materials

Article
Effects of Mechanical Ball Milling Time on the
Microstructure and Mechanical Properties of
Mo2NiB2-Ni Cermets
Lei Zhang 1 , Zhifu Huang 1, *, Yangzhen Liu 2 , Yupeng Shen 1 , Kemin Li 1 , Zhen Cao 1 ,
Zijun Ren 3 and Yongxin Jian 1, *
1 State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an 710049, China;
xjtulei@outlook.com (L.Z.); ilovehit3@126.com (Y.S.); kmlixjtu@outlook.com (K.L.); zcao_xjtu@163.com (Z.C.)
2 Institute of Advance Wear & Corrosion Resistant and Functional Materials, Jinan University, Jinan 510632,
China; liuyangzhen626@163.com
3 Instrumental Analysis Center of Xi’an Jiaotong University, Xi’an 710049, China; renzijun@xjtu.edu.cn
* Correspondence: zhifuhuangxjtu@163.com (Z.H.); yxjian@xjtu.edu.cn (Y.J.); Tel.: +86-29-82665479 (Z.H.)




Received: 22 May 2019; Accepted: 13 June 2019; Published: 14 June 2019


Abstract: Mo2 NiB2 -Ni cermets have been extensively investigated due to their outstanding properties.
However, studies have not systematically examined the effect of the powder milling process on the
cermets. In this study, Mo, Ni, and B raw powders were subjected to mechanical ball milling from
1 h to 15 h. XRD patterns of the milled powders confirmed that a new phase was not observed at
milling times of 1 h to 15 h. With the increase in the mechanical ball milling time from 1 h to 11 h, raw
powders were crushed to small fragments, in addition to a more uniform distribution, and with the
increase in the mechanical ball milling time to greater than 11 h, milled powders changed slightly.
Mo2 NiB2 -Ni cermets were fabricated by reaction boronizing sintering using the milled powders at
different ball milling times. The milling time significantly affected the microstructure and mechanical
properties of Mo2 NiB2 -Ni cermets. Moreover, the Mo2 NiB2 cermet powder subjected to a milling
time of 11 h exhibited the finest crystal size and the maximum volume fraction of the Mo2 NiB2 hard
phase. Furthermore, the cermets with a milling time of 11 h exhibited a maximum hardness and
bending strength of 87.6 HRA and 1367.3 MPa, respectively.

Keywords: powder metallurgy; ball milling; Mo2 NiB2 -Ni cermets; microstructure;
mechanical properties

1. Introduction
Mo2 FeB2 -, Mo2 NiB2 -, and WCoB-based cermets have been successfully fabricated by reaction
boronizing sintering [1–4], generating a ternary boride hard phase by liquid-phase sintering.
Boride cermets can be prepared by using a metal binder phase. These cermets exhibit outstanding
mechanical properties, including high hardness, good bending strength, and excellent fracture
toughness. In particular, Mo2 NiB2 -Ni cermets [5,6], comprising the Mo2 NiB2 hard phase that
contributes to the hardness and the Ni binder phase that contributes to the toughness. These cermets
have attracted considerable attention due to their excellent abrasion resistance, anti-oxidation properties,
and superior corrosion resistance [7,8]. With those unique properties, Mo2 NiB2 -Ni cermets exhibit
good prospects for engineering materials in wear- and erosion-resistant applications, and cermets are
expected to replace conventional materials, including nickel-based alloys in injection molding machine
parts and copper alloys in pumps used in oil extraction and transport.
Previous studies of Mo2 NiB2 -Ni cermets have focused on the improvement of mechanical
properties. A series work of Takagi et al. [9] has reported the effect of element doping on the

Materials 2019, 12, 1926; doi:10.3390/ma12121926 www.mdpi.com/journal/materials

Materials 2019, 12, 1926 2 of 17

microstructure and mechanical properties of Mo2 NiB2 -Ni cermets. Their study revealed that the
addition of Cr and V elements can render remarkable strength to the solid solution by the dissolution of
the Mo2 NiB2 hard phase. Hence, the mechanical properties of Mo2 NiB2 -Ni cermets, such as hardness
and fracture toughness, can be significantly improved by the addition of Cr and V elements, which
refines the grains of Mo2 NiB2 . However, the addition of other elements, e.g., Fe, Co, Ti, and Mn, can
reduce the performance of the cermets [10,11], mainly related to the contamination of grain boundaries.
Furthermore, Takagi et al. [12] have revealed that the properties of bulk Mo2 NiB2 can be improved by
the adjustment of Mo/B due to the change in the Mo2 NiB2 cell lattice. However, marginal attention
has been focused on the fabrication of Mo2 NiB2 -Ni cermets. In particular, a systematic investigation
focusing on the powder milling process has not been reported.
For powder metallurgy, planetary ball milling is typically employed for preparing milled powders
due to its convenience and cost-effectiveness. During milling, to achieve the optimization of milled
powders, several milling parameters need to be adjusted, such as the ball-to-powder weight ratio,
milling speed, and milling time. Milling time is a key parameter, and it directly affects the morphology
and distribution of the milled powders. The relationships between the specific surface area of the
powder and milling can be expressed as follows [13]:

Sm − S0
In = kt (1)
Sm − S

Here, k is the constant of dispersion velocity, t is the milling time, and Sm is the specific surface area
of the material after ultimate milling, S0 is the specific surface area of the material before milling, and S
is the specific surface area of the material after milling. Equation (1) revealed that a long milling time
can refine the powders. However, it does not indicate that the powder can be incessantly crushed at an
infinite milling time. With the extension of the milling time, the powder properties may deteriorate,
including poor fluidity and agglomeration, due to the cold weld and work hardening. Meanwhile,
large particles in powders can worsen the fluidity, compaction, and sinterability during sintering,
ultimately affecting the properties of cermets. Besides, in the case of a milling time longer than the
required time, the contamination in milled powders will increase. Hence, the powder should be
appropriately milled to achieve a favorable composition, ensuring that the powder is in equilibrium
between fracture and cold welding [14–16].
Thus, it is crucial to reveal the effect of milling times on the properties of milled powders and
Mo2 NiB2 -Ni cermets. In this study, the microstructure and distribution of milled powders at different
milling times were examined, as well as the mechanical properties of Mo2 NiB2 -Ni cermets prepared
using milled powders at different milling times.

2. Materials and Methods

Commercially available Mo, Ni, and B powders were used as raw materials. Table 1 and Figure 1
summarize the characteristics and microstructures of these powders, respectively. B exhibits an irregular
morphology; meanwhile, Ni and Mo exhibit a near-spherical morphology. Table 2 summarizes the
elemental compositions of the raw materials.

Table 1. Characteristics of Mo, Ni, and B powders.

Chemical Composition
Powder Particle Size (µm) Manufacturer
(wt%)
Fe < 0.005, Si < 0.002, Changsha Tianjiu Metal Material
Mo 3.4–8.6
Mg < 0.002 Corp., Ltd., Changsha, China
Changsha Tianjiu Metal Material
Ni Fe < 0.006, Mg < 0.002 2.8–6.8
Corp., Ltd., Changsha, China
Baoding Zhongpuruituo Technology
B Fe < 0.005, Mn < 0.003 5.4–10.1
Corp., Ltd., Baoding, China
Materials 2019, 12, 1926 3 of 17
Materials 2019, 12, x FOR PEER REVIEW 3 of 17

Figure
Figure 1.
1. Micrographs
Micrographs of
of raw
raw powders:
powders: (a)
(a) Mo,
Mo, (b)
(b) Ni,
Ni, and
and (c)
(c) B,
B, and
and (d)
(d) their
their particle
particle size.
size.

Table 2.
Table Compositions (wt%)
2. Compositions (wt%) design
design of
of the
the milled
milled powders
powders in
in this
this study.
study.

