"Laboratory Services" pp 328-361 from
Kirk's Fire Investigation, 4/E by John D. DeHaan, ©1997.
Reproduced by permission of Prentice-Hall, Inc., Upper Saddle River, NJ.
Permission from Prentice-Hall is required for all other uses.
[interFIRE VR Note: Figures and Tables are not reproduced.]
There are instances where even an experienced fire investigator requires
an analysis or interpretation of the physical evidence relating to fire
causation or spread which entails the services of a laboratory specialist.
In fact, it is the wise and experienced investigators who recognize the
limitations of their own knowledge, beyond which they must take advantage
of other experts to ensure that the investigation is complete and accurate.
The less experienced may try to overextend their expertise and offer erroneous
or misleading interpretations of evidence which can embarrass themselves
and their clients. It is well worth the time and effort for the investigator
to anticipate the need for such experts* and determine their availability
before they are needed in a crucial case.
AVAILABILITY OF LABORATORY SERVICES
Most states in the United States have at least one state criminalistics
laboratory that can provide most of the analytical services a fire investigator
will need. In fact, many states have a system of regional laboratories to
minimize travel time and expense. These laboratories, operating under the
auspices of the state's department of justice, public safety, state prosecutor,
state police, or similar law enforcement authority, work in cooperation
with the city or county laboratories found in many areas to provide a wide
range of services. On a national level, the laboratories of the Federal
Bureau of Investigation and Bureau of Alcohol, Tobacco, and Firearms are
well known. In Canada, in addition to city or provincial laboratories in
Vancouver, Montreal, Ottawa, and the Centre of Forensic Sciences in Toronto,
there is a system of excellent laboratories operated by the Royal Canadian
Mounted Police. With some exceptions, the services from these public laboratories
are available at no cost to personnel from public fire or police agencies
and to investigators from some semiprivate agencies such as transit authorities,
public utilities, railroads, and the like. Due to limited laboratory resources,
there may be restrictions on which agencies they are permitted to serve.
In addition, there may be a case priority scheme operating which relegates
fire-related evidence to low-priority status. Since the priority is often
set on the basis of whether there were deaths or injuries or on a court
date having been set, that information should be provided to the lab when
the evidence is brought in to ensure fastest possible service time. The
investigator should contact the nearest laboratory before its services
are needed, so that time is not wasted just in finding a laboratory.
Private investigators, whether retained by public agencies such as public
defenders or by private or corporate clients in civil or criminal cases,
have a more difficult and costly problem. Although a few public laboratories
will service public defenders and even private-sector investigators on a
contractual basis, for the most part, private-sector investigators must
turn to private laboratories. Many large insurance and engineering firms,
in addition to providing laboratory services to their own investigators,
offer such services to outside clients on a fee basis. There are also small,
private laboratories that can perform appropriate tests on a per-case, per-analysis,
or hourly fee basis and provide expert testimony as to their results. Finally,
there are qualified consultants who will evaluate the evidence and provide
interpretation and expert testimony while contracting with others for the
actual laboratory work. There are many more qualified, private-sector facilities
available now than just a few years ago. Many also do work (on a fee basis)
for public-sector agencies whose own labs cannot provide timely services.
The most important consideration is not whether a lab has the necessary
equipment but the experience necessary to provide accurate interpretation.
The effects of fire, evaporation, exposure, and even sample collection and
storage on physical evidence from fires must be well understood by the analyst
if critical errors in interpretation are to be avoided. Simply having a
chart or numerical reading does not do the investigator any good. Only a
qualified analyst will be able to correctly interpret that data and guide
the investigator in assaying the contribution of the evidence to the case.
Unfortunately, judging someone's qualifications is much more difficult than
dialing a phone number, and there have been instances of malfeasance and
even fraud on the part of some laboratories. The International Association
of Arson Investigators (IAAI) first addressed this problem by establishing
a Forensic Science Committee and developing consensus guidelines on the
analysis and interpretation of fire-related evidence.
American Society for Testing and Materials (ASTM) used that IAAI document
as the foundation for what is now a series of test methods, guides, and
practices covering many of the techniques used in fire debris analysis,
including ASTM "E 1492: Practice for Receiving, Documenting, Storing,
and Retrieving Evidence in a Forensic Science Laboratory,"
"E 1387, Standard Test Method for Ignitable Liquid Residues in Extracts
from Samples of Fire Debris by Gas Chromatography."
methods were developed by consensus of the scientists involved and they
can be used by any analyst. Adherence to them indicates a professional attitude
on the part of the analyst. The American Society of Crime Laboratory Directors
(ASCLD) has established a protocol for accreditation of forensic laboratories
that calls for use of such standardized methods for analysis and interpretation,
as well as record keeping. Finally, the American Board of Criminalistics
(ABC) has established a certification program for individuals involved in
fire debris analysis. Scientists who pass the General Knowledge Examination
in forensic science and a specialty knowledge exam and who participate in
a quality assurance testing program may be designated to be a Fellow of
the ABC in that specialty. In addition, active membership in IAAI or one
of the regional or national forensic science organizations is a hallmark
of a professional forensic scientist. These qualifications should aid the
investigator in finding a suitable laboratory and a qualified analyst. The
selection of a qualified lab is not easy and there are too many variables
of expertise, qualifications, analysis time, and costs to provide specific
guidelines here. The laboratory services discussed here cover the entire
range of those available from many sources. With the information offered
on types of evidence, possible analyses, and strengths and weaknesses of
evidence and techniques, the investigator should be able to confer knowledgeably
with various experts and determine their suitability for the particular
problem at hand.
GENERAL FIRE EVIDENCE
Identification of Charred or Burned Materials
One of the most common questions a fire investigator asks is, "What
is (or was) this stuff-and does it belong at the scene?" In earlier
chapters we discussed the importance of looking for things that appear to
be out of place. In a fire, many items, both trivial and significant, can
be sufficiently damaged that a casual examination will not result in their
identification. Plastic, rubber, or metal objects having a relatively low
melting point are typical problems. The identification may be based on the
shape of the remaining portion, portions of labels or details cast into
the object that survive, or a determination of the type of material present.
Even though the plastic or metal can be chemically identified, if only minute
amounts remain unburned this information may not be of use in establishing
the specific item. Some materials, like polyethylene, polystyrene, or zinc
diecast, seem to be universal in their applications. Some plastic bottles
may be identified by their shape or casting marks, but the knowledge that
a puddle of debris is melted polyethylene is of little use in figuring out
which of thousands of polyethylene objects it may represent. Glass objects,
whether broken by mechanical or thermal shock, can be pieced back together
(as in Figure 14-1) if there has not been excessive melting of the pieces.
Laboratory identification may consist of visual examination, reassembly,
cleaning, or microscopic, chemical, or instrumental tests of considerable
complexity, depending on the nature of the material, the quantity recovered,
and the extent of damage. By verifying the composition of even partial remains,
these services may be crucial in detecting the substitution of less valuable
items for heavily insured objects like jewelry, art, or clothes in cases
of insurance fraud, or even in confirming their complete removal in cases
of fraud or theft. If an ignition or time-delay device is suspected, the
laboratory may be able to reconstruct the device. The nature or specific
details of the device may then be of use in identifying the perpetrator,
linking him or her with the scene, or in interrogating suspects.
The reconstruction and identification services described earlier can
extend to fire-damaged documents. Writing paper has an ignition temperature
of approximately 230°C (450°F), and single sheets of paper exposed
to fire in air will readily char and burn almost completely. However, most
documents in files, stacks, or books will not burn completely but will char
and, if intact, may be identified. Identification in the laboratory may
employ visual examination under visible, ultraviolet, or infrared light;
photographic techniques; or treatment with glycerine, mineral oil, or organic
solvents to improve the contrast between the paper and the inks. Some clay-coated
papers and writing papers with a high percentage of rag content are more
resistant to fire and are more easily restored than cheaper wood-pulp papers
like newsprint. The most critical problem is the fragility of the charred
papers. Even fragmentary remains can sometimes be identified, but the more
intact the document, the better the chances of identification. When charred
paper is to be recovered, the investigator should take every precaution
against physical destruction of the remains. Documents should be disturbed
as little as possible and carefully eased, using a piece of stiff paper
or thin cardboard, onto a cushion of loosely fluffed cotton wool. This cotton
is placed into a rigid box of suitable size and a layer of fluffed cotton
wool placed over the documents to keep them in place. (Rigid plastic containers
like those used for cottage cheese are excellent for small items like matchbooks.)
The box should then be hand-carried to the laboratory. Reconstruction is
best done under controlled laboratory conditions by experienced personnel
with proper photographic support in case characters are visible only fleetingly.
For these reasons, the investigator should not attempt to perform any tests
in the field. Virtually anything done to the document in the way of coating,
spraying, or treating it with solvent will interfere with laboratory tests.
Many general crime laboratories and qualified document examiners have the
expertise and equipment needed to perform such tests properly.
