Simultaneous spark plasma synthesis and densification of TiC-TiB 2 composites

Article abstract of DOI:10.1111/j.1551-2916.2005.00832.x

The simultaneous synthesis and densification of dense TiC-TiB₂ composite has been investigated starting from Ti, B₄C, and C as reactants, and using the spark plasma sintering (SPS) technique. The optimal conditions for complete conversion of the reactants to the composite were determined for different applied DC current levels. A kinetic investigation performed allows us to conclude that solid-state diffusion is the mechanism governing the synthesis process. It is seen that TiC is the first phase formed, while TiB₂ formation occurs afterward. Two intermediate boride phases, i.e. TiB and Ti₃B₄, are also formed but, as the SPS holding time was augmented, they were gradually and completely converted to TiB₂. Moreover, it is found that in order to reach relatively high dense products, an electric current needs to be applied for time intervals longer than those required for obtaining complete conversion. A pure dense product (relative density ～98%) was obtained when an electric pulsed current of 1100 A and a mechanical pressure of 20 MPa were applied for about 4 min.

Full text of DOI:10.1111/j.1551-2916.2005.00832.x

J. Am. Ceram. Soc., 89 [3] 848–855 (2006)
DOI: 10.1111/j.1551-2916.2005.00832.x
r 2006 The American Ceramic Society
ournal
J
Simultaneous Spark Plasma Synthesis and Densification of
TiC–TiB2 Composites
Antonio M. Locci, Roberto Orru,w and Giacomo Caow,z
Dipartimento di Ingegneria Chimica e Materiali, Centro Studi sulle Reazioni Autopropaganti (CESRA), and
Unita di Ricerca del Consorzio Interuniversitario per la Scienza e Tecnologia dei Materiali (INSTM),
Universita degli Studi di Cagliari, Piazza d’Armi, 09123 Cagliari, Italy
Zuhair A. Munir
Facility for Advanced Combustion Synthesis, Department of Chemical Engineering and Materials Science,
University of California, Davis, California 95616
The simultaneous synthesis and densification of dense TiC–TiB2
composite has been investigated starting from Ti, B4C, and C as
reactants, and using the spark plasma sintering (SPS) technique.
The optimal conditions for complete conversion of the reactants
to the composite were determined for different applied DC cur-
rent levels. A kinetic investigation performed allows us to con-
clude that solid-state diffusion is the mechanism governing the
synthesis process. It is seen that TiC is the first phase formed,
while TiB2 formation occurs afterward. Two intermediate boride
phases, i.e. TiB and Ti3B4, are also formed but, as the SPS
holding time was augmented, they were gradually and complete-
ly converted to TiB2. Moreover, it is found that in order to reach
relatively high dense products, an electric current needs to be
applied for time intervals longer than those required for obtain-
ing complete conversion. A pure dense product (relative density
B98%) was obtained when an electric pulsed current of 1100 A
and a mechanical pressure of 20 MPa were applied for about
4 min.
were synthesized by reactive sintering of powder mixtures of Ti
and B4C (3:1 molar ratio).6 Titanium carbide–titanium boride
composites of different compositions were also produced by
transient plastic-phase processing (TPPP) starting from Ti, B4C,
TiC0.5, and TiB2 powders.1,3,7
The self-propagating high-temperature synthesis (SHS) cou-
pled with explosive shock consolidation (ESC) was used to syn-
thesize dense TiC–TiB2 composites starting from Ti, C, and
B13C2 powders.8 Similar processes were utilized by Song et al.,9
who used a combination of SHS and dynamic compaction start-
ing from B4C, C, and Ti, and by Zhang et al.,2 who took ad-
vantage of pseudo-hot isostatic pressing (PHIP). TiC–2TiB2
composites were also fabricated from blends of Ti–B4C by us-
ing either thermal explosion (10001C, 50–150 MPa) or reactive
hot-pressing (11001C, 150 MPa) methods.10–12 Dense TiB2–15%
TiC composites were obtained by high-pressure sintering (HPS)
of TiC and TiB2 powders as well as by high-pressure self-com-
bustion synthesis (HPCS) starting from elemental reactants.4
Two different Ti–B–C composites were fabricated using reaction
synthesis of Ti and B4C in 3/1 and 4.8/1 molar ratios by hot
pressing (35 MPa, 18001C, 1 h).13 Dense nanometric TiB2–TiC
composites were recently generated by the field-activated pres-
sure-assisted synthesis apparatus (FAPASA) using elemental Ti,
C, and B powders that had been mechanically activated by high-
energy ball milling.14 A Ti–B–C composite was also prepared by
reaction hot pressing at 18001C and 35 MPa starting from Ti/
B4C powders premixed at a molar ratio of 4.8/1.15
In the present study, the simultaneous synthesis and consol-
idation of the titanium carbide–titanium diboride (TiC–
TiB2 5 1:1 molar ratio) composite using the spark plasma
sintering (SPS) method was investigated from a Ti, B4C, and
C powder mixture. This technique, which is based on the simul-
taneous application of a pulsed DC current and a uniaxial me-
chanical load through the reacting sample, has been previously
shown to be effective in the preparation of dense, bulk materi-
als.16–19
I. Introduction
ERAMIC composites containing carbides and borides of tran-
sition metals possess such desirable properties as high melt-
C
ing temperatures, low densities, good thermal and chemical
stability, high hardness, and excellent wear resistance.1,2 In par-
ticular, these composites have superior mechanical properties
(e.g., wear resistance and fracture toughness) relative to their
constituent ceramic components.3,4 This fact makes them at-
tractive as structural materials, where good mechanical proper-
ties and high-temperature resistance are required. In addition,
because of their high electrical conductivity, they can also be
used as materials for ‘‘Hall–Heroult’’ electrolytic cells for the
production of aluminum.2
Several methods have been utilized for the preparation of
dense TiC–TiB2 composites with varying proportions of carbide
and boride constituents.