Mo
Mo Ni
Ni BB
60.5
60.5 33.5
33.5 6.0
6.0

The
The raw
raw powders
powders were
were subjected
subjected toto ball
ball milling
milling by by using
usingaaplanetary
planetaryball ballmill
mill(Focucy,
(Focucy, F-P400).
F-P400).
To
To prevent the powders from oxidation during milling, a stainless-steel milling pot subjected to
prevent the powders from oxidation during milling, a stainless-steel milling pot subjected to
vacuum
vacuum was was used.
used. A A polytetrafluoroethylene
polytetrafluoroethylene (PTFE) (PTFE) liner
liner was
was added
added in in the
the pot,
pot, which
which can can avoid
avoid
damage
damage from from stainless
stainless steel
steel during milling, caused
during milling, caused by by the
the ultrahard
ultrahard B B powder.
powder. In In addition,
addition, wetwet
grinding with ethanol and a relatively low speed of 250 rotations
grinding with ethanol and a relatively low speed of 250 rotations per minute were employedper minute were employed to
to reduce
reduce
the damagethe damage to theliner.
to the PTFE PTFESiliner. Si3N4 grinding balls (Φ5 mm) with high hardness were used to
3 N4 grinding balls (Φ5 mm) with high hardness were used to insure
insure that the ultrahard B powder
that the ultrahard B powder can be crushed can be crushed
during during ball milling
ball milling at a ratio atof
a 3:1
ratio
forofthe
3:1Sifor the Si3N4
3 N4 balls to
balls to the
the raw raw powders.
powders.
The
The raw
raw powders
powders were
were milled
milled atat eight
eight milling
milling times,
times, from
from 11 hh to
to 15
15 h,
h, with
with aa constant
constant increment
increment
of 2 h. The milled powder slurry was dried by using a rotary evaporator
of 2 h. The milled powder slurry was dried by using a rotary evaporator at 50 C under vacuum. at 50 °C under
◦ vacuum. The
milled powders
The milled were were
powders cold-isostatically shaped
cold-isostatically into cylindrical
shaped compacts
into cylindrical (~Φ44 mm
compacts (~Φ44 × 8 mm
mm)×under8 mm)a
pressure of 200 MPa at 120 s. The compacts were sintered in a vacuum furnace
under a pressure of 200 MPa at 120 s. The compacts were sintered in a vacuum furnace with graphite with graphite heaters
(Chenhua, ZT-40, Shanghai,
heaters (Chenhua, China),China),
ZT-40, Shanghai, followed by placing
followed them them
by placing in a graphite crucible
in a graphite during
crucible the
during
process.
the process.TheThefurnace chamber
furnace chamber waswasevacuated
evacuated and
andremained
remainedtoto10 10 Pa
−2
−2
Pathroughout
throughout the the sintering
sintering
process. heating process, the chamber was heated to 550 °C
process. During the heating process, the chamber was heated to 550 C and preserved for 30 min
During the ◦ and preserved for 30 min to
to
remove the PTFE in the compacts. After the heat preservation, the chamber
remove the PTFE in the compacts. After the heat preservation, the chamber was sequentially heated to was sequentially heated
to
1200 ◦ C °C
1200 at the
at the heating
heating raterate 10 °C/min
of 10of◦ C/min and sintered
and sintered for 60formin.60 The
min.specimens
The specimens were naturally
were naturally cooled
cooled to room temperature inside the furnace. The master sintering curve
to room temperature inside the furnace. The master sintering curve was designed (Figure 2). was designed (Figure 2).
Materials 2019, 12, 1926 4 of 17
Materials 2019, 12, x FOR PEER REVIEW 4 of 17

Figure
Figure 2. Sintering process
2. Sintering process of
of Mo
Mo22NiB
NiB22-Ni
-Nicermets.
cermets.

Scanning electron microscopy (SEM, Hitachi, SU3500, Tokyo, Japan), equipped with an
Scanning electron microscopy (SEM, Hitachi, SU3500, Tokyo, Japan), equipped with an energy-
energy-dispersive X-ray spectrometry (EDS), was employed to observe the microstructure of the
dispersive X-ray spectrometry (EDS), was employed to observe the microstructure of the milled
milled powders and Mo2 NiB2 -Ni cermets. Electron Probe Microanalyzer (EPMA, JEOL, JXA-8230,
powders and Mo2NiB2-Ni cermets. Electron Probe Microanalyzer (EPMA, JEOL, JXA-8230, Tokyo,
Tokyo, Japan) was used to confirm the compositions of elements of the milled powders and Mo NiB -Ni
Japan) was used to confirm the compositions of elements of the milled powders and Mo22NiB22-Ni
cermets at different milling times. X-ray diffraction (XRD, Bruker, D8 Advance, Billerica, MA, USA)
cermets at different milling times. X-ray diffraction (XRD, Bruker, D8 Advance, Billerica, MA, USA)
was employed to identify the phase compositions of the milled powders and Mo NiB -Ni cermets,
was employed to identify the phase compositions of the milled powders and Mo22NiB22-Ni cermets,
and the results were obtained by using the XRD analysis software (MDI, Jade 6.0).
and the results were obtained by using the XRD analysis software (MDI, Jade 6.0).
The relative densities and densities of Mo NiB -Ni cermets were estimated by a balance with a
The relative densities and densities of Mo22NiB22-Ni cermets were estimated by a balance with a
sensitivity of 0.01 mg at room temperature by the Archimedes’ method. In addition, the hardness, as
sensitivity of 0.01 mg at room temperature by the Archimedes’ method. In addition, the hardness, as
Rockwell hardness (HRA), at room temperature was estimated. The Rockwell hardness tester (MX1000,
Rockwell hardness (HRA), at room temperature was estimated. The Rockwell hardness tester
Jinan, China) was utilized to estimate the hardness under a load of 50 kg and a dwell time of 10 s.
(MX1000, Jinan, China) was utilized to estimate the hardness under a load of 50 kg and a dwell time
The three-point bending strength was estimated using a universal material testing machine
of 10 s.
(INSTRON, 1195, Boston, MA, USA). According to the test methods, the cermet specimens were cut
The three-point bending strength was estimated using a universal material testing machine
into dimensions of 3 mm × 4 mm × 22 mm using 20-mm spans with a cross-head speed of 0.5 mm/min.
(INSTRON, 1195, Boston, MA, USA). According to the test methods, the cermet specimens were cut
The results obtained from the above-mentioned properties corresponded to the statistical average
into dimensions of 3 mm × 4 mm × 22 mm using 20-mm spans with a cross-head speed of 0.5 mm/min.
of five specimens for each test to insure precision.
The results obtained from the above-mentioned properties corresponded to the statistical
average of and
3. Results five specimens
Discussionfor each test to insure precision.

3. Results
3.1. and Discussion
XRD Analysis of Powders at Different Milling Times
To analyze
3.1. XRD Analysistheofchange
Powders inatthe crystal Milling
Different structures of the milled powders during milling, XRD patterns
Times
of Mo, B, and Ni powders at different milling times were recorded with the increase in the milling
timeTo fromanalyze
0 to 15the
h. change
Figure 3a in shows
the crystal
the XRDstructures of the
patterns. Frommilled
the XRDpowders during
patterns, milling,peaks
diffraction XRD
patterns of Mo, B, and Ni powders at different milling times were recorded
corresponding to Mo and Ni powders are clearly observed. Besides, a broad scattering peak with with the increase in thea
milling time from 0 to 15 h. Figure 3a shows
◦ the XRD patterns. From the XRD
relatively low intensity is observed at 26 (Figure 3b), and the intensity slightly increases with the patterns, diffraction
peaks
increase corresponding
in the millingtotimeMo from
and Ni 0 topowders are clearly
15 h, possibly observed.
related Besides,
to the change in athe
broad scattering
brittle B powders peak to
with
a non-crystallized powder by long-time milling. Moreover, new phases were not observed, andwith
a relatively low intensity is observed at 26° (Figure 3b), and the intensity slightly increases the
the increasepeaks
diffraction in theshifted
millingwith
timethe from 0 to 15 in
increase h, the
possibly
millingrelated
time tofromthe0change
to 15 h,inindicating
the brittle that
B powders
a solid
to
solution does not form in the milled powders during milling. In addition, the result also revealed and
a non-crystallized powder by long-time milling. Moreover, new phases were not observed, that
the diffraction
mechanical peaksoccurs
alloying shiftedby with the increase
mechanical in the milling
ball milling time from
in this study. 0 to 15 h, the
Furthermore, indicating that a
XRD patterns
solid
revealedsolution does
that the not formpeaks
diffraction in the milled
of Mo andpowders
Ni become during
shorter milling. In addition,
and wider with thethe resultinalso
increase the
revealed that mechanical alloying occurs by mechanical ball milling in this
milling time from 0 to 7 h, which gradually change by the increase in the milling time from 7 h to 11study. Furthermore, theh.
XRD
With patterns
the increaserevealed
in the that the time
milling diffraction peaks
to greater thanof11
Mo h,and Ni becomepeaks
the diffraction shorter didand
notwider
clearlywith the
change.
increase in the milling time from 0 to 7 h, which gradually change by the increase in the milling time
from 7 h to 11 h. With the increase in the milling time to greater than 11 h, the diffraction peaks did
Materials 2019, 12, 1926 5 of 17