Failure Analysis--Forensic Engineers
If a mechanical system has failed and has caused a fire, it may be necessary
to determine the causation of the original mechanical fault. A registered
mechanical or materials engineer is often able to tell if a driveshaft or
bearing failed as a result of misuse, overloading, poor maintenance, or
design error. The failure of hydraulic systems may produce leaks of flammable
hydraulic fluid that can contribute to the initiation or spread of a fire,
and the experience of a hydraulic engineer may be required. Because the
civil liability of the manufacturer or user of a mechanical system may be
in the millions of dollars, it is very important to have an accurate estimation
of the condition of that system. Such determinations are beyond the capabilities
of most crime laboratory personnel since they require specialized knowledge
of the materials or processes involved. A registry of qualified professional
engineers is offered in many states to assist investigators.
Evaluation of Appliances and Wiring
The guidelines offered in Chapter 10 will assist the investigator in
making an accurate assessment of the condition of wiring and the contribution
it may have made to a fire. Many diagnostic signs, however, require microscopic
examinations, metallurgical tests, or continuity and conductance tests that
require laboratory examination and interpretation by a specialist. For instance,
whether a switch was on or off at the time of the fire requires careful
dismantling and examination of the contacts, housing, and switch mechanism.
Due to fire damage, such dissection may easily result in movement of the
components and destruction of their precise relationship. X-ray analysis
may be used if the contacts and housing are arranged to permit cross-sectional
viewing. Twibell and Lomas described a novel resin-casting technique that
allows cross-sectioning of the switch unit while ensuring that the components
have not moved.4
Soft x-rays (of the energies used for medical procedures, 7585 kV)
have also been demonstrated to be of value in examining thermal cutoffs
(TCOs) used to protect coffee makers and other similar appliances from overheating,
and cartridge-type fuses. Such x-rays can be used to determine their continuity
and often to establish the mechanism of their operation or cause of failure.
In the case of fuses, those that failed due to moderate overcurrents were
often differentiable from those that failed in a short circuit.5 Under some conditions,
whether a wire was melted by overheating due to excessive current or by
the heat of the fire may be determined by the degree of crystallization
within the conductor, because the entire cross section of a conductor will
be heated by an overcurrent, where the surface of a conductor is likely
to be affected to a greater degree than the core if the heat is applied
from the exterior. It has been suggested that analysis of the elemental
contaminants trapped within a copper conductor should reveal a record of
the environment which surrounded it. If the analysis can be carried out
to examine elemental content as a function of depth within a melted copper
or aluminum bead, a sequence of events might be reconstructed.6,7 Auger electron spectroscopy
(AES) and electron spectroscopy for chemical analysis (ESCA) both allow
elemental analysis of a surface followed by etching away of the surface
(using argon atoms) so that analysis at various depths can be carried out.
Scanning electron spectroscopy with x-ray analysis (SEM/EDX) is also capable
of elemental mapping. It has been suggested that if a conductor is subject
to a direct fault, the absence of insulation in the vicinity is reflected
by carbon, chlorine, and calcium (from PVC insulation) and oxygen (from
the normal air environment) being found in the outermost layer of the bead.
If the fire has created the fault by charring the insulation and the wire
together, the carbon, chlorine, and calcium should be found at greater depths
into the conductor, while oxygen concentration should be diminished due
to the surrounding flames.6,7
Metallurgists can conduct crystallographic tests but, unfortunately, the
elemental analysis (AES or ESCA) requires expertise and equipment available
only in fairly sophisticated materials- or surface-science laboratories.
This method is presently awaiting confirmation by SEM/EDX. It should be
kept in mind that, although the laboratory specialist can perform tests
far beyond the capability of the investigator, there are practical limits.
It is possible for a fire to be sufficiently hot and prolonged to destroy
enough of the diagnostic signs that may have once been present that the
required determination cannot be made with any degree of certainty. Every
laboratory has its quota of miracles that it can perform!
Miscellaneous Laboratory Tests
There are a number of other determinations having an impact on the assessment
of a fire that the laboratory is well-suited to perform. The temperatures
produced during the fire may give an indication of the use of an incendiary
ignition device or an unusually high fuel load. These temperatures may be
indicated by the melting (or melted state) of various materials in the vicinity.
Since small variations in composition may result in considerable differences
in melting points, the objects of interest should be identified in the laboratory
and their melting points measured directly if at all possible. Scanning
electron microscopy (SEM) has been shown to be of value in establishing
the nature of interactions between aluminum and copper in forming eutectic
alloys during a fire.8 SEM can show differences in crystal structure, and when
teamed with elemental analysis (SEM/EDX) it can demonstrate mixing of various
Flash Point. The flash points of materials found at the fire
scene, either as part of the scene itself or as part of an incendiary device,
should also be determined in a laboratory. Flash point determinations are
made using one of several types of apparatus: Tagliabue (Tag) Closed Cup
(FP below 175°F), the Pensky-Martens (FP above 175°F), Tag Open
Cup, or Cleveland Open Cup. The American Society for Testing and Materials
(ASTM) has established standard methods for the use of these devices (previously
described) and the reader is referred there for more details.9 Unfortunately, many laboratories
cannot measure the flash point of a flammable liquid if that point is below
ambient (room) temperature. In addition, the flash point apparatus most
commonly found in analytical labs requires at least 50 milliliters (2 fluid
ounces) for each determination. This is far more than is normally recovered
as residue from an incendiary device and precludes such analysis. However,
a flash point apparatus, called the Setaflash Tester, has found acceptance
in forensic laboratories because it will accept samples as small as 2 milliliters
as well as permit subambient measurements.10
Mechanical Condition. In reconstructing the scene, visual
examination is often adequate to ascertain whether a door lock mechanism
is locked or unlocked at the time of the fire (see Figure 14-2). The lock
mechanisms are usually of brass or zinc diecast, which have relatively low
melting points. As a result, they will often seize at temperatures produced
in normal structure fires. When recovered at a scene, they should be photographed
in place and then worked to clearly indicate their orientation. If it has
not been exposed to too much direct fire, it is possible to ascertain whether
a door was open or closed at the time of the fire by an examination of its
hinges. In a closed door, the plates of a typical butt hinge are protected
from fire by the door jamb and the edge of the door. The spine of the hinge
will be exposed to the fire and will, therefore, be more heavily damaged
than the more protected plates, as in Figure 14-3. In an open door, the
spine and plates of the hinges will be exposed to roughly the same amount
of heat, and damage will be uniform on both. If the fire is not too extreme,
discoloration of the metal and remnants of paint on the hinge may indicate
the relative positions of the two plates. It is possible for the fire damage
to be severe enough to erase the diagnostic signs for this determination.
Since such fire damage is more likely near the ceiling of a structure, if
it is important to know the position of a door, be sure that all (usually
three) hinges from a door are recovered and labeled as to top, bottom, and
middle. Photos of the hinges in place may also be helpful (along with photos
of any protected area on the floor or carpet). These determinations, like
the others in this section, are within the capabilities of many experienced
Fire and Smoke Hazards. In accidental fires, the specialist
laboratory can be of assistance in assessing the contribution made by various
materials to the fuel load. Carpets, particularly early polyester materials,
can represent a hazard because of their flammability. Although furnishings
sold in some states must pass flammability tests before they can be sold,
this is not the case nationwide. Furnishings with synthetic or cotton coverings
and cotton or polyurethane fillings can, therefore, present a likely starting
point for an accidental fire. If any of the coverings or fillings remain
unburned, it may be possible to identify them, duplicate them, and test
the combination for susceptibility to accidental ignition (see Chapter 11).
Such testing, although it can be done informally by most laboratories, requires
considerable experience if the results are to be used in civil litigation.
Some states have agencies, like California's Department of Consumer Affairs,
Bureau of Home Furnishings, which will test such hazards on a contract basis;
otherwise, a private consumer or materials laboratory will have to be contacted.
As part of a reconstruction of the cause and manner of death in fatal
fires it may be necessary to determine the origin of toxic gases that caused
the death or heavy smoke which prevented the escape of still-conscious victims
from a small fire. Although the production of toxic gases and dense smoke
is largely a variable of the temperatures and ventilation during the actual
fire, some laboratory tests can be carried out to evaluate potential hazards.