For example, 2TiB2–TiC composites were formed through a
directional reaction between molten Ti and B4C preforms in a
furnace at 16001–18001C.5 Also, 2TiB2–TiC dense composites
Specifically, the effect of boron carbide particles size on the
conversion degree of reactants to the desired composite, the in-
fluence of the applied electric current level on the synthesis proc-
ess dynamics, and the effect of the processing time on product
densification were studied. Moreover, a kinetic investigation of
the synthesis process where the initial powders are transformed
to the final product is performed.
L. Klein—contributing editor
Manuscript No. 20568. Received May 16, 2005; approved October 11, 2005.
This work was financially supported by the MIUR-PROMEA (D.M. n.1015, 04/10/
2001), MIUR-PRIN (2002), and PRISMA-INSTM (2003), Italy. The support of this work
by the Army Research Office (ARO) to one of us (Z. A. M) is acknowledged.
wPROMEA Scarl, c/o Dipartimento di Fisica, Cittadella Universitaria di Monserrato,
Monserrato, Italy.
II. Experimental Materials and Methods
The characteristics of the starting powders used in the present
investigation along with their sources are reported in Table I.
Titanium, boron carbide (B4C), and amorphous carbon pow-
zAuthor to whom correspondence should be addressed. e-mail: cao@visnu.dicm.unica.it
848

March 2006
Simultaneous SPS and Densification of TiC–TiB2 Composites
849
Table I Starting Powders Used in the Present Investigation
Reactant
Vendor
Particle size
Purity (%)
Titanium (Ti)
Sigma-Aldrich (366994)
Alfa Aesar (40504)
Alfa Aesar (10922)
Alfa Aesar (39724)
ꢀ325 mesh
1–7 mm
22–59 mm
0.042 mm
99.98
99.4
991
Boron carbide (B4C)
Boron carbide (B4C)
Carbon black (C)
99.91
ders were dry mixed in the stoichiometric ratio corresponding to
the following chemical reaction:
two different types of boron carbide powders, with particle sizes
of 22–59 and 1–7 mm, were investigated. Figure 2 shows the
X-ray diffraction (XRD) patterns of samples obtained using the
larger (a) and the smaller size (b). The synthesis in both cases
was performed under SPS conditions of 800 A (total current)
and a sintering/reaction time of 10 min. It was seen that, when
the relatively coarse B4C particles were used, the conversion of
the initial reactants to the final products TiC and TiB2 was not
complete. In addition to the desired phases (TiC and TiB2), the
product contained unreacted B4C as well as other, undesirable,
titanium borides of TiB and Ti3B4. However, when the finer B4C
powder was used under the same SPS operating conditions, the
XRD pattern analysis of the product revealed the presence of
TiC and TiB2 only.
4Ti þ B4C þ C ! 2TiC þ 2TiB2
(1)
Powder mixing was performed in an SPEX 8000 mill for 30 min
using a stainless-steel jar and alumina balls.
An SPS 515 apparatus (Sumitomo Coal Mining Co. Ltd) was
used to synthesize and simultaneously densify the TiC–TiB2
composite. A schematic representation of the machine with de-
tails of the experimental setup is shown in Fig. 1. It combines a
50 kN uniaxial press with a DC-pulsed current generator (10 V,
1500 A, 300 Hz) to provide simultaneously a pulsed electric
current through the sample and the graphite die containing it,
together with a mechanical load through the die plungers.
About 4 g of powder mixture was first cold compacted inside
the die (outside diameter, 35 mm; inside diameter, 15 mm;
height, 40 mm). In order to protect the die and facilitate sam-
ple release after synthesis, a 99.8% pure graphite foil (0.13 mm
thick, Alfa Aesar, Karlsruhe, Germany) was inserted between
the internal surfaces of the die and the top and the bottom sur-
face of the sample and the graphite plungers (14.7 mm diameter,
20 mm height). Both the die and the plungers were made of
E-940 graphite and provided by Electrodes Inc. In addition, with
the aim of minimizing heat losses by thermal radiation, the die
was covered with a layer of graphite felt. The die was then placed
inside the reaction chamber of the SPS apparatus and the system
was evacuated to 10 Pa. This step was followed by the applica-
tion of a mechanical pressure of 20 MPa through the plungers.
The experiment was initiated with the application of a previ-
ously set constant value of the electric current for a given time.
