Flash! Bang! Whiz!

An introduction to propellants, explosives, pyrotechnics and fireworks


  1. Introduction
  2. Reactions and Heat
  3. Black Powder
  4. Fulminate
  5. High Explosives
  6. Smokeless Powder
  7. Aromatic Explosives
  8. Pyrotechnics
  9. Fireworks
  10. Safety
  11. References


This article reviews the technological use of exothermic chemical reactions that release their energy in a very short time interval. There are three primary fields of application for these effects: propellants, explosives and pyrotechnics. Propellants create a high gas pressure for driving projectiles or rockets and for similar uses. Explosives create a disruption of solid or liquid bodies, as in construction, mining or warfare. Pyrotechnics have effects that are mainly sound and light, but include many other varied applications, mainly on a small scale. Fireworks is an application that is entertainment--a show of light, noise and motion. "Fireworks" is an almost exact translation of the Greek roots of "pyrotechnics." Perhaps "pyrotechnics" should be reserved for the serious applications and "fireworks" used for the entertainment side.

The chemical reactions we shall consider are reactions between an oxidizer that supplies oxygen or receives electrons, and a fuel that combines with the oxygen or releases electrons, and is a reducing agent. These two functions may reside in the same molecule, or in different molecules. Some constituents, such as sulphur, may serve as either a fuel or an oxidizer. In any case, both functions are present in every explosive (a general term for all three kinds of applications), and the oxygen of the atmosphere plays no role as an oxidizer, as it does in combustion or respiration, two other chemical sources of energy. For rapid reaction, the oxidizer and the fuel must be intimately mixed before the reaction occurs. In some cases, atmospheric oxidation may play a minor role.

The reactions we use must certainly be spontaneous, yet must not begin until the proper moment, even though all necessary ingredients are in close contact. This means that there must be some energy barrier to initiation of the reaction, which will not occur until this energy is supplied. This is only meant in a general sense; no such unique barrier can be identified. The agents will exhibit a range of sensitivity from the exquisitely sensitive that will be set off by the slightest shock, such as the decomposition of NCl3, to an almost total insensitivity, like that of TNT. In most cases, we search for a degree of sensitivity that will not respond to the usual shocks of handling and transport, but will detonate reliably when a definite stimulus is applied.

The energy to start a reaction may be supplied by impact, friction, heat, flame, spark, radiation, shock wave or deformation. Each explosive agent has its own set of sensitivities to the various stimuli, so there is no single detonation energy that can be supplied by multiple means. The device supplying the initiating reaction is called the detonator, initiator, primer, first fire or some other descriptive name. The reaction in this device then initiates the main charge. A match is a simple example. The head of the match is the first fire, lighting by friction. Its heat then ignites the wood of the match, which represents the main charge. Often the "first fire" is not the initiator, but is ignited by it and strengthens its effect.

Nuclear explosions will not be included here, but they are precisely analgous to chemical explosions, in that the energy is liberated in a very short time interval by a very exothermic reaction. There are also mechanical "explosions" where the energy transfers occur in brief intervals. These, also, will not be included.

Reactions and Heat

A chemical reaction can be described by a balanced chemical equation, such as C + O2 → CO2. This means that one atom of carbon reacts with one diatomic molecule of oxygen to produce one molecule of carbon dioxide. In masses, 12 g of carbon combines with 32 g of oxygen to make 44 g of carbon dioxide. Of course, here we are taking the weight of a fixed number of atoms or molecules, 6.02 x 1023 of them. The ratio of 12 g of carbon to 32 g of oxygen is called stoichiometric, meaning that nothing is in deficiency or excess for a reaction according to this equation.

The writing of an equation does not necessarily mean that the reaction will take place in that way. For example, carbon could react as 2C + O2 → 2CO, producing carbon monoxide instead. Here, 24 g of carbon react with 32 g of oxygen to make 56 g of carbon monoxide. Also, the equation may not show the mechanism of a reaction, or intermediate states. Reactions may not go to completion, but reach an equilibrium state with all of the reactants and products present. Furthermore, just because a reaction will go does not mean it will go rapidly. In fact, it may not proceed at all unless a catalyst is present, or some other necessary condition is established.

In the first reaction above, 94 kcal of heat are evolved for each gram-mole of carbon (12 g) that reacts. This is the heat of reaction. In this case, CO2 is being formed from its elements, so it is also the heat of formation of carbon dioxide. We presume that the carbon, oxygen and carbon dioxide are all in some standard state when we determine the heat. If they are hotter or colder, or in some other form, corrections must be made to the heat of reaction. We shall not bother with such refinements here, since they are rarely necessary for general conclusions, and shall also take rounded values for the heats. A reaction that evolves heat, as this one does, is called exothermic. In explosives and pyrotechnics, we are concerned mainly with exothermic reactions.

The heat of reaction Q is the decrease in the enthalpy H of the system, if the reaction takes place at constant pressure, or Q = -ΔH. Enthalpy changes are what are given in handbooks, so they are the negatives of the heat of reaction. For the formation of CO2, then, ΔH = -94 kcal/mol. An exothermic reaction means a negative ΔH, while an endothermic reaction has a positive ΔH. Some compounds, potentially unstable ones, have negative heats of formation, or positive enthalpy changes, which are the same thing.

The heat of formation of carbon monoxide, in the second equation, is 26 kcal/mol. Now suppose we are burning carbon monoxide, and want to find the heat of reaction. The equation is 2CO + O2 → 2CO2. Imagine that we make each side from its elements. To make the left-hand side, we get 26 kcal/mole, as just stated, or 52 kcal. To make the right-hand side, we get 94 kcal/mole, or 188 kcal. By the conservation of energy, we will get the difference, 188 - 52 = 136 kcal, when we burn two moles, 56 g, of carbon monoxide. We can say, then, that burning a mole of carbon monoxide to the dioxide gives us 68 kcal. This can be done with any equation, so long as we know the heat of formation of each molecule involved.

Let's take another common reaction, 2H2 + O2 → 2H2O. Per molecule of water, the heat of formation is 58 kcal/mol if the water is left as a vapor. If we condense the water to liquid, we recover the heat of condensation as well, so the heat of formation is now 68 kcal/mol. When water is involved, we always have these two choices of the final state, and so the heat values will differ by 10 kcal/mol depending on which we choose.

So far we have worked on a per mole basis, but it may be more convenient to work on a per kilogram basis. In burning hydrogen to water vapor, we get 58 kcal for each 2 g of hydrogen. This is 1000/2 = 500 mol per kg, so the heat is 500 x 58 = 29,000 kcal/kg. In burning carbon to carbon dioxide, 12 g of C gave us 94 kcal. A kg is 83.3 mol, so the heat is 83.3 x 94 = 7833 kcal/kg. In both these cases, we assumed that the oxygen came free from the air. If we include the weight of oxygen as well, then 18 g of hydrogen and oxygen give us 58 kcal, so the heat per kg is 3220 kcal/kg. In the case of the carbon, we find 2140 kcal/kg. From the equation, then, we can figure out the heat evolved in a reaction per kg of reactants.

It is very handy to use a concept called formal charge to keep track of electrons in a molecule. The oxygen atom is assigned a formal charge of -2 (as if it gained two electrons, which it would like to do), and hydrogen a formal charge of +1 as if it had lost its electron. We make the rule that in a molecule, the net formal charge is zero (if the molecule has remained neutral). In H2O, then, we have +2 for the two H, and -2 for the oxygen, so 2 - 2 = 0 and all is right. In CO2, the two oxygens have formal charge -4, so the carbon must be +4. The carbon has not actually lost four electrons, and the oxygen has not gained four electrons, but they are shared. In CO, the carbon must have formal charge +2, since the oxygen is -2. This is just accounting, but is of some help in understanding what goes on in reactions. In an ionic reaction, the formal charges are actual charges and reflect valence. For example, in NaCl, the Na has formal charge +1, the Cl -1, and these correspond to the actual ionic charges in the crystal. In Cl2, formal charge is meaningless, as it is in CH4. However, we usually say that Cl has formal charge 0 in the first case, and C formal charge -4 in the second case if H has +1, but these are not actual charges. There may be fractional charges if the electrons are not symmetrically distributed in the covalent bonds.

Now let us consider a pyrotechnic reaction used to produce heat and "flash," KClO3 + 2Al → KCl + Al2O3. KClO3 is potassium chlorate, a powerful oxidizing agent, and Al is powdered aluminium. This is a relatively dangerous mix, that can be set off by heat or shock. The heat of formation of potassium chlorate is 94 kcal/mol, and the heat of formation of potassium chloride, KCl, is 104 kcal/mol. The heat of formation of aluminium oxide is 385 kcal/mol. Therefore, the heat of this reaction is 385 + 104 - 94 = 395 kcal/mol, which is quite large. The formal charge on the chlorine goes from +5 to -1, while that on the aluminium goes from 0 to +3, for two atoms. The chlorine is said to be reduced, while the aluminium is oxidized, on the basis of the direction of the change of formal charge.

