Homogeneous WC-Co-cemented carbides from a cobalt-coated WC powder produced by a novel solution-chemical route

Article abstract of DOI:10.1111/j.1551-2916.2007.01948.x

A solution chemical route to cobalt-coated WC-powder is described that allows for the preparation of WC-Co powders and compacts having a carbon content very close to the desired carbon content even under an inert atmosphere. The microstructural homogeneity in the sintered WC-Co composites when using the Co-coated grains was found to be superior as compared with conventionally mill-mixed powders, and the structural changes in the individual WC-grains were found to be much smaller, which is ascribed mainly to the fact that the coated grains do not require a grinding step leading to the formation of a tail of smaller WC grain sizes.

Full text of DOI:10.1111/j.1551-2916.2007.01948.x

J. Am. Ceram. Soc., 90 [11] 3449–3454 (2007)
DOI: 10.1111/j.1551-2916.2007.01948.x
r 2007 The American Ceramic Society
ournal
J
Homogeneous WC–Co-Cemented Carbides from a Cobalt-Coated WC
Powder Produced by a Novel Solution-Chemical Route
Asa Ekstrand, Gunnar Westin, and Mats Nygreny
z
w,z
˚
z
˚
Angstro
¨
m Laboratory, Department of Materials Chemistry, Uppsala University, Uppsala, Sweden
yArrhenius Laboratory, Department of Inorganic Chemistry, Stockholm University, Stockholm, Sweden
A solution chemical route to cobalt-coated WC-powder is de-
scribed that allows for the preparation of WC–Co powders and
compacts having a carbon content very close to the desired car-
bon content even under an inert atmosphere. The micro-
structural homogeneity in the sintered WC–Co composites
when using the Co-coated grains was found to be superior as
compared with conventionally mill-mixed powders, and the
structural changes in the individual WC-grains were found to
be much smaller, which is ascribed mainly to the fact that the
coated grains do not require a grinding step leading to the for-
mation of a tail of smaller WC grain sizes.
A way to overcome the problems of the grinding/mixing
process and to obtain a higher homogeneity of the WC and
Co mixture is to coat the WC powder with the cobalt phase, but
in spite of its importance, there are quite a few such routes re-
ported as far as we can find, and most are in the patent liter-
ature. Chemical vapor deposition and other gas phase-based
routes are of less interest due to the high costs, low flexibility,
and low rate of deposition. Solution-based routes have been re-
1
1,12
ported using the polyol precipitatation process,
techniques, and just recently a route based on precipitation by
electro-less
13
1
4
pH adjustment was patented. All of them use thermal treat-
ment under a reducing atmosphere to convert the cobalt com-
pounds to metal and have extra steps introduced to remove
different kinds of additives, such as large amounts of organic
groups, phosphates, Pd, Sn, and alkali –salts, which are detri-
mental to the materials’ properties even in very small amounts,
which makes the processes tedious. It is normally not reported
how successfully the additives have been removed and it is diffi-
cult to make detailed comparisons, due to the very nature of
the patent literature mainly being judicial. The complexity, and
in some cases the reduced yields and large volumes of solvents
containing complexes that are difficult to purify by precipita-
tion, e.g. in the polyol route, make some of these routes
more difficult to apply with environmental restraints, even
though water is sometimes put forward as an environmentally
benign solvent.
I. Introduction
HE WC–Co composites are frequently used materials for
cutting tools in the machining and drilling of materials
T
ranging from electric circuits boards to rock. The high perfor-
mance and versatility of the cemented carbides arises from a
combination of the very hard WC grains and the tough and
well-adhering cobalt binder phase, which is also essential for the
1
,2
sintering process. The composites normally contain 5–10 wt%
cobalt as a binder phase between 1- and 10-mm-sized WC-
grains. Commercially, a powder-metallurgical route is used in
the production of the cemented carbides: the precursor powder
mixture is prepared by milling Co and WC powders together
with an organic binder in a rotating ball mill. The milled pow-
ders are then spray dried into agglomerates, about 100 mm in
size, which are then pressed into green bodies of various sizes
and shapes. The binder is removed in a hydrogen-containing
atmosphere at ca. 4001C, and then follows the sintering process,
comprising heating to ca. 14501C and an isothermal heat treat-
In a patent, we have described an efficient solution-based route
1
5
to nonagglomerated Co-coated WC. In this case, the precursor
solution should preferentially consist of Co(OAc) O,
2
ꢀ 4H
2
triethanolamine (TEA), and methanol to which WC powder is
added. The methanol solvent is evaporated under moderate
stirring, and the powder lumps thereby obtained are heated to
remove the organic residues. The burning-off procedure is carried
out under special conditions, in order to ensure that the carbon
content of the final compact will not deviate from that set by the
chemical composition of the WC powder.