Temperatures were measured during synthesis by a C-type ther-
mocouple, which was inserted inside a small hole on one side of
the graphite die. Temperature, applied current, voltage, mechan-
ical load, and the vertical displacement (d) of the lower electrode
were recorded in real time. It is worth noting that the parameter
d is a measure of compact densification. However, because of the
thermal expansion of the sample as well as that one of both
electrodes, graphite blocks, spacers, and plungers, the displace-
ment d is not exactly equal to sample shrinkage. Therefore, the
real degree of consolidation was determined by measuring the
density of the sample at the end of the process. For the sake of
reproducibility, each experiment was repeated at least twice.
After the synthesis process, the sample was allowed to cool
and then removed from the die. The relative density of the
product was determined by geometric and gravimetric measure-
ments and also by the Archimedes method. Phase identification
was performed using a Philips PW 1830 X-rays diffractometer
(Eindhoven, the Netherlands) with CuKa radiation (l 5 1.5405
The above results can be explained on the basis of the longer
diffusion paths involved in the chemical transformation when
relatively coarse B4C powders are used. Therefore, in order to
investigate the reaction kinetics, the finer B4C powders (particle
size 1–7 mm) were used.
(2) Kinetic Study of the Synthesis Process
The mechanism of formation of the TiC–TiB2 composite by the
SPS process was studied by simultaneously applying a constant
pulsed electric current (I 5 800 A) and mechanical pressure
(P 5 20 MPa), for different time intervals (tIa0), ranging from
1 to 10 min.
Fixed Stainless
Steel Electrode
I
Graphite Block
Graphite
Spacer
DC Pulse
Current
Generator
Graphite Felt
~
Graphite Die
Sample
Graphite
Plungers
˚
A) and an Ni filter. The microstructure of end products was ex-
amined by scanning electron microscopy (SEM) using a Hitachi
S4000 microscope (Tokyo, Japan).
Knoop microhardness measurements of the SPS composite
using a 0.9806 N (100 gf) indentation load and 15 s dwell time
were also performed.
Moving Stainless
Steel Electrode
III. Results and Discussion
(1) Effect of Boron Carbide Particle Size
Mechanical Load
Fig. 1. Schematic representation of the spark plasma sintering apparatus.
The influence of B4C particles size on the TiC–TiB2 synthesis
process by the SPS method was investigated first. Specifically,

850
Journal of the American Ceramic Society—Locci et al.
Vol. 89, No. 3
respondingly, the voltage showed a rapid increase, reaching a
maximum value of about 6.4 V within about 20 s, and then de-
creasing rapidly to a constant value of about 4.5 V (Fig. 3(b)).
The decrease in voltage was a result of the increase in temper-
ature of the reacting system, which produces an increase in its
electrical conductivity. Under the given experimental conditions
(I 5 800 A, P 5 20 MPa, tIa0 5 10 min), the temperature in-
creased rapidly during the first 2 min, and then it increased at a
slower rate, approaching an asymptotic value of about 17001C
after a holding time of about 5 min (Fig. 3(c)). The conditions
indicate an approach to thermal equilibrium. Regarding the
profile of sample displacement (Fig. 3(d)), no significant changes
are observed during the first 60 s. A linear increase (to about
0.5–0.6 mm) then takes place during the next minute (60–120 s)
and is followed by a steep increase (to 1.25 mm) in the time
range of 120–150 s. The displacement then decreases gradually
to reach an approximately constant value at about 300 s. The
observed increase of d when the application of the electric cur-
rent is interrupted (Fig. 3(d)) was a manifestation of the shrink-
age of the sample ensemble (die/plungers/sample) because of
thermal shrinkage accompanying the cooling of the system. A
qualitatively similar behavior was also observed when current
values other than 800 A were applied. In order to investigate the
kinetics associated with the SPS process of TiC–TiB2 composites
based on Eq. (1), the reacting system composition was moni-
tored at different time intervals (tIa0 5 1, 2, 3, 4, 5, 6, 7, 8, and
10 min). The XRD patterns, including the one related to the
starting powder mixture, are shown in Figs. 4(a) to (j), respec-
tively. The range for the diffraction angle (2y) in these patterns
(331 and 461) includes, in addition to the main peaks of TiB2 and
TiC, the main peaks of Ti and B4C. As the carbon was amor-
(b)
(a)
33 34 35 36 37 38 39 40 41 42 43 44 45 46
Diffraction angle, 2Θ
Fig. 2. Effect of B4C particle size on the composition of the final spark
plasma sintering product (I 5 800 A, P 5 20 MPa, tIa0 510 min): (a)
22–59 mm; (b) 1–7 mm ( & , B4C; J, TiC; n, TiB; &, Ti3B4;
ꢁ
, TiB2).
The applied current, voltage, temperature, and shrinkage dis-
placement profiles measured during the synthesis process for the
case of tIa0 5 10 min are shown in Fig. 3. It can be seen from
Fig. 3(a) that the current increased from 0 to 800 A in about 20 s
and remained constant until the end of the desired time. Cor-
900
800
700
600
500
400
300
200
100
0
7
(a)
(b)
6
5
4
3
2
1
0
t
t
t
t
t
t
t
t
8
1
2
3
4
5
6
7
0
60 120 180 240 300 360 420 480 540 600 660 720 780 840 900
0
60 120 180 240 300 360 420 480 540 600 660 720 780 840 900
Time, t [s]
Time, t [s]
2000
1800
1600
1400
1200
1000
800
2.5
2.0
1.5
1.0
0.5
0.0
(c)
(d)
600
400
200
0
0
60 120 180 240 300 360 420 480 540 600 660 720 780 840 900
0
60 120 180 240 300 360 420 480 540 600 660 720 780 840 900
Time, t [s]
Time, t [s]
Fig. 3. Temporal profiles of spark plasma sintering outputs: (a) current intensity, (b) voltage, (c) temperature, and (d) sample displacement (I 5800 A,
P 5 20 MPa); t1 51 min, t2 5 2 min, t3 5 3 min, t4 5 4 min, t5 5 5 min, t6 56 min, t7 5 7 min, t8 5 8 min.

