Combustion synthesis of silicon carbide assisted by a magnesium plus polytetrafluoroethylene mixture

Article abstract of DOI:10.1016/j.materresbull.2009.07.002

In this study, the use of SiC combustion synthesis for immobilization of ¹⁴C was considered. Due to the low exothermicity of the reaction between silicon and graphite, a highly exothermic mixture (magnesium and polytetrafluoroethylene) was used

Full text of DOI:10.1016/j.materresbull.2009.07.002

Materials Research Bulletin 44 (2009) 2134–2138
Contents lists available at ScienceDirect
Materials Research Bulletin
journal homepage: www.elsevier.com/locate/matresbu
Combustion synthesis of silicon carbide assisted by a magnesium plus
polytetrafluoroethylene mixture
b
R.M. Ayral a, F. Rouessac a, , N. Massoni
*
a Institut Charles Gerhardt Montpellier, PMOF-UM2-CNRS Pl. E. Bataillon, 34095 Montpellier Cedex 5, France
b CEA, DEN, DTCD, 30207 Bagnols-sur-Ce`ze, France
A R T I C L E I N F O
A B S T R A C T
In this study, the use of SiC combustion synthesis for immobilization of 14C was considered. Due to the
low exothermicity of the reaction between silicon and graphite, a highly exothermic mixture
(magnesium and polytetrafluoroethylene) was used both as a chemical oven and activate additive in the
mixture. With this configuration the reaction between graphite and silicon was initiated and propagated
on the whole sample. The self-propagating high temperature synthesis samples were characterized by
using scanning electron microscopy and X-ray diffraction.
Article history:
Received 30 April 2009
Received in revised form 29 June 2009
Accepted 2 July 2009
Available online 10 July 2009
Keywords:
ß 2009 Elsevier Ltd. All rights reserved.
A. Carbides
A. Ceramics
C. Electron microscopy
C. X-ray diffraction
1. Introduction
(Ti + Si + C), the temperature and the rate of propagation of the
combustion reaction decrease giving rise at least to an extinction of
Silicon carbide could be a promising matrix for immobilization
of 14C due to its high thermal stability, good resistance and
corrosion [1]. This material can be synthesized through many
methods; most of it is produced by the Acheson process where
thermal reduction of silicon dioxide by carbon is carried out at
temperatures ranging from 2400 8C to 3000 8C. This process is
quite expensive and now other synthesis methods are investi-
gated. Among the variety of methods, self-propagating high
temperature synthesis (SHS) has become interesting because of
energy savings and shorter processing times [2–4].
An empirical criterion typically adopted for determining the
feasibility of combustion synthesis is an adiabatic temperature
exceeding 1800 K [5]. For the Si–C system, the calculated adiabatic
temperature is 1860 K which is too low to enable the combustion
synthesis reaction. Hence, an additional energy source must be
introduced into the system like preheating temperature [6,7], field
assisted synthesis [8,9] or chemical oven for instance [10]. One
other possible way is to use active gas like nitrogen [11] to form
nitrides whose adiabatic temperatures can reach 5000 K.
the reaction when 80 at% of silicon is added. In this case, silicon
acts as a diluent and high percentages of this element extinguishes
the reaction.
Another way to enhance the reactivity of the powders is to use a
high caloric mixture like (Mg + polytetrafluoroethylene (PTFE)) for
example. Several papers referring to the use of Teflon in
exothermic reactions exist in the literature [13–16]. Studies made
by Zhang et al. [17] concern the use of this mixture as an activate
additive in the reaction Si + C with different types of carbon (black
carbon, activated carbon and graphite). In that paper, the authors
show that the reaction between silicon and graphite is very
difficult to ignite even with the mixture Mg + PTFE.
To investigate the reaction with graphite, which has been
thoroughly used for nuclear applications in the past, we tried to
change the conditions established by Zhang and to use both
activate additive and chemical oven protected by a thermal
insulator in order to reduce heat losses. This paper will describe the
experimental procedure and the influence of composition of
mixture on temperatures and compositions of the product. The
characterizations will essentially be made by X-ray diffraction and
scanning electron microscopy.
Previous studies were undertaken using a powder mixture
containing titanium, graphite and silicon powders [12]. We have
shown that when percentage of silicon increases in the mixture
2. Experimental procedure
Silicon (Alfa Aesar 1–5
mm), graphite (Alfa Aesar, 44 mm),
* Corresponding author. Tel.: +33 0 467143355; fax: +33 0 467144290.
