Ammonium nitrate: Thermal stability and explosivity modifiers

Article abstract of DOI:10.1016/S0040-6031(01)00775-4

Two basic approaches to diminish the explosivity of AN have been suggested: dilution of ammonium nitrate (AN) with a chemically inert material or incorporation of small amounts of material which increases the chemical reaction zone [Method of Desensitizing AN and the Product Obtained, United States Patent Office, No. 3,366,468 (1968); Fertilizer additives: can AN be defanged? C&E News (1995) 6]. While small-scale tests of these deterred AN formulations appeared promising, larger amounts (30 pounds under confinement) were found to be detonable, though with reduced performance [Institute of Makers of Explosives Report, 1995]. In this study thermal analysis was used to screen a large number of AN formulations in search of possible deterrents. The sodium, potassium, ammonium and calcium salts of sulfate, phosphate, or carbonate as well as certain high-nitrogen organics (urea, oxalate, formate, guanidinum salts) were chosen because they should enhance AN thermal stability and because they could be used with agricultural products. This study considers whether laboratory tests can be used as benchmarks in evaluating explosivity.

Full text of DOI:10.1016/S0040-6031(01)00775-4

Thermochimica Acta 384 (2002) 23–45
Ammonium nitrate: thermal stability and explosivity modifiers
Jimmie C. Oxley*, James L. Smith, Evan Rogers, Ming Yu
Chemistry Department, University of Rhode Island, Kingston, RI 02881, USA
Abstract
Two basic approaches to diminish the explosivity of AN have been suggested: dilution of ammonium nitrate (AN) with a
chemically inert material or incorporation of small amounts of material which increases the chemical reaction zone [Method
of Desensitizing AN and the Product Obtained, United States Patent Office, No. 3,366,468 (1968); Fertilizer additives: can AN
be defanged? C&E News (1995) 6]. While small-scale tests of these ‘‘deterred’’ AN formulations appeared promising, larger
amounts (30 pounds under confinement) were found to be detonable, though with reduced performance [Institute of Makers of
Explosives Report, 1995]. In this study thermal analysis was used to screen a large number of AN formulations in search of
possible deterrents. The sodium, potassium, ammonium and calcium salts of sulfate, phosphate, or carbonate as well as certain
high-nitrogen organics (urea, oxalate, formate, guanidinum salts) were chosen because they should enhance AN thermal
stability and because they could be used with agricultural products. This study considers whether laboratory tests can be used
as benchmarks in evaluating explosivity. # 2002 Elsevier Science B.V. All rights reserved.
Keywords: Explosivity; Ammonium nitrate; Thermal stability; Decomposition
1. Introduction
comparison with other compounds of equal effect
[7,8]. Over the past century there have been many
large accidents as well as terrorist bombings involving
AN. Table 1 summarizes some of the AN accidents
[9,10]. Terrorist use of AN began in the bombing
campaign of the Provisional Irish Republican Army
(PIRA) (1969–1994). During that period there were
14,000 bombing incidents, most involving commer-
cial explosives or sodium chlorate/nitrobenzene (NB).
At the peak of the campaign in the early 1970s, the
British government issued a ban on the sale of chlo-
rate, NB, and pure AN in Northern Ireland. (Similar
restrictions were imposed in Ireland.) Nevertheless,
large fertilizer bombs were used in the City of London.
Approximately 1000 pounds were used at St. Mary le
Axe (April 1992) and about 3000 pounds at Bishops
Gate (April 1993) [11]. In other countries, AN has
been used less frequently in terrorist bombings; a
notable exception was the bombing of the Murrah
Federal Building in Oklahoma City, 19 April 1995.
The US Census Bureau data for 1998 showed the
total ammonium nitrate (AN) melt production to be
about 18.2 billion pounds. About 13.8 billion pounds
were produced for fertilizer, about 5.2 billion pounds
in solid form and 8.6 billion pounds in various solu-
tions formulations [1–3]. Only, ꢀ3.3 billion pounds
were converted to solid products for use in ANFO
(AN/fuel oil) and related blasting products; another 1
million pounds were used in emulsion and slurry
formulations for blasting purposes [4,5]. Pure AN is
considered to be highly stable and relatively safe. The
Department of Transportation has classified it as an
oxidizing agent for transport, storage and handling
purposes [6]. Combined with fuel, AN is used in many
explosive applications, and it is relatively safe in
* Corresponding author.
E-mail address: joxley@chm.uri.edu (J.C. Oxley).
0040-6031/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 4 0 - 6 0 3 1 ( 0 1 ) 0 0 7 7 5 - 4

Table 1
History of AN accidentsa
Where
When
Amount AN
Unit
lb
Material
Circumstance/damage
Mechanism
Speculation
Dead Reference
[43]
Stolberg, Germany
Vergiati, Italy
12 April 1920
Waggon-load
Caked AN
Blast ! explode
Blasting to break caked AN
First explosion in magazine
of bombs, third reported here
Blasting to break caked AN
AN/AS detonable or
26 November 1920
4.4E6þ1.2E5
AN þ TNT in magazine
Destroyed all 1600 in.; windows Blast ! explode
0
[47]
broke 21 million
Kriewald, Germany
Oppau, Germany
26 July 1921
21 May 1921
2 car-loads
900000
Caked AN
AN/AS 2/1
60 in. Â 20 in. crater
Buildings flatten to 3000 in.;
2000 injured
Blast ! explode
Blast ! explode
19
[43–46]
lb
T
561
[9,43,45,46]
hiding explosives
Tessenderloo, Belgium
Oakdale, PA
29 April 1942
150–200
3000
AN pile
Plant destroyed, crater 164 in.
 36 in.
Blast ! explode
Fire ! explode
Fire ! explode
Blasting caked AN; or tools
ignited AN and CO3/PO4
NA from TNT plant and
possibly tar
200
[9,43,45–47]
[43,46]
15 September 1916
7 November 1925
lb
lb
AN IN preboil tank
Emporium, PA Hercules
500
AN with wax in water,
heel in evaporation pan
AN
[43,46]
Gibbstown, NJ DuPont
Texas City, TX
26 October 1932
High pan
Fire ! explode
Fire ! explode
[43,46]
16–17 April 1947
2–5.5 K
T
lb
T
Waxed prill
1000 injured; US$ 60 million
damage, destroyed 5 million
Wax coated prills on Grand
Camp and highflyer
600
13
[9,43,45,47]
Pinole, CA Hercules
Roseburg, OR
12 February 1953
7 August 1959
2400–3600
4.5 þ 2
AN, NaNO3, organic—mixed
to start making dynamite
AN car-prill þ dynamite
Fire ! explode
Fire ! explode
Fire ! explode
Fire ! explode
[43,46]
[9,45]
125 injured, US$ 9 million;
stored downtown
Explosives involved
Traskwood, AR
Norton, VA Atlas
Mt. Vernon, MO
17 December 1960
27 December 1961
9 November 1966
23 boxcars
20 þ 35
1 car AN bag, 1 bulk, fertilizer
solution, NA
Petroleum and paper in
cars—involved
[9,46]
T
AN þ ANFO
Five injured
Welding AN/FO transfer
pipeline
0
0
0
[9,43,46,47]
50
T
T
Bagged AN fertilizer
10 in. deep carter
Eight injured
Fire ! explode
Fire ! explode
[43]
Prvor, OK Cherokee N Plant 17 January 1973
14000
High density AN in PE bags
10% detonated
[43,47]
destroying building
Barksdale, WI
Brooklyn, NY
April 1920
30000
lb
lb
Fire no explosion
Fire no explosion
[43,44,46]
[43,44,47]
14 April 1920
2–5E6
AN
AN
Hold of SS Hallfried
High pan
Fire started in hold with
chlorate and newsprint
Sinnnemahoning, PA
Cleveland
10 September 1922
1922
4300
lb
Fire no explosion
Fire no explosion
Fire no explosion
[43]
[46]
Muscle Shoals, AL
4 April, 3 May, 1925
2 events
Boxcars with 220 flour
barrels of AN
Same batch AN as in
Hercules fire below
[43,44,46]
Gibbstown, NJ DuPont
St. Stephens, Canada
April 1940
1.5, 6.5E5
400
lb
T
AN in paper sack and steel drums
AN fertilizer bagged
Fire no explosion
Fire no explosion
[43,44,46]
1 September 1947
B said like Benson, AZ;
[43,44,46,47]
DoD said fire in warehouse
Presque Isle, ME
Independence, KS
26 August 1947
14 October 1949
80000
lb
T
AN þ superphosphate, K,
meal, etc. fertilizers
FGAN (4% wax)
Fire no explosion
Fire no explosion
[43,44,46]
1.4–2.7 K
Piled in paper bags 15 ft high
Fire in mixer in dope house
2.5 in. separation vented, fire
in next warehouse ignited AN
[43,44,46,47]
Mt. Braddock, PA Olin
Boron, CA
16 January 1958
10 May 1960
–
AN þ NaNO3
Fire no explosion
Fire no explosion
0
[47]
20
T
Prilled AN in 80 lb bags
Fire hot enough to collapsed
steel roof
[43,44]
Peytona, WV
Potosi, WI
9 November 1966
24 October 1967
Boxcar at dock with AN grade
100 lb bags
Fire no explosion
Steel car hot enough to turn red
Fire in consumed wood lining
left some AN
[43]
50
T
Steel boxcar of AN, wood lined Fire no explosion
[43,44]
Bucharest, Romania
Tahawas, NY IRECO
20 December 1974
1 March 1976
33
lb
lb
AN þ wood powder
Two slight injuries
Fire no explosion
Fire no explosion
Fire in mixer of 15 kg
wood powder þ trace AN
Two explosions, cause:
electrical or arson
0
0
[47]
[47]
100
AN
No injuries, minor damage

