nanomaterials
Article
Improving Powder Magnetic Core Properties via
Application of Thin, Insulating Silica-Nanosheet
Layers on Iron Powder Particles
Toshitaka Ishizaki *, Hideyuki Nakano, Shin Tajima and Naoko Takahashi
Toyota Central R&D Labs., Inc., 41-1 Nagakute, Aichi 480-1192, Japan; [email protected] (H.N.);
[email protected] (S.T.); [email protected] (N.T.)
* Correspondence: [email protected]; Tel.: +81-561-71-7654; Fax: +81-561-63-6135
Academic Editor: Thomas Nann
Received: 19 October 2016; Accepted: 15 December 2016; Published: 23 December 2016
Abstract: A thin, insulating layer with high electrical resistivity is vital to achieving high performance
of powder magnetic cores. Using layer-by-layer deposition of silica nanosheets or colloidal silica over
insulating layers composed of strontium phosphate and boron oxide, we succeeded in fabricating
insulating layers with high electrical resistivity on iron powder particles, which were subsequently
used to prepare toroidal cores. The compact density of these cores decreased after coating with
colloidal silica due to the substantial increase in the volume, causing the magnetic flux density
to deteriorate. Coating with silica nanosheets, on the other hand, resulted in a higher electrical
resistivity and a good balance between high magnetic flux density and low iron loss due to the
thinner silica layers. Transmission electron microscopy images showed that the thickness of the
colloidal silica coating was about 700 nm, while that of the silica nanosheet coating was 30 nm.
There was one drawback to using silica nanosheets, namely a deterioration in the core mechanical
strength. Nevertheless, the silica nanosheet coating resulted in nanoscale-thick silica layers that are
favorable for enhancing the electrical resistivity.
Keywords: powder magnetic core; silica nanosheet; insulating layer; electrical resistivity; iron loss;
magnetic flux density
1. Introduction
Powder magnetic cores have been extensively studied in recent years due to the many advantages
they offer over electromagnetic steel sheets with respect to their isotropic magnetic properties, high
electrical resistivity, flexible design, potential for size reduction, and high design flexibility [1]. In order
to use powder magnetic cores in AC magnetic field applications, however, it is important to reduce
the iron loss, i.e., the sum of eddy current loss and hysteresis loss. Pure iron powders, in particular,
are well known to offer a high magnetic flux density at low cost, but with the drawback of low
electrical resistance and an associated high eddy current loss [2]. Powder magnetic cores are, therefore,
usually fabricated by compacting magnetic powder particles and coating them with insulating layers
to prevent an eddy current from forming. This technique of coating powder particles with appropriate
insulation is crucial to improving the magnetic properties of a core.
Powder magnetic cores generally need to be annealed after press forming to reduce plastic strain,
as this accumulation of strain causes a high hysteresis loss. However, if the insulating layers used
do not have sufficient thermal resistance they may break after annealing, resulting in a decrease in
electrical resistivity and a very high eddy current loss. Although an increase in the thickness of the
insulating layers would enhance their thermal resistance, this would also reduce the magnetic flux
density due to the increased volume of non-magnetic phase. Consequently, the powder particles need
Nanomaterials 2017, 7, 1; doi:10.3390/nano7010001
www.mdpi.com/journal/nanomaterials
Nanomaterials 2017, 7, 1
2 of 13
to be coated with insulating layers that have high thermal and electrical resistance, and which are also
thin and homogeneous.
Several insulating materials have been used to coat powder particles, including silicone resins [3],
phosphates [4–7], silicate glasses [8,9], magnesium oxide [10], aluminum oxide [11–13], ferrites [14–16],
sodium salts [17,18], and silica [19–23]. A homogeneous and thin layer can be realized by using
silicone resin or phosphate, because these materials undergo chemical reactions with the particle
surface. However, the thermal resistance of these materials is not very high, and so they are prone
to breakage during annealing. An insulating layer composed of silicate glass offers a high thermal
resistance, but the thermal stress generated by the difference in the coefficients of thermal expansion
between it and iron causes the magnetic properties to deteriorate. Magnesium and aluminum oxides
also have high thermal resistance, but as the deposition of thin, homogenous layers is challenging,
the magnetic flux density generally deteriorates due to the extra volume of non-magnetic layers.
Ferrites have the advantage of being magnetic and insulating, but their nano-scale deposition with
high coverage is still difficult to achieve by ordinary methods involving the mixing of ferrite particles
or chemical precipitation. An insulating layer composed of sodium salts has a high thermal resistance
and can be formed homogeneously, but pure iron is prone to corrosion by sodium salts. Of the
various insulating materials used for magnetic materials, silica provides excellent stability against
corrosion, oxidation, and a high thermal-resistance [22–26]. Silica coating of magnetic powder particles
is often carried out via the hydrolyzation of tetraethoxysilane (TEOS), but this method makes it
difficult to achieve thin, homogeneous layers of silica [19–23]. An appropriate technique for creating
the homogenous, nanoscale silica layers needed for magnetic powder cores has, to the best of our
knowledge, not yet been reported.
Tajima et al. developed a new type of insulating layer composed of strontium phosphate and
boron, and showed that this has a higher electrical resistivity than conventional insulating layers
composed of phosphates, even after annealing at 673 K (this layer is hereafter referred to as a Sr-B-P-O
insulating layer) [27]. However, the thermal resistance of this Sr-B-P-O insulating layer was still
insufficient to remove strain by annealing at high temperature, and so an attempt was made to enhance
its thermal resistance by mixing silica compounds into iron powders coated with Sr-B-P-O insulating
layers. This resulted in a high electrical resistivity, but at the cost of a substantial decrease in the
compact and magnetic flux densities. The thickness of the silica coating, therefore, needs to be reduced,
and its homogeneity increased, by adopting an effective technique for precise deposition on iron
surfaces. There have, however, been very few studies in which the deposition of thin insulating layers
of silica have been deposited on iron powder particles for fabricating a magnetic core, as there has
been no effective method of achieving thin silica layers with precise control over their deposition.
A new inorganic-layer coating technique that involves the alternating deposition of cationic
polymers and negatively-charged colloids on a solid surface has attracted much attention of
late because it allows the layer thickness to be controlled by simply changing the number of
coatings. This coating technique is based on the layer-by-layer (lbl) assembly of oppositely-charged
polyelectrolyte and colloidal inorganic nanoparticles to create a multilayer film [28,29]. Using this
technique, oxide nanosheets were coated onto inorganic substrates in a colloidal suspension, from
which a variety of unique optical and electronic properties were obtained by precisely controlling
the coating thickness [30–33]. It can, therefore, be hypothesized that thin insulating layers with high
thermal and electrical resistivities could be realized through the precise deposition of nanoscale silica
onto magnetic powder particles.
Although commercial suspensions of colloidal silica are currently available, their particle diameter
is normally more than 10 nm. Thus, in order to reduce the size of silicon-based materials, Nakano et al.
synthesized nanosheets with a nanometer scale thickness composed of silicon and amorphous
silica by exfoliating layered polysilane to create colloidal monolayers [34–36]. If these very thin
silica nanosheets could be coated onto magnetic powder particles, they should provide an effective
insulating layer. To test this idea, silica was coated onto iron powder particles via the lbl method using
Nanomaterials 2017, 7, 1
3 of 13
Nanomaterials 2017, 7, 1
3 of 13
poly(diallyldimethylammonium chloride) (PDADMAC) as the cationic polymer, and the magnetic
properties of the magnetic cores produced from this material were examined.
2. Results and Discussion
2. Results and Discussion
2.1. Colloidal Silica Coating of Iron Powder Particles with and without Sr-B-P-O Insulating Layers
2.1. Colloidal
Coating
Iron Powder Particles
without
Sr-B-P-O
Insulating
Layers of 20–
In orderSilica
to assess
the of
adhesiveness
of silica,with
pureand
iron
powder
particles
with a diameter
160 μm
wereto used
or withoutofSr-B-P-O
insulating
layers,
which
produced
usingµm
a
In order
assess with
the adhesiveness
silica, pure
iron powder
particles
withwas
a diameter
of 20–160
previously
methodSr-B-P-O
[27]. The insulating
colloidal silica
withwhich
a diameter
of 10–20 nm
wasacoated
onto
were used described
with or without
layers,
was produced
using
previously
pure
iron
powder
particles
with
or
without
Sr-B-P-O
insulating
layers
after
coating
of
PDADMAC,
described method [27]. The colloidal silica with a diameter of 10–20 nm was coated onto pure iron
and
this process
repeated
1–5 times.
Toroidal
coreslayers
with outer
innerofdiameters
of 39and
andthis
30
powder
particleswas
with
or without
Sr-B-P-O
insulating
after and
coating
PDADMAC,
mm
andwas
a thickness
5 mm
were
fabricated
bywith
warm
compaction
ofdiameters
the coatedofiron
powders
at and
423
process
repeatedof1–5
times.
Toroidal
cores
outer
and inner
39 and
30 mm
K
with
a
pressure
of
1176
MPa
using
the
die
wall
lubrication
method
[37,38].
These
toroidal
cores
a thickness of 5 mm were fabricated by warm compaction of the coated iron powders at 423 K with
were
then annealed
for 30
minthe
at 773
K in alubrication
nitrogen atmosphere
to remove
a pressure
of 1176 MPa
using
die wall
method [37,38].
