Hungary
3
4
5
6
Department of Inorganic and Analytical Chemistry, University of Szeged, Szeged, H-6720, Hungary
Samples of copper powder was milled with varied grinding frequencies in the presence of various
organic agents (oleylamine, ethylene glycol or dimethyl sulfoxide) or without additives. The effects
of experimental conditions were investigated by X-ray diffractometry, scanning electron microscopy
and dynamic light scattering measurements. The aggregation of particles were supressed by added
IP: 185.46.84.155 On: Sat, 08 Dec 2018 12:49:41
organics. The catalytic activities of the variously treated samples were measured in the Ullmann-
Copyright: American Scientific Publishers
type reaction of iodobenzene and 1H-pyrazole.
Delivered by Ingenta
Keywords: Cu Powder, Dry or Wet Milling, Aggregation Suppression by Organics,
XRD-SEM-DLS, Ullmann Reaction.
or in microemulsion is as typical as pyrolysis9 and
8
1
. INTRODUCTION
1
0
11
The main interest towards the use of copper nanoparti-
cles and fine powders stems from their peculiar physi-
cal and chemical properties. They have increased surface
area, controllable size and shape being crucial in their cat-
alytic reactivities. Besides being used as catalysts, they
are applied as ingredient of paints for centuries, ceramic
capacitors and anti-corrosion/antibacterial coatings. Cop-
per is frequently considered as a secondary option to
replace silver, gold, palladium or platinum, especially in
electrolysis. Physical methods including laser ablation,
1
2
13
the atomization and vacuum vapor deposition are
also often used. The mechanochemical techniques can be
regarded as mixed chemical and physical methods. The
starting materials are deformed, welded, fractured and
transformed to or reacted with other crystals due to the
generated local high pressure and temperature from the
1
ꢀ2
3
4
5
1
4–16
kinetic energy of the grinding medium.
The colli-
sions result in continuous deformation and fracturing, and
thus fine powder can be obtained. However, cold welding
can have predominant role when the mechanical treatment
reaches the adequate intensity and the agglomeration of
the particles becomes the favorable process, especially for
ductile and soft materials as metallic copper is. In order
to prevent the sticking of the particles and stabilize them
at the nanoscale, one can select either thermodynamic or
kinetic way of stabilization. The former technique requires
adding solutes to decrease the grain boundary energy, and,
6
microelectronic applications. Due to its relatively high
abundance (50 ppm) in the lithosphere of earth and its
easy recyclability without loss in quality, copper could
be economically and environmentally viable alternative of
scarcer and more expensive precious metals.
Copper powders and nanoparticles have multifarious
synthesis methods including chemical as well as physi-
cal routes. Chemical reduction with ultrasonic irradiation7
1
7ꢀ18
thus help the segregation of particles.
route, secondary particles are necessary to wedge between
In the other
∗
Author to whom correspondence should be addressed.
J. Nanosci. Nanotechnol. 2019, Vol. 19, No. 1
1533-4880/2019/19/389/006
doi:10.1166/jnn.2019.15784
389

Ball Milling of Copper Powder Under Dry and Surfactant-Assisted Conditions
Musza et al.
the grain and reduce their mobilities.19ꢀ20 As the result of
stabilization, the agglomeration processes are suppressed,
and the unimodal dispersity of the milled powders can be
enhanced.2
long (Agilent HP-1). During the chromatographic mea-
surement, the temperature was increased stepwise from
100 to 180 C. Authentic samples were used to identify
the starting materials and the product using benzene as the
internal standard.
ꢀ
1
In experimental work leading to this contribution,
highly viscous organic materials, oleylamine and ethylene-
glycol surfactant and a frequently used organic solvent,
dimethyl sulfoxide were added to the copper powders. The
mechanochemical treatment was carried out in a mixer
mill. In this mill type, the balls impinged from the rounded
ends of the jars causing mill-dependent, characteristic
2.4. Methods of Structural Characterisation
Powder X-ray diffractograms were recorded on a Rigaku
ꢀ
Miniflex II instrument in the 2ꢂ = 5–80 range using
ꢀ
4 /min scan speed and employing CuKꢃ (ꢄ = 1ꢅ5418 Å)
radiation. The reflections of the prepared materials were
identified using the JCPDS (Joint Committee of Powder
Diffraction Standards) database.
