311% {IIIHIWIIWlHllUHtillHNliHM!“WIHIHHWI

   

”Lace: 3 w

&-.£:>7

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
thesis entitled

THE PERSISTENCE OF GUNSHOT RESIDUE ON
DECOMPOSING TISSUE

presented by

LUTHER STERLING SCHAEFFER

has been accepted towards fulfillment
of the requirements for the

MS degree in CRIMINAL JUSTICE

 

 

MLAJMQC é/
'T/lajor PMnature

4151/ 23/04 j

Date

MSU is an affirmative-action, equal-opportunity employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/ClRC/DaleDue.indd-p.1

 

The Persistence of Gunshot Residue on Decomposing Tissue
By

Luther Sterling Schaeffer

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Criminal Justice

2007

ABSTRACT
The Persistence of Gunshot Residue on Decomposing Tissue
By

Luther Sterling Schaeffer

The research presented herein describes the development and
optimization of scanning electron microscopy / energy-dispersive spectroscopy
(SEM/EDS) and inductively-coupled plasma mass spectrometry (ICPMS) for the
analysis of gunshot residue (GSR) on decomposing tissue. Two euthanized pigs
were obtained from the MSU Swine Facility. One pig was shot 10 times with a
Glock 9mm pistol and the other served as the source of the control tissue
samples. Each day, a gunshot wound and a control tissue sample was excised
from each pig over a 7-day sampling period. Three methods of sample
preparation were attempted prior to SEM/EDS analysis; air-drying, freeze-drying,
and tape-lifting. GSR consisting of Sb, Ba, and Pb was found by SEM and
analyzed by EDS on the tape-lifts from days 1 and 2. ICPMS was used to
analyze gunshot wounds and control tissue for 121Sb, 13883, and 208Pb after
preparing the samples by microwave digestion. GSR was determined in the
gunshot wounds throughout the course of the 7-day sampling period in
concentrations ranging from 9.19 pg of 121Sb per gram of tissue (pg/g) to 0.15

ug/g, 138Ba was determined in the range of 30.67 — 0.06 pig/g, and 208Pb was

determined in he range of 26.71 — 0.10 uglg. ICPMS had a linear response
between 0 - 1000 ppb for 121Sb, 13883, 208Pb and the limits of detection were

0.002, 0.033, and 0.009 ppb, respectively.

ACKNOWLEDGEMENTS

First and foremost, I owe a great bit of gratitude to everyone who had contributed
to this project.

Dr. Ruth Waddell, thanks for reading through this time and time again, you made
it through all the tens of us racing theses for corrections. Thanks for all the
guidance and suggestions on the project ranging from the research to the AAFS
presentation and finally the written work.

Dave Szymanski (soon to be Dr.), thanks for all the help, guidance, and work for
the ICPMS portion of this study, without your help this study would have been
lacking.

Dr. Stanley Flegler, Carol Flegler, and everyone at the Center for Advanced
Microscopy, you all have been invaluable resources throughout this work. Thank

you for all your patience and passing on some of your knowledge to me.

Thanks also to Michael Blaskovitz, Dr. Brian Hunter, Dr. Rich Merritt, Dr. Norm

Sauer, Al Snedegar, Mr. Firearms Specialist of Local PD.

Thanks to my family for once again enduring my frustrated diatribes. And finally
thanks to all of my climbing friends for providing me with an essential outlet,

support, and the opportunity to forget about my work.

Anyone who I may have left out, just because your name is not here does not

mean you were forgotten, thank you.

iii

TABLE OF CONTENTS

List of Tables ........................................................................................................... vi
List of Figures ........................................................................................................ vii
Chapter 1: Introduction ........................................................................................... 1
1.1 Introduction .......................................................................................................... 1
1.2 Anatomy of a Cartridge ........................................................................................ 2
1.3 Gunshot Residue ................................................................................................. 4
1.4 Methods of Elemental GSR Analysis ................................................................... 7
1.5 Research Objectives ............................................................................................ 8
Chapter 2: Instrumental Theory ............................................................................ 10
2.1 Scanning Electron Microscopy / Energy Dispersive Spectroscopy (SEM/EDS). 10
2.2 Analysis of GSR Using SEM/EDS ...................................................................... 16
2.3 Inductively-Coupled Plasma Mass Spectrometry (ICPMS) Theory .................... 17
2.4 ICPMS for GSR Detection .................................................................................. 21
Chapter 3: Materials and Methods ........................................................................ 23
3.1 Experimental Design .......................................................................................... 23
3.2 SEM/EDS Sample Preparation and Instrumental Parameters ........................... 24
3.3 ICPMS Sample Preparation ............................................................................... 26
3.4 ICPMS Operating Parameters ............................................................................ 28
Chapter 4: Results and Discussion ...................................................................... 31
4.1 Introduction ........................................................................................................ 31
4.2 Decomposition Observations During Sampling Period ....................................... 31
4.3 Analysis of tissue for GSR by SEM/EDS ............................................................ 36

