Bell Chapter 11
Bell Chapter 11
Bell Chapter 11
11
The Chemistry of
Combustion and Arson
Burning/Flame/Deflagration Explosion/Detonation
Hydrocarbons C C C O
Fuel C H, H H C N
C C C O
Type of
initiation Thermal Mechanical (shockwave)
process
FIGURE 11.1 The continuum of combustion. The dividing line between burning, deflagration,
and detonation is the speed at which the reaction front propagates.
EXHIBIT A
Salt, Peter?
An older name for the salt KNO 3 (potassium nitrate) is saltpeter (also spelled salt Peter and salt
peter), which is used as a chemical oxidant in gunpowder and explosives. Saltpeter is a mineral
that forms when organic material such as waste, decaying plants, and animal manure is placed
in contact with soil high in alkali content (such as limestone). Saltpeter is found on the earths
surface as well as in caves and is easily collected and mined. The first recorded mention of salt-
peter was by a Taoist alchemist writing and working in the ninth century. Although the mineral
was once used for medicinal purposes, demand skyrocketed with the invention of gunpow-
der and explosives. Lammont du Pont, of the famous du Pont family of chemists, was able to
patent a method for making blasting powder without saltpeter, a much more economical ap-
proach. Unfortunately, he died in 1884 while experimenting with dynamite and sulfuric acid. Du
Pont and several of his assistants were literally blown apart in the accident.
Sources: Harmon, M. B. Gunpowder, Ingenuity, Madness, and Murder: The Saga of the du Ponts.
Biography, November 2002, 92; Morrison, P., and P. Morrison. Nitrogen: The Dark Side. Scientific
American, 281 (October 1999): 12526.
Combustion requires reactants and enough energy to exceed the energy of acti-
vation (Ea) required to initiate the reaction. The reaction profile shown in Figure 11.2
illustrates the exothermic nature of a combustion reaction, as well as the need for
enough energy to initiate it. Once initiated, enough energy is produced to supply the
necessary Ea to sustain the reaction until one of the reactants is exhausted. With a sim-
ple flame, the fuel is exhausted first, since the oxidant is atmospheric oxygen. When
chemical oxidants are employed, either fuel or oxidant may be the limiting reagent.
The energy released in a combustion reaction results from the increased stability
(lower potential energy) of the products relative to that of the reactants. The energy
1 DH 2 released can be estimated with the two methods shown in Figure 11.3. The first
method is based on table values (found in Appendix 5) of thermodynamic quantities
under standard conditions of temperature and pressure (STP, 25C and 1 atm). The sec-
ond method calculates how much energy is required to break the chemical bonds in the
reactants and form the bonds of the products. As shown in the figure, this result should
be relatively close to that determined from the table values; however, the quantities cal-
culated are used only as starting points, because the combustion reactions encountered
in forensic chemistry are complex and rarely occur close to standard conditions. Even
if the reaction starts at near standard conditions, the reaction itself quickly pushes the
system well beyond standard conditions.
Ea
Reactants
H (exothermic)
Energy
Products
Time
Reaction profile
Oxidized
FIGURE 11.2 A generic reaction profile for
combustion, along with a simple example of a CH4(g) + 2O2(g) CO2(g) + 2H2O(g)
combustion reaction. Combustion is a specialized Oxidizing
type of redox reaction. agent
(moles)(kJ/mol)
H oreaction = [ 393.5 + 2(241.82)] [ 74.8]
FIGURE 11.3 Two methods of estimating how much energy is released by a combustion reaction.
In the top frame (A), H is calculated from table values, whereas in (B), the energy balance between
bonds broken and bonds formed is used.
Some readers may be familiar with the fire triangle, which is one way of summa-
rizing the requirements for a combustion reaction. Such a triangle is divided into three
regions, identified as fuel, oxidant, and heat (the last of which supplies Ea). Building on
the concept of the triangle, we will consider the requirements for combustion to be
1. fuel and oxidant in appropriate quantities and concentrations,
2. a source of Ea, and
3. sufficient contact time for the energy source to initiate the reaction.
The absence of any one factor prevents combustion. We will examine the particu-
lars of each factor next.
The first requirement, fuel and oxidant in proper proportions, illustrates key
points and unmasks common misconceptions. Wood does not burn; rather, what burns
are the vapors emanating from heated wood. Gasoline in a can will not explode be-
cause the proper fuelair mixture does not exist. A cigarette tossed into a pool of gaso-
line usually smothers before it has a chance to ignite the vapor above it. Similarly, the
Hollywood staple of exploding gas tanks in cars is more fiction than fact. Rapid burning
can occur, but only when the gas tank is ruptured, the contents leak and vaporize, and
the proper airfuel vapor mixture is created at the same time and place as a source of
ignition that stays in contact long enough to spark the reaction.
C
H2O
Convective
CO2 plume
HCN
HCN
H2O
CO CO2
CO2
H2O
C CO2
CO2
CO H2O
H2
H2 C
CH3OH CH
CH HCN CO
CH3 C
CHO
Radiant heat CN CH2O Radiant heat
C
CH4
H2
CH C
CO
CH CH3
C C C
CH2
CH2 CH
CHO H
H CH2
CH CH2
CH3
CH2O C CH3
O2 C H O2
Air CH Decomposition region H Air
N2 N2
Vaporized fuel
FIGURE 11.4 The combustion of wood (fuel),
Pyrolysis zone
illustrating aspects of combustion addressed in
this chapter. Note the complexity of what would
Fuel
at first seem to be a simple chemical reaction.
(speed) of combustion depends on the rate of formation of free radicals in the flame
flickering above the wood. The reaction speed also depends on multiple rate constants
and reactant concentrations in multiple connected chain-reaction pathways. The heat
evolved, favored pathways, and the balance of products will all depend on thermody-
namic considerations, including stoichiometric ratios and equilibria. Thus, although
we will address each of these topics individually, all interact to control and define the
complex process of combustion.
