Journal of Environmental Chemical Engineering 6 (2018) 281–288

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

journal homepage: www.elsevier.com/locate/jece

The mechanism of the spontaneous detonation of ammonium nitrate in T

reactive grounds
⁎
Alex M. Djerdjeva, , Pramith Priyanandaa, Jeff Goreb, James K. Beattiea, Chiara Netoa,
Brian S. Hawketta
a
School of Chemistry, The University of Sydney, NSW, 2006, Australia
b
Dyno Nobel Asia Pacific Pty Ltd, Mt Thorley, New South Wales, 2330, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: The mechanism of the decomposition of ammonium nitrate fuel oil (ANFO) and of ammonium nitrate (AN)
Ammonium nitrate emulsions has been investigated. The formation of NO gas bubbles at the end of an induction period was ob-
Reactive ground served, which led to the evolution of brown NO2 gas on contact with air. Inhibition of the decomposition
NOx reaction could be achieved by reducing the NO concentration in the gas phase, with inhibitors not in physical
Hydrotalcite
contact with the reaction mixture. NO adsorbing clay minerals (hydrotalcite) were effective inhibitors, in-
Nitrous acid
dicating multiple mechanisms of inhibition. A central role is assigned to nitrous acid (HNO2): sequestering NO
prevents its formation, base inhibitors keep it in the non-reactive form of the nitrite ion (pKa of HNO2 ∼ 3), and
urea in acidic conditions decomposes it to harmless N2. Our results have implications in industries where NOx
emissions can cause significant environmental problems for both aquatic and marine life as well as for human
health.

1. Introduction 1.1. Chemistry background

Ammonium nitrate, or AN (NH4NO3), is the basis of most explosives
used in the mining industry. Its detonation usually must be deliberately
initiated, but there have been a number of instances when it has
spontaneously exploded, at great risk to life and property. One of the
causes has been identified as the presence of pyrite (FeIIS2) in contact
with the explosive. Another is the presence of naturally high tempera-
tures in the ore body, from some geothermal source. These causes are
termed ‘reactive ground’ and ‘hot ground’, respectively [1].
The explosives industry has developed emulsion explosives due to
their superior water-resistance, and low environmental impact [2,3]
over traditional explosives. However, when sulfide ores are present, or
in hot grounds, inhibited emulsion explosives must be used to prevent
premature detonation of the explosive. The explosives industry achieve
this by the addition of urea, to extend the ‘sleep time’ of the product-
the quiescent time between contact with the reactive ground and the
exothermic process that leads to detonation [4,5]. The role of urea as an
inhibitor was the original subject of this research. The work led to a
new, alternative understanding of the mechanism of the spontaneous,
and of the inhibited, decomposition reactions.
Ammonium nitrate decomposes in an exothermic reaction
(ΔH = –36 kJ mol−1) to produce three moles of gaseous products for
each mole of solid reactant:
NH 4 NO3(s) → N2 O(g ) + 2H2 O(g) (1)
−1
The reaction can be made more exothermic (ΔH = –280 kJ mol ),
with more gaseous products, if some oxidisable fuel is added:
2NH 4 NO3(s) + C→ 2N2(g ) + 4H2 O(g) + CO2(g ) (2)
Hence, the standard ammonium nitrate explosive mixture is termed
ANFO, for an “ammonium nitrate fuel oil” mixture.
The decomposition temperature of pure ammonium nitrate is
170 °C, but recently Gunawan and Zhang [6] calculated that an in-
timate mixture of ammonium nitrate and pyrite can decompose at
temperatures as low as 50 °C in blast holes more than 0.2 m in dia-
meter. This is consistent with many field observations of detonations at
low ambient temperatures. We have analysed the heat flow in these
systems elsewhere [7].
The same initial reactions occur in acid mine drainage, which has

⁎
Corresponding author.
E-mail address: alexdjerdjev@gmail.com (A.M. Djerdjev).