2
2
deformations in the sample. Finally, the catalytic perfor-
mance of dry- and wet-milled copper powders were tested
2
3ꢀ24
in the Ullmann reaction of iodobenzene and pyrazole.
A Hitachi S-4700 scanning electron microscope (SEM)
was used to characterize the morphologies of the sam-
ples. Images were obtained at several magnifications and
at 10 kV acceleration voltage. The surfaces of the sam-
ples were coated with thin gold layer in order to avoid
charging. The elemental analysis was performed by energy
dispersive X-ray analysis (EDX). The Röntec QX2 spec-
trometer (installed with Be window) was coupled to the
microscope.
2
. EXPERIMENTAL DETAILS
2
.1. Materials
In the experiments, copper powder 99% purity, average
diameter <75 ꢁm, Sigma-Aldrich), dimethyl sulfoxide
(
99.9%, Sigma-Aldrich), 1-phenylpyrazole (97%, Sigma-
Aldrich), iodobenzene (98%, Sigma-Aldrich), ethylene
glycol (99,8% Reanal Private), 1H-pyrazole (98%, Alfa
Aesar), sodium methoxide (98%, VWR International),
benzene (99,9%, VWR International) and oleylamine (C18
content 80–90%, Acros Organics) were employed. All
chemicals were used without further purification.
Malvern NanoZS dynamic light scattering (DLS) instru-
ment with 4 mW helium-neon laser light source (ꢄ =
6
33 nm) was applied to evaluate size distribution of the
samples at room temperature. Detection was made in back
ꢀ
scattering mode at 173 and the powders were ultrasoni-
IP: 185.46.84.155 On: Sat, 08 Dec 2018 12:49:41
cally dispersed in ethylene glycol.
Copyright: American Scientific Publishers
.2. Mechanical Treatment of Copper Powder
Delivered by Ingenta
2
For the grinding of the copper nanoparticles, a mixer mill
(
3
3. RESULTS AND DISCUSSION
Retsch MM 400) having two stainless steel with 50 cm
3
.1. XRD Analysis of the Dry- and
grinding jars and two stainless steel grinding balls (vol-
3
Wet-Milled Cu Powders
ume: ∼8.2 cm , diameter: 25 mm) were applied. The
The diffractograms of the dry-milled copper powders are
depicted in Figure 1 with varying milling frequencies.
Three reflections were observable, the (111), (200) and
mill was operated using constant ball/sample weight ratio
(
100), and 120 min milling time, and the grinding fre-
quency was systematically varied between 3 and 21 Hz.
Both dry and wet grinding was performed in air.
(
220) they were, which confirmed that the face-centred
cubic structure of the copper powder (JCPDS#04-0836)
During the treatment, 0.6 g of copper powder was
placed in the grinding jar. Wet grindings were carried out
via adding oleylamine, ethylene glycol or dimethyl sul-
foxide, between 1 and 200 ꢁl, to the reaction mixture.
Following the milling, the particles were washed by ace-
tone, collected on 0.45 ꢁm filter and stored in closed glass
was intact after milling and other impurities, such as Cu O
2
(111) (200)
(220)
21 Hz
d=11.0 nm
d=12.0 nm
18 Hz
tubes under N atmosphere.
2
d=13.3 nm
d=13.7 nm
d=20.3 nm
15 Hz
12 Hz
9 Hz
2
.3. Catalytic Test Reaction
The following reactant quantities were used: iodobenzene:
mmol; 1H-pyrazole: 7.5 mmol; Cu powder: 0.2 mmol;
5
d=27.0 nm
d=31.0 nm
d=37.3 nm
6 Hz
3 Hz
0 Hz
sodium methoxide: 10 mmol. The reactants were dis-
solved in 5 ml DMSO, and placed in a continuously
stirred glass reactor with PTFE septum. The reaction mix-
ture was heated to 130 C in oil bath, and quantitatively
analyzed using a gas chromatograph (Hewlett-Packard
ꢀ
20
30
40
50
60
70
80
2
θ (degree)
5
890 Series II). The gas chromatograph was equipped
Figure 1. XRD measurements of the dry-milled powder with various
grinding frequencies (the indexed lines in the diffractogram are of Cu ꢆ.
