Homogeneous TiB2 ceramics achieved by electric current-assisted self-propagating reaction sintering

Article abstract of DOI:10.1111/j.1551-2916.2007.01856.x

Using spark plasma sintering techniques, homogeneous microstructures of titanium diboride (TiB₂) ceramics were obtained by sintering of boron and titanium powder mixtures. The results show that an additional electric current is essential for achieving a large number of evenly distributed ignition points that ensure that the self-propagating reaction simultaneously takes place within the entire volume. The effects of the electric current, the use of Mg additions, and the heating rates on the resulting TiB₂ ceramic densities and microstructures are discussed.

Full text of DOI:10.1111/j.1551-2916.2007.01856.x

J. Am. Ceram. Soc., 90 [10] 3303–3306 (2007)
DOI: 10.1111/j.1551-2916.2007.01856.x
r 2007 The American Ceramic Society
ournal
J
Homogeneous TiB Ceramics Achieved by Electric Current-Assisted
2
Self-Propagating Reaction Sintering
w
David Salamon, Mirva Eriksson, Mats Nygren, and Zhijian Shen*
Arrhenius Laboratory, Department of Inorganic Chemistry, Stockholm University, SE-106 91 Stockholm, Sweden
Using spark plasma sintering techniques, homogeneous micro-
structures of titanium diboride (TiB ) ceramics were obtained by
are the possibilities to apply very high heating rates, up to sev-
eral hundred degrees per minute, and subsequently to achieve
full densification within minutes.
The aim of this work is to examine the influence of electric
2
current on the self-propagation reaction Ti12B-TiB using
2
sintering of boron and titanium powder mixtures. The results
show that an additional electric current is essential for achieving
a large number of evenly distributed ignition points that ensure
that the self-propagating reaction simultaneously takes place
within the entire volume. The effects of the electric current, the
use of Mg additions, and the heating rates on the resulting TiB2
ceramic densities and microstructures are discussed.
amorphous boron powders containing 0.8 and 11 wt% Mg. The
beneficial effects of current on chemical reactions have been re-
5
–9
ported previously, but now we also report the effect of DC
current under SPS conditions on nonconductive reactants. The
resulting solidification and microstructures of the TiB
formed were studied, and are reported below.
2
ceramics
I. Introduction
ELF-PROPAGATING high-temperature synthesis (SHS), reactive
synthesis, and combustion synthesis are methods by which
II. Experimental Procedure
S
To examine the effect of the electric current on the reaction be-
tween boron and Ti powders, slightly different starting mixtures
were used in this study compared with previous investigations.
ceramic materials can be produced. All these processes involve
exothermic reactions between starting powders as a driving force
yielding the desired final product. Commonly, these processes
comprise an ignition of the reaction, followed by self-propagat-
ing combustion fronts that transform the green body to a solid
piece. At the reaction front, very high heating rates are gener-
ated, resulting in high thermal gradients within short distances.
This implies that extremely nonisothermal conditions are pres-
2,4
Instead of a pure a-Ti powder, an a-Ti (6Al/4V) powder was
chosen due to its higher plasticity at lower temperatures. This
might lower the onset temperature of densification, increase the
densification rate, and result in a higher electrical conductivity of
the sample. In addition, Mg-containing amorphous boron pow-
ders were used. They may increase the overall electrical conduc-
tivity of the compacts during processing as the electrical
ent during the formation of the new phase(s) found in the final
body. The formation of titanium diboride (TiB ) ceramics has
1
2
1
2
resistivity is 4.45, 42, and 1.8 Â 10 mO Á cm, respectively, for
previously been intensively studied because TiB ceramics com-
bine superior hardness and corrosion resistance with a high
melting point and good oxidation resistance. In addition, the
2
1
0
Mg, Ti, and B at room temperature.
To ensure that the final product had an overall composition
close to TiB , the Ti and the amorphous boron powders were
2
2
TiB ceramic is electrically conductive, which makes it suitable
for electrical applications or enables the solid bodies formed to
be shaped into complex geometry using electrical discharge ma-
1
1
mixed at a ratio of 33/67 at.%. The mixtures of b-Ti 6/4 (Al/V)
powder (TLS Technik GmbH & Co., Bitterfeld, Germany) and
amorphous boron powders (ABCR GmbH & Co., Karlsruhe,
Germany: Grade I 95%–97% and Grade III 87.5% containing
2
chining (EDM). However, in the past, TiB ceramics have been
difficult to sinter by traditional sintering techniques to an ade-
quate density, e.g., hot pressing of TiB -based materials does not
0
.8% and 11% Mg, respectively) were dry planetary ball milled
2
for 1 h using WC/Co milling media. The aluminum and vana-
dium present in the Ti powder were mixed at the same ratio and,
therefore, some AlB and VB were expected to form together
give a density over 70% of the theoretical value even despite
using pressures up to 110 MPa. A recent work reported success
2
2
2
2
in preparing a dense TiB –Ni cermet by a pulse current process
3
2
with the TiB . Minor impurities present in boron powder start-
with the participation of an SHS reaction.
Spark plasma sintering (SPS), also known as the field-assisted
ing material (mainly oxygen) were not expected to have any sig-
nificant influence on the reaction. Two different compositions
were investigated and labeled as Ti B (Ti1B Grade I, mol ratio
sintering technique (FAST), has been used previously to syn-
thesize TiB , but using a relatively slow heating rate. In general,
4
2
2
1
2
:2), respectively, Ti BMg (Ti1B Grade III, mol ratio 1:2). The
SPS is a sintering technique with some similarities to conven-
tional hot pressing. However, in SPS, a strong-pulsed current is
directly passed through the electrically conducting pressure die
instead of using an external heating source. In appropriate cases,
electric current also passes through the sample, implying self-
heating of the sample. The unique features of the SPS process
mixed powders were stored in an oxygen-free dry box and the
specimens were placed in graphite dies in the same protective
environment. The graphite die had an inner diameter of 12 mm,
an outer diameter of 30 mm, and a height of 30 mm. The graph-
ite punches had a height of 20 mm. The graphite die with the
specimen was protected by a paraffin film when transported and
placed in the SPS apparatus (Dr. Sinter 2050; SPS Syntex Inc.,
Kanagawa, Japan). The SPS processing was carried out in vac-
uum at heating rates of either 20 or 1001C/min. The temperature
was automatically increased to 6001C during a 3-min period;
thereafter, the heating process was monitored and regulated by
an optical pyrometer focused on the outer surface of the die. A
pulse sequence of 12:2 (12 pulses on and 2 off) was used in all
cases. This setup allows a cooling rate of B4001C/min in the
A. Passerone—contributing editor
Manuscript No. 22966. Received March 21, 2007; approved May 1, 2007.
Supported by the Swedish Research Council (through Grants 621-2002-4299 and
21-2005-6290).
Member, American Ceramic Society.
Author to whom correspondence should be addressed. e-mail: mirva@inorg.su.se
6
*
w
3
303