These plastic
toroidalstrain.
cores were then
Figure
provides
theKcomparison
the electrical
resistivities
and
compact densities of the
annealed
for130
min at 773
in a nitrogenofatmosphere
to remove
plastic
strain.
annealed
toroidal
cores the
fabricated
from iron
powder
particles
coated and
withcompact
colloidaldensities
silica with
Figure
1 provides
comparison
of the
electrical
resistivities
ofand
the
without
layers from
increasing
numberparticles
of silica coatings
fromcolloidal
1 to 5. Note
when
annealedSr-B-P-O
toroidalinsulating
cores fabricated
iron powder
coated with
silicathat
with
and
the
iron powder
were
directly
coatednumber
with colloidal
the from
electrical
did
not
without
Sr-B-P-Oparticles
insulating
layers
increasing
of silicasilica,
coatings
1 to resistivity
5. Note that
when
increase
appreciably
with
an
increasing
number
of
silica
coatings,
but
the
compact
density
decreased
the iron powder particles were directly coated with colloidal silica, the electrical resistivity did not
proportionally.
Whenwith
Sr-B-P-O
insulating
layersofwere
prior
with
colloidal
silica
increase appreciably
an increasing
number
silicaapplied
coatings,
but to
thecoating
compact
density
decreased
the
electrical
resistivity
increased
by
a
much
greater
extent,
though
the
decrease
in
compact
density
proportionally. When Sr-B-P-O insulating layers were applied prior to coating with colloidal silica
was
much theresistivity
same in both
cases. This
that amount
of colloidal
silica deposited
was density
almost
the electrical
increased
by a indicates
much greater
extent, though
the decrease
in compact
the
of both
whether
Sr-B-P-O
insulating
were
used or not,
but
those particles
with
wassame
muchregardless
the same in
cases.
This indicates
thatlayers
amount
of colloidal
silica
deposited
was almost
Sr-B-P-O
insulating
layers
could
be
coated
with
silica
more
effectively.
A
higher
resistivity
could,
the same regardless of whether Sr-B-P-O insulating layers were used or not, but those particles with
therefore,
be obtained
by coating
over
Sr-B-P-O
insulating
layers,Aand
afterresistivity
every cycle
of
Sr-B-P-O insulating
layers
could besilica
coated
with
silica more
effectively.
higher
could,
coating,
the
compact
density
was
found
to
be
slightly
lower
than
the
particles
without
Sr-B-P-O
therefore, be obtained by coating silica over Sr-B-P-O insulating layers, and after every cycle of coating,
insulating
conclude
from lower
this that
insulating
caninsulating
accelerate
the
the compactlayers.
densityWe
wascan
found
to be slightly
thanSr-B-P-O
the particles
withoutlayers
Sr-B-P-O
layers.
adsorption
of colloidal
silica,
which
should
result inlayers
a core
with
a higher
resistivity
and
We can conclude
from this
that
Sr-B-P-O
insulating
can
accelerate
theelectrical
adsorption
of colloidal
lower
compact
density.
silica, which should result in a core with a higher electrical resistivity and lower compact density.
7.9
colloidal silica
160
Compact density (Mg/m3)
Electrical resistivity (μΩ m)
200
Sr-B-P-O + colloidal silica
120
80
40
0
colloidal silica
7.8
Sr-B-P-O + colloidal silica
7.7
7.6
7.5
7.4
1
3
Number of Coats
(a)
5
1
3
Number of Coats
5
(b)
Figure
Figure 1.
1. (a)
(a)Electrical
Electrical resistivities;
resistivities; and
and (b)
(b) compact
compact densities
densities of
of the
the annealed
annealed toroidal
toroidal cores
cores fabricated
fabricated
from
from iron
iron powder
powder particles
particles coated
coated with
with colloidal
colloidal silica
silica with
with and
and without
without Sr-B-P-O
Sr-B-P-O insulating
insulating layers
layers
increasing
number
of
silica
coatings
from
1–5.
increasing number of silica coatings from 1–5.
The obtained toroidal cores were used for evaluation of the magnetic properties because they
The obtained toroidal cores were used for evaluation of the magnetic properties because they
can prevent the influence of a demagnetizing field. The magnetic properties were measured by DC
can prevent the influence of a demagnetizing field. The magnetic properties were measured by DC
and AC B-H curve tracers. Magnetic flux densities and maximum permeabilities were estimated at a
and AC B-H curve tracers. Magnetic flux densities and maximum permeabilities were estimated at
magnetic field strength of 10 kA/m from the DC B-H curves. Coercivities were also estimated from
a magnetic field strength of 10 kA/m from the DC B-H curves. Coercivities were also estimated from
the DC B-H curves applying a magnetic field strength to be 2 kA/m. Hysteresis and eddy current
the DC B-H curves applying a magnetic field strength to be 2 kA/m. Hysteresis and eddy current
losses were estimated at the maximum magnetic flux density of 1 T and frequency of 400 Hz by an
losses were estimated at the maximum magnetic flux density of 1 T and frequency of 400 Hz by an AC
AC B-H curve tracer. An iron loss was the sum of hysteresis and eddy current losses.
B-H curve tracer. An iron loss was the sum of hysteresis and eddy current losses.
Figure 2 shows the comparison of the iron losses, magnetic flux densities, coercivities and
maximum permeabilities of the annealed toroidal cores fabricated from iron powder particles coated
with colloidal silica with and without Sr-B-P-O insulating layers increasing number of silica coatings
from 1–5. The magnetic hysteresis curves measured by a DC B-H curve tracer are shown in Figures
Nanomaterials 2017, 7, 1
4 of 13
Figure 2 shows the comparison of the iron losses, magnetic flux densities, coercivities and
maximum permeabilities of the annealed toroidal cores fabricated from iron powder particles coated
with colloidal silica with and without Sr-B-P-O insulating layers increasing number of silica coatings
from 1 to 5. The magnetic hysteresis curves measured by a DC B-H curve tracer are shown in
Figures
S1 and 2017,
S2 for
colloidal
Nanomaterials
7, 1the particles with and without Sr-B-P-O insulating layers after coating of4 of
13
silica, respectively. The decreasing trend in magnetic flux density with increasing number of silica
S1 and
for the the
particles
and without
Sr-B-P-O
insulating
colloidal
silica,
coatings
wasS2almost
samewith
whether
Sr-B-P-O
insulating
layerslayers
wereafter
usedcoating
or not,ofbut
as both
coatings
respectively. The decreasing trend in magnetic flux density with increasing number of silica coatings
are non-magnetic, there was greater degradation in magnetic flux density when Sr-B-P-O insulating
was almost the same whether Sr-B-P-O insulating layers were used or not, but as both coatings are
layers were used in addition to silica. This agrees well with the decreasing trend in the compact
non-magnetic, there was greater degradation in magnetic flux density when Sr-B-P-O insulating
densities,
which is closely related to the magnetic flux density. The eddy current losses also decreased
layers were used in addition to silica. This agrees well with the decreasing trend in the compact
with increasing
number
of silica
coatings,
regardless
whether
insulating
layers
were used
densities, which
is closely
related
to the magnetic
fluxof
density.
TheSr-B-P-O
eddy current
losses also
decreased
or not,
and
the
hysteresis
losses
were
also
almost
the
same.
However,
when
the
particles
with
Sr-B-P-O
with increasing number of silica coatings, regardless of whether Sr-B-P-O insulating layers were used
insulating
layers
were
used,
the
eddy
current
loss
after
every
coating
cycle
was
substantially
less than
or not, and the hysteresis losses were also almost the same. However, when the particles with Sr-Binsulatingwithout
layers were
used, the
eddy current
loss This
after every
cycle was
substantially
lesseddy
with P-O
the particles
Sr-B-P-O
insulating
layers.
can becoating
explained
by the
fact that the
than
with
the particles
withoutproportion
Sr-B-P-O insulating
layers. This
can be explained
by the
fact that
thesilica
current
loss
decreases
in inverse
to the electrical
resistivity,
which was
higher
when
eddy
current
loss
decreases
in
inverse
proportion
to
the
electrical
resistivity,
which
was
higher
when
was coated over Sr-B-P-O insulating layers. The coercivities of the cores decreased with increasing
silica was coated over Sr-B-P-O insulating layers. The coercivities of the cores decreased with
number of silica coatings, as did the compact and magnetic flux densities. This suggests that silica
increasing number of silica coatings, as did the compact and magnetic flux densities. This suggests
layers provide a barrier against plastic strain during fabrication, and as coercivity is dependent on
that silica layers provide a barrier against plastic strain during fabrication, and as coercivity is
plastic
strain, iton
is plastic
gradually
reduced
with anreduced
increase
in the
thickness
of the
silica coating
applied.
dependent
strain,
it is gradually
with
an increase
in the
thickness
of the silica
The permeabilities
of
the
cores
also
decreased
with
the
increasing
number
of
silica
coatings,
because
coating applied. The permeabilities of the cores also decreased with the increasing number of silica
a permeability
is generally
reducedisowing
to increase
of an amount
of of
silica.
coatings, because
a permeability
generally
reduced owing
to increase
an amount of silica.