https://doi.org/10.1016/j.jece.2017.12.003
Received 1 September 2017; Received in revised form 14 November 2017; Accepted 2 December 2017
Available online 05 December 2017
2213-3437/ © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).
A.M. Djerdjev et al.
been extensively studied [8,9]. Parallels can be made between the two
processes and analogies usefully drawn. Water is required in both cases,
implying that soluble species are involved. The first step in the process
is the oxidation of pyrite by air. The oxidation product of the sulfur
could be various substances, such as SO2, SO3, thiosulfate, etc. For il-
lustrative purposes we choose SO2, because it is detected as a product in
reactive ground environments; however, this choice does not affect the
conclusions of the argument. For example, oxygen from the air oxidizes
the disulfide anion to SO2:
2FeS2 + 5O2 + 4H+ → 2Fe2 + + 4SO2 + 2H2 O
The Fe(II) is further oxidised to Fe(III), which precipitates as the
insoluble hydroxide in near neutral pH solutions.
2Fe2 + + 5H2 O+ 1 2 O2 → 2Fe(OH)3 + 4H+
The sum of these two reactions neither consumes nor produces
protons,
2FeS2 + 5 1 2 O2 + 3H2 O→ 2Fe(OH)3 + 4SO2
but the SO2 is readily soluble in water to produce sulfurous acid,
with pKa1 of 2.
SO2 + H2 O→ H+ + HSO−3
Both of the oxidation reactions above are relatively slow, but, as
they proceed and the acidity increases, some of the Fe(OH)3 begins to
dissolve. It turns out that the oxidation of pyrite by Fe(III) is a lot faster
than by oxygen [8,10,11].
FeS2 + 10Fe3 + + 4H2 O→ 11Fe2 + + 2SO2 + 8H+
The process now becomes autocatalytic, as more acid is produced
and more Fe(III) dissolves. The rate-limiting step in this inorganic cycle
then becomes the oxidation of Fe(II) to Fe(III) by oxygen, but in the
field this is accomplished rapidly by bacteria. In mine sites where
bacteria are present, pH values can range from 0.7–3.0 [12] and ferric
(Fe(III)) concentrations from 1 to 20 g/L [13,14]. Such conditions have
been shown to lower the thermal decomposition of ammonium nitrate
[15,16].
The thermal profile of the decomposition process comprises three
stages: an induction period, an intermediate stage and the final highly
exothermic decomposition. (Fig. 1) The reactions described above
could explain the observation of the induction period in the thermal
decomposition of ammonium nitrate explosives caused by reactive
ground. Some preliminary studies have indicated an inverse correlation
between initial acidity and the induction time. According to Rumball,
acid accelerates the rate of the initial stage and has little or no effect on
the intermediate stage [17]. The initial stage of the process is
Fig. 1. Typical temperature versus time trace for reaction between pyritic black shale and
AN. A-B is the initial stage (the induction stage or Stage 1), B-C is the intermediate stage,
or Stage II. Ignition stage starts at C (Stage III). After Rumball (1991) [17].
Journal of Environmental Chemical Engineering 6 (2018) 281–288
interpreted as the slow reduction in pH until the rapid and exothermic
oxidation by Fe(III) accelerates.
The preferred method of controlling both acid mine drainage and
reactive ground has been to maintain a high pH through the use of
alkaline substances. The use of solid bases, such as limestone, is not
effective, however, as the Fe(III) precipitates on the surface and passi-
vates the remaining solid base, a process termed ‘armoring’, rendering
it ineffective.
Accordingly, the use of urea is preferred, which homogeneously
generates the weak bases ammonia and carbonate by hydrolysis, and
(3)
hence consumes protons (Eq. (8))
CO(NH2) 2 + 2H+ + 2H2 O → 2NH+4 + H2 CO3 (8)
There is compelling empirical evidence in the industry that urea is
(4) an effective inhibitor of the thermal decomposition of AN in reactive
ground. The mechanism of this inhibition is uncertain. The hydrolysis
of urea is known to be a slow reaction, which proceeds at a rate that is
independent of pH [18]. The length of the induction period could be
(5) limited by the total consumption of the urea. Alternatively, if the rate of
acid generation is greater than the rate of urea hydrolysis, then the pH
of the system could slowly drop, despite the partial neutralisation by
the urea hydrolysis, until it reaches an acidic condition that allows an
(6) autocatalytic runaway decomposition. Finally, the urea could act as an
inhibitor by a mechanism not involving its acid-base chemistry. This
third alternative is in fact our preferred explanation from the evidence
we have accumulated.
2. Experimental
(7)
2.1. Materials
Several sources of pyrite were used. RG1 was supplied by Dyno
Nobel and is a reactive grade ground sample containing ∼2.50% by
weight of adsorbed water, and a pyrite content of less than 10 wt%. The
remaining material is a mixture of clays, quartz and organic matter. The
particle size is less than 50 μm on average. Pure pyrite (PY) was ob-
tained from Spectrum Chemicals and is 100% oxidized pyrite with a
grain size of 200–400 μm.
Ammonium nitrate, AN, (Acros Organics, 99 +%) was used as re-
ceived but was ground in a mortar prior to use to break up any large
clumps. Dodecane (Sigma, ≥99%), iron (II) sulfate 7 hydrate
(BDH, > 99.5%), iron (III) sulfate 5 hydrate (Fluka), urea (Ajax che-
micals, 99.5%), hydrazinium sulfate (Ajax chemicals, > 99.5%), zinc
oxide (Aldrich, 99.9%, < 1 μm), kaolin (Kaolin Australia, Pty Ltd,
Eckafine BDF), hydrotalcite (Sigma) and Basolite C300 (BASF) were
used as received. Sodium nitrite (Mallinckrodt), diacetyl monoxime,
DCM (Fluka), thiosemicarbazide, TSC (BDH), phosphoric acid (85%,
Ajax Finechem Pty. Ltd), sulfuric acid (96%, Ajax Finechem Pty. Ltd),
iron (III) chloride 6 hydrate (Merck), and PIBSA-DEEA (Clariant) were
also used as received.
2.2. Urea determination
Urea was determined by UV–vis spectroscopy at a wavelength of
525 nm using diacetly monoxime, DCM and thiosemicarbazide, TSC
[19]. An acidic ferric solution was made containing phosphoric acid
(100 ml), sulfuric acid (300 ml), water (600 ml) and ferric chloride
(0.10 g). DCM and TSC were mixed (0.50:0.01 g) and made to volume
(100 ml). When ready to use, the chromogenic reagent containing the
acid solution (2 parts) and DCM/TSC solution (1 part) were mixed. Urea
stock solutions were prepared containing ∼ 20 ppm urea.
Standard urea solutions were prepared by diluting stock urea solu-
tions in water. The urea solution (0.32 ml) was mixed with the chro-
mogenic solution to 10 ml, covered with aluminum foil and heated in
boiling water for 10 min. The sample was cooled rapidly in ice and the
UV–vis spectrum was measured from 400 to 600 nm.
282
A.M. Djerdjev et al. Journal of Environmental Chemical Engineering 6 (2018) 281–288