Work of this type has been done to test upholstery materials for closed
environments such as airplane interiors, but it is costly and time consuming
and may not be available to most fire investigators. The NFPA has developed
standard tests for evaluating the fire hazards of interior finishes and
the reader is referred to its publications for specific guidance.11
IDENTIFICATION OF VOLATILE ACCELERANTS
By far the predominant service provided by laboratories to fire investigators
is the analysis of fire debris for suspected volatile accelerants. In one
laboratory study, volatile accelerants were found in 49% of all arson cases
submitted to a major crime laboratory over a three-year period.12 The use of volatile
accelerants, most commonly petroleum products, has been common throughout
the history of twentieth-century incendiarism. Many accelerants are used
in excess and their residues remain for ready detection by instrumental
analysis. Unfortunately, they can be used in moderation or on flammable
substrates like paper or plastic, and the resulting fire can evaporate all
traces of the accelerant. The author has participated in a number of structure
fire experiments where a petroleum distillate such as gasoline had been
used to ignite a fire and the debris, even though collected directly after
the fire, packaged properly, and analyzed by the best available methods,
failed to reveal any accelerant. A negative laboratory finding is not proof
that an accelerant was not used, it merely fails to establish its presence
after the fire. In addition, petroleum products change in their physical
and chemical properties as they burn or evaporate during (and after) a fire,
and their specific identification can be a challenge even to experienced
Gas chromatography is used both for the screening of samples to determine
which contain adequate volatile ignitable liquids for identification and
for the identification itself once the volatile has been isolated. Although
gas chromatographs (called GCs) vary considerably in their complexity, their
basic principles of operation are quite simple. All types of chromatography
are means of separating mixtures of materials on the basis of slight differences
in their physical or chemical properties. Gas chromatography uses a stream
of gas (nitrogen or helium) as a carrier to move a mixture of gaseous materials
along a long column or tube filled or coated with a separating compound.
The components in the mixture interact with the separating compound by alternately
dissolving in it and then volatilizing to be swept farther along the column
by the carrier. To ensure that all the compounds in the unknown mixture
remain in the gaseous state, the whole column is kept in an oven to maintain
it at a precisely controlled temperature. The mixture is injected onto the
column at one end and at the other there is a detector, usually a small
flame whose electrical properties, monitored by electronic circuits, change
whenever a compound comes off the column. The whole process may be likened
to the analogy of a stream of vehicle traffic being routed through a forest.
The motorcycles speed right through, zigzagging around the trees and come
out the other side of the forest first to encounter a radar speed trap.
The small cars have more trouble finding a path and come out shortly afterward
to encounter the same radar. The big cars take a long time to find spaces
big enough to allow passage and they exit after a longer delay, followed
by the trucks that take the longest of all.
In the case of volatile hydrocarbon analysis, our "traffic"
would be our unknown hydrocarbon mixture and our "forest" would
be the GC column. We can adjust the spacing of our "trees" by
choosing the correct chemicals for the separating compound and carefully
controlling the temperature and carrier flow so that each type of "traffic"
comes out in a neat, readily identifiable group. To extend the "forest-traffic"
analogy further, the proper choice of GC column and conditions for the hydrocarbons
found in fire debris allow the lab to identify the type of "traffic"
introduced by analysis of the groups of "vehicles" that result.
Gas chromatography as an accepted laboratory procedure has been in use for
the last 35 years or so. From simple, insulated box ovens that kept the
column at a set temperature, we have progressed to ovens that are programmed
to provide steady or increasing temperatures at preset time intervals to
maximize the separation of similar compounds. From simple thermal detectors,
technology has progressed to flame detectors that can be made sensitive
only to compounds containing specific elements.
Although the column can be selected to separate a mixture of compounds
on the basis of their chemical properties, most separations in volatile
accelerant analysis are made on the basis of differences in the boiling
points of the various components. Nearly all the components of petroleum
distillates of interest are either aliphatic (straight-chain) or aromatic
(ring) in structure, and thus will have very similar chemical properties
within those two groups. A typical separation of a series of aliphatic hydrocarbons
is shown in Figure 14-4. In this typical chromatogram, one can see that
the heavier a molecule is, the less volatile it is, and the later it will
appear on the chromatogram. Columns for GC analysis of suspected arson residues
are usually set up to offer the best separation of components having volatilities
between those of pentane (C5H12)
and triacontane (C30H62), since
that range includes all of the commonly encountered petroleum products.
Most examiners prefer a silicone-based stationary phase, like OV-1, OV-101,
or SP-2100, on an inert filler or support medium. Although the mainstay
of columns was for many years a packed tubular column, averaging 1/8 inch
(3 millimeters) in diameter and 10 feet (3 meters) in length, this has now
been supplanted by a new generation of very narrow (approximately 0.1 millimeter
in diameter) capillary columns of various lengths.13,14 Early capillary columns
required very long lengths (50 meters or more) to achieve good separations,
but these extreme lengths usually required long analysis times. Such columns
were used primarily for research or development where short analysis times
could be sacrificed for sensitivity and selectivity. The newest columns
provide extremely high performance on short columns (1025 m) using
high temperature-programming rates. The result is analysis times that are
shorter than those possible with packed columns and faster turnaround times.
Capillary columns also use smaller sample sizes than was ever possible with
packed columns, an advantage of much merit in forensic work where the sample
size is often critically small. Sensitivity is now in the nanoliter range
Examples of capillary chromatographs are seen in Figure 14-5.
Some analysts prefer to repeat the GC analysis on another column using
a stationary phase that discriminates by chemical properties. A packing
such as Bentone 34 affects aromatics (in gasoline these would be benzene,
toluene, xylenes, and alkylbenzenes) differently than it does the aliphatics.
As a result, the aromatics show up on the chromatogram in a discrete "envelope"
of peaks which is readily identifiable even at very low concentrations,
thereby improving the detectability of gasoline in complex or "dirty"
debris. Other columns can separate polar (oxygenated) species from nonpolar
In addition to the flame ionization detector, which has been standard
for analysis of hydrocarbons for a number of years, there are now specialized
detectors that can be used in arson work. Electron capture detectors, which
use a minute radioactive source as a monitor, are very sensitive to particular
chemical species such as organometallics and halogenated hydrocarbons. They
can be used to characterize a petroleum distillate more precisely than is
possible with flame ionization alone. These detectors can be used to characterize
the lead additive package sometimes still found in gasolines as an octane
Naturally such additives are not found in unleaded gasolines, and some petroleum
companies tend to share the same additive package, making brand discrimination
uncertain; however, there are occasions when such information can be of
use. There are nitrogen-phosphorus detectors that are not sensitive to "normal"
hydrocarbons but will select those compounds in a complex mixture that contain
nitrogen or phosphorus. By monitoring the color of the detector flame, specialized
detectors can be made specific for compounds containing sulfur. Since the
crude oil feedstocks for gasoline vary considerably in their nitrogen-,
phosphorus-, and sulfur-containing components, it is very likely that such
techniques may allow discrimination between gasolines from different origins.
Such discrimination has been greatly needed in the past when comparing the
gasoline from an incendiary device to gasoline from a suspected source.
In addition, progress in microelectronics has dramatically changed the
data collection and handling in gas chromatography. Runs (and irreplaceable
samples) are no longer lost when the recording conditions are not set just
right, since the data system collects all the data and then displays whatever
portion the analyst wants. The displayed data can expand or condense portions
of the chromatogram to make identification easier. Collected chromatograms
are also filed on minicomputer discs and can be compared against a library
of stored chromatograms of known products. Such libraries can be swapped
between instruments and even between labs, permitting ready intercomparison.
Gas Chromatography/Mass Spectrometry (GC/MS)
Flame ionization detectors (FIDs) are sensitive and stable detectors
but they yield little chemical information about the compounds that pass
through to produce a signal. Mass spectrometry allows the analyst to break
apart each compound into small submolecular pieces and, by counting those
pieces, establish the chemical structure of the original molecule. Mass
spectrometry alone is an analytical technique that needs to be fed one compound
at a time, while gas chromatography is a separation technique that is good
at separating things into groups. Since most of the volatile residues in
fire debris involve complex mixtures, coupling the two together gives the
best tool. While GC/MS has been available for more than 20 years, only recently
has it been made small, inexpensive, and user-friendly enough to make it
convenient for fire debris analysis. In addition to the capability of displaying
a mass spectrum for each peak, it can be asked to scan for selected ions,
those that are characteristic for a particular chemical species of interest.
This means that it can display a chromatogram just for aromatics, for instance,
ignoring all the peaks that do not yield that specific ion. This can make
it possible to pick out characteristic patterns even amidst complex chromatograms
with high backgrounds. The technique is well established and there is an
ASTM method for its use (ASTM E 1618).17,18
As of this writing most forensic gas chromatographic analyses for suspected
acclerants are being carried out using a capillary column of moderate length,
using SP-2100 (or OV-1 or an equivalent bonded phase), in a programmed temperature
range of 50°C to 250°C, and employing a flame ionization detector.
The discussions that follow are based on the results of such analyses.
Sample Handling and Isolation of Volatile Residues
As described in previous chapters, the best containers for debris suspected
of containing volatile ignitable liquids are clean metal paint cans, sealable
glass jars, or bags of nylon or a suitable copolymer (nylon/polyester/polyolefin
or nylon/polyethylene). The first step in the normal examination protocol
of such debris is to briefly open the container (in the case of bags, this
will usually not be necessary) and to inspect the contents. The nature of
the contents (wood, carpet, soil, etc.) may provide guidance in selecting
an analytical technique or predicting what interferences are likely to be
encountered. The contents should be checked against the labeled or indicated
contents to ensure there have been no errors in packaging or labeling that
could compromise the value of the evidence. The examination must be brief
to minimize losses of volatiles that may be present only at trace levels.