1
–10
ment at this temperature for typically up to 1 h.
The milling
procedure has several drawbacks: (i) the wear of the milling
bodies may lead to contamination of the milled mixture. (ii)
Even after extensive milling, due to crushing of the WC powder,
a wide size distribution of the particles is found rather than a
narrow size distribution, which implies that larger amounts of
Co, higher sintering temperatures, and longer sintering times
must be used. (iii) Grain growth during sintering cannot be
avoided due to the more rapid dissolution and reprecipitation of
the small particles, which makes the microstructure obtained
strongly dependent on the milling parameters applied. There-
fore, very coarse-grained cemented carbide compacts, needed in
the mining applications, or compacts containing grains of a
narrow size distribution are difficult or impossible to prepare
with conventional techniques.
In this article, we will describe the chemistry and microstruc-
ture development in some detail but instead of Co(OAc)
2
ꢀ 4H
2
O,
a mixture of Co(OAc) ꢀ 4H O and Co(NO ) ꢀ 6H O is used as a
2
2
3 2
2
precursor for Co, which permits a more straightforward burning-
off procedure of the organic residues, allowing for the production
ofcoated powderswithvery close tothe desired stoichiometric W,
Co, andC contents, and very low O (ca. 0.02 wt%) and N (ca. 0.01
wt%) contents. This thus enables a direct formation of Co-coated
WC powder without introducing foreign elements and treatment
with reducing gases such as H
trations of metal salts in the solvent and the yield is very close to
00%. The evaporated methanol can be recycled, and the gases
2 2
/N . This route uses high concen-
1
evolved on heat treatment are similar to car exhaust gases such as
NO , CO , and organic molecules and should easily be taken care
x
x
R. Koc—contributing editor
of with a catalytic converter. There is no strong complex binder
that inhibits the easy precipitation of the Co salts from a simple
pH adjustment of cleaning solutions, etc., meaning that this route
not only produces very controllable metal coatings, but is also
very environmentally controllable.
Manuscript No. 22245. Received September 14, 2006; approved June 26, 2007.
Author to whom correspondence should be addressed. e-mail: gunnar.westin
mkem.uu.se
w
@
3
449

3
450
Journal of the American Ceramic Society—Ekstrand et al.
Vol. 90, No. 11
A study comprising carbon content analysis and phase-de-
Dissolve Co(NO ) 6H O and
velopment studies of Co precursor mixtures with different
nitrate-to-acetate ratios was conducted to gain control of the
carbon content. The details of the process used for the prepa-
3 2
2
Co(OAc) 4H O in methanol
2
2
1
6
ration of nanostructured cobalt was recently described.
The microstructure and phase development of the coated pow-
der were investigated. The sintering process to form compacts
and their microstructures obtained from cobalt-coated WC
powder and ball-milled powder mixtures, respectively, have
been compared. The experimental techniques used were scan-
ning electron microscopy-energy dispersive X-ray spectroscopy
Add 0.5 TEA per Co
Add WC powder
(SEM-EDS), transmission electron microscopy-EDS (TEM-
EDS), powder-X-ray diffraction (XRD), chemical analysis
(
W, Co, C, N, O), thermogravimetric analysis, and dilatometry.
Evaporation of solvent at ~35°C using
an air flow. Mechanical working of the
mixture until clay-like state in the final
part of the drying
II. Experimental Procedure
(
1) Equipment and Chemicals Used
The Co(NO ) ꢀ 6H O (Acros Organics, Geel, Belgium; 99%),
3
2
2
Co(OAc)
2
ꢀ 4H
2
O (Fluka, Chemie, Buchs, Switzerland; 991%),
methanol (MeOH) (Merck p.a., Darmstadt, Germany), and TEA
Fluka, ca. 99%) were used as purchased. The water contents of
Heat-treatment of the powder-lumps
−1
(
to >500°C, typiacally at 10°C min ,
the Co salts were determined by a standard thermogravimetric
procedure. The WC powder had a main size fraction of 5–8 mm
and close to 100% of the particles fell between 3 and 11 mm, ac-
cording to the analysis supplied (H. C. Starck GmbH & Co. KG,
Laufenburg, Germany). The mixed WC1Co powder made for
under flowing Ar or N2
Cobalt coated WC powder
¨
comparison was obtained from Sandvik Coromant (Vastberga,
Sweden). It was produced by standard milling with the same
WC powder source as was used for the coating procedure.