Six samples containing 5 wt% urea (based on AN), ammonium ni-

trate (0.9 g), reactive ground 1 (0.9 g) and weathering solution con-
taining dissolved urea (0.047 g) were prepared and placed in two
thermos flasks and heated to 55 °C in a sand bath. Samples were re-
moved at selected time intervals, the first after five minutes and the last
after 20 days. The samples were quenched with water (8.4 g) and the
pH measured. More water was added (30 g total) and the slurry was
then filtered through a 0.2 μm filter in a 50 ml volumetric flask con-
taining a drop of concentrated sulfuric acid. The solution was further
diluted (1.0 ml into 50 ml) and 0.32 ml was dispensed into 10 ml vo-
lumetric flasks to which the chromogenic solution was added to vo-
lume. The sample was heated, cooled and then the UV–vis spectrum
was measured from 400 to 600 nm. Inhibition studies involved adding
inhibitors to the PY/AN/WS mixture, sealing and capping with a syr-
inge, then heating to 55 °C in a glycerol bath. The reaction was noted
when NOx began to be generated in the syringe. Fig. 2. XRD spectra of pure pyrite (SPC) and reactive ground (RG 1).

2.3. Weathering solution

The x-ray diffraction data were collected with CuKa radiation, using an
Synthetic weathering solution was freshly made, containing iron (II) X’Pert Pro diffractometer (Pan analytical). The copper source was run at
sulfate 7 hydrate (0.245 g), iron (III) sulfate 5 hydrate (0.50 g), and 45 KeV and 45 mA and measured between 5 and 90°.
water (3.3 g). The mixture was gently sonicated until fully dissolved.

2.4. pH measurements 3. Results and discussion

Mixtures of reactive ground (RG1), ammonium nitrate (AN), and
weathering solution (WS) were reacted (0.9:0.9:0.2 g) in small glass
vials and heated in a water bath at 55 °C or 80 °C for five minutes. After
this time the sample was quenched with ∼6.5 g water and the pH
measured.
2.5. NO adsorption
Reactive ground was mixed with AN, and WS (0.9:0.9:0.2 g) and
placed in the bottom of a small 5 ml glass tube. Potential inhibitors
were physically separated from the reactive mixture, so that they were
only in contact through the gas phase. The solid inhibitors (∼0.7 g)
were dispersed in dodecane (∼40 wt%, based on the mass of the in-
hibitor). The reactive mixture was mixed until a uniform paste was
achieved, then added to the bottom of the reaction tube. A polyethylene
foam support, cut to size, was then placed half-way up the tube on
which a glass fibre filter disc (250 μm pore size) was placed. The in-
hibitors were placed on top of this filter to prevent small particles from
being in direct contact with the reactive mixture and inhibiting the
reaction. We tested kaolin along with zeolite A, hydrotalcite and zinc
oxide. A blank was made by adding a similar quantity of dodecane to
the glass filter. The reaction tubes were closed with a plastic cap con-
taining a small pin hole and immersed in a water bath at 55 °C. The
reaction began at the first visible sign of brown NO2.
3.1. Tests of the acid neutralisation hypothesis
Reactive ground and pure pyrite was used and characterized by XRD
(Fig. 2). The reactive ground sample contained mixtures of minerals
consisting predominantly of quartz (Q), with some clinochlore (C), as
well as some pyrite mineral (P) [20]. The spectrum pyrite was 100%
pyrite. Six reactions containing ammonium nitrate (AN), reactive
ground (RG 1) and weathering solution (WS), with 5 wt% urea, were
prepared and sampled after 5 min, 1, 3, 6, 10 and 17 days. After
quenching the samples with water the pH was measured and the total
urea analysed by UV–vis. After 17 days of reaction, the consumption of
urea was only partial; the urea decreased from an initial mass of 0.046 g
to ∼ 0.02 g. During this time, the pH of the slurry decreased from 1.5 to
1.3 (Fig. 3). If urea were hydrolysing to produce base, then the pH
should be greater than 1.3 after this time. Similar results were found at
lower urea concentrations; with 0.2 wt% the urea deceased by one-third
prior to the decomposition process at the end of the induction period.
We conclude that the hydrolysis of urea is too slow to neutralise acid
significantly, especially during the early stages (< 1 day) of the reac-
tion, and that the presence of excess urea does not increase the hy-
drolysis rate. This is consistent with literature reports that the hydro-
lysis reaction is slow, with a rate constant of 8.4 × 10−10 s−1 at 25 °C
[21].