E-mail address: florence.rouessac@univ-montp2.fr (F. Rouessac).
polytetrafluoroethylene (PTFE) (Acros Organics) and magnesium
0025-5408/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.materresbull.2009.07.002

R.M. Ayral et al. / Materials Research Bulletin 44 (2009) 2134–2138
2135
powders (Fisher Scientific) were used as reactants. They were
mechanically milled in a jar closed under argon during 12 h. Si, C
reactants were weighed with different molar ratios of 1:2 and 1:1
and Mg and PTFE powders at a ratio of 23 wt.%. In the following, the
mixtures will be labelled as Si:C = 1:2 (molar ratio) À Mg:PTFE
= 23:23 (wt.%) and Si:C = 1:1 À Mg:PTFE = 23:23. Then, these
mixed powders were pressed into pellets of 13 mm diameter
under various compaction pressures (from 0 to 375 MPa).
The pellet was placed in a SHS reactor under an argon
atmosphere and ignited by a tungsten wire connected to a power
supply. The device used in this study was composed of a steel
cylindrical reactor whose volume is 1 L. The pressure inside the
reactor must not exceed 7 MPa and the temperatures of the walls
cannot be higher than 50 8C. Any overheating problem was
prevented by a water cooling device. The lower part of the device,
constituted by ceramic support made in pyrophyllite is remo-
vable. Fig. 1 presents the experimental system: the compacted
mixture made of silicon, graphite, magnesium and PTFE was
covered by the highly exothermic mixture made of Mg and PTFE
(chemical oven).
In order to study the evolution of temperatures during
combustion synthesis, thermocouples are positioned in the upper
and down side of the sample. The temperature control is realized
thanks to an acquisition card ‘‘Hi-speed USB Carrier National
Instruments’’, with Labview software. Then, the whole is main-
tained in a Papyex confinement in order to avoid heat losses.
The composition of the products is determined by X-ray
diffraction, with a Philips Expert device. In order to make accurate
measurements, patterns are collected using capillary and following
The reaction of formation of silicon carbide can be assisted by
using the heat liberated by the reaction between magnesium
and PTFE whose enthalpy of formation is
D
fH0 (300 K) =
À1518 kJ molÀ1. This reaction is very exothermic due to the great
affinity of fluorine with magnesium. The literature [17] for this
system shows the influence of three parameters: (a) The ratio Si:C, it
is equal to 1:2 or to 1:1. Indeed, Zhang et al. [17] have shown that the
stoichiometry 1:1 did not allow to obtain a complete reaction. They
attributed this result to the fact that stoichiometric amount of
carbon was not sufficient for the complete reaction due to the
evaporation of this component. Hence, excess carbon was added to
the reactant for the compensation of the carbon loss and they
worked with the ratio Si:C equal to 1:2. (b) The percentage of the
mixture Mg + PTFE: it varies between 20% and 25 wt.%. For graphite,
the percentage is fixed to 25%. Below this value the reaction is not
complete. (c) The nature of carbon used. These authors show that
with the carbon black the results are the best. In our application, the
nature of the carbon is imposed with graphite. So, in order to get
initiation and propagation of the reaction in the entire sample, the
idea was to use (Mg + PTFE) mixture according two ways: As a
chemical oven: in this case, the mixture surrounds the compact, the
high exothermic reaction between magnesium and PTFE allows to
add some amount of heat tothe compact toignite the reaction and to
propagate it all along the sample. ÀMixed to the mixture
silicon + graphite: It helps for the complete propagation of the
combustion wave. After processing, magnesium fluoride by-product
is removed with a chemical treatment with diluted fluorhydric acid
(10%HF). The efficiency of this method is evidenced in Table 1 where
thepercentageofMgF2 intheproduct afterwashingisreduced to0%.
conditions (Cu K radiation, 2u = 20–808, angle pitch 0.0178). A
1
a
semi-quantitative analysis in X-Pert High score works on basis of
the RIR (reference intensity ratio) values. It determines the
estimated mass fractions of the identified phases. The normal-
ization used in this method assumes that the sum of all identified
phases is 100%. The morphology of the fractured surface of the
pellet is analysed by observation with a scanning electron
microscope (SEM).