0
with flame ionization detector and the column was 50 m
390
J. Nanosci. Nanotechnol. 19, 389–394, 2019

Musza et al.
Ball Milling of Copper Powder Under Dry and Surfactant-Assisted Conditions
(
111) (200)
(220)
00 μl
or CuO was not generated in detectable amounts during
milling. The X-ray diffraction pattern at 0 Hz represents
the as-purchased (and unmilled) commercial copper pow-
der with an average grain size under 75 ꢁm (indicated on
the label, possibly the grain consists of highly agglomer-
ated particles); however, the calculated particle size was
around 37 nm (it was obtained from the X-ray diffrac-
togram). X-ray line profile analysis determines the sizes of
coherently scattering domains, which is often in good cor-
relation with the crystallite or subgrain sizes obtained by
other characterising techniques like transmission electron
microscopy. However, these numbers cannot be regarded
as particle sizes, when the powder was prepared by plas-
1
d=26.0 nm
50 μl
25 μl
10 μl
5 μl
1 μl
d=21.3 nm
d=21.0 nm
d=18.7 nm
d=15.7 nm
d=14.0 nm
d=13.7 nm
0
μl
2
0
30
40
50
60
70
80
2θ (degree)
2
5
tic deformation. The generated defect boundaries and the
dipolar walls inside the grains provide smaller values than
the real ones, as revealed by for instance the SEM images
Figure 3. X-ray diffractograms of the wet-milled powders with ethy-
lene glycol (the indexed lines in the diffractogram are of Cu ꢆ.
0
(
they showed copper particles in the micrometer range).
The size of coherently scattering domains was reduced
monotonously by growing the grinding frequency from
During grinding, on liquid addition, copper(I) oxide was
formed as a side product. The extent of oxidation increased
on the increasing amount of organics to a point. The gener-
3
to 12 Hz. On further increasing the frequency, this ten-
dency slowed down. In order to avoid the amortization of
the grinding jars, higher frequency than 21 Hz was not
applied, and the wet milling experiments were performed
at 12 Hz.
Applying dry milling, the ductile copper powders
adhered, and formed thin foils on the internal surface of
the grinding jars. The addition of organic liquids could
ation of Cu O (JCPDS#78-2076) reached the maximum at
2
2
5 ꢁl added organics in all diffractograms. When milling
was performed in the presence of DMSO, its reflections
Fig. 4) were clearly observable to estimate its wt% from
(
2
6
the integrated peaks intensities. The crystalline phase of
cuprous oxide was only 1 wt%, probably it is located
on the surface of copper particles. Curiously, the sizes of
coherently scattering domains have changed and followed
maximum curve, which seemed to be contradicting to for-
mer experiences.
IP: 185.46.84.155 On: Sat, 08 Dec 2018 12:49:41
prevent this phenomenon, the samples kept their powder
Copyright: American Scientific Publishers
consistency allowing easier handling. This observation was
Delivered by Ingenta
used in further experiments, the volume of the liquids
applied were varied from 1 to 200 ꢁl, and in all cases,
even 1 ꢁl proved to be sufficient to avoid foil formation.
The diffractograms registered after the addition of oley-
lamine (Fig. 2) and ethylene glycol (Fig. 3) displayed sim-
ilar behaviour: the coherently scattering domain sizes grew
permanently due to the increasing amount flexible organic
molecules, which could absorb part of mechanical energy
efficiently.
3
.2. Characterization by Scanning
Electron Microscopy
In order to follow the size and morphology alterations,
SEM images of the milled powders were taken (Fig. 5).
