Thermal analysis of ammonium dinitramide (ADN)

Article abstract of DOI:10.1016/S0040-6031(00)00387-7

Kinetic constants of decomposition were determined for ADN and prilled ADN by DSC and TG. Pressure DSC at 550 psi air or helium had similar kinetic constants for prilled ADN, namely 29 kcal/mol activation energy and a log frequency factor of ~14.0 min^-1. These values were similar to those found in sealed sample pans in which the decomposition gases were confined. DSC and TG runs in open sample pans in nitrogen or helium had higher activation energies and frequency factors, namely 40-43 kcal/mol and 18-20 min^-1, respectively. Thus, it appears that confinement of the ADN decomposition gases either under pressurization or a sealed environment accelerates the decomposition of prilled ADN. Kinetics of decomposition of unprilled ADN at atmospheric pressure in nitrogen or helium by DSC and TG were E(a)=37 kcal/mol and log A=17 min^-1. Unprilled ADN had lower activation energies and frequency factor for decomposition than prilled ADN which contained stabilizers. The effect of sample containment (i.e. aluminum, coated aluminum, gold, and glass ampoules) on the shape of the DSC curve of ADN was investigated. Isothermal TG in vacuum of ADN in the 45-75°C region and the DSC analysis of the TG residues showed unusual results in the 60°C region. The rate and enthalpy of decomposition of four dinitramide salts and their dielectric relaxation were related to the basicity of the cation. (C) 2000 Published by Elsevier Science B.V.

Full text of DOI:10.1016/S0040-6031(00)00387-7

Thermochimica Acta 357±358 (2000) 177±193
Thermal analysis of ammonium dinitramide (ADN)
Albert S. Tompa*
Indian Head Division, Naval Surface Warfare Center, Indian Head, MD 20640-5035, USA
Received 5 October 1998; accepted 26 July 1999
Abstract
Kinetic constants of decomposition were determined for ADN and prilled ADN by DSC and TG. Pressure DSC at 550 psi
air or helium had similar kinetic constants for prilled ADN, namely 29 kcal/mol activation energy and a log frequency factor
of ꢀ14.0 min 1. These values were similar to those found in sealed sample pans in which the decomposition gases were
con®ned. DSC and TG runs in open sample pans in nitrogen or helium had higher activation energies and frequency factors,
namely 40±43 kcal/mol and 18±20 min 1, respectively. Thus, it appears that con®nement of the ADN decomposition gases
either under pressurization or a sealed environment accelerates the decomposition of prilled ADN. Kinetics of decomposition
1
of unprilled ADN at atmospheric pressure in nitrogen or helium by DSC and TG were Eaˆ37 kcal/mol and log Aˆ17 min
.
Unprilled ADN had lower activation energies and frequency factor for decomposition than prilled ADN which contained
stabilizers. The effect of sample containment (i.e. aluminum, coated aluminum, gold, and glass ampoules) on the shape of the
DSC curve of ADN was investigated. Isothermal TG in vacuum of ADN in the 45±758C region and the DSC analysis of the
TG residues showed unusual results in the 608C region. The rate and enthalpy of decomposition of four dinitramide salts and
their dielectric relaxation were related to the basicity of the cation. # 2000 Published by Elsevier Science B.V.
Keywords: Ammonium dinitramide; Prilled ADN; Kinetics of decomposition
1. Introduction
study of ADN in vacuum (20±35 mm) was followed
in the 45±758C region. A DSC analysis of the TG
residue was carried out. The TG/DSC curves were
unusual. DSC and TG kinetic analyses of prilled/
unprilled ADN under different experimental condi-
tions were performed. Finally, a relationship was
observed between the rate and enthalpy of decom-
position of four dinitramide salts and their low tem-
perature dielectric relaxation with the basicity of the
cation.
Ammonium Dinitramide [NH4N (NO2)2] abbre-
viated ADN is a relatively new oxidizer which may
be environmentally friendly because it does not have a
halogen acid decomposition product. The decomposi-
tion reactions of ADN have been discussed by many
researchers [1±3]. In the present study, the effect of
sample pans, i.e. aluminum, coated aluminum, gold,
and glass ampoule on the decomposition curve of
ADN along with the effect of pressure and sealed
sample pan will be discussed. Processing of ADN
based propellant has been carried out under vacuum in
the 608C, region. Consequently, an isothermal TG
2. Experimental
Three samples of ADN, namely prilled, unprilled
NSWC Pilot Plant Lots 50-4, 50-6 and Lot K (recrys-
* Tel.: ‡1-301-743-4671.
0040-6031/00/$ ± see front matter # 2000 Published by Elsevier Science B.V.
PII: S 0 0 4 0 - 6 0 3 1 ( 0 0 ) 0 0 3 8 7 - 7