Rocky Mountain, NC
Moreland, ID
Boxcars
1978
500
T
T
lb
AN fertilizer
Fire no explosion
Fire no explosion
Runaway ! fire
! no explosion
Thermal runaway
! detonation
Storage facility destroyed by fire
[44]
1979
200
AN fertilizer
[44]
1946–1949
80–100 K
13 railroad cars caught fire
AN loaded hot 93 8C
[43,44,46]
Kensington, UK
Gibbstown, NJ DuPont
NewBrunswick, NJ Nixon
Merano, Italy
2 March 1896
14 January 1916
1 March 1924
1936
18
lb
lb
lb
AN for N2O production
AN decomposed purposely;
pressure rupture
[43,46]
[43,46]
[43,46,47]
[46]
400
Molten AN, 160 8C
High pan
Thermal runaway
! detonation
NA from TNT plant
4800 þ 35À50 K
AN in grain kettle and 35–50 K warm
Thermal runaway
! detonation
AN being recycled from
Amatol 2% TNT and wax
7
4
AN at 0.3% paraffin
Thermal runaway
! detonation
Milan, TN
2 March 1944
2 December 1944
28 July 1947
23 January 1953
28 December 1956
9 January 1963
1977
4800
lb
lb
T
Molten AN, 154 8C
AN 149 8C
High pan; 17 injured;
damage past 600 in.
High pan
Thermal runaway
! detonation
Oil in air lines
[43,46,47]
[43]
Benson, AZ Apache
Brest, France
8500
Thermal runaway
! detonation
3.3–6 K
4–8 K
5200
Waxed prill
250 injured; SS Ocean Liberty
Destroyed SS Tirrenia
Thermal runaway
! detonation
21
0
[43,45,47]
[43,47]
[47]
Red Sea, Israel
New Castle, PA
Typpi, Oy, Finland
Delaware City, DE
Morgan, NJ
T
AN fertilizer–paper bag
AN
Thermal runaway
! detonation
Spontaneous ignition of
paper bags
lb
T
One injured, exposion in
evaporator
Thermal runaway
! detonation
8–10
Molten AN, 160 8C
Thermal runaway
! detonation
Dodecyl benzene sulphonate
added to prevent caking
Heated to 518 8F (270 8C),
production exceeded equipped
Amatol loading, explosive
detonated
[43]
83% AN solution for N2O
Thermal runaway
! detonation
production
AN
1918
9Eþ06
lb
lb
150 in. Â 30 in. crater
Fire ! blast
[43,46]
[43]
under AN pile
! detonation
Amboy, IL
7 April 1966
1989
30000
AN in truck next to bunker at
2000 lb TNT and 10000 lb AN
AN production in Stengel Reactor
Blast no detonation Explosion of bunker did not
set off truck of AN
Joplin, MO
High temperature rx 450–500 8F
a High pan is for evaporation water from AN before AN goes to graining kettles.

26
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
This event generated concern in the US regarding the
explosive nature of AN. Because AN explosives are
easily prepared and the Oklahoma City bombing was
so devastating, a number of research programs aimed
at desensitizing commercially available AN were
developed.
carbonate). This mixture has enhanced thermal stabi-
lity. (Unfortunately, PIRA found a way to detonate this
mixture, albeit with diminished performance [11].)
2. Experimental section
AN melts at 169 8C and begins to decompose as
soon as it melts, the first step being dissociation into
ammonia and nitric acid Eq. (1). Studies have shown
that the decomposition of AN operates by an ionic
mechanism in the temperature range 200–300 8C and
formation of nitronium ion is rate-determining
[12,13]. Acidic species, such as ammonium ion,
hydronium ion or nitric acid, increase the rate of
AN decomposition dramatically, Eq. (2), while bases
such as ammonia or water retard decomposition.
Above 290 8C, a free-radical decomposition mechan-
ism dominates and homolysis of nitric acid is the rate-
controlling step [12,13].
Relative thermal stabilities were determined by
differential scanning calorimetry (DSC) with a TA
Instruments DSC 2910, scanning from 50 to 500 8C, at
a scan rate of 20 8C/min and using thin walled 1.47–
1.50 mm o.d. glass capillary tubing [14]. Three repli-
cates were run; the average temperature of the
exotherm or endotherm maxima and the heat of reac-
tion are given in tables. A shift from the position of the
exothermic maximum of the mixture from that of neat
AN (326 8C) to lower temperatures was interpreted as
destabilization, while a shift to higher temperatures
was considered indicative of stabilization. Consider-
ing experimental errors, only shifts of 10 8C or greater
were generally considered significant. Additives were
ground together with the AN; the amount added to AN
was generally calculated in weight percent (wt.%)
(where 5 wt.% is defined as 5 g additive in 100 g
AN). Higher percentages of additives were also exam-
ined. In comparing additives, mole percent (mol%)
was also used to ensure that differences in molecular
weights of the additives did not influence observed
trends. A translation between our definition of wt.%
and mol% is shown in Table 2 along with the DSC
characteristics of the neat additives.
Isothermal decompositions of AN and AN mixtures
(0.5–1.0 mg) were conducted in 200 ul glass tubes
(2.0–2.2 mm i.d.), which were flame sealed. After the
samples were heated for given time intervals at 260 or
320 8C in molten metal baths, the glass tubes were
broken and rinsed with water to quantitatively remove
the contents. The contents were diluted to 10 ml for
quantitative analyses of ammonium ion (NH4þ) and
nitrate ion (NO3À). The fraction of nitrate remaining
was determined using a Hewlett-Packard (HP) 1084B
liquid chromatograph (LC) equipped with a Dionex
Ionpack AG4A guard column and AS4A analytical
column. The mobile phase was 1.8 mM Na2CO3/
1.7 mM NaHCO3 at a flow rate of 1.5 ml/min. Detec-
tion was with a Waters 486 Tunable Absorbance Detec-
tor (214 nm). Injection volumes were 20 ul. The
fraction of ammonium ion remaining was determined
NH4NO3 , NH3 þ HNO3
(1)
ÀA
þ
HNO3 þ HA,H2ONO2 ! NO2þ þ H2O;
where HA ¼ NH4þ; H3Oþ; HNO3
slow
NO2þ þ NH3 ! ½NH3NO2 Š ! N2O þ H3Oþ
(3)
(4)
þ
NH4NO3 , N2O þ 2H2O
As a fertilizer and an explosive, AN is often mixed
with a number of additives. A low temperature, ionic
decomposition pathway for AN suggests acidic addi-
tives destabilize while basic additives stabilize thermal
decomposition. This study sought to thermally stabi-
lize AN using additives compatible with fertilizer use.
Our hypothesis was that increased thermal stability
could lead to reduced detonability. The basis for this
premise is two-fold. First, a material which retards AN
decomposition may increase the reaction zone of the
composition to the point that a detonation cannot be
propagated. Second, two AN formulations, which
have been considered non-detonable, contain species
that thermally stabilize AN. Experiments following
the Oppau explosion strongly suggested that the 50/50
mixture of AN and ammonium sulfate was non-deton-
able. (The accident was explained as resulting from a
pocket of concentrated AN [9].) To thwart PIRA, AN
sold in Northern Ireland is sold as a mixture containing
21% dolomite (the double salt calcium/magnesium

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
27
Table 2
DSC peaks (maxima or minima) of neat additives and conversion of weight (wt.%) to mole (mol%)a
Sample
F.W. Endotherms
Peak 1
Exotherms
Endotherm Peak 4
Tmin (8C) E (J/g) Tmax (8C) E (J/g)
wt.%
5
mol%
10
Endotherm Peak 2
Endotherm Peak 3
Exotherm
Tmin (8C) E (J/g)
128 51
Tmin (8C) E (J/g)
20
20
AN
80
85
168
335
68
89
326
1182
NaNO3
308 149
4.5 15.8 10.6 21.0
3.8 13.7 12.3 24.0
KNO3
101
164
106
84
132
46
Ca(NO3)2
Na2CO3
NaHCO3
K2CO3
2.4
8.9 18.6 33.9
3.6 13.1 12.8 24.9
4.5 16.0 10.4 20.8
2.8 10.4 16.1 30.1
3.8 13.8 12.2 23.8
3.8 13.8 12.2 23.8
4.5 16.0 10.4 20.8
4.0 14.3 11.8 23.1
2.7 10.1 16.5 30.7
3.2 11.8 14.3 27.3
200
231
683
326
138
100
100
84
KHCO3
CaCO3
290
116
MgCO3
(NH4)2CO3
Na2SO4
NaHSO4
K2SO4
405
100
257
180
1025
713
73
96
142
120
174
136
132
115
136
164
142
120
212
174
136
136
132
115
60
116
2.2
8.4 19.5 35.2
KHSO4
212
366
140
145
18
2.9 10.5 15.9 29.8
2.9 10.8 15.5 29.2
3.4 12.2 13.8 26.4
2.9 10.5 15.9 29.8
(NH4)2SO4
(NH4)HSO4
CaSO4
95
Na3PO4
Na2HPO4
NaH2PO4
K3PO4
2.4
8.9 18.6 33.9
312
213
36
340
335
11
416
12
2.7 10.1 16.5 30.7
3.2 11.8 14.3 27.3
185
128
1.9
2.2
7.0 22.7 39.8
8.4 19.5 35.2
K2HPO4
KH2PO4
CaHPO4
(NH4)2HPO4
(NH4)H2PO4
Urea
314
230
27
34
369
275
66
241
2.9 10.5 15.9 29.8
2.9 10.5 15.9 29.8
2.9 10.8 15.5 29.2
3.4 12.2 13.8 26.4
220
213
136
242
215
414
279
229
911
192
274
349
216
76
430
401
291
73
6.3 21.1
7.7 15.8
Guanidinum CO3 181
Guanidine NO3 122
2.2 8.1 20.1 36.1
390
295
805
650
3.2 11.6 14.5 27.6
2.8 10.3 16.3 30.4
HMTA
140
77
NH4CH3CO2
KCH3CO2
NH4CHO2
KCHO2
120
167
211
200
119
136
366
321
194
707
4.9 17.2
3.9 14.0 12.0 23.4
6.0 20.3 8.0 16.4
9.7 19.4
98
63
122
160
130
168
174
238
83
84
4.5 16.0 10.4 20.8
3.1 11.4 14.7 27.9
(NH4)2C2O4
K2C2O4
124
184
13
247
392
682
103
117
25
2.1
8.0 20.4 36.5
Icing sugar
Nitromethane
Nitrobenzene
Sulfur
285
433
472
523
2889
861
61
6.2 20.8
7.8 16.0
123
3.1 11.5 14.6 27.8
117
56
a Carbon and diesel no peaks in DSC 50–500 8C.
using a ConstaMetric 3200 MS low pulsation solvent
delivery system equipped with a Waters IC-Pak Cation
M/D column and Waters 431 conductivity detector. A
mobile phase consisting of 0.1 mM EDTA/3.0 mM
HNO3 and a flow rate of 1 ml/min wasþused. Injection
in the additive, that amount of the ion was assumed
to be inert and subtracted from the fraction remain-
ing before calculating the rate constant. Decomposi-
tion gases contained within in the reaction tubes at
260 and 320 8C were quantified by direct injection
into a HP5890 gas chromatograph (GC) with thermal
À
volumes were 10 ul. If NO3 or NH4 were present

