There is lots of stuff in the rec.pyrotechnics archive from google.
I remembered seeing something about blue strobe rockets using silicone. So I looked around and found this:
<a href="http://www.google.ca/groups?q=blue+silicone+rocket+group:rec.pyrotechni cs&hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=20010530053906.15535.00003513%40ng-fo1.aol.com&rnum=4" target="_blank">rec.pyrotechni cs</a>
"
Blue strobe rocket propellant
Source: Greg Gallacci psygreg@u.washington.edu
Comments: The GE silicone II is noted for having an ammonia-like odor, where
the GE silicones smell more like vinegar. The dimensions of the rocket made
with this propellant were 1 1/8 inch ID, with a 1/2 inch core.
Preparation: Mix the copper oxide, PVC and silicone first, in a plastic bag.
Then mix in the ammonium perchlorate. The stuff is said to be somewhat crumbly,
and presses well.
Ammonium perchlorate..............................63
Silicone II.......................................22
Copper(II)oxide................................... 10
PVC............................................... 5
"
I guess this isn't what you were looking for, but the propellant certainly sounds good from the description so I thought I'd bring attention to it.
Edit: Oh, oh! I found something good
:) . Google is your friend:
<a href="http://www.google.ca/groups?q=blue+silicone+rocket+group:rec.pyrotechni cs&hl=en&lr=&ie=UTF-8&oe=UTF-8&selm=392E064A.5AC9B17%40the.sig&rnum=5" target="_blank">rec.pyrotechnics</a>
Silicone II a New Fuel and Binder for Fireworks
presented by Ken Burdick
July 17, 1996
at the Summer Fireworks Festival
Weedsport, New York
WARNING
The information in this paper is based on the limited experience of the
author, using specific tools, ingredients and methods under specific
conditions not necessarily described in full detail. The information
supplied is not represented as being either completely researched nor
totally accurate. The subject highly experimental and not well
understood.
Readers are cautioned that they must form their own opinion as to the
application of anything found herein.
NOTICE of AGREEMENT
By accepting this document, you agree to accept responsibility for
consequences of the use of the information contained herein, and accept
full responsibility for consequences of inappropriate dissemination of
the
information to third parties.
COPYRIGHT
Copyright 1996 by Kenneth Burdick. No portion of this document including
formulae contained herein is to be copied, transmitted, stored in any
retreival system or published without the written permission of the
author.
The COMPLETE document may be copied and transferred to experienced
fireworks experimenters per the conditions in the above Notice of
Agreement
if no fee is involved.
Introduction
This paper presents some initial work on the use of GE Silicone II
(GE280)
as fuels and binders in fireworks. Some of the pros and cons of
silicones
are covered, as well as a partial list of safety concerns. A brief
review
of silicone chemistry is included as a background to the choice of
suitable
silicone materials. This is followed by some pyrotechnic cheniistry,
observations, methods and ideas for how to use silicone compositions.
Since
I have been unable to do any real work on this subject in the last few
years, I hope that this presentation will stimulate others to carry the
concept forward. I believe this material has great promise for certain
applications., and xill prove to be safer and lower cost than some of
the
chemistries it could replace. Critical review, and conunentary are
welcome.
Safety Issues
I have seen no published material on the use of silicone in fireworks.
although it is highly likeiv that militan- research has covered this
area
at least once. A literature search needs to be performed to uncover
previous work. The information presented here is based on verv limited
experiments. seldom involving the testing of more than one device. These
combined facts leave this line of investigation severely lacking in the
knowledge of other's mistakes or in a histon. of safe use. Until some
history is established the experimenter must accept thal in spite of
good
experience to date, we simply don't know whether this stuff is safe.
Hobbyists who don't understand the chernism should avoid this work. or
at
nunimum, restrict his investigation to the GE280 material in combination
Aith perchlorates. Those looking to tn other silicones should, at
minimum,
call the manufacturer and obtain a Material Safetv Data Sheet (MSDS) and
consider possible chemical reactions before proceeding.
In chemical engineering, new chemistries are carefully put through a
series
of experiments where the size of the operation is made Tepeatediv
larger.
for the purpose of revealing unexpected behavior. Most of the mixtures
presented here have onlv been scaled to 10 grams-, a few to 200 grams.