178
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
tallized), and BOFORS (Department of Energetic
Materials, Sweden) were investigated. DSC, TG,
and DEA experiments were carried out on TA Instru-
ments models 2910, 2950 and 2970, respectively with
2100 and 3100 thermal analysts. To test the precision
of the DSC analysis ®ve runs at 58C/min in unsealed
sample pans with a pin-hole in the cover were carried
out and the peak temperature, rate, and enthalpy of the
melting peak and decomposition peak were recorded
for the different sample pans in a nitrogen atmosphere.
Pressure DSC runs were made in a TA pressure DSC
cell at 550 psi air and helium. Kinetic analyses were
carried out employing a variable heating rate method
[4].
decomposition exotherm, and sublimation endotherm
due to ammonium nitrate [5]. With a coated (ano-
dized) aluminum pan prilled ADN showed a doublet
decomposition peak (Fig. 2). While in a gold pan the
shape of the decomposition and sublimation peaks
were different (Fig. 3). It was thought that the dis-
sociation of ADN into ammonia and dinitramide acid
(strong acid) may react with the regular aluminum
sample pans. However, a reaction was only apparent
with coated aluminum pans and possibly with gold
pans. Ammonium nitrate (AN), a decomposition pro-
duct of ADN showed an endothermic sublimation
peak in 2508C region (Fig. 4) with regular Al pans
and an exotherm peak in coated Al pans (Fig. 5).
AN dissociates into ammonia and nitric acid (strong
acid).
3. Discussion
3.2. Sealed glass ampoules
3.1. DSC sample pans
In a sealed glass ampoule, prilled ADN displayed a
much sharper exotherm peak (Fig. 6) with a three-fold
increase in the reaction rate at the peak maximum
The DSC curve in Fig. 1 is for BOFORS ADN with
the following thermal events, melting endotherm,
Fig. 1. DSC curve of BOFORS ADN in nitrogen at 58C/min.

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
179
Fig. 2. DSC curve of prilled ADN in a coated Al pan in nitrogen at 58C/min.
(11.2 vs. 4.5 W/g). Interestingly, a dark brown gas
(NO2) was observed in the ampoule when it was hot
(ꢀ3008C) which turned colorless (N2O4) as the
ampoule cooled. NSWC ADN Lot 50-6 was slightly
broader than prilled ADN with a reaction rate only
doubled (Fig. 7). BOFORS ADN in Fig. 8 was dif-
ferent from the other two and resembled more closely
the decomposition exotherm in an aluminum pan
(Fig. 1).
3.3. Pressure DSC
In an atmosphere of 550 psi air prilled ADN in Fig. 9
resembled prilled ADN in a sealed glass ampoule
Table 1
Prilled ADN in standard aluminum pans with pin-hole in cover at 58C/min in nitrogen
Melta
Decompositionb
Endothermc
Mass (mg)
Peak (8C)
Enthalpy (J/g)
Peak (8C)
Rate (J/g s)
Enthalpy (J/g)
Peak (8C)
Enthalpy (J/g)
1.37
1.96
2.37
2.83
3.10
92.8
92.8
93.2
92.0
92.5
124.5
120.0
118.3
114.3
115.6
182
180
179
180
±
4.51
4.52
4.61
4.42
4.62
1826
1884
1876
1873
1894
236
243
241
250
248
±
289
249
288
275
a Standard deviations: melt (8C): 92.7Æ0.48C; melt (J/g): 118Æ4.
b Decomposition (8C): 180Æ1; decomposition (J/g s): 4.54Æ0.08; decomposition (J/g): 1871Æ26.
c Endotherm (8C): 244Æ5; endotherm (J/g): 275Æ19.

