mr.crow
National Hazard
Posts: 884
Registered: 9-9-2009
Location: Canada
Member Is Offline
Mood: 0xFF
|
|
Nitrates and Aluminum question
I finished my experiment reacting Ammonium Nitrate with Sodium Hydroxide to form Sodium Nitrate. After boiling down and crystallizing the solution 3
times I took the remaining liquid and heated it to a paste.
This is when something unexpected happened.
Heating the paste with a butane torch on a sheet of aluminum foil caused a self sustaining reaction producing fizzing, steam and the smell of ammonia.
It would spread to the entire surface covered by the liquid, not just the area heated by the torch. The bubbling reaction turned gray and ate a hole
in the foil.
Heating a large crystal of the purified product and a few drops of water produced no reaction other than boiling of the water leaving the salt behind.
We can conclude that I had an excess of NH4NO3 and not NaOH. Any NH3 would have been driven off a long time ago. The ammonium salt reacts with the
aluminum and not the sodium.
Does anyone know what is going on?
Double, double toil and trouble; Fire burn, and caldron bubble
|
|
|
The WiZard is In
International Hazard
Posts: 1617
Registered: 3-4-2010
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by mr.crow  |
We can conclude that I had an excess of NH4NO3 and not NaOH. Any NH3 would have been driven off a long time ago. The ammonium salt reacts with the
aluminum and not the sodium.
Does anyone know what is going on? |
----
Introduction
EXPLOSION HAZARDS OF PYROTECHNIC ALUMINIUM COMPOSITIONS
Stig R Johansson and K Goran Persson
Industrial Laboratories,
The Swedish Match Co , Jonkoping
Föredrag vid PYROTEKNIKDAGEN [Pyrotechnic Day] 1971
The small scale handicraft production of generator-gas matches and
other pyrotechnic items based on aluminium containing compositions takes
place occasionally, for instance, in some match factories An explosion
which occurred on the 7th of May 1970 in connection with manufacture of
this kind brought into focus the question as to whether self-ignition and the
subsequent detonation of pyrotechnic aluminium compositions are possible
on the whole.
The high reactivity of aluminium powders is utilized, inter alla in
thermite reactions, explosives, and propellants, but very little can be found
in literature about fire and explosion hazards, especially as far as
aluminium compositions are concerned (cf ref 1, p. 328)
The course of events of the present accident was as follows. At the
end of the day before (i.e., the 6th of May) a stainless steel bucket
containing about 12 kg residual aluminium composition was placed in water
in a vat without being fully immersed, The composition was covered by a
layer of water, and in addition a loose lid was placed on the top of the
bucket.
On the following day, which was a holiday, an explosion occurred in
the afternoon. The explosion resulted in considerable structural damage to
the building, as is shown by the picture in figure 1. Fortunately, nobody was
in the area.
After the place had been cleared an impact mark could be seen in the tiled
floor where the vat, which was made of stainless steel, had been placed,
Two possible explanations can be imagined, The explosion could have been
caused either by a gas phase detonation, or by detonating composition, In both
cases a chemical reaction involving the consumption of aluminium is likely to
have started the process,
According to the gas hypothesis hydrogen was evolved in accordance with the
reaction
2 Al(s) + 6 H2O(1) - 2 Al(OH)3(s) + 3 H2(g) R,1.
The resulting explosive mixture of hydrogen and air filled the room (the
volume of which was 73 m3) and exploded after having been ignited, for
instance by an electric spark,
However, a stoichiometric calculation shows that the lower
explosion limit (4 % by volume) could not possibly have been reached,
even if all available alurninium had reacted according to Reaction 1.
Another important fact is that no obvious ignition source could be
pointed out. For these reasons the hypothesis of gas phase detonation
was discarded at an early stage.
According to the second hypothesis it was the composition that was
brought to detonation and in the bucket it acted as a bomb. The process
probably started by self-heating, perhaps due to Reaction 1, which is
exothermal enough. After attaining the ignition temperature, ', a combustion
reaction was set off. Finally, the initial low velocity combustion or defla-
gration underwent a transition to high velocity combustion or detonation.
Experimental
The main inorganic components of the present composition,
which turned out to be an explosive with self-igniting properties and
thus very dangerous, were, in decreasing order of quantity, potassium
perch orate, water, aluminium powder, barium nitrate and potassium
nitrate,
The purpose of this anything but comprehensive investigation was to
find out about the conditions precedent for self-ignition. The experiments
therefore include introductory studies on self-heating, ignition, and initiation
of detonation.