One
known problem of scale is the evolution of methanol and ammonium
hydroxide
during cure. These must be vented to avoid personal injury.
Anunoriium Perchlorate is known to act as a high explosive under certain
conditions. Conditions where severe overpressure can develop showd be
avoided. No guidance is offered here.
Silicone pyro compositions burn hot and fast in the uncured state, and
can
be readily ignited. These compositions also adhere readily to the hands
or
gloves. Combustible quantities must not be allowed to accumulate on the
digits or under the fingernails.
Silicone - Perchlorate mixtures generate hydrochloric acid gas when
burnt.
which is extremely, corrosive to mucosa. Most compositions generate
silicon
dioxide (silica) smoke as a combustion product. Repeated exposure will
result in loss of lung capacitv (silicosis).
Chlorates, perchlorates and chlorides can undergo oxidation-reduction
reactions, yeilding extremely reactive chiontes and hypochlorites. These
reactions can be encouraged by the presence of acids Or high chloride
concentrations. Because of the manufacturing process for silicones, we
must
assume that both chlorides and acids are present unless proven
otherwise.
In abbreviated form these reactions are:
Cl- + C103-<=> ClO- + ClO
Cl- + C104- <=> ClO- + C103-
OK <=> BOOM
Both require little energv to drive to the wrong state. The first is the
reaction of chlorate with chloride. The only reasonable defense is to
avoid
using chlorates. The second equation is the breakdown of Perchlorate.
which
under normal conditions will proceed from right to left. The presence of
something to buffer the pH is desirable. Silicones which evolve acids
are
to be avoided.
Perchlorates can also be reduced bymetals to form dangerous species.
Unless
one has full understanding of the chemistry of the silicone
involved,-the
use of metals is an invitation for disaster. Note that the formulae
presented do not contmn free reducing agents such as sulfur, sulfides
and
metals.
Cl04- + Me => Cl03- + Meo
Cl03- + Me => Cl02- + MeO
Many of the formulae presented contain black copper oxide CuO. Mixture
with
chlorates presents the possibility. of creating copper chlorate, a very
unstable explosive.
When pressing rockets. don't place vour hand over the rammer, in case
the
composition ignites.
Pressing rockets with hydraulic rams might confine combustion products
during accidental ignition, resulting in an explosion rather than just a
fire.
Unless vou can handle the consequences, don't make flash powder by
accident
- such as bv combining prechlorate with sodium benzoate before mixing
with
the silicone.
For more information on safety with fireworks, please see the United
States
Naval Ordnance Laboratory, publication NOL TR 61-138 or-its successor.
Advantages and Disadvantages of Silicones
Perhaps the most g advantage of silicone is that it is verv unreactive
at
normal ambient temperatures. Most silicones are totally stable at 300
degrees Fahrenheit and some as high as 500 degrees. It is somewhat
surprising that it burns at all, but is in fact a fairly energetic fuel
in
pyrotechnic compositions. Some formulations are difficult to ignite,
while
others take fire readily.
Like Parion, silicone rubber can serve as both a fuel and a binder.
Unlike
parlon. no solvents need to be added. but the silicone rubbers generally
lack halogens chlorine and fluorine. Silicones burn colorless but
somewhat
luminous like metal. All types tried so far are free of sodium which
would
interfere with color puritv.
The chemical properties, lubricity of uncured material, and elasticicitv
of
cured silicone rubber seem to reduce sensitivity
of compositions during and after manufacture. The water resistant nature
of
the silicone also serves to encapsulate the grains of the composition,
not
only protecting them from water, but slowing reactions such as ion
exchange
between salts.
Good pyrotechnic mixes also usually have good physical properties for
processing. Ten to fifteen percent silicone compositions tend to have a
dry
crumblv texture, and can be pressed into objects which hold their shape
during and after cure. Compositions with 15 to 25 percent silicone
usually
have a workable puttv-like plastic texture, and probably can be pumped.
These could be used in low cost, high volume automated production. After
cure, they can be bent and struck vathout loosing physical integrity.
The
lubricity of the uncured material allows devices such as rockets to be
easily removed from forming tools.