Very strong odors of a volatile detected during this examination may indicate
that a particular test protocol should be used. The major problem with fire
debris analysis is having a small quantity of product mixed with a great
deal of solids or liquids that may provide contamination. Each of the techniques
described here has advantages and disadvantages that must be evaluated in
light of the character of the sample.
Heated Headspace. Since any volatiles present will reach a vapor
equilibrium in a sealed container, we can take advantage of the vapors in
the air cavity (headspace) of the container. Heating the sample container
to 5060°C (120140°F) will ensure at least a partial vaporization
of heavier hydrocarbons prior to sampling. Some analysts prefer heating
the container above 100°C to force even higher boiling materials (such
as fuel oil) into the vapor stage. It is the experience of most other analysts,
however, that even the heavy ends of fuel oil will be detectable in samples
heated to moderate temperatures and, more importantly, that such excessive
temperatures encourage the degradation of synthetics (carpet, upholstery,
etc.) in the debris. These degradation products add to the complexity of
the chromatogram, possibly masking low levels of real accelerants, while
not contributing significantly to the sensitivity of the method.
Once the sample container has equilibrated at a moderate temperature,
it is punctured and a small (0.13 milliliter) sample of the headspace
vapor is drawn off using a gas-tight singe. This vapor is then injected
directly into the gas chromatograph. For many mixtures, positive results
of this injection may be adequate to permit characterization of the volatiles
present (as described in ASTM "E 1388: Standard Practice for Sampling
of Headspace Vapors from Fire Debris Samples"). It also signals the
presence of very volatile residues that may be lost be careless handling,
or alcohols or ketones that require special handling. Since this technique
is not as sensitive as others to be described, a negative result here merely
indicates that no high concentration of volatile exists, and that another
isolation technique is suggested. It is fast, convenient, and requires virtually
no sample handling, while providing useful information.
Passive Headspace Diffusion (Charcoal Sampling). Another technique
that capitalizes on the vapor equilibrium in a sealed container while concentrating
some of the vapors in a more condensed form is passive adsorption. This
actually constitutes several similar techniques that rely on the adsorptivity
of activated charcoal. Originally developed by forensic scientists in England
using a charcoal-coated wire, this technique requires no manipulation of
the sample or its original container. As originally described, the method
called for a coated wire to be inserted in the air space of the sample container
(as shown in Figure 14-6) and allowed to equilibrate for at least 2 hours.
The wire is then removed from the container and inserted into the pyrolysis
unit of a GC, where it is raised instantly to a very high temperature, driving
the absorbed hydrocarbons off into the column.19 This technique has been
adapted in a variety of methods described in ASTM "E 1412: Practice
for Separation and Concentration of Ignitable Liquid Residues from Fire
Debris Samples by Passive Headspace Concentration." These adaptations
include the insertion of a plastic or glass bead coated with activated charcoal
or a sampling tube or bag containing charcoal into the airspace of the container.
The container is resealed and allowed to equilibriate, usually at room temperature
for 12-15 hours or at 50°C-60°C for 2 hours. The sampling device
is removed and adsorbed volatiles are extracted from the charcoal with a
small quantity of solvent such as pentane, diethyl ether, or carbon disulfide
for injection into the GC. These techniques are very sensitive to all types
of volatiles, require no heating or manipulation of the samples, and can
be repeated again and again since they take up only a small fraction of
the volatiles each time.20,21
They are suitable for all kinds of debris, and are especially valuable for
debris that will be examined for trace evidence or latent fingerprints.
A variant of this technique has become the single most widely used sampling
and concentration method for fire debris. In this form, a small strip of
carbon fiber mat is suspended in the container, at room temperature or at
moderately elevated temperatures. It is then extracted with carbon disulfide
or diethyl ether. It has excellent sensitivity (0.1 microliter under some
conditions) and no interferences. It is described in ASTM E 1412.
A new variant, called solid-phase microextraction, uses a solid phase
absorbent bonded to a fiber that is inserted into the can of fire debris
through a hollow needle. After it has been exposed to the heated contents
of the can for 5 to 15 minutes, it is withdrawn and then inserted directly
into the injection port of a GC. The heat of the injection port then thermally
desorbs the volatiles directly into the GC column. The technique offers
extreme simplicity (with no manipulation of the sample or any extract) and
low cost, since it involves no solvents and no modifications to existing
Dynamic Headspace (Swept Headspace). This technique,
sometimes called the charcoal trap method, employs a charcoal or polymeric
matrix through which the air from the evidence container is drawn. It was
adapted from methods developed for trapping ultralow concentrations of hydrocarbons
for environmental monitoring.22
The sample container is fitted with a modified lid or a septum that allows
for the introduction of a heated carrier gas (purified room air or nitrogen).
The container is warmed while vapors are drawn off by a low vaccuum through
a cartridge filter containing activated charcoal or a molecular trapping
agent such as TenaxTM
(see Figure 14-7). Once extraction is complete (20-45 minutes), the charcoal
can be washed with carbon disulfide or a like solvent, and the resulting
solutions is ready for GC analysis. TenaxTM or charcoal may be thermally
desorbed to avoid the handling of toxic solvents, contamination, or any
further dilution. This technique requires the use of traps (commercially
prepared or laboratory prepared) and specially built apparatus, described
in ASTM "E 1413, Standard Practice for Separation and Concentration
of Ignitable Liquid Residues from Fire Debris Samples by Dynamic Headspace
Concentration." It is ver useful because of its sensitivity (less than
1 microliter) and its applicability to all ignitable liquid residues (from
gasoline to fuel oil) including alcohols and ketones. The GC analysis of
traps, which are thermally desorbed, has been automated, making it more
suitable for laboratories with a large caseload. It does have the drawbacks
that all of the volatiles in a sample can be swept out (and pulled through
the trap) if the time and temperature conditions are not monitored closely
and are too severe for the sample size. Excessive extraction produces loss
of the lighter fractions and the process can generally be carried out only
once on a sample.
Steam Distillation. The oldest technique, originating
from classical chemistry, requires only modest equipment. The specimen is
removed from its container and boiled with water in a glass flask (as in
Figure 14-8). The steam, carrying with it the volatiles, is condensed and
trapped. It is suitable for volatiles that form an azeotropic (nonmixing)
layer with water; therefore, it cannot be used to recover alcohol or acetone.
It is suitable for petroleum distillates as heavy as paint thinner; heavier
products (e.g., fuel oil) require the use of ethylene glycol, which boils
at a higher temperature.23
Light hydrocarbons may be lost in recovery. Detection limits are on the
order of 50 microliters (for gasoline), so it is not as sensitive as charcoal-trapping
methods, but it has the advantage of providing a neat liquid sample with
no extraneous solvents.24
The method is described in ASTM "E 1385, Standard Practice for Separation
and Concentration of Ignitable Liquid Residues from Fire Debris Samples
by Steam Distillation."
Vacuum Distillation. This well-known technique common
to organic chemistry requires only modest equipment (see Figure 14-9) and
a supply of refrigerant such as dry ice or liquid nitrogen. It is suitable
for the recovery of any volatile hydrocarbon, but excessive water in the
specimen may interfere with recovery. It is suitable for debris that is
susceptible to water damage, including fragile, charred documents. Recovery
yield is on the order of 60% for gasoline, while detection limits are on
the same order as steam distillation.25
Solvent Extraction. This very direct and simple method
for extracting volatiles has been in use for many years. Debris is extracted
or washed with a small quantity of n-pentane, n-hexane, dichloromethane,
or carbon disulfide that has been checked to ensure its purity. The extract
is then concentrated by evaporation in a stream of warm air until a small
quantity remains for testing. It is especially suitable for treating small,
nonabsorbent specimens (glass, rock, metal) or washing the interior of containers
used to transport ignitable liquid accelerants. It provides good recovery
for most petroleum products, especially the heavier ends of kerosene and
diesel fuel, but it is not suitable for isolating the light petroleum distillates
due to evaporative losses. Its overall sensitivity is about the same as
for steam distillation with the main drawback that it will extract residues
of partially pyrolyzed carpet or foam rubber, which often interfere with
the detection of low levels of hydrocarbons of evidentiary significance.
This technique requires only simple glassware and high-purity solvents.
It is described in detail in ASTM "E 1386, Standard Practice for Separnitable
Liquid Residues from Fire Debris Samples by Solvent Extraction."
It should be noted that most of the newer isolation techniques, especially
when combined with capillary column gas chromatography, are more sensitive
than the techniques used less than a decade ago. In many cases they are
more sensitive than the human nose, often relied upon to screen for the
best debris samples at the scene. This means that sample collection may
not depend on just the detection of suspicious odors but also on the presence
of suspicious burn patterns and on aids such as canine detection. Comparison
samples of floor coverings or other fuels involved are more desirable now
than ever before because of the contributions made by the pyrolysis of synthetic
materials, binders, and adhesives, which can interfere.25,26
Identification of Volatile Residues
Once the suspected ignitable liquid residues have been isolated, attention
can turn toward identifying them as precisely as possible. Because recoveries
are typically on the order of several microliters, gas chromatography will
be the technique of first choice because of the quantity of information
it yields while consuming 1 microliter or less of the sample. The identification
of the volatiles (usually a petroleum distillate or product) present is
then usually based on a comparison of the pattern of peaks generated by
a flame ionization detector (FID) in the chromatogram of the unknown against
a library of chromatograms of known materials made under the same conditions.