28
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
conductivity detection and a Heyesep DB 100/120
column (9.1 mÂ32 mm).
the II to I phase change endotherm. The more additive
mixed with AN, the more the melting point of AN was
depressed, which was in keeping with colligative
properties. Of principal interest was the effect of
various additives on the position of the AN exotherm.
Most of the anions were available as the Naþ, Kþ,
NH4þ, or Ca2þ salts. Among the halide salts, the
positions of the DSC exothermic maxima were nearly
the same for all four ClÀ salts, varying by only about
6 8C (Table 3). We chose the Kþ salt for most of our
studies since potassium is often used for agricultural
products. With the exception of fluoride, the halides
had a destabilizing effect on the thermal stability of
AN. Chloride lowered the exothermic maximum the
most (about 70 8C). Added nitrate had no significant
effect on the temperatures of the exothermic maxima,
compared to neat AN; nor was there any appreciable
difference observed when the cation was varied from
Naþ to Kþ to Ca2þ (Table 4). Table 5 summarizes
DSC exotherms for other oxide salts mixed with AN.
Most of the inorganic additives (sulfates, carbonates or
3. Results
3.1. Differential scanning calorimetry studies
Neat AN scanned at 20 8C/min from 50 to 450 8C
exhibited two endotherms and one exotherm followed
by a final endotherm (Fig. 1). The first endotherm
(ꢀ125 8C) is the result of the II to I phase change and
the second at about 169 8C is the melting point of AN
[15]. The endotherm, following the exotherm, has
been previously examined in our lab and found to
be the vaporization of water formed during AN
decomposition [13]. The observed exotherm was
about 100 8C wide and had an exothermic maximum
of 326 8C. The various additives had either no effect
on the observed AN endotherms or depressed the AN
melting point, in some cases, to where it merged with
Fig. 1. DSC thermogram of neat AN.

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
29
Table 3
Effect of halide salts on AN, DSC exotherm maximum (208/min)a
3.1.2. In situ AN
While it is clear that some of the additives chosen for
this study impart thermal stability to AN, it is not clear
that they will make it less detonable. A more radical
approach to reducing AN detonability might be to use a
formulation consisting of two compounds, one contain-
ing the ammonium cation and the other the nitrate
anion (e.g. AX þ EN or A2Y þ GN2). These examples
would be equivalent to a 50/50 mixture AN with EX or
a 2/1 mixture of AN with GY, respectively. For ferti-
lizer purposes it should make no difference, but we
speculate that detonability would be greatly decreased
by the less intimate contact between oxidizer and fuel.
The first step to confirming this hypothesis was to
determine thermal stabilities. Eight formulations were
examined and compared to the equivalent AN formu-
lation (Table 9). Addition of about 33 mol% CaCO3
substantially stabilized AN (increase in exothermic
maximum is greater than 100 8C). Using the in situ
approach of generating this formulation (A: ammon-
ium carbonate/calcium nitrate) appeared to provide no
further stability (i.e. the exotherm of AN/CaCO3 was
433 8C while that of the in situ mixture Awas 440 8C).
Likewise, the addition of 50% sodium or potassium
hydrogen carbonate or potassium acetate or formate so
stabilized AN that additional stability was not evi-
denced in the in situ formulations (D–G). The addition
of 33 mol% potassium oxalate so greatly enhanced AN
thermal stability that no peak was observed at the limit
of the DSC scan (500 8C) in the mixture or in the in situ
formulation (H). The cases where the in situ approach
appeared to be successful were with additives that did
not dramatically improve AN thermal stability. Addi-
tion of CaSO4 to AN had almost no stabilizing effect,
while, the in situ formulation, ammonium sulfate and
calcium nitrate (B) provided a significant stabilizing
effect. Similarly, when diammonium phosphate and
calcium nitrate (C: exothermic maxima 373 8C) were
mixed, there was a greater stabilizing effect than when
33 mol% calcium phosphate was added directly to AN
(exothermic maxima 341 8C). To simulate the effect of
rain, three of the formulations (A, B, C) were dissol-
ved in water, mixed, and thoroughly dried. For formula-
tion A, when the mixture had been wetted then dried,
most of the enhanced stability was lost, presumably
through decomposition of the ammonium carbonate.
However, the exothermic maxima of formulations B
and C were essentially unchanged by moisture.
wt.% NaCl KCl NH4Cl CaCl2 KF KCl KBr KI
256 319 319
5
10
20
262
263
264
256
254
256
257
256
256
261
262
266
348
381
a Neat AN exothermic maximum: 326 8C.
Table 4
Effect of nitrate salts on AN, DSC exotherm maximum (208/min)
5 wt.%
20 wt.%
AN
326
328
323
328
NaNO3
KNO3
CaNO3
328
326
330
phosphates) raised the decomposition exotherm of
AN. The exception was bisulfate. All the organic
compounds examined raised the temperature of the
AN exotherm. In addition, they lowered the melting
point of AN so that the sharp AN phase transition at
129 8C and its melt at 169 8C became a single, broad,
poorly defined endotherm. For guanidinum carbonate,
the endotherm began as low as 83 8C (Table 6).
3.1.1. Isothermal kinetics studies
Because DSC is a temperature scan technique, it is
difficult to ascribe the observed relative thermal sta-
bility to a specific temperature range. For this reason,
the rate constants of AN with a number of select
additives were determined from the slope of first-order
plots of isothermal thermolyses at 320 and 260 8C. For
the additives that contained additional nitrate or
ammonium ions, the contribution of the additive
was subtracted so that the relative loss of that ion
did not appear disproportionately slow. The plots
exhibited varying degrees of linearity, depending on
the additive. Only the initial linear section of the curve
was used to determine the slope. Tables 7 and 8 show
the linearity of fit (R2), the number of data points (out
of eight data points) and percent of decomposition for
the portion of the plot used to calculate the indicated
rate constant. Only neat AN and AN with CaSO4
demonstrated linear first-order plots to greater than
50% decomposition at both temperatures examined.

Table 5
Effect of oxide salts on the DSC exotherm (208/min) of AN (neat ¼ 326 8C)
E2CO3
EHCO3
E2SO4
EHSO4
E3PO4
E2HPO4
EH2PO4
5 wt.% mol% 20wt.% mol% 5wt.% mol% 20wt.% mol% 5wt.% mol% 20wt.% mol% 5wt.% mol% 20wt.% mol% 5wt.% mol% 20wt.% mol% 5wt.% mol% 20wt.% mol% 5wt.% mol% 20wt.% mol%
Na
K
354
351
3.6
2.8
4.0
3.8
381
374
374
389
378
13
10
14
348
345
4.5
3.8
378
372
16
14
333
332
337
333
2.7
2.2
2.9
2.9
349
343
348
334
10
8
331
332
331
3.2
2.9
3.4
329
328
322
12
11
12
329
339
2.4
1.9
345
362
9
7
338
339
340
336
2.7
2.2
2.9
2.9
366
364
364
337
10
8
328
337
336
3.2
2.9
3.4
339
348
347
12
11
12
NH4 349
11
11
11
11
Ca
360
Mg
pH
10
7.4–8.5
7.3–7.4
4.8–5.6
1.7–1.8
8.2–8.3
6.4–7.4
4.2–4.3

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
31
Table 6
AN exotherm in the presence of selected organic species
Sample name
Exotherm Tmax (8C)
Heat released
E (cal/g)
E (J/g)
pH
Neat AN
326
370
384
394
402
372
377
357
384
368
385
383
341
408
283
350
368
395
415
368
675
243
332
245
362
308
355
358
1182
1465
1539
1653
1738
1541
2824
1016
1389
1026
1513
1287
1486
1496
4.9
AN þ 5% urea
AN þ 10% urea
AN þ 20% urea
5.0
AN þ 30% urea
AN þ 20% NH4 acetate
AN þ 10 mol% hexamethylenetetramine
AN þ 10 mol% K acetate
AN þ 20 wt.% NH4 formate
AN þ 10 mol% K formate
AN þ 20 wt.% NH4 oxalate
AN þ 10 mol% K oxalate
AN þ 10 mol% guanidine NO3
AN þ 10 mol% guanidine CO3
6.3
6.5
6.3
5.9
5.9
5.9
6.4
4.8
8.2
3.1.3. Decomposition gases
produced/mol of AN, and the principal decomposition
gas was N2O
Table 10 shows the gaseous decomposition pro-
ducts of AN and its various formulations. For neat
AN and most of its mixtures about 1.0 mol of gas were
NH4NO3 ! N2O þ 2H2O
(5)
Table 7
Rate constant (sÀ1) of AN formulation at 320 8C
Additive
103 k/sa
R2
Data
Up to
103 k/s
R2
Data
Up to
Amount
Pure
Exothermic
À
þ
(NO3
)
% dec.
(NH4
)
% dec.
T (8C)
AN
10.9
9.42
9.28
8.88
4.23
2.10
9.41
6.68
19.6
30.1
19.3
22.9
21.9
14.4
11.4
27.4
28.6
13.8
8.53
11.2
7.03
6.74
22.3
1
6
4
4
4
5
6
4
4
3
5
4
3
3
4
4
4
5
4
7
4
8
8
13
67
34
42
55
29
40
43
33
44
74
58
37
35
47
40
55
57
34
92
48
63
60
72
13.2
11.2
1
1
7
4
80
39
327
AN þ NaNO3
AN þ KNO3
0.99
0.99
1
10 mol% 328
10 mol% 323
AN þ Ca(NO3)2
AN þ Na2CO3
AN þ CaCO3
AN þ KHCO3
AN þ K2CO3
AN þ Na2SO4
AN þ (NH4)2SO4
AN þ CaSO4
11.8
5.29
2.10
9.30
0.99
0.95
0.97
0.99
4
5
6
5
65
35
53
53
20 wt.%
10 mol% 381
20 wt.% 389
328
1
1
1
10 mol% 372
10 mol% 374
0.99
1
21.3
23.6
22.4
21.4
1
3
4
4
4
47
51
64
49
20 wt.%
20 wt.%
20 wt.%
349
348
334
1
0.99
1
1
AN þ NaH2PO4
AN þ KH2PO4
AN þ Na2HPO4
AN þ K2HPO4
AN þ (NH4)2HPO4
AN þ CaHPO4
AN þ K3PO4
0.99
1
0.94
10 mol% 339
10 mol% 348
10 mol% 366
10 mol% 364
10 mol% 364
10 mol% 337
10 mol% 362
0.99
0.98
1
14.4
1
4
47
21.6
34.1
13.5
4.08
0.99
1
4
6
4
7
48
74
34
70
1
0.99
1
0.97
0.99
AN þ urea
20 wt.%
394
AN þ guanidinum CO3
AN þ NH4CHO2
AN þ (NH4)2C2O4
AN þ HMTA
1
10 mol% 408
1
20 wt.%
20 wt.%
384
385
0.99
0.99
10 mol% 377
a k/s is rate constant k in units of 1/seconds.