11.2.1 Thermodynamics
Thermodynamics relates concepts of energy flow, enthalpy (H ), entropy (S), free en-
ergy (G), and equilibrium. The first law of thermodynamics states roughly that energy
is neither created nor destroyed but only changes form. In combustion and in explo-
sions, potential energy in chemical bonds (chemical energy) is converted to heat and
Equilibrium is defined as the point at which G 0.
work. Energy is defined as the ability to do work and can be categorized by the type of
work done. For example, there is chemical energy, potential energy, mechanical energy,
and kinetic energy. The second law of thermodynamics relates to entropy and (again,
roughly) states that in any spontaneous process, the disorder of the universe increases.
Entropy increases during combustion because gaseous products are formed and heat is
released by the exothermic reaction. Molecules move faster at higher temperatures rela-
tive to molecules at lower temperatures, resulting in increased disorder.
The free-energy change of any reaction, including combustion, is defined in terms
of enthalpy and entropy:
DG 5 DH 2 TDS (11.1)
where S is entropy and G is the Gibbs free energy. In a flame, the combination of an
exothermic reaction with increasing disorder leads to a large negative value for the
change in free energy 1 DG 2 . DG also is a measure of how much work can be done by
a system in a spontaneous reaction. Work can be divided into two components: actual
work (w) and heat (q). Both aspects come into play across the combustion continuum.
In a fire, such as an intentionally set arson fire, heat and gases are produced but are
not exploited to do work. In propellants and explosives, heat and work play critical
roles and the work done is the central issue.
In chemical applications, the most common type of work is PV work. Examples
are shown in Figure 11.5. The top frame illustrates how the evolution of a gas 1 H2 2 in a
spontaneous process is used to do PV work in a system in which heat is a minor contrib-
utor. The lower frame depicts the relationship between PV, work, and force. To move
the piston, workspecifically, PV workmust be done, and the force acting inside the
cylinder must exceed the force exerted by atmospheric pressure. If the piston was re-
moved from the system, the gasoline would still ignite, burn, and generate hot expand-
ing gases; however, this expansion is not purposely directed into doing PV work. This is
the situation we see in arson fires.
Forensic examples involving energy and illustrating how work is done are
shown in Figures 11.6 and 11.7. In Figure 11.6, a bullet is propelled out of a gun by
PV work done by expanding gases produced by the burning propellant. Here, the
bullet is analogous to the piston (Figure 11.5), which is moved as a result of the pro-
duction of gas. In the gun, primer ignites when struck by the hammer, provides the
initial Ea spark, and initiates combustion, which produces heat and hot expanding
gases. The bullet is held in the cartridge by compression and friction, but the joint
is designed to give way once sufficient pressure builds up. The result is movement
of the bullet down the barrel, just like movement of the piston shown in Figure 11.5,
except that the force must overcome the compression and friction forces holding the
bullet in place. As gas expansion continues, much of the energy is transferred to the
bullet as kinetic energy. The energy trace of a gun firing can be summarized as a me-
chanical energy (hammer striking primer) h chemical energy h heat and work
(heat and mechanical) h kinetic energy. This progression can be simplified to
ME h CE h ME 1 and heat 2 h KE.
For a crude pipe bomb (Figure 11.7) made of galvanized steel pipe and gunpow-
der, the energy pathway is the same. What differs is that no joint is designed to fail, as
was the case with the gun firing. A pipe bomb is a mechanically stronger containment
device that allows pressure to build until it exceeds the strength of the container at its
weakest point. As we will see in Chapter 13, pressure and confinement are critical fac-
tors in explosions and detonations. The pipe shatters and ejects shrapnel in all direc-
tions. In contrast, a gun is designed to focus and direct the gas expansion and work in
order to impart the most kinetic energy to a single projectile traveling in a controlled
trajectory in one direction.
a) b)
H2 gas
plus original
atmosphere
Zn
Zn
HCl solution HCl solution
P = F/A P = F/A
Gun barrel
Primer
Bullet KE = 1/2 mv 2
Ea
Ea
Vessel ruptures
Shrapnel
KE = 1/2mv2
FIGURE 11.7 A pipe bomb is designed to rup-
ture and not to fail at a specific joint. The result
is catastrophic destruction of the container and
the ejection of sharp shrapnel moving at high
speed.
The value of Q (used to heat the products to Te) is central to describing and esti-
mating explosive power. Figure 11.3 illustrated two methods of calculation of the heat of
reaction, but of more interest here is gauging the relative heat generated per gram of fuel.
Also, the calculations shown in Figure 11.3 assume that the fuel and oxidant are in stoi-
chiometric equivalence (which we abbreviate as STE), meaning that the molar amounts
of fuel and oxidant are in exactly the correct proportions according to the balanced equa-
tion for the reaction to go to completion. In the case of methane combustion (Figure 11.2),
if 1.0 mol of methane was present when the reaction started, then at STE there would
be exactly 2.0 mol of O2 present. This situation, however, is rarely encountered, and as a
result, both the products of the reaction and the heat evolved will be affected.
Assuming that conditions are such that methane (CH4) and methanol (CH3OH) combust via
an explosion, which is the more powerful per gram detonated?
Answer:
To answer this question, the value of heat released (Q or H) for each compound is needed,
as well as the number of moles of gas produced. Use the thermodynamic values provided in
Appendix 5 to determine Q per mole combusted. Some assumptions are necessary, but for a
rough comparison, this approach is reasonable. Here, we will treat all species in the gas phase.