¨
PowderXRDphotographswereobtainedwithaGuiner–Hagg
Fig. 1. Preparation scheme for Co-coated WC powder with 6 wt% Co.
Table I. The amounts of chemicals used in a normal batch.
focusing camera (Stockholm University, Stockholm, Sweden),
using CuKa1 radiation and Si as the internal standards. The SEM
studies were carried out with JEOL 820 and JEOL 880 micro-
scopes (Tokyo, Japan), equipped with EDS, Link 10000AN, and
Link ISIS, respectively, and the TEM imaging was made with a
JEOL 2000FX microscope. The thermogravimetric (TG) analys-
es were performed with a Perkin-Elmer TGA7 apparatus (Wal-
tham, MA), which was thoroughly flushed with nitrogen before
the measurements were started. The dilatometer curves were ob-
tained with a Setaram DHT 2050 dilatometer (Caluire, France),
in an argon atmosphere. The W, Co, C, N, and O contents were
determined by an external analytic laboratory, with the C, N, O
contents determined using standard LECO combustion-spectro-
scopic techniques (Stockport, U.K.), and the W and Co contents
using the X-ray fluorescence technique. Small batches of the WC–
Co mixtures for the studies of the phase evolution were prepared
in the TG apparatus, while the larger batches of ca. 10 g for Co
and 50 g for WC–Co dilatometry and chemical analysis were pre-
pared in a closed alumina container in a top-loaded pit furnace
with flowing nitrogen gas entering from the bottom. A heating
rate of 101C/min to 7001C was used in both cases.
Table I. The Amounts of Chemicals Used in a Normal Batch
Chemical
Formula
Amount
Tungsten carbide
Cobalt nitrate
Cobalt acetate
Triethanolamine
Methanol
WC
Co(NO ) ꢀ 6H O
18.80 g
5.333g
0.507 g
1.35 mL
100 mL
3 2
2
Co(OAc) ꢀ 4H O
2
2
N(C H OH)
3
2
5
CH OH
3
precursor mixture is given in Fig. 3. Powder-XRD patterns were
recorded of samples heated to 5001, 7001, and 10001C in the TG
unit, andto7001Cinthefurnace. TEMstudieswereperformedon
3
a samplewiththeNO :OAccomposition9:1andheatedto5001C.
(3) Preparation of Co-Coated WC Powder and WC–Co
Compacts
Co-coated WC powder (6 wt% Co) was prepared according to
the route outlined in Fig. 1, Table I. In order to study how the
1
00
(
2) Preparation of Cobalt Metal
a
In order to study how cobalt metal with low contents of addi-
tional elements, i.e. C, N, and O, could be prepared from
the Co salt precursors, a series of samples with varying
8
6
0
0
Co(NO
3
)
2
ꢀ 6H
2
O:Co(OAc)
2
ꢀ 4H O composition ratios were prez-
2
pared as outlined in Fig. 1, but without addition of WC powder.
A more detailed account of the phase development with an aim to
prepare nanocrystalline cobalt is published elsewhere. The ra-
16
40
b
tios studied were [Co(NO ) ꢀ 6H O:Co(OAc) ꢀ 4H O]5 1:0, 9:1,
3
2
2
2
2
4
:1, 2:1, 7:6, 1:1, 1:2, and 0:1. The thermal decomposition of these
c
d
2
0
mixtures was studied in the TG unit in a nitrogen atmosphere at a
heating rate of 101C/min. The weight–loss curves are given in
Fig. 2. The C, N, and O contents were determined for the larger
batches prepared in the top-loaded pit furnace. A graph of the
1
00 200 300 400 500 600
Temperature (°C)
3
obtainedcarbon contents plotted versusthe NO :OAcratio in the
Fig. 2. Thermogravimetry graphs obtained at a heating rate of 101C/
min in a nitrogen atmosphere of a mixture of WC powder and Co con-
zWarning: Although we have not observed any tendency toward explosive behavior
during the many studies on the cobalt and WC-Co mixtures, any mixture of nitrate and
organic compounds should be regarded as potentially hazardous to explosion.
centrate (NO
3
:OAc ratio of 9:1) for formation of WC–6 wt% Co (a);
:OAc ratios 0:1 (b), 9:1 (c), and 1:0 (d).