Materials 2019, 12, x FOR PEER REVIEW 5 of 17

In particular, with the increase in the milling time from 0 to 11 h, the height of the (111) peak of Mo
(Figure 3c) sharply decreases, and the width rapidly increases and then slightly changes. In addition,
not clearly change. In particular, with the increase in the milling time from 0 to 11 h, the height of the
from the XRD patterns, the full-width at half-maximum (FWHM) of the Mo and Ni diffraction peaks
(111) peak of Mo (Figure 3c) sharply decreases, and the width rapidly increases and then slightly
exhibits remarkable broadening, and the diffraction peak intensities of Mo and Ni are visibly weakened,
changes. In addition, from the XRD patterns, the full-width at half-maximum (FWHM) of the Mo and
mainly related to the distortion of the grain refining and lattice distortion caused by the ball-milling
Ni diffraction peaks exhibits remarkable broadening, and the diffraction peak intensities of Mo and
action. This phenomenon is in good agreement with the results reported by Sharafi et al. [17]. In this
Ni are visibly weakened, mainly related to the distortion of the grain refining and lattice distortion
study, the decrease in the average crystallite size is typically reflected by the broadening of FWHM.
caused by the ball-milling action. This phenomenon is in good agreement with the results reported
Furthermore, according to the Williamson and Hall plot [18], the relationship between the crystallite
by Sharafi et al. [17]. In this study, the decrease in the average crystallite size is typically reflected by
size and FWHM can be concluded as follows:
the broadening of FWHM. Furthermore, according to the Williamson and Hall plot [18], the
relationship between the crystallite size β2 and FWHM
Kλ β be concluded
can as follows:
= · + 16e2 (2)
2
tan θ L tan𝜷θ sin θ
𝑲𝝀
= ∙ 𝟏𝟔𝒆𝟐 (2)
𝑳 𝜽 𝜽
where L is the crystallite size, and λ is the FWHM, K is determined by the crystallite configuration. β is
where L is the crystallite size, and λ is the FWHM, K is determined by the crystallite configuration. β
the spread that is related to the lattice strain and crystallite size alone. e is the wavelength of the X-ray
is the spread that is related to the lattice strain and crystallite size alone. e is the wavelength of the X-
radiation, and θ is the Bragg diffraction angle.
ray radiation, and θ is the Bragg diffraction angle.

Figure 3. (a) XRD patterns of milled powders with milling from 0 h to 15 h, (b) evolution of peak of B
Figure 3. (a) XRD patterns of milled powders with milling from 0 h to 15 h, (b) evolution of peak of B
with milling times, (c) evolution of (110) peak of Mo with milling times.
with milling times, (c) evolution of (110) peak of Mo with milling times.
To further reveal the change in the microscopic crystal of the powder with the increase in the
To further reveal the change in the microscopic crystal of the powder with the increase in the
milling process from 1 h to 15 h, the FWHM of the (111) crystal plane and the average crystallite size of
milling process from 1 h to 15 h, the FWHM of the (111) crystal plane and the average crystallite size
Mo (calculated by Equation (2)) are shown in Figure 4. The variation trend of FWHM revealed that
of Mo (calculated by Equation (2)) are shown in Figure 4. The variation trend of FWHM revealed that
with the increase in the milling time from 3 to 11 h, the FWHM of the Mo (111) peak rapidly increases
with the increase in the milling time from 3 to 11 h, the FWHM of the Mo (111) peak rapidly increases
from 1 to 3 h and then gradually increases (Figure 4). With the increase in the milling time to greater
from 1 to 3 h and then gradually increases (Figure 4). With the increase in the milling time to greater
than 11 h, the FWHM of the Mo (111) peak slightly decreases. It also indicates that the milling time has
than 11 h, the FWHM of the Mo (111) peak slightly decreases. It also indicates that the milling time
a significant effect on the microstrain of Mo (110). The microstrain increases from 0.016% to 0.208%
has a significant effect on the microstrain of Mo (110). The microstrain increases from 0.016% to
with the milling time increasing from 1 to 15 h, mainly due to the releasing of the microstrain of the
0.208% with the milling time increasing from 1 to 15 h, mainly due to the releasing of the microstrain
refining powders and the deformation accumulation caused by impact energy during the milling
of the refining powders and the deformation accumulation caused by impact energy during the
process. From the variation trend of microstrain, it can be seen that although microstrain increases as
milling process. From the variation trend of microstrain, it can be seen that although microstrain
the milling progresses, its effect on crystallite size is negligible. Meanwhile, the average crystallite size
increases as the milling progresses, its effect on crystallite size is negligible. Meanwhile, the average
of the Mo powder changes with the opposite trend. With the increase in the milling time from 1 to 3 h,
crystallite size of the Mo powder changes with the opposite trend. With the increase in the milling
the average crystallite size of the Mo powder sharply decreases from 94.6 to 48.2 nm. With the increase
time from 1 to 3 h, the average crystallite size of the Mo powder sharply decreases from 94.6 to 48.2
in the milling time to greater than 3 h, the average crystallite size of the Mo powder continuously
nm. With the increase in the milling time to greater than 3 h, the average crystallite size of the Mo
decreases, achieving the minimum value of 44.6 nm at a milling time of 11 h. However, with the
powder continuously decreases, achieving the minimum value of 44.6 nm at a milling time of 11 h.
However, with the increase in the milling time from 11 to 15 h, the average crystallite size of the Mo
powder exhibits a slight upward trend. Moreover, the sharp decrease in the average crystallite size
of the Mo powder with the increase in the milling time from 1 h to 3 h is related to the fact that the
Materials 2019, 12, 1926 6 of 17

increase in the milling time from 11 to 15 h, the average crystallite size of the Mo powder exhibits a
slight upward trend. Moreover, the sharp decrease in the average crystallite size of the Mo powder
Materials
with the2019, 12, x FOR
increase PEER
in the REVIEW
milling 6 of 17
time from 1 h to 3 h is related to the fact that the powders are severely
crushed and broken by the high-speed grinding balls with a high collision energy. Furthermore, the
powders are severely crushed and broken by the high-speed grinding balls with a high collision
average crystallite size of the Mo powder slightly decreases with the increase in the milling time from
energy. Furthermore, the average crystallite size of the Mo powder slightly decreases with the
3 h to 11 h due to the agglomeration of the powders caused by cold welding. In addition, the increase
increase in the milling time from 3 h to 11 h due to the agglomeration of the powders caused by cold
in the surface energy and plastic deformation of powders, which are caused by working hardening,
welding. In addition, the increase in the surface energy and plastic deformation of powders, which
may also make it difficult to further refine the powders. On the other hand, the more obvious effect of
are caused by working hardening, may also make it difficult to further refine the powders. On the
cold welding turning is the possible reason for the slight increase in the average crystallite size of the
other hand, the more obvious effect of cold welding turning is the possible reason for the slight
Mo powder with the increase in the milling time from 11 h to 15 h. Hence, the Mo powder exhibits the
increase in the average crystallite size of the Mo powder with the increase in the milling time from
finest size at a milling time of 11 h, attributed to achieving the stable state between the powders of
11 h to 15 h. Hence, the Mo powder exhibits the finest size at a milling time of 11 h, attributed to
crush and cold welding.
achieving the stable state between the powders of crush and cold welding.

4. The
Figure 4. Thefull-width
full-widthat at
half-maximum (FWHM),
half-maximum microstrain,
(FWHM), and crystallite
microstrain, size ofsize
and crystallite Mo with milling
of Mo with
from 0 h to 15 h.
milling from 0 h to 15 h.