CH4 1 2O2 h CO2 1 g 2 1 2H2O 1 g 2
DHg
274.8 2393.5 2241.82
kJ/mol
moles 1.0 1.0 2.0 (3.0 total)
grams 1.0
moles/gram 0.063 0.063 0.125 (0.19 total moles per gram)
22.4 L
0.19 3 > 4.3 L/g
mol
DHreaction
o
5 3 2393.5 1 2 1 2241.82 2 4 2 1 274.8 2
2802.3 kJ 250.1 kJ 50 kJ
5 5 ;Q>
mol CH4 g CH4 1 g 2 g CH4
2CH3OH 1 g 2 1 3O2 h 2CO2 1 g 2 1 4H2O 1 g 2
2 mol h 6 mol gas
moles/gram 0.0313 h 0.0313 mol 1 0.0625 mol or,
2.1 L
calculated as above,
g CH3OH
DHoreaction 5 3 2 1 2393.5 2 1 2 1 2241.82 2 4 2 1 2201.2 2
21069 kJ 2535 kJ 16 kJ
DHoreaction 5 5 >
2 mol CH3OH mol CH3OH g CH3OH
QVCH4 > 50 3 4.3 5 215 QVCH3OH 5 1 16 2 1 2.1 2 <36
The difference is ~ a factor of 6 for 1 g of each substance.
Notice that, in balancing the equations, clearing the fractions was not necessary, since the oxy-
gen does not contribute to Q. The combustion of a gram of methanol produces about half the
gas volume of the combustion of a gram of methane. Thus, ethanol is a better fuel, at least
using this one metric.
Using the ratio of fuel to air at STE as a reference point, we can describe situa-
tions in which we do not have STE. When the amount of fuel relative to the amount of
oxidant decreases, the mixture becomes a lean mixture. In chemical terms, the system
is overoxidized, and if the fuel concentration drops too low relative to the concentra-
tion of oxidant, combustion cannot take place. When there is more fuel relative to the
oxidant, the mixture is rich, and the system is underoxidized. At the extreme of rich-
ness, combustion is impossible. An underoxidized system favors production of the less
oxidized product and releases less heat.
The calculation of the fuel/air ratio (F/A) is an important one and can assist
in determining such properties as potential flammability of mixtures. Returning to
the combustion of methane as our example, the molar ratio of methane to oxygen in
a balanced equation at STE is 1:2, respectively. Because the types of combustion of
interest here take place in air, additional corrections are required. Assuming that the
atmosphere is 21% oxygen, the number of moles of air supplied must be adjusted up-
ward. A mole of air contains 0.21 mol of O2; to obtain 1 mol of O2, we need to multiply
this value by 4.76 1 0.21 3 4.76 5 1.0 2 . One last adjustment is to multiply this result
by 2, since the complete combustion of 1 mol of methane requires 2 mol of oxygen.
This value is easily converted to a mass ratio using the formula weight of methane
1 16.0 g/mol 2 and the weighted average mass of air, which is generally taken to be
28.85 g/mol:
Thus, at stoichiometric equivalence, the mass ratio F/A is 0.0583. It is important to note
that here we are assuming a closed system, in that the fuel and air are being held at a
fixed volume at 1.0 atm. We will see many instances in which we have to make some
basic assumptions about conditions, and in many cases, our calculations are reasonable
estimates, not hard values. However, these types of calculations are the foundations of
complex models used to describe combustion, deflagration, and detonation.
Now, consider an example in which 5.0 g of methane is released into a container of
air with a volume of 20.0 L at a temperature of 25C. To determine whether the result-
ing mixture is combustible, rich, or lean, the mass ratios of fuel to air are calculated and
compared with the ratio at STE just calculated. The first step is to obtain mole ratios via
partial pressures and the ideal gas law:
5.0 g L # atm
a b a0.0821 b 1 298 K 2
nRT 16.0 g/mol mol # K
PCH4 5 5 5 0.38 (11.6)
V 20.0 L
The last step is to determine the F/A ratio relative to that at STE:
The mixture in the container is rich relative to the STE ratio. As seen in Figure 11.8,
deviations from STE affect the temperature of the flame; we will see why momentarily.
Staying with our methane example, Figure 11.8 illustrates how rich and lean mix-
tures alter the heat evolved. The heat released at stoichiometric equivalence was previ-
ously calculated (Figure 11.3); note that the heats of formation for water (in gas or liquid
form) and methane are constants in this combustion, and the contribution of elemental
O2 is zero. As a result, the expression for calculating DH to compare the three reactions
depicted in the figure can be simplified as follows:
where Xi represents the mole fraction of each species produced. The most negative
value (the largest Q) occurs when only CO2 is produced. Any carbon monoxide pro-
Temperature
h mol
de
c
Ri
Un
Stoichiometric equivalence
CH4(g) + 2O2(g) CO2(g) + 2H2O(g) + heat
393.5 kJ
mol
Ov
ox
er
L e id iz e
an d
CO2(g) + 2H2O(g) + O2(g) + heat
excess
duced decreases the heat released, since the overall heat of the reac- CH3
tion becomes more positive (DHreaction becomes less negative). Thus,
O2N NO2
underoxidized systems, which favor CO production, release less heat
than do systems at stoichiometric equivalence. However, if there is
excess oxidant, some of the heat evolved is diverted to heat that oxi-
dant, rather than just the products, as we assumed in the adiabatic
combustion model. As a result of heat diversion, Q decreases from the
maximum produced at stoichiometric equivalence, at which Q heats NO2
only the products.
Trinitrotoluene (2,4,6 TNT) C7H5N3O6
The discussion becomes more interesting when we examine
reactions in which the oxidant is not atmospheric oxygen or in
which the source of oxygen is chemical, as in the case of explosives
NO2
such as TNT or nitroglycerin (Figure 11.10). With these explosives, O
part of the oxidant is supplied by the molecule, part is supplied by
the atmosphere, and the ratio is expressed as the oxygen balance. O O
As demonstrated in Example Problem 11.3, nitroglycerin has a posi- O 2 N NO2
tive oxygen balance, meaning that when the explosive decomposes
Nitroglycerin C3H5N3O9
to gaseous products, the explosive molecule itself can supply all
the needed oxygen, with some to spare. The reverse is true for ex- FIGURE 11.10 Two common explosives.
plosives such as TNT, which require oxygen from the atmosphere
or another chemical source. When the oxygen balance is negative
and relatively large, CO will form in preference to CO2. In other words, the system
is underoxidized and lean. The oxygen balance can also be expressed as a weight-
percentage-like quantity derived from the ratio
Answer:
The oxygen balance is always negative. There is no oxygen in the molecule.