March 2006
Simultaneous SPS and Densification of TiC–TiB2 Composites
851
ꢃ
ꢁ
(f)
(a)
ꢁ
ꢀ
ꢀ
ꢃ
ꢃ
ꢂ
ꢂ
ꢂ
ꢂ
(g)
(b)
ꢃ
ꢀ
ꢁ
ꢁ
ꢀ
ꢃ
ꢃ
ꢂ
ꢂꢂ
ꢂ
33
(h)
(c)
(d)
(e)
ꢁ
ꢀ
ꢀ
ꢀ
ꢁ
ꢁ
ꢁ
ꢁ
ꢃ
ꢁ
ꢀ
ꢃ
ꢃ
ꢀ
ꢀ
ꢂ
ꢂ
ꢂ
ꢂ
ꢂ
(i)
ꢁ
ꢁ
ꢁ
ꢀ
ꢀ
ꢀ
ꢂ
ꢂ
(j)
ꢁ
ꢁ
ꢁ
ꢀ
ꢀ
ꢀ
ꢂ
ꢂ
ꢂ
33 34 35 36 37 38 39 40 41 42 43 44 45 46
33 34 35 36 37 38 39 40 41 42 43 44 45 46
Diffraction Angle, 2θ
Diffraction Angle, 2θ
Fig. 4. X-ray diffraction (XRD) patterns of spark plasma sintering products obtained for different values of the time interval during which the pulsed
electric current is applied (I 5 800 A, P 5 20 MPa). (a) Reactants; (b) tIa0 5 1min; (c) tIa0 5 2 min; (d) tIa0 5 3 min; (e) tIa0 5 4 min; (f) tIa0 5 5 min; (g)
tIa0 5 6 min; (h) tIa0 5 7 min; (i) tIa0 58 min; (j) tIa0 5 10 min (^, Ti; & , B4C; J, TiC; n, TiB; &, Ti3B4;
ꢁ
, TiB2).
phous, it was not identified by this analysis. The application of a
current (800 A) for 60 s showed that no significant interaction
between the reactants took place, as seen from the pattern of
Fig. 4(b), which depicts peaks for Ti and B4C only. This obser-
vation is related to the corresponding temperature profile
(Fig. 3(c)). The maximum temperature measured during this

852
Journal of the American Ceramic Society—Locci et al.
Vol. 89, No. 3
stage of the SPS process was less than 6001C, too low to activate
the solid-state diffusion processes necessary for the synthesis re-
action. When the reaction time is increased to 2 min, the desired
end phases, TiB2 and TiC, were now present in the final product,
particularly the latter one (Fig. 4(c)). However, a complete con-
version to the desired product had not yet been achieved, as can
be seen by the presence of titanium and boron carbide peaks,
along with peaks of the intermediate phases of TiB and Ti3B4, in
the XRD pattern of the SPS sample. The increase in the inter-
action between reactants was a result of the increase in temper-
ature to about 12001C (Fig. 3(c)).
tion to TiB2 through two reaction steps, as follows:
16TiB þ B4C ! 5Ti3B4 þ TiC
3Ti3B4 þ B4C ! 8TiB2 þ TiC
(4)
(5)
In the second path, titanium monoboride is directly converted
into the diboride phase according to
6TiB þ B4C ! 5TiB2 þ TiC
(6)
Both reaction paths, i.e. Eqs. (3)–(5) and the one involving Eqs.
(3) and (6), give rise to the same final composite.
Although the reaction investigated by Zhao and Cheng6 was
different from that examined in the present work (cf. Eq. (1)) the
latter one can be considered as the sum of Eq. (2) and the fol-
lowing:
On the basis of the results described above, it can be con-
cluded that the main reaction taking place during the first stage
of the SPS process was the interaction between titanium and
amorphous carbon to form TiC.
As the reaction time is increased to 3 min (Fig. 4(d)), the
product contained large amounts of both TiC and TiB2, with
only small traces of B4C and the secondary products TiB and
Ti3B4. Elemental titanium was not present in the product as
judged by the XRD pattern. This result is in good agreement
with the experimental evidences reported by Zhao and Cheng,6
who investigated the formation of TiC–TiB2 composites by re-
active sintering. In fact, they found that the interaction between
Ti and B4C becomes significant only when T413001C. Accord-
ingly, our results show that the complete conversion of titanium
takes place in the time interval of 2–3 min, during which this
temperature threshold (13001C) is reached (cf. Fig. 3(c)). In ad-
dition, as the corresponding temperature for this stage is con-
siderably lower than the melting point of Ti, this observation
indicates that the synthesis reaction is a solid-state process.