3.2. EDS Analysis of Powders at Different Milling Times

3.2. EDS Analysis of Powders at Different Milling Times
To reveal the effect of the milling time on the distribution of the Mo, Ni, and B powders in milled
To reveal the effect of the milling time on the distribution of the Mo, Ni, and B powders in milled
powders, EDS mapping analysis was employed. Figure 5 shows the surface distributions of the Mo,
powders, EDS mapping analysis was employed. Figure 5 shows the surface distributions of the Mo,
Ni, and B elements in milled powders at milling times of 0 h, 1 h, 3 h, 7 h, 11 h, and 15 h. Blue, green,
Ni, and B elements in milled powders at milling times of 0 h, 1 h, 3 h, 7 h, 11 h, and 15 h. Blue, green,
and red areas represent Mo, Ni, and B elements in Figure 5, respectively. With the increase in the
and red areas represent Mo, Ni, and B elements in Figure 5, respectively. With the increase in the
milling time from 0 h to 3 h, clearly, the regions of the three colors were remarkably homogenized,
milling time from 0 h to 3 h, clearly, the regions of the three colors were remarkably homogenized,
indicating that the distributions of Mo, Ni, and B powders clearly improved, in addition to the ball
indicating that the distributions of Mo, Ni, and B powders clearly improved, in addition to the ball
milling. This phenomenon is probably related to the fact that raw powders are crushed into fragments
milling. This phenomenon is probably related to the fact that raw powders are crushed into fragments
under the impact of high-speed balls and are evenly mixed in the vial as a result of the movement of
under the impact of high-speed balls and are evenly mixed in the vial as a result of the movement of
the balls. Moreover, the EDS results observed at 7 h and 11 h revealed that the blue and green regions
the balls. Moreover, the EDS results observed at 7 h and 11 h revealed that the blue and green regions
are continuously homogenized with smaller particles than the previous time, indicating that Mo and
are continuously homogenized with smaller particles than the previous time, indicating that Mo and
Ni powders are not only more evenly distributed, but also gradually become finer. Meanwhile, the
Ni powders are not only more evenly distributed, but also gradually become finer. Meanwhile, the
distribution of B powders does not exhibit any clear improvement with the increase in the milling time
distribution of B powders does not exhibit any clear improvement with the increase in the milling
to greater than 3 h because it is difficult to crush the B powders and further make them thin due to their
time to greater than 3 h because it is difficult to crush the B powders and further make them thin due
extremely high hardness, causing slight homogenization. Furthermore, as the milling time reaches
to their extremely high hardness, causing slight homogenization. Furthermore, as the milling time
15 h, the improved homogenization of the Mo and Ni powders becomes less obvious than before as
reaches 15 h, the improved homogenization of the Mo and Ni powders becomes less obvious than
the regions of elements no longer tend to be homogenized, mainly because the homogenization of the
before as the regions of elements no longer tend to be homogenized, mainly because the
milled powder achieves the limitation during milling.
homogenization of the milled powder achieves the limitation during milling.
Compositions of milled powders at different time is exhibited in Table 3 by using EMPA. Except
Mo, Ni, and B elements, C, F, and Si elements in impurities also can be detected in the milled powders
(not at 0 h). The source of C and F is the PTFE liner in the milling pot, which is peeled off during
milling, related to the continuous impact of the ultrahard B powders and Si3N4 grinding balls.
Meanwhile, Si elements originate from the damage of the Si3N4 grinding balls, which is related to the
Materials 2019, 12, x FOR PEER REVIEW 7 of 17

ball. Moreover, with the increase in the milling time from 1 h to 15 h, the impurity content increases
to some degree. In particular, with the increase in the milling time to greater than 11 h, the impurity
content rapidly increases. Notably, to reduce the impurities, a relatively slow milling speed and a
Materials 2019, 12, 1926
low ball-to-powder 7 of 17
weight ratio are utilized for ball milling although this may lead to a lower
limitation of size and the distribution of the milled powder.

Figure 5. Energy-dispersive
Figure 5. Energy-dispersive X-ray X-ray spectrometry
spectrometry (EDS)
(EDS) mapmap analysis
analysis of
of the
the milled
milled powders
powders of
of milling
milling
times:
times: (a)
(a) 00 h,
h, (b)
(b) 11 h,
h, (c)
(c) 33 h,
h, (d)
(d) 77 h,
h, (e)
and11(e)
h,11
and
h, (f)
(f) 15
15 h.
h.

Compositions of milled powders at different time is exhibited in Table 3 by using EMPA. Except
Table 3. Compositions of the milled powders at different milling times.
Mo, Ni, and B elements, C, F, and Si elements in impurities also can be detected in the milled powders
(not at 0 h). The source of C and F is the PTFE liner Compositions (wt%)
in the milling pot, which is peeled off during milling,
Milled Powders
Mo
related to the continuous impact of the ultrahard Ni B andCSi3 N4 Fgrinding
B powders Si balls. Meanwhile, Si
elements originate from the damage
0h of the Si3 N433.02
61.26 grinding balls, which
5.72 - is- related
- to the mutual impact
between the grinding balls and 1 h the wear60.50
of the 32.66
ultrahard B powders
5.48 1.36 to- the grinding
- ball. Moreover,
with the increase in the milling
3 h time from 1 h to30.24
57.43 15 h, the impurity
5.20 4.28 content
1.17 increases
1.68 to some degree.
In particular, with the increase
7 h in the milling
56.75 time to greater
28.64 5.15 than
6.0711 h,1.52the impurity
1.87 content rapidly
increases. Notably, to reduce 11the
h impurities,
55.86a relatively slow milling
27.21 5.07 7.62 speed
2.18 and 2.06a low ball-to-powder
weight ratio are utilized for15ball
h milling 53.97
although this may
25.83 4.69 lead to a lower
10.04 3.16 limitation
2.31 of size and the
distribution of the milled powder.
3.3. Microstructure of Powders at Different Milling Times
To investigate the effect of the milling time on the microstructure of the milled powders, the
morphological images of the milled powders at milling times of 0 h, 1 h, 3 h, 7 h, 11 h, and 15 h were
Materials 2019, 12, 1926 8 of 17

Table 3. Compositions of the milled powders at different milling times.

Compositions (wt%)
Milled Powders
Mo Ni B C F Si
0h 61.26 33.02 5.72 - - -
1h 60.50 32.66 5.48 1.36 - -
3h 57.43 30.24 5.20 4.28 1.17 1.68
7h 56.75 28.64 5.15 6.07 1.52 1.87
11 h 55.86 27.21 5.07 7.62 2.18 2.06
15 h 53.97 25.83 4.69 10.04 3.16 2.31

3.3. Microstructure of Powders at Different Milling Times

To investigate the effect of the milling time on the microstructure of the milled powders, the
morphological images of the milled powders at milling times of 0 h, 1 h, 3 h, 7 h, 11 h, and 15 h were
recorded (Figure 6). B powders exhibited irregular morphologies before milling, while Ni and Mo
exhibit a globular morphology (Figure 6a). Moreover, the particles sizes of Mo, Ni, and B powders
clearly decreased at the early stage of milling, with 1 h and 3 h (Figure 6b,c). Besides, the morphologies
of B powders changed to a near-square morphology as the milling time reached 3 h, possibly related
to the high brittleness of B powders. This brittleness in turn caused the B powders to be crushed to
near-square fragments by the impact of hard Si3 N4 grinding balls. Furthermore, with the increase
in the milling time up to 7 h and 11 h (Figure 6d,e), Ni and Mo powders were continuously refined,
mainly related to the occurrence of plastic deformation of the soft Mo and Ni powders caused by
the sustained impact of high-speed grinding balls. Eventually, the big flakes in the powders were
crushed into fragments, indicating that the milled powders become more brittle with the increase in
the milling times from 3 h to 11 h. Besides, this phenomenon is consistent with the result reported
by Ozkaya and Canakci [19]. At the initial stage of ball milling, the soft powders became brittle
due to the accumulation of plastic deformation, and this transformation was mainly determined by
the continuous impact between the grinding balls and powders. Meanwhile, the morphologies of B
powders gradually tended to be near-spherical (Figure 6e), which is related to the continuous impact
from the hard Si3 N4 grinding balls. However, by comparing Figure 6e,f, the powders slightly changed
as the milling time reached 15 h, as the powders getting the balance between crushing and cold welding
due to the limitation of the relatively low ball-grinding speed and ball-to-batch ratio. In addition, Ni
and Mo powder agglomerates were observed, which is consistent with the increase tendency of the
average crystallite size in Figure 4.
The laser particle sizer was utilized to further estimate the particle size of the milled powders
at 11 h, and Figure 7a shows the result. The particle size mainly ranged between 3 µm and 10 µm.
The particle size interval was considerably greater than the theoretical size of Mo powders calculated
by Equation (2). To explain this anomaly, the morphology of the cross-section of milled powders
at 11 h was observed (Figure 7b). This figure revealed that the particles of milled powders were
aggregated from fine particles of Mo, Ni, and B powders. This agglomeration caused the experimental
values for the powder size to be greater than the calculated ones. Ozkaya and Canakci [19] have
reported that the properties of milled powders, including size, distribution, and morphology, crucially
affect the microstructure and mechanical properties of the ceramics prepared by powder metallurgy.
To epitomize, with the increase in the milling time from 1 to 7 h, the milled powders became finer in
addition to the more uniform distribution. With the increase in the milling time to 11 h, the milled
powders did not clearly exhibit further refinement. As the milling time reached 15 h, agglomeration
was observed in the milled powders. Hence, a milling time of 11 h was selected as the optimum time
in mechanical ball milling for the preparation of Mo2 NiB2 -Ni cermets.
determined by the continuous impact between the grinding balls and powders. Meanwhile, the
morphologies of B powders gradually tended to be near-spherical (Figure 6e), which is related to the
continuous impact from the hard Si3N4 grinding balls. However, by comparing Figure 6f and Figure
6e, the powders slightly changed as the milling time reached 15 h, as the powders getting the balance
between crushing and cold welding due to the limitation of the relatively low ball-grinding speed
Materials 2019, 12, 1926 9 of 17
and ball-to-batch ratio. In addition, Ni and Mo powder agglomerates were observed, which is
consistent with the increase tendency of the average crystallite size in Figure 4.

