COMMUNICATION

www.aem-journal.com

Microspheres with Diverse Material Compositions Can be

Prepared by Mechanical Milling
Mehnaz Mursalat, Daniel L. Hastings, Mirko Schoenitz, and Edward L. Dreizin*

properties of the material milled and evolve

This study introduces a novel, simple, and scalable method of preparing spherical during milling until a steady state size
particles from a range of materials of potential interest in food, pharmaceuticals, is reached.[17,18] Yet, particle shapes are
energetics, and additive manufacturing. Spherical particles with dimensions in important for many applications.[13,19,20]
Spherical particles are often desired for
the range of
1–100 μm are prepared by mechanical milling of precursor
flowability, ease of transport, processing,
materials in the presence of a blend of immiscible liquids. Microspheres of hard packing, and mixing.[13,21–24]
and ductile materials including metals (aluminum, titanium), metalloids (boron), Spherical particles in the micrometer
oxides (of iron or silicon), organic compounds (melamine, sucrose), and com- size range are most commonly prepared
posites (aluminum–boron, aluminum–titanium, aluminum–copper oxide, alu- from melts or solutions.[25–27] However,
minum–iron oxide) were prepared. The proposed mechanism leading to the challenges arise for refractory compounds
or composites comprising components
formation of spheres includes formation of a Pickering–Ramsden emulsion with diverse melting temperatures. This
coexisting with a dense suspension of solids in the continuous phase. Milling method is also inapplicable to materials
continuously transfers energy to the multiphase mixture, destabilizing particles sensitive to heating, as they may react
located on the liquid interface. This causes a net transport of solids from the before they melt. We introduce a new tech-
continuous phase into the emulsion droplets where solids accumulate and form nique of preparing spherical composite
powders by mechanical milling. It is shown
microspheres that can be recovered after milling. The process continues until the
in this article that mechanical milling in the
solid loading of the droplets exceeds a limit, or until the continuous phase presence of a liquid PCA comprising two
suspension is depleted. Microspheres prepared by this method may be of interest immiscible liquids can produce nearly per-
as feedstock for additive manufacturing, for drug formulations, catalysts, fect microspheres that consist of densely
membranes, and in various other technologies. compacted micro- or nanoparticles of one
or more material components. There are
multiple potential applications for such
microspheres, including feedstock pow-
Mechanochemical processing and synthesis of new materials ders for additive manufacturing (when flowability attained with
have attracted significant attention because of the versatility spherical particles is of critical importance), materials for drug
and simplicity of this approach.[1–4] Products are typically pow- formulation, materials for joining, multifunctional porous com-
ders. Their compositions and structures are often metastable ponents, catalysts, membranes, and so on. Mechanical milling
and not attainable by other synthetic routes.[5,6] Particle shapes used to prepare the microspheres is a readily scalable, industri-
can be described as roughly equiaxial rock-like compacts or ally flexible, and adaptable technology, suggesting that practical
flakes.[7] Although compositions of materials prepared by milling quantities of the microspheres can be readily produced once the
are effectively unlimited, control of particle morphology is diffi- mechanism of their formation is fully elucidated.
cult and may require additional treatment.[8–10] Indeed, the pow- Electron images of microspheres prepared by mechanical
ders are formed by repeated fracturing and cold welding or milling in the presence of a blend of hexane and acetonitrile
agglomeration, starting typically from solid powder precursors. are shown in Figure 1 and compared with their precursor pow-
Common but relatively ineffective ways to affect particle sizes ders. Images for additional materials are shown in Figure S1,
are to change the size of milling balls[11–13] and to select Supporting Information. The microspheres consist of densely
process control agents (PCAs)[14–16] that affect cold welding. packed fine precursor particles but have a strikingly different
However, particle shapes are mostly determined by mechanical morphology. The sizes of the spheres vary in the range of

1–100 μm (Figure S2, Supporting Information). Precursor
M. Mursalat, D. L. Hastings, Prof. M. Schoenitz, Prof. E. L. Dreizin materials tested in this work include metals, metalloids, oxides,
O.H. York Department of Chemical and Materials Engineering and organic compounds. Mechanical properties vary from very
New Jersey Institute of Technology hard (boron) to ductile (aluminum) or brittle (melamine).