The fact that a reaction evolves heat when it proceeds does not mean that it is spontaneous. Chemical reactions are driven not by simple energy, but also by entropy, and seek for a minmum of the free energy U + PV - TS, not just a minmum of the energy U alone. In fact, the heats of reactions are changes in the quantities H = U + PV, the decrease in H (the enthalpy) when the reaction takes place. We see that if the entropy S decreases in a reaction, then the free energy is raised and the reaction is less favorable. A reaction can proceed spontaneously even if absorbs energy, provided the entropy increase is large enough. An example is the evaporation of a liquid. It takes energy to boil off some liquid, but the increase in entropy is very satisfying. With explosives and pyrotechnics, the entropy increases are usually substantial, so most of the reactions go easily to completion.

Black Powder

Black powder was the sole propellant, explosive and pyrotechnic agent for 500 years, from 1300 to 1800, and is still in use for certain applications. It is a unique and fascinating compound chemically, technologically and socially. It was invented as a pyrotechnic substance, then applied as a propellant in firearms, and finally used in engineering and mining. The history of black powder and firearms is treated in Cannon. Some authors make assertions about the history of black powder that are not supported by good evidence, and should not be accepted without better proof. An egregious assertion is that Chinese alchemists experienced a black powder explosion in 220 BCE. There is no evidence of "black powder" in China, and this is about 1200 years before nitrates were first discovered and used, according to more reliable sources. The great Chinese invention was pure nitrates, which they used in pyrotechnic devices, arrow throwers and rockets. The invention of black powder is shrouded in mystery; neither Roger Bacon nor Berthold Schwartz invented it, but high-nitrate powder is probably a European invention. Black powder is not a simple mixture of nitrate, charcoal and sulphur.

The composition of ordinary black powder is 65-75 KNO3, 15-20 C, 10-15 S, which is close to the "stoichiometric" ratio of 84:8:8 that gives the ideal reaction 10KNO3 + 8C + 3S → 2K2CO3 + 3K2SO4 + 6CO2 + 5N2. The heat released is 685 kcal/kg, and the volume expansion factor is 5100. The solid products make the characteristic white smoke. The actual reaction depends on the exact constitution of the powder, how it is prepared, and how it is detonated. The density of gunpowder is about 1.04 g/cc. Black powder is the safest of all explosives. It is insensitive to shock and friction or to electric spark. It must be initiated by heat or flame. Moisture renders black powder useless, and drying does not restore its properties.

The nitrogen in KNO3 has a formal charge of +5, which is reduced to 0 in N2 (in such molecules the formal charge is taken as zero, its average value). The carbon is oxidized from 0 to +4 in CO2 and the carbonate, and the sulphur from 0 to +6. KNO3 is a stable and safe oxidizing agent, not capable of explosion on its own. Black powder is a very stable explosive, insensitive to shock or friction, but sensitive to heat and flame. Like all explosives, it supplies its own oxygen and does not rely on the atmosphere. Note that it is much less efficient as a heat source than carbon and oxygen, which gives 2140 kcal/kg. Its utility lies in its ability to furnish its energy in a very short time, while the carbon will take a good while to burn.

How the powder burns is affected by the grain size. The larger the grain, the slower the powder burns. Fine powder is used for blasting, small grain for firearms, and large grain for cannon. A large variety of black powders are manufactured, and each type has a special designation and use. Black powder is essentially a propellant that burns at a rapid but finite rate determined mainly by its temperature. It is often said that gunpowder will only burn in the open, but explodes when confined. This is much too simple a statement. When in the open, the unburnt powder never becomes hot enough to burn rapidly. When confined, as in a firecracker, the powder quickly becomes hot enough to burn very rapidly, releasing all the energy in a very short time, quickly enough to make a loud report. Pressure does raise the rate of burning, but gunpowder has the least pressure effect of any common explosive, and for this reason is gentle to guns. A thread of gunpowder, wrapped in paper or other covering, burns at a slow and reliable rate, making a delay element or fuse.

Because of its safety and reliability, pressed black powder is used as the propellant in small rockets. A powder for this service has less KNO3 and S, and more C. Its rate of burning can be slowed with chalk, wax or talc. A typical mix is 91 black powder, 9 chalk. No more than 3% of the powder can be stopped by a #20 sieve (0.84 mm) and no less than 60% must be stopped on a #40 sieve (0.42 mm). It is compressed to 1.82-1.89 g/cc, and contains 1.8%-2.5% moisture. This propellant grain is burned in a chamber with a ceramic choke in army signal rockets, which reach 700 ft. altitude. A bursting charge expels 5 white stars that free-fall, or else a red star with parachute, that burns for about 50 sec. and falls at 10-15 ft/s. A model rocket has a pressed black powder propellent grain, and a granular black powder ejection charge. There is a delay element between the two charges, so that the rocket coasts to its maximum altitude before releasing the payload. The fuel for solid-fuel rockets, though called the "powder grain," is a cast plastic cylinder of the fuel material. The word "grain" does, in fact, seem to come from the grains of black powder that are used in a pressed charge, and has been transferred to the whole fuel assembly of any type.

Black powder is an oxidizer--fuel mixture of the type we shall discuss at more length under pyrotechnics. The sulphur and charcoal are first ground together, so that the thixotropic sulphur coats the colloidal charcoal intimately. Then the nitrate is mixed in by wet grinding. The nitrate produces oxygen to oxidize the sulphur and carbon, catalyzed by the large active surface of the charcoal, while releasing the nitrogen with the evolution of heat. The reaction begins at a temperature where there is a change in the crystal structure of the nitrate, which creates lattice defects that encourage the solid-state reaction.

War rockets were not extensively developed in China, and were used only incidentally in the West. Rockets for pleasure pyrotechnics did, however, become widely used, and were the basis for later war rockets. William Congreve developed his war rockets in the late 18th century, but they were only successfully used first in 1807 at Copenhagen. They were difficult to control, and not very effective. However, they could be fired with a light launcher instead of a heavy cannon, a principle later extensively applied. The last major use of Congreve rockets was in the Zulu war of 1879. Rockets appeared again in the Second World War for use in mass flights from landing ships and to support infantry. They also were used in the recoilless rifles and antitank rocket launchers that are still valuable, providing powerful artillery without the weight and recoil.


In 1799 or 1800, Edward Howard discovered the second explosive substance to be found, mercuric fulminate. The name comes from Latin fulmen, fulminis, "lightning." To make fulminate, dissolve 1 part mercury in 13 parts nitric acid (sp.gr. 1.36), then pour the solution in 8 parts of alcohol. A vigorous reaction occurs, with frothing and emission of nitrogen oxides. Finally, the fulminate separates as small white needles. The solution is filtered, and the precipitate thoroughly washed until all the acid is gone. The salt is dried, but should remain moist until it is used. Do not, incidentally, do this at home! The crystalline salt is heavy, 4.42 g/cc, and explodes when exposed to flame, impact or friction. It produces sufficient flame to initiate black powder. Percussion caps were made from about 1815 in which a little fulminate, perhaps with a little chlorate, were sandwiched between thin sheets of copper. When struck by a firing pin, they released a puff of flame that could discharge a firearm. The pinfire breech-loading shotgun appeared in France in 1836, and the "needle gun" in Germany in 1841. Percussion caps displaced the flintlock only gradually.

Justus Liebig studied the fulminates in 1822, determining that the formula for mercuric fulminate was Hg(ONC)2. The fulminate radical ONC- is an isomer of the cyanate radical, NCO-, but very different in behavior. Cyanates are stable (and nonpoisonous), while all fulminates are explosive. Note that in the cyanate, the carbon is in the middle, and we can write the valence structure -N=C=O where all the atoms show their usual valence. For the fulminate, we must write -O-N=C, and the carbon is left with only two bonds, instead of the usual four. The actual behavior depends on the electronic structure, of which this is only a crude representation, but the result is the same. Mercuric fulminate decomposes by Hg(ONC)2 → Hg + N2 + 2CO (or CO2 + C), so that all the products are gases or vapors. Mercuric fulminate "puffs off" at 195°C in a test. It is easily detonated by a hot wire.

It is said that the alchemist Johann Kunkel von Löwenstein tinkered with fulminate in the 17th century. This fact was discovered only long after Howard's discovery had been put to practical use, and Leibig had studied fulminates. Other discoveries have similar precursors, only found after the fact and often by enthusiastic misinterpretation of the sources. Sometimes isolated Russians and Americans, both ingenious peoples, did make earlier discoveries at least partially like later and better-known ones that were not only conceived, but put to practical use, but many are only the results of wishful thinking. William Kelley of Kentucky may have refined pig iron by blowing air through it, but only Henry Bessemer solved the difficult problems of making the process a practical one.

Radicals like ONC- that are associated with explosive potential are called explosophores. We shall see that NO2-, NO3-, N3-, ClO3- and ClO4- are other examples frequently seen. The nitrogen radicals seem to release active oxygen when N2, a very stable molecule, is formed. The chlorates and perchlorates are also easy sources of oxygen, the chlorates more so than the perchlorates. Explosive reactions are usually solid-state, or at least liquid-state reactions, in which the crystal structures and molecular neighborhoods play an important role. Free radicals (groups with unpaired electrons) and chain reactions are probably important. The mechanisms of explosive reactions are by no means well known, and those that are known are often surprising.