On the negative side, many compositions generate large amounts of smoke.
but some do not. Silicone also has a no-stick property like teflon. That
is
nothing wants to stick to it. This makes priming cured stars extremely
difficult. The water resistance and chemical stability also assure that
any
unconsummed iters such as unlit stars will remain as reactive
pyrotechnic
compositions indefinately.
Chemistry of Silicone Manufacture
The manufacture of silicones is a multi-stage process which involves
several foreign materials which could find their way into the finshed
product. There are many processes. The following is typical.
Silicon metal is reacted with chlorinated hydrocarbons with a catalyst
to
form single units which can be polymerized This forms compounds with
varying degree of chlorine saturation. so the result is distilled to
obtain
the desired product shown.
Si + RCI => SiCl2R2
The "R" in the equation above represents a hydrocarbon radical such as
-CH3- The type or types chosen for R detemine the chemical and physical
properties of the final product. These become the side chains of the
polymer in the final product. A catalvst such as copper is used in this
reaction. The result typically dimethyl dichloro silane.
The dimethvl dichloro silane is then hvdrolyzed by adding water.forming
short chains and rings.and evolving hydrochloric acid.
The rings are broken and the chains further polymerized by the addition
of
catalyst. Strong acids like sulfuricc and nitric acid, strong bases like
sodium hydroxide, amides and other materials are used for this purpose.
The
result is a silicone oil or grease with chains terminated with hydroxvl
groups. These terminations may be replaced to alter the curing
chemistry.
For a silicone grease, this is the final product.
Silicone Rubber Curing Chemistry
The curing of silicone rubber is similar to building of Tinkertovs. The
chains described above are the green sticks, except that they bend
easilv
and are very long. The round knob of the Tinkertoy corresponds to a
cross-linking material in the silicone. This material is typicallv a
silicon atom with three or four reactive groups attached. Thus the
tinkertoy knobs only have three or four holes During reaction the sticks
get connected to the knobs until most of the ends are connected. That
leaves some unreacted groups which might be a problem if they were
present
as a large percentage.
The cure reaction of one-part silicones is initiated bv the addition of
atmospheric water vapor. This reaction is where various acids, bases.
esters, alcohols and such are released during the cure. These by
products
are what gievs a silicone its characteristic odor. Some typical
reactions
are as follows
-Si-O-CH3 + H20 => -Si-OH + CH30H
-Si-NH, + H20 => -Si-OH + NH3
The first equation is howthe methanol byproduct is generated. The second
is
my conjecture as to how the aammino is generated bv GE280. Note that
both
reactions leave hydroxyl groups attached to the exposed silicon atoms.
The hvdroxyl groups then react using a catalyst yeilding water. This
catalyst is problematic to the pyro, because it is often listed as a
trade
secret ingredient by the manufacturer. It also may remain in the cured
product.
-Si-OH + OH-SI- => -Si-O-Si- + H20
The cured product typically does not have any carbon-carbon bonds and
few
reactive groups remaining, which makes it stable when attacked bv strong
chenucals, heat or LTV light. Possible chemicals which remain in the
cured
product are
copper
HCI, H 2so 4, NAOH. KOH. etc
unreacted organic groups from the crossings catalysts
cure reaction bvpr(>ducts ( animonia and methanol for GE280 acetic acid
for
traditinoal RTV any possible reaction products of the above
Choice of Materials
There is a wide selection of silicone materials avaiable from greases to
nibbers to casting resins. Each has its own manufacturing process and
chemistry, and they should not be assumed to be equal. The
manufacturer's
Matenal Safety Data Sheet (MSDS) lists the ingredients in a product
although some mav be listed as "industry secret". The MSDS, the
manufacturer's technical support engineers. and basic silicone chemistry
texts should be used to gain information about the chemistry of a
material.
The chemistry of the silicone should then be considered in respect to
the
other ingredients.
Because of the possible reactions when oxidizers are combined with
acids,
silicones which evolve acids should not be used. DO NOT USE RTV WMCH
GIVES
OFF A VINEGAR ODOR- Silicones which give off weak bases like ammonium
hydroxide are interesting because they buffer the pH to about 10 during
cure, and can neu any acids which might be present in the unreacted
silicone. Silicones which evolve alcohols are probably least reactive,
and
should be sought out. RTVs made for electronic circuit assemblv are an
area
for future research. because thev don't evolve acids or bases. I would
avoid anything which gives off amides. keytones or species with double
bonds. General Electric GE280 gives off methyl alcohol and ammonia
(which
is quicklv hydrolyzed to ammonium hydroxide).