A standardized approach to the classification and characterization of volatiles
is described in ASTM "E 1387: Standard Test Method for Ignitable Liquid
Residues in Extracts from Samples of Fire Debris by Gas Chromatography."
Classification as to whether the volatile fits one of six categories--light,
medium, or heavy petroleum distillates, gasoline, kerosene, or miscellaneous--is
based on the presence of particular sequences of hydrocarbons or oxygenated
compounds. This classification is intended to make description and comparisons
easier for the laboratory analyst, but because consumer products may represent
different classes of distillates, and a single distillate may have more
than one product designation, the system is not always well understood by
the investigator. For this reason, it is recommended that when an analyst
describes an ignitable liquid residue as a particular classification, examples
of some common consumer products should be included, as in Table 14-1.
As discussed in a previous chapter, isoparaffinic hydrocarbon products
are being used in many consumer products from copier toners to insect sprays.
Since they are not true petroleum distillates, they should not be classified
as such, although in some instances they display a boiling point range that
fits classes 3 or 4. Examples of these products should give the investigator
a clue as to what product may have been used to start the fire or that a
product normally at the scene may have become involved accidentally in the
fire. Since the contents of many products are not always described accurately
on the labels, containers of any product in the vicinity of the fire's origin
should be collected for analysis if there is a chance they could be involved
in the fire.
While the GC/FID analysis described here is adequate for most products,
there is a growing number of ignitable liquids that are not petroleum distillates
and, therefore, do not lend themselves to a classification scheme based
on peak pattern; for instance, there are aromatic specialty solvents used
in insecticides, enamel reducers, and pavement sealers, and naphthenic/paraffinic
petroleum products used in lamp oils. The GC/FID patterns of these products
are not enough to permit their identification and GC/MS is required.
Special detectors may also be used to gain more information about a volatile.
Each requires injection on a separate GC column with a specific detector
in place of the FID, but some GCs have the ability to split a single injection
onto two separate columns, each with its own detector. Photoionization detection
(PID), for example, can be used in parallel with FID since it can be made
specific to aromatic, olefinic, or aliphatic hydrocarbons. Another detector,
originally intended for the detection of nitrogen- and phosphorus-containing
compounds, has been modified so it will detect specifically the oxygenated
volatiles (ketones, alcohols, and ethers) that have become widely used as
fuel additives to minimize air pollution.27
When distillation is used to isolate a neat volatile, infrared spectroscopy
can yield considerable information about its chemical structure. About 20
microliters of sample are needed for conventional liquid-cell spectrophotometry
but a specialized FTIR microcell has been designed for use as a detector
on a GC column when smaller quantities are recovered. While its sensitivity
is limited, it can aid in the identification of gasoline, complex pyrolysis
products, and some natural products (terpenes).28
If sufficient quantities of the accelerant are recovered, physical properties
like flash point and refractive index can be measured. Refractive index
can be measured on samples as small as 20 microliters, but it is generally
of little value in characterizing a volatile petroleum distillate. The criminal
charges against a suspected arsonist may depend on the flammability of the
accelerant in his or her possession, and a flash point determination may
be required. A precise determination requires a considerable sample and
specialized equipment, but a rough estimate can be made by placing several
drops of the liquid on a watch glass and placing it in a freezer. On removal
from the freezer, the glass will warm up gradually and a small ignition
source passed over the specimen will cause a flash when the flash point
is exceeded. If the liquid has a flash point above ambient (usually 25°C,
77°F), the glass can be placed on a hot plate and the sample tested
in the same manner. Recent advances in ultraviolet (UV)/visible fluorescence
technique permit extensive characterization of petroleum distillates and
this technique, along with high-pressure liquid chromatography, may be useful
in special circumstances because it provides information that supplements
the gas chromatography.29
Petroleum products behave in a fairly complex but predictable manner
when burned or simply exposed to air to evaporate. Due to their higher vapor
pressures, the most volatile components evaporate more quickly than those
less volatile, and this changes the peak profile of the product when it
is subjected to gas chromatography. As we can see from Figure 14-5, a gasoline
profile changes dramatically as it evaporates, complicating its identification.
When the gasoline is burned, this same process occurs, only much more rapidly.
It is not possible to determine whether a petroleum product burned or merely
evaporated at room temperature by examination of its chromatogram.30 In the case of gasoline,
there are characteristic aromatic compounds that are found in residues of
gasoline that have evaporated down to less than 3% of its original volume.
Therefore, evaporated or "weathered" gasoline can still be identified
if those key or target compounds are identified in the debris, as in Figure
14-10. Unfortunately, this is not true for simple petroleum distillates
like mineral spirits (paint thinners). As they evaporate, the residues assume
more and more of the characteristics of the next heavier class of petroleum
distillate. Thus, it is not always possible to conclusively establish the
exact nature of the original product. In addition, petroleum products when
exposed to moist garden soil can undergo microbiological degradation. This
degradation involves the consumption of both aliphatic and aromatic compounds,
so the peak profile of gasoline, for instance, can be significantly changed.31 These effects combined
with contamination from pyrolysis products in the debris can make the identification
of a specific petroleum product from postfire debris difficult. In most
cases, the laboratory can be expected to characterize the product or its
general family--for example, light (petroleum ether), medium (kerosene),
or heavy (fuel oil) petroleum distillates; lacquer thinners; and so on--from
residues left in the debris unless the accelerant has been nearly or totally
consumed. Any petroleum distillate, when used as an accelerant, can be burned
away completely, leaving no readily detectable traces, if the fire is hot
enough and long enough, and the accelerant is exposed directly to its effects.
Fortunately, in most arson-caused fires, accelerants are used in excess
and the resulting fires do not burn with sufficient intensity to completely
Laboratories are often asked to determine the presence of petroleum distillate
accelerants from the soot accumulated on window glass taken from the scene.
As we discussed in Chapter 7, the appearance of the soot is most directly
linked to the state of combustion in the room and the amount of ventilation.
Fuel-rich fires produce heavy soot; lean, or well-ventilated fires produce
little or none. The synthetic fibers and rubbers that constitute a large
percentage of the fuel load in modern structures produce an oily soot very
similar chemically to that produced by petroleum distillate accelerants.
It has been suggested that there should be chemical species in the soot
produced by gasoline which are characteristic for that fuel and are not
produced by any "normal" fuel. It was once thought that the combination
of bromine and lead in leaded automotive gasoline would produce a soot with
a unique elemental profile. Unfortunately, some carpets contain both elements
so there is a potential for background interference.32 More importantly, lead
and its related additives are being phased out of modern fuels, making them
less likely to be detected. There are several organic molecules, chiefly
polynuclear aromatic hydrocarbons, that appear to be produced by burning
gasoline but are not produced by burning synthetic materials.33 The
National Bureau of Standards conducted preliminary evaluations of this idea,
but the work was never completed and was not suitable for adaptation to
The most extensive work done in this area is that of Pintorini et al., who
used gas chromatography, combined with elemental analysis, x-ray diffraction,
and SEM to characterize soot from various ignitable liquids and that from
synthetic fibers and materials.35 They discovered that under controlled conditions, soot
from the liquids could be distinguished from that produced by solid fuels,
using multiple techniques. Unfortunately, the presence of water from suppression
produced nonreproducible condensation of the soot, and limited the discrimination
value of even the combination of techniques. Some analysts have reported
that freshly made carbon soot will absorb petroleum distillate fumes from
the room air just as activated charcoal absorbs them in laboratory extractions.
If a fire occurs and a too-rich mixture of fuel is present, the soot formed
during the early stages may retain enough of the accelerant to permit identification
upon extraction of the soot.
In spite of partial evaporation, most common petroleum distillates have
a distinctive GC pattern that can be compared with the patterns of known
compounds. The experienced analyst will know what they effective of fire
and evaporation can be and will take them into account when comparing chromatograms.
All petroleum distillates are overlapping "cuts" or fractions
of a continuous spectrum of compounds. Therefore, it is possible for a complex
hydrocarbon mixture to be sufficiently affected by fire that its chromatogram
cannot be readily identified. It could, for instance, represent a heavily
evaporated product such as kerosene, or a relatively undamaged product like
a heavy solvent.
As to the identification of a specific source for a product, the range
of techniques available now enables the analyst to rule out many potential
sources and thereby shorten the list of "possibles." The sophistication
of chromatography and today's data handling systems makes it possible to
recognize delicate shifts in the patterns of GC peaks between products that
were thought to be indistinguishable. Determining the significance of these
shifts is entirely another matter, requiring extensive field sampling and
validation of the reproducibility of patterns detected. As with many mass-produced
materials that start from different feedstocks in different manufacturing
facilities, changes can be produced by many factors. To complicate matters,
when gasoline, for instance, is distributed, it mixes with other gasolines
already in storage and transportation facilities and its composition changes.