32
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
Table 8
Rate constant (sÀ1) of AN formulation at 260 8C
Additive
104 k/s
R2
Data
Up to
% dec.
104 k/s
R2
Data
Up to
% dec.
Amount
Pure
Exothermic
T (8C)
À
þ
(NO3
)
(NH4
)
AN
AN þ NaNO3
AN þ KNO3
1.95
1.79
1.72
2.75
1.26
0.067
2.00
1.18
1.17
1.97
2.27
2.02
1.57
1.05
0.56
1.29
2.80
1.17
0.34
1.57
0.38
0.12
0.31
0.997
0.99
1
8
3
6
4
3
4
4
5
4
5
8
4
6
4
5
4
4
3
3
5
3
6
7
56
20
66
39
34
9
2.20
1.89
0.984
0.99
8
4
59
29
327
10 mol% 328
10 mol% 323
AN þ Ca(NO3)2
AN þ Na2CO3
AN þ CaCO3
AN þ KHCO3
AN þ K2CO3
AN þ Na2SO4
AN þ (NH4)2SO4
AN þ CaSO4
0.99
1
2.57
1.70
0.09
2.01
0.99
1
4
3
6
4
36
46
18
52
20 wt.%
10 mol% 381
20 wt.% 389
328
0.93
0.98
0.99
0.98
0.99
0.996
1
0.800
0.960
51
49
35
38
62
31
40
18
34
38
40
25
12
53
13
21
43
10 mol% 372
10 mol% 374
1.42
1.79
2.33
2.24
0.993
0.967
0.989
0.994
4
4
8
4
40
28
64
34
20 wt.%
20 wt.%
20 wt.%
349
348
334
AN þ NaH2PO4
AN þ KH2PO4
AN þ Na2HPO4
AN þ K2HPO4
AN þ (NH4)2HPO4
AN þ CaHPO4
AN þ K3PO4
10 mol% 339
10 mol% 348
10 mol% 366
10 mol% 364
10 mol% 364
10 mol% 337
10 mol% 362
10 mol% 394
10 mol% 408
0.95
0.95
0.98
0.98
0.97
0.99
0.97
0.98
1
1.01
0.9
4
18
1.14
3.06
1.16
0.16
1
4
5
3
3
34
50
25
6
0.97
1
AN þ urea
1
AN þ guanidinum CO3
AN þ NH4CHO2
AN þ (NH4)2C2O4
AN þ HMTA
20 wt.%
20 wt.%
384
385
0.96
0.97
10 mol% 377
Table 9
Exothermic maxima of mixtures of AN and in situ formulations
In situ mixture AN ! additive
Nitrogen (wt.%)
a
A
B
C
ðNH4Þ2CO3 þ CaðNO3Þ2 ! 2NH4NO3 þ CaCO3
21.7
19.0
19.0
433 ꢁC ð397 J=gÞ
440 ꢁC ð979 J=gÞ; 364 ꢁC ð929 J=gÞ
a
ðNH4Þ2SO4 þ CaðNO3Þ2 ! 2NH4NO3 þ CaSO4
333 ꢁC ð669 J=gÞ
355 ꢁC ð561 J=gÞ; 356 ꢁC ð494 J=gÞ
a
ðNH4Þ2HPO4 þ CaðNO3Þ2 ! 2NH4NO3 þ CaHPO4
341 ꢁC ð682 J=gÞ
373 ꢁC ð665 J=gÞ; 361 ꢁC ð573 J=gÞ
D
E
ðNH4ÞHCO3 þ NaNO3 ! NH4NO3 þ NaHCO3
17.1
15.6
>500 ꢁ
C
>500 ꢁ
C
ðNH4ÞHCO3 þ KNO3 ! NH4NO3 þ KHCO3
>500 ꢁ
C
>500 ꢁ
C
F
ðNH4ÞCH3CO2 þ KNO3 ! NH4NO3 þ KCH3CO2
15.7
17.1
17.3
416; 471 ꢁC ð1600 J=gÞ
409; 465 ꢁC ð1400 J=gÞ
G
H
ðNH4ÞCHO2 þ KNO3 ! NH4NO3 þ KCHO2
337; 409 ꢁC ð310 J=gÞ
326 ꢁC ð79 J=gÞ
ðNH4Þ2C2O4 þ KNO3 ! 2NH4NO3 þ K2C2O4
>500 ꢁ
C
No exotherm
a Sample was evaporated from aqueous solution.

Table 10
Decomposition gas of neat AN and AN with additives
AN with 10 mol%
additive indicates
20 wt.%a
Time (h)
Gas product mol/mol AN
Gas ratio Time (h)
Gas product mol/mol AN
Gas ratio
(N2O/N2)
at 320 8C
(N2O/N2) at 260 8C
N2
CO2
CO
N2O
Total
N2
CO2
CO
N2O
Total
AN
1
5
0.19
0.14
0.21
0.17
0.32
0.50
0.33
0.22
0.26
0.22
0.26
0.23
0.18
0.27
0.28
0.27
0.23
0.16
0.28
0.93
0.53
0.81
0.60
0.54
0.59
0.78
0.87
0.71
0.85
0.62
0.40
0.63
0.70
0.62
0.74
0.79
0.77
0.81
0.67
0.66
0.71
0.84
0.83
0.64
0.25
0.78
0.25
0.41
0.52
0.61
0.97
1.01
0.91
1.02
1.03
1.01
1.04
0.93
0.93
0.96
1.05
0.99
0.99
0.94
0.93
0.98
1.07
0.99
0.93
1.38
1.48
1.48
1.32
1.24
1.43
4.1
6.1
3.4
4.9
1.9
0.8
1.9
3.3
2.4
3.4
3.1
3.4
4.5
2.5
2.4
2.7
3.7
5.2
2.3
0.3
1.5
0.3
0.7
1.0
1.0
93
93
93
93
72
417
0
0.26
0.22
0.25
0.29
0.36
0.60
0.60
0.60
0.59
0.66
0.48
0.35
0.86
0.82
0.84
0.95
0.84
1.07
2.3
2.8
2.4
2.3
1.3
0.59
AN þ NaNO3
AN þ KNO3
18
24
18.7
24
20
24
43
1
a
AN þ Ca(NO3)2
AN þ Na2CO3
0.10
0.10
0.07
0.02
0.05
0.00
0.12
a
AN þ CaCO3
AN þ MgCO3
AN þ KHCO3
AN þ K2CO3
AN þ Na2SO4
AN þ (NH4)2SO4
AN þ CaSO4
72
72
72
68
93
93
72
89
235
68
93
68
235
72
93
0
0.24
0.26
0.38
0.37
0.41
0.29
0.30
0.43
0.43
0.45
0.28
0.37
0.63
0.62
0.72
0.01
0.01
0.58
0.62
0.56
0.55
0.54
0.51
0.51
0.45
0.43
0.51
0.57
0.44
0.39
0.59
0.07
0.83
0.89
0.94
0.91
0.95
0.81
0.81
0.88
0.86
0.96
0.85
0.81
1.08
1.37
1.05
2.4
2.3
1.5
1.5
1.3
1.8
1.7
1.1
1.0
1.1
2.0
1.2
0.61
0.95
0.10
24
1
AN þ NaH2PO4
AN þ KH2PO4
AN þ Na2HPO4
AN þ K2HPO4
AN þ (NH4)2HPO4
AN þ CaHPO4
AN þ K3PO4
16
43
4
18
16
16
2
AN þ ureaa
24
0.20
0.17
0.28
0.30
0.16
0.21
0.06
0.17
0.18
AN þ guanidinum CO3 24
AN þ HMTA
24
20
24
24
0.14
0.01
0.03
0.03
0.09
AN þ K2C2O4
AN þ NH4CHO2
a
235
72
0.50
0.52
0.26
0.24
0.30
0.34
1.06
1.10
0.60
0.66
a
AN þ (NH4)2C2O4
a Most additives present at 10 mol%, but at 320 8C. Additives are at 20 wt.% and at 260 8C formate and oxalate are 20 wt.%.