Silicone casting compounds are chemically more complex than RTVS. They
niav
contain carbon-carbon bonds, or other opportunities for chemical
reactions.
Silicone greases, especially electronic or medical grades, offer the
promise of a totally stable fuel. They have no cure process, which means
thev won't serve as a strong binder, but may work well in rockets.
Fuel resistant grades may containing side chains loaded with halogens -
fluorine more likely - but possibly chlorine. These could provide
additional halogen donors for color compositions.
Users not able to understand the chemistry or accept the risk of
spontaneous combustion should restrict their choice to materials with
some
history of use (currently GE280). This, however, is no guarantee of
safety.
The GE280 material can be positively identified by comparing the list of
ingredients on the container to the listing in the appendix.
Other Ingredients
Other fuels should be chosen carefully to avoid unwanted
oxidation-reduction reactions with the oxidizers. I personally do not
understand the chemistry well enough to risk using metals or sulfur or
nitrogen bearing fuels. Polymers such as Parlon. PVC and polvethylene
are
at the other extreme and are probably quite safe to use.
For Oxidizers, only potassium and ammonium percholorates are considered
acceptable at present. Potassium nitrate simply does not burn well.
Other
nitrates have not been tried. CHLORATES SHOULD BE AVOIDED. because it is
too easy for chlorate to proceed through the exothermic redox reaction
veilding chlorites.
2(Cl03-) <=> Cl02- + Cl04-
Chlorate samples made for safety testing were more sensitive to hammer
blow
on an anvil than perchlorates. The burn rate for potassium chlorate was
actually equal to to slower than that for equivalent ammonium
perchlorate
mixes.
Sulfates are the safest color donors, but did not generate good colors
in
limited testing. It is unknown why these did not work, possiblv there
was
inmfficient temperature, or the metal oxides generated dissolved into
the
slag forming a glass. Black copper oxide (CuO) and sodium benzoate can
hardly be considered safe, but they work. Copper oxide has a tendencv to
oxidize things around it at room temperature. The formulae listed here.
however, lack anything which can be oxidized. The copper oxide is
ceramics
grade (very fine). It is not known which process it was made with. but
is
free of acid when tested with pH paper.
Chlorine donors should be restricted to polymers (PVC and Parlon) plus
totally halogenated hydrocarbons. Chlorides should be avoided.
Characteristics of Pyrotechnic Formulations
As mentioned above, potassium chlorate is not considered suitable. and
was
not investigated in anN. detail. The chlorate @xes would snap when
pounded
on an anvil with a force which might be used to drive a large finishing
nail. Nitrate mixes required a severe blow to cause a reaction, but were
nearly refractory in reaction. For some reason. the silicone and nitrate
do
not want to react.
Potassium Perchlorate
Potasium Perchlorate creates slow burning compositions which might be
suitable for gerbs and lance. Smoke tends to be a problem, though.
Excellent blues and good yellows can be created with simple
compositions.
Some blues were tested as comets and in aerial shells, but they tended
to
wash out to a purplish blue, and did not look quite as good as then. did
on
the ground.
The base fire for silicone assuming complete balanced reaction with
silicone is represented in its approximate, simplified form is
2(KC]04) ' SiOC2H6 => 2(KCI) + Si02 + 2(CO2) + 3(H20)
That is a 78.9 % to 21.1 % weight ratio
The silicon accepts a second oxygen creating silicon dioxide (sand)
mostly
in a molten state, forming a slag. The potassium tends to hold onto its
chlorine, so a chlorine donor is needed. During the burn. it is
plausible
that methn I groups are initially left unreacted. leaving the slag as a
gas. They can burn in the atmosphere Aithout consuming Oxygen from the
oxidizer. This means that higher fuel loadings may be possible.
For blue compositions. the copper oxide serves as an optional oxidizer,
sometimes being reduced. and sometimes not. The presence of elemental
copper can easily be detected in ash of a test, indicating that the
copper
oxide is being reduced. The base fire when CuO is fully reduced is
8(CuO) + SiOC2H6 => 8Cu + SiO2 + 2(CO2) + 3(H20)
giving a 89.8 % to 10.2 % weight ratio.