Hysteresis loss
1.7
2000
Magnetic flux
density
1.6
1500
1.5
1000
1.4
500
1.3
0
1.2
1
3
5
1
3
Number of Coats
Coercivity (A/m)
2500
colloidal silica
300
1.8
Magnetic flux density (T)
Iron loss (kW/m3)
Sr-B-P-O + colloidal silica
Eddy current loss
colloidal silica
3000
Sr-B-P-O + colloidal silica
250
200
150
100
Coercivity
50
0
5
1
3
(a)
5
1
3
Number of Coats
5
(b)
Maximum permeability
Sr-B-P-O + colloidal silica
colloidal silica
1000
Maximum
permeability
800
600
400
200
0
1
3
5
1
3
Number of Coats
5
(c)
Figure 2. (a) Iron losses, magnetic flux densities; (b) coercivities; and (c) maximum permeabilities of
Figure 2. (a) Iron losses, magnetic flux densities; (b) coercivities; and (c) maximum permeabilities of
the annealed toroidal cores fabricated from iron powder particles coated with colloidal silica with and
the annealed toroidal cores fabricated from iron powder particles coated with colloidal silica with and
without Sr-B-P-O insulating layers increasing number of silica coatings from 1 to 5.
without Sr-B-P-O insulating layers increasing number of silica coatings from 1 to 5.
The above results suggest that irrespective of the surface condition, the similar amount of silica
is deposited
onto suggest
surfaces that
withirrespective
each increase
in surface
the number
of silica
coatings.
XPS (X-ray
The
above results
of the
condition,
the similar
amount
of silica is
photoelectron
spectroscopy)
analysis
was
conducted
on
uncoated
iron
powder
particles
and
iron
deposited onto surfaces with each increase in the number of silica coatings. XPS (X-ray photoelectron
powder
particles
coated
five
times
by
colloidal
silica
with
and
without
Sr-B-P-O
insulating
layers
in
spectroscopy) analysis was conducted on uncoated iron powder particles and iron powder particles
order to analyze the surface chemical states. The Si2p and Fe2p3/2 core level spectra in Figure 3 show
coated five times by colloidal silica with and without Sr-B-P-O insulating layers in order to analyze
a strong peak corresponding to silica (SiO2) at around 103 eV for the particles with and without Sr-BP-O after coating with colloidal silica, which indicates that silica can be deposited regardless of
whether an insulating layer exists or not. A strong peak corresponding to iron oxides (Fe2O3, Fe3O4)
was observed at around 710–712 eV in the case of the uncoated particles, and this is attributed to
Nanomaterials 2017, 7, 1
5 of 13
the surface chemical states. The Si2p and Fe2p3/2 core level spectra in Figure 3 show a strong peak
corresponding to silica (SiO2 ) at around 103 eV for the particles with and without Sr-B-P-O after coating
with colloidal silica, which indicates that silica can be deposited regardless of whether an insulating
Nanomaterials
of 13
layer
exists or2017,
not.7,A1 strong peak corresponding to iron oxides (Fe2 O3 , Fe3 O4 ) was observed at 5around
710–712 eV in the case of the uncoated particles, and this is attributed to naturally-formed surface
naturally-formed surface oxides. A small peak corresponding to iron oxides was also observed with
oxides. A small peak corresponding to iron oxides was also observed with the particles coated only
the particles coated only with colloidal silica, but this was weaker than that of the uncoated particles.
with colloidal silica, but this was weaker than that of the uncoated particles. This peak corresponding
This peak corresponding to iron oxides was barely visible for the particles coated with colloidal silica
to iron oxides was barely visible for the particles coated with colloidal silica over Sr-B-P-O insulating
over Sr-B-P-O insulating layers. As the analysis depth of XPS is only about 2 or 3 nm, iron oxides
layers. As the analysis depth of XPS is only about 2 or 3 nm, iron oxides cannot be detected if the
cannot be detected if the surface is thoroughly coated with colloidal silica. This means that the
surface
is thoroughly
coatedinsulating
with colloidal
This
means that
the particles
Sr-B-P-O
particles
without Sr-B-P-O
layers silica.
were not
completely
covered
by silica, without
as the uncoated
Nanomaterials 2017, 7, 1
5 of 13
insulating
layers
were not
completely
by silica,
as the uncoated
surface
was detected
by XPS,
surface was
detected
by XPS,
whereascovered
no surface
was exposed
on particles
with Sr-B-P-O
insulating
whereas
no
surface
was
exposed
on
particles
with
Sr-B-P-O
insulating
layers.
This
difference
agrees
naturally-formed
A small
peak
correspondingresults
to iron oxides
was
also observed
with
layers. This
difference surface
agreesoxides.
well with
the
experimental
for the
electrical
resistivities
and
well
with
the
experimental
results
for
the
electrical
resistivities
and
eddy
current
losses
between
the particles
only with
silica, but from
this was
weaker than
that with
of the colloidal
uncoated particles.
eddy current
lossescoated
between
the colloidal
cores prepared
particles
coated
silica with andthe
This peak
corresponding
to iron oxides
barely visible
theand
particles
coatedSr-B-P-O
with colloidal
silica
cores
prepared
from
particles coated
with was
colloidal
silica for
with
without
insulating
layers.
without
Sr-B-P-O
insulating
layers.
over Sr-B-P-O insulating layers. As the analysis depth of XPS is only about 2 or 3 nm, iron oxides
cannot be detected if the surface is thoroughly coated with colloidal silica. This means that the
no coat
noparticles
coat
without Sr-B-P-O insulating layers were not completely covered by silica, as the uncoated
silica
colloidal
surfacesilica
was detected by XPS, whereas no surface was exposed on particles with Sr-B-P-Ocolloidal
insulating
Sr-B-P-O
colloidal
silica agrees well with the experimental results for the electrical resistivities
layers. +This
difference
Sr-B-P-Oand
+
colloidal
silica
eddy current losses between the cores prepared from particles coated with colloidal silica
with and
without Sr-B-P-O insulating layers.
108
107
no coat
no coat
colloidal silica
colloidal silica
Sr-B-P-O + colloidal silica
106
105
104
103
102
101
100
714
713
712
+
710Sr-B-P-O
709
711
colloidal silica
Binding Energy (eV)
Binding Energy (eV)
(a)
(b)
708
707
Figure 3. X-ray photoelectron spectroscopy (XPS) (a) Si2p; and (b) Fe2p3/2 core level spectra of
Figure 3.108X-ray
photoelectron
spectroscopy
(XPS)
(a) Si2p;
and
(b)
Fe2p3/2
core 708
level 707
spectra of
714
713
712
711
710
709
107
106
105
104
103
100
uncoated iron
powder
particles
and 102
iron 101
power
particles coated five times with colloidal silica with
Binding
Energywith
(eV)
uncoated iron powderBinding
particles
times
colloidal silica with
Energyand
(eV) iron power particles coated five
and without Sr-B-P-O insulating layers.
(a) layers.
(b)
and without Sr-B-P-O insulating
Figure
3. X-rayfor
photoelectron
spectroscopy
(XPS) (a)via
Si2p;
Fe2p3/2 core
level
spectra
of
Schematic
models
coating powder
particles
theand
lbl(b)method,
with
and
without
Sr-B-P-O
uncoated
ironfor
powder
particles
and iron particles
power particles
coated
five method,
times with colloidal
silicawithout
with
Schematic
models
coating
powder
via
the
lbl
with
and
insulating layers, are shown in Figure 4. Note that when no Sr-B-P-O insulating layers areSr-B-P-O
used,
and without Sr-B-P-O insulating layers.
insulating
layers,
areadequately
shown inattracted
Figure 4.
Note
that surface
when no
Sr-B-P-O
insulating
layers
are used,
PDADMAC
is not
to the
overall
during
the first
coating step.
However,
PDADMAC
notare
adequately
to the
overall
surface
during
the
first
coating
step. However,
Schematic
models by
forattracted
coating powder
particles
the lbl method,
without
Sr-B-P-O
those areasisthat
covered
PDADMAC
succeed
invia
attracting
silica,with
andand
become
repeatedly
coated
insulating
layers,
are
shown
in
Figure
4.
Note
that
when
no
Sr-B-P-O
insulating
layers
are
used, which
those
areas
areeach
covered
bycycle.
PDADMAC
succeed
in attracting
and
repeatedly
coated
with
silicathat
with
coating
This coating
process
results insilica,
islands
of become
insulating
layers,
PDADMAC
not adequately
attracted
to the overall
surface
during
first coating
step. However,
with
with each
coating
cycle.
This even
coating
process
results
inthe
islands
of insulating
layers, with
which
dosilica
not increase
theiselectrical
resistivity,
though
the
density
does
decrease.
Powder
particles
those areas that are covered by PDADMAC succeed in attracting silica, and become repeatedly coated
insulating
layers,
on
the
other
hand,
attract
much
more
PDADMAC
during
the
first
step,
do Sr-B-P-O
not increase
the
electrical
resistivity,
even
though
the
density
does
decrease.
Powder
particles
with
with silica with each coating cycle. This coating process results in islands of insulating layers, which
with
silica
being
subsequently
deposited
on
these
PDADMAC-covered
surfaces
to
produce
robust
Sr-B-P-O insulating
layers,
on the
other hand,
attract
PDADMAC
duringwith
the first step,
do not increase
the electrical
resistivity,
even though
themuch
density more
does decrease.
Powder particles
insulating
layers.
insulating layers,deposited
on the otheron
hand,
attract
much more PDADMACsurfaces
during theto
first
step,
with
silicaSr-B-P-O
being
subsequently
these
PDADMAC-covered
produce
robust
silica being subsequently deposited on these PDADMAC-covered surfaces to produce robust
insulating with
layers.
PDADMAC (+)
insulating layers.