Pure concentrates of the NO
3

November 2007
Novel Solution-Chemical Route
2) Thermal Treatment
3451
(
The decomposition of organic and nitrate groups to form the Co
metal was monitored in the TG unit (see Fig. 2). The decom-
position was found to be finished at about 4701C for all
1
0
5
0
NO
from high to low NO
tio of NO groups exhibited a very rapid decomposition step at
3
:OAc ratios, but the decomposition patterns changed
3
:OAc ratios. The samples with a high ra-
3
1
4
301–1501C, followed by very slow decomposition ending at ca.
701C. Samples with a high ratio of OAc-groups showed a more
complex pattern, without the rapid decomposition found for the
nitrate-rich samples. The very rapid reaction of the nitrate-rich
concentrates at 1301–1501C is probably due to combustion of
the organic parts with the nitrate groups as oxidizers, causing
internal heating of the sample to ca. 4501C. This would explain
the TG graph plateau found between 1501 and 4501C with these
samples. At higher OAc concentrations, less net heat should be
evolved by the combustion and decomposition, implying that
several relatively well-defined decomposition steps can be dis-
cerned in the interval 1501–4501C. Above 4701C, the weight loss
was almost zero for all samples, and the total weight loss
decreased with increasing acetate content. The powder-XRD
patterns of all mixtures heat treated to 5001, 7001, and 10001C
0
20
40
60
80
100
mol% OAc to NO3
Fig. 3. Carbon content of powders obtained by heating Co precursors
with different ratios of acetate to nitrate (mole% OAc in NO ), to 7001C
at 101C/min, in nitrogen.
3
1
6
contained peaks which could all be ascribed to FCC cobalt.
The powder-XRD patterns of the samples heated to 7001C in
a furnace showed FCC cobalt to be the only crystalline phase
present. Low O and N contents were observed in samples heated
to 7001C, ca. 0.23 wt% O, and ca. 0.02 wt% N for the 1:0 sam-
ple. If the samples were taken out from the furnace at temper-
atures of 1001C or more, oxygen contents up to Co:O5 1:1
could be found, i.e. full oxidation to CoO by air-oxygen took
place. Likewise, long-time storage or samples heated only to
morphology and phase contents of the cobalt layer vary with the
heat treatment, nonheat-treated powders were heated in the TG
apparatus, quenched at 5001, 7001, 8001, 9001, and 10001C, and
subsequently investigated by powder XRD. The XRD peaks,
but those of WC are given in Table II. In addition, the samples
quenched at 5001, 7001, and 10001C were studied with SEM-
EDS and that from 7001C in the pit furnace also by C, N, and O
analysis. A thoroughly ball-milled mixture of Co-metal and
WC was prepared for comparison. The coated and ball-milled
WC–Co powders were mixed with ethanol and ultrasonically
deagglomerated, dried, and then uniaxially pressed under 10
MPa to cylinders with 6-mm diameter and ca. 5-mm length,
without a pressing aid. These compacts were studied in a dilato-
meter using a heating rate of 51C/min up to 14501C and iso-
thermal heat treatment at this temperature for 10 min, followed
by rapid cooling. The microstructures obtained were investigat-
ed with SEM-EDS on polished cross sections.
5001C normally had increased oxygen contents due to air
oxidation, and storage should therefore preferably be made in
an inert atmosphere. The carbon contents of the samples heated
to 7001C decreased from 13 wt% with increased NO
and approached zero close to the 1:0 and 9:1 NO
3
:OAc ratio,
:OAc com-
3
positions, as shown in Fig. 3. The fact that the oxide was not
found even in the 1:0 sample might be explained by the presence
of 0.5 equivalents of TEA per Co, implying that TEA is a com-
plexing agent strong enough to remain bound to the cobalt
when heated and contains enough reducing components to com-
pensate for the oxidizing power of the NO -groups. Consider-
ꢁ
3
ing only the carbon content, the 1:0 mixture seems to be slightly
better than the 9:1 and 4:1 mixtures but, as discussed above,
crystallization of a Co-nitrate–TEA complex easily occurred at
the 1:0 composition, inducing an inhomogeneous distribution of
the oxidation and reduction agents in the precursor concen-
trates, which would lead to an elementally inhomogeneous
product. Accordingly, the 9:1 composition was used for the
preparation of cobalt-coated WC–Co powders.