3
304
Communications of the American Ceramic Society
Vol. 90, No. 10
Table I. Summary of the Experimental Conditions Used and Obtained Densities
3
Density (g/cm )
Name
Mixture
T (1C)
Heating rate (1C/min)
Dwell (min)
Pressure (MPa)
Reaction T (1C)
Ti BMg
2
Ti B
2
Ti B–20C
2
Ti BMg
2
Ti BMg–20C
2
Ti BMg–BN
2
Ti BMg
2
Ti BMg
2
Ti BMg
2
Ti B
2
Ti B
2
Ti BMg
2
Ti BMg
2
Ti BMg
2
Ti BMg
2
Ti BMg
2
1350
1500
1500
1500
1500
1500
1700
1900
100
100
20
100
0
3
3
3
3
3
0
2
50
50
50
50
50
50
50
50
1230
1330
1280
1260
1280
1360
1230
1260
3.90
4.33
4.29
4.13
4.25
3.99
4.22
4.36
20
100
100/20w
100/20w
wThe heating rate was slowed down at temperature above 14001C.
temperature range of 19001–10001C. A uniaxial mechanical
pressure of 50 MPa was applied from the beginning of the sinte-
ring cycle. In one experiment, in order to prevent the electrical
current from passing through the sample and to prevent direct
contact between the sample and the graphite tooling, the
Ti BMg green body was embedded in BN (labeled Ti BMg–
2 2
BN below). A summary of the experimental conditions used and
the densities obtained is given in Table I.
ple, where the current was not passing through the sample, and
the graphite die was the only source of heat before the self-
propagating reaction started. The ignition sites were located
on the contact points between the BN powder and the sample.
Thus, it is reasonable to assume that the current that passes
through the samples gives rise to local hot spots at the neck-
ing points, which in turn serve as ignition points of the self-
propagating reaction. The presence of encapsulated Mg in the
vicinity of these necking points seems to increase the rate of the
self-propagating reaction substantially.
The hardness and fracture toughness of the selected samples
were evaluated by Vickers indentation measurements using a
1
2
load of 98 N. Phase compositions were determined by powder
X-ray diffraction (XRD) using a Guinier–Hagg camera with
CuKa radiation and Si as an internal standard. The micro-
2
The XRD analysis showed that no phase other than TiB was
¨
detected in any of the cases. However, SEM pictures, recorded
in the backscattered electron (BSE) mode, did reveal the pres-
ence of unreacted amorphous boron and Ti (6Al/4V) particles.
As illustrations of the microstructures, the SEM micrographs of
several fracture surfaces of samples sintered at 15001C (with a
1
structures of the samples obtained were examined by a scanning
electron microscope (SEM 820; JEOL, Tokyo, Japan).
3
-min dwell) and at 19001C (with a 2-min dwell) are shown in
III. Results and Discussion
Fig. 2. It is obvious that the high heating rate (1001C/min)
yielded more homogeneous microstructures than those achieved
with the slow heating rate (201C/min), and therefore a rapid re-
action yields a more homogeneous microstructure.
Temperature, current, voltage, and pressure profiles were re-
corded during the SPS process. These profiles for the sample
Ti BMg sintered at 15001C/min with a 3-min dwell are shown in
2
Fig. 1(a). In this case, the shrinkage occured within a very nar-
row temperature range, implying that the chemical reaction oc-
curred within seconds and that more heat is produced than
needed for maintaining the preset heating rate of 1001C/min.
Accordingly, a sudden decrease in both the current and voltage
was recorded, as can be seen in Fig. 1(a).
The presence of Mg in the starting powder affected the elec-
trical conductivity of the green bodies and the shrinkage behav-
ior. Mg has a high evaporation rate at elevated temperatures.
However, due to the use of high heating rates, which results in
closed porosity, a large part of Mg is trapped in the pores be-
tween particles. The conditions for magnesium boride formation
are unfavorable as the temperature rapidly reaches the temper-
ature region where magnesium boride is not stable and the va-
1
3
por pressure is low (2–10 Pa). On the other hand, when a
slower heating rate (201C/min) is applied, the Mg has a longer
time to escape from the system and accordingly the electrical
conductivity of the compact is reduced in this case. On using a
lower heating rate, the shrinkage curve resembled that obtained
by sintering of Ti B composition; see Fig. 1(b). The finally sin-
2
tered material does not contain any traces of Mg due to a rapid
particle rearrangement during the fast shrinkage. The residual
Mg may also leave through diffusion. The Mg evaporation is
clearly revealed by the formation of a Mg-containing layer cov-
ering the graphite sintering die. The effect of Mg addition is only
noted during the sintering process, and the final properties are
not affected.
The reaction temperatures defined as the onset temperature of
the shrinkage are summarized in Table I. Apparently, when a
slower heating rate of 201C/min is applied, the onset tempera-
ture is almost independent of the Mg content, whereas when a
higher heating rate of 1001C/min is applied, the onset temper-
ature is shifted by approximately 1001C toward a lower tem-
perature for the Mg-rich sample. The highest reaction
Fig. 1. (a) Recorded current, voltage, temperature, and shrinkage ver-
sus time curves for Ti BMg heat treated to 15001C using a heating rate
2
of 1001C/min and dwell time of 3 min. (b) Recorded current, voltage,
temperature, and shrinkage versus time curves for TiB Mg heat treated
2
temperature, 13601C, was observed for the Ti BMg–BN sam-
to 15001C using a heating rate of 201C/min and dwell time of 3 min.
2