With an increase in reaction time to 4 min (Fig. 4(e)), the
relative amount of TiB2 increased while the TiB phase appeared
to decrease and the Ti3B4 increased. In addition, traces of B4C
were still found in the product.
When the synthesis time was increased to 5 min, B4C was not
detected in the XRD pattern (Fig. 4(f)) and the amounts of TiB2
and Ti3B4 still increased at the cost of TiB. The monoboride (TiB)
completely disappeared when the reaction time was increased to 6
min (Fig. 4(g)). Moreover, and in contrast to the previous reac-
tion times, the relative abundance of the phase Ti3B4 is now lower
with only small amounts present in the product. This trend con-
tinues as the reaction time was increased to 7 min, with only a
trace of Ti3B4 found in the product (Fig. 4(h)). Ti3B4 disappeared
when the reaction time was increased to 8 min, with the XRD
pattern now showing peaks for TiC and TiB2 only (Fig. 4(i)).
As the results presented above indicate, the composition of
the product depends on the reaction time under the influence
of the pulsed electric current. In addition, the rate of conversion
of the starting reactants is different for the two ceramic com-
ponents of the composite. Under the experimental conditions
utilized, titanium reacted completely while in the solid state.
This observation, along with the fact that the maximum tem-
perature measured during the process was about 18001C and
that the eutectic temperature of the TiC–TiB2 system was ap-
proximately 25001C,20 allows us to conclude that the formation
of the desired composite by the SPS process is governed by a
solid-state diffusion mechanism.
Ti þ C ! TiC
(7)
which represents the direct carburization of elemental Ti by
amorphous carbon.
Based on the results obtained during the kinetic investigation
of the SPS process, it is possible to postulate that Eq. (7) is the
first step during the spark plasma synthesis of TiC–TiB2. More-
over, Eqs. (3)–(5) are more likely responsible for the subsequent
TiB2 formation, as well as the intermediate phases, i.e. TiB and
Ti3B4, detected before the synthesis reaction was completed.
In fact, while the possibility of Ti3B4 formation is excluded by
Eq. (6), this phase was the last to disappear from the product
(Figs. 4(g)–(h)), suggesting that the sequence of Eqs. (4) and (5)
is the likely path in the synthesis of the composite.
The presence of B4C is then required to guarantee the com-
plete transformation of TiB and Ti3B4 into TiB2. The fact that
boron carbide was not detected by XRD for t ꢂ 5 min is con-
sistent with its relatively low peak intensity even when the cor-
responding content was maximum, i.e. in the starting mixture
(cf. Fig. 4(a)).
It is now possible to correlate the results shown in Fig. 4 with
those reported in Fig. 3(d), where the displacement profile meas-
ured during the synthesis/densification process is reported. In
particular, the linear increase in d, observed in the range 60–120
s, corresponds to the formation of titanium carbide. Therefore,
the sample displacement recorded during this stage can be re-
lated primarily to the carburization process of elemental titani-
um by amorphous carbon. Although the first evidence of
formation of titanium borides (TiB, Ti3B4, and TiB2) was indi-
cated, it was at a much limited level.
In addition, it should be noted that this stage corresponds to
the maximum transformation rate of the starting reactants.
However, this does not lead to the complete conversion into
the desired phases, which is reached only when the slow con-
version of the titanium borides TiB and Ti3B4 to the desired
TiB2 takes place. The completion of such a conversion occurs at
8 min (cf. Fig. 4(i)), which therefore represents the required du-
ration of current application leading to the formation of the
desired compounds, i.e. TiC and TiB2. This aspect will be fur-
ther examined in the next section, where different current levels
have been applied.
By studying the formation of the 2TiB2–TiC composite by
reactive sintering based on the following reaction:
(3) Effect of Current Intensity
3Ti þ B4C ! 2TiB2 þ TiC
(2)
The influence of the pulsed electric current intensity (I) on the
simultaneous synthesis and densification of the TiC–TiB2 com-
posite was investigated in the range 700–1100 A, while the
mechanical pressure was maintained constant at 20 MPa in all
SPS experiments. In particular, the dependence of the synthesis
time, i.e. the minimum time interval during which the applied
current results in complete conversion, on the current intensity
was investigated first. To this aim, the same approach described
in the previous paragraph, i.e. monitoring the evolution of the
synthesis process through quenched reactions at different inter-
vals (cf. Figs. 3 and 4), was adopted in the range 700–1100 A.
Zhao and Cheng6 suggested that the first step occurring during
the transition from the starting powders to the final composites
is represented by the formation of TiB and TiC, through the
following reaction:
5Ti þ B4C ! 4TiB þ TiC
(3)
Then, two reactions are proposed as possible paths. The first
involves the formation of Ti3B4 and its subsequent transforma-

March 2006
Simultaneous SPS and Densification of TiC–TiB2 Composites
853
1000
900
800
700
600
500
400
300
200
100
7000
2.8x106
2.4x106
2.0x106
6500
6000
5500
5000
4500
4000
3500
3000
2500
1.6x106
1.2x106
700
800
900
Current, I [A]
1000
1100
700
800
900
1000
1100
Fig. 6. Effect of applied current on the total energy (Eq. (8)) and the
average power (Eq. (9)) needed to achieve the full conversion of the in-
itial reactants into the final desired product (P 5 20 MPa).