III. Results and Discussion
(1) Studies on the Co-Metal Formation; Preparation of
Cobalt Precursor Mixtures
Cobalt precursor mixtures from TEA, Co(NO ) ꢀ 6H O, and
3
2
2
Co(OAc)
2
ꢀ 4H
2
O with nitrate-to-acetate ratios from 1:0 to 0:1
were prepared according to the scheme given in Fig. 1, but
without addition of WC and heat treatment. The TEA was
added in order to inhibit crystallization of the Co precursor,
which otherwise easily occurred, especially for the 1:0 nitrate-to-
acetate ratio. From IR and TG studies, it was shown that the
(3) Preparation of Co-Coated WC Powder
The TG graph of WC powder coated with an NO :OAc (9:1)
3
21
multidentate TEA is sufficiently strongly bonded to the Co
mixture corresponding to an addition of 6 wt% Co showed
a decomposition pattern very similar to that of WC-free
NO :OAc (9:1 and 1:0) mixtures, as shown in Fig. 2. The stud-
ions to remain in the dried concentrates, and the thermal de-
composition of the acetate and nitrate groups during drying is
believed to be at most very small.
3
ies of pure Co materials thus also seem to be relevant for the
WC–Co materials. As discussed above, the TG studies showed
the decomposition of the organic and nitrate groups to be fin-
ished at temperatures below 4701C.
Table II. XRD Peak Maxima for Phases Other than WC in
Coated WC–6%Co Powder Heated at 101C/min
(
A) 5001C: After heating to 5001C, the material consist-
ed of nonagglomerated or loosely agglomerated Co-coated WC
grains and a very small amount of isolated, ca. 1-mm sized, clus-
ters of coalesced nano sized Co particles (see Fig. 4). According
to the powder-XRD study, the only crystalline phases present
˚
d-value (A) (hkl)
Temperature (1C)
Phase
5
7
8
9
1
1
00
00
00
00
000
000
Co (fcc)
Co (fcc)
Co (fcc)
Co (fcc)
Co (fcc)
2.04 (111)
2.04 (111)
2.05 (111)
2.05 (111)
2.07 (111)
2.23 (422)
17
18
were WC and FCC Co. The X-ray intensity from the latter
phase was so weak that only the (111) peak could be observed
1.78 (200)
1.77 (200)
(
see Table II). Almost all WC grains were covered with a cobalt
coating consisting of coalesced 10–100 nm-sized Co-particles.
SEM-EDS studies showed average metal contents of 60 mol%
Co and 40 mol% W. The high value of the Co content is due to
Co W C (Z)
2.10 (511)
6
6
XRD, X-ray diffraction.

3
452
Journal of the American Ceramic Society—Ekstrand et al.
Vol. 90, No. 11
Fig. 4. (a) Scanning electron microscopy image of the WC–6 wt% Co sample heated to 5001C, obtained in secondary-electron imaging mode.
b) Close-up with transmission electron spectroscopy imaging of Co metal obtained at 5001C.
(
the screening effect from surface cobalt. Approximately the
same amount of cobalt was observed in the material that was
not burnt off. Spot analyses of the WC–Co grains showed cobalt
contents from 11 to 90 mol%, and in no case was a sole tungsten
EDS spectrum obtained.
very loosely agglomerated. Deagglomeration could easily be
achieved by gentle hand pressing of the samples or treatment
in an ultrasonic bath containing ethanol and coated WC parti-
cles. The powder-XRD pattern showed the presence of the Z
1
9
6 6
phase, W Co C, implying that Co had reacted with the WC
(
B) 7001–9001C: Heating to 7001C caused densification
grains, although not completely, because the powder-XRD pat-
tern also contained the Co (111) peak. The SEM–EDS analyses
yielded a mean surface composition of 21 mol% Co, which is
much lower than those found in the 5001 and 7001C samples.
This lower Co content accords with the powder-XRD studies,
of the Co coatings into smoother surfaces, as can be seen
in Fig. 5. The grains were still nonagglomerated or loosely
agglomerated, and the amount of isolated cobalt clusters was
reduced. The powder-XRD studies showed the presence of WC
and Co only, the latter through weak peaks assignable to (hkl)
6 6
which showed the presence of W Co C through two weak peaks
given in Table II.
1
11 and 200 (the latter discernable only in the range 8001–9001C
as given in Table II). This indicates that no reaction between Co
and WC had occurred although, according to the SEM study,
the wetting of the Co phase seemed to be good. The average Co
and W contents were very similar to those obtained in the EDS
studies of the 5001C sample. The spot analyses gave a somewhat
lower scatter of Co contents, compared with the 5001C sample,
and the lowest values observed were slightly higher, i.e. ca. 17
mol% Co. This indicates that Co in this case is more evenly
distributed on the WC surface. From the SEM-EDS studies, it
seems that the free cobalt clusters become more and more at-
tached to the WC-coated grains at elevated temperatures. This
was also indicated by the surface structure of the coated particles
that seemed to have a double coating in some places, although
they might also be due to loss of the cobalt coating from weakly
attached neighboring coated WC particles. The powder-XRD
results remained the same even after heating to 8001 and 9001C.