The azide radical, N3-, with a structure something like -N=N≡N, releases N2 to give the free radical -N, which can do further execution. Mercurous azide, HgN3 decomposes according to 2HgN3 → 2Hg + 3N2, giving only mercury vapor and nitrogen. It is one of the rare explosives that contains no oxygen. Lead azide, Pb(N3)2, was discovered in 1891 by Curtius, and is a useful fulminating substance, less sensitive than mercuric fulminate. Indeed, it cannot be exploded with a firing pin or a safety fuse, and must usually be mixed with a more sensitive compound. If copper is around, lead azide forms a dangerously sensitive compound. It is even heavier than mercuric fulminate, 4.8 g/cc. Most of the fulminates and azides of heavy metals are sensitive explosives. These materials do not make efficient explosives; their sole attraction is their sensitivity.

Lead styphnate, the lead salt of 2,4,6-trinitro,1,3-dihydroxybenzene, or lead 2,4,6-trinitroresorcinol, is now used, in combination with other ingredients (to give it more bang), in electrical detonators, replacing mercuric fulminate. It is non-hygroscopic and stable, having a positive heat of formation, but very sensitive to flame or spark. It is unusually sensitive to static electricity, so it is dangerous to handle. It is insensitive to nuclear radiations.

Detonators, or blasting caps, are given numbers from 1 to 10 depending on tbe detonating charge. No. 1 is the weakest, with 0.3 g, while No. 10 is the strongest, with 3.0 g. No. 3 is adequate for gelatin dynamite, No. 6 for Gelignite, and No. 8-10 for ammonium nitrate. This is only for orientation; ratings may well be quite different at the present time. Blasting caps are usually fired electrically, the most effective and reliable method. They are kept shorted until just before firing. Detonators must not be stored near explosives.

It was long a mystery why one mass of some explosives would detonate another some distance away. This effect, called "detonation by influence," was employed usefully in practice, although it was not understood. It did not occur with black powder, but only with the newer explosives. Some investigators thought there might be resonances of molecular frequencies, or other arcane influences, but it was shown that the detonations were not specific to particular compounds. Brisant explosives, however, were more effective than less brisant. We now know that this effect was due to the propagation of shock waves, that travelled faster than sound, but rapidly decayed to sound waves with distance.

High Explosives

In 1838 M. Pélouze treated cotton with concentrated nitric acid, producing cellulose nitrate, a substance later called guncotton. In 1845, Schönbein showed that the nitration could be accelerated by mixed nitric and sulphuric acids, and dreamed that guncotton could replace black powder. This new explosive proved very unpredictable. In 1847, a guncotton factory exploded in England. Von Lenk worked for years in Austria to adapt guncotton to ballistics, but explosions in 1862 and 1865 put an end to his experiments. Not until 1865 did Sir Frederick Abel at Woolwich Arsenal finally produce a safe guncotton, by rigorously purifying the raw materials and carefully controlling the manufacturing process. Guncotton was a much more powerful explosive than black powder, but it was difficult to use. E. O. Brown showed that moist guncotton (which was relatively safe) could be exploded by a little dry guncotton (which was sensitive to shock) and a detonator. Guncotton releases about 1100 kcal/kg, nearly twice the heat of black powder, and two-thirds that of nitroglycerine.

There is no molecule of cellulose, but the formula C24H40O20 describes its composition. It is, of course a carbohydrate, since H and O are in the ratio 2:1, consisting of linked sugar molecules. Sugar is rich in -OH groups. Nitration replaces -OH by -NO3, so if n hydroxyls have been replaced, the formula becomes C24H40-nO20-n(NO3)n. The percentage of nitrogen is easily worked out as %N = 1400n/(648 + 45n). The usual range of n for manufactured cellulose nitrate is 8-12. The n = 12 product cannot be made in reliably stable form, so n = 11 is the usual maximum. In this case, %N = 13.47%, and the result is called guncotton. For n = 8-10, the nitrogen ranges from 11.11% to 12.76% and the result is called pyrocotton, used for making collodion cotton, or pyroxylin, for gelatinizing nitroglycerine and for smokeless powder.

Highly nitrated guncotton is insoluble in 2:1 ether-alcohol mixtures, but pyroxyline is completely soluble. Pyroxyline, dissolved in ether, makes films called collodion that were used for early motion picture film. It can be plasticized by a hot mixture of camphor and alcohol. Fillers and pigments can be incorporated, and the mixture hardens as celluloid or xylonite, the first plastic, which filled a long-standing need. It was invented simultaneously in the United States and Britain, which was the reason for the two different trade names given above. In the United States, it was invented by J. W. Hyatt (1837-1920), who was trying to win a prize for an alternative to ivory for billiard balls. The one disadvantage of celluloid was its inflammability.

Meanwhile, in 1846, Antonio Sobrero in Milan synthesized glyceryl trinitrate by treating glycerol with concentrated nitric and sulphuric acids. Fortunately for him, he did not synthesize much before discovering that it was a powerful and sensitive explosive, which was named nitroglycerine. This was a much purer and more controllable explosive than guncotton, but it was too sensitive to be generally used, and too powerful for guns. Alfred Nobel, who had become wealthy through the family oil wells in Russia, became fascinated with this powerful explosive and looked for ways to employ it. In 1865, he discovered that nitroglycerine could be detonated by a mercury fulminate primer in a copper tube. The copper cap containing mercury fulminate for detonating gunpowder had been invented in 1816. The next year, he found that nitroglycerine could be rendered insensitive to shock by adsorption in diatomaceous earth, or kieselguhr. 75% nitroglycerine in 25% kieselguhr made an explosive that you could use as a hammer, but would explode with full power when detonated by a fulminate primer. This explosive was called dynamite. He went on to discover that ammonium nitrate could also be detonated, and was a powerful and useful explosive. In 1875, he discovered that colloidal "collodion cotton" (11.2%-12.2% N2) would dissolve in nitroglycerine and form a gel that was nearly as stable as dynamite. This blasting gel is the most powerful chemical explosive known. Colloidal guncotton was made by plasticizing it in ether and alcohol, and then allowing the solvents to evaporate.

Glyceryl trinitrate is the nitric ester of glycerol, a thick liquid with density 1.6 g/cc, melting at 13°C and becoming rather volatile above 50°C. It invariably explodes at 200°C to 260°C, with a propagation velocity of 7450 m/s. It is difficult to detonate when frozen (below 13°C!), which has been the source of much difficulty with dynamite. Its vapors cause headache. It is used in small amounts as a medicine to dilate cardiac blood vessels and relieve angina pectoris. The stoichiometric reaction is 4C3H5(ONO2)3 → 12CO2 + 10H2O + 6N2 + O2, which provides 330 kcal/mol, or 1470 kcal/kg. Note that the reaction products are completely gaseous, and that there is excess oxygen, so it produces little smoke. In the diagram, the nitrate group is written ONO2 instead of NO3 to show that O is bonded to the carbon, not N. The NO3's have replaced the OH groups of glycerol.

Straight dynamite is made by absorbing 15%-60% nitroglycerine in wood meal, which is an active ingredient that releases more energy than the inert kieselguhr, but also renders the nitroglycerine insensitive. Nitroglycerine produces a little free oxygen on explosion, which burns the carbon in the wood meal. Some nitrate can be added to give more oxygen. It is sensitive and shattering. Ammonia dynamite replaces some or all of the nitroglycerine with ammonium nitrate and sodium nitrate. It is cheaper, but not as shattering, as straight dynamite. A typical gel dynamite has 62.5% nitroglycerine, 2.5% collodion cotton, 25.5% NaNO3, 8.75% wood meal, and 0.5% soda to prevent acidity and stabilize the nitrate. A famous type of gel dynamite is Gelignite, 54%-63% nitroglycerine, 3%-5% collodion cotton, 26%-34% KNO3, 6%-9% wood meal, and 0.5% chalk. Another composition of the same name is 70% NH4NO3, 29.3% nitroglycerine, and 0.7% collodion cotton. This gives some idea of the variations in composition of these explosives. Dynamite is rated on the equivalent nitroglycerine percentage in straight dynamite. A 30% dynamite is as powerful as 30% straight dynamite.

Guncotton and nitroglycerine are high explosives, which means that they decompose at very high rates, and have a property called brisance, a somewhat foggy concept expressing the shattering power of an explosive. Brisance is a combination of a fast rise of pressure and rapid projection of mass, probably equivalent to the creation of a strong shock front. Black powder is very low in brisance, while guncotton and nitroglycerine are high. This rules out using high explosives for propellants in guns. These compounds also contain everything necessary in a single molecule, both oxidizer and fuel, unlike most pyrotechnic mixtures. They are relatively insensitive to heat and flame, but respond to shock and friction. For this reason, they must be detonated by detonators instead of by powder fuses.