Under some circumstances, such as having a liquid gasoline sample that has
not undergone significant burning or evaporation, and a small number of
potential sources, it is possible to state with some confidence that certain
sources can be completely excluded and that another represents a source
which is indistinguishable by any of its individual features.36 Rather than give the
investigator a false lead with a "possible" identification the
laboratory will usually suggest several alternative identifications in its
report. The investigator can evaluate the suggestions and determine which,
if any, fit the circumstances best. If the investigator does not clearly
understand the identification or characterization, the problem should be
discussed with the laboratory before proceeding.
Volatile acclerants are not the only arson evidence that is amenable
to laboratory testing. A small but significant percentage of all detected
arsons use a chemical incendiary as an initiator or primary accelerant.37 Fortunately, most chemical
arson sets leave residues with distinctive chemical properties, although
physically the residues may be nondescript and easily overlooked.
Safety flares or fusees are the most commonly used chemical incendiary,
presumably because they are effective, predictable, and readily available.
Red-burning flares contain strontium nitrate, potassium perchlorate, sulfur,
wax, and sawdust.38
Their postfire residue consists of a lumpy inert mass, white, gray, or greenish-white
in color. It is not soluble in water, but prolonged exposure to water turns
it to a pasty mush.
The residue contains oxides of strontium and various sulfides and sulfates.
Although strontium is found in all natural formations of calcium, only flare
residues contain strontium salts in nearly pure form. Elemental analysis
by emission spectroscopy, atomic absorption, or x-ray analysis will readily
reveal these high concentrations of strontium in suspected debris. The igniter-striker
button on the cap and the wooden plug in the base of the stick are resistant
to burning and may be found in the fire debris and may be identified as
to manufacturer by their physical features.39
The solid chlorine tablets or powder used for swimming pool purification
will react with organic liquids such as the glycols in automotive brake
fluid or some hair dressings to produce a very hot flame that lasts for
several seconds. This occurs after a time delay that varies with the concentration
of the chlorine-based components and the degree of mixing between the solid
and liquid phases. This reaction has been studied and a mechanism for the
production of ethylene, acetaldehyde, and formaldehyde by the action of
the hypochlorite on the ethylene glycol has been suggested and verified.40 The production of a
cloud of readily ignitable vapors would certainly account for the dramatic
yellow-orange fireball of flames that occurs in these reactions. The starting
solid compounds are usually not completely consumed and are not immediately
water soluble and may be visibly detected after the fire. The presence of
hypochlorite salts can be detected by infrared spectrometry or chemical
spot tests to confirm the identity of the oxidizer.
Potassium permanganate is used in a variety of chemical incendiary mixtures,
including a hypergolic (self-igniting) mixture with glycerine. It is soluble
in water and can be washed away. Even a weak water solution of it, however,
will have a pronounced red, green, or brown color, depending on the oxidation
state of the manganese ion. Such chemical ions are readily identified by
chemical test or by infrared if even a minute chip of the solid chemical
Flashpowders contain a fine metallic powder, usually aluminum, and an
oxidizer such as potassium perchlorate or barium nitrate. Elemental analysis
of the powdery residue left after ignition will reveal the aluminum, chlorine,
potassium, or barium. The perchlorate or nitrate residues are detectable
y chemical test if any of the mixture remains unreacted.41
Potassium chlorate is easily identified by chemical or instrumental tests
if it can be recovered in its unreacted state. When it reacts in an incendiary
mixture, however, its residues are almost entirely chloride salts, which
are innocuous and which, if recovered from debris, could not be readily
identified as having come from a device.
Sugar reacts with sulfuric acid to leave a puffy brown or black ash of
elemental carbon. It may have a faint smell of burning marshmallows, but
for the most part it is innocuous in appearance and is easily overlooked.
The acid present may char wood, paper, or cloth in its concentrated form.
It is very soluble in water, but mineral acids like sulfuric (H2SO4) do not evaporate and remain active and corrosive for
some time. A burning sensation in the skin on contact with debris indicates
that the debris should be checked for acids. The pH (acidity) can be easily
measured, and chemical tests for sulfate or nitrate ions carried out. Reactive
metals such as potassium or sodium can be used as an ignition device because
they generate great amounts of heat and gaseous hydrogen on contact with
water. The residues will contain potassium hydroxide or sodium hydroxide
(lye)--both very strong caustics that can cause skin irritation and burns.
The pH of water containing such residues will be very alkaline (basic),
and chemical or spectrophotometric tests will reveal the metal present.
It should be noted that the residues from these and similar devices are
very corrosive to metal. Debris suspected of containing such corrosives
should be placed in glass jars with plastic or phenolic lids, or in nylon
or polyester/polyolefin bags--not in metal cans, paper bags, or polyethylene
Since white phosphorus ignites on contact with air, it has been used
as an incendiary device in the past. (Feenian fire is a suspension of crushed
white phosphorus in an organic solvent. When the container is smashed, the
solvent evaporates, exposing the phosphorus to air with immediate ignition.)
Elemental analysis of the debris may reveal an elevated concentration of
phosphorus, and there are chemical spot tests, although it tends to be a
difficult element to detect and confirm.
A common source of phosphorus in incendiary mixtures is the household
match. Characterization of the components in match heads can be carried
out by elemental analysis alone or combined with microscopic analysis using
Many improvised chemical incendiaries are described in the literature
(too many to describe here).
Whenever such an incendiary is suspected, samples of nearby
materials should be collected for comparison purposes. Sometimes it is not
the presence of a particular element or ion but an elevated concentration
with respect to that found in the background that betrays the presence of
NON-FIRE-RELATED CRIMINAL EVIDENCE
The most frequent mistake made by arson investigators is a single-minded
focus on the fire itself and its origin, while ignoring other typs of physical
evidence. It is up to the arson investigator to construct an entire scenario
of a fire setter's activities at the scene. These activities are not necessarily
limited to the setting of the fire itself. Since arson may be used to destroy
evidence of other criminal acts, the careful investigator must be always
alert to the existence of evidence of "normal" criminal activity.
Many investigators assume, as do arsonists, that a fire destroys all
fingerprints. That is not true. While itt is true that many fingerprints
will be lost during a fire, many will survive, even on ignition sources,
as in Figure 14-11. Plastic (three-dimensional) impressions in window putty
or patent (visible) impressions in paint or blood may remain even after
diirect exposure to fire. In fact, they may be permanently fixed in place
by the heat. Plastic containers used to carry or hold gasoline in an arson
fire have been seen to be softened sufficiently by the gasoline to allow
the plastic to take on the fingerprint impressions of the person holding
the container. If such a container is not exposed directly to the flames,
these plastic impressions may survive. The plastic window attacked with
a butane cigarette lighter in Figure 14-12 bore both plastic impressions
in the softened plastic and patent impressions in the soot.
Latent fingerprints, those requiring some sort of physical or
chemical treatment to make them visible, consist basically of five components:
water, skin oils, proteins, salts, and contaminants. The water evaporates
rapidly or soaks into absorbent surfaces and is usually of little value.
Skin oils are the residues normally detected by dusting with a soft brush
and powdered carbon. These oils will remain on hard surfaces for some time
or will soak into absorbent materials, like paper. They can be evaporated
by elevated temperatures and such "dried out" prints may not respond
to dusting. The soot produced by a smoky fire (such as a gasoline pool)
may condense on a glass or metal surface to protect the latent prints, however,
needing only to be gently brushed or washed in a stream of tap water to
reveal the impressions. Proteins degrade to amino acids that are detected
by reaction with the chemical ninhydrin. They are easiest to detect when
the latent impressions are on clean, light-colored, porous surfaces like
paper or cardboard. They can be denatured by high temperatures so that they
no longer react, but if the paper has not been charred by the fire it may
be worth testing. However, exposure to water (from suppression, condensation
of water vapor from the fire, or from wet weather) will dissolve the proteins
and amino acids and blur the fingerprints. When this occurs the fatty deposits
in the latents (skin oils and cosmetics) are not affected by the water.
They can be detected by the physical developer method. Nonporous surfaces
that are or have been wet can be examined using small particle reagent (SPR).
The salts are the most fire resistant of all latent fingerprint residues,
resisting very high temperatures. Like amino acids, however, they are susceptible
to water damage. Salts may react with the inks or coatings on paper or with
the metal surfaces of cans. In doing so, they may leave an identifiable
patent (visible) print in corrosion. This effect may be enhanced by the
heat and moisture of the fire, leaving a visible print requiring no further
treatment. Other contaminants that can produce patent prints include cosmetics,
food residues, grease, motor oil, blood, and other products.