Starting from the basic formula (which cannot be given for secrecy
reasons) compositions were prepared with four different aluminium powders
(a survey is given in table 1). Further, one or two chemicals were excluded
or exchanged in some compositions Thus, in one experiment the nitrates
were excluded, in another they were exchanged for the corresponding
chlorides.
The powder designated XP 62000 has been treated with stearine,
Whether this is true for the powders "hell" and "dunker" as well is not
known to us.
Table 1. Data of investigated aluminium powders
Manufacturer Designation Type Particle Size
A - dunkek flake formed 1 - 10 um
B - _ hell " " 50 _ 100 um
C Carlfors Bruk XP 62000 " " 2 - 40 um
D " " A 100 atomized 60 % < 45 um
The aluminium powders have not been analysed chemically.
Self_heating. The self-heating properties of the different compositions were
studied in the experimental arrangement shown in figure 2.
A piece of plastic foam, "Styropor", with the dimensions 13 cm 13 cm x
13 cm was bored so that a cavity with a volume of 5 cm3 remained for the
sample when the Styropor lid was put into place (cf. figure 2). In this way the
sample was surrounded by Styropor walls with a minimum depth Of 4 cm.
An iron_constantan thermocouple, the tip of which was coated with
epoxy resin, was placed in the middle of the sample. The leads, which were
wound with teflon tape, were connected to a potentiometer recorder on which
a continuous temperature-time curve was obtained.
Attempts to measure the pH value of the moist composition with a glass
electrode indicated that the pH of the water phase, if any, was between 6.4
and 6.7.
The initial temperature of the compositions was equal to the
temperature of the room, which varied between 22 C and 26 C
A typical temperature-time curve for self-heating compositions is shown
in figure 3 Substantial self-heating occurs after about 30 hours with a rapid
temperature increase at the end.
The maximum temperature observed was limited to about 130 C. This
limit was set by escaping gases in cases where the lid was blown off, or by
melting of the wall material. Due to -the small amount of sample - ca. 10 g -
the heat evolved by chemical reaction was limited as well
The primary experimental result was that self-heating could occur if, and
only if, the composition contained aluminium powder A (cf. table 1)
Measurements on such compositions further revealed that nitrate and water
had to be present as well Ignition did not occur in any of the experiments
Self-heating was always accompanied by the evolution of ammonia The analysis
was restricted to a qualitative one (Drager tubes were used).
Ignition: Compositions which had been subjected to self_heating appeared dry
and somewhat loose in their consistency. After they had been cooled down to
room temperature, the samples could easily be ignited with a match or by a hot
body. As perceived by eye and ear the combustion was vigorous and subsonic;
we 1herefore prefer to describe it as a deflagration (cf. ref. 2, p. 407, and ref. 3,
p. 1008).
Even samples which had not been subjected to self-heating, but which had
been dried in air at room temperature, burned freely when ignited. This
behavriour was observed in all compositions, irrespective of the type of
aluminium powder used in the preparation.
The ignition temperature of air dried compositions was determined in an
electrically heated oven. Data obtained for standard compositions
containing either A, B or D powder (cf. table 1) are shown in table 2.
.
Table 2 Ignition temperature data for different Al compositions
A1 Amount of Ignition delay in min. at T* in oC
powder comp., mg 260o C 325oC 400oC (estimated)
A 172 2.4
A 178 3.0 250
A 481 no igntion 2.5
B 217 no ignition no ignition 1.2 400
D 371 no ignition 3.1 300
D 288 no ignition no ignition 1.1
The D compositions, i.e., compositions containing atomized aluminium powder,
did not burn as vigorously as did the compositions containing flake formed
powder.
Initiation of detonation So far, none of the samples attained supersonic
combustion velocities, i e., they did not detonate.
In order to establish whether the aluminium compositions were at all prone
to detonate, experiments involving initiation by percussion cups were
performed. These experiments showed that neither self-heated nor air dried
samples weighing 10 g or thereabout could be brought to detonation.
However, when a sample weight of 50 g was used detonation occurred.
This happened when A and B compositions and a composition containing an
atomized aluminium powder designated A 80 (Carlfors Brak) were tested.
Discussion
A complete process including (1) self-heating, (2) deflagration, and (3)
detonation was not observed in any of the crude laboratory experiments
described above. However, each of these three types of reaction could be
brought about separately.
When discussing the growth of explosion in solids, Yoffe summarizes
the different stages which can occur between initiation of reaction and
detonation in the following way (4, p. 254).