PDADMAC (+)
PDADMAC (+)
PDADMAC (+)
Silica (-)
Silica (-)
Silica (-)
Silica (-)
Fe
Fe Fe
Fe
Sr-B-P-O
Sr-B-P-O
(a) (a)
(b) (b)
4. Schematic
models
for the
silica
coatingof
of iron
particles
by the
method
(a) without;
Figure 4.Figure
Schematic
models
for the
silica
coating
ironpowder
powder
particles
bylbl
the
lbl method
(a) without;
Figure 4. Schematic
the silica
coating of iron powder particles by the lbl method (a) without;
and (b) withmodels
Sr-B-P-Ofor
insulating
layers.
and (b) with Sr-B-P-O insulating layers.
and (b) with Sr-B-P-O insulating layers.
2.2. Comparison between Coloidal Silica and Silica Nanosheet Coatings
2.2. Comparison between Coloidal Silica and Silica Nanosheet Coatings
Although coating with Sr-B-P-O was very effective in increasing the electrical resistivity of the
toroidal coating
core prepared
silica-coated
iron effective
particles, the
compact andthe
magnetic
fluxresistivity
densities of the
Although
with with
Sr-B-P-O
was very
in increasing
electrical
toroidal core prepared with silica-coated iron particles, the compact and magnetic flux densities
detailed characterization of silica nanosheets was also shown in the same paper [35]. The thickness
per sheet was 0.68 nm and the length was in the range of 100–200 nm, which were observed by TEM
(transmission electron microscopy) and AFM (atomic force microscopy). The selected-area electron
diffraction indicated that the crystal structure was amorphous state.
Nanomaterials 2017, 7, 1
6 of 13
Figure 5 shows cross-sectional TEM images of iron powder particles with Sr-B-P-O insulating
layers after five coatings of colloidal silica and silica nanosheets. These images reveal that the silica
layer
thicknessesbetween
were about
700
nm and
in the
case
of the colloidal
2.2. Comparison
Coloidal
Silica
Silica
Nanosheet
Coatings silica layer, but just 30 nm for the
silica nanosheet layer. This thickness of the silica nanosheet layers was much less than has been
Although
with Sr-B-P-O
was
very effective
in increasing
the electrical
resistivity
of the
achieved
via thecoating
hydrolyzation
of TEOS
[20,21],
which means
that significantly
thinner
silica layers
toroidal
core
prepared
with
silica-coated
iron
particles,
the
compact
and
magnetic
flux
densities
can be obtained by using silica nanosheets compared with the conventional method. We consider
deteriorated
considerably
increasing
number
of colloidal
silica coatings.
The layers
colloidal
that
our process
by the lblwith
method
of silica
nanosheets
over Sr-B-P-O
insulating
hassilica
the
particles
used
in
the
experiment
had
an
average
diameter
of
10–20
nm,
and
so
the
resulting
silica
advantage that deposition of silica can be more precisely controlled, resulting in thinner layers.layers
wereFigure
simply6 too
thickthe
to comparison
achieve a high
compact
density
with increasing
coatingdensities
cycles. Ideally,
shows
of the
electrical
resistivities
and compact
of the
any
increase
in
the
electrical
resistivity
should
not
reduce
the
high
compact
and
magnetic
flux
annealed toroidal cores fabricated from iron powder particles coated only with Sr-B-P-O insulating
densities.
This
can
be
made
possible
by
using
silica
nanosheets
to
coat
the
iron
powder
particles.
layers, and those with five coatings of colloidal silica or silica nanosheets over Sr-B-P-O insulating
Silica nanosheets
synthesized
the same
described
in a previous
paper
layers.
When only were
Sr-B-P-O
insulatingbylayers
were procedure
used, the electrical
resistivity
was low,
but [35].
the
The
detailed
characterization
of
silica
nanosheets
was
also
shown
in
the
same
paper
[35].
The
thickness
compact density was high. The Sr-B-P-O insulating layers were thin, but could break after annealing
per sheetofwas
0.68
nm and the
lengthresistance,
was in theresulting
range of in
100–200
nm, which
were observed
by TEM
because
their
insufficient
thermal
a low electrical
resistivity.
However,
the
(transmission
electron
microscopy)
and
AFM
(atomic
force
microscopy).
The
selected-area
electron
electrical resistivity increased and the compact density decreased significantly after a coating of
diffraction
indicated
that theover
crystal
amorphous
state.
colloidal
silica
was applied
thestructure
Sr-B-P-Owas
insulating
layers.
This indicates that the insulative
Figure
5
shows
cross-sectional
TEM
images
of
iron
powder
Sr-B-P-O
properties were reinforced by colloidal silica, resulting in a higherparticles
electricalwith
resistivity,
butinsulating
the extra
layers
after
five
coatings
of
colloidal
silica
and
silica
nanosheets.
These
images
reveal
that
the silica
silica
volume of silica reduced the compact density because the silica layers were too thick. When
layer
thicknesses
were
about
700
nm
in
the
case
of
the
colloidal
silica
layer,
but
just
30
nm
for
the
silica
nanosheets were coated over Sr-B-P-O insulating layers instead, the electrical resistivity increased
nanosheet
layer.
This
thickness
of
the
silica
nanosheet
layers
was
much
less
than
has
been
achieved
via
remarkably, while the compact density decreased very little compared to the particles with only Srthe hydrolyzation
of TEOS
[20,21],
which
thatwere
significantly
silicalayers
layers thin
can be
obtained
B-P-O
insulating layers.
This
means
that means
not only
the silicathinner
nanosheet
enough
to
by
using
silica
nanosheets
compared
with
the
conventional
method.
We
consider
that
our
process
by
maintain a sufficiently high compact density, they also had a higher thermal resistivity than the Srthe lblinsulating
method oflayers
silica nanosheets
over
Sr-B-P-O
insulating
layers has
the advantage
deposition
B-P-O
and colloidal
silica
over Sr-B-P-O
insulating
layers.
In fact, it isthat
worth
noting
of
silica
can
be
more
precisely
controlled,
resulting
in
thinner
layers.
that the electrical resistivity was more than 800 μΩm, even after annealing.
100 nm
SiO2
SiO2
Fe
100 nm
Fe
(a)
(b)
Figure 5. TEM images of silica layers on iron powder particles with Sr-B-P-O insulating layers coated
Figure 5. TEM images of silica layers on iron powder particles with Sr-B-P-O insulating layers coated
five times by (a) colloidal silica; and (b) silica nanosheets.
five times by (a) colloidal silica; and (b) silica nanosheets.
Figure 6 shows the comparison of the electrical resistivities and compact densities of the annealed
toroidal cores fabricated from iron powder particles coated only with Sr-B-P-O insulating layers,
and those with five coatings of colloidal silica or silica nanosheets over Sr-B-P-O insulating layers.
When only Sr-B-P-O insulating layers were used, the electrical resistivity was low, but the compact
density was high. The Sr-B-P-O insulating layers were thin, but could break after annealing because
of their insufficient thermal resistance, resulting in a low electrical resistivity. However, the electrical
resistivity increased and the compact density decreased significantly after a coating of colloidal silica
was applied over the Sr-B-P-O insulating layers. This indicates that the insulative properties were
reinforced by colloidal silica, resulting in a higher electrical resistivity, but the extra volume of silica
reduced the compact density because the silica layers were too thick. When silica nanosheets were
coated over Sr-B-P-O insulating layers instead, the electrical resistivity increased remarkably, while the
compact density decreased very little compared to the particles with only Sr-B-P-O insulating layers.
This means that not only were the silica nanosheet layers thin enough to maintain a sufficiently high
Nanomaterials 2017, 7, 1
7 of 13
compact density, they also had a higher thermal resistivity than the Sr-B-P-O insulating layers and
colloidal silica over Sr-B-P-O insulating layers. In fact, it is worth noting that the electrical resistivity
Nanomaterials
2017,800
7, 1 µΩm, even after annealing.
7 of 13
was
more than
7.9
Electrical resistivity
Compact density (Mg/m3)
Electrical resistivity (μΩ m)
1000
800
600
400
200
Compact density
7.8
7.7
7.6
7.5
7.4
0
Sr-B-P-O
Sr-B-P-O
+colloidal silica
Sr-B-P-O
Sr-B-P-O
+silica nanosheet
(a)
Sr-B-P-O
+colloidal silica
Sr-B-P-O
+silica nanosheet
(b)
Figure6.6.(a)
(a)Electrical
Electricalresistivities;
resistivities;
and
compact
densities
of annealed
toroidal
fabricated
Figure
and
(b)(b)
compact
densities
of annealed
toroidal
corescores
fabricated
from
from
iron
powder
particles
coated
only
with
Sr-B-P-O
insulating
layers,
and
with
five
coatings
of
iron powder particles coated only with Sr-B-P-O insulating layers, and with five coatings of colloidal
colloidal
silicananosheets
or silica nanosheets
over insulating
Sr-B-P-O insulating
silica
or silica
over Sr-B-P-O
layers. layers.