180
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 3. DSC curve of prilled ADN in a gold pan in nitrogen at 58C/min.
(Fig. 6) except its melting peak was slightly broader
and it was shifted to a lower temperature (838C).
NSWC ADN Lot K in 550 psi air in Fig. 10 resembled
an atmospheric run except the peak temperatures for
melting and decomposition were shifted to a lower
temperature, namely 86 and 1678C, respectively.
Switching the atmosphere to 550 psi helium in
Fig. 11 broadened and lowered the melting peak to
648C and produced a broader decomposition peak
with a pronounced high temperature shoulder at
1798C. In a vacuum (35 mm), the DSC curve of
NSWC ADN Lot K in Fig. 12 again resembled that
in an atmospheric run as far as peak temperatures for
melting and decomposition were concerned but the
enthalpy of decomposition was lowered from ꢀ1871
to 1315 J/g because of sublimation. Reproducibility of
®ve ADN (prilled) runs are presented in Table 1 and
the data showed good precision. The effect of sample
containment on ADN (prilled) are given in Table 2 and
the data were able to differentiate between the sample
pans, i.e. aluminum, gold vs. coated aluminum vs.
glass ampoule.
Table 2
Comparison of DSC data for prill ADN obtained in different sample containers in nitrogen at 58C/min
Melt
Decomposition
Sample pan
Peak (8C)
Enthalpy (J/g)
Peak #1 (8C)
Rate (J/g s)
Enthalpy (J/g)
Peak #2 (8C)
Aluminum
93
93
93
91
118
126
124
120
180
179
181
176
4.54
5.08
4.47
15.2
1871
2048
1836
2126
Coated Aluminum
Gold
195
Glass ampoule

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
181
Fig. 4. DSC curve of AN in nitrogen at 58C/min.
4. Isothermal TG
ADN exothermic decomposition peak temperature,
rate, and enthalpy remained unchanged (compare to
Fig. 1). Surprisingly, the DSC curve in Fig. 18 for the
BOFORS TG residue at 778C/1600 min did not show
the presence of AN at 598C and the curve was similar
to untreated ADN in Fig. 1. Isothermal TG weight-loss
data in vacuum are presented in Table 3 for prilled
Isothermal TG curves in a vacuum of 20±35 mm for
BOFORS at 52, 62, and 778C revealed some unex-
pected results (Figs. 13±15). The percent weight loss
was ꢀ26% after 1400±2500 min at 52 and 628C but
only ꢀ4% at 778C. The experiments were repeated
and the results were the same. Prilled ADN produced
the same results as BOFORS. NSWC ADN Lot 50-4
was different (Fig. 16); its isothermal curve at 628C
only resulted in a 7% weight loss, but Lot 50-6 was
similar to BOFORS ADN. The only differences
between NSWC Lots 50-4 and 50-6 were particle size
(i.e. small clumps vs. powder) and perhaps absorbed
moisture (<0.5%). DSC runs were carried out on the
isothermal TG residues at the different temperatures.
Typical DSC curve for 52/628C TG residue is pre-
sented in Fig. 17 and at 778C in Fig. 18 for BOFORS
ADN. Fig. 17 showed the presence of AN with an
endotherm peak at 598C which subsequently lowered
the ADN melting peak from 94 to 838C. However, the
Table 3
Isothermal TG weight loss rate data for prill ADN in vacuum
(20 mm)
Temperature (8C)
Rate (10 2%/min)
K1 1st step
K2 2nd step
40
45
50
55
60
65
70
75
0.05
0.3
0.5
3.4
8.6
4.5
0.5
0.7
0.2
1.9
4.3
2.0
2.7
1.4
0.2
0.3

182
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 5. DSC curve of AN in coated Al pans in nitrogen at 58C/min.
ADN. It showed that the rate increased in the 50±
658C, region with a maximum in the 608C region.
Russell [1] found that ADN-An formed a eutectic in
this temperature range; Manelius [6] observed a beha-
vioral change in this region; Tompa et al. [7] reported a
rate change by microcalorimetry in this range. Thus,
isothermal heating in vacuum of ADN at temperatures
in the region of the AN phase transition at 598C favors
the formation of AN in ADN, while at temperatures
ca. 108C higher do not. Since ADN exothermic
decomposition peak temperature, rate and enthalpy
of reaction were apparently unaffected by treatment in
both temperature regions, the phenomenon may be
due to sublimation and not decomposition. Observa-
tion of behavioral change in 608C region were also
found by MDSC dynamic heat capacity measurements
of NSWC ADN Lot 50-6 which showed a small
plateau in this region (Fig. 19); and a TMA expansion
curve (Fig. 20) which showed probe penetration in this
region.
4.1. Kinetics
Typical DSC Arrhenius plot for ADN is presented
in Fig. 21 and by TG in Fig. 22. The TG plots showed
slope changes with extent of conversions; activation
energies decreased as the extent of conversion
increased which was in agreement with that reported
by Vyazovkin [3]. DSC and TG kinetic data are
reported in Tables 4 and 5 for prilled and unprilled
ADN. Pressure DSC at 550 psi air or helium had
similar kinetic constants for prilled ADN, namely
29 kcal/mol (122 kJ/mol) activation energy and a
log frequency factor of ꢀ14.0 min 1. These values
were similar to those found in sealed sample pans in
which the decomposition gases were con®ned. Kinetic
runs for prilled ADN in open sample pans in nitrogen
or helium had higher activation energies and fre-
quency factors, namely 40±42 kcal/mol (168±
176 kJ/mol) and 18±20 min 1, respectively. Thus,
its appeared that con®nement of the ADN decomposi-