4.3.1 Optimizing sample preparation for SEM/EDS analysis ......................... 36

4.3.2 SEM/EDS Analysis of Tape-lifts ........................................................... 43
4.4 Summary of SEM/EDS Results .......................................................................... 51
4.5 Analysis of Tissue for GSR by ICPMS ............................................................... 52

4.5.1 ICPMS Calibration ................................................................................ 52

4.5.2 ICPMS Precision .................................................................................. 54

4.5.3 ICPMS Analysis of Control Tissue Digests ........................................... 54

4.5.4 ICPMS Analysis of Gunshot Wound Tissue ......................................... 59

4.5.5 Concentrations of the Analytes in Tissue Digests ................................ 61

4.5.6 ICPMS Analysis of Procedural Blanks .................................................. 62

iv

4.6 Summary of ICPMS Results ............................................................................... 63

Chapter 5: Conclusions and Further Work .......................................................... 65
5.1 Conclusions ........................................................................................................ 65
5.2 Further Work ...................................................................................................... 68
References .............................................................................................................. 70

LIST OF TABLES

Chapter 3: Materials and Methods

Table 3.1 ICPMS instrumental parameters .............................................................. 28

Chapter 4: Results and Discussion

Table 4.1 The weather, dates, and temperature for the course of the seven day
collection period ....................................................................................................... 33

Table 4.2 Average weight percent of Sb, Ba, and Pb in GSR particles collected
from gunshot wound tissue sample of day 1. ........................................................... 46

Table 4.3 ICPMS data for 10 individually prepared gunshot wound dilutions from
day 3. ....................................................................................................................... 54

Table 4.4 ICPMS data for the gunshot wounds and control tissue collected
during the course of the 7-day sampling period. ...................................................... 62

vi

LIST OF FIGURES

Chapter 1: Introduction

Figure 1.1 A cut-away diagram of a typical firearm ammunition cartridge
revealing its basic components. ................................................................................. 3

Figure 1.2 A handgun firing a bullet. The white cloud trailing the bullet is the
plume of GSR ............................................................................................................. 5

Chapter 2: Instrumental Theory

Figure 2.1 Block diagram of a scanning electron microscope. PMT —
photomultiplier tube, CRT — cathode ray tube. Illustration was adapted from Dr.
Stanley F legler’s text, Scanning and Transmission Electron Microscopy ................. 11

Figure 2.2 A schematic diagram of an ICP torch and its orientation with regards
to the sampling cone of the MS ................................................................................ 18

Chapter 3: Materials and Methods

Figure 3.1 An illustration of the spacing and arrangement of the gunshot wounds
on the pig. The illustration is not drawn to scale ...................................................... 24

Chapter 4: Results and Discussion

Figure 4.1 A picture of the resultant wounds and spacing on the pig. ..................... 32
Figure 4.2 Pictures collected on day 1 (a.) and 8 (b.) illustrating the
decomposition of the shot pig ................................................................................... 35

Figure 4.3 SEM images of freeze-dried (top) and air-dried (bottom) wound tissue
prepared for analysis ................................................................................................ 38

Figure 4.4 SEM secondary-electron image of a GSR particle from day 1 on an
AD wound tissue sample .......................................................................................... 40

Figure 4.5 Images taken by SEM. A secondary electron image of carbon-coated
carbon tape (top) and a back-scattered electron image of a tape-lift from a
gunshot wound collected on day one (bottom) ......................................................... 42

vii

Figure 4.6 BSE (left) and SE (right) images for a single GSR particle on the tape-
lift from a gunshot wound of day 1. The EDS spectrum below the images shows
the elemental composition of the particle ................................................................. 44

Figure 4.7 A SE electron image of a GSR particle from the tape lift of the
gunshot wound from day 1 (above) and three elemental maps acquired by EDS
detection showing the relative intensity of Sb, Ba, and Pb (L to R). ......................... 45