Example Problem 11.3 illustrates how oxygen balance is calculated, and Table 11.1
gives the oxygen balance of some representative explosives.1
The concept of oxygen balance is analogous to the definition of a rich or lean
mixture of fuel versus oxidant. The difference is that the oxygen balance is internal
Two components: x = fraction NG to the fuel molecule; in other words, the molecule supplies both fuel
1 x = fraction TNT and at least part of the oxidant. Because the oxygen ratio is related
to stoichiometric ratios, it also relates to the heat release Q. When a
4x + (1 x)(74) = 0 given explosive is the only material combusted, the more positive
negative
the oxygen balance, the greater is the heat released. By itself, TNT
(+) Oxygen balance does not generate as much heat as a compound such as nitroglycerin.
However, explosives are often formulated such that their combined
4x + 74x 74 = 0 oxygen balance approaches zero, which corresponds to stoichiometric
78x = 74
equivalence (STE). For example, suppose an explosive mixture con-
sists of TNT and nitroglycerin. To maximize Q, the combined oxygen
x = 0.95 95% NG
balance should be as close to zero as possible. TNT has a large nega-
1 x = 0.05 5% TNT tive oxygen balance, whereas nitroglycerin has a small positive bal-
ance. Clearly, then, the mixture should contain a little TNT and lots of
4(0.95) + 0.05(74) nitroglycerin. The calculation is shown in Figure 11.11.
0
3.8 3.7
FIGURE 11.11 Oxygen balance calcu-
lations for a mix of explosives.
Calculate the oxygen balance for the following materials: aluminum metal and nitroglycerin
(NG, C3H5N3O9).
Answer:
The first step is to balance the combustion equation, with products going to their completely
oxidized state. Any nitrogen is assumed to go to N2. The relative excess or deficit of oxygen is
calculated as a weight percent.
For Al:
Al 1 O2 h Al2O3
Balanced:
4Al 3O2
Moles 4 3
Formula weight (g/mol) 27.0 32.0
Grams 108 96
Oxygen balance:
96 g O2
3 100 5 289%
108 g Al
The balance is negative because additional oxygen is required for the reaction to proceed.
For NG:
This part of the problem is more complicated because oxygen is present in the molecule. We
need to determine whether the amount is sufficient by itself to support the reaction, so we first
balance the equation:
4C3H5N3O9 O2
Moles 4 1
Formula weight (g/mol) 227.0 32.0
Grams 908 32
Oxygen balance:
32 g O2
3 100 5 13.5%
908 g NG
Combustion does not involve simple single-step collisions between a fuel mol-
ecule and an oxidant. Rather, the reactions that occur during combustion are based on
free radicals. In a free-radical2 mechanism, three generic steps take place:
initiation, in which the first free radicals are formed
propagation, in which reactions among radicals produce more radicals
termination, which results from the combination of two free radicals to form a
neutral species
Because radicals react with neutrals to create new radi-
cals, a chain reaction results. Each step in a generic free-radical
CO + O2 CO2 + O Initiation
reaction has an associated rate constant. One of these steps
will be the slowest and is termed the rate-limiting step. Just as
O + H2O OH + OH the slowest member of a relay team limits its performance, the
Propagation
CO + OH CO2 + H rate-limiting step limits the speed of the chemical reaction. The
rate-limiting step is generally the step with the highest energy
H + OH H2O Termination of activation. Partial and simplified steps for the combustion
oxidation of hydrogen are shown in Figure 11.12.
FIGURE 11.12 Sample free-radical reactions that
occur in a flame.
EXHIBIT B
Our discussion here will focus on heat transfer via convection (matter transport), but heating by conduction
and radiation also take place.
which were seen in Figure 11.4 and which are also shown in simplified form in
Figure 11.13. Heated air is less dense than cooler air, and as a result, much of the radiant
heat produced in a simple combustion like a burning candle is carried away in rising air
and gases. This kind of combustion is also called a buoyant flame. Heat is also required
for phase transitions, as shown in Figure 11.14. In the candle, the wax must first be melt-
ed to liquid and vaporized before combustion occurs. For some heavier hydrocarbons,
the melting point and heat needed to vaporize them can be quite high. If Q dwindles, so
will the supply of vaporized fuel.
Heat transfer to the substrate has interesting effects on fire behavior and arson
investigation. As shown in Figure 11.15 (top frame), heat may reach deep into a sub-
strate such as wood even when oxygen cannot. The result is pyrolysis (fire cutting)
or decomposition in a reducing environment. The products of pyrolysis are different
from those of oxidation and can be identified as a layer in burned wood. Typically, the
pyrolysis zone is some distance below the burned surface. The layer is defined by the
availability of oxygen and by the depth to which heat can penetrate.
The bottom frame of Figure 11.15 illustrates a situation seen in many arson cases:
a liquid accelerant such as gasoline is poured over a surface and the vapors are ignited.
If the pool is deep enough, it insulates the substrate below and limits the temperature
increase. Just as the temperature of liquid water (not steam) cannot exceed 100C, the
temperature of the liquid accelerant cannot exceed its boiling point. Consequently, the
pattern of burning and scorching at the edge of the pool and away from it will be differ-
ent from the patterns directly beneath.
Heat transfer is directly related to the concept of mass transfer. Q is used to heat
products of the reaction, which, on a molecular level, means that the greater Q is, the
more kinetic energy is transferred to the product molecules. This kinetic energy can be
transferred to other molecules via collision. However, for that energy to be transferred
to key molecules such as those found in the vaporized candle wax, the energy has to
be delivered, via fast-moving molecules, to the right place. That movement of mass is
called mass transfer.