October 2007
Communications of the American Ceramic Society
3305
Fig. 2. Scanning electron micrographs recorded in BSE mode of: (a) Ti
using a heating rate of 1001C/min; (c) Ti BMg- and (d) Ti B- composition samples sintered at 15001C/3 min using a heating rate of 201C/min;
e) the Ti BMg–BN sample sintered at 15001C/3 min using a heating rate of 1001C/min; (f) the Ti BMg sample sintered at 19001C/2 min.
2 2
BMg- and (b) Ti B-composition samples sintered at 15001C/3 min
2
2
(
2
2
It could also be seen that the microstructure of the Ti
BN sample, heated at 1001C/min, resembled the microstruc-
ture of the Ti BMg sample heated at 201C/min more than the
2
BMg–
still some unreacted boron left in the samples. Moreover, the
electrical conductivity of the green body is enhanced by the
presence of small amounts of Mg, which is not found in mea-
surable amounts in the material after the sintering. The current
that passes through the samples gives rise to local hot spots at
the necking points, which serve as ignition points of the subse-
quent self-propagating reaction. The Mg addition and the im-
proved conductivity of the green body result in more evenly
distributed ignition points that ensure that the self-propagating
reaction takes place in the entire volume within a short time
period.
2
one heated at 1001C/min. However, this observation is not in
agreement with the finding reported above that the shape of
the sintering curve of the Ti BMg–BN sample is similar to that
2
2
of the Ti BMg sample heated at 1001C/min. The similarities
in shrinkage curves, but significant differences in the micro-
structures between Ti BMg and Ti BMg–BN specimens, demon-
2
2
strate the importance of the electrical current for the homogeneity
of the chemical reaction. In order to obtain coarser-grained mi-
crostructures, the samples have to be heat treated at tempera-
tures exceeding 17001C; see Fig. 2(f).
All the samples were nearly fully dense. The densities varied
between 86.5% and 96.7%, which is higher than the previously
reported ones of 70% prepared by hot press. The hardness val-
References
1
J. B. Holt and S. D. Dunmead, ‘‘Self-Heating Synthesis of Materials,’’ Annu.
Rev. Mater. Sci., 21, 305–34 (1991).
D. A. Hoke and M. A. Meyers, ‘‘Consolidation of Combustion-Synthesized
Titanium Diboride-Based Materials,’’ J. Am. Ceram. Soc., 78 [2] 275–84
(1995).
J. Jaroszewicz and A. Michalski, ‘‘Preparation of a TiB Composite with
2
a Nickel Matrix by Pulse Plasma Sintering with Combustion Synthesis,’’ J. Eur.
Ceram. Soc., 26, 2427–30 (2006).
ues of Ti BMg compositions sintered at 15001C were not sig-
2
2
nificantly dependent on the heating rate. The measured hardness
was 2471 GPa and the fracture toughness was 3.670.4
1
/2
3
MPa Á m . The measured values were probably affected by un-
reacted amorphous boron.
4
J. Schmidt Anwendung des Spark Plasma Sinterverfahrens fu¨r Pra¨para-
tionszwecke in der Festko¨rperchemie, Ph.D. Thesis, Dresden University of Tech-
IV. Conclusion
nology, 2003.
Z. A. Munir, ‘‘The Effect of External Electric Fields on the Nature and Prop-
5
In conclusion, it has been shown that the heating rate of the self-
propagating reaction has to be as high as possible to favor the
formation of a homogeneous microstructure of TiB from a
2
powder mixture of Ti and B. Thus, the use of high heating rates
that promote high reaction rates is essential even though there is
erties of Materials Synthesized by Self-Propagating Combustion,’’ Mater. Sci.
Eng., A287, 125–37 (2000).
H. Conrad, ‘‘Effects of Electric Current on Solid State Phase Transformations
6
in Metals,’’ Mater. Sci. Eng., A287, 227–37 (2000).
Z. A. Munir, U. Anselmi-Tamburini, and M. Ohyanagi, ‘‘The Effect of
Electric Field and Pressure on the Synthesis and Consolidation of Materials:
7