Figure 4.8 An x-ray spectrum for GSR located on the tape-lift from the gunshot
wound sample collected on day 2. Of note is the aluminum peak in the
spectrum. ................................................................................................................. 47

Figure 4.9 An example SEM image of the Fe-Cr-Ni metal particle discovered on
a tape lift from day 5. The x-ray spectrum below shows the highest elemental
abundance due to Fe ............................................................................................... 49

Figure 4.10 Calibration curves for the target analytes, 121Sb, 138Ba, and 2”Pb. ...... 53

Figure 4.11 ICPMS spectra of the digests of a gunshot wound and control tissue
from day 3. The spectra show the m/z range of 120 to 140 amu ............................ 56

Figure 4.12 ICPMS spectra of the digests of a gunshot wound and control tissue
from day 3. The spectra show the m/z range of 200 to 220 amu ............................ 57

Figure 4.13 Plot of concentration of 121Sb, 138Ba, and 2°8Pb versus the day of
collection. ................................................................................................................. 58

Figure 4.14 Plots of GSR concentration versus the day of collection. The upper

plot is the full-scale version showing all of the data points and below is a
magnified scale to show the values of GSR on later days of collection .................... 60

viii

Chapter 1

Introduction

1.1 Introduction

In 2005, the United States Department of Justice Bureau of Justice
Statistics, reported that index crimes (rape, murder, robbery, aggravated assault,
and property crimes) involving a firearm comprised about 68% of all murders,
42% of all robberies, and 21% of all aggravated assaults that were reported to
the police nationally.1 Approximately 30,000 deaths annually are a result of guns,
and greater than 10,000 of those deaths are ruled homicide.2 Of the 4.7 million
violent crimes consisting of rape and sexual assault, robbery, and aggravated
and simple assault that were reported in 2005, a total of 9% involved a firearm.

From the statistics given above, it is obvious that firearms related incidents
are common occurrences in the US. Therefore it is of utmost importance for
trauma specialists, medical examiners, and pathologists to be able to identify a
gunshot wound when they are presented. Depending on the distance from which
the gunshot wound was inflicted, the caliber of the firearm, and the state of bodily
decomposition, the identification of a wound may not be a simple task due to
reasons such as the lack of a discernable entrance wound, or in the case of
decomposition, blackening of the skin. A study conducted by Collins et al. in
1994 revealed the difficulty trauma specialists face when attempting to
distinguish the entrance from exit wound on mortally wounded bodies. The study

reported that of 46 victims presenting gunshot wounds, 52% were misinterpreted

in terms of differentiating exit from entrance holes and identification of a gunshot
wound.3

Physical examination of gunshot wounds can consist of many
observations recorded by the consulted practitioner; the wound itself, fracturing
of underlying bone, as well as gunshot residue (GSR) around the wound.4 In a
badly decomposed body, cause of death by gunshot wound may be difficult to
discern if the wound itself is not visible and may be misidentified. In the absence
of any other pertinent evidence, this can make cause of death determinations
difficult for the forensic pathologist. In decomposed bodies where it is impossible
to visually identify gunshot wounds, detection of GSR may allow for
determination of the cause of death. Additionally, the examination of wounds for
GSR may be of assistance to determine entrance from exit holes as well as the

distance between firearm and target.5

1.2 Anatomy of a Cartridge

To understand GSR and its relevance to a forensic investigation it is
imperative to first develop a general familiarity with firearms and ammunition.
The lethality of firearms is derived from the ammunition cartridge. The cartridge
unit is made up of four primary parts; the bullet, cartridge casing, propellant, and
the primer, as seen in Figure 1.1. The bullet is securely housed in the cartridge
casing, which is commonly made of brass and filled with a propellant. The
propellant is a low explosive, either black powder or as is more commonly

encountered, smokeless powder. Black powder is a mixture of potassium nitrate,

charcoal and sulfur in an approximate 15:3:2 ratio. Smokeless powder is made
from nitrocellulose (single-base) or mixtures of nitrocellulose and nitroglycerine
(double—base) or the previous two in addition to nitroguanidine (triple-base).
Smokeless powders have a greater energy density than black powder and the
combustion products are mainly gaseous, whereas black powder combustion
produces approximately 33% solid products.6 Bullets come in many different
shapes, calibers, and are made from a variety of materials. Bullets are most
often made from lead and may be covered in a layer of copper plating called a
jacket. The jacket can partially cover the bullet, called a semi-jacketed bullet, or

fully encapsulate it, termed a full metal-jacketed bullet.