For the dry-milled samples, particles sizes varied in the
5
0–200 ꢁm range, which is larger than that of the initial
(111) (200)
(220)
00 μl
0 μl
25 μl
♣ Cu2O
(111) (200)
(220)
1
200 μl
00 μl
0 μl
25 μl
0 μl
μl
d=27.0 nm
d=12.7 nm
d=13.0 nm
d=16.7 nm
d=22.3 nm
d=18.7 nm
d=15.3 nm
d=14.3 nm
d=13.7 nm
1
5
d=24.7 nm
d=23.3 nm
5
♣
♣
♣
1
0 μl
5 μl
1 μl
d=21.0 nm
d=20.3 nm
d=19.0 nm
d=13.7 nm
1
5
1 μl
0 μl
0
μl
20
30
40
50
60
70
80
20
30
40
50
60
70
80
2θ (degree)
2θ (degree)
Figure 2. X-ray diffraction pattern of the wet-milled powders in the
Figure 4. XRD study of the wet-milled powders with dimethyl sulfox-
ide (beside the indexed lines in the diffractogram are of Cu , the lines of
0
presence of oleylamine (the indexed lines in the diffractogram are
0
of Cu ꢆ.
Cu O oxide also appeared).
2
J. Nanosci. Nanotechnol. 19, 389–394, 2019
391

Ball Milling of Copper Powder Under Dry and Surfactant-Assisted Conditions
Musza et al.
A
B
C
IP: 185.46.84.155 On: Sat, 08 Dec 2018 12:49:41
Copyright: American Scientific Publishers
Delivered by Ingenta
D
Figure 5. SEM images of the milled powders: A—dry grinding, wet-ground samples in the presence of 25 ꢁl of organics: B—oleylamine, C—ethylene
glycol, D—dimethyl sulfoxide.
powder. The reason is that the particles flattened due to the
cold welding and the mechanical compressive stress. Using
wet milling, the planar morphology with rounded edges
remained; however, the average particle sizes were reduced
under 30 ꢁm, especially on DMSO addition, where the
size of larger part of the grains did not reach 10 ꢁm. These
observations spectacularly indicate the stabilizing effect of
added organic solvents, even that of dimethyl sulfoxide,
which commonly is not considered to be a surface active
agent. EDX measurements did not show any sign of iron
atoms verifying that the grinding jars were not degraded
at 12 Hz either under dry- or wet-milling conditions.
3.3. Size Distributions of the Milled Powders
From the commercial and dry-milled samples, we were
unable to create stable, non-settling dispersion because of
the voluminous particles obtained; however, the wet-milled
powders could be dispersed successfully in 1 h continu-
ous ultrasonic treatment in ethylene glycol (Fig. 6). The
curves displayed bimodal distribution with progressively
reduced particles sizes. Two main routes of mechanical
energy mediation can be envisaged; the direct energy trans-
fer between the powders and grinding medium (balls, inner
surface of the jars) or the indirect one, when particles
transfer their kinetic energy to other particles. The direct
392
J. Nanosci. Nanotechnol. 19, 389–394, 2019

Musza et al.
Ball Milling of Copper Powder Under Dry and Surfactant-Assisted Conditions
620
710
Ethylene glycol
Oleylamine
110
140
100 μl
170
710
100 μl
220
1700
25 μl
25 μl
400
2000
3000
250
1
μl
1 μl
10
100
1000
10
100
1000
Diameter (nm)
Diameter (nm)
230
DMSO
40
100 μl
430
70
25 μl
800
140
1
μl
10
100
1000
Diameter (nm)
IP: 185.46.84.155 On: Sat, 08 Dec 2018 12:49:41
Copyright: American Scientific Publishers
Figure 6. Bimodal aggregate intensity-weighted size distribution curves of wet-milled copper powders.
Delivered by Ingenta
and indirect energy transfers are connected to elastic
and inelastic collisions, respectively. In the latter, the
soft and ductile copper particles absorbed a part of the
energy and transferred a smaller proportion to deforma-
tions. These effects may induce the aggregation of the
primary particles in two different size ranges. In the pres-
ence of oleylamine and ethylene glycol, these ranges were
maximum yield decreased due to polymerization of the
primary product falling to zero after 48 h reaction time.