34
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
At 320 8C, 0.97 mol gas were produced/mol neat AN;
at 260 8C, neat AN produced slightly less gas
(0.86 mol/mol). When inorganic species were added
the total gas did not vary by more than 10% from the
amount produced by neat AN. The nitrogen-contain-
ing organic species significantly contributed to the
amount of gas produced. The moles of gas/mol AN
increased as much as 50% with some of these additives
present. Except when calcium carbonate or one of the
high-nitrogen organics was added to AN, the primary
decomposition gas was nitrous oxide (N2O) with some
nitrogen (N2), the ratio being sensitive to the tempera-
ture of decomposition. At 320 8C, the final ratio N2O/
N2 was 4.1 for neat AN, while at 260 8C it was 2.3.
This trend was followed in most of the AN mixtures.
At 320 8C the N2O/N2 ratio ranged from 2 to 6, while
at 260 8C the range was 1–3 (Table 10). The excep-
tions were the formulations where AN was mixed with
20 wt.% calcium carbonate or some of the high-nitro-
gen organics; these produced more nitrogen gas than
AN. At 320 8C, gases were sampled at various stages
during the decomposition. The data indicate that
nitrous oxide is the first and main gas produced by
AN, and nitrogen is primarily formed later in the
decomposition. This was particularly true of the
high-nitrogen compounds, urea, guanidinum carbo-
nate and hexamethylenetetramine (HMTA). At
roughly 30% decomposition, the N2O/N2 ratio is three
to four times higher than it is at complete reaction,
presumably because at long reaction times these nitro-
gen-containing additives begin to decompose and
contribute nitrogen gas.
3.1.4. Effect of fuels on AN
Although AN can act as its own oxidizer and fuel,
for practical explosive use it is mixed with added fuel.
Historically, a variety have been used. Carbon black
was the first fuel to be used with AN; later liquid
hydrocarbons were employed [4]. To examine the
effect of various fuels on AN thermal stability, AN
and potassium nitrate (KNO3) were mixed with
various fuels and screened by DSC (Table 11). In
addition, DSC traces were obtained for AN with
varying amounts of fuel: diesel, mineral oil (MO),
and sugar (powdered or icing sugar) were examined
(Table 12). As the diesel fuel was increased from 2 to
10 wt.% (10 to 36 mol% CH2) and the sugar was
increased from 5 to 20 wt.% (12 to 35 mol% OCH2),
the total energy released increased from 1184 J/g
(neat AN) to more than 3138 J/g. With mineral oil
a similar increase in heat released was observed up
to the addition of 15 wt.% (46 mol% CH2) mineral
oil. However, at 30 wt.% (63 mol% based on CH2)
mineral oil, a substantially fuel-rich formulation, the
heat released begins to decrease. As previously obser-
ved, addition of liquid hydrocarbon split the DSC
Table 11
DSC exotherms for ammonium and potassium nitrates with fuelsa
5 wt.% additive
First exotherm
Second exotherm
Total
(cal/g)
Total
(J/g)
8C
cal/g
J/g
8C
cal/g
J/g
Pure AN
326
211
268
284
185
248
334
331
332
283
434
585
483
365
528
285
228
184
1184
1816
2448
2021
1527
2209
1192
954
Pure AN þ carbon
Pure AN þ mineral oil
Pure AN þ diesel
Pure AN þ sugar
303
221
1268
925
334
358
344
343
301
131
364
548
1523
112
469
253
1059
Pure AN þ sulfur
Pure AN þ nitrobenzene
Pure AN þ nitromethane
Pure AN þ aluminum
285
228
184
1192
954
770
770
Pure KNþ
Carbon
460
407
75
314
703
Mineral oil
Sulfur
221
331
315
6
47
25
197
858
KN þ 15% C þ 10% S
205
168
528
2209
a No exotherms were observed for diesel nor aluminum alone.

Table 12
DSC peaks for AN with various amounts of fuelsa
Endotherms
Sample name
Peak 1
Endotherm
Peak 2
Endotherm
Peak 3
Exotherm
Peak 4
Exotherm
Standard
No. Total % first
(J/g) exotherms
Tmin Standard
(8C)
E
Standard Tmin Standard E
Standard Tmax Standard E
Standard Tmax Standard E
(J/g)
(8C)
(J/g)
(8C)
(J/g)
(8C)
(J/g)
Neat AN
128
0
1
0
0
0
0
0
6
0
0
0
0
1
0
0
0
51
53
54
49
52
50
51
0
0
0
0
0
4
0
168
169
169
170
169
168
169
1
0
1
0
1
0
0
68
73
73
62
68
66
70
0
4
0
4
0
4
4
326
331
330
1
1
1
1182 149
1187 59
1541 64
900 25
3
3
3
3
3
3
2
2
3
3
3
3
3
3
3
3
1182
AN þ 2% diesel 129
AN þ 4% diesel 129
AN þ 5% diesel 129
AN þ 6% diesel 128
AN þ 8% diesel 128
AN þ 10% diesel 129
1187
1541
284 10
344
341
348
358
2
0
4
1
1121 84
1188
2020 45
2191 45
2750 53
3210 53
525
294
296
291
285
185
183
183
186
185
268
292
0
1
0
0
0
1
0
1
1
1
7
996
0
0
1464 209
1699 121
525 27
468 29
711 16
802 27
1340 50
1660 46
924 83
898 92
955 50
1284 75
1510 33
Sugar
174
112 17
AN þ 5% sugar 128
AN þ 7% sugar 130
AN þ 8% sugar 129
AN þ 15% sugar 130
AN þ 20% sugar 129
35
36
32
66
51
48
46
38
0
0
0
8
4
4
0
0
157
155
153
0
0
0
30
20
20
0
0
0
343
345
348
364
364
358
363
375
1
0
1
0
1
1
2
3
1057 27
1135 72
1257 72
1486 19
1486 38
1521 63
1877 99
1509 94
1525 31
1846 39
2059 39
2827 47
3146 53
2445 38
2774 32
2464 39
AN þ 5% MO
129
169
169
169
0
0
1
61
57
55
4
4
4
AN þ 15% MO 129
AN þ 30% MO 129
308 12
a Mineral oil and diesel showed no peaks in DSC between 50 and 500 8C.

36
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
exotherm into at least two peaks, one reaching max-
imum at higher and one at lower temperatures than
neat AN [13].
bases of weak acids and, as such, they are somewhat
basic. A mode of action for these anions is as follows:
NH4þ þ AÀ ! NH3 þ HA ðA ¼ any anionÞ
(7)
The carbonate anion is especially basic and in the
presence of acid forms carbonic acid, which is
unstable and decomposes to water and carbon dioxide
4. Discussion
À
gas. In contrast, HSO4 which is formed from the
4.1. Thermal stability
dissociation of a strong acid (sulfuric acid) is a weak
base. Rubtsov et al. [18,19] also observed that salts
that possessed anions more basic than nitrate caused a
marked reduction in the decomposition rate. Included
in Table 5 are pH values of solutions consisting of 5 ml
of water and 0.100 g of specified AN formulations (i.e.
20 wt.% additive). SolutÀions containing the most basic
anions (CO32À, HCO3 and HPO4À) yielded pH
values in the range 7.4–8.3. These species, when
added to AN at the 20 wt.% level, produced exother-
mic maxima 40–60 8C higher than neat AN. Those
With the exception of fluoride, the halides had a
destabilizing effect on the thermal stability of AN.
Chloride lowered the exothermic maximum the most
(about 70 8C). The destabilizing effect of chloride is
well known [16–31]. Nevertheless, chloride is some-
times added to fertilizer formulations and to permis-
sible AN explosives. Many researchers have noted
the synergism between HNO3 and ClÀ [18,19,23–25],
and several claimed acid is essential for chloride
catalysis, but the acid may come from the decom-
position of AN, itself [26,27]. The catalytic activity
of chloride is generally acknowledged to require an
induction period, which acidity obviates or alkalinity
lengthens indefinitely (presumably by suppressing
AN decomposition) [23,24]. The presence of chloride
is thought to generate acidity, and many have observed
the increase in nitrogen rather than nitrous oxide
decomposition gas formation [16,17,20,21,27]. There
appears to be no general agreement on the exact
mechanism by which chloride causes destabilization.
Keenan and coworkers suggested the key step to the
destabilization is
2À
additives producing pH values in the range 4–5 (SO4
and H2PO4À) only raised the maximum by 10–20 8C
when included with AN at the 20 wt.% level (Table 5).
The bisulfates (HSO42À) produced strongly acidic
aqueous solutions. Although they did not lower the
temperature of the AN exotherm when added at the
5 wt.% level, at the 20 wt.% level they did. The
trivalent phosphate anion, PO43À, though basic is
not stable; and gas evolution was observed when its
sodium salt was mixed with AN and an odor was
detected when high concentrations of the potassium
salt were added to AN.
Most of the anion additives were examined in the
presence of several different counter ions. Varying the
cation (Naþ, Kþ, NH4þ, and Ca2þ) did not greatly
affect the position of the exotherm nor the rate con-
stant. However, it has been noted previously that
ammonium ion is not entirely innocuous, being the
salt of a weak base [13], and some anomalies were
observed with the calcium salts. With most of the
additives, if the AN exotherm was shifted by 5 wt.%
additive, the effect was magnified when 20 wt.%
additive was used. However, this was not the case
with CaSO4 and CaHPO4. At 5 wt.% these calcium
salts increased the AN exotherm to about the same
degree as the other sulfates or hydrogen phosphates;
but addition of 20 wt.% showed no further increase
(Table 13). We interpreted this result to mean these
salts do not have a stabilizing effect on AN. The
NO2þ þ ClÀ ! NO2 þ Clþ
(6)
They noted that there must be an initial excess of acid
to allow formation of nitronium ion [22]. We have
observed that the relative destabilizing effect of the
halide, XÀ, is in line with the acidity of HX. The pKa
values of HCl, HBr and HI are À7.0, À9.5, and À10.0,
respectively [28]. The fluoride ion is a much stronger
base than the other halides (pKa of HF is 3.2) [28]
which probably explains its stabilizing influence on
AN.
Most of the oxo-anions, with the exception of
bisulfate, raised the temperature of the AN exotherm
(Table 5). These observations are in line with AN
thermal stability being affected by the acidity of the
3À
additive [18,19,29–31]. The anions SO42À, PO4
,
HPO42À, H2PO4À, CO32À, and HCO3 are conjugate
À