Bullet _

Casing

  
 

Propellant

Primer

Figure 1.1. A cut-away diagram of a typical firearm ammunition cartridge

revealing its basic components.

The primer is a small, compact explosive charge, which is packed into a

primer cup at the base of the cartridge casing. The primer consists of 4

components; an initiating explosive, oxidizing agent, fuel, and sensitizer. The
initiator is typically lead styphnate, the oxidizing agent is typically barium nitrate
however barium peroxide, lead nitrate, or lead peroxide may also be used. The
fuel is most commonly antimony sulfide, and the sensitizing agent often is
tetracene.7 When the primer is struck by the firing pin, the primer detonates
which then ignites the propellant. The ignited propellant deflagrates into a rapidly
expanding gas, which accelerates the bullet out of the cartridge casing and down
the barrel of the gun. Spiraled grooves in the barrel of the firearm, termed rifling,
cause the bullet to spin as it is propelled from the firearm. Rifling provides the
bullet gyroscopic stability resulting in a straighter trajectory and maximum

velocity on leaving the barrel.

1.3 Gunshot Residue (GSR)

Gunshot residue consists of a gaseous mixture of combustion products
from the primer and propellant, as well as particulates from the bullet, cartridge
casing, and the barrel of the firearm. The gaseous GSR is expelled from the
barrel of the firearm along with the bullet, as shown in Figure 1.2.7'9 GSR
deposits on nearby objects such as the target as well as the shooters hands and

clothes.

 

 

Figure 1.2. A handgun firing a bullet. The white cloud trailing the bullet is the

plume of GSR.1°

Expulsion of GSR causes a visible blackened circle of GSR to deposit on
objects close to the end of the firearm barrel. At greater distances, the ring
becomes more difficult to distinguish and larger in diameter as it becomes more
diffuse.11 A characteristic of GSR deposition, called stippling, is a pattern
resulting from unburnt and partially burnt gunpowder embedding in the target.
The mass of the gunpowder particles is greater than that of the smoke
particulates and thus will travel further to distances of approximately 3 through 5
feet. GSR can be observed by the naked eye from firearms discharged at
distances less than 12 inches from their respective target. At distances greater
than 12 inches, it becomes much more difficult to discern GSR without the aid of
staining or a microscope.12

GSR, as mentioned above, consists of trace materials derived from the
primer, propellant, cartridge casing and the bullet itself. It has also been found to

contain distinguishable trace elements from the cast material of the barrel of the

firearm.13 The components of GSR which are considered characteristic for
identification are derived from the primer of the cartridges, which contains
compounds of barium (Ba), lead (Pb), and antimony (Sb). These materials
combust and form the basis for the characteristic elemental profile of GSR;
elemental Pb and Sb and compounds of Ba.7 GSR also contains organic
components from the propellant and primer; however, these are usually analyzed
as supportive evidence when determined along with Ba, Pb, and Sb and will not
be discussed any further in this text.7

Another characteristic beyond the elemental profile used for concluding
the presence of GSR is the morphology of the particulates. GSR coalesces while
in the gaseous state to form small spherical droplets in the range of 0.1 (M to 10
(M. Although larger sizes are occasionally seen, GSR is most commonly found
in the latter size range.8' 9 The particle morphology in conjunction with the
elemental composition provides a distinct profile for GSR versus various other
environmental or occupational contaminants such as automobile brake particles
and exhaust particles as well as lead aerosols.” ‘5 Torre et al. showed that
although analysis of individual particles of automobile brakes revealed an
elemental profile consistent with GSR particles, the individual particles could be
discriminated because they lacked the spherical morphology characteristic of
GSR.15 The studies by Torre and Wolmer illustrate the power of scanning
electron microscopy/energy-dispersive spectroscopy (SEM/EDS) to analyze
individual GSR particles and discriminate non-GSR particles of similar elemental

profiles based on morphology.15

1.4 Methods of Elemental GSR Analysis

Many analytical methods and instrumental techniques have been

described over the last 30 years for GSR detection7' 8' ‘6

including: neutron
activation analysis (NAA), capillary electrophoresis (CE), flameless atomic
absorption spectroscopy (FAAS), SEM/EDS, and recently, inductively-coupled
plasma mass spectrometry (ICPMS).5'17°24 The methods will be briefly discussed
below, however, the reader is referred to one of several recent review articles in
publication for additional details.7' 8' ‘6