The new explosives required nitric and sulphuric acids for their manufacture, and both acids had been expensive and available only in small quantities, especially the nitric. The chamber process for making sulphuric acid rendered that raw material plentiful and cheap. Once you have sulphuric acid, you can make any of the strong inorganic acids. In 1850, the great reserves of Chile saltpeter, NaNO3, were discovered, and in 1863 the first nitric acid was made from it, solving the greatest bottleneck in the manufacture of all explosives, including black powder. Especially in Germany, the manufacture of organic chemicals from the distillation of coal, notably the aromatic compounds, became a major industry. This not only began the manufacture of synthetic dyes, so much superior to the natural product, but also supported the explosives industry. At the time of the First World War, the Haber process for fixing atmospheric nitrogen freed Germany and Europe from dependence on the Chilean nitrates, and brought down their cost steeply.

The high melting point of nitroglycerine has been a source of difficulty. Miners are instructed not to use frozen dynamite for blasting, since it may not detonate reliably and cause "hung shots" and other dangerous situations. To thaw the dynamite quickly, miners sometimes heated it in warm water. This is the wrong thing to do, because nitroglycerine is released from the absorbent in preference to water, and may run out and collect in the bucket holding the warm water. When the water is thrown out, the nitroglycerine explodes to the detriment of the miners present. Leaking dynamite presents the same hazard. Dynamite will burn on an open fire.

In the 1920's and 1930's, liquid nitroglycerine was used for "shooting" oil wells to stimulate production. The productive formation might have a large porosity, so it held a lot of oil, but might be relatively impermeable, or "tight," so the oil would not flow into the small hole with sufficient speed. In limestone, hydrochloric acid was often used, but this was not useful in sandstones. By exploding from 2 to 200 quarts of nitroglycerine, the rock could be fractured for a considerable distance, greatly enlarging the surface through with the oil would flow, equivalent to making a much larger hole. The "shooter" drove alone in a Ford coupe, with the "soup" in the back where the rumble seat used to be, from his source of supply. Nitroglycerine could not be commercially shipped, of course. He poured the "soup" into tin "torpedoes" and lowered them one by one, each fitting into the top of the one below. On the top went a time fuze that ticked away and exploded the charge at a reasonable interval. Then everyone filtered back to the well from their places of refuge. Occasionally, all did not go well, but the "shooters" were well paid and their widows had insurance. There are few graves of "shooters."

Shooting should not be confused with perforating, which also used explosives. When the hole had been "cased" with pipe and the section at the productive horizon cemented in solidly, a device (a gun perforator) was lowered with bullets aimed radially in short guns. These guns were set off, and the bullets punched holes in the casing for the oil to enter. This is still common practice. Shooting was later mostly replaced by a process known as "hydrofrac" in which high pressure water was used to fracture the rock around the base of the hole. A sort of apotheosis of shooting was the use of nuclear explosives in Project Rulison in an attempt to loosen up very tight gas formations in Western Colorado. It was unsuccessful, since it largely destroyed the hole in the process.

Sometimes a burning well was attacked by drilling another hole at some distance that was deflected to some point close to the hole of the burning well. This new hole was then "shot" as described above, which often extinguished the fire, but at the expense of considerable subsurface damage.

Because nitroglycerine is so hazardous to transport, portable nitroglycerine plants were popular in the 19th century. These were small nitrating and washing facilities that produced the "soup" close to where it was used, in mines, construction projects, oil wells and so forth. The raw materials were anhydrous glycerol, fuming sulphuric acid, and 100% nitric acid. One model was produced by a telegraphic equipment manufacturer, one of the few "science-based" industries of the time, which also supplied electrical detonating apparatus.

Mannitol is a hexahydroxy alcohol, like two glycerols end to end. It is found in manna (the sweet juice from Tamarix gallica), celery, rye bread and cane sugar. Mannitol hexanitrate (HNM) is a white powder or colorless crystals melting at 112°C, also called nitromannite. It is less sensitive than mercuric fulminate, and often replaces fulminate in detonating caps. Its explosion point is 160°C - 170°C, about the same as fulminate or nitroglycerine. It is the most brisant explosive, slightly more than even nitroglycerine. It is said, however, to be inefficient as an initiator, which is probably why it is not used alone as a detonator.

Guncotton, nitroglycerine and nitromannite have nitrate groups bound directly to carbons. The free nitrate group is a symmetrical structure with all the oxygens equivalent and the whole group with charge -1. This group is still free to move when bound to a carbon, and there is a very low barrier to exchange of electrons that are closest to the carbon. When a neighboring nitrate group tries to form N2, an oxygen is right there to begin oxidizing the carbon. The reaction probably proceeds due to free radicals in the liquid association of the molecules.

Ammonium nitrate, NH4NO3, is also an excellent explosive, used in certain dynamite mixes (Nobel, 1879) and as a nitrate oxidizer in pyrotechnics. It is very hygroscopic and must be protected against moisture. It decomposes to nitrogen and water, giving very little smoke, by the ideal reaction 2NH4NO3 → 2N2 + 4H2O + O2. The excess oxygen can be used to oxidize some organic material mixed with the nitrate, such as wood meal, starch or diesel oil. Explosives of this type have been widely used since 1867. It is rather insensitive, and must be strongly detonated, perhaps by Primacord or a similar booster. Its density is 1.725 g/cc. Amatol is a mixture of ammonium nitrate and TNT (see below), either 80:20 or 50:50. The nitrate oxidizes the TNT so that no smoke is produced. It was a popular shell filling, economizing on the expensive TNT and stretching out toluene supplies.

Ammonium nitrate is an excellent nitrogen fertilizer, supplying immediately usable nitrate and time-release ammonia. For this reason, it is readily available in bulk. After World War II, fertilizer-grade ammonium nitrate (FGAN) was shipped in large quantities from Texas to France to aid the recovery of European agriculture. On 16 April 1947, SS Grandchamp blew up at Texas City, followed by the SS Highflyer on the 17th. On 28 July 1947, SS Ocean Liberty blew up in the harbor of Brest, France. These disastrous explosions demonstrate the power of ammonium nitrate, and led to more careful handling of this cargo.

Permissible or permitted explosives, also called safety explosives, are explosives approved for use in coal mines where there is a hazard of methane (fire damp) explosions. One modern permitted explosive is ANFO, an ammonium nitrate-fuel oil mixture. The idea is to minimize the flame on explosion, and keep it below the temperature that will ignite the methane. These explosives usually contain mainly ammonium nitrate, sensitized with nitroglycerine so they can be exploded with normal detonators (No. 6), and cooling salts, such as sodium nitrate or sodium chloride, and some wood meal or other organic fuel. Low-velocity grades are specially useful for producing lump coal, since they will not shatter the coal as much as the more powerful explosives. It is not a good idea to do blasting in a gassy mine anyway, so it is better to avoid explosions by adequate ventilation than to rely on permissible explosives, which might ignite the methane anyway.

Smokeless Powder

Vielle discovered how to make a propellant from cellulose nitrate in 1886. He started with low-nitrogen guncotton, or pyrocotton, with 11%-12% of nitrogen, and plasticized it with ether and alcohol. Pyrocotton will dissolve completely in this solvent, unlike guncotton. The gel was rolled out into sheets the sheets were broken up into powder, and the powder formed into grains. These grains, with various additives to control the rate of burning, chemical properties and stability in storage, made a propellant called smokeless powder that could replace gunpowder, and was more powerful. Because no solids are produced in the reaction, there is no smoke, which is a great benefit. Smokeless powder made entirely from pyrocotton is called single-base powder.

Nobel mixed gelatinized the pyrocotton with nitroglycerine to make a smokeless powder of different constitution, called double-base powder. Smokeless powder has completely replaced black powder in ballistics because of its superior power and lack of smoke. It is easily detonated by fulminate caps.

The high temperature of the smokeless powder detonation made erosion of the gun barrel more serious, and the gases resulting could contain flammable constituents such as CO, CH4 and H2. Four molecules of 13.2% guncotton, which is approximately C24H29(NO2)11O20, decomposes to 30CO2 + 71CO + 41H2 + CH4 + 35H2O + 22N2, where the flammable gases are evident. On 13 April 1904, the U.S.S. Missouri, during gunnery practice off Florida, suffered an explosion in the aft 12" gun turret that killed 32 men. The "flareback" when the breech was opened reached a powder magazine, causing a powerful explosion. Flame can often be seen issuing from the muzzle of a gun that has just fired.

Cordite was a famous British double-base smokeless powder, British Service Powder, used in everything from small arms to naval guns until the 1930's. It was 30-40% nitroglycerine, 55-65% guncotton and 5% paraffin grease. It was colloidalized in acetone, and extruded in the form of cords, hence the name. Cordite MD, with 30% nitroglycerine, was a rifle powder. United States Service Powder was single-base, in cylindrical perforated grains. The grains are of different sizes and shapes for different applications. Rifle grains have one perforation, others seven. Small grains may be coated with graphite to facilitate loading of cartridges (to discharge static electricity as well as lubrication). Ball powder is a single-base powder manufactured in spherical grains, for small cannon. Sporting powder is double-base powder that ignites more easily and burns faster than single-base rifle powder.