University Heights, Newark, NJ 07102, USA Hexane and acetonitrile have previously been used as PCAs
E-mail: dreizin@njit.edu
for mechanical milling of energetic composites.[13,28–31] Their
The ORCID identification number(s) for the author(s) of this article combination forms an emulsion that minimizes the milled mate-
can be found under https://doi.org/10.1002/adem.201901204. rial’s exposure to oxygen. Composite microspheres have been
DOI: 10.1002/adem.201901204 prepared by mechanical milling as well, e.g., combining a metal

Adv. Eng. Mater. 2020, 22, 1901204 1901204 (1 of 4) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Figure 1. Electron images of starting materials (top) and resulting microspheres (bottom) of a) Al, 1.25 h; b) B, 4 h; c) Fe2O3, 1 h; and d) melamine,
1 h after premilling.
and oxide powders and combining metal and metalloid. For
some precursors, e.g., fumed silica, the resulting spheres are
fragile and tend to break upon recovery and drying. However,
none of the microspheres prepared to date, including those that
retain their shapes after drying, as shown in Figure 1, have
included any binder, which can be readily added when structural
integrity of the microspheres needs to be improved. The spheri-
cal particles shown in Figure 1 consist of smaller primary
particles that could represent the starting material directly
such as in the case of the iron oxide used here, or that could form
from initially coarser starting powder, such as in the case of Al.
A comparison of particle size distributions for initial materials
and resulting spherical particles is shown in Figure S2,
Supporting Information.
Figure 2 shows cross-sections of microspheres of Al, Fe2O3,
and composite microspheres of Al·CuO, Al·Fe2O3, and Al·B.
For the ductile aluminum, typically forming flakes during
mechanical milling, the microspheres contain packed flakes,
which are, however, not fused together. The flakes packed inside
the spheres are positioned at different angles to the sectioning
plane, explaining a significant spread of their cross-section areas,
as shown in Figure 2. For iron oxide, consisting of nanosized
primary particles, such particles are packed to nearly full density
inside the microspheres. Among the composites, the Al·CuO
thermites assumed spherical shape with visible pores, whereas
Figure 2. Backscattered electron images of cross-sections of microspheres of a) Al, 1.25 h; b) Fe2O3, 1 h; c) Al·CuO, 1 h; d) Al·Fe2O3, 1 h; and e) Al·B, 1 h.
Adv. Eng. Mater. 2020, 22, 1901204
www.aem-journal.com
Al·Fe2O3 and Al·B appear denser in comparison. For some mate-
rials, porosity of the microspheres is found to depend on the
milling time. Longer milling times can lead to finer and denser
microspheres; supporting evidence is shown in Figure S2,
Supporting Information. However, if milled for too long, the
microspheres break.
The formation mechanism of microspheres is not presently
understood. Microspheres are observed with about equal volume
fractions of solid powder precursor and acetonitrile and with a
hexane volume fraction greater than 50%. An overview of solvent
compositions and solid volume fractions where microsphere for-
mation was observed is shown in Figure 3a. In the resulting
emulsion, the mildly polar acetonitrile typically wets the solids
better than hexane. Particles suspended in acetonitrile tend to
deagglomerate, whereas solids suspended in hexane agglomerate
readily. The specific pattern of mechanical agitation does not
appear to be important because spheres are observed in both
shaker and planetary mills. The milling times needed to generate
microspheres vary for different precursors. However, proper
conditions for microsphere formation were found for all precur-
sor powders tested so far.
Formation of composite spherical particles when a powder is
combined with two immiscible liquids bears similarity to the for-
mation of Pickering–Ramsden (PR) emulsions[32] and colloido-
somes.[33,34] In all such cases, emulsion droplets are stabilized
1901204 (2 of 4) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Figure 3. a) Summary compositional map of solvent compositions for all materials discussed in the text. Filled symbols show where microsphere
formation was observed; open symbols show where microspheres did not form. Dashed lines indicate 50 vol% hexane and an acetonitrile/solids volume
ratio of one, respectively. b) Schematic illustration of microsphere formation mechanism.
by solid particles located at the liquid interface. These particles,
partially wetted by both liquids, become immobilized and are dif-
ficult to remove from the interface. In PR emulsions, there are no
solids in the droplet interiors. Partially filled droplets or colloido-
somes were formed using double emulsions.[34,35] In contrast,
microspheres prepared here are filled to much greater extent
and have distinctly different structures.