The untreated commercial copper powder was of low
activity, which could be enhanced significantly (by more
than three times) on dry milling. The activity of the cat-
alyst could be enhanced further applying wet milling in
oleylamine and even further using DMSO. The significant
activity increase on using oleylamine may be connected
to suppressing the aggregation processes, and thus, to the
larger accessible catalytically active surface. This is a sig-
nificant factor for the DMSO-milled materials as well;
2
000–3000 nm and 250–400 nm. The distribution was
bimodal in the presence of DMSO as well; however, the
particles were smaller, around 800 and 140 nm. Apply-
ing 100 ꢁl DMSO the particle sizes further decreased
under 100 nm. DLS measurements verified that the DMSO
was an excellent stabilizing agent mitigating aggregation
induced by cold welding or van der Waals interactions.
however, as it has been mentioned previously, Cu O was
2
Table I. Ullman-type reaction of iodobenzene and 1H-pyrazole in the
synthesis of 1-phenylpyrazole.
3
.4. Catalytic Test
Derivatives of pyrazole are widely investigated as biolog-
ically active heterocyclic chemicals. They have potential
in the synthesis of anti-inflammatory, antimicrobial agents,
herbicidal or even A3 adenosine receptor antagonists. In
our work, the catalytic activity of the milled and com-
mercial copper powders were compared in the Ullmann-
type synthesis of 1-phenylpyrazole (Table I). The duration
of the reaction was limited to 24 h, in longer runs, the
Catalyst
Yield (%)
2
7
Without catalyst
Commercial Cu powder
Dry-milled Cu powder
Cu powder milled with ethylene glycol
Cu powder milled with oleylamine
Cu powder milled with DMSO
0
9
28
14
49
86
J. Nanosci. Nanotechnol. 19, 389–394, 2019
393

Ball Milling of Copper Powder Under Dry and Surfactant-Assisted Conditions
Musza et al.
also formed on Cu particles during wet milling, and it is
9. V. Rosenband and A. Gany, J. Mater. Process. Technol.
53–154, 1058 (2004).
2
8–30
1
known to aid the synthesis.
This promoting effect was
1
1
1
0. N. D. Nikoli c´ , L. J. Pavlovi c´ , M. G. Pavlovi c´ , and K. I. Popov,
the highest in DMSO indicating the presence of increased
Powder Technol. 185, 195 (2008).
amount of Cu O phase as a side effect of wet milling.
2
1. M. S. Yeh, Y. S. Yang, Y. P. Lee, H. F. Lee, Y. H. Yeh, and C. S.
Yeh, J. Phys. Chem. B 103, 6851 (1999).
2. N. B. Dhokey, M. G. Walunj, and U. C. Chaudhari, Adv. Powder
Technol. 25, 795 (2014).
4
. CONCLUSIONS
By increasing the grinding frequency, the coherently scat-
tering domain size was reduced monotonously; however,
but the cold welding processes resulted in heavily aggre-
gated particles, predominant during dry milling. Using
added organics could decrease the aggregation tendency
of milled copper powder. Dimethyl sulfoxide proved to be
excellent protective agent, the as-prepared particles were
generated in nano- or micrometer scale. Even dry milling
was able to increase the catalytic activity; however, wet
milling, especially in DMSO led to remarkable further
increase due to the significant increase in the accessible
13. A. I. Ustinov, T. V. Melnichenko, K. V. Liapina, and A. E. Shishkin,
Vacuum 141, 272 (2017).
1
1
4. C. Suryanarayana, Prog. Mater. Sci. 46, 1 (2001).
5. B. Madavali, J.-H. Lee, J. K. Lee, K. Y. Cho, S. Challapalli, and
S.-J. Hong, Powder Technol. 256, 251 (2014).
1
6. M. S. Khoshkhoo, S. Scudino, T. Gemming, J. Thomas,
J. Freudenberger, M. Zehetbauer, C. C. Koch, and J. Eckert, Mater.
Design 65, 1083 (2015).