Ea
Boiling
point Vaporization
Hv
Q T Liquid Gas
Vaporized
wax components
Ea Melting
gas
gas point
Melting
Hf
liquid Melted wax
Solid
wax (solid)
Hf Hv Q added
solid liquid gas
FIGURE 11.14 Some of the heat generated by the
FIGURE 11.13 Simplified heat flow paths in a combustion is consumed in the necessary phase transitions
candle flame. of the fuel.
O2 O2
Scorching
Optimal [fuel]
for combustion
[Fuel]
0 Distance from surface
At surface
PO2~.2latm
[O2]
0
FIGURE 11.16 Concentration gradients of fuel and oxidant
as a function of distance. The [ ] notation refers to concentration.
Lean 21%
Saturation
Concentration of
Concentration of
fuel vapors
oxygen
Combustion
Rich
zone
Combustion zone Vertical
fuel
0 (solid)
0 Close Far
Distance from fuel surface
Combustion
Radiant
zone
heat
Oxygen
in air
Oxygen
Vapors in air
from fuel
Flames
Fuel
Fuel Vapors generated
surface by pyrolysis of
fuel surface
FIGURE 11.17 Different views of the concentration gradients and the zones at which the
concentrations of the fuel and the oxidant support combustion. Figure 11.8 is superimposed
for reference.
Finally, in cases involving a poured liquid accelerant, mass transport of the fuel
occurs in a lateral direction, controlled by the characteristics of the surface. Gasoline on
a nonporous surface like concrete will diffuse easily, whereas gasoline on a porous sur-
face like wood or carpet will tend to be absorbed. As a result, porous and semiporous
surfaces should be sampled in depth, since the chances of finding residual accelerants
is increased in such cases.
Source: Reprinted, with permission, from Journal of Forensic Sciences, copyright ASTM International, 100 Bar
Harbor Drive, West Conshohocken, PA 19428.
Smoke
(residual soot)
O2 diffusion
Oxidation
Growth
Diffusion Heat
products
diffuse Soot formation
(particulates)
Premix
zone
Fuel (diffusion)
Residual
fuel/air Heat, compression
Combustion wave
~15 45 cm/s
As long as the fuelair mixture remains in the combustible range, the flame will
be self-sustaining. The events that bracket the flame event are initiation (ignition) and
quenching or suppression. The range of combustibility is referred to as the flammable
range and is defined as the fueloxidant ratios that permit steady propagation of the
flame. The lower end of the scale is called the lower flammability limit (LFL) or the
lean limit, whereas the upper range is the upper flammability limit (UFL) or rich limit.
The terms lower explosive limit and upper explosive limit (LEL and UEL, respec-
tively) are also used. Some examples are presented in Table 11.2.
You will note that in Table 11.2, the flammability limits are presented in volume %.
How could you use this data to assist in a fire investigation? This information is essen-
tial to deciding whether a combustible mixture could have existed in the right place and
at the right time to support combustion. For example, assume you are investigating a
fire that occurred in a small storage shed with dimensions of 4 ft 3 4 ft 3 6 ft. Inside the
destroyed shed, you find what is left of a propane bottle, labeled as containing 1 lb of
propane. Is it possible that the F/A ratio could have supported combustion? Having the
dimensions allows you to calculate the total volume of the shed:
4 ft 3 4 ft 3 6 ft 5 96 ft3
1 ft3 5 28.32 L
28.32 L
96 ft3 3 5 2718 L
1 ft3
Table 11.2 indicates that the flammability range of propane is 2.2%9.6% by volume. We
can use this information and the ideal gas law to estimate what these limits correspond
to in liters of propane. Initially, we will assume a temperature of 25C and a pressure of
1.0 atm:
0.022 3 2718 L 5 60 L
0.096 3 2718 L 5 261 L
These values provide an estimate of the volume of fuel that would have to be present.
Using the ideal gas law, we can estimate how many moles of propane these volumes
correspond to:
PV 1.0 atm 3 60 L
n5 5 5 2.4 mol at LFL
RT L # atm
0.0821 298 K
mol # K
PV 1.0 atm 3 261 L
n5 5 5 10.7 mol at UFL
RT L # atm
0.0821 298 K
mol # K
The molar mass of propane (C3H8) is 44.09 g/mol, and 1.0 lb = 453.59 g, so we can
convert the moles to pounds. Of course, you could convert the propane can weight to
grams as well; both approaches are valid.
44.01 g 1.0 lb
2.4 mol 3 3 5 0.2 lb
mol 453.59 g
44.01 g 1.0 lb
10.7 mol 3 3 5 1.0 lb
mol 453.59 g
The empty bottle was labeled as containing 1.0 lb of propane, indicating that the
contents released into the shed could have supported combustion. You will notice we
are being a bit cavalier about significant figures; there is a reason for this. This approach
is by its nature an estimate and does not consider many factors that would be important
in an arson investigation. For example, we made the implicit assumption that the shed
represented a closed system, all the propane left the bottle, and none of the contents
escaped the shed. We also estimated temperature and pressure. In this type of situa-
tion, all we are after is a reasonable and defensible answer to an investigative forensic
question: Could the propane have supported combustion? In this case, there was suffi-
cient propane to sustain combustion. However, this would represent a starting point for
further investigation, not the end. Alone, this calculation would not prove or disprove
arson, but it would provide investigators with important information.
As demonstrated and discussed already, many other variables will alter the fuel/
air ratio that exists in a given place at a given time. Because the fuel and oxidant are
gases, pressure and temperature are among the most important variables to be consid-
ered in a fire investigation. For example, the higher the temperature, the wider is the
flammable range. In arson cases, homogeneous mixtures rarely exist, and the relative
densities of materials become critical factors, as shown in Figure 11.20.
The weighted-average formula weight of air is taken to be approximately 29 g/mol.