We propose that microspheres form as a result of high-energy,
high-shear interactions of a PR emulsion and a high-concentration
suspension. A PR emulsion forms quickly when two immiscible
liquids are combined and agitated if both liquids partially wet
the powder(s) being milled. Because the volume fraction of
acetonitrile in the present effort is always smaller than that of
hexane, it is proposed that acetonitrile forms droplets in the
continuous phase of hexane. Considering typical mass load of
powder found to be suitable to prepare the microspheres, solid
particles stabilized at the liquid interface account for only a small
fraction of the powder loaded in the mill. The rest of the powder
remains suspended in the continuous phase. The resulting
system including both a PR emulsion and a dense suspension
in the continuous phase is unusual and has not been studied
previously.
Milling subjects the liquid interface to significant shear stress.
It has been reported that agitation generates defects and leads to
removal of particles from the liquid interface, and destabilization
and destruction of PR emulsion droplets take place.[36–38]
However, in this work, the interface is stressed in a continuous
liquid densely filled with suspended particles; such particles are
expected to rapidly replace particles at the liquid interface, and to
effectively “repair” any damage to the PR emulsion droplets. This
process is schematically shown in Figure 3b. As the droplet sur-
face is strained, a particle from the surface may move to the drop-
let interior. If the particle does not return to the surface, the
remaining particles rearrange. As continued energy input
removes more particles from the surface, the number density
of particles at the surface decreases until the droplet either breaks
up or other particles from the bulk liquid become trapped at the
liquid interface. It is proposed that in the presence of an agitated,
densely loaded suspension, particle movement from the bulk
suspension to the interface may be faster than either breakup
of the droplet, or return of particles from the droplet interior
to the liquid interface. As a result, solid particles accumulate
in acetonitrile droplets and eventually form filled microspheres
as shown in Figure 1 and 2.
Adv. Eng. Mater. 2020, 22, 1901204 1901204 (3 of 4)
www.aem-journal.com
There could be both thermodynamic and rheological (kinetic)
reasons supporting the proposed mechanism of formation of
filled microspheres. If the liquid inside the droplet wets the par-
ticle better than the continuous liquid, there is an energetic ben-
efit of transferring the particles from the continuous phase into
the droplet. In addition, when the concentration of particles in
the droplet is sufficiently high, capillary forces may hold these
particles together.
Kinetically, particles in the continuous phase and those in drop-
lets have different motion patterns and thus have a different like-
lihood of interacting with the surface. Within the continuous
phase, particles follow the fluid flow, while particles in the droplet
move with the droplet. With any slip between droplets and the
suspension in the continuous phase, the speed of particles relative
to droplets can be substantial, as expected to occur during
mechanical milling. Internal flows are much weaker than the
flows in the continuous phase and increasingly impeded as more
particles accumulate in the droplets. As a result, the emulsion
droplets may become mechanically stable as the particles in the
droplet accumulate. In addition, if the suspended particles are
substantially denser than the liquid of the continuous phase, they
could move by inertia, impacting droplets ballistically. A related
mechanism was recently proposed to explain penetration of sus-
pended oil droplets by solid debris subject to a turbulent flow.[39]
During milling, the amount of powder suspended in the con-
tinuous fluid progressively diminishes. Mechanical agitation may
lead to destabilization and repair of the liquid interface as long as
the concentration of the suspended particles in the continuous
phase remains sufficiently high. In a batch process, once the sus-
pension becomes used up, the droplets may break more readily.
For a given powder mass load, the filled spherical particles can
form in a specific time interval during processing. Once the desta-
bilized droplets break apart, it is unlikely for the released and
rather agglomerated particles to be suspended again. Instead, they
may form clusters of suspended particles too large to stabilize the
liquid interface and to form emulsion droplets.
Observation of any of the described processes in situ is chal-
lenging due the complex motion of the milling containers and
lack of optical access. Experiments investigating transport of
particles across strained liquid interfaces may help in develop-
ing a model. In the first step, PR emulsion droplets were
formed from hexane, acetonitrile, and suspended aluminum
particles by ultrasonic agitation, as shown in Figure S3,
Supporting Information. It may be possible in future work to
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
immobilize such droplets in a microfluidic device and observe
the interaction with a suspension directly.