Current, I [A]
Fig. 5. Effect of the applied current on the time needed to obtain the
full conversion of the initial reactants into the final desired product
(P 5 20 MPa).
of the SPS samples (theoretical value equal to 4.678 g/cm3) ob-
tained by applying the electric current I for the corresponding
This procedure allowed us to identify the minimum time interval
needed to achieve the full conversion, hereafter referred to with
Ã
minimum time t . It can be seen that higher applied currents
Ã
produced denser TiC–TiB2 composites. Specifically, while the
relative density of the end-product obtained when a current in-
tensity of 800 A is applied for 8 min is about 84%, it increases up
to 87% when a current intensity of 1100 A is applied for 3 min.
This result may be justified on the basis of the considerations
made above, and related to the synthesis time. In fact, when
relatively higher current densities are applied, the averaged elec-
the symbol t . For instance, as observed previously, for the case
Ã
of I 5 800 A, t was found to be equal to 480 s. The effect of the
applied current intensity on t is shown in Fig. 5. It can be seen
Ã
that the conversion time decreased monotonically as the applied
current level increased. Full conversion times decreased from
960 to 180 s as the current intensity increased from 700 to 1100
A.
ꢀ
trical power, W, supplied to the system increases, thus increas-
ing its temperature level and the heating rate, which would
enhance the sintering process.
This finding may be explained on the basis of the different
temperature conditions observed when applying different cur-
rent intensities. In fact, the increase of the applied current re-
sulted in both higher heating rate and temperature levels
reached, both of which accelerate synthesis processes.
The results reported in Fig. 5 can be used as a basis for eval-
uating energy costs for the production of a TiC–TiB2 ceramic
composite by the SPS process using Ti, B4C, and C powders as
reactants. Using the electric current (I) and the corresponding
(4) Effect of SPS Processing Time on Product Densification
From the results shown above, it can be seen that the product
densities are relatively low (the maximum value reached was
about 87%). Regarding this aspect, it should be noted that a
further increase in the current was not considered, as it would
cause several practical problems, particularly related to safety,
because of higher process temperatures, as well as difficulties in
sample extraction from the graphite die, and increased wear of
the die itself. However, with the aim of improving the densifi-
cation of the TiC–TiB2 composite, the influence of the time in-
terval during which the pulsed current is applied on the final
density of the SPS product was then investigated.
Ã
minimum application time (t ) from Fig. 5, it is possible to
calculate the energy consumed during the reaction synthesis
through the following expression:
Z
tꢃ
E ¼
IðtÞVðtÞdt
(8)
0
where V is the voltage, whose time profile is known (Fig. 3(b) for
the case of I 5 800 A). Therefore, the minimum energy required
to obtain the desired composite can be calculated from Eq. (8)
for each current level. Besides the total energy requirement for
the synthesis process, another parameter that can be considered
useful is the averaged electrical power evaluated as follows:
90
88
86
84
82
Z
tꢃ
1
tꢃ
ꢀ
W ¼
IðtÞVðtÞdt
(9)
0
Ã
i.e. the total energy supplied divided by the time t needed for
obtaining the complete conversion. Both the parameters defined
above in Eq. (8) and Eq. (9) are reported in Fig. 6 as a function
of the applied current. It can be clearly seen that the total elec-
trical energy E necessary to obtain the complete conversion of
the reactants to the desired phases decreased as the current level
increased. This was a consequence of the corresponding reduc-
Ã
ꢀ
tion of t . On the other hand, Wincreased as I increased. This
result is consistent with the relatively higher heating rates and
temperature levels reached during the process, and clearly jus-
tifies the corresponding shorter synthesis time, i.e. higher reac-
tion rates, observed experimentally.
600
700
800
900
1000
1100
1200
Current, I [A]
Fig. 7. Effect of applied electric current on the density of the obtained
spark plasma sintering samples (P 5 20 MPa, theoretical density 5 4.678
g/cm3).
The effect of the applied electric current on product densifi-
cation was also investigated. Figure 7 shows the relative density

854
Journal of the American Ceramic Society—Locci et al.
Vol. 89, No. 3
tensity of 1100 A for 4 min are shown in Fig. 9. Two different
regions can be observed easily. Specifically, the dark one corre-
sponds to the TiB2 phase, while the light one is TiC.
100
98
96
94
92
90
88
86
84
Knoop microhardness measurements using a 0.9806 N in-
dentation load and a 15 s load duration time were performed for
the case of the TiC–TiB2 composite synthesized under the fol-
lowing SPS conditions: I 5 1100 A, tIa0 5 4 min, and P 5 20
MPa. The hardness value obtained in this work, 25007100 HK,
was compared with the corresponding values available in the
literature. Specifically, our result is in good agreement with the
reported values of the two ceramic constituents of the synthe-
sized composite, i.e. 2470 HK for TiC and 2710–3000 HK for
TiB2,22 both obtained using a 100 gf indentation load as in our
case.
Moreover, the obtained values are very similar to, and in
some cases higher than, the best results reported in the literature
for TiC–TiB2 dense composites having the same stoichiometry.