1.6
200
1.4
100
0
1.2
Sr-B-P-O
Sr-B-P-O
Sr-B-P-O
+colloidal silica +silica nanosheet
Coercivity (A/m)
300
Magnetic flux density (T)
Iron loss (kW/m3)
Figure 7 presents the comparison of the iron losses, magnetic flux densities, coercivities, and
Figure 7 presents the comparison of the iron losses, magnetic flux densities, coercivities, and
maximum permeabilities of the annealed toroidal cores fabricated from iron powder particles coated
maximum permeabilities of the annealed toroidal cores fabricated from iron powder particles coated
only with Sr-B-P-O insulating layers, and with five coatings of colloidal silica or silica nanosheets
only with Sr-B-P-O insulating layers, and with five coatings of colloidal silica or silica nanosheets
over Sr-B-P-O insulating layers. The magnetic hysteresis curves measured by a DC B-H curve tracer
over Sr-B-P-O insulating layers. The magnetic hysteresis curves measured by a DC B-H curve tracer
are shown in Figure S3. In the case of powder particles with only Sr-B-P-O insulating layers, the
are shown in Figure S3. In the case of powder particles with only Sr-B-P-O insulating layers, the
magnetic flux density was high despite a high iron loss caused by the high eddy current loss. This
magnetic flux density was high despite a high iron loss caused by the high eddy current loss. This eddy
eddy current loss was reduced by coating colloidal silica over Sr-B-P-O insulating layers, but the
current loss was reduced by coating colloidal silica over Sr-B-P-O insulating layers, but the magnetic
magnetic flux density also decreased considerably. However, the coercivity also decreased with
flux density also decreased considerably. However, the coercivity also decreased with colloidal silica
colloidal silica coating, because the thick silica layers could prevent the accumulation of plastic strain.
coating, because the thick silica layers could prevent the accumulation of plastic strain. The silica
The silica nanosheet coating, on the other hand, produced the lowest iron loss. The hysteresis loss
nanosheet coating, on the other hand, produced the lowest iron loss. The hysteresis loss
was the
was the similar magnitude between applying only Sr-B-P-O insulating layers (332 3kW/m3), colloidal
similar magnitude
between applying only Sr-B-P-O insulating layers (332 kW/m ), colloidal silica
silica (373 kW/m3) and silica nanosheets over Sr-B-P-O insulating layers (340 kW/m3). However, the
(373 kW/m3 ) and silica nanosheets over Sr-B-P-O insulating layers (3403 kW/m3 ). However, the
eddy current loss with silica nanosheets was significantly low (12 kW/m3 ) than with only Sr-B-P-O
eddy current loss with silica3nanosheets was significantly low
(12 kW/m ) than with only Sr-B-P-O
insulating layers (114 kW/m3 ) and colloidal silica (21 kW/m33). Since an eddy current loss is dependent
insulating layers (114 kW/m ) and colloidal silica (21 kW/m ). Since an eddy current loss is dependent
on a resistivity, the eddy current loss became the lowest value after coating with silica nanosheets. It
on a resistivity, the eddy current loss became the lowest value after coating with silica nanosheets.
is significant for the magnetic core that the iron loss could be reduced about 20% by coating silica
It is significant for the magnetic core that the iron loss could be reduced about 20% by coating silica
nanosheets over Sr-B-P-O insulating layers, resulting in improving the performance. The magnetic
nanosheets over Sr-B-P-O insulating layers, resulting in improving the performance. The magnetic
flux density with silica nanosheet coating over Sr-B-P-O insulating layers was also high due to the
flux density with silica nanosheet coating over Sr-B-P-O insulating layers was also high due to the low
low compact density. These magnetic losses and flux densities agree well with the electrical
compact density. These magnetic losses and flux densities agree well with the electrical resistivities and
resistivities and compact densities for the three powders. The fact that the coercivity was more greatly
compact densities for the three powders. The fact that the coercivity was more greatly reduced with
reduced with silica nanosheet coating over Sr-B-P-O insulating layers compared to using only Sr-Bsilica nanosheet coating over Sr-B-P-O insulating layers compared to using only Sr-B-P-O insulating
P-O insulating layers indicates that that the silica nanosheet layers also prevented the accumulation
layers indicates that that the silica nanosheet layers also prevented the accumulation of plastic strain.
of plastic strain. However, the coercivity after coating with silica nanosheets was slightly higher than
However,
the coercivity
after
coating
withthe
silica
nanosheets
wasofslightly
higher silica
than when
when colloidal
silica was
used,
because
greater
thickness
the colloidal
layerscolloidal
allowed
silica
was
used,
because
the
greater
thickness
of
the
colloidal
silica
layers
allowed
more
plastic
more plastic strain to be removed. The result about the permeabilities also indicates the samestrain
trend
to
removed.
The result
the permeabilities
indicates
the same trend
did the compact
asbe
did
the compact
and about
magnetic
flux densities,also
because
a permeability
is as
dependent
on the
and
magneticratio
flux densities,
composition
of iron. because a permeability is dependent on the composition ratio of iron.
Figure 8 shows the Si2p and Fe2p3/2 core level spectra of iron powder particles that were coated
five times with
colloidal silica or silica nanosheets
insulating layers. The strong peak
600
2.0 over Sr-B-P-O
300
Eddy current loss
that was observed
at around 103
eV after
Hysteresis
loss coating with colloidal silica confirms that SiO2 layers were
500
Magnetic flux density
formed on400
the powder surface [39]. In contrast,1.8a weak200
peak at around 101 eV was observed after
coating with silica nanosheets. This peak obviously represents a shift in the SiO2 peak to a lower
100
Coercivity
0
Sr-B-P-O
Sr-B-P-O
Sr-B-P-O
+colloidal silica +silica nanosheet
Iron loss (kW/m3)
2.0
Eddy current loss
400
Hysteresis
loss
Magnetic flux density
500
400
300
200
200
1.6
0
Sr-B-P-O
100
0
Sr-B-P-O
1.8
Sr-B-P-O
Sr-B-P-O
+colloidal silica +silica nanosheet
300
Coercivity (A/m)
600
600
Magnetic flux density (T)
Maximum permeability
silica (373 kW/m3) and silica nanosheets over Sr-B-P-O insulating layers (340 kW/m3). However, the
eddy current loss with silica nanosheets was significantly low (12 kW/m3) than with only Sr-B-P-O
insulating layers (114 kW/m3) and colloidal silica (21 kW/m3). Since an eddy current loss is dependent
on a resistivity, the eddy current loss became the lowest value after coating with silica nanosheets. It
is significant for the magnetic core that the iron loss could be reduced about 20% by coating silica
Nanomaterials 2017, 7, 1
8 of 13
nanosheets over Sr-B-P-O insulating layers, resulting in improving the performance. The magnetic
flux density with silica nanosheet coating over Sr-B-P-O insulating layers was also high due to the
low
compact
density.
These
losses and
densities
agree well
with
thelower
electrical
binding
energy,
indicating
thatmagnetic
the Si oxidation
stateflux
in the
silica nanosheet
layer
was
than
resistivities
compact densities
for the that
threesilica
powders.
The factwere
that amorphous
the coercivity
was
more
that of SiO2and
. A previous
paper reported
nanosheets
and
had
an greatly
atomic
reduced
withofsilica
nanosheet
Sr-B-P-O
insulating
layersthat
compared
to that
using
only
Sr-Bratio of Si:O
about
1:1.8 [35].coating
This is over
consistent
with
another study
reported
the
peak
for
P-O
insulating
layers
the silica nanosheet
layers
the accumulation
amorphous
silica
wasindicates
observedthat
at a that
lower-energy
than that of
SiO2 also
[39].prevented
As amorphous
silica is also
of
plastic to
strain.
the coercivity
after coating
with
silica nanosheets
highermore
than
reported
be aHowever,
superior insulator,
we believe
that the
electrical
resistivitywas
canslightly
be enhanced
when
colloidal
silica
wasnanosheet
used, because
thethan
greater
the colloidal
silica
layers
allowed
significantly
with
coating
withthickness
a colloidalofsilica
coating [40].
The
Fe2p3/2
core
Nanomaterials
2017,
7,a 1silica
8 of 13
more
plastic strain
toshow
be removed.
Theinresult
about
permeabilities
indicates silica
the same
level spectra
did not
any peaks
the case
of the particles
coatedalso
by colloidal
and trend
silica
800
as
did the over
compact
andinsulating
magnetic
flux densities,
a produced
permeability
is dependent
on the
nanosheets
Sr-B-P-O
layers,
suggestingbecause
that both
a thorough
surface coating
permeability
Maximum
composition
ratio of the
iron.eddy current loss.
that greatly reduced
200
100
1.4 Sr-B-P-O
Sr-B-P-O
+colloidal silica +silica nanosheet
1.2(c)
Coercivity
0
Sr-B-P-O
Sr-B-P-O
Sr-B-P-O
+colloidal silica +silica nanosheet
Nanomaterials
2017,
7, 1Iron losses, magnetic flux densities; (b) coercivities; and (c) maximum permeabilities of
Figure
7. (a)
8 of 13
Maximum permeability
(a)
annealed toroidal cores fabricated
from iron powder particles coated only (b)
with Sr-B-P-O insulating
800of colloidal silica or silica nanosheets over Sr-B-P-O insulating layers.
layers, and with five coatings
Maximum permeability
Figure 8 shows the Si2p and
600 Fe2p3/2 core level spectra of iron powder particles that were coated
five times with colloidal silica or silica nanosheets over Sr-B-P-O insulating layers. The strong peak
that was observed at around 103
400 eV after coating with colloidal silica confirms that SiO 2 layers were
formed on the powder surface [39]. In contrast, a weak peak at around 101 eV was observed after
coating with silica nanosheets.
This peak obviously represents a shift in the SiO2 peak to a lower
200
binding energy, indicating that the Si oxidation state in the silica nanosheet layer was lower than that
of SiO2. A previous paper reported that silica nanosheets were amorphous and had an atomic ratio
0
Sr-B-P-O
Sr-B-P-O
Sr-B-P-O
of Si:O of about 1:1.8 [35]. This is consistent
with
another study
that reported that the peak for
+colloidal silica +silica nanosheet
amorphous silica was observed at a lower-energy than that of SiO2 [39]. As amorphous silica is also
(c) the electrical resistivity can be enhanced more
reported to be a superior insulator, we believe that
significantly
with
a
silica
nanosheet
coating
than
with
a colloidal and
silica
coating
[40].permeabilities
The Fe2p3/2 core
Figure 7.