The larger batches of Co-coated WC powder, prepared in the
furnace, showed morphology and phase contents very similar to
those of the powder prepared in the TG unit at 7001C. This in-
dicates that the TG study is also relevant for larger batches. The
powder heated to 7001C was chosen as a precursor for the prep-
aration of WC–Co compacts, because the Co was relatively
smooth but still not alloyed with WC, and had a low surface
area reducing its oxidation sensitivity. The carbon content of the
sample heated to this temperature was 5.84 wt%, compared with
the theoretical value of 5.78 wt% for a mixture of WC and 6
wt% Co. Part of the ca. 1% excess carbon probably stems from
the WC powder, which had a slightly higher than stoichiometric
carbon content (o0.03% units). If the data of the carbon an-
alyses from the furnace prepared Co metal are used, an excess
of 0.03 wt% (using the measured value for the 9:1 composition)
to 0.1 wt% (using an imaginary tie line in the plot at the 9:1
composition) carbon stemming from the Co addition is expect-
(C) 10001C: The Co phase sintered to dense, smooth
coatings, completely surrounding each WC grain as can be seen
in Fig. 6, and very few isolated Co clusters could be found after
heating up to 10001C. The grains were still nonagglomerated or
Fig. 5. Scanning electron microscopy image of the WC–6 wt% Co
sample heated to 7001C, obtained in secondary-electron imaging mode.
Fig. 6. Scanning electron microscopy image of the WC–6 wt% Co
sample heated to 10001C, obtained in secondary-electron imaging mode.

November 2007
Novel Solution-Chemical Route
3453
at 12751C found in the Co–WC phase diagram, and 13601C is
close to equilibrium of the Co, WC, graphite, (W,Co) C-form-
However, much of the reactions taking
place at lower temperatures during the sintering of Co–WC
compacts from powders remain unrevealed and kinetic and geo-
metric factors are probably important, making it difficult to give
any detailed conclusions on the reactions taking place in the
present system.
The difference in sintering behavior between the ball-milled
powder and coated powder might stem from the fact that some
WC dissolves into the cobalt phase at lower temperatures in the
coated powders. The cobalt particles may be expected to have
comparatively very small contact areas with the WC grains
in the ball-milled powders. This would lead to less diffusion
between the WC and Co particles at temperatures up to where the
main sintering step starts. The smaller sintering step at 13601C,
found for some of the coated powders, might then be due to the
presence of some isolated cobalt particles or due to the circum-
stance that the diffusion between WC and the Co may have been
less complete in some, perhaps thicker parts of the coating.
Further studies are needed for more detailed and safe conclu-
sions concerning the presence of different maximum sintering
temperatures. The shrinkage was rather similar in the two pow-
ders: 20% and 23% for the ball-milled and solution-processed
powders, respectively, as averages of three samples each.
0
5
1
1
1
1
1
1
1
1
500
400
300
200
100
000
Der.
6
2
0,21
ing liquid at 13551C.
0
5
9
8
7
6
00
00
00
00
20
2
5
Displ.
30
1
000 2000 3000 4000 5000 6000 7000 8000 9000
Time (s)
Fig. 7. Shrinkage and time derivative of shrinkage curves of two typical
runs of cylinders of a coated WC–6 wt% Co powder: one with a single
sintering peak at 13001C (solid line) and one with an additional sintering
peak at 13601C (dotted line). Heating rate 51C/min, in an Ar
atmosphere.
ed. This shows that it should be possible to adjust the carbon
contents of the powder within less than ca. 0.1% unit of the
desired value. However, in order to control the carbon contents
even further to the exact desired composition, several parame-
ters have to be extremely well controlled, e.g. small oxygen
sources reducing the carbon contents and the presence of resid-
ual organic decomposition products in the furnace from earlier
runs increasing the carbon contents. The final adjustments of the
carbon contents are expected to be made during the sintering,
making it easier to take care of variations in the WC source and
surface oxidation from storage of the powder, but this has not
been the aim of the study presented here.