Mixtures of explosive compounds with a polymerizing binder, making PBX (polymerized-binder explosives) materials, are now popular. The explosives used for these modern material are HMX (cyclotetramethylene tetranitramine, Octogen), PETN (pentaerythrytol tetranitrate), and RDX (cyclotrimethylene trinitramine, Hexogen). Some of these compounds are discussed below, or in Organic Chemistry. The polymer binder acts like the gelling of nitroglycerine, making the explosive safer, which is important for the use of the explosives mentioned, which are quite sensitive. The first PBX was developed at Los Alamos in 1952, using RDX in polystyrene. PTFE (Teflon) and other polymers are also used as active binders.

Aromatic Explosives

What I shall call "aromatic explosives" are not those that smell good, though they may, but those that contain the benzene ring in their structure. Such compounds were discovered in the liquids condensed when distilling coal, of which the archetype is benzene, which has a very pleasant aroma. Unfortunately, benzene causes liver cancer, so it is best avoided. I do not know if just smelling it is dangerous, but I suppose it is. Anyway, one whiff to see what you are missing is probably without hazard.

The formula of benzene is C6H6, so it is quite different from the cycloalkane C6H12, cyclohexane. It is a very stable molecule, the motif of graphite, which consists of fused benzene rings, and no hydrogen at all. The functional groups -OH for an alcohol, -COOH for a carboxylic acid, -NO2 for a nitrate, and -NH2, as well as hydrocarbons such as -CH3 for methyl, can be attached to the ring with little problem, replacing a hydrogen. Some typical aromatic compounds are shown at the right. The circle in the hexagon represents the delocalized electrons that confer stability. Hydrogens are not shown. C6H5OH is phenol, or carbolic acid. C6H5COOH is benzoic acid, C6H5NH2 is aniline, the basis for many dyes, C6H5NO2 is nitrobenzene, and C6H5CH3 is toluene. All of these, including benzene, are liquids. More than one functional group may be present. C6H4(OH)2 is resorcinol, hydroquinone, or pyrocatechin, depending on whether the OH groups are adjacent (ortho), separated by one C (meta), or by two C's (para). C6H3(OH)3 is gallic acid, or (3,4,5)trihydroxybenzoic acid, with the three OH's on adjacent carbons. Two CH3 groups give us ortho-, meta- and paraxylene. All of these compounds are frequently seen and are useful. Benzene rings may be fused in pairs to form anthracene, or in triplets to make naphthalene, famous from moth balls. These are the first steps on the way to graphite. Aromatic compounds are these days more frequently obtained from petroleum than from coal.

One of the first aromatic explosives was picric acid, or trinitrophenol, C6H2(NO2)3OH. It was first prepared in 1771 by Woulfe as a dye, and was also used in medicine, long before it was first employed as an explosive in 1830. The name comes from its extremely sharp or bitter taste, from the Greek pikros, "sharp." It forms pale yellow crystals of density 1.76 g/cc, melting at 122°C and exploding above 300°C. It is too sensitive to heat to be poured into shells, and must be press-loaded. It corrodes metals, forming sensitive picrates. The OH group makes it easier to nitrate the benzene ring. For picric acid to decompose to N2, CO2 and H2O, 27 oxygens would be required, but only 14 are available. Even if all the C goes as CO, one O is still lacking. It is typical of aromatic explosives to be short on oxygen, so they make black smoke. In 1886, France adopted picric acid as the standard bursting charge for shells, under the name of Melinite. In Britain, it was called Lyddite. Picric acid releases 810 kcal/kg on explosion, about half the yield of nitroglycerine. It is a relatively stable explosive, and of low brisance (about equivalent to ammonium nitrate, and half that of nitroglycerine). An explosive more sensitive than picric acid cannot be used in artillery shells.

The most famous aromatic explosive, however, is trinitrotoluene, called TNT for short. TNT is deficient in oxygen, so makes a cloud of black smoke. It is a popular bursting charge for shells and bombs, replacing picric acid after World War I. Picric acid seems still to be used in armor-piercing shells, however, which must delay before exploding. TNT, like picric acid, forms yellow crystals, density 1.654 g/cc, melts at 80.8°C and explodes at 240-280°C. TNT is very toxic; its dust and vapor must not be inhaled. It is rather insensitive to shock, and requires considerable energy to detonate, especially when cast. To avoid the hazards of a large amount of detonator, a booster charge is used that is first detonated, and then detonates the TNT. TNT cannot be detonated with ordinary blasting caps. TNT has been used in some smokeless powders.

The standard booster for TNT is Tetryl, or trinitrophenyl methylnitramine. This molecule is like picric acid, but instead of the OH, has an N(CH3)(NO2) group. Tetryl is fairly stable, but is more sensitive than TNT to shock and friction. It cannot be cast, but must be pressed into pellets. Like TNT and picric acid, it is toxic. Tetryl caps, used in the early 20th century, contained mercury fulminate and potassium chlorate to insure detonation, and are not as safe as ordinary blasting caps, which contain no chlorate. Tetryl boosters may be used in artillery shells. Tetryl was first obtained by Mertens in 1877, and its structure determined by Romburgh in 1883.

Detonating tubes or cordeaux, also called Primacord, are cords filled with explosives used for detonating charges. They are easily set off with a blasting cap. Do not confuse Primacord with slow fuse! One type has a lead covering and is filled with TNT, detonating at 5200 ft/s. Another has a waterproof textile covering and is filled with PETN, pentaerythritoltetranitrate, detonating at 6200 ft/s. PETN is C(CH2ONO2)4, where the four groups are bonded tetrahedrally to a central C atom. It is very sensitive to impact, and has a high detonation rate, 8000-8300 m/s. PETN is a very powerful explosive, and was mixed with TNT for bursting charges in World War II. The 50:50 mixture was called Pentritol.

The aromatic explosives have NO2 groups bound to a benzene ring. If the nitrogen is attracted away by a free radical nitrogen to form N2, then the oxygens may attack and oxidize the benzene ring, releasing more free nitrogen radicals that pick apart other nitro groups. The mechanism does not seem to be known, but lone nitro groups do not appear to be explosive.

All the explosives we have so far discussed depend on the nitro or nitrate groups. As mentioned above, nitrates were often scarce and expensive, especially in time of war, so alternatives to nitrate explosives were sought. The only real alternative was to mixes containing chlorates, which are widely used in pyrotechnics. French Cheddite was an example of such explosives, which used 60%-80% ammonium, sodium or potassium chlorate or perchlorate, with some fuel such as carbon, sulphur, aluminium or vegetable meals. Some aromatic nitro compounds improved flame propagation, and paraffin or castor oil was added as a desensitizer. In Germany, a little nitroglycerine or collodion cotton was added to increase the brisance. These explosives were used in mining and quarrying, not for military purposes, for which they released scarce nitrates.


In pyrotechnics we are principally concerned with solid state oxidizer-fuel mixtures in relatively small amounts. The heat produced by the reaction is used to drive other chemical reactions, to change the physical or chemical state of some other substances, or to create some desired physical effect. Small amounts of gas may be the object, but equally often the reactions may be gasless, or produce a slag. The mixtures are usually much more sensitive than those used for propellants or explosives, which is tolerable because of the smaller amounts involved. However, there may be great hazards in manufacturing and storage when large quantities are present, although each individual is small. A firecracker containing 1 g of explosive is a trivial hazard, but a million such firecrackers in a warehouse would make a bang that can be heard a ways off.

Propellants and explosives may be used in pyrotechnics, of course. For example, a rocket to lift a device to an altitude may use a black powder motor grain, and another black powder explosive charge to deploy the payload, with a pyrotechnic delay charge between them.

Some of the effects produced by pyrotechnic devices are: light, sound, smoke, delays, and motion. The light may be a brief, intense flash, a continued illumination of an area, or a flare to be seen, a signal light, perhaps colored, a shower of sparks or stars, or an intermittent light. The sound can be a bang, a whistle or a crackle. The smoke can be for concealment or for signalling, and of various colors. The motion can be the inversion of a dimple or the extension of a bellows, or the expulsion of a projectile, or a rotation. A pyrotechnic reaction can close an electric circuit, or can open it. It can produce a puff of flame. It can produce gas for pressurizing a safety bag in a car, or for some other instantaneous use. The gas must usually be at low temperature. A pyrotechnic device is for one-time use, of course. Its virtue is that it holds itself in quiet readiness until it is needed.

A squib was a little black powder twisted up in a piece of tissue paper. The paper was lighted on one end, and the squib was thrown. The powder burned, sending out a shower of sparks, and the gas emitted sent the squib darting here and there on the ground. This device was called a serpenteau in French, either from its darting motion, or from an old word for tissue paper, serpente. In Spanish, the firework was called a buscapiés, a foot-chaser, for obvious reasons, a word now used for a firecracker. Later, in English, a squib became a written lampoon, to chase its target in print. A tube filled with powder for igniting artillery came to be called a squib, like the French pétard or amorce. Today, a squib is an electrically-ignited device that ejects a puff of flame, used for igniting rocket propellant, black powder and similar materials sensitive to flame. A squib will generally not ignite shock-sensitive high explosives, such as dynamite or TNT, for which a detonator is necessary.