7. (a)
(a) Iron
Iron losses,
losses, magnetic
magnetic flux
flux densities;
densities; (b)
(b) coercivities;
coercivities; and
(c) maximum
maximum permeabilities
of
Figure
(c)
of
level spectra did not show any peaks in the case of the particles coated by colloidal silica and silica
annealed
toroidal
cores
fabricated
from
iron
powder
particles
coated
only
with
Sr-B-P-O
insulating
annealed toroidal
cores fabricated
from
iron powder
particles
with Sr-B-P-O
insulating
nanosheets
over Sr-B-P-O
insulating
layers,
suggesting
thatcoated
both only
produced
a thorough
surface
layers, and
and with
with five
five coatings
coatings of
of colloidal
colloidal silica
silica or
or silica
silica nanosheets
nanosheets over
over Sr-B-P-O
Sr-B-P-Oinsulating
insulatinglayers.
layers.
layers,
coating that greatly reduced the eddy current loss.
Figure 8 shows the Si2p and Fe2p3/2 core level spectra of iron powder particles that were coated
colloidal silica coating
coating
five timescolloidal
withsilica
colloidal
silica or silica nanosheets over Sr-B-P-O insulating layers. The strong peak
silica nanosheet coating
silica nanosheet coating
that was observed at around 103 eV after coating with colloidal silica confirms that SiO2 layers were
formed on the powder surface [39]. In contrast, a weak peak at around 101 eV was observed after
coating with silica nanosheets. This peak obviously represents a shift in the SiO2 peak to a lower
binding energy, indicating that the Si oxidation state in the silica nanosheet layer was lower than that
of SiO2. A previous paper reported that silica nanosheets were amorphous and had an atomic ratio
of Si:O of about 1:1.8 [35]. This is consistent with another study that reported that the peak for
714
713
712
711
710
709
708
707
108
107
106
105
104
103
102
101
100
amorphous silica was
observed
at
a
lower-energy
than
that
of
SiO
2 [39]. As amorphous silica is also
Binding
Energy
(eV)
Binding Energy (eV)
reported to be a superior insulator, we believe that the electrical resistivity can be enhanced more
(a)
(b)
significantly with a silica nanosheet coating than with a colloidal silica coating [40]. The Fe2p3/2 core
Figure 8.
XPS (a)show
Si2p; and
(b) Fe2p3/2
core level of
spectra
of iron powder
particles
coated five
times
levelFigure
spectra
any
in the
the particles
coated
by colloidal
and silica
8. did
XPS not
(a) Si2p; and
(b)peaks
Fe2p3/2
corecase
level spectra
of iron powder
particles
coatedsilica
five times
with colloidal silica or silica nanosheets over Sr-B-P-O insulating layers.
nanosheets
over silica
Sr-B-P-O
insulating
suggesting
that layers.
both produced a thorough surface
with colloidal
or silica
nanosheetslayers,
over Sr-B-P-O
insulating
coating that greatly reduced the eddy current loss.
The mechanical strength of annealed toroidal cores was measured by radial crushing tests. As
shown in Figure 9, the highest strength was achieved with the core prepared using uncoated powder
colloidal silica coating
colloidal silica coating
particles,
with Sr-B-P-O coating and subsequent silica coating both reducing the strength. The
silica nanosheet coating
silica nanosheet coating
toroidal core fabricated using particles coated with silica nanosheets exhibited the lowest strength of
less than 60 GPa. The reason for this reduction in strength when using silica nanosheets is not yet
understood, and so will be studied in the future.
Nanomaterials 2017, 7, 1
9 of 13
The mechanical strength of annealed toroidal cores was measured by radial crushing tests.
As shown in Figure 9, the highest strength was achieved with the core prepared using uncoated
powder particles, with Sr-B-P-O coating and subsequent silica coating both reducing the strength.
The toroidal core fabricated using particles coated with silica nanosheets exhibited the lowest strength
of less than 60 GPa. The reason for this reduction in strength when using silica nanosheets is not yet
Nanomaterials 2017, 7, 1
9 of 13
understood, and so will be studied in the future.
Radial crushing strength (MPa)
180
150
120
90
60
30
0
no coat
Sr-B-P-O
Sr-B-P-O
+colloidal silica
Sr-B-P-O
+silica
nanosheet
Figure 9.
9. Radial
Radial crushing
crushing strengths
strengths of
of annealed
annealed toroidal
toroidal cores
cores fabricated
fabricated from
from uncoated
uncoated iron
iron powder
powder
Figure
particles,
particles
with
only
Sr-B-P-O
insulating
layers,
and
particles
with
five
coatings
of
colloidal
particles, particles with only Sr-B-P-O insulating layers, and particles with five coatings of colloidal
silica or
or silica
silica nanosheets
nanosheets over
over Sr-B-P-O
Sr-B-P-O insulating
insulating layers.
layers.
silica
The
The above
above results
results prove
prove that
that toroidal
toroidal cores made from iron powder particles
particles with Sr-B-P-O
insulating layers
layers and
andaasilica
silicananosheet
nanosheetcoating
coating
provide
a good
balance
between
loss high
and
provide
a good
balance
between
low low
iron iron
loss and
high
magnetic
flux
density.
These
excellent
properties
are
attributed
to
the
very
thin,
homogeneous
magnetic flux density. These excellent properties are attributed to the very thin, homogeneous and
and
highly
thermal
resistant
that
is obtained
by coating
precisely
than
is possible
with
highly
thermal
resistant
layerlayer
that is
obtained
by coating
moremore
precisely
than is
possible
with other
other
methods.
The
low
thickness
results
in
a
high
compact
density,
while
the
homogeneity
of
the
methods. The low thickness results in a high compact density, while the homogeneity of the coating
coating
results
in
a
high
electrical
resistivity.
Notably,
coating
with
silica
nanosheets
can
overcome
results in a high electrical resistivity. Notably, coating with silica nanosheets can overcome the usual
the
usual problem
trade-offof
problem
of an increasing
of insulating
layersa creating
a higherresistivity
electrical
trade-off
an increasing
thickness thickness
of insulating
layers creating
higher electrical
resistivity
and
lower
magnetic
loss,
but
the
extra
volume
of
insulating
layers
reducing
the
compact
and lower magnetic loss, but the extra volume of insulating layers reducing the compact density
and
density
and
magnetic
magnetic
flux
density.flux density.
Materials and
and Methods
Methods
3. Materials
3.1. Preparation
Preparation of
of Polycation
Polycation and
and Silica
Silica Solutions
Solutions
3.1.
A 20
of PDADMAC
(molecular
weight
200,000–350,000)
produced
by Sigma-Aldrich
A
20 wt
wt%%solution
solution
of PDADMAC
(molecular
weight
200,000–350,000)
produced
by Sigma(St.
Louis,
MO,
USA)
was
used
as
the
polycation
solution.
The
concentration
of
this
PDADMAC
Aldrich (St. Louis, MO, USA) was used as the polycation solution. The concentration of this
solution was solution
adjustedwas
to 0.1
wt % bytomixing
with
the pH
was adjusted
to 11
by
PDADMAC
adjusted
0.1 wtit%
by distilled
mixing itwater,
with and
distilled
water,
and the pH
was
adding 1 to
N 11
KOH
prevent
the
severe
of severe
iron that
occurs inofsolutions
a pH
11.
adjusted
by to
adding
1N
KOH
to corrosion
prevent the
corrosion
iron thatwith
occurs
in <solutions
A
20
wt
%
suspension
of
colloidal
silica
produced
by
Nissan
Chemical
Industries
(Chiyoda,
Tokyo,
with a pH < 11.
Japan)
was
used
for
the
coating
experiment.
The
average
diameter
of
these
colloidal
silica
particles
A 20 wt % suspension of colloidal silica produced by Nissan Chemical Industries (Chiyoda,
was about
10–20
nm.used
The for
concentration
the silica solution
was adjusted
to 2ofwtthese
% bycolloidal
mixing itsilica
with
Tokyo,
Japan)
was
the coatingofexperiment.
The average
diameter
distilled
water,
and
the
pH
was
adjusted
to
11
by
the
addition
of
1
N
KOH
solution.
particles was about 10–20 nm. The concentration of the silica solution was adjusted to 2 wt % by
A itsuspension
containing
silica
nanosheets
was to
synthesized
using aofpreviously
described
mixing
with distilled
water, and
the pH
was adjusted
11 by the addition
1 N KOH solution.
procedure
[35],
whereby
1
g
of
CaSi
crystallites
produced
by
Sigma-Aldrich
(St.
Louis,
MO,
USA) was
2
A suspension containing silica nanosheets was synthesized using a previously described
first
immersed
in
100
mL
of
37%
HCl
at
243
K.
After
stirring
continuously
in
a
dark
room
for seven
procedure [35], whereby 1 g of CaSi2 crystallites produced by Sigma-Aldrich (St. Louis, MO,
USA)
daysfirst
under
an Ar atmosphere,
the37%
mixture
washed
with
concentrated
HCl atin243
K, rinsed
was
immersed
in 100 mL of
HCl was
at 243
K. After
stirring
continuously
a dark
roomwith
for
acetone,
and
then
dried
at
300
K
to
synthesize
laminated
polysilane
(SiH)
.