In the last 15 years, there have been dozens of techniques developed
for the detection of latent prints that may aid the inspection of fire-related
exhibits. Examination using lasers and high-intensity tunable light sources
(including ultraviolet and infrared wavelengths) makes it possible to penetrate
contaminants and make prints visible against a variety of backgrounds.
Cyanoacrylate ester fuming and a variety of fluorescent dye
stains and powders further expand the possibilities of recovery of latent
All these developments permit the development of latent finger
prints on many surfaces thought previously to be incapable of retaining
such impressions. These include surfaces such as leather, wood, vinyl plastics,
smooth fabrics, even charred paper and metal. Latent prints have even been
recovered from human skin. Recovering such prints requires appropriate skin
conditions (smooth, dry, free of hair) and the contact has to be of adequate
duration. Such prints lose detail after a few hours even if the victim is
deceased. If a fire victim has not been dead more than a few hours and the
skin has not been damaged by the fire, a search for such latents should
There is no more conclusive proof of
identity than a fingerprint, making it worth the time and effort to search
for it. Because of the rapidity with which these new techniques are being
put into service, the investigator should never conclude that latents are
impossible to recover before consulting with a knowledgeable specialist
in the field.
Bloodstains left at a crime scene have much greater evidentiary value
now than ever before because of great strides made in determining subtle
variations in the proteins and enzymes of human blood. Not long ago typing
was limited to ABO, MN, and Rh types. They can be detected by the physical
developer method. Nonporous surfaces that are or have been wet can be examined
using small particle reagent (SPR). The salts are the most fire resistant
of all latent fingerprint residues, resisting very high temperrred to exclude
witnesses or bystanders as sources of bloodstains at a scene or on a suspect's
clothes or vehicle. All 16 commonly used systems can be done on a fresh,
dried stain the size of a dime, making it possible to get conclusive results
even from small stains. These variants are, however, susceptible to deterioration
by age, bacteria, or heat. If the temperature of a bloodstain does not exceed
50°C (120°F), almost all typings can be carried out on a fresh stain.
Between 50° and 100°C (212°F), many off the enzymes degrade
and can no longer be identified. Between 100° and 200°C (425°F),
the proteins degrade and the species and ABO blood group can no longer be
determined. Above 200°C, the stain degrades and probably cannot be even
identified as blood. If the pros and probably cannot be even identified
as blood. If the pros and probably cannot be even identified as blood. If
the proteins in a blood or semen stain (or tissue fragment) have not degraded,
and the evidence is critical enough, the stain can be subjected to DNA typing.
This new approach is only recently tested in criminal cases and not all
forensic labs are equipped or prepared to take it on. DNA typable material
is more resistant to thermal and biological degradation than are the enzymes
and protein factors. DNA does not result in a unique identification but
it does promise a means of associating a stain or tissue fragment with an
individual with considerable certainty. If you are in doubt about a suspected
stain, collect it, keep it dry and as cool as possible, and submit it to
your laboratory as quickly as possible.
Often overlooked, impressions of tools on windows or doors can confirm
the fact of forced entry (Figures 14-13 and 14-14), identify points of entry
(or attempted entry), and, if tools are recovered, permit identification
of the tools responsible. Such tool marks may also be found on desk drawers,
filing cabinets, locks, or chains at a scene, confirming that a burglary
took place before the fire. The investigator should learn to recognize signs
of forcible entry and be able to account for all such damage at a fire scene.
The best kind of tool-mark evidence is the impression itself. After it has
been sketched and photographed, the object (or suitable portion) bearing
the impression should be collected. If its size makes recovery impractical,
a replica may be cast in silicone rubber or dental impression material.
Photos of a striated impression are of no use in a laboratory comparison
other than to establish the location and orientation of the mark. The submitted
tools are tested on a variety of surfaces to duplicate the manner of use
and the qualities of mark-bearing material as closely as possible without
risking damage to the tool. The evidence and test marks are examined under
40 to 400X magnification to try to match the striations or contours present.
See Figure 14-15 for a typical comparison.
Impressions of footwear and vehicle tires may be found on the periphery
of the fire scene. Shoe prints have also been linked to forcible entry,
as shown in Figure 14-16. Although prints in soil cannot generally be collected
themselves, they are worth photographing in place and possibly casting with
plaster of Paris. Photographs must be taken using low-angle (oblique) lighting
to highlight the three-dimensional details; a scale or ruler must be included
in the photo, near (not on!) and parallel to the impression; and the film
of the camera must be parallel to the ground to prevent distortion. See
Figure 14-17 for a good example. Although particular tires can sometimes
be matched back to their impressions, it is more common to establish only
a correspondence of size and tread pattern. Footwear, due to its characteristic
details and slower rate of change, is more amenable to identification and
offers a means of placing the subject right at the fire.
One of the few conclusive identifications possible with typical evidence
is the physical comparison of torn, cut, or broken edges or surfaces with
each other to establish a "jigsaw" fit between them. Pieces of
glass from shoes or garments have been matched back to windows, vehicle
lamps, or bottles to connect a victim or suspect with a scene. Wrappings
from incendiary or explosive devices have been matched to source materials
in the possession of the suspect, as have pieces of tape, wire, wood, and
rope. The evidentiary value of such positive identifications is very high
and the investigator must be aware of the possibilities when searching a
scene for anything that looks out of place or when searching a suspect's
property or vehicle. Close cooperation and communication between the lab
and the investigator has produced some excellent evidence (see Figure 14-18
for an example).
Trace or transfer evidence such as paint, fibers, soil, or glass can
be used to link a suspect with a crime scene even in arson cases. Paint
can be transferred to the clothing or tools when a forcible entry is attempted.
Flakes as small as 1 millimeter square are adequte for color or layer sequence
comparisions or complete analysis by infrared, spectrography, or x-ray analysis.
Such properties are generally class characteristics; that is, even a complete
correspondence between the chemical properties of known and questioned paints
does not imply that the two paints must have come from the same object.
Instead, it confirms that they came from the same class of objects painted
with the same paint. Multilayered paint chips, however, may be linked to
a particular source if they have enough layers to demonstrate a unique origin.
Glass is found in nearly all structures and can tell a good deal about
the events of the crime. It can be broken by either mechanical force or
thermally induced stress. Curved fracture ridges on the edges of glass fragments
reveal that the breaking force was mechanical. The shape of these conchoidal
fracture lines can help establish from which side of the window the force
came. The absence of these lines usually means the fracture was caused by
thermal stress. The presence or absence of soot, char, or fire debris on
the broken edges can establish whether breakage occurred before or after
the fire. When a glass window is broken, minute chips of glass are scattered
as much as 3 meters (10 feet) away, especially back in the direction of
the breaking force. If a subject is standing nearby, minute glass chips
may be found in his or her hair, clothes, hat, pockets, or cuffs. Even very
small (less than 1 millimeter) fragments can be compared to a suspected
origin, such as a bottle or window. Once again, this usually results in
a class characteristics comparison. Glass is made in large batches that
vary little in their physical or chemical properties. But such transfer
evidence does not require direct or prolonged contact with the source to
make it valuable to the investigator.
Fibers can be transferred from the furnishings of the fire scene to the
subject or from the subject's clothes to the scene, particularly at the
point of entry. Such two-way transfers can be very useful in linking suspect
and scene, even though fibers are usually another type of class characteristic
evidence. Fibers are compared by microscopic, chemical, elemental, or spectrophotometric
tests. Even the most sophisticated analyses must deal with the fact that
synthetic fibers are produced in such large quantities that they are used
mainly as corroborative or circumstantial evidence.
In a similar way, soils from shoes or clothing can be compared to soil
from a fire scene. This may be especially helpful in a wildlands fire where
a distinctive or unusual type of natural soil is found in the vicinity of
the fire's origin or along an access road. In urban environments, artificial
contaminants such as slag, cinders, metal filings, or paint chips may contribute
to the uniqueness of a soil. Contact with the soil near a workshop or industrial
site may result in the transfers of soils that are much more unique than
the natural soils nearby. Normally, soil comparisons are of marginal evidential
value, but in instances such as those offered, individual soils, especially
artificial ones, do occur and should not be overlooked.
SUMMARY: THE EXPERT WITNESS
The forensic scientists or criminalists who perform the examinations
discussed in this chapter are a breed apart from scientists in research
or industrial laboratories. Although they may share the same science or
engineering backgrounds as the others, forensic scientists relate their
work directly to the real world and its criminal activities. They are uniquely
capable of applying scientific tests to evidence, interpreting the results,
and reconstructing a sequence of events that led to the production of that
evidence. They cannot work in a vacuum, however. To do their job properly,
they must interact with the investigators, exchanging information and testing
ideas and hypotheses. Investigators should encourage this kind of communication;
only in this way can we obtain better evidence through fuller use of all
that the scientific expert witness has to offer.
Forensic Science and Engineering Committee. "Guidelines for Laboratories
Performing Chemical and Instrumental Analysis of Fire Debris Samples."
Fire and Arson Investigator, 38, June 1988.