The relative weight of hydrogen 1 H2; formula weight FW , 2 2 is therefore 2/29, or
about 0.07. If hydrogen is released into the air, it rapidly dissipates upward and away
from the release point. Conversely, gasoline vapors tend to sink. If gasoline is crudely
represented by n-octane 1 C8H18, FW , 114 2 , the weight ratio relative to air is 114/29,
or ,3.9. Therefore, in a quiescent or nearly quiescent environment, fuel vapors
disperse according to their weight and density. This phenomenon is illustrated in
Figure 11.20.
The presence of vapors within the explosive limits is a necessary condition for
combustion to occur, but it is not a sufficient one. The initial Ea barrier must be over-
come, and the energy to do so is supplied by an ignition source that must have suffi-
cient energy available and that must remain in contact with the flammable mixture long
enough to ignite it. In arson investigations, the ignition source is called an incendiary
device, such as a match, a cigarette, or a more sophisticated apparatus. As demonstrated
in Figure 11.20, this device must also be in the right place at the right time. In that ex-
ample, gasoline has been poured down the basement stairs and enough time has passed
that a stratification has developed in this closed, quiescent system. The gasoline vapors,
which are heavier than air, move lower in the room and create a rich layer near the floor.
Someone wanting to ignite the mixture would have to place the incendiary device in the
flammable region. A match tossed to the floor would not work because the mixture is too
rich, and a spark at the top of the stairs would fail because there the mixture is too lean.
In addition, the incendiary device would have to produce sufficient energy to initiate the
Gasoline,
poured down the stairs
Lean
Source
FIGURE 11.20 In a quiescent or nearly quiescent environ-
ment, the density of the fuel relative to the atmosphere
will dictate the places where fuel vapors accumulate and
Flammable Liquid the zones where combustion is supported. In this example,
gasoline which is heavier than air, will settle into a pattern
such as shown here, with the zone next to the floor being
Rich
relatively rich and the zone at the level of the can being
Drain
relatively lean.
reaction and would have to be in contact with the reaction mixture long enough to en-
sure the combustion had become self-sustaining. The energy needed to ignite a mixture
is usually thermal. Kinetic energy is transferred to the reactants via collisions among the
molecules of the mixture. The collisions may be instantaneous, as in the case of someone
using a striker and spark to ignite a Bunsen burner, but even the biggest spark will not
ignite a mixture that will not support combustion. In other cases, ignition takes much
more time, as when a smoldering cigarette is placed between the cushions of a couch.
With explosives, pressure can provide sufficient energy for combustion.
Intentionally set fires, or arson, usually involve an accelerant of some type, as well as
an incendiary device used to ignite it, and these two components create the physical
evidence forensic chemists work with. Arson fires are also referred to as incendiary
fires. For fires set in homes, a bedroom was the most common point of origin; fires set
in public buildings were usually started in bathrooms. From a law enforcement point of
view, arson is a difficult crime to clear; in 2008, only 18% of arson cases were cleared by
arrest or other means. Nearly half of those arrested were under 18, and 3% were under
10 years of age.5,6
One of the challenges of fire investigation is the classification of a fire as natu-
ral, accidental, or incendiary (arson). Fire investigators utilize evidence at the scene,
as well as forensic analysis, to make such determinations. One of the most important
pieces of information required in making a determination of arson is the location of
the point or points of origin of the fire. Multiple points of origin are strongly indicative
of an intentionally set fire, whereas a point of origin at an electrical outlet suggests an
accidental fire. As discussed in Section 11.3, the behavior of a fire creates predictable
damage that is useful in locating a point of origin. Although such fire and fire scene
investigations are critical in determinations of arson, we will focus on the chemical
analysis aspects of fire and arson investigation. The former is fire investigation; the lat-
ter is forensic chemistry.
Recall that one measure of the efficiency of a chromatographic column is the number of theoretical plates.
This analogy is drawn from distillation: the more collection plates in the column, the more effective is the
separation. The maximum number of plates (N) is achieved when the distance between them (i.e., the smallest
possible height of a plate, HETP) is minimized.
Source: ASTM Standard E 1618, Test Method for Ignitable Liquid Residues in Extracts from Fire Debris
Samples by ChromatographyMass Spectrometry.
and extracted with a solvent such as CS2. Active or dynamic headspace (DHS), also
called purge-and-trap, is a methodology widely used in environmental analysis. In
DHS, an inert gas constantly flows through the heated container, carrying volatiles
downstream to a trap. Because the equilibrium is constantly disturbed by removal of
GC-FID
Petroleum
Distillates
GC-MC
Trap
E1385 Steam distillation
Heat
GC-FID
Debris Concentrate
GC-MS
Headspace GC-FID
Syringe
Sample GC-MS
GC-FID
Charcoal
strip
GC-MS
Extraction
Heat E1412 Passive headspace absorption/elution
Absorbent
(Charcoal)
GC-FID
Ne/He
CS2
Extraction GC-MS
GC-FID
GC-MS
SPME
fiber
E2154 Solid-phase microextraction
product, the volatiles are efficiently extracted and trapped. Solvents are used to desorb
the sample traps. DHS methods are particularly effective with low concentrations of
residual accelerants.
analytes (debris) d
S analytes (vapor)
Another passive method of vapor preconcentration is solid phase microextrac-
tion (SPME). The procedure is much like that used for the charcoal strip, except that
other adsorbents are used to coat the silica needle. The needle can be directly intro-
duced into the gas chromatograph for thermal desorption, or a solvent extraction can be
employed.7 An added advantage of SPME is versatility: The fiber can be immersed in
an aqueous matrix if the fire debris is waterlogged. Currently, ASTM lists as a screening
technique standard E 2154-01, but high sensitivity and a solventless approach make the
SPME method increasingly attractive. A chromatogram of the same sample prepared
by different methods is presented in Figure 11.25. Note the peak to the far left on the
lower two frames, attributable to the solvent used in the extraction. Detection was with
a flame ionization detector (FID), and the sample was gasoline on charred carpet.