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
183
Fig. 6. DSC curve of prilled ADN in sealed glass ampoule at 58C/min.
tion gases either under pressurization or a sealed
environment accelerated the decomposition reaction.
Kinetic of decomposition of NSWC and BOFORS
ADN in nitrogen or helium by DSC were 36±38 kcal/
mol (151±159 kJ/mol) and log A of 17±18 min 1. In a
vacuum of 35 mm, the DSC kinetic constants for
Table 4
DSC and TG kinetic data for prilled/unprilled ADN
1
)
ADN
Method
Ea (kcal/ml)
Log A (min
1. Prill
PDSC, 550 psi air
PDSC, 550 psi He
DSC, N2
29.3
29.5
41.9
43.1
30.8
40.2
13.92
14.22
19.97
20.32
14.67
18.67
DSC, He
DSC, sealed pan
TG, He
2. NSWC LOT K
DSC, N2
TG, He
36.8
37.9
17.31
17.62
3. NSWC LOT 50-6
DSC, N2
38.6
31.2
20.9
43.2
18.00
14.42
11.47
20.64
DSC, vacuum (35 mm)
DSC, glass ampoule
TG, N2
Aged in TG 25 mm vacuum at 568C/2500 min
DSC, N2
39.1
18.38

184
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 7. DSC curve of NSWC ADN Lot 50-6 in a sealed glass ampoule at 58C/min.
Table 5
DSC/TG dynamic kinetic data for BOFORS ADN
unprilled NSWC ADN was similar to those for con-
®ned/pressurization prilled ADN. The lowest value for
decomposition kinetics were found for NSWC/
BOFORS ADN in sealed glass ampoules. Dark brown
gas (NO2) was visible in the free air space in the
heated ampoule which turned colorless (N2O4) as the
ampoule cooled. Prilled ADN which contained a
stabilizer had higher activation energy and frequency
factor than unprilled ADN in nitrogen/helium.
DSCa
DSCb
TGc
Ea (kJ/mol)
161
38
89
168
40
Ea (kcal/mol)
21
1
)
Log A (min
18.89
9.80
19.32
a DSC in unsealed pan with pin-hole in cover in nitrogen ¯ow.
b DSC in sealed glass ampoule.
c TG in nitrogen ¯ow.
Table 6
MDSC and DSC data for dinitramide salts
Sample
MDSC
DSC
Tr (8C)a
Rate (W/g)d
Enthalpy (J/g)e
b
Cp at TG
c
Td Peak
AND
HDN
KDN
NaDN
47
25
5
1.3
1.2
0.7
0.9
180
5.6
4.7
1.6
0.9
2508
1731
652
203
220, 230f
212, 162f
2
574
a Relaxation transition temperature (8C).
b Speci®c heat capacity (J/g 8C).
c Decomposition peak temperature (8C).
d Rate, decomposition rate at peak maximum (W/g).
e Enthalpy, exothermic heat of decomposition (J/g).
f Minor decomposition peak (8C).