The oxidizers used in pyrotechnics are mainly the high-energy nitrates, chlorates, perchlorates, and the low-energy metal oxides. The fuels are metals, such as Zn, Al and Mg, carbon, phosphorus and sulphur, and various organic materials. The preferred potassium nitrate, chlorate, and perchlorate are often replaced by the cheaper sodium and ammonium salts where the hygroscopic nature of these salts is not detrimental. Antimony sulphide, Sb2S3, calcium silicide, CaSi2, and other easily-oxidized substances are often seen. These are very sensitive substances, and their mixing is not something that should be done in the home or general laboratory. Those who manufacture the devices are aware of the dangers, and know how to meet them.

Potassium chlorate, KClO3, is a kind of wonder ingredient in pyrotechnics, a powerful oxidizer that give a low-temperature reaction, so that colors will be distinct. It was discovered by Berthollet in 1785, and is the most popular oxidizer in pyrotechnics. It is too unstable for use in explosives. All pyrotechnic colors tend to bleach to white in a hot reaction, as emission through the spectrum is enhanced relative to the specific colored emission bands and lines. It is only relatively recently that strongly-colored displays have been possible through the use of chlorate. However, chlorate is unstable and a very treacherous friend. With sufficient stimulus, it can explode all by itself. If ground dry with any organic material, it is sure to explode. It is this sensitivity, however, that makes it valuable, as well as its peculiarly loose crystal structure.

Potassium chlorate is the substance used for laboratory preparation of oxygen. The chlorate decomposes slowly on heating, giving off oxygen (and a little chlorine, making the oxygen unsuitable for breathing). The reaction is accelerated by a little MnO2 as a catalyst. This is a safe experiment (chlorate will not explode on its own in small amounts), but the chlorate must be respected and never allowed to approach any oxidizable material, especially things like sulphur or phosphorus. Filter paper soaked in 40% chloric acid bursts into flame when it dries. The same might happen with potassium chlorate solution. Potassium chlorate and sugar is called partisan's mixture from it use by revolutionaries and such. It can be exploded by concentrated sulphuric acid that eats through a metal container in a few hours or days. Any revolutionaries using chlorates usually blow themselves up sooner or later.

A typical reaction is 2KClO3 + 3S → 2KCl + 3SO2, often used in smoke mixtures. Using tabulated heats of formation, this reaction yields 223 kcal, or 653 kcal/kg. Potassium chlorate alone gives only 10 kcal/mol or 82 kcal/kg, so the addition of a fuel makes considerable difference. One might think that the sulphur reaction began with 2KClO3 → 2KCl + 3O2, and continued with S + O2 → SO2, but this is not the case. Before the temperature has increased enough for the chlorate to decompose, the sulphur has melted and S3 molecules insinuate themselves into the loose crystals of the chlorate to begin the reaction at just below 160°C. That is, the sulphur picks apart the chlorate, and the reaction begins at a low temperature.

Perchlorates are much less sensitive than chlorates, and should replace them wherever possible. Nitrates are still less sensitive, and can be used with confidence. Strontium nitrate gives the flame a red color, while barium nitrate gives a green color. Strontium or barium anywhere in a mix gives the same color effects as the nitrate. Yellow is made with sodium oxalate, Na2C2O4. Copper gives a blue flame, the most difficult color to produce. Paris Green, a copper arsenate, makes a fine blue, but the arsenic is toxic. A golden color can be made with Fe sparks. These are all emission colors, so they can be mixed additively. The reaction temperature must be kept low, or the colors will be bleached by thermal radiation.

Other oxidizers sometimes used are potassium permanganate, KMnO4, barium peroxide, BaO2, barium chromate, BaCrO4, potassium dichromate KCr2O7, as well as many metallic oxides, such as those of iron, lead and manganese. The insoluble chromates, as desired for pyrotechnic mixes, are easily available in pure form as yellow pigments. Oxidizers are generally of intermediate cost, while fuels can be cheap (C, S, asphalt, sugars, rosin) or very costly (B, Ti, Zr). Mg and Al powders are of moderate cost. Boron is used as the dark-brown powder with particles of about 1μm in size that is 84%-90% B, the rest O. Silicon is ground into a dark-grey powder of size 10μm and larger. Metallic powders can be very hazardous substances, since they react readily, and may even be pyrophoric. In general, pyrotechnic materials are insoluble, non-hygroscopic, not hydrated, and have high melting points. They are held by a binder, or as a compressed powder.

Pyrotechnics are very useful for the production of smoke. Smoke is a colloidal suspension of solid particles in air, an aerosol, that can be created in various ways. A fog is a suspension of liquid droplets in air, and we'll make no distinction between smokes and fogs here, calling them both smoke. Classified by use, smokes can be screening or signalling. A screening smoke is intended to be dense and obscuring. A signalling smoke is intended to be easily seen, and perhaps colored. Evaporating a liquid that then condenses in small droplets is one way to make a smoke (water, for example), or partial burning of carbonaceous matter that leaves carbon particles is another. They are combined in the smoke produced by a fire in damp wood. Crude oil was atomized into the funnels of destroyers, partially burning to make a dense black screening smoke. Burning white phosphorus makes a dense cloud of white phosphorus pentoxide, which combines with the moisture in air to make a fog of phosphoric acid. Silicon tetrachloride and ammonia can be atomized into aircraft exhaust to form, with the moisture, a smoke of silicic acid and ammonium chloride that is used in skywriting. Smokes are also made from chlorsulfonic acid-sulphur trioxide solutions (agent FS) and titanium tetrachloride (agent FM). With the moisture of the air, FM makes titanic acid and hydrogen chloride, and is intensified by being used with ammonia, which forms NH4Cl. These are chemical, not pyrotechnic, smokes.

Military smokes have been dominated by a zinc chloride aerosol produced by a pyrotechnic reaction between zinc dust and hexachloroethane, 3Zn + C2Cl6, that was discovered in 1920 by Capt. Henri Berger in France. These are called, in general, HC smokes. The reaction is 3Zn + C2Cl6 → 3ZnCl2 + 2C, which also produces a good amount of heat. The carbon makes the smoke dark gray, and the rapid reaction made so much heat that convection carried the smoke up into the air. Adding some potassium perchlorate to oxidize the carbon, and some chalk and ammonium chloride to slow down the reaction and consume some of its heat made a useful smoke by 1941. The chalk neutralized any acid that might make the mix extra sensitive (it is found in many pyrotechnic mixes for this purpose). Modifications during World War II substituted ZnO as an oxidizer instead of chlorate, and CCl4 instead of the hexachloroethane. TiO2 was found not to be as satisfactory as the zinc. Manganese was also added in place of the Al tried in earlier experiments. The heat of reaction was brought down from 645 kcal/kg to 365 kcal/kg, reducing the lifting of the smoke as well as the revealing light by night. ZnCl2 smokes cause headaches when personnel are exposed to them for longer than about 20 minutes.

Colored smokes are produced with the potassium chlorate-sulphur reaction discussed above, which vaporizes organic dyes that are about 40% of the mix. The low temperature of the reaction prevents decomposition of the dyes and bleaching of the light. Sodium bicarbonate is added as an alkalizer, flame suppressor and coolant. The mix is about 30% chlorate, 10% sulphur, 20% bicarbonate and 40% dye. Green, red, blue, yellow, violet and orange are the usual colors. In place of the sulphur, sucrose or lactose are used as fuels in standard smoke mixes to avoid the effects of the SO2 on personnel. Smokes colored by dyes may be largely governed by subtractive color mixing, so combinations of dyes must be carefully chosen or the result will only be a muddy darkness if the smokes are well mixed. Smokes that are not mixed may well give an additive color at a distance. Red smoke was colored with Para Red (paranitraniline red), or 1-methylaminoanthraquinone. Blue and yellow dyes will give green only if neither absorbs strongly in the green. The usual blue dye is Indigo, the yellow dye Auramine, and they do give green when mixed. Indigo will sublime at 300°C, but decomposes at 392°C, so the smoke must be kept cool (this is true of all the dyes). Red and blue dyes give violet, while yellow and red produce an orange color.

Among the most curious pyrotechnic devices are those which will act as electrical switches, opening or closing a circuit when they are ignited. For example, the reaction 2PbO + Si → SiO2 + 2Pb begins with nonconducting reactants, but ends with conducting Pb that will join wires embedded in the mixture. The reaction Fe + 3BaO2 → Fe2O3 + 3BaO begins with conducting iron filings, and ends with an insulating slag. Boron with metallic oxides, in a thermite-like reaction, behaves the same way, going from nonconducting to conducting. Lead or tin, with selenium or tellurium and a barium peroxide oxidizer, also becomes conducting after the reaction.