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
37
Table 13
Effect of calcium salts on the DSC exothermic maximum of AN
Additives to AN
CaNO3
CaCO3
CaSO4
CaHPO4
AN neat
326
DSC 5 wt.% additive (8C)
330
328
294
11.5
360
389
2.8
333
334
227
15.0
336
337
220
22.5
DSC 20 wt.% additive (8C)
Rate constants (260 8C) 106 k/s
Rate constants (320 8C) 103 k/s
195
1.59
10.9
isothermal rate constants at 260 8C confirm this
(Table 13). It may be that the slight stabilizing effect
of the oxo-anion is overcome by a destabilizing effect
from the cation. Calcium being a small cation with
high charge density may possess a measure of acidity,
such as we have previously described for metals with
high charge-to-radius ratios [32]. Such acidity would
be deleterious to the thermal stability of AN.
times or less than half that of neat AN. We observed
that if the DSC exothermic maximum of the AN/
additive mixture was at least 20 8C higher than neat
AN, then that mixture would exhibit a reduced rate
constant at 260 8C, but the relative stability ordering
among the various admixtures did not hold constant
between DSC and the 260 8C rate constants.
Examining the assignments in Table 14, it is
obvious that the isothermal data at 260 8C are more
comparable to the DSC data than those obtained at
320 8C. This is surprising since the DSC exotherms of
AN and all the mixtures occurred above 320 8C, but it
is in line with our previous observation that the
composition of gaseous products produced by the
decomposition of AN during a DSC scan were closer
to the composition of AN decomposed at 270 8C
than at 320 8C [13]. Furthermore, the additives which
stabilize AN at 260 8C have much less of an effect at
320 8C (Table 14). This is reasonable since the decom-
position of AN is more susceptible to acids and bases
at low temperatures where an ionic decomposition
mechanism is dominant than at high temperature
(above 290 8C) where a free-radical decomposition
pathway becomes important. Thus, species which
retarded AN decomposition at 260 8C were not as
effective at 320 8C (e.g. HMTD).
Ca2þ þ 2H2O ! CaðOHÞ2 þ 2Hþ
(8)
However, with this interpretation the high thermal
stability imparted to AN by calcium carbonate is
difficult to explain. It is possible that the acidic nature
of the calcium cation is counteracted by the extremely
basic nature of the carbonate anion.
The oxalate, formate, and acetate salts, added to
AN at the 10 mol% level, substantially raised the pH
(5.9–6.3 pH units) and the temperature of the DSC
exotherm. Unfortunately, both the ammonium and
potassium acetate and ammonium formate increase
the hygroscopicity of what was already a hygroscopic
material. Oxlate salts significantly increase the stabi-
lity of AN, but there is some concern about its toxicity
[33]. HMTA and guanidinum carbonate also stabilize
AN, but the stabilizing effect of HMTD appears to
disappear at 320 8C.
Trends suggested by DSC were confirmed by iso-
thermal studies (Tables 7 and 8). Table 14 compares
the relative stabilities of the various AN mixtures to
AN as determined by the positions of the DSC
exothermic maxima and by the isothermal rate con-
stants at 260 and 320 8C. For DSC the results were
assigned as ‘‘ND’’ (decomposition not much different
from neat AN) if the exothermic maximum was within
20 8C of that of AN; ‘‘S’’ (decomposition slower than
neat AN) if the exotherm was more than 20 8C above
that of AN. Isothermal rate constants are given a
similar subjective rating, with the rating ‘‘ND’’ (not
different from AN) if the rate were not more than two
The additives which best stabilize AN are the
carbonates and the organic bases. Comparing the
260 8C isothermal rate constants of the carbonate
salts, calcium carbonate clearly stabilizes AN better
than the sodium and potassium salts. This is especially
surprising in light of the destabilization apparent with
other calcium salts. In an attempt to understand the
effect of the calcium carbonate, we examined the
effect of added magnesium carbonate using DSC
and 260 8C thermolysis. The DSC exotherm of AN
mixed with magnesium carbonate was similar to the
other carbonates (Table 14). Its 260 8C isothermal rate
constant indicates more stabilization than the alkali

38
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
Table 14
Ammonium nitrate mixtures: relative decomposition rates as predicted by the DSC exotherm or by the isothermal rate constants (additives
about 10 mol%)a
DSC
260 8C 104 k/s
320 8C 102 k/s
pH
N2O/N2 CO2
Ammonium nitrate
326
328
323
328
349
348
334
339
348
366
364
364
337
362
381
374
389
378
372
408
384
385
383
394
377
2.0
1.4
1.0
0.78
0.74
0.68
0.9
2.0
1.5
1.6
1.3
0.9
0.8
2.1
1.9
1.0
0.4
0.5
0.2
4.9
4.7
4.8
4.9
5.1
4.8
5.6
4.3
4.2
7.3
7.4
7.4
6.4
8.2
8.5
8.5
7.4
7.9
7.3
8.2
5.9
5.9
4.1
6.1
3.4
4.9
3.4
3.1
3.4
4.5
2.5
2.4
2.7
3.7
5.2
2.3
1.9
2.4
0.8
1.9
3.3
1.5
1.0
1.0
0.70
0.30
0.30
0
Ammonium nitrate þ NaNO3
Ammonium nitrate þ KNO3
Ammonium nitrate þ Ca(NO3)2
Ammonium nitrate þ Na2SO4
Ammonium nitrate þ (NH4)2SO4
Ammonium nitrate þ CaSO4
Ammonium nitrate þ NaH2PO4
Ammonium nitrate þ KH2PO4
Ammonium nitrate þ Na2HPO4
Ammonium nitrate þ K2HPO4
Ammonium nitrate þ (NH4)2HPO4
Ammonium nitrate þ CaHPO4
Ammonium nitrate þ K3PO4
Ammonium nitrate þ Na2CO3
Ammonium nitrate þ K2CO3
Ammonium nitrate þ CaCO3
Ammonium nitrate þ MgCO3
Ammonium nitrate þ KHCO3
Ammonium nitrate þ guanidinum CO3
Ammonium nitrate þ NH4CHO2
Ammonium nitrate þ (NH4)2C2O4
Ammonium nitrate þ K2C2O4
Urea
ND
ND
ND
ND
ND
ND
ND
ND
S
ND
ND
ND
S
ND
0
2.2
ND
0
2.9
ND
0
1.0
1.6
ND
F
0
ND
ND
ND
ND
S
0
2.3
ND
0
1.4
1.4
ND
ND
0
0
0.5
ND
0
S
S
0.5
ND
0
S
S
0.9
2.2
F
0
ND
S
ND
S
F
0
0.7
ND
0
S
S
1.0
S
0.10
0.05
0.10
0.07
0.02
0.17
0.16
0.21
0.30
0.20
0.28
S
S
1.0
S
S
S
0.03
0.17
1.4
S
S
S
Not run
S
ND
ND
S
S
0.7
0.8
0.7
0.7
S
1.3
S
S
0.11
0.01
0.01
0.09
0.28
S
S
S
S
S
S
Not run
S
S
S
F
0.7
2.2
5
HMTA
S
S
6.5
a ND: no great difference; S: slower; F: faster.
metal salts, but somewhat less than that of the calcium
salt. Possibly it is less stabilizing than the calcium salt
because as a smaller cation it is more acidic, but this
does not explain the reason either calcium or magne-
sium carbonate should be more stable than the carbo-
nates of sodium or potassium. (One reviewer has
suggested their lower solubility in the AN melt may
explain their enhanced thermal stability compared to
the alkali carbonate.)
The decomposition gases are also shifted from N2O
toward N2 in the presence of hydrocarbons [13].
Table 14 suggests there is a general trend that the
more stabilizing the additive, the lower the N2O/N2
ratio and the higher the production of CO2. This
appears to be a better predictor of an additives stabi-
lizing ability than pH (note urea). The additives that
fall in this category are the carbonates and the organic
additives. Given the composition of these, it was
expected that they would produce carbon dioxide;
and in producing carbon dioxide they consumed the
oxygen normally used to produce nitrous oxide. While
the organic species are oxidizing by nitric acid to
produce carbon dioxide and nitrous acid, the carbo-
nates react with the acid proton of nitric acid (or of
ammonium, Eq. (2) to form carbonic acid and subse-
quently carbon dioxide and water. Either event results
in the retarding of AN decomposition. Although we
could not detect ammonia by our chromatographic
method, urea [34] and possibly guanidinum ion [35]
4.2. Gaseous decomposition products
Table 14 compiles the data on the various AN
mixtures—the temperature of the DSC exotherm,
the rate constants, pH, and gaseous decomposition
products at 260 8C. We have previously noted that at
lower temperatures, AN produces more nitrogen than
it does at higher temperatures [12]
5NH3 þ 3HNO3 ! 4N2 þ 9H2O
(9)

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
39
and HMTD produce ammonia during their decompo-
sition; this gas would not have been detected. The
formation of ammonia from the decomposition of urea
would explain the fact that AN/urea decomposition at
260 8C slowed dramatically after about 12% decom-
position. Calcium carbonate was the most stabilizing
of the carbonates, and its N2O/N2 ratio is the lowest
among the carbonates. The shift in decomposition
gases is rather large considering the additive only
comprised 10 mol% of the AN formulation. If this
unique stabilizing effect of calcium carbonate is con-
trasted with the unique destabilizing effect exhibited
by the other calcium salts Eq. (9), the apparent dif-
ference is that the decomposition of calcium carbonate
forms the basic oxide CaO.
Some of the fuels reacted with AN in the temperature
range of the DSC scan (carbon, mineral oil, diesel,
sugar, and sulfur) and some did not (NB, nitromethane
(NM) and aluminum). The thermal reaction of AN
with the organic fuel is based on the interaction of
nitric acid, formed by reaction (1), with the fuel
1
2
HNO3 þ ‘‘CH’’ ! CO2 þ H2O þ N2
(10)
The presence of carbon black in the early reaction
category and of NB and NM in the later reaction
category prevents speculation that liquid additives
facilitate early decomposition. The DSC results
appear to depend on the relative decomposition tem-
peratures of the oxidizer and the fuel. Of the fuels
examined with AN (Table 11), sugar decomposed at
the lowest temperature. Clearly, with AN/fuel mix-
tures, the relative position of the exothermic maxi-
mum is not an indication of explosivity. AN and sugar
can be made detonable, and AN with NM (Kinepak1)
is a commercial high explosive more powerful than
4.3. Effect of fuels on AN
Table 11 summarizes the DSC results for AN/fuel
mixtures, most of which have been used as explosives.
Fig. 2. DSC of AN with mineral oil.