Materials 2019, 12, x FOR PEER REVIEW 9 of 17

Figure 6. SEM pictures of milled powders of milling times: (a) 0 h, (b) 1 h, (c) 3 h, (d) 7 h, (e) 11 h, and
Figure 6. SEM pictures of milled powders of milling times: (a) 0 h, (b) 1 h, (c) 3 h, (d) 7 h, (e) 11 h, and
(f) 15 h.
(f) 15 h.

Figure 7. Particle size (a) and SEM picture of cross section (b) of milled powders at 11 h.

The laser particle sizer was utilized to further estimate the particle size of the milled powders at
Materials 2019,6.12,
Figure SEM1926pictures of milled powders of milling times: (a) 0 h, (b) 1 h, (c) 3 h, (d) 7 h, (e) 11 h, and10 of 17

(f) 15 h.

Particlesize
Figure7.7.Particle
Figure size(a)
(a)and
andSEM
SEMpicture
pictureof
ofcross
crosssection
section(b)
(b)of
ofmilled
milledpowders
powdersat
at11
11h.
h.

3.4. Phase Composition of Mo2 NiB2 -Ni Cermets of Powders at Different Milling Times
The laser particle sizer was utilized to further estimate the particle size of the milled powders at
To further
11 h, and Figureinvestigate
7a shows the the result.
impactThe of the milling
particle sizetime on the
mainly performance
ranged betweenof3Mo μm2 NiB 10 μm.
and2 -Ni cermets,
The
bulks were
particle size prepared
interval wasusing milled powders
considerably greateratthan
different milling times
the theoretical size (1
of h,
Mo 7 h, 11 h, and
powders 15 h). XRD
calculated by
patterns were
Equation (2). Torecorded
explain to investigate
this anomaly,the thephase composition
morphology of the of the four cermets
cross-section (Figure
of milled 8). From
powders at 11this
h
figure,
was the Mo2(Figure
observed NiB2 hard
7b).phase and Nirevealed
This figure binder phase
that the were detected
particles ofin all of the
milled four cermets.
powders Moreover,
were aggregated
with fine
from the increase
particlesin ofthe
Mo,milling
Ni, and time from 1 toThis
B powders. 11 h,agglomeration
the peak height of thethe
caused Moexperimental
2 NiB2 phase,values
especially
for
thatpowder
the of the (211)
size topeak, clearlythan
be greater increased, indicative
the calculated of aOzkaya
ones. high peak andintensity.
Canakci [19]Subsequently,
have reported the peak
that
intensity
the of Moofslightly
properties milled weakened as the milling
powders, including size,time reached 15and
distribution, h; this variation was
morphology, possibly
crucially related
affect the
to the incomplete
microstructure andgeneration
mechanical of the Mo2 NiB2 phase
properties of thewith the milled
ceramics powders
prepared by at millingmetallurgy.
powder times of 1 h and To
epitomize,
7 h, due towith the increase
the relatively smallin contact
the milling
area time
betweenfromthe 1 to 7 h, theasmilled
powders well as powders
the uneven became finer in
distribution.
addition to the time
As the milling morereaches
uniform 11distribution.
h, the reactions With forthe increase in
generating thetheMomilling time togradually
2 NiB2 phase 11 h, the became
milled
powders did not clearly
more consummate, exhibittofurther
attributed refinement.
the relatively largeAs the milling
contact time reached
area between the fine15milled
h, agglomeration
powders in
was observed
addition to theinuniform
the milled powders. Hence,
distribution. a milling
Furthermore, timethe
with of continuous
11 h was selected
increaseas in
thethe
optimum
millingtimetime
in
tomechanical
up to 15 h, ball milling forwere
the reactions the preparation of Mo2NiB
slightly restrained, due 2-Ni cermets.
to the powder agglomeration. Hence, the
reaction completion of the Mo2 NiB2 phase reaction for the milled powders at a milling time of 15 h
3.4.
wasPhase
worse Composition
than that at of Mo 2NiB2-Ni
a milling Cermets
time of 11ofh.Powders
Besides, at aDifferent
small amountMilling of
Times
the third phase was also
detected,
To further investigate the impact of the milling time on the performance ofthat
which was examined to be Mo-Ni (Figure 7). Yuan et al. [14] reported Mo2the
NiBMo2-Ni2 NiB 2 hard
cermets,
phasewere
bulks is formed
prepared before
usingthemilled
liquid powders
phase byatthe reaction
different Mo + times
milling B = MoB (1 h,and
7 h,later
11 h,byand 15 h).+XRD
2MoB Ni =
patterns were recorded to investigate the phase composition of the four cermets (Figure 8). From than
Mo 2 NiB 2 . However, the formation mechanism of the Mo 2 NiB 2 phase at temperatures greater this
900 ◦ C the
figure, Mo2+NiB
is 2Mo Ni2 +hard
2B +phase
Mo2 NiB and2 =Ni 2Mo 2 NiB2phase
binder , wherewere the previously
detected informedall of Mo NiB2 cermets.
the 2four particles
act as seedwith
Moreover, crystals, allowinginMo,
the increase the Ni, and B
milling elements
time from 1totodirectly
11 h, theproduce new particles.
peak height of the MoBased on the
2NiB2 phase,
above findings, the Mo-Ni phase is not generated with the formation of the Mo2 NiB2 phase. Kubliy et
al. [13] have reported that the Mo–Ni phase is produced by the dissolution of Mo in Ni at the Ni-rich
region between 1160 ◦ C and 1231 ◦ C, which is consistent with the extensive homogeneity region of the
nickel-based solid solution in a Mo-Ni binary system. This theory is consistent with the variation trend
of the peak intensity of Mo-Ni, which gradually decreases in addition to the more uniform distribution
of milled powders with the increase in the milling times from 1 to 11 h. The uniform distribution of
milled powders can reduce the high Ni content region during sintering. On the other hand, the Ni-rich
regions may also be formed during the sintering of the cermets with milled powders at 15 h, due to the
agglomeration as a result of long-time milling.
Ni at the Ni-rich region between 1160 °C and 1231 °C, which is consistent with the extensive
homogeneity region of the nickel-based solid solution in a Mo-Ni binary system. This theory is
consistent with the variation trend of the peak intensity of Mo-Ni, which gradually decreases in
addition to the more uniform distribution of milled powders with the increase in the milling times
from 1 to 11 h. The uniform distribution of milled powders can reduce the high Ni content region
Materials 2019, 12, 1926 11 of 17
during sintering. On the other hand, the Ni-rich regions may also be formed during the sintering of
the cermets with milled powders at 15 h, due to the agglomeration as a result of long-time milling.