In fact, the hardness of the synthesized SPS material is signif-
icantly higher than the values reported by Zhang et al.2 and Lee
et al.14 The observed differences may be explained on the basis
of the higher product density achieved during the present inves-
tigation, i.e. about 98% of the theoretical value. The densities of
the samples reported by Zhang et al.2 and Lee et al.14 were about
91% and less than 96.5%, respectively. On the other hand,
Yuan et al.23 have produced TiC–TiB2 composites characterized
by a very high density (99.8%) and hardness. However, the syn-
thesis method suggested in this case is very complex and involves
three different steps, i.e. thermite-type SHS of TiC/TiB2/oxides
powder mixture, leaching to remove the oxide phases, and sub-
sequent hot pressing (HP) of the leached powders. Therefore,
the above comparison suggests that through the SPS method, it
is possible to obtain a dense TiC–TiB2 composite characterized
by superior hardness properties using a one-step synthesis proc-
ess.
3
4
5
6
7
8
time [min]
Fig. 8. Densities of end products obtained by spark plasma sintering as
a function of the application time of the applied current (I 5 1100 A,
P 5 20 MPa).
Based on the results obtained previously, this study has been
carried out using an electric current equal to 1100 A, because it
corresponded to the higher relative density reached at this stage
(Fig. 7). In particular, as under this current level condition the
desired composition was reached after 3 min (Fig. 5), the time
interval investigated for densification purposes was in the range
3–8 min. The results are reported in Fig. 8. It can be seen that as
the time was increased from 3 to 4 min, the relative density of
the product increased markedly, from 87% to about 98%. A
further increase of the synthesis time up to 8 min did not result
in significant changes in the product density.
These observations are important as they provide evidence
that the conversion into the product contributes only modestly
to the densification. The goal of higher density is achieved after
the completion of the reaction through the sintering of the prod-
uct. This is another, albeit indirect, evidence for the occurrence
of the synthesis reaction between solid-state phases. If a liquid
had been present during the reaction, significant densification
would have been observed during that stage. The observation in
this study is quite different from the one that was found recently
when synthesizing dense WC-6%Co by SPS,21 where synthesis
and densification phenomena to produce fully dense composites
were observed to occur simultaneously, and the complete con-
version of starting reactants to the carbide phase corresponded
to the production of a fully dense material. Two back-scattered
SEM micrographs at different magnifications of the SPS com-
posite synthesized by applying a pulsed electric current of in-
IV. Concluding Remarks
A TiC–TiB2 dense composite has been synthesized starting with
Ti, B4C, and C by a one-step process using the SPS apparatus.
It was shown that the use of a finer B4C powder resulted in
higher conversion levels to the desired product.
An investigation of the evolution of the synthesis process
(through quenched reactions at different intervals) provided an
insight into the mechanism of product formation. It was seen
that TiC is the first phase formed during the process as a con-
sequence of the carburization of Ti, while TiB2 formation took
place subsequently. In addition, by examining product compos-
itional changes with sample displacement in parallel, it was
found that the latter stage was associated with the most rapid
and significant sample shrinkage. Two intermediate boride
phases, i.e. TiB and Ti3B4, were also formed but, as the SPS
Fig. 9. Scanning electron microscopy back-scattered micrograph of spark plasma sintering end product at different magnifications: (a) ꢄ 200 and (b)
ꢄ 1500 (I 5 1100 A, tIa0 5 4 min, P 5 20 MPa).

March 2006
Simultaneous SPS and Densification of TiC–TiB2 Composites
855
5S. K. Lee, D. H. Kim, and C. H. Kim, ‘‘Fabrication of TiB2/TiC Composites
by the Directional Reaction of Titanium with Boron Carbide,’’ J. Mater. Sci., 29
[15] 4125–30 (1994).
holding time was increased, they were gradually converted to
TiB2. Moreover, because of the relatively low temperature levels
recorded during the entire process, it is possible to conclude that
all transformations during the synthesis of the desired composite
by SPS are governed by a solid-state diffusion mechanism.
A systematic investigation of the effect of the electric current
on the synthesis process made it possible to identify the optimal
time intervals during which the pulsed current is applied, thus
leading to the achievement of full conversion for all current in-
tensity levels considered. Specifically, it was found that the ap-
plication of relatively higher currents produces not only the
reduction of the time intervals mentioned above but also a de-
crease of the corresponding energy provided to the reacting sys-
tem for the synthesis, as well as the fabrication of denser
material. However, in order to obtain a relatively high-density
product, the electric current needs to be applied for time inter-
vals longer than those required for obtaining complete conver-
sion. A pure dense product (relative density of about 98%) was
obtained when an electric pulsed current of 1100 A and a me-
chanical pressure of 20 MPa were applied for about 4 min, i.e.
1 min after full conversion was achieved.
6H. Zhao and Y. B. Cheng, ‘‘Formation of TiB2–TiC Composites by Reactive
Sintering,’’ Ceram. Int., 25 [4] 353–8 (1999).
7D. Brodkin, A. Zavaliangos, S. R. Kalidindi, and M. W. Barsoum, ‘‘Ambient-
and High-Temperature Properties of Titanium Carbide Titanium Boride Com-
posites Fabricated by Transient Plastic Phase Processing,’’ J. Am. Ceram. Soc., 82
[3] 665–72 (1999).
8R. Tomoshige, Y. Kakoki, K. Imamura, and A. Chiba, ‘‘Effect on Addition of
Titanium Diboride to Titanium Carbide Produced by the SHS Shock Consolida-
tion Method,’’ J. Mater. Process. Technol., 85 [1–3] 105–8 (1999).