None of the reactions in the preceding paragraph emits a lot of gas, so the slag is not blown away and the device is not disrupted. Such gasless reactions are very useful in delay mixtures. In fact, the lead oxide-silicon reaction can be made to burn at 1.5 - 2 cm/s. The similar Si + PbO2 → SiO2 + Pb reaction burns at 5-6 cm/s. Silicon reacting with red lead, PbO·PbO2, burns at an intermediate rate. KMnSO4-Sb mixes burn at a very low rate, and can give delays on the order of seconds. It is very important for a delay mix to burn at a predictable rate. Gasless reactions are insensitive to pressure variations, and this is one reason for preferring them. Early delays with black powder were often no more than a train of powder in the open. In 1831, Bickford discovered the slowing effect of wrapping the powder with fabric or paper, which conducted away the heat that would otherwise accelerate the reaction. The powder may be slowed by additions of chalk or sodium bicarbonate. Bickford fuses burn at about 1 ft/min (0.5 cm/s), and a puff of flame shoots out of the end of the fuse when it is consumed, ideal for initiating black powder. Incidentally, the spelling "fuse" is commonly used for a chemical initiating device, and "fuze" for a hardware device that may use any principle of operation, including electronics or clockwork. Most present-day fuzes are electronic.

Some pyrotechnic devices provide moderate heat in circumstances where building a fire would be inconvenient. The self-heating food can developed during World War II used a 50:50 iron oxide-calcium silicide mix, Fe3O4 and CaSi2. The first fire for this device replaced the iron oxide with Pb3O4 and added a little china clay to make a more sensitive mix that was easily set off by the lighting match (included with the can). Calcium silicide is a very active reducing agent, so you will not see such cans in the supermarket. The M1 fire starter was essentially a match head of KClO3 and Sb2S3 glued together with dextrin, which, when scratched, ignited a small reservoir of napalm-thickened kerosene. A "Heat Block" used granulated iron as a fuel, and potassium perchlorate and barium chromate as oxidixers, with some other ingredients, to make 210-230 kcal/kg.

Napalm is a soap of aluminium NAphthenate and aluminium PALMitate that can be used to thicken hydrocarbon liquids. It was patented by Prof. L. Fieser in 1952. It forms a gel with the hydrocarbon that suffers from syneresis (evolution of fluid as the gel ages), and is not an explosive or incendiary in itself. The gel is thixotropic, which means that it can be thrown as a liquid, and then gels again. Napalm is used for all sorts of nasty purposes, but was mainly intended for incendiary bombs.

The Goldschmidt process, patented in 1895, has been widely used since then. It is better known as thermite or Thermit, a trade name. A mix mainly of aluminium powder and iron oxide is packed into a crucible and ignited by a magnesium ribbon and a starter mass, or some other means. It is not easy to light. The vigorous reaction results in a slag of aluminium oxide on very hot molten iron, up to 3000°C. The iron flows into a mold that is part of the crucible, usually to perform a weld. The iron oxide is the oxidizer, and the aluminum metal the fuel, in the gasless reaction. The same idea can be used for other metals, such as chromium, manganese and ferrotitanium, and also for a nefarious incendiary device.

Railways have used pyrotechnics as detonators or torpedoes, and as fusees. Detonators are placed on the rail, held by soft metal wings (Pb or Al) that are shaped around the rail head, and explode when run over as an audible signal of danger. They were invented by E. A. Cowper in England in 1842, when they included match chemicals (a crushable bottle of sulphuric acid) and gunpowder. The potassium chlorate-sulphur reaction is now more generally used, mixed with sand and chalk. Potassium perchlorate-antimony sulphide-sulphur mixes are also found. They are rather inert, and difficult to set off except in the intended way. Fusees are cylindrical, with a spike on one end to hold them erect after they are thrown or dropped (they come down spike lowermost). They are especially effective in falling snow, fog, and other difficult situations. The cap contains a scratcher with red phosphorus, and the top of the fusee when the cap is removed contains chlorates and perchlorates. As in a safety match, the reaction is begun by scratching the top of the fusee. The flare mix is mainly Sr(NO3)2 and perchlorate oxidizer, with sulphur, sawdust and miscellaneous ingredient as fuel. The strontium gives the red color. Stearic acid is added in small amounts to adjust the burning rate. A standard fusee burns for 10 minutes. In the past, green fusees were also used (when green was the signal for caution), as well as 5-minute fusees. The green was obtained by substituting barium nitrate for the strontium nitrate.

Some mixes, when pressed into a metal tube about an inch in diameter and ingnited, produce a very loud whistle, descending quickly from a high pitch to a lower as the mix burns. The rate of burning of the mix is strongly affected by pressure. The open part of the tube acts as a resonator, with a velocity node at the surface of the mix, and a pressure node at the open end, so the length is a quarter wavelength at the sound speed in the gas produced. The pressure is maximum at the velocity node, and varies at the resonant frequency of the tube, radiating strongly at the open end. Typical frequencies are from 5 kHz to 1 kHz. 25:75 gallic acid (its structure is mentioned above in connection with aromatic explosives) and KClO3 was an early whistling mix. 70:30 Potassium perchlorate and potassium benzoate, 60:40 potassium picrate and potassium nitrate, and 72.5:27.5 potassium perchlorate and sodium salicylate (aspirin) are others. The last is the standard U.S. whistling mix. Whistling mixes have the unfortunate tendency to explode without warning.

The German Pfeifpatrone of World War II was a handgun-launched parachute flare that gave an intense flash of light, a cloud of smoke, and a shrill whistle in rapid succession. The U.S. had a Day-Nite signal, a metal can that belched red fire from one end, and orange smoke from the other. These, apparently, are still manufactured for use as an emergency signal. As we have seen in the past few paragraphs, pyrotechnics is an excellent means for attracting attention. The Very pistol was patented in 1878, and adopted by the British services in 1888. It had a 1" bore, and fired a cartridge like a shotgun shell, which contained a pyrotechnic star (of various colours) instead of shot.

Pyrotechnics uses one-shot devices that cannot be tested before use. Nevertheless, their reliability must be satisfactory, especially for military and safety use. This can be assured through the use of statistics and rigorous destructive testing. Methods are presented in any text on practical statistics. Redundancy also decreases the probability of failure. If a device fails only once in 1000 times, then two redundant devices will fail only once in 1,000,000 times (on the average).


Fires and illuminations have always been an entertainment, a celebration, or a display of religious superstition. The summer son et lumière displays are examples. Pyrotechnics is a perfect medium for these festivities, combining light, sound and motion in a colorful and impressive way. The displays celebrating the millennium a year early on New Year's 2000 were memorable. A good pyrotechnic display is majestic and stirring when the observer is close to it, so that it happens not only in all directions, but also overhead, and can be smelled and occasionally felt. Although this can be managed with safety, most displays are now seen from a distance, like a picture in a museum, because of the obsession with zero-risk. Residential neighborhoods are not a suitable venue for fireworks, it is clear. Loutish excesses have led to severe legal prohibitions, which are necessary but unfortunate. Until the 19th century, there was little color in pyrotechnic displays, but the brilliance was still impressive. The Chinese were the first to brighten their celebrations by pyrotechnic fires. Incidentally, the original Chinese firecracker was not an explosive, but a green bamboo joint that made a loud crack when thrown on a fire. Many fireworks still come from China and Japan.

Until the 19th century, European fireworks displays were based on an elaborate "machine" or "temple," an architectural display which could involve transparencies or cutouts illuminated from the back, as well as fountains of fire and other effects, which could be very bright and impressive, but not coloured. A common type of machine was in the shape of a pointed obelisk, which erupted in fire. These displays were generally close to the ground. After the introduction of chlorate oxidizers and colour effects, displays were mounted on "lancework" that merely supported the pyrotechnics and presented images outlined in fire. The use of rockets and mortars with air bursts became the principal part of a display, as it is at the present time, so the display was mainly in the air.

The common sparkler is a steel wire coated with a KClO4-Al mixture that makes white sparks. Added Ba(NO3)2 makes the sparks green, and SrCO3 makes red sparks. Fe filings and the barium nitrate make golden sparks. The binder used is dextrine, a gum made from starch. The sparks can be received on the skin with equanimity. Wowsers have attempted to ban sparklers as dangerous, though a more innocuous firework that gives so much pleasure can hardly be imagined. Any child waving a wire can probably cause injury sometime, but sparklers are probably less dangerous than rubber ducks. They can certainly be allowed in residential areas, and are a safe outlet for festive feelings.

The M-80 firecracker is a flash-and-sound device for army training. It is about 0.5 inch in diameter and 1.25 inch long, with a fuse coming out of the middle. The body is spiral-wound chipboard covered with Kraft paper. The charge is 2.5 g of 4:1:1 KClO4-Al-Sb2S3 mix, or a 1:1 Sb2S3-S mix. The Al is the dark colloidal powdered aluminium, which is mixed with the antimony sulphide first. Then the perchlorate is put on top of it, and the tube sealed. Then the tube is "rumbled" in a barrel of sawdust to mix things well. The mixture is quite hazardous if mixed by hand. A fuse is then put in a hole punched in the side. Firecrackers that can be sold legally today are very feeble things, with a maximum load of 50mg. You can hold them in your hand when they explode. More people probably injure themselves trying to make proper firecrackers at home than would be injured by more realistic 1-gram firecrackers. Firecrackers generally contain a flash-and-sound mix, not black powder. The crusade against fireworks is strange in a country where 40,000 die annually in car crashes and over a million are seriously injured. It's probably the usual envy of people having too much fun. I, personally, would not use fireworks, am annoyed by hearing them go off, and consider them vulgar, but I am not a great fan of compulsion or interfering with the course of evolution. It is impossible to buy a firecracker with more than 50mg of charge, but ammunition is easy to obtain everywhere. Insanity.