2.5.1. NOx analysis

The build-up of NOx during Stage I and into Stage II was determined
with a Kane, Quintox flue gas analyser. Four duplicate samples were
prepared, to which were added reactive ground, ammonium nitrate and
water (0.9:0.9:0.2 g) in 16 mm (i.d.) glass test tubes (15 cm long). The
samples were sealed with a rubber stopper and heated in a water bath
at 55 °C. At designated time intervals, each sample was analysed for NO
and NO2 in the headspace above the sample, then continued to be he-
ated and resampled until the end of Stage I.

2.6. IR spectroscopy, UV–vis and XRD

IR spectra were recorded with a Bruker Tensor 27 spectro-

photometer using the DRIFTS method between 400 and 4000 cm−1 Fig. 3. Change in urea concentration and pH with time for RG 1, AN, and WS mixtures
using KBr as background. UV–vis absorbance spectra were recorded containing 5 wt% urea and heated at 55 °C. The end of the induction time has not been
reached. The dashed line represents the starting urea concentration.
with a UV–vis spectrophotometer (Cary 1E) between 200 and 700 nm.

283
A.M. Djerdjev et al.
Fig. 4. Reaction between ammonium nitrate and reactive ground (RG 1) showing a) the
hot stage microscope and b) the generation of gas bubbles.
3.2. Generation of NO product
Preliminary studies were conducted in a glass reaction cell which
was placed on a microscope hot stage at 55 °C (Fig. 4), as described in
reference [7]. An interface was formed between reactive ground (RG 1)
and an AN emulsion. The initial stages of the reaction were directly
observed using a video microscope. Gas bubbles form rapidly in the
sample after an induction period of approximately 20 min, when the
emulsion had no inhibitor. The colourless gas in the bubbles im-
mediately became brown when it came in contact with O2, indicating
that it was nitric oxide.
3.2.1. IR identification and quantification of NO
The IR spectra of ammonium nitrate, pyrite and their mixtures are
shown in Fig. 5. Bands due to surface bound NO species are seen in the
region 1750–1800 cm−1. The presence of the vibrational mode at
1776 cm−1 is due to the stretching vibration of N]O of adsorbed NO
[22,23]. To confirm the initial formation of NO as the precursor to NO2
gas, a reaction was performed with Reactive ground (RG1) and AN in
the presence of ∼2% water. The sample was mixed and then sealed
with a rubber septum and placed in a water bath at 55 °C for 1 h.
Oxygen was generated by mixing permanganate ions with peroxide and
collecting the gas in a syringe. The oxygen gas was then injected
through the rubber septum (Fig. 6). Brown NO2 immediately formed in
the vial. The NO gas was even formed at room temperature without
heating. After ∼1 h when the cap was removed, a clear gas was dis-
charged that turned brown on exposure to air.
Further experiments were conducted in reaction tubes maintained
at 55 °C in a water bath. The induction time was taken as the time at
which the first indication of brown gas was evident above the slurry.
This point (end of Stage I) closely coincided with an expansion of the
284
Journal of Environmental Chemical Engineering 6 (2018) 281–288
Fig. 5. IR spectra of pure ammonium nitrate and ammonium nitrate/pyrite mixtures.
sample volume by a factor of ∼2 (start of Stage II). After this initial
expansion, the volume further increased by ∼4 times with the evolu-
tion of more dark brown gas (Stage II). Stage III began when the volume
increased further with violent bubbling, followed by vigorous evolution
of dark brown gas, and sometimes thick white smoke. The presence of
inhibitors generally reduced the severity of Stage II (and III) and ex-
tended its length. Here, the induction time was still taken as the time at
which brown NO2 gas was initially evolved, despite the runaway being
further delayed. To measure the formation of NO directly, instead of
observing its oxidation to brown NO2, a combustion gas emissions
monitor was employed.
The reaction was conducted at 55 °C in a water bath, with the gas
atmosphere withdrawn for measurement at each data point. This re-
moval of the gas also inhibits the reaction, which reached Stage II only
after four hours when sampled ten times throughout the induction
period; when NO was not removed the induction time was only ∼100
min. The results indicate that the accumulation of NO remains low, at
least in the gas phase, until the end of the induction period (Fig. 7),
when it was quantitatively formed accompanied by NO2.
3.3. Inhibition
3.3.1. Nitric oxide adsorption
The results described above suggest that the decomposition of AN
occurs in the presence of NO, but that NO and NO2 only accumulate
significantly at the end of the induction stage, and their removal delays
the reaction. To test this inference, the standard reactive ground test
with AN, RG 1 and WS was performed, but with potential inhibitors
physically separated from the reactive mixture, so that they were only