XRD patterns
Figure 8. XRD patterns of Mo22NiB
NiB22-Ni
-Nicermets
cermetsprepared
preparedby
bymilled
milledpowders
powdersof
offour
fourmilling
millingtimes.
times.

To further confirm the phase formation during sintering, DTA analysis was employed using the
milled powders heated to 1250 ◦ C under Ar at 10 ◦ C/min. The curve is relatively flat at temperatures
ranging from 100 to 700 ◦ C (Figure 9). A weak endothermal peak is observed at 520 ◦ C, corresponding
to the volatilization of a small amount of PTFE in milled powders. At temperatures greater than 800 ◦ C,
a wide exothermic peak is observed, which is consistent to the study reported by Bo Yuan. This peak
revealed that raw powders start to react at 800 ◦ C. Moreover, an endothermal peak is observed at
1100 ◦ C; this temperature is in accordance with that of the liquid phase observed according the study
reported
Materials by 12,
2019, Takagi. K. REVIEW
x FOR PEER 11 of 17

◦ C under
Figure 9.
Figure Differentialthermal
9. Differential thermalanalysis
analysis(DTA)
(DTA)cover
cover
ofof milled
milled powders
powders from
from 100100 °C under
to 1250
to 1250 Ar
Ar atmosphere.
atmosphere.

To further confirm the phase formation during sintering, DTA analysis was employed using the
milled powders heated to 1250 °C under Ar at 10 °C/min. The curve is relatively flat at temperatures
ranging from 100 to 700 °C (Figure 9). A weak endothermal peak is observed at 520 °C, corresponding
to the volatilization of a small amount of PTFE in milled powders. At temperatures greater than 800
°C, a wide exothermic peak is observed, which is consistent to the study reported by Bo Yuan. This
peak revealed that raw powders start to react at 800 °C. Moreover, an endothermal peak is observed
Materials 2019, 12, 1926 12 of 17

3.5. Microstructure of Mo2 NiB2 -Ni Cermets of Powders at Different Milling Times
Figure 10 shows the SEM photographs of the Mo2 NiB2 -Ni cermets prepared from milled powders
at milling times of 1 h, 7 h, 11 h, and 15 h. Three phases are clearly observed in this figure: The grey
and dark phases are the Mo2 NiB2 hard phase and the Ni binder phase, respectively. Meanwhile, the
few white third phases are also confirmed to be the Mo-Ni phase. In addition, voids and cracks are
clearly observed in specimens at milling times of 1 h and 7 h (Figure 10a,b). The possible reason for
the defects could be that, during liquid-phase sintering, the densification of cermets mainly depends
Materials
on 2019, 12, x FORflow
the liquid-phase PEERand
REVIEW
particle rearrangement. However, the interdiffusion in the liquid12phase
of 17

is limited due to the non-uniform distribution as well as the large particles in the milled powders at
Equation (3) revealed that the saturation solubility is inversely proportional to the particle
milling times of 1 h and 7 h. Besides, the complete liquid-phase sintering process depends on the
radius. Hence, the solubility increases with the fine powder. In other words, the relative density of
wettability and solubility between the milled powders and the liquid phase, and these two points are
Mo2NiB2-Ni increases along with the finer milled powders. Hence, voids in the specimens with
related to the size of the milled powders. According to the Young equation [13], only γSL < γL + γS
powders milled at a milling time of 11 h (Figure 10c) clearly decrease as the interdiffusion enhanced
(γSL is the interfacial energy between the solid and liquid phases, γL and γS are the specific surface
during sintering with the powders becomes fine and uniform. Hence, the densification of the
energies of the liquid and solid phases, respectively), the wettability of the liquid phase with solid
specimen at a milling time of 11 h remarkably increases, with barely any defects. Furthermore, by
particles is good. Thus, the wettability can be improved in the case of finer powders, attributed to the
comparing Figure 10d with Figure 10c, the grain growth of Mo2NiB2 is clearly significant as the
decrease of γSL . Moreover, the relationship between the solubility of particles and their size can be
milling time reaches 15 h, but the specimen is not densified further. The reason for the merging and
expressed as follows [13]:
growing of Mo2NiB2 grains is that the high surface 2γSL δ3energy and the relatively larger area of contact
1
between the powders caused by long-time ∆L =balling limit
· · L∞
the viscous flow of the liquid phase and(3)
kT r
shrinkage of voids.

Figure 10.10. SEM

SEM pictures
pictures of
of Mo
Mo2NiB 2-Ni cermets prepared by milled powders of four milling times: (a)
Figure 2 NiB2 -Ni cermets prepared by milled powders of four milling times: (a)
1 h, (b) 7h, (c) 11 h, and (d) 15
1 h, (b) 7h, (c) 11 h, and (d) 15 h.h.

To further investigate the effect of the milling time on the microstructure of Mo2NiB2-Ni cermets,
the volume ratio and grain size of the Mo2NiB2 hard phase in specimens at different milling times
were estimated (Figure 11). The figure revealed that the volume fraction of the Mo2NiB2 hard phase
significantly increases with milled powders with the increase in the milling time to 11 h, followed by
a slight decrease with the increase in the milling time up to 15 h. In addition, the volume fraction of
the hard phase reaches the maximum with milled powders at a milling time of 11 h, which is
consistent with the change in the peak intensity of the Mo2NiB2 phase in the above-mentioned XRD
Materials 2019, 12, 1926 13 of 17

Here ∆L is the saturated solubility, γSL is the interfacial energy, T is the temperature, and r is the
particle radius.
Equation (3) revealed that the saturation solubility is inversely proportional to the particle radius.
Hence, the solubility increases with the fine powder. In other words, the relative density of Mo2 NiB2 -Ni
increases along with the finer milled powders. Hence, voids in the specimens with powders milled at
a milling time of 11 h (Figure 10c) clearly decrease as the interdiffusion enhanced during sintering with
the powders becomes fine and uniform. Hence, the densification of the specimen at a milling time
of 11 h remarkably increases, with barely any defects. Furthermore, by comparing Figure 10d with
Figure 10c, the grain growth of Mo2 NiB2 is clearly significant as the milling time reaches 15 h, but
the specimen is not densified further. The reason for the merging and growing of Mo2 NiB2 grains is
that the high surface energy and the relatively larger area of contact between the powders caused by
long-time balling limit the viscous flow of the liquid phase and shrinkage of voids.
To further investigate the effect of the milling time on the microstructure of Mo2 NiB2 -Ni cermets,
the volume ratio and grain size of the Mo2 NiB2 hard phase in specimens at different milling times
were estimated (Figure 11). The figure revealed that the volume fraction of the Mo2 NiB2 hard phase
significantly increases with milled powders with the increase in the milling time to 11 h, followed by a
slight decrease with the increase in the milling time up to 15 h. In addition, the volume fraction of the
hard phase reaches the maximum with milled powders at a milling time of 11 h, which is consistent
with the change in the peak intensity of the Mo2 NiB2 phase in the above-mentioned XRD analysis
results of cermets. Meanwhile, the grain size of the Mo2 NiB2 hard phase gently decreases with milled
powders with the increase in the milling time from 1 h to 7 h and then clearly decreases with the
increase in the milling time from 7 to 11 h. However, the grain size significantly increases as the milling
time reaches 15 h. The possible reason for these variation trends is that the specific surface area of
Materials 2019, 12, x FOR PEER REVIEW 13 of 17
the powders increases with the refining, and the wettability and solubility of the milled powders in
the liquid phase are simultaneously improved. Hence, reactions are sufficiently carried out during
powders in the liquid phase are simultaneously improved. Hence, reactions are sufficiently carried
the solution–precipitation stage, leading to the increase in the volume fraction of the Mo2 NiB2 hard
out during the solution–precipitation stage, leading to the increase in the volume fraction of the
phase. In addition, with the more adequate reactions of the Mo2 NiB2 hard phase, more crystal nuclei
Mo2NiB2 hard phase. In addition, with the more adequate reactions of the Mo2NiB2 hard phase, more
of Mo NiB2 are formed in the liquid phase. Thus, the size of Mo2 NiB2 particles reaches the minimum
crystal2 nuclei of Mo2NiB2 are formed in the liquid phase. Thus, the size of Mo2NiB2 particles reaches
with cermets prepared using powders at a milling time of 11 h.
the minimum with cermets prepared using powders at a milling time of 11 h.