Ruch et al. first documented analysis of GSR by NAA in 1964.25 NAA
provided a sensitive means to analyze GSR from shooters hands and clothing for
Ba and Sb, but not Pb. The method was used for over a decade but suffered
from major drawbacks such as requiring access to a nuclear reactor and
specially trained individuals to perform the analysis.8 Application of FAAS for
GSR analysis followed in the early 19705 boasting the capability for sensitive
analysis of Sb, Ba, and Pb.7 Koons et al. reevaluated methods for GSR
extraction followed by detection by FAAS. GSR was collected by scrubbing
shooters hands with swabs moistened with 5% nitric acid and subsequent
analysis was performed after extracting the sample from the cotton.26 FAAS is a
method possessing adequate sensitivity, however it lacks simultaneous multi-
elemental analysis. Another method for simultaneous inorganic (Sb, Ba, and Pb)

and organic components of GSR by CE was introduced.17 The method utilized

cotton swabbing as described previously for sample collection. The CE method

was found to lack adequate sensitivity for real samples and would require sample
preconcentration or large sample injections.17

The most widely used and accepted method of GSR analysis is SEM/EDS
due to its ability to differentiate individual GSR particles by their morphology as
well as their elemental composition.16' 17' 27 Morphological and elemental analysis
allows for the discrimination of particles, which may have the same elemental
profile but clearly are not morphologically characteristic of GSR.15 The large
drawback of SEM/EDS analysis of GSR is it is a cumbersome method; it can
take several hours to perform elemental and morphological analysis.
Advancement of this method include programs that automate the procedure of
particle analysis and low vacuum systems which require very little sample
pretreatment.“ ”'30

A survey of crime labs conducted by Singer et al. revealed that bulk
analysis of elemental GSR is most often performed by FAAS.16 Recent
publications, however, describe another alternative, ICPMS, as a promising
method for trace elemental analysis.5' 3‘ Benefits of ICPMS are its low detection
limits, rapid-analysis time, isotopic analysis, wide linear dynamic range, and

multi-elemental analysis capabilities.

1.5 Research Objectives
Numerous papers have been published in which the detection of GSR has
been conducted from materials such as hand swabs, tape lifts, and clothing

materials, however there is a very limited assortment of papers which address

GSR determined from biological tissues. The primary objective of this research is
to evaluate the applicability of SEM/EDS and ICPMS for the detection of GSR
located directly on decomposing tissue since the use of either of these
techniques may be a useful tool for identifying gunshot wounds on bodies in
advanced stages of decay. The details of appropriate tissue sample collection
methods and storage procedures, as well as sample preparation and

instrumental analysis procedures will be discussed.

Chapter 2

Instrumental Theory
2.1 Scanning Electron Microscopy / Energy Dispersive Spectroscopy

(SEM/EDS)

The credit for building the first functional scanning electron beam
microscope is given to Manfred von Ardenne.32 Vast improvements of the design
and functionality of the SEM followed over the course of the next 15 years by
researchers such as Vladimir Zworykin, James Hillier, Gerald Snyder, and Sir
Charles Oatley.32' 33 Cambridge Instruments introduced the first commercially
produced model in 1964. Since its commercial production, there have been
many more developments for electron microscopy, including x-ray microanalysis
for elemental identification.33

Compared to the light microscope, SEM provides an analyst not only with
a superior resolution and magnification, but also the ability to identify elemental
metal composition of samples when equipped with an x-ray analyzer. SEMs are
continually being applied to new applications in failure analysis and in pathology,
forensic, geological, metallurgical, and environmental labs.28 There are an
estimated 50,000 operating SEMs worldwide.” The SEM has the ability to
image samples down to 3 nanometers with magnifications up to 300,000x
making it very attractive to demanding detail oriented analyses like those

commonly encountered in forensic trace analysis.28 Forensic applications

10

include identification of diatoms, fiber and hair analysis, and trace material
analysis, such as GSR.