40
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
ANFO. Typically, ANFO and AN/sugar have detona-
tion velocities around 3.5–4.0 km/s, but mixed with an
energetic liquid such as NM or hydrazine (Astrolite1
G) detonation velocity of up to 8.6 km/s are reported
[36].
the fuel, the relative ratio of the low and high tem-
perature exotherms remained constant as the amount
of fuel increased (Fig. 3). AN mixed with sugar
gave a thermogram with two distinct exotherms
(185 and 343 8C) (Fig. 4). The latter moved to higher
temperature (364 8C) as the level of sugar was
increased to 20 wt.%, but the former remains at
185 8C and becomes a larger percentage of the total
heat released. Both exothermic regions increased
markedly as the amount of sugar was increased. Sugar,
like NM and NB, released heat upon decomposi-
tion (exothermic maximum, 285 8C and heat, 523 J/
g); but it was not nearly as energetic as NM and
other species added to AN (Table 12). (Surprisingly,
addition of 5 wt.% NB did not increase the amount
of heat released (1192 J/g) nor substantially change
the exotherm of AN. Apparently, it was too stable
to react in the DSC temperature range.) Sugar, in the
The fuels which reacted with AN generally exhib-
ited at least two exothermic regions. Addition of the
fuel increased the overall heat release of the reaction,
but the majority of heat was released in the higher
temperature exotherm. In a detailed study of the
effect of mineral oil on AN, we determined that the
first exothermic region was associated with disso-
ciation of AN Eq. (1) and nitric acid attack of the
fuel Eq. (10). The second exothermic region we
interpreted as reaction of ammonia and complete
oxidation of the fuel [13]. As the amount of mineral
oil was increased, the first exotherm decreased, while
the second one increased (Fig. 2). When diesel was
Fig. 3. DSC thermogram of AN with increasing amounts of diesel fuel.

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
41
Fig. 4. DSC thermogram of AN with increasing amounts of sugar.
presence of AN, released heat at lower tempera-
4.4. Explosivity testing
tures than either species did neat. Raising the percen-
tage of sugar in the AN mixture from 5 to 20 wt.%
increased the heat released in the first exotherm
from 31 to 53% of the total. The first exotherm for
the AN/sugar mixture immediately followed their
endothermic melt (AN, 169 8C; sugar, 174 8C). These
observations suggest a low temperature AN/sugar
reaction.
To further shed light on the role of fuel, KNO3 was
examined. KNO3, not possessing the source of hydro-
gen provided by AN, did not decompose nor react with
fuels as readily as AN. It was so stable that it did not
exhibit an exotherm in the range of the DSC (up to
500 8C); and only with a few fuels (carbon, sulfur, and
their mixture in black powder) was an exotherm
observed below 500 8C. In some cases, the amount
of heat released was so small that they may be baseline
deviations rather than exotherms. As with AN/fuel
mixtures, we must conclude the position of the
exotherm is not indicative of explosivity.
It is desirable to conduct explosive testing on a
large-scale (100–1000 lb), especially for such a poor
explosive as AN. However, laboratory-scale testing is
convenient and large-scale may not be possible due to
lack of funds or materials. Correlation of small- and
large-scale test results is a frequently sought and
seldom achieved goal. We have experimented using
the small-scale explosivity device (SSED) [37–41] to
obtain a relative ranking of the explosivity of energetic
materials and have been successful in comparing
SSED results (2 g scale) with pipe bomb fragmenta-
tion data (800 g scale). The SSED attempts to detonate
2 g samples of energetic materials confined in a 0.303
(British) brass cartridge case. The entire set-up was
contained in a heavy-walled, bolted-closure, stainless
steel vessel. Violence of an event was judged by the
weight of the main body of the casing remaining
attached to the base after detonation. Obviously, set-
ting off a detonator (no. 8), even in an empty cartridge,

42
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
Table 15
Explosivity measure on a small-scale explosivity device (% case shattered)a
Test no. No.
average
Tested material
% shattered AN in
DSC exothermic DSC heat
released (J/g)
mixture Tmax (8C)
1
2
3
NaCl
HMX
40
80
56
58
56
65
53
64
61
43
47
36
39
37
39
34
41
32
30
42
36
32
3
2
3
3
3
3
3
3
3
1
3
4
1
3
3
3
3
3
3
3
3
3
AN
100
95
95
90
79
76
69
90
90
67
67
67
50
50
50
30
30
10
10
10
328
1255
4
AN þ 5% mineral oil
268, 358
268, 358
296, 348
393
2448
2448
2749
1439
5
AN þ 6% fuel oil
6
90% AN þ 10% sugar
7
79% AN þ 21% Ca/MgCO3
96.5% (79% AN þ 21% Ca/MgCO3) þ 3.5% mineral oil
87.7% (79% AN þ 21% Ca/MgCO3) þ 12.3% sugar
9 AN þ 1 NaNO3
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
328
383
433
341
333
971
1289
397
AN þ 10 mol% K2C2O4
2 AN þ 1 CaCO3
2 AN þ 1 CaHPO4
682
2 AN þ 1 CaSO4
669
1 AN þ 1 NaNO3
1 AN þ 1 Al2O3
1 AN þ 1 ASO4
364
640
360
30 AN þ 10 ASO4 þ 30 CaCO3 þ 30 urea (mol%)
30% AN þ 10% ASO4 þ 30% CaCO3 þ 30% urea
1 AN þ 9 NaNO3
357, 388, 463
10% AN þ 30% urea þ 30% CaCO3 þ 30% ASO4
10% AN þ 30% urea þ 30% A2HPO4 þ 30% CaCO3
In situ formulations
303
No exotherm
A
C
B
3
4
3
2 AN þ 1 CaCO3
67
67
67
440
373
355
979
665
561
2 AN þ 1 CaHPO4
2 AN þ 1 CaSO4
a A: ammonium; AN: ammonium nitrate; all formulations 2.00 g and cartridges 11.0 g.
will result in recovery of less than 100% of the
cartridge. In tests where the cartridge was filled with
2 g of inert NaCl ꢀ60% of the cartridge remained
attached to the base, whereas with HMX, only 20% of
the original case weight was recovered. Since only
60% of the casing remained attached to the base when
the test material was an inert substance, the differ-
entiation between a good explosive and a weak one is
difficult. One solution might be to use smaller deto-
nators. Reducing the amount of explosive in the
detonator would reduce the damage to the cartridge/
inert system, expanding the low-reactivity end of the
test. However, decreasing the power of the detonator
would increase the likelihood that insensitive explo-
sives like AN-based materials would not propagate the
detonation wave. For this reason no. 8 detonators were
used.
Those with fuel (mixtures 4–9, Table 15) left only 35–
47% of the cartridge attached to base, while those
without fuel retained greater than 50% of the car-
tridge. Among the non-fuel containing formulations,
results suggest that the degree of destruction directly
correlated with the amount of AN in the formulation.
However, this may reflect a dilution effect. Low ther-
mal stability appears to be related to high explosivity,
but high thermal stability is no guarantee of low
explosivity. Insufficient tests have been run to make
firm correlations, and no tests have been run to
correlate these SSED results to large-scale explosive
testing. Neither of the mixtures discussed earlier as
relatively non-detonable—AN with 21% dolomite nor
a 50/50 mixture of AN and ammonium sulfate—were
markedly less violent; this is backed by field experi-
ence showing them to be detonable. However, the test
does indicate both formulations are less explosive than
AN. Thus, at 1/3 dilution there are additives which
The differentiation in SSED destruction among AN
formulations with fuel and those without was clear.

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
43
significantly stabilize AN. Whether nitrogen content
of 17–22% is sufficiently high to make these viable
fertilizer mixtures is not known. In Switzerland and
Ireland the only AN sold has been diluted to 27%
nitrogen [42].
acidic media Eq. (2). A number of additives were
identified which dramatically raised the temperature
of the AN DSC exotherm. In general, an increase in
the DSC exothermic peak maximum signaled a
decrease in the rate of AN decomposition, but the
increase in the exotherm had to be greater than 20 8C
to have notable effect on the decomposition rate.
Ordering of thermal stability by the position of the
DSC exotherm was similar but not exactly the same as
an ordering in terms of 260 8C first-order rate con-
stants. Materials that dramatically raised the tempera-
ture of the AN exotherm, thus, enhancing AN stability,
were basic in nature: HMTA, and salts of carbonate,
formate, oxalate, and mono-phosphate. However,
basicity is not the only factor associated with stabiliz-
ing AN. Urea had a large stabilizing effect but did
not raise the pH of an AN solution dramatically; rather
on decomposition urea forms ammonia, thus generat-
ing a basic media. For the most part organic species
5. Conclusions
Both DSC and isothermal results have shown a
general and strong correlation between the basicity
of an additive and its ability to stabilize AN to thermal
decomposition. Interestingly, DSC scans of AN more
closely reflect isothermal studies at 260 8C than those
at 320 8C. In fact, additives had more effect on the low
temperature decomposition of AN than they did on its
high temperature decomposition. This is in line with
the accepted low temperature route of AN decomposi-
tion which involves formation of nitronium ion from
Table 16
Survey of best additives
Formula Additive DSC
(mol%)
Rate constant wt.% N
260 8C,
pH
4.9
Cartridge
shattered
Dec. gas 320 8C
exotherm
Tmax (8C)
106 k/s
N2O/N2 CO2
Ammonium nitrate
NH4NO3 100
326
200
35
4.1
0
AN þ single additive
CaSO4
In situ
CaSO4
11
33
11
33
8
334
355
337
373
364
351
378
389
440
408
368
377
384
383
385
230
220
29
19
29
19
29
29
35
29
22
37
36
36
33
28
34
5.0
7.4
37
35
39
38
3.4
5.2
0
0
CaHPO4
CaHPO4
In situ
K2HPO4
K2CO3
K2HPO4
K2CO3
MgCO3
CaCO3
49
100
17
7.4
8.5
7.9
7.4
2.7
2.4
1.9
0.8
0
10
10
14
33
0.05
0.07
0.10
MgCO3
CaCO3
2.8
36
In situ
Guanidinum CO3
Urea
C3O3N6H12 10
130
9.2
8.2
5.0
6.5
5.9
6.4
5.9
1.5
0.3
0.3
1.0
0.7
1.0
0.17
0.20
0.28
0.16
0.30
0.21
CON2H4
C6N4H12
CO2NH5
C2O4K2
10
10
20
10
8
HMTA
29
11
NH4 formate
K oxalate
NH4 oxalate
9.9
9.9
47
32
C2O4N2H8
AN þ multiple additives (wt.%)
30% AN/30% urea/30%
CaCO3/10% (NH4)2SO4
357
394
404
27
33
37
48% AN/20% urea/16%
A2HPO4/16% A2SO4
68% AN/36% urea/16% A2HPO4