Figure 11. The

The morphology
morphology evolution
evolution of Mo22NiB
NiB22-Ni
-Nicermets
cermetsprepared
prepared by
by milled
milled powders
powders of
of four
four
milling times.

Figure 12 shows the EPMA map analysis of Mo22NiB NiB22-Ni

-Nicermets
cermetsprepared
prepared using
using milled
milled powders
powders
at a milling time of 11 h, indicative of the elemental distribution in cermets. The distributions
distributions of the
Mo22NiB
NiB22hard
hardphase
phaseand
andthe
theNi
Nibinder
binderphase
phaseininthe
thecermets
cermetsare
areclearly
clearlyuniform
uniform(Figure
(Figure12).
12).Moreover,
Moreover,
the chemical constituents and elemental composition of different phases in the specimen are
confirmed by EPMA point analysis. Table 4 summarizes the results obtained: The atomic ratio of Mo,
Ni, and B in the Mo2NiB2 hard phase is ~2:1:2, which is consistent with the theoretical composition of
the Mo2NiB2 hard phase. Besides, few Mo elements are detected in the Ni binder phase due to the
interdiffusion of elements through a liquid phase during solid-phase dissolution–reprecipitation. In
Materials 2019, 12, 1926 14 of 17

the chemical constituents and elemental composition of different phases in the specimen are confirmed
by EPMA point analysis. Table 4 summarizes the results obtained: The atomic ratio of Mo, Ni, and B in
the Mo2 NiB2 hard phase is ~2:1:2, which is consistent with the theoretical composition of the Mo2 NiB2
hard phase. Besides, few Mo elements are detected in the Ni binder phase due to the interdiffusion
of elements through a liquid phase during solid-phase dissolution–reprecipitation. In addition, the
elemental composition of the third phase is further confirmed by the above analysis as the Mo-Ni
phase
Materialsis2019,
formed by the
12, x FOR dissolution
PEER REVIEW of Mo in Ni. 14 of 17

12. Electron probe

Figure 12. probe microanalyzer
microanalyzer (EPMA)
(EPMA) map
map analysis
analysis of
of Mo
Mo22NiB22-Ni cermets prepared by
11 h.
milled powders of milling 11 h.

Table 4. Composition analysis of the three different phases in Figure 12.

3.6. Mechanical Properties of Mo2NiB2-Ni Cermets of Powders at Different Milling Times
Figure 13 shows the densitiesPoint and relativeComposition
densities of(at%)
the specimens prepared using milled
powders at milling times of 1 h, 7 h, 11 h, and Mo 15 h.Ni Clearly, Bthe density and relative density of
Mo2NiB2-Ni cermets prepared using 1milled powders 40.31 at milling
21.36 times of 1 h and 7 h are relatively low
38.33
because large, nonuniform powders limit the interdiffusion in -the liquid phase, leading to a slow
2 3.98 96.02
liquid-phase flow and incomplete 3particle 64.19 35.81
rearrangement. -
Moreover, the densities and relative
densities of the cermets are clearly high for powders milled at a milling time of 11 h. On the other
3.6.
hand,Mechanical Properties
the density of Mo2 NiB
and relative 2 -Ni Cermets
density slightly of Powdersas
decrease at the
Different Milling
milling time Times
reaches 15 h, which is
related to the limitation of the viscous flow of the liquid phase and shrinkage of voids by large
Figure 13 shows the densities and relative densities of the specimens prepared using milled
Mo2NiB2 grains.
powders at milling times of 1 h, 7 h, 11 h, and 15 h. Clearly, the density and relative density of
Mo2 NiB2 -Ni cermets prepared using milled powders at milling times of 1 h and 7 h are relatively
low because large, nonuniform powders limit the interdiffusion in the liquid phase, leading to a
slow liquid-phase flow and incomplete particle rearrangement. Moreover, the densities and relative
densities of the cermets are clearly high for powders milled at a milling time of 11 h. On the other hand,
the density and relative density slightly decrease as the milling time reaches 15 h, which is related to
the limitation of the viscous flow of the liquid phase and shrinkage of voids by large Mo2 NiB2 grains.
Mo2NiB2-Ni cermets prepared using milled powders at milling times of 1 h and 7 h are relatively low
because large, nonuniform powders limit the interdiffusion in the liquid phase, leading to a slow
liquid-phase flow and incomplete particle rearrangement. Moreover, the densities and relative
densities of the cermets are clearly high for powders milled at a milling time of 11 h. On the other
hand, the density and relative density slightly decrease as the milling time reaches 15 h, which is
Materials 2019, 12, 1926
related to the limitation of the viscous flow of the liquid phase and shrinkage of voids by15large
of 17

Mo2NiB2 grains.

Figure 13. Density and relative density of Mo

Mo22NiB22-Ni
-Ni cermets
cermets prepared
prepared by
by milled
milled powders
powders of four
Materials 2019, 12, x FOR PEER REVIEW 15 of 17
milling times.

The mechanical properties, such as the hardness and bending strength of Mo2NiB2-Ni cermets
The mechanical properties, such as the hardness and bending strength of Mo2 NiB2 -Ni cermets
prepared using milled powders at different milling times, are measured. Figure 14 shows the results
prepared using milled powders at different milling times, are measured. Figure 14 shows the results
obtained. The figure indicated that the hardness clearly increases with the increase in the milling time
obtained. The figure indicated that the hardness clearly increases with the increase in the milling
from 1 h to 11 h. Subsequently, the hardness slightly decreases as the milling time reaches 15 h. In
time from 1 h to 11 h. Subsequently, the hardness slightly decreases as the milling time reaches 15 h.
addition, the variation trend of the bending strength of Mo2NiB2-Ni cermets with the milling time is
In addition, the variation trend of the bending strength of Mo2 NiB2 -Ni cermets with the milling time
similar to the trend of hardness. Moreover, the trends of the mechanical properties revealed that
is similar to the trend of hardness. Moreover, the trends of the mechanical properties revealed that
cermets prepared using milled powders at a milling time of 11 h exhibit the best mechanical
cermets prepared using milled powders at a milling time of 11 h exhibit the best mechanical properties.
properties. In other words, the hardness and bending strength of the cermets at a milling time of 11
In other words, the hardness and bending strength of the cermets at a milling time of 11 h reach the
h reach the maximum, possibly related to the fact that the relative density of the specimen at a milling
maximum, possibly related to the fact that the relative density of the specimen at a milling time of 11 h
time of 11 h is clearly improved in comparison with that of the specimen at milling times of 3 h and
is clearly improved in comparison with that of the specimen at milling times of 3 h and 7 h (Figure 10).
7 h (Figure 10). However, the two properties deteriorate as the milling time reaches 15 h, and this
However, the two properties deteriorate as the milling time reaches 15 h, and this observation is
observation is consistent with the growth of the grains for the specimen milled at a milling time of 15
consistent with the growth of the grains for the specimen milled at a milling time of 15 h (Figure 10c).
h (Figure 10c).

14. Mechanical
Figure 14. Mechanical properties of the Mo22NiB22-Ni
-Ni cermets
cermets prepared
prepared by
by milled
milled powders
powders of four
milling times.

Figure 15 shows the fracture morphologies of Mo2NiB2-Ni cermets at four milling times. Cracks
and pores are clearly observed on the fracture surface (Figure 15a,b). The main fracture mode is brittle
fracture with an intergranular fracture. Moreover, intergranular and transgranular fractures are
clearly observed on the fracture surface of cermets at a milling time of 11 h (Figure 15c). In addition,
tearing ridges are observed (Figure 15c). In addition, a small amount of pull out and debonding is
Materials 2019, 12, 1926 16 of 17
Figure 14. Mechanical properties of the Mo2NiB2-Ni cermets prepared by milled powders of four
milling times.
Figure 15 shows the fracture morphologies of Mo2 NiB2 -Ni cermets at four milling times.
CracksFigure
and 15 shows
pores are the fracture
clearly morphologies
observed of Mosurface
on the fracture 2NiB2-Ni cermets at four milling times. Cracks
(Figure 15a,b). The main fracture mode
and pores are clearly observed on the fracture surface (Figure 15a,b).
is brittle fracture with an intergranular fracture. Moreover, intergranular The main fracture modefractures
and transgranular is brittle
fracture with an intergranular fracture. Moreover, intergranular and transgranular fractures
are clearly observed on the fracture surface of cermets at a milling time of 11 h (Figure 15c). In addition, are
clearly
tearing observed
ridges areonobserved
the fracture surface
(Figure 15c).ofIn
cermets at aamilling
addition, time of 11
small amount of h (Figure
pull 15c).debonding
out and In addition,is
tearing ridges are observed (Figure 15c). In addition, a small amount of pull out and debonding
observed (Figure 15c), indicative of good bending strength. Furthermore, the intergranular fracture of is
observed (Figure 15c), indicative of good bending strength.
abnormal big Mo2 NiB2 particles is observed with no voids (Figure 15d).Furthermore, the intergranular fracture
of abnormal big Mo2NiB2 particles is observed with no voids (Figure 15d).