in contact through the gas phase. The solid inhibitors were dispersed in
oil containing surfactant to simulate the actual condition inside a water-
in-oil emulsion.
The reactive mixture consisting of RG 1, AN, and WS was mixed,
then added to the bottom of the reaction tube. Since particulate in-
hibitors would be present in the oil phase of an emulsion, the inhibitors
were dispersed in dodecane to make a thick paste, which was placed on
the top of the filter as described in Section 2.5. Kaolin, zeolite A, hy-
drotalcite and zinc oxide were tested and the reaction tubes were closed
with a plastic cap containing a small pin hole and immersed in a water
bath at 55 °C. After 71 min of heating the zeolite A sample had already
reacted, and the kaolin was beginning to react along with the blank as
indicated by the evolution of brown NO2 gas. After 90 min the zinc
oxide sample began to react leaving only hydrotalcite unreacted. Fi-
nally, after 130 min the hydrotalcite sample began to react (Fig. 8). The
slight differences in times between the blank and inhibitors are due to
different amounts of inhibitor and oil present initially; zinc oxide also
adsorbed ∼40% more oil to achieve the same consistency.
The only mechanism for inhibition in these systems is gas
A.M. Djerdjev et al.
Fig. 6. Reactive mixture containing RG 1, AN, and water after a) heating for 1 h at 55 °C; and b) after injecting O2.
Fig. 7. NO and NO2 observed by sampling the atmosphere above the reaction mixture
containing AN, RG 1 and WS at 55 °C. The start of Stage II occurs after ∼260 min.
adsorption, since the inhibitors were not in direct contact with the re-
active mixture, but separated by some distance. In the absence of in-
hibitor, the decomposition reaction occurs after about 20 min. The in-
crease in the induction time in the presence of dodecane to 71–90 min
indicates that gas adsorption occurs into the oil phase with a significant
solubility. It is well known that nitric oxide has a much greater solu-
bility in oil than in water [24]. The oil-dissolved NO apparently does
not participate in the AN decomposition reactions.
Selected inhibitors, which now included a metal/organic framework
(MOF, Basolite C300) and urea, were then heated with reactive ground,
AN and WS at 55 °C (Fig. 9) to demonstrate inhibition by NOx removal,
compared to acid neutralisation. Slight variations in the reproducibility
of the results arose due to the state of dispersion of the particulate in-
hibitors. As such, the results serve to demonstrate the overall trend of
285
Journal of Environmental Chemical Engineering 6 (2018) 281–288
the role of NOx on the inhibition of the reaction between ammonium
nitrate and pyrite.
3.3.2. Acid neutralisation-Zinc oxide
A role for zinc oxide as an acidity inhibitor could be to keep the pH
above pH 4 and inhibit the catalytic action of HNO2. Zinc oxide is a
weak base, which is typically used by incorporating it into the oil
continuous phase of an emulsion. The pH of the reactive ground mix-
ture with ZnO after heating to 55 °C for 5 min increased gradually from
its initial value of 2.2 (with no inhibitor) to pH 2.5 with ∼2.5 wt% ZnO
and to ∼3.0 with 4.2 wt% ZnO. This final pH increased further to ∼4.7
when heated to 80 °C. The induction time similarly increases gradually
up to ∼4 wt% then increases more rapidly, consistent with the increase
in pH (Fig. 9). Zinc oxide is not known to sequester NO, so that its
function as an inhibitor could either occur through the acid-base
equilibria or be coupled in another mechanism in some way to the re-
actions of NOx. For comparison, the pH of the hydrotalcite mixture at
55 °C was also measured. The pH increases quickly to 2.4 with 0.5 wt%
HT, but remains constant even with 4.25 wt% added.
3.3.3. Nitrous acid
It has been proposed by Rumball that nitrous acid (HNO2, pKa = 2.8
[25]) is a catalyst for AN decomposition that accumulates during the
induction stage and that bases could act to neutralise the acid form
[17]. The effect of adding nitrite ions to the RG 1/AN/Water system
was investigated to determine if the build-up of nitrite ions/nitrous acid
within the system would accelerate the rate of reaction in the same way
that ferric ions do. The induction time was found to be dramatically
shortened with increasing nitrite ion concentration (100 min in the
absence of added nitrite; 58 min at 0.02 M; 6 min at 0.05 M).
At ∼0.075 M sodium nitrite the reaction spontaneously occurred at
55 °C.
Further evidence for nitrous acid formation can be seen in the IR
(Fig. 10). Surface peaks assigned to nitrite-PY adsorption [23] can be
seen in the region 1200–1300 cm−1 when pyrite and ammonium
A.M. Djerdjev et al. Journal of Environmental Chemical Engineering 6 (2018) 281–288

Fig. 8. Effect of particulate inhibitors present above the reactive mixture of RG 1, AN, and WS after heating at 55 °C for the given times.