To illustrate the dangers of fireworks, the explosion in a Paris toy store on 14 May 1878 can be adduced. It happened in a storage room where six to eight million amorces, paper caps each containing only 10 mg of explosive, were stored. Individually, these were innocuous devices. Somehow, a few caught fire (probably someone was smoking) and although they were soon extinguished, the fire caught on a storage box, and all the millions of caps, containing about 64 kg of explosive, went up in an explosion that killed 14 people. Such high-order accidents are the reason why manufacturing and handling fireworks is a job for experts only.

The Roman candle illustrates a lot about pyrotechnics. When lighted, it expels variously-colored stars at intervals, usually six stars, as it is held in the hand. There are successive modules consisting of delay mix, star and expelling charge. The gas-producing delay mix makes a colored flame shoot from the end of the tube as it burns. The flame burns around the outside of a star to reach the black powder expelling charge behind it, which ignites and propels the star out of the tube, to burn in its trajectory. Then the next module of mix burns, and so on. The tube is convolute-wound (like a cylinder, overlapping, parallel to the grain) so it is strong and will not burst. The end of the chipboard is feathered so the inside is smooth and round. Roman candles are not safe in the hands of idiots. Roman candles are named after the use of candles in Mardi Gras festivities in Rome, where people tried to extinguish each other's candles and relighted them, and were first used in Britain.

A stick rocket, shown in the diagram at the left, has a stout paper body, convolute wound. A stick about three times as long as the rocket body is firmly attached to stabilize the flight. The propellant grain, ignited by a squib or fuse inserted in the nozzle, produces gas that exits from the ceramic nozzle at high velocity, propelling the rocket. The design of the grain is critical, so that as much of it burns evenly without burning through to the body. The grain ignites the delay mixture, which allows the rocket to coast to the top of its trajectory before it ignites the bursting charge, which expels the payload in the nose cone, which may be a sound-and-flash device or a parachute star.

An aerial shell is fired from a short mortar using coarse black powder. The black powder is ignited by a fuse, or, better, by an electrical squib. Shells are usually 2" to 8" in diameter, but can be much larger, up to 24". A delay fuse is lighted when the black powder burns, which detonates a bursting charge of black powder at the height of the trajectory. This disperses the stars and salutes (another name for firecrackers) radially. Concentric shells can give multiple bursts timed by delay fuses. A good spherical Japanese or Chinese shell gives a beautiful crysanthemum effect to the burst. American shells are cylindrical, because cylinders are easier to make. Shells are another example of how black powder is still used for many purposes in pyrotechnics. It is much safer than the other pyrotechnic mixes.

The British "cracker" or American "party popper" contains a string that when pulled at both ends makes a small report and liberates a party favor from the burst paper container. To make a cracker, the string is laid out, a loop is formed in it, and a drop of 68% KClO3, 12% red P, 9% S, and 11% chalk is dropped on it (there are other mixes). The mix is unstable, but only 16 mg is used, so it can do no harm. When it dries, the device is wrapped as required. When the string is pulled, the crumbling of the drop detonates it, and it makes a small report. "Snap n'Pops" consist of a cigarette paper wrapped around a little sand coated with 0.8 mg of silver fulminate. When this is thrown down, it makes a little bang. This is a very safe firework, also suitable for residential areas. "Caps" used to be available in rolls for use in "cap pistols." They were on red paper, with little dots where the mix was placed between the layers. When struck, it would make a bang and a little smoke. The mix was Armstrong's Mixture, 67% KClO3, 27% red phosphorus, 3% sulphur, 3% CaCO3 dissolved in water with some gum arabic or similar binder. A small dab, containing a few mg of mix, was put on a paper backing and allowed to dry. Such devices are individually wholly innocuous, but in millions may pose a hazard, as in the explosion in Paris mentioned above. I would confidently bet that they have now disappeared from toy stores. Years ago, boys made a device from two bolts and a nut that held match heads, scraped off strike-anywhere matches, between the two bolts. When thrown rather firmly, a bang was produced.

Novelty fireworks include the cigarette load. This is a small wooden peg coated with a little lead azide (Pb(N3)2). When the fire reaches it, it makes a little bang and blows tobacco about. Candles with wicks soaked in perchlorate ignite again when blown out. Snakes are small pellets that expand greatly when lighted. They were originally made with mercuric compounds, but these have been banned, probably in an excess of regulative mania, since they contained very little mercury and were very seldom used. Substitutes are available, however, so those with a passion for them are not frustrated. The American Hotfoot uses a match head inserted between sole and upper of a shoe. The match is surreptitiously lighted, and when the fire reaches the match head, the result is a source of merriment.


Safety with all of the materials mentioned in this article results from appreciation of the lessons learned from two centuries of practical experience. The greatest hazards, by far, are in the manufacturing processes where materials are combined and the finished devices produced. The ingredients are individually rather harmless. In some cases, they present a poisoning hazard, and in others a fire hazard, but only under exceptional conditions an explosion hazard. The finished devices are also safe, and can often be mistreated in shocking ways without danger. It is in the mixing that all the danger arises, and the manufacturers have been made well-acquainted with the hazards by hard experience. Special measures are taken to ensure as much safety as possible in the manufacturing procedures, and to limit the damage in case of surprise.

Storage and transport of explosives and pyrotechnics is carried out safely by observing the lessons of past experience, codified in manuals of procedure. Pyrotechnic devices stored in large quantities may be as dangerous as explosives, and in some cases even more dangerous because of the more sensitive compounds that may be used, in small quantities in the individual device, but in dangerous amounts in mass. The lesson of the caps in Paris is instructive.

The user is exposed to the least hazard of all. If good procedures are used, then there is little danger to be anticipated. Dynamite will not explode until the detonator is placed. Detonators will not explode if shorted and not exposed to extreme conditions (but they are more dangerous to workers than the dynamite).

A person may want to experiment with these mixtures to find out more about them, to obtain entertainment and to gain practical experience. With electronics, or chemistry in general, this is to be commended. With pyrotechnic mixtures, it is folly. This is not the usual case of wowsers trying to spoil people's fun (as is so very frequent today), but something completely different. You may hear warnings about "only experts should do this" in connection with dangerous things, but here any expert would unequivocally reject casual experimentation. The problem is that the situations that you would establish, in things like purity of materials, grain sizes, order of procedure steps, and so forth, would lead to unpredictable outcomes. This is only exacerbated by amateurism, and even professionalism would be in danger. For example, if you try to make gunpowder with "flowers of sulphur" you will be in danger of premature explosion, since flowers of sulphur contain a little acid, and is much more easily oxidized than the "flour of sulphur" that you really want. Pyrotechnics contains many such lovely traps as this. Of course, you will not succeed in making black powder, but very possibly will succeed in blowing off parts of your body. Some student's last experiments have been finding out what happens when potassium chlorate is substituted for potassium nitrate. Explosives research laboratories and manufacturers have procedures for detecting and eliminating dangers of this type that the individual cannot emulate. Experiments that you may think about doing they would do by remote control behind sturdy blast shields.

One way to learn a lot more about pyrotechnics might be to volunteer to help people who present public fireworks displays. You would get valuable safety training, and could work with these devices yourself. I do not know how practical this suggestion may be, but would like to know if it is possible. The hobby of model rocketry may still be available, but I have not heard much about it lately. Anything involving black powder will probably be relatively safe, and will teach a lot about explosives.


J. H. McLain, Pyrotechnics (Philadelphia: Franklin Institute Press, 1980). This is an absolutely excellent and extremely informative book that suffers only from an inadequate index, though references to the literature are extensive.

H. Ellern, Modern Pyrotechnics (New York: Chemical Publishing Co., 1961). Another excellent book, with much interesting practical information. Particularly good on spontaneous combustion.

H. Brunswig (C. E. Munroe and A. L. Kibler, transl.), Explosives (New York: John Wiley & Sons, 1912).

A. St. H. Brock, A History of Fireworks (London: Harrap, 1947). History of pleasure fireworks by a member of the prominent British fireworks family.

J. Bebie, Manual of Explosives, Military Pyrotechnics and Chemical Warfare Agents (Boulder, CO: Paladin Press, 1942). An excellent dictionary of all the terms, trade names, code names and other lore pertaining to explosives.

J. Akhavan, The Chemistry of Explosives, (London: The Royal Society of Chemistry, 1998).

Return to Physics Index

Composed by J. B. Calvert
Created 21 December 2002
Last revised 18 March 2004