The
chemical
reaction
for
n
seven days under an Ar atmosphere, the mixture was washed with concentrated HCl at 243 K, rinsed
this
production
of
laminated
polysilane
is
as
follows:
with acetone, and then dried at 300 K to synthesize laminated polysilane (SiH)n. The chemical reaction
for this production of laminated polysilane is as follows:
3CaSi2 + 6HCl → Si6H6 + 3CaCl2
About 0.5 g of polysilane was mixed with 100 mL of sodium dodecylsulfate (SDS,
C12H25OSO3Na) solution produced by Sigma-Aldrich (St. Louis, MO, USA). After shaking this
solution at 100 rpm and 300 K for 7 days, a suspension of silica nanosheets was obtained. This
Nanomaterials 2017, 7, 1
10 of 13
3CaSi2 + 6HCl → Si6 H6 + 3CaCl2
About 0.5 g of polysilane was mixed with 100 mL of sodium dodecylsulfate (SDS, C12 H25 OSO3 Na)
solution produced by Sigma-Aldrich (St. Louis, MO, USA). After shaking this solution at 100 rpm
and 300 K for 7 days, a suspension of silica nanosheets was obtained. This suspension was washed
with distilled water and centrifuged at 4000 rpm for 10 min in order to remove the SDS and HCl.
The concentration of the silica nanosheets was then adjusted to 0.2 wt % by mixing with distilled water,
and the pH was adjusted to 11 by the addition of 1 N KOH solution.
3.2. Coating Process
Water-atomized powders of pure iron (Fe 99.9 wt %) with an average particle diameter of
20–160 µm produced by JFE Steel (Chiyoda, Tokyo, Japan) were employed to fabricate the powder
magnetic cores. Plain powders and powders with Sr-B-P-O insulating layers were prepared in order
to compare the adhesiveness of silica. The Sr-B-P-O coating was carried out using the same method
described in another paper [27]. The ratio of Sr-B-P-O insulating layers to iron powders was 2 wt %.
The powders were coated with silica layers by first mixing 45 g of the iron powder and 50 mL of
PDADMAC solution in a glass bottle by shaking using a rotor at 60 rpm for 20 min. After removing
the PDADMAC solution, the powder was mixed with 50 mL of colloidal silica suspension or silica
nanosheet suspension and shaken by the same process. The powders were then rinsed using distilled
water and collected after removing the distilled water by decantation. This coating process was
repeated 1–5 times to prepare multilayer coatings, after which the powders were dried at 373 K
under vacuum.
3.3. Preparation of Powder Magnetic Cores and Evaluation of Their Magnetic Properties and Mechanical Strength
Powder magnetic cores were prepared via warm compaction by the die wall lubrication
method [37,38]. The heating temperature of the die and powder was 423 K. The compacting pressure
was 1176 MPa. Every specimen was fabricated into a ring shape to create a toroidal core and prevent the
influence of a demagnetizing field; the outer and inner diameters of this core were 39 mm and 30 mm,
and the thickness was 5 mm. Every specimen was annealed for 30 min at 773 K in a nitrogen atmosphere
to remove plastic strain. The compact density after annealing was measured by Archimedes’ method,
and the electrical resistivity was measured with a micro-ohmmeter using a four-point method. Cu wires
with a diameter of 0.5 mm were wounded around the toroidal core. The numbers of wire turns were
180 for excitation and 90 for detection, respectively. The magnetic properties of the toroidal core were
measured by using DC and AC B-H curve tracers with a TRF-5A-PC (Toei Industry, Machida, Tokyo,
Japan) and SY8232 (Iwatsu Test Instruments, Suginami, Tokyo, Japan). Magnetic flux densities and
maximum permeabilities were measured under a magnetic field of 10 kA/m using the DC B-H curve.
Coercivities were also estimated from the DC B-H curves applying a magnetic field strength to be
2 kA/m. Hysteresis and eddy current losses were measured at the maximum magnetic flux density of
1 T and a frequency of 400 Hz using the AC B-H curve tracer. It is known that an iron loss (Pc ) can be
calculated by the sum of a hysteresis loss (Ph ) and an eddy current loss (Pe ) with a frequency (f ) as
a parameter by the following Equation (1):
Pc = Ph × f + Pe × f 2
(1)
Hysteresis and eddy current losses were separately estimated from frequency dependency of iron
loss by the Equation (1). The radial crushing strength of the toroidal core was measured using the
press mode of an electronic universal testing machine (Shimadzu, Kyoto, Kyoto, Japan).
Nanomaterials 2017, 7, 1
11 of 13
3.4. Observation and Surface Analyses
TEM observation with a JEM 2010FEF (JEOL, Akishima, Tokyo, Japan) was conducted to observe
the thickness of the silica layers formed using colloidal silica and silica nanosheets. Samples for
this TEM observation were cross-sectionally polished using the FIB (focused ion beam) method
with a FB-2000A (Hitachi High-Technologies, Minato, Tokyo, Japan). XPS analysis was conducted
to investigate the chemical state of the coating layers and compare the surface conditions of the
uncoated powder particles, powder particles directly coated with colloidal silica, and powder particles
with Sr-B-P-O and colloidal silica or silica nanosheet coatings. The analysis region and depth
were approximately φ100 µm and 2–3 nm, respectively. The XPS instrument used was a Quantera
SXM (ULVAC-PHI, Chigasaki, Kanagawa, Japan) with a monochromatic Al Kα source. Charge-up
compensation was made using C1s (285 eV).
4. Conclusions
When iron powder particles were directly coated by colloidal silica, the compact density of the
toroidal core fabricated from them decreased due to the silica deposition, but the electrical resistivity
could not be increased because the silica layers existed as islands over the iron surface and did not
provide complete insulation. In contrast, Sr-B-P-O insulating layers on iron particles were more
capable of attracting silica and became completely coated, resulting in an effective insulation layer.
The magnetic core fabricated from iron powder particles with Sr-B-P-O insulating layers and colloidal
silica coating exhibited low iron loss due to the high electrical resistivity, but the magnetic flux density
decreased considerably because of the excessively thick silica layers (about 700 nm). When silica
nanosheets were used as the coating material instead of colloidal silica, the silica layer thickness was
reduced to only 30 nm. This thickness was much less than has been achieved by the conventional
method. A lower iron loss could therefore be achieved due to the higher electrical resistivity and
higher magnetic flux density made possible by the thinner silica layers. The drawback to this approach
was that the strength of the core was reduced after silica nanosheet coating. Nevertheless, it can be
concluded that silica nanosheet coating results in amorphous, nanoscale-thickness silica layers that are
favorable for enhancing the electrical resistivity of iron powder magnetic cores.
Supplementary Materials: The following are available online at http://www.mdpi.com/2079-4991/7/1/1/s1,
Figure S1: Magnetic hysteresis curves of the annealed toroidal cores fabricated from uncoated iron powder
particles after coating one, three, and five times with colloidal silica, Figure S2: Magnetic hysteresis curves of the
annealed toroidal cores fabricated from iron powder particles with Sr-B-P-O insulating layers after coating with
one, three, and five times with colloidal silica, Figure S3: Magnetic hysteresis curves of annealed toroidal cores
fabricated from iron powder particles with only Sr-B-P-O insulating layers, and those with colloidal silica and
silica nanosheets coated five times over the Sr-B-P-O insulating layers.
Acknowledgments: The authors would like to thank Kayo Horibuchi, Shoji Hotta, and Takeshi Hattori for
assisting with the TEM observations, radial crushing tests, and technical advice, respectively.
Author Contributions: Toshitaka Ishizaki conceived, designed and performed the experiments and wrote the
paper; Hideyuki Nakano, Shin Tajima and Naoko Takahashi performed the experiments and analyzed the data.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
4.
Bas, J.A.; Calero, J.A.; Dougan, M.J. Sintered soft magnetic materials. Properties and applications. J. Magn.
Magn. Mater. 2003, 254–255, 391–398. [CrossRef]
Wu, L.Z.; Ding, J.; Jiang, H.B.; Chen, L.F.; Ong, C.K. Particle size influence to the microwave properties of
iron based magnetic particulate composites. J. Magn. Magn. Mater. 2005, 285, 233–239. [CrossRef]
Zhang, L.; Yang, B.; Cao, Y.; Yu, R.H. Preparation and magnetic properties of novel hybrid magnetic powder cores.
Mater. Res. Innov. 2014, 18, 610–614. [CrossRef]
Liu, H.J.; Su, H.L.; Geng, W.B.; Sun, Z.G.; Song, T.T.; Tong, X.C.; Zou, Z.Q.; Wu, Y.C.; Du, Y.W. Effect of
particle size distribution on the magnetic properties of Fe-Si-Al powder core. J. Supercond. Nov. Magn. 2016,
29, 463–468. [CrossRef]
Nanomaterials 2017, 7, 1
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
12 of 13
Xie, D.-Z.; Lin, K.-H.; Lin, S.-T. Effects of processed parameters on the magnetic performance of a powder
magnetic core. J. Magn. Magn. Mater. 2014, 353, 34–40. [CrossRef]
Liu, Y.; Yi, Y.; Shao, W.; Shao, Y. Microstructure and magnetic properties of soft magnetic powder cores of
amorphous and nanocrystalline alloys. J. Magn. Magn. Mater. 2013, 330, 119–133. [CrossRef]
Ding, W.; Wang, X.; Feng, Z.K.; Gong, R.Z. FeSiAl powder core coated by Ba-B-P-O system insulating layer.
Appl. Mech. Mater. 2013, 303–306, 2469–2473.