Fig. 9. Induction time of particulate inhibitors present in the reactive mixture of RG 1, Fig. 10. IR spectra of ammonium nitrate, pyrite and ammonium nitrate pyrite mixtures.
AN, and WS after heating at 55 °C for the given times. In the absence of inhibitors the
induction times is approximately 20 min.
economical due to the low cost, high water solubility, moderate pH and
neutral by-products formed. It reacts to reduce nitrous acid to dini-
nitrate were mixed. The accumulation of nitrous acid during the in-
trogen:
duction period between ammonium nitrate and pyrite was also mea-
sured by UV–vis. A sample containing pure washed pyrite, ammonium CO(NH2) 2 + 2HNO2 → 2N2(g ) + CO2(g ) + 3H2 O (9)
nitrate and water (2, 2, 6 g respectively) were mixed and heated to
Several other scavengers act in a similar fashion. According to
55 °C isothermally. After selected time intervals during the induction
Fitzpatrick [27] the reaction of sulphamic acid with nitrous acid in
period a 0.6 ml aliquot of the supernatant was removed and filtered
acidic media produced nitrogen gas, water and more acid, as does the
through a 0.4 μm filter into sulfuric acid (∼10−2 M, 3.6 ml). The fil-
reaction with hydrazinium ion.
trate was measured by UV–vis between 200 and 700 nm using an am-
monium nitrate blank in sulfuric acid (∼10−2 M). The initial broad NH2 SO3 H + HNO2 → N2(g ) + HSO−4 + H+ + H2 O (10)
band at ∼ 340 nm increased in intensity over time and the emergence
N2 H+5 + HNO2 → NH3(g ) + H+ + 2H2 O (11)
of bands around 370 nm confirm the build-up of nitrous acid prior to
the acceleration of the reaction (Fig. 11) [25,26]. If acid neutralisation were the key to the inhibition mechanism
Consequently, nitrous acid scavengers were investigated. Nitrous between the reaction of pyrite and ammonium nitrate then this should
acid scavengers are common and well known and typically include accelerate the reaction as more acid is produced. When we tested the
amines, amides and related products. Of these, urea is the most reaction with these inhibitors we found increases in the induction time.