No matter what sample preparation method is used, discrimination occurs, and
there are inherent limitations. Analyses of fire debris samples are designed to detect a
wide range of compounds and, as a result, are not optimized for any one compound.
Headspace methods will be biased toward the more volatile materials, even under con-
ditions of gentle heating. Excessively aggressive heating can drive off the more vol-
atile fractions, so heating temperature is limited and typically falls into the range of
approximately 70C. The efficiency of a solvent extraction, like any other partitioning,
will depend on the relative polarities of the solutes and solvent; in a complex hydrocar-
bon mixture, discrimination is inevitable, and some compounds will be desorbed more
efficiently than others. Similarly, not all components of an accelerant will be adsorbed
with equal efficiency onto charcoal or other solid phases.8 These caveats do not mean
that the methods are fatally flawed, but they do mean that the limitations of each tech-
nique must be understood and that validated methods are essential. In the case of fire
debris, the conditions that are optimal for the collection of gasoline components are
likely not optimal for heavy distillates, and vice versa; the discrepancy is even more
critical when nonpetroleum products, such as methyl ethyl ketones or industrial clean-
ing solvents, are involved. However, keep in mind that the transfer (from debris to in-
strument) does not have to be 100% efficient for every compound that might be in the
matrix. Rather, the transfer of each target compound must be in an acceptable range and
also must be reproducible. The analysis of fire debris is qualitative, not quantitative,
and is based on pattern matching, not the presence or absence of any one component.
As we will see, pattern matching is widely used in materials analysis as well.
Regardless of the type of sample preparation, the instrumental method employed
for fire debris analysis is gas chromatography, coupled to either a flame ionization de-
tector or a mass spectrometer. Unlike other chromatographic methods used in forensic
chemistry, the primary goal (in most cases) is to recognize patterns rather than identify
specific compounds. The pattern of gasoline (Figure 11.25) differs significantly from the
pattern of diesel fuel (Figure 11.26, bottom frame), which is composed of heavier and less
volatile hydrocarbons. With the use of mass spectrometry, the patterns can be further
analyzed to identify significant groups of compounds within a sample, such as aromat-
ics, alkanes, and branched alkanes.911 In addition to recognizing patterns and groups
of compounds, the analyst must consider environmental factors. Accelerants undergo
weathering, and their composition changes over time. The changes are predictable in
that the more volatile a compound is, the more quickly it will be lost. For gasolines, then,
weathering is more of an issue than it is for diesel fuel. Understanding weathering and
the analysis of weathered samples is essential to interpreting analytical results.
55
Also essential in any fire debris analysis is the collection and analysis of background
samples (matrix controls).1213 For example, if debris suspected of containing an accelerant
is collected on a carpet, samples of undamaged carpet should be collected as well, if at all
possible. Carpeting is manufactured from synthetic fibers (Chapter 16), the raw materials
of which are polymers, many derived from petroleum products. As seen in Figure 11.26,
many common materials produce patterns that could be confused with accelerants. The
data shown in this figure were collected by means of passive headspacecharcoal strips
and heating at 80C for 16 hours. The instrument used was a GC-MS, but the patterns of
the total ion chromatogram (TIC) are comparable to patterns that would be obtained from
a GC-FID. The figure illustrates the vital nature of controls in fire debris analysis.
An officer submitting fire debris requests that you analyze it for the presence of gasoline with
ethanol. Would you use GC-FID or GC-MS?
Answer:
GC-MS. The FID detector would respond to ethanol (it is widely used for blood alcohol analy-
sis), but the response is not specific. Because a complex pattern of peaks would be expected
from such a sample, it would be difficult to definitively identify one as ethanol, although it
would likely be one of the earliest eluting peaks. A mass spectrometer could provide definitive
identification of ethanol via the compounds mass spectrum coupled with retention time data
and comparison to reference ethanol standards.
Mineral Spirits
T1C C11
C10 C12
Time 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0
C9
Time 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0
C12
Time 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0
WD 40 Spray Lubricant
T1C C10
C9 C11
Time 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0
Raid Insecticide
T1C C12C13
C11 C14
C15
C10
Time 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0
Summary
From flame to bomb, the underlying chemistry of apply to intentionally set fires in arson. We also dis-
combustion is the same: a rapid oxidative decompo- cussed how forensic analytical chemistry approaches
sition that produces large volumes of hot expanding arson investigations, taking advantage of the volatile
gases. A slower reaction front produces deflagration, nature of many accelerants. In the next chapter, we
whereas a rapid and confined mixture can detonate. will move into the chemistry and evidence associated
In this chapter, we discussed combustion chemistry with firearms.
generically and then focused on how these principles
Problems
FROM THE CHAPTER with a mixture of about 94% NH4NO3 and 6% fuel oil.
What is the approximate oxygen balance of fuel oil?
1. Suppose that 1.00 gram of nitroglycerin is used in a 4. Three compounds that have been used as fuel in cars
firearm as a propellant. Suppose also that combustion and racing cars are ethanol, nitromethane (CH3NO2),
is 100% efficient and that 65% of the chemical energy is and gasoline. Using octane as representative of gasoline,
transferred to a bullet that weighs 115 grains. How fast calculate the following quantities for each compound:
will the bullet be moving? Will it exceed the speed of DH of combustion per mole of the fuel, moles of gas pro-
sound? duced per mole of fuel at 25C and 1 atm pressure, Q pro-
2. How is a shotgun like a pipe bomb in terms of energy duced per gram of fuel, liters of gas produced per gram
conversion? How is it different? of fuel at 25C and 1 atm, and the QV value. Assume that
3. ANFO is a powerful explosive mixture containing am- all components are in the gas phase. Which fuel would
monium nitrate and fuel oil. It was used in the 1995 be the best choice based on these considerations? You
bombing of the Alfred P. Murrah Federal Building in may need to look up some heat of formation values.