Experimental Section
Two types of ball mills were used to prepare microspheres from differ-
ent materials. A SPEX Sampleprep 8000D Mixer/Mill with 50 mL flat-
ended steel vials was used to prepare aluminum–copper(II) oxide thermite
microspheres. A Retsch PM 400 MA planetary mill was used to prepare
microspheres of aluminum, titanium, boron, iron(III) oxide, fumed silica,
and melamine, as well as composite microspheres of aluminum–
copper(II) oxide, aluminum–iron(III) oxide, aluminum–titanium, and
aluminum–boron. The planetary mill was equipped with custom hardened
steel milling jars, designed to withstand accidental combustion of reactive
composite materials, such as thermite compositions (although no such
accidents occurred during this effort, where liquid PCA also served as a
coolant). Samples were milled at 350 RPM with 90 g (25 g for the shaker
mill) of 9.525 mm (3/8 00 ) diameter hardened steel balls, typically for 1 h.
Milling time for selected materials (aluminum, titanium, boron, and iron
oxide) varied up to 4 h.
Starting materials included powders of aluminum and titanium (both
by Atlantic Equipment Engineers, 325 mesh, 99.5%), copper oxide
(Sigma Aldrich, 10 μm, more than 99%), fumed silica (Sigma-Aldrich,
99.5%), iron oxide (Alfa Aesar, 325 mesh, 99.5%), boron (SB95 by SB
Boron, <1 μm, 95%), melamine (Sigma-Aldrich, 100–500 μm [est.],
99%), and food grade sucrose.
For the PCAs, hexane and acetonitrile (both by Alfa Aesar, 99.5%) were
used. In most experiments involving the planetary mill, each milling jar
was charged with a total of 24 mL PCA. In experiments with aluminum,
the total volume of PCA was 37 mL. Mixed solvent compositions ranged
from 6.25 to 75 vol% acetonitrile. Milling containers were loaded in a pro-
tective argon atmosphere. A parameter found to be important for the for-
mation of spherical particles was solid to acetonitrile volume ratio. This
ratio varied from 0.06 to >99 (for milling in pure hexane); microsphere
formation was observed in the range of 0.75–2.
Cross-sections were prepared by drying samples, embedding in epoxy
and hand polishing using silicon carbide polishing paper.
Supporting Information
Supporting Information is available from the Wiley Online Library or from
the author.
Acknowledgements
This work was funded in parts by Defense Threat Reduction Agency (grant
no. HDTRA1-15-1-0024), Air Force Office of Scientific Research (grant no.
FA9550-16-1-0266), and Office of Naval Research (grant no. N00014-19-1-
2048).
Conflict of Interest
The authors declare no conflict of interest.
Keywords
composite materials, emulsions, mechanochemistry, suspensions
Received: October 4, 2019
Revised: December 1, 2019
Published online: December 13, 2019
Adv. Eng. Mater. 2020, 22, 1901204 1901204 (4 of 4)
www.aem-journal.com
[1] P. Baláž, M. Achimovičová, M. Baláž, P. Billik, Z. Cherkezova-Zheleva,
J. M. Criado, F. Delogu, E. Dutková, E. Gaffet, F. J. Gotor, R. Kumar,
Chem. Soc. Rev. 2013, 42, 7571.
[2] J. L. Do, T. Friščić, ACS Cent. Sci. 2017, 3, 13.
[3] J. L. Howard, Q. Cao, D. L. Browne, Chem. Sci. 2018, 9, 3080.
[4] D. Tan, T. Friščić, Eur. J. Org. Chem. 2018, 2018, 18.
[5] E. L. Dreizin, M. Schoenitz, in Metal Nanopowders: Production,
Characterization, and Energetic Applications, Wiley-VCH Verlag
GmbH & Co. KGaA, Weinheim 2014, pp. 227–278.
[6] M. Schoenitz, T. Ward, E. L. Dreizin, MRS Proc. 2011, 800, AA2.6.
[7] C. Suryanarayana, Mechanical Alloying and Milling, Marcel Dekker,
New York 2004, pp. 1–472.
[8] T. Kozawa, K. Fukuyama, A. Kondo, M. Naito, ACS Omega 2019, 4, 5690.