286
A.M. Djerdjev et al.
Fig. 11. UV–vis spectra versus time for PY (SPC), AN, and water mixtures after heating at
55 °C.
Fig. 12. Induction time versus inhibitor concentration for PY (SPC), AN, and WS mixtures
after heating at 55 °C. ■ Urea, ● Hydrazinium sulfate.
In Fig. 12 we show the results for hydrazinium sulfate and urea for
comparison. Both of these salts show a similar effect on the induction
time at low concentrations despite one salt being an acidic salt (hy-
drazinium ion) and the other hydrolysing slowly to produce base
(urea). However, at higher concentrations the generation of excess acid
negates the inhibition mechanism as shown for the hydrazinium ion.
3.4. Proposed mechanisms
Mechanisms can now be advanced for the multiple roles of in-
hibitors of the decomposition of ammonium nitrate as illustrated in
Scheme 1:
Pyrite and/or Fe2+ react with the nitrate ion to form NO. In the
presence of NO3− and acid some of the dissolved NO will form HNO2
through the reversible equilibrium (Eq. (12)) [28,29], or by the oxi-
dation with molecular dioxygen (Eq. (13)) [30]
H2 O+ 2 NO + H+ + NO−3 ⇔ 3 HNO2
4 NO + O2 + 2 H2 O⇔ 4 HNO2
Reaction of ferrous to ferric occurs via oxidation by nitrous acid
leading to the reformation of NOx. NO is a powerful auto-catalyst which
accelerates the reaction between pyrite and nitrates. The autocatalytic
and rate-enhancing power of NO has been utilized to extract valuable
metals trapped within sulphide minerals as inclusions by annihilating
the sulphide lattice through rapid oxidation [31].
When NO is dissolved in water and converted to HNO2, it can be
reduced by urea to produce N2 and CO2 at low temperatures (5–60 °C)
[32–34].
Journal of Environmental Chemical Engineering 6 (2018) 281–288
Scheme 1. Thermal decomposition pathway of ammonium nitrate and pyrite showing
NOx generation/sequestering and nitrous acid formation/reduction.
2HNO2 + NH2 CONH2 → 2N2(g ) + CO2 + 3H2 O (14)
This method of removal of NO by converting it first to nitrous acid
and then reacting nitrous acid with urea, has been patented by
Warshaw in 1971 as a process to remove NO from industrial waste
gases (urea acidic process) [35]. Since at low acidity levels HNO2 de-
composes to form gaseous NO, the urea oxidation process is carried out
at pH of about 1 to prevent decomposition. As pH increases above 2, the
efficiency of the process decreases sharply [36]. Therefore, when
emulsions are used, the urea in the emulsion droplets (at ∼pH 5) does
not scavenge NO diffusing into them via the oil phase of the emulsion.
Our new understanding of the mechanism of the reaction between
ammonium nitrate and pyrite enable us to better control the potential
environmental hazards associated with mining of reactive grounds.
Highly toxic gases such as NOx and CO can be formed during blasting or
during deflagration, which can be dangerous for human health at
concentrations of 20, 100, and 1200 ppm, for NO2, NO and CO re-
spectively [37,38]. Other hazards include nitrate leaching into
groundwater from emulsion deflagration, which is toxic to both aquatic
and human life. An illness known as “Blue Baby Syndrome” arises from
high nitrate levels, above the WHO guideline for maximum con-
taminant level (MCL) of 50 mg/L [39].
4. Conclusions
Our results, together with the literature indicate that the active
species for the decomposition appears to be HNO2, with a pKa of 2.8
[25], but not the nitrite ion NO2−. This has revealed several mechan-
isms for the inhibition of the decomposition of AN. This enables weak
bases such as ZnO to inhibit the decomposition without being otherwise
involved in the reaction chemistry. The nitrous acid is formed from NO,
so sequestering this species provides another means of inhibition. Hy-
(12) drotalcite appears to work by this mechanism, and other modified clay
minerals could be even more effective. Sequestering NO only provides a
(13)
reservoir which ultimately can become saturated. A permanent solution
is the decomposition of the nitrous species to inert N2 and H2O, which
can be effected by urea. Under conditions of moderately low tem-
perature (< ∼60 °C), urea acts as an inhibitor by scavenging nitrous
acid, not by slowly hydrolyzing to produce base, as originally sug-
gested. The kinetics of this reaction are likely to determine the sleep-
time of an inhibited product, and is the subject of future work.
By using gas adsorbers such as hydrotalcite, emulsion explosives are
more stable and less prone to react prematurely, so that nitrates, NOx
and other greenhouse gas emissions can be minimised.
287
A.M. Djerdjev et al.
Author contributions
The manuscript was written through contributions of all authors. All
authors have given approval to the final version of the manuscript.
Acknowledgments
This work was supported by the Australian Research Council and
Dyno Nobel Asia Pacific, Pty Ltd.
References
[1] T.J. Briggs, I.J. Kelso, Ammonium nitrate-sulfide reactivity at the century Zn- Pb-Ag
mine, northwest queensland, Australia, Explor. Min. Geol. 10 (2001) 177–190.
[2] X. Wang, Emulsion Explosives, Metallurgical Industry Press, Beijing China, 2008.
[3] E.G. Mahadevan, Ammonium Nitrate Explosives for Civil Applications: Slurries,
Emulsions and Ammonium Nitrate Fuel Oils, Wiley-VCH Verlag GmbH & Co KGaA,
2013.
[4] D.H. Cranney, B. T. Maxfield, Emulsion that is compatible with reactive sulfide/
pyrite ores, in: US Patent, 5159153, 1992.
[5] Z. Xu, D. Liu, Y. Hu, Investigation of ammonium nitrate based emulsion ignition
characteristic, J. Loss Prev. Process Ind. 26 (2013) 994–1001.
[6] R. Gunawan, D. Zhang, Thermal stability and kinetics of decomposition of ammo-