PVC is used for the chlorine donor here. For PVC with KCl04 the base
fire
is
5(KCl04) + 4(CH2CHCI) => 5(KCl) + 4(HCI) + 8(CO2) + 4(H20)
giving a 73.5 % to 26.5 % weight ratio.
The triangle diagrams show the range of blue color generation with 10 %
content of copper oxide. 5 % CuO was also tested yeilding a similar
pattern. but with inferior color quality. The nomenclature of Shimzu in
Pyrotechnica VI is used, and his formula B11 was used as a standard to
judge results. Shimizu's data corresponds with mine along the right hand
edge of the triangle, where the silicone content is zero. His results
are
shown there for reference. Little data was taken on burn rates. The
fastest
point measured was at 36 sec/inch for 75:10:10:5 ox:sil:PVC:CuO.
decreasing
for points below there on the diagram
[Image]
A few formualae were tried for yellow. Attempts to correct the color
using
barium sulfate werenot successful. The following formula (KY07) gives a
good sodium yellow using sodium benzoate-. KCl04 (70) NaBz (10) SiOC2H6
(15) Charcoal, air float (5). The burn rate is an impressive 14.5
sec/in.
Note that Perchlorate plus benzoate is basically a whistle mix. Safe
procedure would be to mix in the benzoate after the other ingredients
have
been combined with the silicone.
Ammonium Perchlorate
Ammonium percworate seems to be the oxidizer most suited to use with
silicone. It burns fast, acts as a chlorine donor, and has a tendency to
strobe. The compositions also takes fire more readily than with
potassium
perchlorate.
The base fire for ammonium perchlorate with silicone is
16(NH4Cl04) + 5(SiOC2H6) => 8(N2) + 16(HCI) + 39(H20) + 10(CO2)
giving a 83.5 % to 16.5 % weight ratio. If the hydrogen is not oxidized,
the ratio becomes 66.5 % to 33.5 %. The oxidizer releases its chlorine
largely as HCI, which is good for color generation. If PVC is used as a
fuel and chlorine donor, the base fire for that reaction is NH4Cl04
(79.3%)
PVC (20.7 %).
Copper oxide is used as a coloring agent in these formulations to make
blues. The copper oxide also serves as a catalyst and sensitizer as
shown
by the following results. The samples have a constant 72:28 ratio of
perchlorate to silicone, with varying percent of CuO. Ignition is tested
with a thin layer of 5FA resting on the surfface. As the copper content
increases. so does the burning rate and ease of ignition.
CuO 5FA Ig. Rate sec/in
NH4WO1 0 0/3 14.8
NH4BO1 5 1/2 13.8
NH4BO2 10 2/2 13.5
The triangle diagram shows the combination of ammonium perchlorate.
silicone and black copper oxide. No chlorine donors are added. Some
samples
showed moderate sensitivity when hammered on an anvil, others were quite
insensitive. The line drawn across the diagram shows complete oxidation
and
reduction per the theoretical base fires. Note that good colors can be
found well on the fuel-heavv side of the line.
[Image]
The second diagram shows burn rate as measured in a I" lance tube. There
is
a large area of nearly constant burn rate which roughly corresponds
.vith
blue color generation. There is probably some physical process such as
melting of the oxidizer which is controlling this rate.
[Image]
The third diagram shows unstabe burning characteristics observed during
small sample testing. Samples which exhibited noticeably unstable
burning
are marked with "S" for strobe. Where there seemed to be a strobe rate,
that number is marked. It is doubtful that any of these would pass for a
blue strobe star, although the thought is tantalizing.
[Image]
No aerial shell testing was done. No clear conclusion was reached from
this
data.
Some samples when burned in the uncured state created an excellent long.
rigid snake-like ash. Data on this property is sketchy, but some mixes
which exhibit the property are NH4B01 (above), NH4B10: NH4Cl04 (20) CuO
(50) SiOC2H6 (30) and NH4Bl9: NH4Cl04 (30) CuO (40) SiOC2H6 (30). This
has
the drawback that large volumes of HCI gas are generated. Perhaps a
potassium Perchlorate / copper oxide / silicone grease/ gassing agent
can
be found which doesn't generate HCI.