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
185
Fig. 8. DSC curve of BOFORS ADN in a sealed glass ampoule at 58C/min.
4.2. Dinitramide salts
5. Conclusions
The dinitramide salts of ammonia (ADN), hexam-
ethylenetetramine (HDN), potassium (KDN), and
sodium (NaDN) were investigated by Tompa [8].
He found that the rate and enthalpy of decomposition
decreased as the cation basicity increased (Table 6)
and that the dielectric relaxation peak temperature and
activation energy for relaxation increased as the cation
basicity increased (Table 7).
Kinetic constants of decomposition were deter-
mined for ADN and prilled ADN by DSC and TG.
Pressure DSC at 550 psi air or helium had similar
kinetic constants for prilled ADN, namely 29 kcal/mol
activation energy and a log frequency factor of
ꢀ14.0 min 1. These values were similar to those
found in sealed sample pans in which the decomposi-
tion gases were con®ned. DSC and TG runs in open
Table 7
DEA data for dinitramide salts
Sample
Tan d
(8C)
Tan d at
Tr
Ea (kJ/mol)
10 Hz
100 Hz
1000 Hz
1000 Hz
AND
HDN
KDN
NaDN
75
31
23
19
68
24
18
10
58
14
6
9.0
7.8
4.8
3.5
96
142
205
302
1

186
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 9. PDSC curve of prilled ADN in 500 psi air at 58C/min.
Fig. 10. PDSC curve of NSWC ADN lot K in 550 psi air at 58C/min.

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
187
Fig. 11. PDSC curve of NSWC ADN lot K in 550 psi helium at 58C/min.
Fig. 12. DSC curve of NSWC ADN lot 50-6 35 mm vacuum at 58C/min.

188
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 13. Isothermal TG curve of BOFORS ADN in 35 mm vacuum at 528C.
Fig. 14. Isothermal TG curve of BOFORS ADN in 35 mm vacuum at 628C.

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 15. Isothermal TG curve of BOFORS ADN in 35 mm vacuum at 778C.
Fig. 16. Isothermal TG curve of NSWC ADN lot 50-4 in 35 mm vacuum at 628C.
189

190
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 17. DSC curve of BOFORS ADN TG residue (528C) in nitrogen at 58C/min.
Fig. 18. DSC curve of BOFORS ADN TG residue (778C) in nitrogen at 58C/min.

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
191
Fig. 19. MDSC heat capacity curve of NSWC ADN lot 50-6 in nitrogen at 18C/min.
Fig. 20. TMA expansion curve of prilled ADN under a load of 0.01 N at 38C/min.

192
A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
Fig. 21. DSC Arrhenius plot of BOFORS ADN in nitrogen.
Fig. 22. TG Arrhenius plot of BOFORS ADN in nitrogen.

A.S. Tompa / Thermochimica Acta 357±358 (2000) 177±193
193
sample pans in nitrogen or helium had higher activa-
tion energies and frequency factors, namely 40±
43 kcal/mol and 18±20 min 1, respectively. Thus, it
appears that con®nement of the ADN decomposition
gases either under pressurization or a sealed environ-
ment accelerates the decomposition of prilled ADN.
Kinetics of decomposition of unprilled ADN at atmo-
spheric pressure in nitrogen or helium by DSC and TG
were Eaˆ37 kcal/mol and log Aˆ17 min 1. Unprilled
ADN had lower activation energies and frequency
factor for decomposition than prilled ADN which
contained stabilizers. The effect of sample contain-
ment (i.e. aluminum, coated aluminum, gold, and
glass ampoules) on the shape of the DSC curve of
ADN was investigated. Isothermal TG in vacuum of
ADN in the 45±758C region and the DSC analysis of
the TG residues showed unusual results in the 608C
region. The rate and enthalpy of decomposition of four
dinitramide salts and their dielectric relaxation were
related to the basicity of the cation.
References
[1] T.P. Russell, A.G. Stern, W.M. Koppes, C.D. Bedford, CPIA,
Publ. II, 1992, p. 339.
[2] J.C. Oxley, J.L. Smith, W. Zheng, E. Rogers, M.D. Coburn, J.
Phys. Chem. 101 (1997) 4375.
[3] S. Vyazovkin, C.A. Wright, J. Phys. Chem. C. 101 (1997)
5653.
[4] ASTM Method B 698-79 (reapproved 1993).
[5] S. Lobbecke, H.H. Krause, A. Pfeil, Propellants, Explosives,
Pyrotechnics 22 (1997) 184.
[6] G.B. Manelius, 26 International Annual Conference of ICT,
April 1994.
[7] R.F. Boswell, A.S. Tompa, A. Chen, D. Ellison, BMDO/ONR
Review Meeting, Chesterstown, MD, October 1995.
[8] A.S. Tompa, R.F. Boswell, P. Skahan, C.J. Gotzmer, Journal
of Thermal Analysis, 49 (1997) 1161.