Effect of different additives on the thermal properties and combustion characteristics of pyrotechnic mixtures containing the KClO4/Mg-Al alloy

Article abstract of DOI:10.1016/j.tca.2010.11.021

The thermal properties and combustion characteristics of KClO ₄/Mg-50%Al, KClO₄/Mg-50%Al with additives such as nitrocellulose (NC), urotropine and guanidine nitrate (GN) were studied experimentally using differential thermal analysi

Full text of DOI:10.1016/j.tca.2010.11.021

Thermochimica Acta 513 (2011) 119–123
Contents lists available at ScienceDirect
Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca
Effect of different additives on the thermal properties and combustion
characteristics of pyrotechnic mixtures containing the KClO4/Mg–Al alloy
Dehua Ouyang∗, Gongpei Pan, Hua Guan, Chenguang Zhu, Xin Chen
308, Chemical Engineering College, Nanjing University of Science and Technology, Nanjing 210094, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
The thermal properties and combustion characteristics of KClO4/Mg–50%Al, KClO4/Mg–50%Al with
additives such as nitrocellulose (NC), urotropine and guanidine nitrate (GN) were studied exper-
imentally using differential thermal analysis (DTA) and thermogravimetry (TG) and closed bomb
experiments. The results of DTA–TG analysis revealed that additives could affect thermal stability and
the decomposition temperature of KClO4/Mg–50%Al alloy. The order of the ignition temperatures of
mixtures (Ti) is KP/Mg–50%Al/urotropine (2#, 557.7 ◦C) > KP/Mg–50%Al/GN(3#, 553.6 ◦C) > KP/Mg–50%Al
(1#, 538.8 ◦C) > KP/Mg–50%Al/NC (4#, 536.5 ◦C). On the other hand, the results of the closed bomb exper-
iments confirm that the combustion pressures were all increased with the addition of additives. The
ignition delay times of these mixtures are in the order KClO4/Mg–50%Al (1#) > KClO4/Mg–50%Al/GN
(3#) > KClO4/Mg 50%Al/NC (4#) > KClO4/Mg–50%Al/urotropine (2#).
Received 6 January 2010
Received in revised form
12 September 2010
Accepted 19 November 2010
Available online 30 November 2010
Keywords:
DTA–TG
Closed bomb
Pyrotechnics
Additives
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
ignition. In this research we have started the initial work towards
understanding how additives affect the stability and gas production
Potassium perchlorate (KClO4, KP) is a powerful oxidizing agent
and decomposes at elevated temperatures to produce oxygen as
one of its major products [1]. Pyrotechnic formulations containing
magnesium–aluminum alloy (Mg–Al alloy) and potassium perchlo-
rate are used for flash pyrotechnic compositions, fireworks [2,3]
and fuel rich propellants [4], because they are easy to ignite. The
thermal properties of KClO4/Mg–Al alloy/Ca resinate mixtures have
been studied [5]. Urotropine is a white powder that has been used
as a stabilizing agent in ADN (ammonium dinitramide) propellants
and its thermal properties have been studied [6]. It is well known
that some of the most successful commercial flame-retardants
for cellulosic materials have been guanidine compounds, which
are effective and economical flame-retardants. The thermal degra-
dation of wood treated with guanidine compounds (guanidine
phosphate, guanidine carbonate and guanidine nitrate) have been
studied [7,8]. Nitrocellulose is a derivative of natural cellulose and
has an outstanding range of properties. With a nitrogen content
above 12.6% it is classed as an explosive [9,10]. The effect of nitrate
content on thermal decomposition of nitrocellulose has been stud-
ied [11].
of standard pyrotechnic mixtures of potassium perchlorate with
magnesium–aluminum alloy. For this initial study we have chosen
as additives substances that are familiar to those working in the
pyrotechnics area. These substances have characteristics that can
produce stability in pyrotechnic materials (urotropine and guani-
dine nitrate) and can add to the production of gaseous products
(nitrocellulose).
In the present work, urotropine, guanidine nitrate and nitro-
cellulose were used as additives. The reaction behavior of
KClO4/Mg–50%Al alloy with them has been studied experimentally
using differential thermal analysis (DTA) and thermogravimetry
(TG) and closed bomb analysis. This was done in order to examine
the influence of additives on the thermal properties and com-
bustion characteristics of KClO4/Mg–Al alloy. Results of this work
should provide valuable information for selecting mixtures to be
used in future studies of additives in these types of pyrotechnic
mixtures.
2. Experimental
Two major properties of pyrotechnic compounds are the quan-
tity of gas released and the stability of the compound towards
2.1. Materials
The materials used were potassium perchlorate (KClO4, pure,
mesh 300) purchased from Richem Company Ltd. (Beijing, China);
magnesium–aluminum alloy (Mg–50%Al alloy, Mg-50%, Al-50%,
mesh 300) purchased from Northeast Light Alloy Company Ltd.
∗
Corresponding author. Fax: +86 25 84315523.
E-mail address: oydh2010@yahoo.com.cn (D. H. Ouyang).
0040-6031/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2010.11.021

120
D. Ouyang et al. / Thermochimica Acta 513 (2011) 119–123
Table 1
Composition of samples.
No.
KP (wt%)
Mg–50%Al (wt%)
Urotropine (wt%)
GN (wt%)
NC (wt%)
1
2
3
4
60
40
0
9.1
0
0
0
9.1
0
0
0
0
9.1
54.5
54.5
54.5
36.4
36.4
36.4
0
(Harbin, Heilong-jiang, China); urotropine (pure, mesh 240) pur-
chased from Zhengzhou Jinfengda Chemical Products Company
Ltd. (Zhengzhou, Henan, China); guanidine nitrate (GN, pure, mesh
240) purchased from Jiangsu Qiangsheng Chemical Company Ltd.
(Suzhou, Jiangsu, China); and nitrocellulose (NC, pure, mesh 240)
purchased from Sichuan Nitrocell Corporation (Chengdu, Sichuan
China).
2.2. Procedure
2.2.1. Preparation of samples
The
pyrotechnic
mixtures
studies,
containing
KP/Mg–50%Al, KP/Mg–50%Al/NC, KP/Mg–50%Al/urotropine, and
KP/Mg–50%Al/GN are show in Table 1 with their weight ratios. The
KClO4 and Mg–50%Al alloy used were sieved through 300 mesh.
The other additives used were sieved through 240 mesh before
being mixed. Pyrotechnic mixtures of KClO4/Mg–50%Al alloy with
different additives (in Table 1) were prepared by carefully sieving
small quantities of the components through a slightly coarser
sieve than 240 mesh, after which the mixtures were further
mixed in a glass lined powder mixer until homogeneity was
achieved.
Fig. 1. Schematic diagram of closed bomb apparatus.
3. Results and discussion
3.1. Thermal behavior of compounds
3.1.1. Thermal behavior of the individual components
The thermal behavior of pure KClO4 has been previously
reported [14,15]. The oxidation characteristics of the Mg–50%Al
alloy were studied by simultaneous DTA–TG in air. The results are
shown in Fig. 2. It can be seen that the melting point of Mg–50%Al
alloy is 461.5 ◦C and there is evidence of a two-stage oxidation pro-
cess. It is expected that the Mg component of the alloy oxidizes
preferentially. There are reports in the literature to this effect. This
is due to the higher reactivity of the magnesium with oxygen com-
pared to aluminum [16,17]. The results observed for the Mg–50%Al
are similar in nature to those found by Pourmortazavi et al. [11]
in their study of Al powders. The results in the current study have
two additional features. First, near 461.5 ◦C there is an endother-
mic peak that corresponds to the melting of the Mg–50%Al alloy.
Second, there is a double peak that corresponds to a two-stage oxi-
2.2.2. Thermal analysis measurements
A thermo-balance (Stanton, Model TR-01, with a sensitivity
of 0.1 mg, with a differential thermal analysis (DTA) attachment)
was used to investigate the thermal properties (DTA–TG) of the
samples. In this study, the heating rate was 10 ◦C/min in air at
1 bar and the sample weight for each analysis was 5.0 mg. ␣-
Al2O3 was used as the reference material for calibration of the
equipment. The crucibles were not covered. The DTA–TG appara-
tus was calibrated with standard samples [12] (indium was used
for low temperature phase, quartz was used for middle temper-
ature phase, strontium carbonate was used for high temperature
phase). In order to investigate the reproducibility of the ther-
mal characteristics for KClO4/Mg–50% Al alloy with additives, the
results were recorded three times for each composition and then
averaged.
2.2.3. Closed bomb measurements
The combustion pressures of pyrotechnics in the closed bomb
(its volume is 50 mL) were recorded in air with a pressure trans-
ducer (American PCB Company, type is M102B, sensitivity is
143.96 mV/MPa). The closed bomb was developed according to the
verification regulation for measuring system of closed bomb for
propellants [13]. The closed bomb is constructed of refractory steel.
The closing bolt and the deaden plug are made of the same steel.
The threads were sealed with a polyurethane sealant. The nichrome
wires were insulated and sealed in bolt with 502 sealant. Each sam-
ple (0.500 g) was weighted into a paper tube (H = 20 mm, R = 5 mm)
and consolidated using a ram under hand pressure and placed into
the closed bomb. The sample was ignited by electrically heating a
coil of nichrome wire buried in the composition at the center of the
tube. The schematic diagram of closed bomb apparatus is shown
in Fig. 1. The measurements were recorded three times for each
composition and then averaged.
Fig. 2. DTA and TG curves for Mg–50%Al alloy (50–50%).

D. Ouyang et al. / Thermochimica Acta 513 (2011) 119–123
121
Fig. 3. DTA and TG curves for mixtures.
dation of the alloy. These peaks are in the same temperature range
as that found by Pourmortazavi et al. [11] for the oxidation of pure
Al powders. The oxidation of the Mg begins below 540 ◦C and peaks
at 542.6 ◦C. This is to be expected, as Mg is more reactive in air than
Al. The Al oxidation process peaks at approximately 599 ◦C. While
this temperature is higher than that obtained by Pourmortazavi et
al. [11], it is consistent with their study of Al powders. All of this
reasonably speculative as the work cover here deals with an alloy of
two metals. The process of oxidizing an alloy which is in the liquid
phase is a much more complicated process than just oxidizing pure
component powders. In addition, the alloy is in the liquid state at
the point of oxidation with the air. This alloy has a complex phase
diagram, so as the Mg is oxidized the liquid alloy will become richer
in Al. The liquid phase will also have the oxide of Mg dispersed in it.
The explanation given here for a two-stage process seems at least
plausible but is recognized as being highly speculative.
assumed uniform for Al and Mg, for simplicity) would result in a
∼17.4% weight increase. Decomposition without oxidation of the
alloy would result in a ∼27.6% weight loss. But the observed 61.2%,
it could indicate that some material was carried out of the crucible.
Fig.
3b
shows
the
DTA–TG
results
for
the
KP/Mg–50%Al/urotropine mixture. There is an endothermic
peak at 306.7 ◦C corresponding to a rhombic–cubic transition of
the KP. Then, an endothermic peak occurs at 467.1 ◦C due to the
melting of the KP. An exothermic peak with a decrease in the mass
(about 48.7%) occurs at 536.5 ◦C, which corresponds to the reaction
of Mg–50%Al and urotropine with the KP. A rough calculation
shows that complete decomposition of KP, complete oxidation of
the alloy and urotropine would result in a ∼15.8% weight increase.
Decomposition without oxidation of the alloy would result in a
∼34.2% weight loss. But the observed 48.7%, it also could indicate
that some material was carried out of the crucible.
In the DTA and TG curves for KP/Mg–50%Al/GN mixture (Fig. 3c),
there are endothermic peaks at 306.4 ◦C, which corresponds to the
rhombic–cubic transition of the KP, and at 480.0 ◦C due to the melt-
ing of the mixture. At the same time, we can see that an intense
decomposition occurred at about 553.6 ◦C from the reaction of
Mg–50%Al and GN with the KP, which corresponds with a decrease
in the mass (about 44%). it may be some material was carried out
of the crucible or vaporized from the crucible.
Fig. 3d shows DTA and TG curves for the sample of
KP/Mg–50%Al/NC. Two endothermic stages were observed at
306.7 ◦C and 476.9 ◦C, corresponding to a rhombic–cubic transition
and the melting point, respectively. Next there is an exothermic
peak at 536.5 ◦C with a reduction (about 74.7%) in the mass of sam-
ple that corresponds to the reaction of the Mg–50%Al alloy and NC
with the KP. It also could indicate that some material was carried
out of the crucible or vaporized from the crucible.
3.1.2. Thermal behavior of the mixtures
Fig. 3 shows TG and DTA curves for KP/Mg–50%Al alloy mixture
in the pure form and with various additives.
DTA and TG curves for the pure KP/Mg–50%Al alloy mixture are
shown in Fig. 3a. An endothermic peak in the DTA curve at 306.3 ◦C
corresponds to a rhombic-cubic transition of KP [14]. The DTA curve
also shows an endothermic peak at 463.3 ◦C, which corresponds
with the temperature of fusion of KP. A sharp exotherm occurs near
538.8 ◦C accompanied by a reduction (about 61.2%) in the mass of
sample that corresponds to the oxidation of the Mg–50%Al alloy by
the KP according to:
5KClO4 + 8Mg–Al → 8MgO + 4Al2O3 + 5KCl
A rough calculation shows that complete decomposition of
KP and complete oxidation of the alloy (1#: KP: Mg–50%Al = 6:4,

122
D. Ouyang et al. / Thermochimica Acta 513 (2011) 119–123
Fig. 4. Curves for pressure (0.500 g).
On the other hand, GN and NC cannot really stable up to 500 ◦C,
the Pmax are 5.23 MPa and 4.92 MPa, respectively. The maximum
pressures are lower for urotropine or GN as the quantity of gas
produced is less than that for the same weight of NC. In the same
manner, we can determine the degree of oxidation for the mixtures
(2#) and (3#).
because the thermal decomposition of NC and GN are ∼200 ◦C (it is
affected by its particle size and so on). This can be seen from Fig. 3(c)
and (d), that their DTA curve are raising and the TG curve are spiral
downward before the 500 ◦C, but the other DTA curve (e.g. (a) and
(b)) are horizontal. Because the decomposition of NC and GN that
cause the ꢀT raised.
From Fig. 4 we also can observe that the ignition delay times (Tig
,
the time for the pressure to increase to 3% of the maximum pres-
sure). As an example, the Tig, Pmax and 0.03 Pmax of 4# are shown in
Fig. 4. The igniter was triggered at t = 0 s. The magnitudes of the Tig
of these mixtures are in the order 1# > 3# > 4# > 2#. This is because
the additives need oxygen when they are reacting, but the oxygen
which is produced by the KP decomposition is limited such that the
Mg–50%Al will not react completely. On the other hand, the addi-
tives will absorb heat and this may also decrease the reaction rate.
Also, this can be proven from Fig. 5 since the pressure variation
speed (dp/dt) decreased, especially for KP/Mg–50%Al/urotropine
mixture (2#).
The results of these experiments are summarized in the table
below.
3.2. Pressure characteristics of mixtures
Pressure is an important parameter in the combustion of
pyrotechnics. It can provide information about the degree and
intensity of reaction. In this work, the combustion pressure of mix-
tures of KP/Mg–50%Al with different additives in a closed bomb
were measured using a pressure transducer as show in Fig. 4.
As shown in Fig. 4 we can see that the combustion pressures of
all KP/Mg–50%Al alloy mixtures increased after adding the differ-
ent additives. For the mixture of KP/Mg–50%Al (1#) the maximum
combustion pressure (Pmax) is 4.58 MPa. Also the result of the
closed bomb experiment for KP/Mg–50%Al/NC (4#) mixture shows
this mixture has a maximum pressure of (6.10 MPa). This may be
related to the decomposition mechanism of the components in this
mixture. For example, a rough calculation shows that complete oxi-
dation of the mixture KP/Mg–50%Al/NC (4#) would result in a Pmax
of 7.42 MPa (perfect gas, pV = nRT) gas. The recorded 6.10MPa can
be matched with an assumed ∼72.1% oxidation of the mixture (4#).
This incomplete reaction may be due to the lack of oxygen resulting
in the NC not being completely oxidized. For the two other mix-
tures, KP/Mg–50%Al/urotropine (2#) and KP/Mg–50%Al/GN (3#),
3.3. Ignition temperatures of mixtures
In this study, the thermal behavior of several pyrotechnic mix-
tures containing Mg–50%Al as fuel, KP as oxidizer and urotropine,
GN, NC as additives was studied under identical conditions. For
the mixture of KP/Mg–50%Al, as seen in Fig. 3a, the ignition reac-
tion occurred at 538.8 ◦C. The ignition temperature was decreased
to 536.5 ◦C by the addition of NC to the KP/Mg–50%Al mixture,
thus demonstrating an increased sensitivity of this mixture to heat.
However, by replacing the NC with GN in this mixture, the sensitiv-
ity was decreased and the ignition occurred at 553.6 ◦C. Comparison
of the two mixtures (see Table 2) shows that KP/Mg–50%Al/ GN is
more stable than KP/Mg–50%Al/NC and has an increased ignition
Table 2
Summary of results.
No.
Ti (◦C)
Average
538.8 ◦
Pmax (MPa)
Average
Tig (s)
Average
0.0453 s
1# – Test 1
1# – Test 2
1# – Test 3
2# – Test 1
2# – Test 2
2# – Test 3
3# – Test 1
3# – Test 2
3# – Test 3
4# – Test 1
4# – Test 2
4# – Test 3
540.1
538.6
537.7
559.5
557.8
555.8
555.0
553.7
552.1
537.0
536.6
535.9
C
4.60
4.59
4.55
5.25
5.22
5.22
4.98
4.90
4.88
6.14
6.10
6.06
4.58 MPa
0.0455
0.0452
0.0452
0.0445
0.0446
0.0450
0.0448
0.0448
0.0451
0.0442
0.0441
0.0446
557.7 ◦C
5.23 MPa
4.92 MPa
6.10 MPa
0.0447 s
0.0449 s
0.0443 s
553.6 ◦
C
536.5 ◦C
1# 95% confidence intervals: 534.6–543.0 (Ti); 4.49–4.67 (Pmax); 0.0448–0.0458 (Tig).
2# 95% confidence intervals: 551.2–564.2 (Ti); 5.18–5.28 (Pmax); 0.0438–0.0456 (Tig).
3# 95% confidence intervals: 548.5–558.7(Ti); 4.74–5.10 (Pmax); 0.0444–0.0454 (Tig).
4# 95% confidence intervals: 534.6–538.4 (Ti); 6.00–6.24 (Pmax); 0.0436–0.0450 (Tig).

D. Ouyang et al. / Thermochimica Acta 513 (2011) 119–123
123
(4#) < KP/Mg–50%Al/ urotropine (2#). The order of the ignition
temperatures of the mixtures (Ti) is KP/Mg–50%Al/ urotropine (2#,
557.7 ◦C) > KP/Mg–50%Al/GN (3#, 553.6 ◦C) > KP/Mg–50%Al (1#,
538.8 ◦C) > KP/Mg–50%Al/NC (4#, 536.5 ◦C).
This study of pyrotechnic compositions is important not only
for understanding the characteristics of their thermal decom-
position, but also for assessing the degree to which additives
can affect their exothermic decomposition and ignition tempera-
tures. This information can lead to pyrotechnic compositions with
improved characteristics to mitigate potential hazards during han-
dling, usage, storage, and to knowledge on how to improve the
sensitivity of igniter compositions and reduce factors for accidents.
Acknowledgements
The authors are grateful to Dr. N.D. Hamer, L. Zheng, F. Wei and
H.Y. Liu for their help. Project supported by the National Natural
Science Foundation of China (Grant no. 51076066) and the 2009
Innovation Foundation of the Jiangsu Higher Education Institutions
of China (No. CX09B 105Z).
Fig. 5. Curves for pressure variation speed.
temperature of 14.8 ◦C above that of the pure KP/Mg–50%Al without
additives.
By replacing of GN with urotropine, the thermal stability of the
pyrotechnic mixture is increased further. The ignition temperature
of the mixture of KP/Mg–50%Al/ urotropine is increased to 557.7 ◦C.
For this mixture, the ignition temperature is ∼18.9 ◦C higher than
the pure KP/Mg–50%Al mixture.
The relationship of ignition temperatures (Ti) of
these mixtures is observed to be in the following order
KP/Mg–50%Al/urotropine (2#, Ti = 557.7 ◦C) > KP/Mg–50%Al/GN
(3#, Ti = 553.6 ◦C) > KP/Mg–50%Al (1#, Ti = 538.8 ◦C) > KP/Mg–
50%Al/NC (4#, Ti = 536.5 ◦C).
References
[1] A.P. Hardt, Pyrotechnics, Pyrotechnica Publications, USA, 2001.
[2] Y.Q. Huang, Chinese Patent CN 1490590A (2004).
[3] F.SH. Qui, United States Patent US 6718882B1 (2004).
[4] T. Kuwahara, J. Prot. Tech. 1 (1992) 2.
[5] I. Lee, S.A. Finnegan, J. Therm. Anal. 50 (1997) 707–717.
[6] T.K. Highsmith, C.S. Mcleod, R.B. Wardle, R. Hendrickson, WO Patent WO001408
(1999).
[7] M. Gao, C.Y. Sun, K. Zhu, J. Therm. Anal. Calorim. 75 (2004) 221–232.
[8] M. Gao, B.C. Ling, S.S. Yang, M. Zhao, J. Anal. Pyrol. 73 (2005) 151–156.
[9] E. Schering, J. Am. Chem. Soc. 1 (1879) 175.
[10] E.L.M. Krabbendam-LaHaye, W.P.C. de Klerk, R.E. Krämer, J. Therm. Anal.
Calorim. 80 (2005) 495–501.
[11] S.M. Pourmortazavi, S.G. Hosseini, M. Rahimi-Nasrabadi, J. Hazard. Mater. 162
(2009) 1141–1144.
4. Conclusions
[12] GJB 5382.11-05, Test methods of physical parameter for pyrotechnic. Part 11.
Determination of Autoignition temperature-DTA and DSC method, 2005, pp.
56–59.
[13] GJB/J 3345-98, Verification regulation for measuring system of closed bomb of
propellants, 1999, pp. 1–6.
[14] S.M. Pourmortazavi, M. Fathollahi, S.S. Hajimirsadeghi, S.G. Hosseini, Ther-
mochim. Acta 443 (2006) 129–131.
[15] J.S. Lee, C.K. Hsu, K.S. Jaw, Thermochim. Acta 367–368 (2001) 381–385.
[16] J.H. McLain, Pyrotechnics From the Viewpoint Of Solid State Chemistry, Franklin
Institute, 1980.
Thermal behavior and combustion characteristics of pyrotech-
nic mixtures of KClO4/Mg–50%Al alloy in
a pure form and
with the different additives were studied by DTA–TG and
closed bomb methods. The results show (see Table 2) that
with the addition of additives the combustion pressure is
increased in the order KP/Mg–50%Al/NC (4#) > /Mg–50%Al/
urotropine (2#) > KP/Mg–50%Al/GN (3#) > KP/Mg–50%Al (1#),
but the order of ignition delay time (Tig) for these mixtures is
KP/Mg–50%Al (1#) < KP/Mg–50%Al/GN (3#) < KP/Mg–50%Al/NC
[17] F.S. Stone, Lattice defects in ionic crystals, in: W.E. Graner (Ed.), Chemistry of
the Solid State, London, 1955.