Some compositions generate a large amount of smoke, while others
generate
relatively little. This seems to depend on the amount of silica which
remains in the slag. The low smoke comps may produce a good low-smoke
lance.
One mixture containing PVC (in an attempt to improve color) NH4BO6 was
pressed into a large rocket. This gave a nice strobe effect and had good
lift and color, despite the fact that the mixture is very fuel rich.
Only
one device was testecl but some more may go up in Weedsport. Formula for
NH4BO6: NH4ClO4 (63) CuO (10) SiOC2H6 (22) PVC(5). The burn rate is 14
sec/inch.
Methods
The dry ingredients should be sifted together 2 to 3 times through a
screen. The silicone is then combined with them. If a hot fuel like
benzoate is being used, it is probablv best to add this after the
oxidizer
is combined with the silicone
During the mixing process, carefully avoid buildup of composition on
vour
hands or clothing. Clean anvthing off immediately with a paper towel to
avoid the possibility of very serious burns. Don't allow composition to
fall on the floor as it will rapidly become part of the floor, and be
nearly impossible to remove.
Small test batches around 10 grams can be niixed bv folding the material
over on itself, first with a wooden spatula. then by folding the
composition over and pressing it flat.
Larger batches around 200 grams can be placed in a Ziplock bag and
kneaded.
Leave a moderate amount of air in the bag. Drop in mixed dry,
ingredients
first, then add the silicone. keeping it away from the corners and the
zipper. Work the powder into the silicone until all is dampened. Then
remove the material sticking to the mfface of the bag bN rubbing other
composition across it.
Rocket Construction
It is not known whether a nozzle is required for this type of rocket. If
it
is, then the lower part of the spindle must be made of metal, Otherwise,
the whole tool may be made of hardwood like maple or cherry. The
silicone
compound is easilv removed from the spindle, so essentially- no taper is
required above the nozzle. This makes the tools easv to build. and also
results in all of the composition burning out at the same time. My tool
has
a taper of about O.02 inch over3-
1/2 inches. The ramming tool seems to work ok if it has some clearance
to
the spindle.
The one sample tested had a tapered nozzle from 0.5" to 0.7" diameter
over
a 0.7" length. The bore through the composition was slightly less than
1/2
inch, and 3.5 inches deep. The composition covered the top of the
spindle
by 0.4". The tube inside diameter was 1-1/8 inch.
The nozzle can be rammed with about 10 % silicone added to the cla-y.
This
makes it remove much easier from the spindle, but seems to reduce
push-out
strengh of the nozzle bv about one third.
Compositon should be broken up by rubbing it through a wire mesh. so
that
it doesn't get stuck on top of the spindle. Increments should be kept
small
to avoid trapping large air pockets between consolidated composition.
The
material does not need to be rammed hard. A long series of light taps
will
drive out the air and leave solid composition. It takes a certain amount
of
time for the air to leak out between the grains, and ramming hard does
not
help. Composition adhering to the inside of the rammer is evidence that
the
increment was fulIy rammed. Never place any part of the body over the
rammer as this will become a high velocity,projectile when premature
ignition occurs.
The spindle is easilv extracted. except for the fact that a partial
vaccuum
is formed at the top. If pulled too rapidly a large movement of the
composition mav occur. A large amount of spinning with a small amount of
pulling will allow air into the space. A small hole through the length
of
the spindle would probably help a lot.
References
Rochow. E., Silicones, in Modem Chemistry for the Engineer and
Scientist.
McGraw-HilI Companv. Inc., 1957. pp 211-231.
Shimizu. T., Studies on Blue and Purple Flame Compositionskfade Ifith
Potassium Perchlorate. Pvrotecnica VI. Pyrotechnica Publications, Jul-v
1980. pp 5-2 1.
United States Naval Ordnance Laboraton Explosives, Propellants and
FINrotec@c Safet-, Covering Laboraton. Pilot Plant, and Production
Operations, NOL TR 61-138, Oct 20, 1961-1 US, US Dept of Commerce.
publication No. AD 272 424.
EDIT: Shortened Link - Zaibatsu
<small>[ April 10, 2003, 09:56 AM: Message edited by: zaibatsu ]</small>