nium nitrate in the presence of pyrite, J. Hazard. Mater. 165 (2009) 751–758.
[7] P. Priyananda, A.M. Djerdjev, J. Gore, B.S. Hawkett, C. Neto, J.K. Beattie,
Premature detonation of an ammonium nitrate emulsion in reactive ground, J.
Hazard. Mater. 283 (2015) 314–320.
[8] V.P. Evangelou, Pyrite Oxidation and Its Control, Chpt 6, C.R.C. Press, 1995.
[9] D.B. Johnson, K.B. Hallberg, Acid mine drainage remediation options: a review, Sci.
Total Environ. 338 (2005) 3–14.
[10] M. Gleisner, R.B. Herbert, P.C.F. Kockum, Pyrite oxidation by Acidithiobacillus
ferooxidans at various concentrations of dissolved oxygen, Chem. Geol. 225 (2006)
16–29.
[11] Z.M. Migaszewski, A. Gałuszka, P. Pasławski, E. Starnawska, An influence of pyrite
oxidation on generation of unique acidic pit water: a case study podwiśniówka
quarry, holy cross mountains (South-Central Poland), Polish J. Environ. Stud. 16
(2007) 407–421.
[12] K.J. Edwards, P.L. Bond, T.M. Gihring, J.F. Banfield, An archael iron- oxidising
extreme acidophile important in acid mine drainage, Science 287 (2000)
1796–1799.
[13] K.J. Edwards, M.O. Schrenk, R. Hamers, J.F. Banfield, Microbial oxidation of pyrite:
experiments using microorganisms from the extreme acidic environment, Am.
Mineral. 83 (1998) 1444–1453.
[14] T.A. Fowler, P.R. Holmes, F.K. Crundwell, Mechanism of pyrite dissolution in the
presence of thiobacillus ferrooxidans, Appl. Environ. Microbiol. 65 (1999)
2987–2993.
[15] Z. Xu, D. Liu, Y. Hu, Influence of iron ion on thermal behavior of ammonium nitrate
and emulsion explosives, Cent. Eur. J. Energetic Mater. 7 (2010) 77–93.
[16] B. Fabiano, R.J.A. Kersten, A. Barbucci, M.N. Boers, Experimental and theoretical
approach to the assessment of stability criteria for safe transport of ammonium
nitrate based emulsions, Chem. Biochem. Eng. Q. 27 (2013) 307–317.
[17] J.A. Rumball, The Interaction of Partially Weathered Sulphides in the Mt. McRae
Shale Formation with Ammonium Nitrate, P. Murdoch University, Australia, 1991
288
Journal of Environmental Chemical Engineering 6 (2018) 281–288
(Libraries Australia ID: 8898701, Perth Australia).
[18] W.H.R. Shaw, J.J. Bordeaux, The decomposition of urea in aqueous media, J. Am.
Chem. Soc. 77 (1955) 4729–4733.
[19] M. Rahmatullah, T.R.C. Boyde, Improvements in the determination of urea using
diacetyl monoxime; methods with and without deproteinisation, Clin. Chim. Acta
107 (1980) 3–9.
[20] M. Fantauzzi, C. Licheri, D. Atzei, G. Loi, B. Elsener, G. Rossi, A. Rossi, Arsenopyrite
and pyrite bioleaching: evidence from XPS, XRD and ICP techniques 2237–2247,
Anal. Bioanal. Chem. 401 (2011) 2237–2247.
[21] P. Hagel, J.J.T. Gerding, W. Fieggen, H. Blumendal, Cyanate formation in solutions
of urea. I. Calculation of cyanate concentrations at different temperature and pH,
Biochim. Biophys. Acta 234 (1971) 366–373.
[22] A.L. Goodman, T.M. Miller, V.H. Grassian, Heterogeneous reactions of NO2 on NaCl
and Al2O3 particles, J. Vac. Sci. Technol. 16 (1998) 2585–2590.
[23] S. Singireddy, A.D. Gordon, A. Smirnov, M.A. Vance, M.A.A. Schoonen,
R.K. Szilagyyi, D.R. Strongin, Reduction of nitrite and nitrate to ammonium on
pyrite, Origins Life Evol. Biospheres 42 (2012) 275–294.
[24] A.C. Denicalo, C. Batthyany, E. Lissi, B.A. Freeman, H. Rubbos, R. Radi, Diffusion of
nitric oxide into low density lipoprotein, J. Biol. Chem. 277 (2002) 932–936.
[25] E. Riordan, N. Minogue, D. Healy, P. O’Driscoll, J.R. Sodeau, Spectroscopic and
optimisation modelling study of nitrous acid in aqueous solutions, J. Phys. Chem. A
109 (2005) 779–786.
[26] T. Arakaki, T. Miyake, T. Hirakawa, H. Sakugawa, pH dependent photoformation of
hydroxyl radical and absorbance of aqueous-Phase N(III) (HNO2 and NO2−),
Environ. Sci. Technol. 33 (1999) 2561–2565.
[27] J. Fitzpatrick, T.A. Meyer, M.E. O’Neill, D. Williams, H. Lyn, Comparison of the
reactivity of nine nitrous acid scavengers, J. Chem. Soc. Perkin Trans. 2 (1984)
927–932.
[28] T.A. Turney, G.A. Wright, Nitrous acid and nitrosation, Chem. Rev. 59 (1959)
497–513.
[29] M.J. Akhtar, D. Axente, F.T. Bonner, Nitrogen isotope exchange between nitric and
nitrous acids, J. Chem. Phys. 71 (1979) 3570–3572.
[30] H.H. Awad, D.M. Stanbury, Autoxidation of NO in aqueous solution, Int. J. Chem.
Kinet. 25 (1993) 375–381.
[31] O. Gok, Catalytic oxidation mechanism of oxy-nitrogen species (NOx) in FeSO4
electrolyte, Nitric Oxide 25 (2011) 47–53.
[32] A. Lasalle, C. Roizard, N. Midoux, P. Bourret, P.J. Dyens, Removal of nitrogen
oxides (NOx) from flue gases using the urea acidic process: kinetics of the chemical
reaction of nitrous acid with urea, Ind. Eng. Chem. Res. 31 (1992) 777–780.
[33] K.N. Dvoeglazov, V.I. Marchenko, Oxidation of urea with nitrous acid in nitric acid
solutions, Radiochemistry 47 (2005) 58–62.
[34] P. Fang, C. Cen, Z. Tang, P. Zhong, D. Chen, Z. Chen, Simultaneous removal of SO2
and NOX by wet scrubbing using urea solution, Chem. Eng. J. 168 (2011) 52–59.
[35] A. Warshaw, Removal of nitrogen oxides from a gas stream, in: US Patent 3565575,
1971.
[36] W. Fitzgibbons, Inexpensive purification of urea waste streams, in: US Patent
4188292, 1980.
[37] Pocket Guide to Chemical Hazards, National Institute for Occupational Safety and
Health [NIOSH], US Department of Health and Human Services, Public Health
Service, Centre for Disease Control and Prevention, Cincinnati, OH USA, 1994.
[38] Blasters' Handbook, 17th ed., International Society of Explosives Engineers,
Cleveland, OH USA, 1998.
[39] R.K. Maheshwari, A.K. Chauhan, B. Lal, A.K. Sharma, Nitrate toxicity in ground-
water: its clinical manifestations, preventive measures and mitigation strategies,
Octa, J. Environ. Res. 1 (2013) 217–230.