4. Combustion synthesis of the chemical oven: the reaction
2Mg(s) + C2F4(s) = 2C(s) + 2MgF2(s)
In a first step, combustion synthesis of the mixture (Mg, PTFE)
was studied. The reactants are introduced in the reactor and
ignited with a tungsten coil. When the reaction between Mg and
PTFE starts, an abrupt increase in temperature occurs giving rise to
the thermal profile shown in Fig. 2(a). The temperature reached in
this case is 1933 K. As this temperature is not high enough, we
investigate a way to lower heat losses during the reaction. Due to
its very specific properties, Papyex1 material can be used in
furnaces as a shield. At high temperatures, it reduces heat losses
due to its reflective capacity. In this case, the temperature of the
reaction was growing up to 2373 K (400 K) (Fig. 2(b)) higher than
3. Results and discussion
The product SiC which is involved in combustion synthesis
presents an adiabatic temperature of 1860 K and an enthalpy of
formation of À69 kJ molÀ1 [18]. This value is a little higher than the
empirically obtained critical adiabatic temperature of 1800 K. This
reaction cannot attain a self-sustaining combustion form without
the addition of energy from an exterior source.
previously without Papyex1
.
Fig. 3(a) shows X-ray pattern of the product. The major phase is
MgF2 and only trace amounts of graphite appear. Indeed, the
reaction between magnesium and PTFE is so exothermic that
graphite vaporizes and then condensates into the walls of the
reactor leaving only in the crucible the phase MgF2.
5. Combustion synthesis in the system Si:C = 1:2 (molar ratio)
S Mg:PTFE = 23:23 (wt.%)
The reactive system surrounded by Papyex1 was composed of
different pellets with the following ratios Si:C = 1:2 (molar ratio)
Table 1
Semi-quantitative analysis of Si:C À Mg:PTFE samples for different molar ratio.
Sample molar ratio
(Si:C/Mg:PTFE)
A.S.: as
% C
% Si
% MgF2
%
b
-SiC
% a-SiC
synthesized
(HF: HF
treatment)
1:2/23:23
1:1/23.23
1:1/23:23
A.S.
A.S.
HF
10
6
8
12
0
20
12
0
43
70
92
19
0
8
Fig. 1. Combustion synthesis device.

2136
R.M. Ayral et al. / Materials Research Bulletin 44 (2009) 2134–2138
Fig. 2. Thermal profiles of the chemical oven mixture obtained on a non-compacted powder: (a) Mg + PTFE without Papyex1 support and (b) with the Papyex1 support.
Fig. 3. X-ray pattern of the obtained powders after reaction (a) between Mg and PTFE, (b) with Si:C = 1:2 À Mg:PTFE = 23:23 mixture, (c) Si:C = 1:1 À Mg:PTFE = 23:23
mixture, and (d) Si:C = 1:1 À Mg:PTFE = 15:15 mixture.
À Mg:PTFE = 23:23 (wt.%). Compared to[17], the quantityof activate
additive is here reduced at 23% weight instead of 25% and in order to
enhance reactivity, a chemical oven made of Mg + PTFE is positioned
around the pellet. The heat liberated by the combustion synthesis of
chemical oven will help the initiation and the propagation of the
reaction between silicon and graphite into the pellet.
rather 150 s) which corresponds to the reaction between silicon and
graphite and to the formation of SiC.
The diffractograms (Fig. 3(b)) evidence the formation of the
phases of
b-SiC phase with trace of a-SiC (polytype 2H). Table 1
gives the semi-quantitative analysis obtained by X-pert high score:
Non-compacted mixture gives rise to formation of SiC 2H, to a low
The whole powder is mixed during 10 h and put in the reactor
without compacting. Thermal profiles recorded during the reaction
are presented in Fig. 4. The maximum temperature reach during the
reaction is 1823 K revealing important heat losses compared to the
reaction between magnesium and PTFE (maximum temperature
2373 K). This is due to the reaction between silicon and graphite
(Fig. 4) which consumes a large amount of heat produced by the
chemical oven reaction. The subsequent temperature rise followed
by a gradual decline represents a prolonged post-combustion stage.