(
5) Microstructures After Sintering
The microstructures of the compacts after heat treatment at
4501C for 10 min of the two differently prepared powders are
1
shown in the Figs. 9 and 10. It can be observed that the compact
obtained from the solution-processed powder has a much more
homogeneous microstructure, i.e. a more narrow distribution of
WC grain sizes and a much better distribution of the Co phase,
and studies of several compacts clearly show that it contains less
pores than that obtained from the ball-milled powder. Another
striking difference is that the grains in the compact obtained
(4) Preparation of Compacts from Co-Coated WC Powder
and Ball-Milled WC1Co Powders; Sintering
The sintering process of the solution-processed and ball-milled
WC–6wt% Co powders was monitored in a dilatometer, using a
heating rate of 51C/min up to 14501C, followed by isothermal
heat treatment at this temperature for 10 min, and rapid cooling.
The selected heating profile is typical for sintering of WC–Co
compacts, but has by no means been optimized for the present
powders. The shrinkage and its time derivative for the two pow-
ders are plotted against temperature in Figs. 7 and 8.
The derivative of the shrinkage curves of the ball-milled pow-
ders shows a maximum close to 13601C, while the corresponding
curves of the solution-processed powders show a maximum at
1
3001C, and sometimes an additional smaller maximum at ca.
3601C, as can be seen in Figs. 7 and 8, respectively. Hence, the
1
two powders undergo different sintering processes. It can be
noted that 13001C is close to the lowest liquid-forming eutectic
1
1
1
1
1
1
500
400
300
200
100
000
0
Der.
5
10
1
5
Displ.
9
8
7
6
00
00
00
00
20
2
5
30
1
000 2000 3000 4000 5000 6000 7000 8000 9000
Time (s)
Fig. 8. Shrinkage and time derivative of shrinkage curve of a cylinder
of mill-mixed WC16 wt% Co powders. Heating rate 51C/min, Ar
atmosphere.
Fig. 9. Microstructure of a compact obtained by sintering a ball-milled
powder to 14501C, and holding at this temperature for 10 min (scanning
electron microscopy, backscattered electron imaging).

3
454
Journal of the American Ceramic Society—Ekstrand et al.
Vol. 90, No. 11
cobalt coating on the powders should also allow for a more even
sintering of the green body and thereby make submicrometer
WC–Co composites attainable. This requires, however, that a
mono-disperse submicrometer WC powder of a narrow grain
size distribution be used. The coating approach should have an
advantage compared with routes using highly agglomerated
WC–Co–C nanophase powders obtained from solution. Such
powders typically sinter firstly within the clusters and make full-
dense compacts difficult to achieve without extensive heating,
which causes extensive and uneven grain growth. Inhomoge-
neously distributed surplus carbon that is typical in these pro-
cesses adds further complications due to the increased sintering
activity in these areas, and grain growth inhibitors are typically
required. A preliminary study has also shown that it is possible
2
2
to prepare homogeneous Co–Ni coatings by this route.
Acknowledgments
We are grateful to Dr. Kjell Jansson at the Department of Inorganic Chem-
istry, Stockholm University, for assistance with the dilatometer runs.
All researchers who were involved in this study are affiliated to the Brinell
Centre, Inorganic Interfacial Engineering—A competence center financed by VIN-
NOVA, and the BRIIE member companies presently being AB Sandvik Corom-
¨ ¨
ant, Hoganas AB, Secotools, Kloster Erasteel AB, and Atlas Copco Secoroc AB.
References
1
P. Ettmayer and W. Lengauer, pp. 519–31 in Encyclopedia of Inorganic Chem-
istry, Vol. 2, Edited by R. B. King. John Wiley and Sons, Chichester.
2
W. M. Stoll and A. T. Santhanam, pp. 841–78 in Encyclopedia of Chemical
Technology, Vol. 4, 4th edition, Edited by J. I. Kroschwitz and M. Howe-Grant.
Wiley-Interscience Publication, John Wiley and Sons, Chicester.
H. E. Exner, ‘‘Physical and Chemical Nature of Cemented Carbides,’’ Int.
3
Fig. 10. Microstructure of a compact obtained by sintering a Co-coat-
ed WC powder to 14501C, and holding at this temperature for 10 min
Metals Rev., 24, 149–73 (1979).
B. H. Kear, G. Skandan,and R. Sadangi, ‘‘FactorsControlling Decarburation in
4
(scanning electron microscopy, backscattered electron imaging).
Hvof Sprayed Nano-WC/Co Hard Coatings,’’ Scripta Mater., 44, 1703–7 (2001).
S. Raghunathan, R. Caron, J. Friederichs, and P. Sandell, ‘‘Tungsten Carbide
5
Technologies,’’ Adv. Mater. Processes, 4, 21–3 (1996).
¨
M. Christensen, G. Wahnstrom, C. Alibert, and S. Lay, ‘‘Quantitative Analysis
from the solution-processed powder are much more angular and
facetted than those in the other compact.