3
306
Communications of the American Ceramic Society
Vol. 90, No. 10
1
0
A Review of the Spark Plasma Sintering Method,’’ J. Mater. Sci., 41 [3] 763–77
2006).
R. C. Weast (ed.), CRC Handbook of Chemistry and Physics, 7th edition, CRC
Press Inc., Boca Raton, FL, 1987 p. F-120.
(
8
11
W. Chen, U. Anselmi-Tamburini, J. E. Garay, J. R. Groza, and Z. A. Munir,
‘Fundamental Investigations on the Spark Plasma Sintering/Synthesis Process-I.
Effect of DC Pulsing on Reactivity,’’ Mater. Sci. Eng., A394 [1–2] 132–8
2005).
J. K. Murray, P. K. Liao, and K. E. Spear, ‘‘The B–Ti (Boron–Titanium)
System,’’ Bull. Alloy Phase Diagrams, 7, 550–5 (1986).
‘
12
G. R. Anstis, P. Chantikul, B. R. Lawn, and D. B. Marshall, ‘‘A Critical
Evaluation of Indentation Techniques for Measuring Fracture Toughness: I,
(
9
U. Anselmi-Tamburini, S. Gennari, J. E. Garay, and Z. A. Munir, ‘‘Funda-
mental Investigations on the Spark Plasma Sintering/Synthesis Process-II.
Direct Crack Measurements,’’ J. Am. Ceram. Soc., 64 [9] 533–8 (1981).
Z-K. Liu, D. G. Schlom, Q. Li, and X. X. Xi, ‘‘Thermodynamics of the Mg–B
13
Modeling of Current and Temperature Distributions,’’ Mater. Sci. Eng., A394
1–2] 139–48 (2005).
System: Implications for the Deposition of MgB
78, 3678–80 (2001).
2
Thin Films,’’ Appl. Phys. Lett.,
&
[