9I. Song, L. Wang, M. Wixom, and L. T. Thompson, ‘‘Self-Propagating High
Temperature Synthesis and Dynamic Compaction of Titanium Diboride/Titanium
Carbide Composites,’’ J. Mater. Sci., 35 [10] 2611–7 (2000).
10I. Gotman, N. A. Travitzky, and E. Y. Gutmanas, ‘‘Dense In Situ TiB2/TiN
and TiB2/TiC Ceramic Matrix Composites—Reactive Synthesis and Properties,’’
Mater. Sci. Eng. A, 244 [1] 127–37 (1998).
11E. Y. Gutmanas and I. Gotman, ‘‘Dense High-Temperature Ceramics by
Thermal Explosion Under Pressure,’’ J. Eur. Ceram. Soc., 19 [13–14] 2381–93
(1999).
12E. Y. Gutmanas and I. Gotman, ‘‘Reactive Synthesis of Ceramic Matrix
Composites Under Pressure,’’ Ceram. Int., 26 [7] 699–707 (2000).
13G. Wen, S. B. Li, B. S. Zhang, and Z. X. Guo, ‘‘Reaction Synthesis of TiB2–
TiC Composites with Enhanced Toughness,’’ Acta Mater., 49 [8] 1463–70 (2001).
14J. W. Lee, Z. A. Munir, and M. Ohyanagi, ‘‘Dense Nanocrystalline TiB2–TiC
Composites formed by Field Activation from High-Energy Ball Milled Reac-
tants,’’ Mater. Sci. Eng. A, 325 [1–2] 221–7 (2002).
15S.-B. Li, B.-S. Zhang, G. Wen, and J. X. Xie, ‘‘Microstructure and Mechanical
Properties of Platelet-Reinforced Ti–B–C Ceramics Prepared by Reaction Hot
Pressing of B4C and Ti Powders,’’ Mater. Lett., 57 [8] 1445–52 (2003).
16M. Omori, ‘‘Sintering, Consolidation, Reaction and Crystal Growth by Spark
Plasma Sintering (SPS),’’ Mater. Sci. Eng. A, 287 [2] 183–8 (2000).
17M. Kawahara, H-T. Kim, and M. Tokita, ‘‘Fabrication of Nano-Materials by
Spark Plasma Sintering (SPS) Method’’; pp. 741–4 in Proc. 2000 Metallurgy World
Congress. Japan Society of Powder and Powder Metallurgy, 2000.
18R. Orru, J. Woolman, G. Cao, and Z. A. Munir, ‘‘The Synthesis of Dense
Nanometric MoSi2 through Mechanical and Field Activation,’’ J. Mat. Res., 16 [5]
1439–48 (2001).
Acknowledgments
The help extended by Mrs. Gloria Arangino is also gratefully acknowledged. In
addition, the authors would like to thank Prof. Tiberio Bacci, University of Flor-
ence (Italy), for structural characterization of SPS samples.
References
1M. W. Barsoum and B. Houng, ‘‘Transient Plastic Phase Processing of Tita-
nium Boron Carbon Composites,’’ J. Am. Ceram. Soc., 76 [6] 1445–51 (1993).
2X. H. Zhang, C. C. Zhu, W. Qu, X. D. He, and V. L. Kvanin, ‘‘Self-Propa-
gating High Temperature Combustion Synthesis of TiC/TiB2 Ceramic–Matrix
Composites,’’ Compos. Sci. Technol., 62 [15] 2037–41 (2002).
19Z. J. Shen, M. Johnsson, Z. Zhao, and M. Nygren, ‘‘Spark Plasma Sintering of
Alumina,’’ J. Am. Ceram. Soc., 85 [8] 1921–7 (2002).
20H. Holleck, H. Leiste, and W. Schneider, ‘‘Significance of Phase Boundaries in
Wear Resistant TiC/TiB2 Materials. I. Investigations with Transmission Electron
Microscopy,’’ Int. J. Refract. H. Met., 6 [3] 149–54 (1987).
3D. Brodkin, S. R. Kalidindi, M. W. Barsoum, and A. Zavaliangos, ‘‘Micro-
structural Evolution During Transient Plastic Phase Processing of Titanium
Carbide–Titanium Boride Composites,’’ J. Am. Ceram. Soc., 79 [7] 1945–52
(1996).
21A. M. Locci, R. Orru, and G. Cao, ‘‘Simultaneous Spark Plasma Synthesis
and Consolidation of WC/Co Composites,’’ J. Mater. Res., 20 [3] 734–41 (2005).
22CRC, Materials Science and Engineering Handbook, 3rd edition, CRC Press
LLC, Boca Raton, FL, 2001.
4S. K. Bhaumik, C. Divakar, A. K. Singh, and G. S. Upadhyaya, ‘‘Synthesis and
Sintering of TiB2 and TiB2–TiC Composite Under High Pressure,’’ Mater. Sci.
Eng. A, 279 [1–2] 275–81 (2000).
23R. Yuan, Z. Fu, W. Wang, and H. Wang, ‘‘SHS TiB2-Based Multiphase Ce-
ramics and Composites,’’ Int. J. SHS, 10 [4] 435–50 (2001).
&