1
1
1
7. P. Wynblatt and R. C. Ku, Surf. Sci. 65, 511 (1977).
8. R. Kirchheim, Acta Mater. 55, 5129 (2007).
9. C. C. Koch, R. O. Scattergood, K. A. Darling, and J. E. Semones,
J. Mater. Sci. 43, 7264 (2008).
20. M. A. Atwater, D. Roy, K. A. Darling, B. G. Butler, R. O.
Scattergood, and C. C. Koch, Mat. Sci. Eng. A 558, 226 (2012).
active surface as well as the promoting effect of Cu O
2
2
1. K. A. Nazari, A. Nouri, and T. Hilditch, J. Alloy. Compd. 615, 47
2014).
even if it was formed in minute amounts during milling.
(
2
2. P. Balaz, Mechanochemistry in Nanoscience and Minerals Engi-
neering: High-Energy Milling, Springer-Verlag, Berlin, Heidelberg
(2008), pp. 111–114.
Acknowledgment: This work was supported via the
grant GINOP-2.3.2-15-2016-00013. The financial help is
highly appreciated.
23. G. Clavé, C. Garel, C. Poullain, B.-L. Renard, T. K. Olszewski,
B. Lange, M. Shutcha, M.-P. Faucon, and C. Grison, RSC. Adv.
6
, 59550 (2016).
References and Notes
24. C. Sambiagio, S. P. Marsden, A. J. Blacker, and P. C. McGowan,
Chem. Soc. Rev. 43, 3525 (2014).
1
2
3
4
. A. Galloa, T. Tsoncheva, M. Marelli, M. Mihaylov, M. Dimitrov,
V. Dal Santo, and K. Hadjiivanov, ApIpPl.:C1a8ta5l..4B61. 28 6 4, .1 16 51 5( 2 0O1 2n ) :. Sat,2 05 8. TD. eU cn g 2á r0, 1G 8. T1 i 2c h: y4 ,9J :. 4 G1 ubicza, and R. J. Hellmig, Powder Diffr.
. F. Alonso, Y. Moglie, and G. Radivoy, ACcc.oCphyermig. hRte:s.A 4m8 ,e2r 5i c1 6a n Scien2 t0i f, i 3c 6 P6 (u2 b0 0l i 5s ) h. ers
(
2015).
26. C. R. Hubbard, E. H. Evans, and D. K. Smith, J. Appl. Crystallogr.
9, 169 (1976).
Delivered by Ingenta
´
. S. Švarcová, Z. Cermáková, J. Hradilová, P. Bezdi c´ ka, and D. Hradil,
Spectrochim. Acta A 132, 514 (2014).
. S. P. Wu, H. L. Qin, and P. Li, BME-MLCC, J. Univ. Sci. Technol. B
27. Z. N. Siddiqui, F. Farooq, T. N. M. Musthafa, A. Ahmad, and A. U.
Khan, J. Saudi Chem. Soc. 17, 237 (2013).
1
3, 250 (2006).
28. S. U. Son, I. K. Park, J. Park, and T. Hyeon, Chem. Commun. 7, 778
(2004).
29. Y.-P. Zhang, Y.-C. Jiao, Y.-S. Yang, and C.-L. Li, Tetrahedron Lett.
54, 6494 (2013).
30. S. B. Ötvös, G. Hatoss, Á. Georgiádes, Sz. Kovács, I. M. Mándity,
Z. Novák, and F. Fülöp, RSC. Adv. 4, 46666 (2014).
5
6
7
. J. Xiong, B. Xu, and H. Ni, Int. J. Min. Met. Mater. 16, 293 (2009).
. L. Lu, M. L. Sui, and K. Lu, Science 287, 1463 (2000).
. R. V. Kumar, Y. Mastai, Y. Diamant, and A. Gedanken, J. Mater.
Chem. 11, 1209 (2001).
8
. L. Qi, J. Ma, and J. Shen, J. Colloid. Interf. Sci. 186, 498 (1997).
Received: 9 October 2017. Accepted: 13 December 2017.
394
J. Nanosci. Nanotechnol. 19, 389–394, 2019