After the passage of the combustion front, the temperature profile
(Tup) reveals a slow decreasing of temperature (whose duration is
percentage of b-SiC and free graphite in the product in large
quantity. One possible explanation of SiC-2H existence lies on
compaction pressure. Indeed, compression of a pellet prior to SHS
enlarges the surface contact and enhances the heat transfer rate
from the combustion front to the unburned region and so increases
the reactivity of the system. In this case, contacts between silicon
and graphite are reduced and do not allow to reach high
temperatures. Moreover, the product obtained after this reaction
is powdered and difficult to handle.
The utilization of compacted powders will give rise to a better
contact between particles of the mixture, would contribute to a

R.M. Ayral et al. / Materials Research Bulletin 44 (2009) 2134–2138
2137
Fig. 6. Thermal profiles obtained during the combustion synthesis of non-
compacted Si:C = 1:1 À Mg:PTFE = 23:23 samples.
Fig. 4. Thermal profiles obtained during the combustion synthesis of non
compacted Si:C = 1:2 À Mg:PTFE = 23:23 samples.
better thermal conductivity and so will favour the propagation of
the reaction. In their study Lee and Chung [19] showed that in Ti–C
system, ignition was controlled by the rate of surface reaction
between titanium and carbon which in turn was determined by the
contact surface area between them. In our experiments, the
compacting pressure varied from 75 to 375 MPa. The first
observation made was that there is no influence of compacting
pressure on maximum temperatures reached during the reaction.
The recorded thermal profiles in each case give the same value of
maximum temperature. However, locally, the temperatures must
quantity of graphite in the reactants powder. Even if this
stoichiometry did not previously give good results according to
[13], our new experimental procedure (chemical oven + activate
additiveinside themixtureSi + C)allowsustowaitfor betterresults.
Fig. 6 shows the thermal profiles obtained with a non compacted
mixture made of Si:C = 1:1 À Mg:PTFE = 23:23 reactants.
In this case, the maximum temperature is 2548 K which is
higher than the previous temperature (1823 K).
Percentages analysis obtained from X-ray diffraction patterns
(Fig. 3(c)) reveals an increase in the amount of
SiC phase was evidenced (Table 1).
b-SiC phase. No a-
be higher because in each case, only b-SiC phase was evidenced by
X-ray diffraction measurements. No trace amount of SiC-2H was
observed in the patterns. Then, the products synthesized at high
compacting pressures were observed by scanning electron
microscopy. Shown in Fig. 5 is the typical SEM photograph of
the synthesized product. SiC grains present an environment
essentially composed of graphite particles which had been spread
over by silicon. So the ratio Si:C = 1:2 gives rise to a large amount of
free graphite in the material. This configuration is not available for
nuclear conditioning which forbids any free graphite in the
ceramic.
To improve these results, next experiments with compacting
pressures equal to 75 and 375 MPa were done. Scanning electron
micrograph of the fractured surface of the product (Fig. 7) shows
groups of sintered grains structure which are identified as SiC. No
graphite surrounded SiC grains.
Table 2 gives the semi-quantitative analysis made from X-ray
pattern. With a compacting pressure of 375 MPa, the percentage of
SiC formed was increased up to 84% and the quantity of free
graphite was very low (3% value equal to the uncertainty of
measurements).
6. Combustion synthesis of mixture
Si:C = 1:1 S Mg:PTFE = 23:23
In the following, we will present results obtained with the
configuration Si:C = 1:1. This configuration reduces twice the
Fig. 7. SEM micrograph for a Si:C = 1:1 À Mg:PTFE = 23:23 sample compacted at
375 MPa.
Table 2
Semi-quantitative analysis of Si:C = 1:1 À Mg:PTFE = 23:23 (Æ5%) compacted at 75
and 375 MPa.
Pressure
% C
% Si
% MgF2
% b-SiC
75 MPa
5
3
9
0
12
13
74
84
Fig. 5. SEM micrograph for a Si:C = 1:2 À Mg:PTFE = 23:23 sample compacted at
375 MPa
375 MPa.

2138
R.M. Ayral et al. / Materials Research Bulletin 44 (2009) 2134–2138
Two further experiments were made with Si:C = 1:1
are more favourable to the process to avoid big quantity of free
graphite in the self-propagating high temperature synthesis
product.
À Mg:PTFE = 15:15 and Si:C = 1:1 À Mg:PTFE = 20:20. The max-
imum temperature with 15% mixture is only 1273 K and in this
case, X-ray pattern does not evidence peaks of SiC (Fig. 3(d)).