A basic SEM design is illustrated below in Figure 2.1. The SEM consists
of many components that need to be in perfect working order in order to generate
a high quality image. The first essential step of successful imaging via electron
microscopy is the production of an electron beam, which requires that the entire
system be under high-vacuum (in the order of <10‘9 atm). The electron beam is
generated by an electron gun located at the top of the SEM column in the
electron gun chamber. The electron gun is typically one of three types; tungsten-
hairpin filament, lanthanum-hexaboride crystal, or field-emission.33

Electron gun EU]

 

Condenser lens
[Scan generator

Scan coils \ /
\ é Computer
\ \
Objective lens~ \ \ . J
\
CRT

Final aperture -—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—_'l
Scintillator , 3:; PMT Preamplifier
Sample \ Light pipe

Faraday cage

 

Figure 2.1 Block diagram of a scanning electron microscope. PMT -

photomultiplier tube, CRT — cathode ray tube. Illustration was adapted from

F legler et. al.33

The tungsten filament emitter is the most cost-effective electron gun and
does not require an ultrahigh vacuum, however it has a relatively short lifetime
and it is not a strong emitter. The lanthanum-hexaboride (LaBe) electron gun
produces a much stronger electron beam under similar conditions, provides a
superior resolution, and has a lifetime ten times that of the tungsten filament.
The LaBs emitter however requires a higher vacuum and is much more
expensive to replace. The third choice is the field-emission emitter, which
produces an electron beam intensity up to 1000 times that of the tungsten
filament with a resolution better than both of the former choices and a lifetime
around 100 times that of the tungsten. Disadvantages of this emitter are its high
cost, the need for a specially designed SEM, and an even higher vacuum than
the LaBs emitter. More importantly, the field-emission emitter does not provide a
stable electron beam.”

The electron beam travels from the gun through an anode aperture,
limiting the beam size and eliminating stray electrons, and into focusing
electromagnetic lenses. The electromagnetic lenses serve as a means to
condense and focus the electron beam into the desired spot size. Focusing of
the beam is performed by changing the amount of current applied to the lenses
via a microscope control. Two types of lenses are used in the SEM; condenser
and objective lenses. The condenser lens forces the initial electron beam into a
smaller diameter. The objective lens then focuses the beam to a fine spot size.

Vlfithin the objective lens, there are scan coils that are connected to a scan

12

generator. The scan coils create a varying electromagnetic field to move the
focused electron beam back and forth over the sample in the form of scan lines.

When the electron beam strikes the sample, a complex series of
interactions occur. These interactions of the electron beam and the sample atom
electrons and nuclei result in the formation of variable energy electrons, x-rays,
heat, and light. The low energy electrons resulting from inelastic collisions are
ejected from the sample, these are termed secondary electrons (SE) and are
used for imaging of the sample. The detector for SEM images is called an
Everhart-Thornley detector. It consists of a Faraday cage with a positive bias of
+300 V in order to attract the SE. The collected SEs are then accelerated by a
+12,000 V bias applied to the scintillator. The scintillator is typically a phosphor
and AI coated disk that converts the SE into photons. Photons are transmitted
through the light pipe to outside of the vacuum of the SEM column into a
photomultiplier tube (PMT). The PMT amplifies the initial photon signal, which
then enters a preamplifier before finally terminating as an image on the cathode-
ray tube (CRT).

The elastic collisions of the electron beam with the sample forms another
type of higher energy electrons, termed back-scattered electrons (BSE). BSE
are high-energy electrons reflected backward from the same side they entered.
Their high energy is enough to overcome the positive bias applied to the Faraday
cage where they strike a BSE semiconductor detector. The detector is mounted
on the final aperture and produces an image that has much less resolution than

the SE image. Though the resolution of BSE images are generally poor, they do

13

provide important information. BSE images allow for one to differentiate
elements of differing atomic number. These images do not however provide the
actual atomic number, but merely relative approximations.33 Elements of higher
atomic number result in brighter contrast than lower atomic numbers.

X-rays are another result of the complex sample-beam interaction. The
detection of X-rays is an important feature of SEM because they allow for the
determination of the spatial distribution of elements within a sample. X-rays are
photons resulting from inelastic scattering events. When an electron strikes the
sample, it can result in the ejection of an electron from an inner orbital shell. An
electron from a higher energy shell then fills the inner orbital vacancy. The
energy difference between the two shells may be emitted in the form of an X-ray.
The energy and the wavelength of the produced X-ray is characteristic of the
element from which it originated.33 X-rays are named according to the shell from
which they originate (K, L, M, N) followed by a subscript denoting the number of
shells the electron which filled the vacancy has jumped (1 = a, 2 = 3).” Each
element typically produces several different energies of X-rays due to the
different, quantized transitions that can occur within the atom. Each type of
electron is called a line; the generation of enough X-rays of a particular line
results in a peak corresponding to the energy of the X-ray, which is characteristic
of the element that produced it.33

The X-ray detector attached to the SEM column can be one of two types,

either wavelength-dispersive (WDS) or energy-dispersive (EDS) spectroscopes.