Materials 2019, 12, x FOR PEER REVIEW 16 of 17

Figure
Figure 15.15. Fracture
Fracture surface
surface ff Mo
Mo22NiB
NiB22-Ni
-Nicermets
cermetsprepared
preparedby
by milled
milled powders
powders of
of four
four milling
milling times:
times:
(a)
(a) 11 h,
h, (b)
(b) 77 h,
h, (c)
(c) 11
11 h,
h, and
and (d)
(d) 15
15 h.
h.

4. Conclusions
4. Conclusions
(1) A
(1) A new
newphasephaseoror a mechanical
a mechanical alloy is not
alloy is detected during
not detected the milling
during of Mo, Ni,
the milling and BNi,
of Mo, powders.
and B
The milledThe
powders. powdersmilled arepowders
crushed areto small fragments,
crushed to smallin addition
fragments, to the more uniform
in addition to thedistribution
more uniformwith
the increase in the milling time from 1 h to 11 h; however, the milled powders
distribution with the increase in the milling time from 1 h to 11 h; however, the milled powders slightly change with the
increasechange
slightly in the milling
with the time to 11 h.
increase in the milling time to 11 h.
(2) Mo NiB -Ni cermets are successfully
(2) Mo2NiB2-Ni cermets are successfully
2 2 prepared
preparedusing milled
using powders
milled at different
powders milling
at different times,
milling
while there is a small amount of the Mo-Ni phase precipitation in specimens milled
times, while there is a small amount of the Mo-Ni phase precipitation in specimens milled at milling at milling times of 1
h, 7 h,of
times and 157hh,due
1 h, andto15the nonuniform
h due distributiondistribution
to the nonuniform and agglomeration of milled powders.
and agglomeration of milledMeanwhile,
powders.
the cermets prepared using milled powders at a milling time of 11 h exhibit
Meanwhile, the cermets prepared using milled powders at a milling time of 11 h exhibit optimal optimal morphology with
an optimum relative density and the smallest crystal
morphology with an optimum relative density and the smallest2crystal size of Mo NiB .
2 size of Mo2NiB2.
(3) The
(3) The mechanical
mechanical properties,
properties, such
such as
as the
the hardness
hardness andand bending
bending strength,
strength, exhibit
exhibit the
the maximum
maximum
value for Mo NiB -Ni cermets prepared using milled powders at a milling
value for Mo22NiB22-Ni cermets prepared using milled powders at a milling time of 11 h, of 87.6 time of 11 h, of 87.6 HRA
HRA
and 1367.3
and 1367.3 MPa,
MPa, respectively.
respectively.

Author Contributions: Formal analysis: L.Z., Y.S. and K.L.; investigation: L.Z., Y.L. and Z.C.; writing—original
araft: L.Z.; writing review & editing: L.Z. and Y.J.; funding acquisition: Z.H.; supervision: Z.H. and Y.J.;
resources: Z.R.

Funding: This research was funded by the National Natural Science Foundation of China (51571159).

Conflicts of Interest: The authors declare no conflict of interest.

Materials 2019, 12, 1926 17 of 17

Author Contributions: Formal analysis: L.Z., Y.S. and K.L.; investigation: L.Z., Y.L. and Z.C.; writing—original
araft: L.Z.; writing review & editing: L.Z. and Y.J.; funding acquisition: Z.H.; supervision: Z.H. and Y.J.; resources:
Z.R.
Funding: This research was funded by the National Natural Science Foundation of China (51571159).
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Takagi, K. High tough boride base cermets produced by reaction sintering. Mater. Chem. Phys. 2001, 67,
214–219. [CrossRef]
2. Takagi, K.; Yamasaki, Y.; Komai, M. High-strength boride base hard materials. J. Solid State Chem. 1997, 133,
243–248. [CrossRef]
3. Takagi, K. Development and application of high strength ternary boride base cermets. J. Solid State Chem.
2006, 179, 2809–2818. [CrossRef]
4. Takagi, K.; Yamasaki, Y.; Hirata, K. Development of high strength complex boride base hard materials
produced by reaction boronizing sintering. J. Jpn. Soc. Powder Powder Met. 1998, 45, 507–514. [CrossRef]
5. Takagi, K.; Ohira, S.; Ide, T. New multiple-boride base hard alloy. Met. Powder Rep. 1987, 42, 483–490.
6. Takagi, K.; Komai, M.; Kondo, Y. Microstructure and properties of sintered Mo2 FeB2 base hard alloys. Sci.
Technol. Sinter. Tokyo 1987, 11, 1296–1301.
7. Yamasaki, Y.; Nishi, M.; Takagi, K. Development of very high strength Mo2 NiB2 complex boride base hard
alloy. J. Solid State Chem. 2004, 177, 551–555. [CrossRef]
8. Komai, M.; Isobe, Y.; Ozaki, S. Effects of Cr on the microstructure and mechanical properties of WCoB base
hard alloys. J. Jpn. Soc. Powder Powder Met. 1993, 40, 38–43. [CrossRef]
9. Wang, Y.; Li, Z. Development of ternary-boride-based hard cladding material. Mater. Res. Bull. 2002, 37,
417–423.
10. Koyama, K.; Morishita, M.; Zhang, G. Calcualated Phase Diagrams of the Ni-Mo-B and Ni-W-B Ternary
Systems on the Basis of the Data Obtained by Thermodynamic Measurement of the Related Materials. J. Jpn.
Soc. Powder Powder Met. 2006, 53, 419–429. [CrossRef]
11. Hirata, K.; Iwanaga, K.; Yamasaki, Y. Development of Mo2 NiB2 base cermets with high strength and excellent
corrosion resistance for plastic injection molding machine parts. J. Jpn. Soc. Powder Powder Met. 2006, 53,
447–451. [CrossRef]
12. Hirata, K.; Iwanaga, K.; Yamasaki, Y. Development of High Corrosion Resistant Mo2 NiB2 Boride Base
Cermets for Plastic Injection Molding Machine Parts. In Proceedings of the Materials Science Forum,
Herceg-Novi, Montenegro, 10–14 September 2007.
13. Huang, P.Y. Principle of Powder Metallurgy, 2nd ed.; Machine Press: Beijing, China, 2012.
14. Yuan, B.; Zhang, G.; Kan, Y. Reactive synthesis and mechanical properties of Mo2 NiB2 based hard alloy. Int.
J. Refract. Met. Hard Mater. 2010, 28, 291–296. [CrossRef]
15. Shen, Y.; Huang, Z.F.; Zhang, L. Effect of milling time on the microstructure and mechanical properties of
Mo2 FeB2 based cermets. Mater. Res. Express 2017, 4, 106518. [CrossRef]
16. Zhang, L.; Huang, Z.F.; Shen, Y. Bulk Mo2 FeB2 based cermets fabricated by mechanical ball milling and
reaction boronizing sintering. Powder Metall. Met. Ceram. 2017, 55, 665–675. [CrossRef]
17. Dorset, D.L. X-ray Diffraction: A Practical Approach. Microsc. Microanal. 1998, 4, 513–515. [CrossRef]
18. Khorsand, Z.A. X-ray analysis of ZnO nanoparticles by Williamson–Hall and size–strain plot methods. Solid
State Sci. 2011, 13, 251–256. [CrossRef]
19. Ozkaya, S.; Canakci, A. Effect of the B4 C content and the milling time on the synthesis, consolidation and
mechanical properties of AlCuMg-B4 C nanocomposites synthesized by mechanical milling. Powder Technol.
2016, 297, 8–16. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).