When adding 20% of additive, maximum temperature is 1473 K
and is yet too low to form silicon carbide phase. The macroscopic
observation of this sample showed that the reaction was initiated
in the top of the sample but did not propagate in the whole sample.
The heat liberated by the chemical oven is not sufficient to
propagate the SHS reaction. So these experiments allow us to
determine that SHS reaction between silicon and graphite can be
complete with a mixture containing 23 wt.% of PTFE + Mg. This
value is necessary to propagate the reaction in the whole sample.
With a compacting pressure of 375 MPa, the quantity of SiC is
84 wt.% and residual graphite is quite negligible.
Acknowledgments
The authors would like to thank Thomas Demars for his help
with experiments and GDR Matinex for its financial support.
References
[1] O.K. Karlina, V.L. Klimov, M.I. Ojovan, G.Y. Pavlova, S.A. Dmitriev, A.Y. Yurchenko, J.
Nucl. Mater. 345 (2005) 84–85.
[2] A.G. Merzhanov, in: Z.A. Munir, J.B. Holt (Eds.), Combustion and Plasma Synthesis
of High Temperature Materials, VCH, New York, 1990.
[3] J.J. Moore, H.J. Feng, Progr. Mater. Sci. 39 (1995) 243–273.
[4] P. Mossino, Ceram. Int. 30 (2004) 311–332.
[5] Z.A. Munir, Am. Ceram. Soc. Bull. 67 (2) (1988) 342–349.
[6] N. Pradeilles, M.C. Record, R.M. Marin-Ayral, J. Eur. Ceram. Soc. 26 (13) (2006)
2489–2495.
7. Conclusion
[7] R. Pampuch, L. Stobierski, J. Lis, Am. Ceram. Soc. 72 (1989) 1434–1435.
[8] A. Feng, Z.A. Munir, Metall. Mater. Trans. B 26B (1995) 587–593.
[9] Z.A. Munir, Solid State Ionics 101–103 (1997) 991–1001.
[10] J. Xu, B. Zhang, G. Jiang, W. Li, H. Zhuang, Ceram. Int. 32 (6) (2006) 633–636.
[11] J.-C. Han, G.-Q. Chen, S.-Y. Du, J.V. Wood, J. Eur. Ceram. Soc. 20 (2000) 927–932.
[12] F. Bosc-Rouessac, R.M. Marin-Ayral, A. Haidoux, N. Massoni, F. Bart, J. Alloys
Compd. 466 (1–2) (2008) 551–557.
The effects of activate additive and chemical oven on the
combustion characteristics of the silicium–graphite system were
studied. It was found that the production of silicon carbide which
has a low adiabatic temperature can be achieved by combustion
synthesis. The factors affecting the reaction are the utilization of
Papyex1 as a way to reduce heat losses during the initiation and
the propagation of the reaction, the presence of chemical oven
(magnesium and polytetrafluoroethylene) and mixing activate
additive (magnesium and polytetrafluoroethylene). Chemical oven
gives rise to better thermal conditions to initiate the reaction and it
is necessary to put up to 23 wt.% (magnesium and polytetrafluor-
oethylene) to allow propagation of the reaction. Moreover, we have
shown that one–one stoichiometries in these previous conditions
[13] J.L. O’Loughlin, C.H. Kiang, C.H. Wallace, T.K. Reynolds, L. Rao, R.B. Kaner, J. Phys.
Chem. B 105 (2001) 1921–1924.
[14] T. Michinori, O. Toshio, K. Takuo, K. Naminosuke, J. Pyrotechn. 10 (1999) 45–48.
[15] M. Szala, Int. J. Self Propagating High Temp. Synth. 17 (2) (1999) 106–111.
[16] D.T. Osborne, M.L. Michelle, Combust. Sci. Technol. 179 (8) (2007) 1467–1480.
[17] J. Zhang, J.C. Jeong, J.H. Lee, C.W. Won, D.J. Kim, C.O. Kim, Mater. Res. Bull. 37
(2002) 319–329.
[18] O. Knacke, O. Kubaschewski, K. Hesselmann, Thermochemical Properties of
Inorganic Substances, Springer-Verlag, 1991.
[19] W.-C. Lee, S.-L. Chung, J. Am. Ceram. Soc. 80 (1) (1997) 53–61.