6
of WC Grain Shape in Sintered WC–Co Cemented Carbides,’’ Phys. Rev. Lett., 94,
0
66105 (2005).
Z. Yao, J. J. Stieglich, and T. S. Sudarshan, ‘‘WC–Co Enjoys Proud History
and Future.
7
8
B. Aronssson, The Origins and Growth of Cemented Carbide. Ord & Form AB,
Uppsala, 2005
IV. Conclusions
9
An efficient solution-based process has been developed that
produces cobalt-coated WC-powder with an adjustable carbon
R. K. Sadangi, L. E. McCandlish, B. H. Kear, and P. Seegopaul, ‘‘Grain
Growth Inhibition in Liquid Phase Sintered Nanophase WC/Co Alloys,’’ Int. J.
Powdermet., 35, 27 (1999).
content from
Co(NO
ꢀ 6H
a
O, and TEA mixed with WC powder, after dry-
methanol solution of Co(OAc) ꢀ 4H O,
10
2
2
˚
Z.-H. Xu and J. Agren, ‘‘A Modified Hardness Model for WC–Co Cemented
Carbides,’’ Mater. Sci, Eng., A 386, 262–8 (2004).
3
)
2
2
11
ing and heating to 4701C or more under inert gas. The micro-
structure of the final WC–Co compact produced by sintering of
a cobalt-coated WC powder (6 wt% Co) at 14501C was superior
to that obtained from a thoroughly ball-milled mixture, in hav-
ing less pores, more evenly sized, and more angular WC grains.
The improved quality is presumably due to the more homoge-
neous mixture of WC and cobalt in the Co-coated WC powder
allowing for a more homogeneous sintering process and the fact
that a ball-milling step, resulting in the formation a tail of small-
er WC particles leading to an inhomogeneous sintering process,
is avoided. Thus, this novel low-cost route produces superior
microstructural homogeneity, removes the need for a ball-mill-
ing step, and improves the control of WC grain-size distribution
in the preparation of WC–Co compacts. This study describes the
coating of large WC grains suitable for preparation of micro-
structures with large WC grains and low cobalt content, which is
highly desired for mining, oil drilling, and tunnel excavation.
But the route should also be useful for coating of smaller WC
grains allowing for sintering into fine-grained cemented carbides
desired in metal machining and fine precision drilling. The even
S. Andersson and M. Muhammed. World Intellectual Property Organization
Patent: ‘‘Method of Making Metal Composite Powder’’ 9503420-3 (1995).
G. Eriksson, H. Siegbahn, S. Andersson, T. Turkki, and M. Muhammed,
12
21
‘
Mater. Res. Bull., 32, 491–9 (1997).
‘The Reduction of Co by Polyalcohols in Presence of WC Surfaces by XPS,’’
13
A. Nakabayashi, ‘‘Co–W Coated WC-Powder’’; JP Patent 62192501 (1987).
D. F. Caroll and J. Jarvi, ‘‘Methods of Producing Composite Powders’’; U.S.
Patent 20030097909 (2003).
14
15
˚
¨
U. Fisher, M. Waldenstrom, S. Ederyd, M. Nygren, G. Westin, and A. Ek-
strand, ‘‘Method of Making Metal Composite Materials’’; U.S. Patent 5,505,902
(1996).
16
˚
A. Ekstrand, K. Jansson, and G. Westin, ‘‘A Solution Synthetic Route to
Cobalt Film and Sponge,’’ Chem. Mater., 17, 199–206 (2005).
JCPDS card No: 25-1047; Bind, McCarty, Pennsylvania State University,
17
University Park, PA, ICDD Grant in Aid, 1973.
18
19
20
JCPDS card No: 15-806; Natl. Bur. Stand. (U.S.) Monogr. 25, 10, 1965.
JCPDS card No: 23-939; Pollock, Stadelmayer Metal Trans. 1, 767, 1970.
˚
L. Akesson, Thesis, Royal Institute of Technology, Stockholm, Sweden,
1
982.
21
B. Uhrenius, B. Carlsson, T. Franzen, ‘‘A Study of the Cobalt-tungsten-
carbon System at Liquidus Temperatures,’’ Scand. J. Metall., 549–56 (1976).
22
˚
A. Ekstrand and G. Westin, ‘‘Highly Homogeneous WC–Ni/Co Composites
Produced from WC-Powder Coated with a Solution Chemical Route,’’ Advances
in Science and Technology CIMTEC 2002 – 10th International Ceramic Congress
[Part C], 489 (2002).
&