"E 1492, Practice for Receiving, Documenting, Storing, and Retrieving
Evidence in a Forensic Science Laboratory." ASTM, Philadelphia, PA,
"E 1387, Standard Test Method for Ignitable Liquid Residues i Extracts
from Samples of Fire Debris by Gas Chromatography." ASTM, Philadelphia,
J. D. and Lomas, S. C. "The Examination of Fire-Damaged Electrical
Switches." Science and Justice, 35, 2, 1995, pp. 113116.
J. D. and Christie, C. C. "The Forensic Examination of Fuses."
Science and Justice, 35, 2, pp. 141149.
R. N. "Surface Analysis of Electrical Arc Residues in Fire Investigation."
Journal of Forensic Sciences, 34, May 1989.
B. "Examination of Arc Beads." Fire and Arson Investigator,
M. J. "The Perforation of a Copper Pipe During a Fire." The/Le
Journal, Canadian Association of Fire Investigators, June 1991, pp.
O. M., Jr. "Flammable and Combustible Liquids." Fire Protection
Handbook, 16th ed., Section 3, Chapter 5, NFPA, Quincy, MA, 1991, pp.
"D 3278, Test Method for Flash Point of Liquids by Setaflash Closed
Tester." ASTM, Philadelphia, PA, 1987.
N. R. "Interior Finish." Fire Protection Handbook, 16th
ed., Section 6, Chapter 4, NFPA, Quincy, MA, pp. 6-376-47.
J. D. "Laboratory Aspects of Arson: Accelerants, Devices, and Targets."
Fire and Arson Investigator, 29, January-March 1979, 39-46.
A. T., and Wittkower, R. S. "Identification of Accelerants in Fire
Residues by Capillary Column Gas Chromatography." Journal of Forensic
Science, 23, October 1978.
D. "A Unified Scheme for the Analysis of Light Petroleum Products Used
as Fire Accelerants." Forensic Science, 10, 243252,
W. Gas Chromatography with Glass Capillary Columns. Academic Press,
New York, 1980.
H. A. "Lead Alkyl Components as Discriminating Factors in the Comparison
of Gasolines." Journal of the forensic Science Society, 20,
J. "An Accelerant Classification Scheme Based on Analysis by Gas Chromatography/Mass
Spectrometry." Journal of Forensic Sciences, 35, 5, September
"E 1618, Standard Guide for Ignitable Liquid Residues in Extracts from
Fire Debris Samples by Gas Chromatography-Mass Spectrometry." ASTM,
Philadelphia, PA, 1994.
J. D., and Home, J. M. "A Novel Method for the Selective Adsorption
of Hydrocarbons from the Headspace of Arson Residues." Nature,
J. "A Method of Absorption of Flammable Vapors by Direct Insertion
of Activated Charcoal into Debris." Bridgeport Crime Lab, Bridgeport
MI. Reprinted in Arson Accelerant Detection Course Manual. U.S. Treasury
Department, Rockville, Maryland. 1980.
J. D., Home, J. M., Smalldon, K. W. "A Comparison of Relative Sensitivities
of the Adsorption Wire and Other Methods for the Detection of Accelerant
Residues in Fire Debris." Journal of the Forensic Science Society,
22, April 1982.
J. E., and Holmes, R. N. "Collection and Determination of Accelerant
Vapors from Arson Debris." Arson Analysis Newsletter, 5,
J. Journal of Criminal Law, Criminology and Police Science, 46,
S., and DeHaan, J. D. "An Evaluation of Some Arson Distillation Techniques."
Arson Analysis Newsletter, 2, September 1978.
J. D. and Bonarius, K. "Pyrolysis Products of Structure Fires."
Journal of the Forensic Science Society, 28, 1989.
P. L. "Oxygenate Fingerprint of Gasolines." DET Report No.
18, October 1990, pp. 812.
S. E. et al. "Evaluation of the GC-FTIR for the Analysis of Accelerants
in the Presence of Background Matrix Materials." MAFS Newsletter,
20, 1, pp. 4876.
J. et al. "Fluorescence of Petroleum Products II: Three-Dimensional
Fluorescence Plots of Gasolines." Journal of Forensic Sciences,
32, January 1987.
D. C. "Comparison of Automotive Gasolines Using Capillary Gas Chromatography
(I and II)." Journal of Forensic Science, 32, May 1987.
D. C. and Gresham, W. R. "Microbial Degradation of Gasoline in Soil."
Journal of Forensic Sciences, 35, 4, 1990.
J. et al. "Practical Experiences with Scanning Electron Microscopy
in a Forensic Science Laboratory." Linkoping University, NCJRS-Microfiche
Program, Report Number 2, 1979.
J. "Analysis of Polycyclic Aromatic Hydrocarbons in Soils and Its Application
to Forensic Science." Linkoping University, NCJRS-Michofiche Program,
Report Number 4, 1979.
Bureau of Standards. Gaithersburg, MD, personal communication, 1986.
M. T., Lennard, C. J., Margot, P., Dustin, I., and Furrer, P. "Soot
as an Indicator in Fire Investigations: Physical and Chemical Analysis."
Journal of Forensic Sciences, 39, 4, July 1994.
(1990) op. cit.
(1979) op. cit.
H. Modern Pyrotechnics. Chemical Publishing Co., New York, 1961.
W. L. "Examination of Fire Debris for Flare (Fusee) Residues by Energy
Dispersive X-Ray Spectrometry." Arson Analysis Newsletter, 7,
K. P. and Kobus, H. J. "The Explosive Reaction Between Swimming Pool
Chlorine and Brake Fluid." Journal of Forensic Sciences, 36,
3, May 1991.
R. E. "A Systematic Approach to the Forensic Examination of Flash Powders."
Journal of Forensic Sciences, 23, October 1978.
J. "Identification of Burnt Matches by Scanning Electron Microscopy."
Journal of Forensic Sciences, 23, October 1978.
Army. Unconventional Warfare Devices and Techniques. TM 3-201-1,
P. and Lennard, C. J. Techniques for Latent Print Development. University
of Lausanne, Lausanne, 1991.
H. C. and Gaensslen, R. E. "Cyanoacrylate 'Super Glue' Fuming for Latent
Fingerprints." The Identification Officer, .Spring 1985.
D. "Some Technical Aspects of the Demonstration and Visualization of
Fingerprints on Human Skin." Journal of Forensic Science, 14,
E. R., and Duff, J. M. "Laser Detection of Latent Fingerprints-Treatment
with Flourescers." Journal of Forensic Science, 24, 96-100,
Bertsch, W., Zhang, Q. W., and Holzer, G. "Using the Tools of Chromatography,
Mass Spectrometry and Automated Data Processing in the Detection of Arson.
Journal of High Resolution Chromatography, 13, September 1990.
Caddy, B., Smith, F. P., and Macy, J. "Methods of Fire Debris Preparation
for Detection of Accelerant." Forensic Science Review, 3,
1, June 1991.
Conkling, J. The Chemistry of Pyrotechnics. Marcel Dekker, New
Coulombe, R. "Chemical Markers in Weathered Gasoline." Journal
of Forensic Sciences, 40, 5, September 1995.
Fultz, M. L. and DeHaan, J. D. "Gas Chromatography in Arson and
Explosives Analysis." Gas Chromatography in Forensic Science,
I. Tebbett, ed., Ell Handbook, 16th ed., Section 3, Chapter 5, NFPA, Quincy,
MA, 1991, pp. 3-433-53.
.Holzer, G. and Bertsch, W. "Recent Advances Toward the Detection
of Accelerants in Arson Cases." American Laboratory, December
Lennard, C. J., Rochaix, V. T. and Margot, P. "A GC-MS Database
of Target Compound Chromatograms for the Identification of Arson Acccelerants."
Science and Justice, 35, 1, 1995.
Lennard, C. J. "Fire (Determination of Cause)--A Review: 199295."
Presented at the 11th INTERPOL Forensic Science Symposium, Lyon, France,
Metson, J. B. and Hobbis, C. M. "The Use of Auger Electron Spectroscopy
in Fire Investigations." Chemistry in New Zealand, July 1994.
Phelps, J. L., Chasteen, C. E., and Render, M. M. "Extraction and
Analysis of Low Molecular Weight Alcohols and Acetone from Fire Debris Using
Passive Headspace Concentration," Journal of Forensic Sciences.
39, 1, January 1994.
Sheff, L. M. and Siegel, J. A. "Fluorescence of Petroleum Products
V: Three-Dimensional Fluorescence Spectroscopy and Capillary Gas Chromatography
of Neat and Evaporated Gasoline Samples." Journal of Forensic Sciences,
39, 5, September 1994.
*Many such experts may be described as forensic scientists--forensic
chemists, forensic engineers, and the like. The term forensic means
having to do with a court of law, so such terms refer to the application
of a scientific discipline to problems having to do with the law.
Kirk's Fire Investigation, 4/E by John D. DeHaan, ©1997.
Reproduced by permission of Prentice-Hall, Inc., Upper Saddle River, NJ.
Permission from Prentice-Hall is required for all