44
J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
enhanced AN stability more than the inorganics. This
cannot be a matter of pH alone since the pH of AN
solutions were similar for both the inorganic and
organic additives. Given that the only outstanding
stabilizers contain carbon, calcium carbonate being
dramatically better than most other inorganic salts, we
speculate that the formation of carbon dioxide may be
a factor in the ability of a compound to stabilize AN.
Table 16 shows characteristics observed for the more
thermally stable AN mixtures. There appears to be a
correlation between a low 260 8C rate constant and a
low N2O/N2 ratio or high carbon dioxide content.
Calcium carbonate in admixture with AN showed
the largest retardation of the 260 8C first-order rate
constant. This AN mixture is used successfully as a
fertilizer. The organic bases were effective in stabiliz-
ing AN, and species could be chosen that contain
nitrogen; however, use of them in fertilizers may be
difficult. For the most part, they are more expensive
than the inorganic additives; several (urea and the salts
of formate and acetate) dramatically increase the
hygroscopicity of AN; the oxalates are toxic to animals;
and HMTA is the chemical precursor of RDX. Of the
group considered urea is currently used. It is inexpen-
sive, but the hygroscopicity of the urea/AN mixture
requires it to be sold in solution. Three other bases
targeted in this study require testing on a large-scale;
these are oxalate, HMTA, and guanidinum carbonate.
In situ, formationofAN byproviding the ions that make
up AN from two different additives (i.e. NH4Xþ ENO3
instead of NH4NO3 þ EX) in some cases resulted in
enhanced thermal stability. This appears to be the case
when the decomposition point of AN plus the additive
is below the melting points of the in situ ingredients.
This approach deserves further study.
References
[1] K. Othmer, Encyclopedia of Chemical Technology, 3rd
Edition, Wiley, New York, 1978.
[2] Study of Imposing Controls on, or Rendering Inert, Fertilizer
Chemicals Used to Manufacture Explosive Materials, Final
Main Report B005, submitted by The International Fertilizer
Development Center (IFDC) to Bureau of Alcohol Tobacco
and Firearms (ATF) under contract Tatf 97-2, 28 March
1997.
[3] Explosives US Department of Interior, US Geological Survey,
Mineral Industry Surveys, Annual Review, 1996.
[4] J.C. Oxley, The Chemistry of Explosives in Explosive Effects
and Applications, in: J.A. Zukas, W.P. Walters (Eds.), Springer,
New York, 1998, p. 159.
[5] R.B. Hopler, Custom-designed explosives for surface and
underground coal mining, Mining Eng. (1993) 1248–1252.
[6] F.L. Fire, ANFO: the tool of destruction fire engineering
1995, in: F.L. Fire (Ed.), AN Fire Eng. (1986).
[7] R.M. Hainer, The Application of Kinetics to the Hazardous
Behavior of AN Symposium Combustion, 5th Edition,
Pittsburgh, 1954, pp. 224–230.
[8] W.G. Sykes, R.H. Johnson, R.M. Hanier, AN explosion
hazard, Chem. Eng. Prog. 59/1 (1963) 66–71.
[9] G.S. Biasutti, History of Accidents in the Explosives Industry,
1980.
[10] M.A. Cook, The science of high explosives, in: M.A. Cook
(Ed.), The Science of Industrial Explosives Graphics Service
and Supply for IRECO Chemicals, Robert E. Krieger
Publishing Co. Inc., Huntington, NY, 1971, p. 13.
[11] B. Foulger, P.J. Hubbard, A review of techniques examined
by UK authorities to prevent or inhibit the illegal use ferti-
lizer in terrorist devices, in: Proceedings of the Interna-
tional Explosives Symposium, Fairfax, Virginia, 1995, p.
129.
[12] K.R. Brower, J.C. Oxley, M.P. Tewari, Homolytic decom-
position of AN at high temperature, J. Phys. Chem. 93 (1989)
4029–4033.
[13] J.C. Oxley, S.M. Kaushik, N.S. Gilson, Thermal decomposi-
tion of AN-based composites, Thermochem. Acta 153 (1989)
269–286.
[14] L. Whiting, M. LaBean, S. Eadie, Evaluation of a capillary
tube sample container for differential scanning calorimetry,
Thermochem. Acta 136 (1988) 231–245.
While this study has successfully identified several
additives which stabilize AN and might be acceptable
in fertilizer formulations, these formulations have not
been shown to be non-detonatable. The SSED clearly
differentiated between fueled and unfueled AN for-
mulations in terms of explosivity, and the device
detected the decrease in detonability accompanying
dilution of AN from 100 to 90, 67, 50, 30, and 10%.
However, within a given dilution of AN there was little
range to distinguish whether a stabilizing additive can
also reduce detonability. At this point, large-scale
testing of some of the best formulations is necessary
to determine failure diameter.
[15] E. Kestila, M.E. Harju, J. Valkonen, Differential scanning
calorimetric and Raman studies of phase transitions V $ IV
of AN, Thermochem. Acta 214 (1) (1993) 67–70.
[16] A.G. Keenan, Differential thermal analysis of the thermal
decomposition of AN, J. Am. Chem. Soc. 77 (1955) 1379.
[17] A.G. Keenan, K. Notz, N.B. Franco, J. Am. Chem. Soc. 91
(1969) 3168.
[18] Y.I. Rubtsov, I.I. Strizhevskii, A.I. Kazakov, Kinetics of the
influence of ClÀ on thermal decomposition of AN, Zhurnal
Prikladnoi Khimii 62 (11) (1989) 2417–2422.
[19] Y.I. Rubtsov, I.I. Strizhevskii, A.I. Kazakov, Possibilities of
lowering the rate of thermal decomposition of AN, Zhurnal
Prikladnoi Khimii 62 (1989) 2169–2174.

J.C. Oxley et al. / Thermochimica Acta 384 (2002) 23–45
45
[20] A.G. Keenan, B. Dimitriades, Mechanism for the chloride-
catalyzed thermal decomposition of AN, J. Chem. Phys. 37/8
(1962) 1583.
[33] S. Budavair (Ed.), The Merck Index, 12th Edition, Merck
Research Lab, Whitehouse Station, NJ, 1996.
[34] B.Y. Rozman, L.I. Borodkina, Inhibition of the thermal
decomposition of AN, J. Appl. Chem. USSR 32 (1959) 291.
[35] J.W. Schoppelrei, M.L. Kieke, X. Wang, M.T. Klein, T.B.
Brill, J. Chem. Phys. 100 (1996) 14343–14351.
[36] Two Component High Explosive Mixtures, Desert Publica-
tions, El Dorado, AR, 1982.
[21] G. Guiochon, Annales de Chimie, Masson & Cie, Librairies
de Lacademe de Medecine, in: L.A. Medard (Ed.), Ammo-
nium Nitrate and its Thermal Decomposition in Accidental
Explosions, Vol. 2, Types of Explosive Substances, Wiley,
New York, 1987 (Chapter 23).
[22] C.I. Colvin, P.W. Fearnow, A.G. Keenan, The induction
period of the chloride-catalyzed decomposition of AN, J.
Inorg. Chem. 4 (1965) 173–176.
[37] D.P. Lidstone, The Assessment of Explosive Performance by
Means of a Cartridge Case Deformation Test, RARDE
Memorandum 15/69.
[23] L.A. Medard, AN and its Thermal Decomposition in
Accidental Explosions, Vol. 2, Types of Explosive Sub-
stances, Wiley, New York, 1989 (Chapter 23).
[38] A.M. Lowe, Cartridge Case Test Validation DRA Report,
1995.
[39] T.R. Gibbs, A. Popolato, U. Calif (Eds.), LASL Explosive
Property Data, LA, 1980, p. 235.
[24] C.H. Solomon, K.S. Barclay, AN: Manufacture and Use
Presented at the Fertilizer Society, London, 1965.
[25] H.L. Saunders, J. Chem. Soc. 121 (1922) 698.
[26] K.D. Shah, A.G. Roberts, Properties of AN, in: C. Keleti
(Ed.), Nitric Acid and Fertilizer Nitrates, Marcel Dekker,
New York, 1985 (Chapter 11).
[40] J.C. Oxley, J.L. Smith, E. Rogers, M.A. Hamad, E. Resende,
Small-scale explosivity testing, J. Energetic Mater. 17 (4)
(1999) 331–343.
[41] J.C. Oxley, J.L. Smith, E. Resende, Determining Explosivity,
Part II, Comparison of Small-Scale Cartridge Test to Actual
Pipe Bombs, J. Forensic Sci.
[27] H. Tramm, H. Velde, On the spontaneous decomposition of
AN melts, Angew. Chem. 47 (24) (1934) 782–783.
[28] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry,
3rd Edition, Wiley, New York, 1972, p. 169.
[42] Marking, Rendering Inert, and Licensing of Explosive
Materials by National Academy Sciences Board on Chemical
Sciences and Technology, National Academy Press, Wa-
shington, DC.
[29] B.J. Wood, H. Wise, Acid catalysis in the thermal decom-
position of AN, J. Chem. Phys. 23/24 (1955) 693–696.
[30] W.A. Rosser, S.H. Inami, H. Wise, The kinetics of
decomposition of liquid AN, Trans. Faraday Soc. 67 (1963)
1753–1757.
[43] A. Bauer, A. King, R. Heater, A Review of Accidents with
Ammonium Nitrate Report to Canadian Fertilizer Institute
(CFI), 1978.
[44] M.L. Greiner, Ammonium nitrate fertilizer—exploding the
myth, Ammonia Plant Safety 25 (1986) 1–9.
[31] K.S. Barclay, J.M. Crewe, The thermal decomposition of AN
in fused salt solution and in the presence of added salts, J.
Appl. Chem. 17 (1967) 21.
[45] G. Perbal, The Thermal Stability of Fertilizers Containing
Ammonium Nitrate, The Fertilizer Society, 1971, pp. 3–63.
[46] W.G. Sykes, R.H. Johnson, R.M. Hainer, Ammonium nitrate
explosion hazards, Chem. Eng. Prog. 59 (1) (1963) 66–71.
[47] DoD Explosive Safety Board Electronic Files.
[32] J.C. Oxley, S.M. Kaushik, N.S. Gilson, AN explosives:
thermal stability and compatibility on small- and large-scale,
Thermochem. Acta 212 (1992) 77–85.