Ding, W.; Wu, R.; Xiu, Z.Y.; Chen, G.Q.; Song, J.B.; Liao, Y.Q.; Wu, G.H. Effect of iron particle size and volume
fraction on the magnetic properties of Fe/silicate glass soft magnetic composites. J. Supercond. Nov. Magn.
2014, 27, 435–441. [CrossRef]
Ding, W.; Jiang, L.T.; Ma, K.; Yang, W.S.; Wu, G.H. Processing and properties of soda lime silica glass glass
insulation soft magnetic composite. Optoelectron. Adv. Mater. 2014, 8, 68–71.
Dhokey, N.B.; Patil, S.; Dhandare, S.; Bandal, V.S. Role of ceramic coating on electrical and magnetic
properties of iron powder. Electron. Mater. Lett. 2014, 10, 591–596. [CrossRef]
Peng, Y.D.; Nie, J.W.; Zhang, W.J.; Ma, J.; Bao, C.X.; Cao, Y. Effect of the addition of Al2 O3 nanoparticles on
the magnetic properties of Fe soft magnetic composites. J. Magn. Magn. Mater. 2016, 399, 88–93. [CrossRef]
Wang, J.; Fan, X.; Wu, Z.Y.; Li, G.Q. Synthesis, microstructure and magnetic properties of Fe3 Si0.7 Al0.3 @SiO2
core-shell particles and Fe3 Si/Al2 O3 soft magnetic composite core. J. Solid State Chem. 2015, 231, 152–158.
[CrossRef]
Yaghtin, M.; Taghvaei, A.H.; Hashemi, B.; Janghorban, K. Effect of heat treatment on magnetic properties
of iron-based soft magnetic composites with Al2 O3 insulation coating produced by sol-gel method.
J. Alloys Compd. 2013, 581, 293–297. [CrossRef]
Wang, M.G.; Zan, Z.; Deng, N.; Zhao, Z.K. Preparation of pure iron/Ni-Zn ferrite high strength soft magnetic
composite by spark plasma sintering. J. Magn. Magn. Mater. 2014, 361, 166–169. [CrossRef]
Wu, S.; Sun, A.; Xu, W.; Zhang, Q.; Zhai, F.; Logan, P.; Volinsky, A.A. Iron-based soft magnetic composites
with Mn–Zn ferrite nanoparticles coating obtained by sol-gel method. J. Magn. Magn. Mater. 2012, 324,
3899–3905. [CrossRef]
Streckova, M.; Hadraba, H.; Bures, R.; Faberova, M.; Roupcova, P.; Kubena, I.; Medvecky, L.; Girman, V.;
Kollar, P.; Fuzer, J.; et al. Chemical synthesis of nickel ferrite spinel designed as an insulating bilayer coating
on ferromagnetic particles. Surf. Coat. Technol. 2015, 270, 66–76. [CrossRef]
Wei, D.; Wang, X.; Nie, Y.; Feng, Z.; Gong, R.; Chen, Y.; Harris, V.G. Low loss sendust powder cores comprised
of particles coated by sodium salt insulating layer. J. Appl. Phys. 2015, 117, 17A921-1-3. [CrossRef]
Wei, D.; Wang, X.; Feng, Z.K.; Gong, R.Z. Effect of annealing temperature on the soft magnetic behavior of
Sendust powder core. Appl. Mech. Mater. 2014, 556–562, 6–10. [CrossRef]
Zhao, Y.; Zhang, X.; Xiao, J.Q. Submicrometer laminated Fe/SiO2 soft magnetic composites—An effective
route to materials for high-frequency applications. Adv. Mater. 2005, 17, 915–918. [CrossRef]
Yang, B.; Wu, Z.; Zou, Z.; Yu, R. High-performance Fe/SiO2 soft magnetic composites for low-loss and
high-power applications. J. Phys. D 2010, 43, 365003-1-6. [CrossRef]
Wu, Z.Y.; Fan, X.A.; Li, G.Q.; Wang, J.; Gan, Z.H. Intergranular insulated Fe-6.5 wt % Si/SiO2 composite
compacts with tunable insulating layer thickness for low core loss applications. RSC Adv. 2015, 5,
67031–67040. [CrossRef]
Liu, Y.D.; Fang, F.F.; Choi, H.J. Core-shell-structured silica-coated magnetic carbonyl iron microbead and its
magnetorheology with anti-acidic characteristics. Colloid Polym. Sci. 2011, 289, 1295–1298. [CrossRef]
Liu, Y.D.; Choi, H.J.; Choi, S.-B. Controllable fabrication of silica encapsulated soft magnetic microspheres
with enhanced oxidation-resistance and their rheology under magnetic field. Colloid. Surf. A Physicochem.
Eng. Asp. 2012, 403, 133–138. [CrossRef]
Desautels, R.D.; Freeland, J.W.; Rowe, M.P.; van Lierop, J. The role of interfacial metal silicates on the
magnetism in FeCo/SiO2 and Fe49% Co49% V2% /SiO2 core/shell nanoparticles. J. Appl. Phys. 2015, 117,
17C728-1-4. [CrossRef]
Desautels, R.D.; Rowe, M.P.; Jones, M.; Whallen, A.; van Lierop, J. Spontaneously formed interfacial metal
silicates and their effect on the magnetism of superparamagnetic FeCo/SiO2 core/shell nanoparticles.
Langmuir 2015, 31, 2879–2884. [CrossRef] [PubMed]
Nakaya, M.; Nishida, R.; Muramatsu, A. Preparation of air-stable iron-oxide-coated metallic iron nanoparticles.
J. Chem. Chem. Eng. 2013, 7, 1050–1053.
Nanomaterials 2017, 7, 1
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
13 of 13
Tajima, S.; Hattori, T.; Kondoh, M.; Kishimoto, H.; Sugiyama, T.; Kikko, T. Properties of high-density magnetic
composite fabricated from iron powder coated with a new type phosphate insulator. IEEE Trans. Mag. 2005,
41, 3280–3282. [CrossRef]
Decher, G. Fuzzy nanoassemblies: Toward layered polymeric multicomposites. Science 1997, 277, 1232–1237.
[CrossRef]
Schmitt, J.; Decher, G.; Dressick, W.J.; Brandow, S.L.; Geer, R.E.; Shashidhar, R.; Calvert, J.M. Metal nanoparticle/
polymer superlattice films: Fabrication and control of layer structure. Adv. Mater. 1997, 9, 61–65. [CrossRef]
Caruso, F.; Shi, X.; Caruso, R.A.; Susha, A. Hollow titania spheres from layered precursor deposition on
sacrificial colloidal core particles. Adv. Mater. 2001, 13, 740–744. [CrossRef]
Caruso, R.A.; Susha, A.; Caruso, F. Multilayered titania, silica, and laponite nanoparticle coatings on
polystyrene colloidal templates and resulting inorganic hollow spheres. Chem. Mater. 2001, 13, 400–409.
[CrossRef]
Sasaki, T.; Ebina, Y.; Tanaka, T.; Harada, M.; Watanabe, M.; Decher, G. Layer-by-layer assembly of titania
nanosheet/polycation composite films. Chem. Mater. 2001, 13, 4661–4667. [CrossRef]
Sasaki, T.; Ebina, Y.; Fukuda, K.; Tanaka, M.; Harada, M.; Watanabe, M. Titania nanostructured films derived
from a titania nanosheet/polycation multilayer assembly via heat treatment and UV irradiation. Chem. Mater.
2002, 14, 3524–3530. [CrossRef]
Nakano, H.; Ishii, M.; Nakamura, H. Preparation and structure of novel siloxene nanosheets. Chem. Commun.
2005, 2945–2947. [CrossRef] [PubMed]
Nakano, H.; Ohtani, O.; Mitsuoka, T.; Akimoto, Y.; Nakamura, H. Synthesis of amorphous silica nanosheets
and their photoluminescence. J. Am. Ceram. Soc. 2005, 88, 3522–3524. [CrossRef]
Nakano, H.; Mitsuoka, T.; Harada, M.; Horibuchi, K.; Nozaki, H.; Takahashi, N.; Nonaka, T.; Seno, Y.;
Nakamura, H. Soft synthesis of single-crystal silicon monolayer sheets. Angew. Chem. Int. Ed. 2006, 45,
6303–6306. [CrossRef] [PubMed]
Tajima, S.; Hattori, T.; Kondoh, M.; Sugiyama, M.; Higashiyama, K.; Kishimoto, H.; Kikko, T. Properties of
high density magnetic composite (HDMC) by warm compaction using die wall lubrication. Mater. Trans.
2004, 45, 1891–1894. [CrossRef]
Tajima, S.; Hattori, T.; Kondoh, M.; Okajima, H.; Sugiyama, M.; Kikko, T. Magnetic properties and
microstructure of high-density sintered iron formed by warm compaction using die wall lubrication.
Mater. Trans. 2005, 46, 1402–1406. [CrossRef]
Ristić, D.; Ivanda, M.; Speranza, G.; Siketić, Z.; Bogdanović-Radović, I.; Marciuš, M.; Ristić, M.; Gamulin, O.;
Musić, S.; Furić, K.; et al. Local size distribution of oxygen in silicon-rich oxide thin films: a tool to investigate
phase separation. J. Phys. Chem. C 2012, 116, 10039–10047. [CrossRef]
Noma, Y.; Saga, Y.; Une, N. Amorphous silica-coated graphite particles for thermally conductive and
electrically insulating resins. Carbon 2014, 78, 204–211. [CrossRef]
© 2016 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/).