Oklahoma City. Optimal power, related to Q, is obtained
5. Through a balanced equation of combustion, calcu- high school welding shop. You call the supplier and
late the oxygen balance of the following compounds. learn that a full tank contains 25 lb of acetylene. The
Assume that the combustion is complete and produces shop teacher says the tank was new and barely used.
fully oxygenated species. Estimate the LFL and UFL in pounds and determine if
a) Ammonium nitrate a combustible mixture would be supported. Assume 1.0
b) HMX (Octogen) atm pressure and a typical indoor temperature of 25C.
c) Picric acid 10. Is SPME a destructive analysis? Justify your answer.
6. The propellant used in the solid rocket boosters dur- 11. Two common solvents used in clandestine drug labo-
ing the space shuttle program was based on aluminum ratories are diethyl ether and acetone. Being less than
metal and ammonium perchlorate. If these were the vigilant in laboratory and safety practices, clandestine
only two ingredients, what composition would pro- chemists often work with leaky equipment. If a person
duce a net zero oxygen balance? was brought to an emergency room under suspicious
7. Assume that the two main ingredients in a gunpowder circumstances, where would you predict the burn pat-
formulation are nitrocellulose (NC) and nitroglycerin terns on the persons body to be most pronounced if
(NG). If these were the only active ingredients, what he or she was injured by a fire or explosion at a clan-
composition would produce a net zero oxygen balance? destine laboratory?
Assume that the oxygen balance of NC is 24%.
8. According to the UFL and LFL for flammables listed in
FOOD FOR THOUGHT
Table 11.2, which would be better choices for setting
an arson fire based strictly on these criteria? Why? 1. Hydrogen is billed as the fuel of the future for au-
9. You are called to a fire scene in which the point of origin tomobiles. A popular misconception, mostly owing to
appears to be in a restroom in a high school. The dimen- films of the Hindenburg disaster, is that cars that store
sions of the room are 25 ft 3 10 ft 3 10 ft. In one corner, hydrogen as fuel will be more likely to explode in an
you find the burned remains of a small tank labeled accident than current cars that use gasoline. Why is this
acetylene that appears to have been stolen from the a misconception?
Further Reading
Akhavan, J. Thermochemistry of Explosives. Chap. 5 in Kelly, J. Gunpowder: Alchemy, Bombards, and Pyrotechnics:
The Chemistry of Explosives. Cambridge, U.K.: Royal The History of the Explosive That Changed the World. New
Society of Chemistry, 1998. York: Basic Books, 2004.
Almirall, J. R., and K. G. Furton, eds. Analysis and Mark, P., and L. Sandercock. Fire Investigation and Ignitable
Interpretation of Fire Scene Evidence. Boca Raton, FL: CRC Liquid Residue AnalysisA Review: 20012007. Forensic
Press, 2004. Science International 176, nos. 23 (2008): 93110.
DeHann, J. D., Kirks Fire Investigation, 7th ed. Upper Saddle Meyer, R., J. Kohler, and A. Homburg. Explosives, 6th ed.
River, NJ: Prentice Hall, 2011. Weinheim, Germany: Wiley-VCH, 2007.
Dolan, J. Recent Advances in the Applications of Forensic Pert, A. D., M. G. Baron, and J. W. Birkett. Review of
Science to Fire Debris Analysis. Analytical and Analytical Techniques for Arson Residues. Journal of
Bioanalytical Chemistry 376, no. 8 (2003): 116871. Forensic Sciences 51, no. 5 (2006): 103349.
Glassman, I. Combustion, 3d ed. San Diego: Academic Press, Turns, S. R. An Introduction to Combustion, 2d ed. Boston:
1996. McGraw-Hill, 2000.
References
1. Akhavan, J. Thermochemistry of Explosives. Chap. 5 3. Turns, S. R. Some Important Chemical Mechanisms.
in The Chemistry of Explosives. Cambridge, U.K.: Royal Chap. 5 in Introduction to Combustion: Concepts and
Society of Chemistry, 1998. Applications, 2d ed. Boston: McGraw-Hill, 2000.
2. Turns, S. R. Laminar Premixed Flames. Chap. 8 in 4. Gamboa, J. A., et al. Rate Controlling Factors in a
Introduction to Combustion: Concepts and Applications, 2d Bunsen Burner Flame. Journal of Chemical Education 80
ed. Boston: McGraw-Hill, 2000. (2003): 52428.
5. NFPA. 2011. An Overview of the U.S. Fire Problem. 9. Dolan, J. A., and Stauffer, E. Aromatic Content in
National Fire Protections Association, 2010. Available Medium Range Petroleum Distillate ProductsPart I:
from http://www.nfpa.org/assets/files/PDF/ An Examination of Various Liquids. Journal of Forensic
FireOverview.pdf. Accessed January 2011; Sciences 49 (2004): 9921004;
6. NFPA. 2011. Intentional Fires Fact Sheet . National 10. Gilbert, M. W. The Use of Individual Extracted Ion
Fire Protections Association, June 2010. Available Profiles versus Summed Extracted Ion Profiles in Fire
from http://www.nfpa.org/assets/files/PDF/ Debris Analysis. Journal of Forensic Sciences 43 (1998):
FireOverview.pdf. Accessed January 2011. 87176.
7. Harris, A. C., and J. F. Wheeler. GCMS of Ignitable 11. Wallace, J. R. GC/MS Data from Fire Debris Samples:
Liquids Using Solvent-Desorbed SPME for Automated Intepretation and Applications. Journal of Forensic
Analysis. Journal of Forensic Sciences 48 (2003): 4146. Sciences 44 (1999): 9961012.
8. Lloyd, J. A., and P. L. Edmiston. Preferential Extraction 12. Lentini, J. J. Persistence of Floor Coating Solvents.
of Hydrocarbons from Fire Debris Samples by Solid Journal of Forensic Sciences 46 (2001): 147073.
Phase Microextraction. Journal of Forensic Sciences 48 13. Lentini, J. J., et al. The Petroleum-Laced Background.
(2003): 13034. Journal of Forensic Sciences 45 (2000): 96889.