(1) Initiation of reaction in the solid by a suitable source of energy, e.g., heat,
light, shock, ionizing radiation, etc.
(2) The growth of reaction from this region of decomposition into an
accelerating burning. This can attain speeds up to several hundred metres
per second.
(3) A sharp transition from burning to low-velocity detonation.
(4) Propagation of low-velocity detonation with a velocity in the region of
1000 m s -1,
(5) Propagation of high-velocity detonation at a velocity of about
5000 m s -1 or higher.
In devising a similar scheme for the purpose of discussing the present
experiments, it should be noted that our observations do not require a
distinction between low- and highvelocity detonation. Thus, we recognize
the following partial processes
(1) Self-heating We assume this process to be due to an
ordinary chemical reaction, i.e., a thermal reaction in the
chemical kinetics sense.
(2) Deflagration, We use this term to designate subsonic
combustion in any system
(3) Detonation In our case only low-velocity detonation caused
by high-velocity (supersonic) combustion is believed to have
occurred
A graphical presentation can be found in figure 4 It is clear that this
picture is essentially a rather crude one, However, we have the feeling that
this diagram summarizes the qualitative behaviour of pyrotechnic
compositions (and maybe other systems in the field of combustion and
explosion as well) in a clear and unifying way. For instance, four of the five
points put down by Yoffe can easily be recognized as follows: "(1)"
corresponds to the transition point B (ignition), "(2)" to B - D - E, "(3)" to the
transition point D, and, finally, "(4)" to D _ E (lIigh-velocity detonation -
Yoffe's point (5) - is probably caused by decomposition reactions and not
necessarily by a combustion process; cf. detonation of azides, for instance.)
In discussing our experimental findings reference will frequently be
made to the temperature-time curve in figure 4.
Self-heating: If metallic aluminium takes part in the self-heating reaction it
is to be expected that the self-heating tendency increases with the specific
surface. The smaller the particles the larger the specific surface and the higher
the reactivity. In addition flake formed particles exhibit a larger specific surface
than do spherical ones of comparable size, It was therefcre not surprising to find
that the small-grained powder desigr~ated A (cf. table 1) was the only one that
turned out to give aluminium compositions capable of self-heating,
The difference in particle size between A and B powders comes out fairly
well from the microscope pictures shown in figures 5 and 6. Still more
impressive are the pictures taken with a scanning electron microscope, figures
7 and 8, Here the three-dimensional appearance of the flakes is clearly to be
seen.
Besides powder Al and water, nitrate turned out to be a necessary ingredient in
self-heating compositions, This observation together with the fact that the
self-heating process was accompanied by the evolution of NH3 suggests that
the following reaction is responsible for the heat generation
8 Al(s) + 3 NO3 - + 18 H2O -> 8 Al(OH)3(s) + 3 OH + 3 NH3(g) R,2.
As pointed out by Ellern (1, p, 302) a nitrate in presence of an active metal
such as aluminium may undergo reduction to ammonia if moisture is present,
Reaction 2 is also discussed by Shidlovsky who writes (5, p, 173): " me
stability of nitrate compositions containing metal powders depends strongly
upon the presence of moisture in the composition and, consequently, upon the
hygroscopicity of the oxidizer. From this we may conclude that with the same
storage conditions, the least significant chemical changes will occur in compo-
sitions containing the least hygroscopic salts - barium nitrates - as an
oxidizer",
Whether Reaction 1 is of importance as well cannot be told from our
experiments since we did not analyse for hydrogen gas.
Our result may be compared with that of Wetterholm (6) who found, in a
study of the reaction between aluminium powders and water, that nitrates in
solution as well as ''highly concentrated and sufficiently diluted" nitric acid
passivates the surface of the aluminium particles. Flake formed powders are
inactivated in the manufacturing process, during which protective agents such
as paraffin, stearic acid, etc are added, In his study Wetterholm found no
correlation between specific surface and reactivity, and points at the difference
in the surface layers as a possible explanation. He found flake formed powders
inactivated by paraffin to be very reactive in contrast to those that had been
treated with stearic acid.
Since the systems studied by Wetterholm and by us are substantially
different one cannot really say that a disagreement exists. However, it is not
easy to find an immediate answer to the question why nitrate ions cause a
protective layer (as they probably do) in one case and consume the aluminium
particles with increasing speed in the other.
The reasons why we failed to achieve temperature values in excess of
130 C have already been given. However, the experimental fact itself is
represented by point A in figure 5.
Ignition. The ignition point is indicated by point B in the diagram. In our
opinion a more rapid reaction with different kinetics starts at this point.