[9] Z. Shalabayev, M. Baláž, N. Daneu, E. Dutková, Z. Bujňáková,
M. Kaňuchová, Z. Danková, Ĺ. Balážová, F. Urakaev, Ĺ. Tkáčiková,
M. Burkitbayev, ACS Sustainable Chem. Eng. 2019, 7, 12897.
[10] C. Suryanarayana, E. Ivanov, in Advances in Powder Metallurgy (Eds:
I. Chang, Y. Zhao), Woodhead Publishing, Cambridge 2013, pp. 42–68.
[11] B. Fullenwider, P. Kiani, J. M. Schoenung, K. Ma, Powder Technol.
2019, 342, 562.
[12] Y. F. Zhang, R. R. Li, X. L. Zhao, S. Z. Diao, P. P. Liu, F. R. Wan,
Q. Zhan, Mater. Res. Express 2019, 6, 066535.
[13] Q. Nguyen, C. Huang, M. Schoenitz, K. T. Sullivan, E. L. Dreizin,
Powder Technol. 2018, 327, 368.
[14] A. P. Radlinski, A. Calka, B. W. Ninham, W. A. Kaczmarek, Mater. Sci.
Eng. A 1991, 134, 1346.
[15] A. Nouri, C. Wen, Crit. Rev. Solid State Mater. Sci. 2014, 39, 81.
[16] N. S. Anas, M. Ramakrishna, R. K. Dash, T. N. Rao, R. Vijay, Mater.
Sci. Eng. A 2019, 751, 171.
[17] X. Shang, X. Wang, S. Chen, Materials 2019, 12, 1224.
[18] S. M. Umbrajkar, M. Schoenitz, S. R. Jones, E. L. Dreizin, J. Alloys
Compd. 2005, 402, 70.
[19] W. Jiajun, Y. Xiao-Su, Compos. Sci. Technol. 2004, 64, 1623.
[20] M. Mosharraf, C. Nyström, Int. J. Pharm. 1995, 122, 35.
[21] X. Fu, D. Huck, L. Makein, B. Armstrong, U. Willen, T. Freeman,
Particuology 2012, 10, 203.
[22] Y. Kawashima, M. Imai, H. Takeuchi, H. Yamamoto, K. Kamiya,
T. Hino, Powder Technol. 2003, 130, 283.
[23] H. Attar, S. Scudino, A. K. Chaubey, L. Löber, P. Wang, H. Attar,
F. P. Schimansky, F. Pyczak, J. Eckert, J. Mater. Sci. Technol. 2015, 31, 1001.
[24] A.-X. Ren, T.-Y. Wang, T.-Q. Tang, Y.-R. He, Pet. Sci. 2019, https://doi.
org/10.1007/s12182-019-00401-4.
[25] A. Ünal, Powder Metall. 1990, 33, 53.
[26] J. Vicente, J. Pinto, J. Menezes, F. Gaspar, Powder Technol. 2013, 247, 1.
[27] J. Yao, L. Kuang Lim, J. Xie, J. Hua, C. H. Wang, J. Aerosol Sci. 2008, 39,
987.
[28] X. Liu, M. Schoenitz, E. L. Dreizin, Combust. Flame 2018, 195, 292.
[29] K. L. Chintersingh, M. Schoenitz, E. L. Dreizin, Combust. Flame 2016,
173, 288.
[30] M. Mursalat, M. Schoenitz, E. L. Dreizin, Adv. Powder Technol. 2019, 30, 1319.
[31] J. Yu, B. W. McMahon, J. A. Boatz, S. L. Anderson, J. Phys. Chem. C
2016, 120, 19613.
[32] Y. Chevalier, M. A. Bolzinger, Colloids Surf. A 2013, 439, 23.
[33] A. D. Dinsmore, M. F. Hsu, M. G. Nikolaides, M. Marquez,
A. R. Bausch, D. A. Weitz, Science 2002, 298, 1006.
[34] M. Williams, S. P. Armes, P. Verstraete, J. Smets, Langmuir 2014, 30,
2703.
[35] D. Lee, D. A. Weitz, Adv. Mater. 2008, 20, 3498.
[36] C. P. Whitby, E. J. Wanless, Materials 2016, 9, 626.
[37] C. Griffith, H. Daigle, J. Colloid Interface Sci. 2019, 547, 117.
[38] S. Melle, M. Lask, G. G. Fuller, Langmuir 2005, 21, 2158.
[39] L. Zhao, M. C. Boufadel, J. Katz, G. Haspel, K. Lee, T. King,
B. Robinson, Environ. Sci. Technol. 2017, 51, 11020.
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim