The thermal behavior of potassium dinitramide. Part 1. Thermal stability

Article abstract of DOI:10.1016/s0040-6031(99)00173-2

In this paper, thermal analysis, X-ray photoelectric spectroscopy (XPS), etc. were used to investigate the thermal stability of potassium dinitramide (KDN). It was found that the thermal decomposition of KDN crystal conformed to the topochemistry. The dec

Full text of DOI:10.1016/s0040-6031(99)00173-2

Thermochimica Acta 335 (1999) 105±112
The thermal behavior of potassium dinitramide. Part 1. Thermal stability
Ming Lei*, Zhi-Zhong Zhang, Yang-Hui Kong, Zi-Ru Liu, Chun-Hua Zhu, Ying-Hui Shao, Pei Zhang
Xi'an Modern Chemistry Research Institute, PO Box 18, Xi'an 710065, China
Received 1 February 1999; accepted 22 June 1999
Abstract
In this paper, thermal analysis, X-ray photoelectric spectroscopy (XPS), etc. were used to investigate the thermal stability of
potassium dinitramide (KDN). It was found that the thermal decomposition of KDN crystal conformed to the topochemistry.
The decomposition product, KNO3, formed eutectic system with KDN and decreased the liquefaction temperature of KDN.
# 1999 Elsevier Science B.V. All rights reserved.
Keywords: Potassium dinitramide; Thermal stability
1. Introduction
thermal decomposition in solid state and the in¯u-
ences on the thermal stability of KDN.
Potassium dinitramide (KDN) is a new type of high
energetic oxidizer of great interest for halogen-free,
high-performance solid propellants. It can be used to
replace AP or AN completely or partly. In recent year,
KDN attracted the attention for more stable than
ammonium dinitramide (ADN) and its phase-stabili-
zation to AN [1]. A great deal of the investigation into
synthesis and performance of KDN have been carried
out by Russian researchers. In addition, thermal
decomposition of KDN in solid and liquid state have
been studied preliminarily [2,3]. They found an abnor-
mal phenomenon that liquid KDN is more stable than
solid state [3]. Furthermore, the US researchers inves-
2. Experimental
2.1. Sample
KDN [KN(NO2)2], pale yellow crystal, was puri®ed
in water±methane alcohol. The sample was synthe-
sized by Xi'an Modern Chemistry Institute.
2.2. Apparatus
The thermal analysis were carried out on a differ-
ential scanning calorimeter, Perkin-Elmer model
DSC-2C and TA model DSC 2910, and a thermogra-
vimetric analyzer, TA model TGA 2950, with sample
mass of about 2±4 mg, and Al cell. The measurements
tigated the geometry of NꢀNO2† in solid state and in
2
solution [4]. However, further investigation on thermal
behavior of KDN should be done.
In this paper, DSC, TGA and XPS technology were
applied to investigate the KDN topochemistry of
were preformed in a dynamic atmosphere of nitrogen
1
with a ¯ow rate of 40 ml min
.
*Corresponding author.
0040-6031/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 0 - 6 0 3 1 ( 9 9 ) 0 0 1 7 3 - 2

106
M. Lei et al. / Thermochimica Acta 335 (1999) 105±112
In addition, an X-ray photoelectric spectroscopy
moreover less decomposition occurred in solid state,
the endothermic peaks at 108.38C and 110.38C would
be insigni®cant, only a fusion peak of KDN could be
seen (see Fig. 1(b)). Thus the fusion peak temperature
(XPS) apparatus, model PE PHI-5400, with Mg target
and in 3 Â 10 3 Pa vacuum, was used to study the
chemical state of solid KDN sample on surface.
of the sample is the real melting point of KDN. It is
1
1328C when the heating rate was 108C min
.
3. Results and discussion
IR and TG analyses (in next part) indicate that
KNO3 and KNO2 are the main products of decom-
position of KDN in solid state. It is seen that a phase
transition peak at 132.28C and a fusion peak at 3358C
appear on the DSC for KNO3 (Fig. 1(c)), but only one
fusion peak at 428.18C for KNO2 (Fig. 1(d)). By
comparison, the DSC of decomposition remainder
of KDN (Fig. 1(e)) is similar to curve C. Merely,
the ®rst endothermic peak is at 106.78C, but the
second endothermic peak is still at 342.58C. From
Fig. 1(f), we can see an eutectic peak existing in
KNO3/KNO2 binary system. The eutectic temperature
is 312.58C. So we believe that the last sets of endother-
mic peaks on the DSC curve of KDN are those of
fusion and liquefaction of KNO3/KNO2 binary sys-
tem. For the eutectic peak, formed by a few residues, is
very weak, it is often overlapped with liquefaction
peak. So more often, only liquefaction peak of KNO2
can be seen. It changes with the proportion of KNO3 to
KNO2.
3.1. Thermal stability of solid KDN
3.1.1. Thermal character on DSC and TGA
The curve in Fig. 1(a) shows that the DSC of KDN
is complicated. There is an exothermic peak and three
endothermic peaks before 1358C; two main exother-
mic peaks between 2288C and 2388C and a set of
endothermic peaks at last. The ®rst three endothermic
peaks are the phase transition peak of KNO3 at
108.38C, eutectic peak of KDN/KNO3 system at
110.38C and liquefaction peak of KDN at 128.08C,
respectively. If the sample contained less KNO3, and
The endothermic peak at 1088C is the phase transi-
tion peak of KNO3 which is formed by early decom-
position of KDN (including that formed at room
temperature during storage time). It could be proved
by the endothermic peak at 106.78C on DSC curve
(Fig. 1(d)) of decomposition residue. Fig. 1(c) curve
shows the phase transition temperature of KNO3
formed by KDN is decreased, or changed by the effect
of other decomposition product as KNO2. The ®rst
two sharp endothermic peaks in Fig. 1(f) could prove
it. In addition to a normal phase transition peak there is
an endothermic peak at 118.18C on DSC curve of
KNO3/KNO2.
The two exothermic peaks between 2288C and
2388C are those of decomposition of KDN in liquid
state.
While, the TG-DTG curves of KDN are relatively
simple (Fig. 2). Besides a very slight mass loss (about
1.5%) at 1208C, the main process of mass loss occurs
at 214.88C, added up to about 35.6%, just between
30.3%, the mass loss of forming KNO3 completely
and 41.4%, that of KNO2 completely.
Fig. 1. The DSC curves of KDN, KDN/KNO3, KNO3 and KNO2:
(a) KDN; (b) pure KDN; (c) KNO3; (d) KNO2; (e) decomposition
residue of KDN; (f) KNO3/KNO2 ˆ 1 : 1.

M. Lei et al. / Thermochimica Acta 335 (1999) 105±112
107
Fig. 2. The TG-DTG curves of KDN (model TA TGA 2950).
3.1.2. The influence of main factors on the thermal
stability of KDN
temperature of liquefaction peak, which is close to the
melting point of KDN for A1, is decreased apparently
to 128.78C and a quite strong exothermic peak appears
before liquefaction of A2, indicating an apparent
decomposition in solid state.
Comparing B2, B3 and B4 with B1 shows: the ®rst
endothermic peak is apparently strengthened as the
storage time increases, indicating the increase in the
decomposition depth and in the product. But the peak
in the second endothermic peak (i.e. the liquefaction
peak of KDN) gradually reduces and the liquefaction
temperature decreases from 124.58C to 118.38C, even
3.1.2.1. Storage time and pretreatment of the sample.
Fig. 3 and Table 1 show that same samples have
various DSC curves for various storage time and
pretreatment. Samples A1, A2 and B1±B4 were all
purified in water±methane alcohol and stored dry.
Comparison of the DSC curve of A2 with that of
A1 shows: as the storage time increases, the
endothermic peak at 109.18C (eutectic peak of
KDN/KNO3) is drastically strengthened for A2, the
Table 1
The DSC data of KDN during various storage time and pretreatment
Sample
Date of
Dry
Date of
testing
Eutectic
Liquefaction peak
synthesis
condition
temperature (8C)
temperature (8C)
a
A1
A2
B1
B2
B3
B4
C1
C2
12.12
6.17
8.28
8.28
8.28
8.28
9.4
12.24
12.25
9.1am.
9.2am.
9.4
109.1
110.8
109.7
109.7
110.1
110.7
110.4
110.9
131.7
128.7
124.5
123.8
121.0
118.3
126.1
121.1
±
±
±
±
±
9.18
9.10
±
b
9.10
9.10
a In dryer at room temperature.
b Dried at 858C in vacuum for 2.5 h.

108
M. Lei et al. / Thermochimica Acta 335 (1999) 105±112
KNO3 in dry condition and room temperature. There-
fore, an eutectic peak of KDN/KNO3 appears in the
DSC curve (including phase transition of KNO3), the
peak of which increases as the storage time and the
decomposition depth of KDN increases. But the peak
of liquefaction peak of KDN reduces with the increase
in storage time, and relative surplus of KDN reduces,
and its liquefaction temperature gradually approaches
eutectic.
For various dry conditions, the fusion and the
liquefaction processes of solid state of sample C1 is
different from sample C2, obviously, the decomposi-
tion process giving KNO3 is accelerated by vacuum
dry condition at 858C, with the eutectic peak strength-
ened, liquefaction peak weakened, and the peak tem-
perature also decreased. So vacuum dry condition
would do harm to the stability of KDN.
As the data of sample A1 and A2 show, although the
thermal stability and the change in solid±liquid state
are strongly affected by storage time, the decomposi-
tion of KDN in liquid state is affected slightly.
3.1.2.2. Particle size. It is seen from Fig. 4 and
Table 2 that for the ground sample, the heat of the
exothermic peak before fusion (or liquefaction)
decreases significantly, as is also the mass loss. So
the decomposition process of KDN is strongly
affected by the particle state. The samples with
large particle size, with many edges, or imperfect
crystal, are liable to decompose and their thermal
stability is poor, the reverse is stable.
Although the decomposition peak of the ground
sample D2 reduces, i.e. the decomposition product
KNO3 reduces, the eutectic peak of KDN/KNO3
strengthened and the liquefaction temperature of
KDN decreased, the reason being the presence of
KNO3 partially formed in storage time. The DSC
curve of the sample D1, which was tested immediately
after grind, was different from that of the sample D2,
which was stored for some time after grind. From this,
it can be concluded that the decomposition of KDN
took place at room temperature, especially in dry
environment.
Fig. 3. The DSC curves of KDN sample during various storage
time and pretreatment.
The sample D3 is solid, formed by cooling the
lique®ed sample. Although this sample underwent
the decomposition in solid state before liquefaction,
its behavior was similar to that of ground sample for it
changed to ®ne crystal by drastically cooling. When
when there is a difference, just 4±5 h interval, of
storage time between sample B1 and B2. Thus it is
believed that the samples decomposed in part to form

M. Lei et al. / Thermochimica Acta 335 (1999) 105±112
109
Fig. 4. The DSC and TGA curves of KDN in various crystalline state.
heated again, the exothermic peak of decomposition in
solid state disappeared, and the eutectic peak reduced.
The results show that its thermal stability improved.
So the ®ne crystal is more stable.
moisture tends towards stability with less
decomposition in solid state. Once the sample was
dried, it would be decomposed significantly. Sample
E2 was prepared by storing sample E1 in dryer for two
days, while sample E3 was prepared by storing sample
E2 in saturated solution of 35% relative humidity for
five days, with 3.2% moisture content. Therefore,
3.1.2.3. Moisture content. The curves in Fig. 5 and the
data of Table 3 show that the KDN containing certain

110
M. Lei et al. / Thermochimica Acta 335 (1999) 105±112
Table 2
The data of DSC and TG of KDN in various particles state
Sample
Particle state
Date of Date of
preparation grind
Date of
testing
ÁHa
(J/g)
Mass
lossb (%)
Eutectic
Liquefaction
temperature (8C) temperature (8C)
D
Large particle
Ground
2.25
2.25
2.25
2.25
±
6.3
6.3
22.2
13.8
17.6
0.0
2.5
±
109.9
109.9
110.0
110.3
127.5
126.6
119.5
123.9
D1
D2
D3c
6.3
5.27
±
Ground
6.1
1.2
±
Liquefied
5.26
a,b They are the exothermic enthalpy and mass loss of decomposition in solid state, respectively.
c Sample D3 was prepared by drastically cooling after heating the sample D to 1408C.
Table 3
The influence of moisture content on KDN
1
)
Sample
Moisture content (wt%)
ÁHa (J g
Eutectic temperature (8C)
Liquefaction temperature (8C)
E1
E2
E3
Certain
4.9
75.6
10.1
109.7
109.9
110.8
130.2
117.6
126.4
0
3.2
a It is the exothermic enthalpy of decomposition in solid state.
when the sample absorbed certain moisture again, the
decomposition in solid state would decrease, so did the
eutectic peak. The sample tended to be stable again.
decomposition on DSC and TGA depended on particle
size, moisture content, and storage time of the sample,
it is considered that the decomposition of KDN crystal
conforms to the law of topochemistry.
3.2. Topochemistry of decomposition of KDN in solid
state
There are many defects and cracks in the KDN
crystal of large size and with striking edges and
corners on surface or with the imperfection, which
forms more reaction center and decomposes quickly.
Under the decomposition process, the solid product
KNO3 would be adhered to these reaction sites to
extend the defects and cracks and to promote the
increase in stress on the crystal. When the stress
increased to a certain degree, the crystal cracked
and formed little crystals with new surface. Then
the reaction would be continued on the increased
speci®c surface. When the product KNO3 has covered
the surface of little crystal completely, the active
center would be restrained and the decomposition
would be stopped completely or partly.
According to above test results, i.e. the stability of
KDN crystal before liquefaction (or fusion) and the
From the ®ne crystal or ground samples of KDN,
whose surface would be completely covered by KNO3
formed owing to its partial decomposition during long
storage time at room temperature, the active center are
eliminated and the crystals do not crack easily due to
its little size. Thereby, the process of ``topochemistry''
not occurs easily for these samples and they have good
thermal stability.
Fig. 5. The influence of moisture content on DSC of KDN.

M. Lei et al. / Thermochimica Acta 335 (1999) 105±112
111
Similarly, the presence of a little of moisture is
favorable to the elimination of the potential reaction
center. Namely, certain moisture can stabilize the
KDN crystal.
The decomposition in solid state and the process of
liquefaction for KDN have been also observed by a
microscope warming table. When temperature raised
up to: 908C, the crystal of sample D began to crack
(the partial decomposition in solid state began); and to
978C, the surface of the crystal began to darken (the
product KNO3 was formed and adhered to the sur-
face). At 1088C, the crystal liquidated partly, and the
liquid ¯owed bubbling up (the decomposition in liquid
state continued); to 1308C, the lique®ed sample
became transparent (the liquefaction stopped, includ-
ing eutectic formed by KNO3) without increase in the
bubble; and also to 1508C, was the same (the liquid
KDN is stable relatively). After 1608C, the bubbles
increased and enlarged with opaque substance formed
(the decomposition in solid state began, KNO3 and
KNO2 are formed). The preceding process of topo-
chemistry can be proved by the observed phenomena,
especially the decomposition in solid state.
In order to prove the above-mentioned facts that the
decomposition of KDN in solid state was closely
related to the coverage of KNO3 on its surface, the
chemical state of K and the ratio of element K to N
on surface of KDN were determined by means of
X-ray photoelectron spectroscopy (XPS).
‡
Fig. 6. Binding energy of K in KDN.
‡
It is shown in Fig. 6 and Table 4, that for all
samples except sample D1, the 2P electron binding
sample is ground and new crystal surface has been
formed, where the surface cannot be covered imme-
diately by the decomposition product KNO3, and the
inner layer structure of KDN crystal can be emerged,
‡
energy of K is in good agreement with that of KNO3,
‡
indicating that the major K on surface of these
samples is in chemical state of KNO3. When the
‡
the electron binding energy of K /2P of D1 has been
Table 4
Electron binding energy of K and K/N on surface for various crystal state
‡
‡
Sample
Date of preparation Date of treatment
Treatment
Date of testing
Electron binding energy of K
K/N
2P3/2
2P1/2
KNO3
D
±
±
±
6.3
6.3
6.3
6.3
6.3
293.0
292.9
292.6
292.9
292.9
295.6
295.7
295.4
295.7
295.7
2.25
2.25
2.25
2.25
Large cry
D ground
D ground
0.47
0.40
0.55
0.41
D1
D2
D3a
6.3
5.27
6.2
b
a Sample D3 was prepared by drastic cooling after heating the sample D to 1358C.
b D after liquefaction.

112
M. Lei et al. / Thermochimica Acta 335 (1999) 105±112
‡
decreased and is close to K chemical state of KDN
structure. When the ground sample has been stored for
several days, where the new crystal surface has been
covered by decomposition product KNO3 again, the
4. Conclusion
1. There is complicated thermal behavior from solid
state to fusion in KDN. For KDN, the thermal
stability, decomposition in solid state, formation
of eutectic, and fusion or liquefaction process are
dependent on particle size, storage environment
and time and whether containing certain moisture,
etc.
2. The decomposition of KDN in solid state conforms
to topochemistry process. Cracking of large crystal
would deepen the decomposition. The product
KNO3 or certain moisture covered on crystal sur-
face would help to eliminate active center. They
play the role as a stabilizer.
‡
ionic state of K on sample D2 crystal surface is as
that of KNO3 again. K/N value of crystal of all the four
samples are greater than 0.33, the theoretical value of
KDN. Thus it is seen that KNO3 has been already
accumulated on the crystal surface. Whereas, the
speci®c surface of sample D2 is larger due to its being
ground, so the K/N value is as high as 0.55. Since all
the newly formed surface of D1 crystal has been not
immediately covered by KNO3, its K/N value is low-
est, but is 0.40 still. Once the sample has been
recrystallised by means of liquefaction, the KNO3,
which was originally accumulated on crystal surface,
would be redistributed to new surface of ®ne crystal
with larger speci®c surface, leading to the decrease of
K/N value to 0.41.
The liquefaction (or fusion) temperature of KDN is
affected by quantity of KNO3 product.
It is shown from the results that the KNO3 on KDN
crystal surface formed by decomposition plays an
important part in thermal behaviors of KDN, such
as topochemistry of KDN in solid state, thermal
stability, forming eutectic and liquefaction process,
etc. Also, certain moisture content could stabilize
KDN, for the polar molecular H2O is of bene®t to
eliminate the reaction center on crystal surface.
References
[1] U.S. Patent 5 292 387.
[2] F.I. Dubovitskii et al., Dokl. Akad. Nauk 347 (1996) 763±765.
[3] F.I. Dubovitskii et al., Dokl. Akad. Nauk 348 (1996) 205 and
206.
[4] K.O. Christe, W.W. Wilson et al., Inorg. Chem. 35 (1996)
5068±5071.