In practice, the intersection between the self-heating curve and the
deflagration curve will surely not be as distinct as indicated in the diagram,
However, if an intersection point obtained by extrapolation of these curves
makes sense in a practical case, the ordinate value of this point might be
used as a well defined ignition temperature.
For self-ignition it is doubtful whether a critical size exists, as probably is the
case for detonation, because the properties of the measuring system may
influence in too high a degree,
Deflagration: After ignition we move from B to D in the diagram, provided
the amount of sample is above the critical size. If not, we stop at some
point C This is what happened to our 10 g samples. All we know about the
critical size at present is that it is to be found somewhere between 10 g
and 5o g.
Contrary to the self-heating process, deflag:ration could be brought about
for all of the powders investigated. We might therefore conclude that the
specific surface of the aluminium powder is of secondary importance as far
as burning is concerned.
Deflagration to detonation transition: In terms of the diagram the use of
percussion cups means that we start the process at D (in the same way as
ignition by external means is equivalent to a start at B) Holzman (2, p 418)
introduces the designation "DDT" - deflagration-to-detonation transition for
this point.
Again, we believe this transition to be due to an abrupt change in the
kinetics of the combustion reaction
Detonation: The fact that detonation could be brought about indicates that
the compositions investigated must be regarded as true explosives We do
not believe that a detonation occurring in a pyrotechnic composition ever
can pass over into the high-velocity detonation region (cf. Yoffe's point (5))
because the required self-decomposition mechanism is not likely to be
found there.
Suggested test procedures
In order to avoid accidents in connection with the handling of aluminium
compositions we recommend that the self-heating tendency should be tested,
e.g., with the aid of a simple apparatus in accordance with figure 2 If a tendency
towards self-heating is observed, it is recommended that a coarser grade of
aluminium powder be employed If this jeopardizes the functioning of the actual
pyrotechnic product it becomes necessary to ascertain the period of time the
composition may be kept in a moist condition.
Summary
Finely divided aluminium powder has been shown to impart
self-heating properties to alurniniurn compositions. Selfheating occurs only if
water and nitrate are present. On standing, such compositions may
self-ignite and detonate after about 30 hours or less.
Only freshly prepared compositions have been studied. However,
attention should be paid to the dried final pyrotechnic product, if stored in a
hot and humid atmosphere.
In discussing our observations an idealized diagram has been used in
order to illustrate the partial processes observed in the laboratory and
characterized as self-heating, deflagration, and detonation as well as the
transitions between these processes. The differences in reaction behaviour
are ascribed to a qualitative difference in the reaction mechanisms (the
details of which are not known).
Acknowledgements
We are grateful to the head of this laboratory, Dr P. Ronstrom, for
permission to publish this work and to Messrs. A Bjork, T Eklund, and K,
Ostman for help with the experiments. Our thanks are also due to Mr.
Michael Callow for linguistical revision of the text.
References
1 Ellern H, Military and Civilian Pyrotechnics, Chem Publ Co, New
York 1968.
2 Holzman R T, Chemical Rockets and Flame and Explosives
Technology, DeRker, New York and London 1969.
3 van Tiggelen A, Balanceanu J C, "Oxydations et combustions", Rev Inst
Franc Petrole, 17:7-8 (1968) 1003-1015.
4 Yoffe A D, "The Growth of Explosion in Solids", Proc Roy Soc A 246
(1958) 254-257.
5 Shidlovsky A A, Fundamentals of Pyrotechnics, Picatinny Arsenal
Dover, New Jersey 1965.
6 Wetterholm A, "Aluminiumpulvers reaktion med vatten (The Reaction
between Aluminium F'owder and Water)", in Reaktionskinetik for
explosivarnnen, Foredrag och diskussionsinlagg 28_29 mars 1957, Svenska
Nationalkommitten for Mekanik, Specialsektionen for Detonik och Forbranning,
Stockholm 1958.
|
|
|
mr.crow
National Hazard
Posts: 884
Registered: 9-9-2009
Location: Canada
Member Is Offline
Mood: 0xFF
|
|
Wow so Al can reduce nitrate to ammonia.
8 Al(s) + 3 NO3 - + 18 H2O -> 8 Al(OH)3(s) + 3 OH + 3 NH3(g) R,2.
Therefore my assumption about excess NH4NO3 could be incorrect. Excess NaOH could expose fresh Al and the sodium nitrate could react and create the
ammonia.
Double, double toil and trouble; Fire burn, and caldron bubble
|
|
|
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