EDS, the detector on the SEM used throughout this research, has several distinct

14

advantages over WDS. A full elemental analysis can be performed with EDS
simultaneously, whereas with WDS a maximum of 4 selected elements an be
analyzed at one time. The cost of the EDS system is also approximately one
quarter of the WDS system. EDS is also beneficial for biological samples due to
its greater sensitivity at low beam currents. WDS however has a greater
selectivity than EDS allowing for 10x the ability to differentiate elements with
similar spectra.33

The EDS system is made up of two main components: the detector and
the analyzer. The detector consists of a solid-state crystal and a thin beryllium
window that is protected by a collimator. The collimator is a housing column that
protects the detector from stray X-rays and BSE. The window isolates the
detector crystal from contaminants in the SEM column. The detector crystal is a
lithium-doped silicon wafer semiconductor which converts the incident X-rays into
a brief voltage pulse. The detector apparatus is kept cooled by liquid nitrogen to
maintain the semiconductor properties of the detector crystal intact, to prevent
the crystal from shorting out, and to eliminate electronic noise. The analyzer for
the EDS system amplifies the voltage pulses of the X-rays, converts the analog
signal into a digital form, and then sorts the digital pulses into channels so that a
computer can then display the resultant spectra. The spectra consist of a plot of

X-ray intensity (counts) versus X-ray energy.

15

2.2 Analysis of GSR Using SEM/EDS

Analysis of GSR by SEM/EDS is recognized as a selective and definitive
method of analysis. SEM/EDS allows an investigator a great deal of specificity to
image and determine morphological and elemental profiles of individual GSR
particles compared to bulk methods of analysis such as atomic absorption
spectroscopy and neutron activation analysis briefly described in Chapter 1.10' ‘6'
3“ The method has been recognized as a invaluable tool for GSR analysis since

1.35 Since its initial introduction, many

it was first demonstrated by Boehm in 197
researchers have described various methods of SEM/EDS analysis for GSR.
The basic instrumental technique remains consistent with minor changes in the
areas of sample preparation and sampling. A great deal of research focuses not
necessarily on optimizing detection conditions but on the sample collection
procedure.22' 27

The most commonly encountered GSR preparation method is via tape lift
where a double-sided adhesive is mounted on an aluminum stub and then used
to dab the sample.1°'22' 36 The aluminum stub is the platform on which specimen
analyzed by SEM are mounted. Using a tape lift has several distinct advantages
such as sample stability with respect to time since the tape does not degrade as
tissue does, simplicity of preparation, minimization of interferences such as
epidermal cells in biological tissues, and a homogenous background signal
contribution. The tape lift method is reported as the most effective sample

preparation approach compared with other lifting techniques such as the glue-lift

and concentrating techniques."22

16

Sample requirements for SEM/EDS analysis dictate that a sample must be
thoroughly dry and devoid of all volatile chemicals, mounted firmly on the sample
stub, and electrically conductive. The presence of volatiles or water in the sample
can contaminate the vacuum of the SEM and an improperly mounted specimen
can move or fall off the aluminum stub during insertion of the sample into the
vacuum tube. The GSR samples collected by tape lifting satisfy the first two
criteria, and to make them electrically conductive, the samples are coated with a
thin layer of carbon via a carbon coater. The thickness of the carbon layer is

typically in the range of 30 — 40 nm.22

2.3 Inductively-Coupled Plasma Mass Spectrometry (ICPMS) Theory

ICPMS was first commercially introduced in the late 1980s and has been
extensively reviewed by researchers for a diverse range of multielemental trace
analysis applications.” 38 ICPMS can be used for analysis of solid and solution-
phase samples, however, the focus of the writing contained herein will describe
its use and fundamental working principles from the solution-phase stand-point.
ICPMS remains a popular technique of elemental analysis because of its low
detection limits, selectivity, wide linear-dynamic range, good precision and
accuracy and its ability to perform rapid multielemental analyses.39' 4°

The ICPMS instrument is composed of two main components: the

inductively-coupled plasma (ICP) source and the mass spectrometer (MS).

Sample ions are generated in the ICP and then accelerated into the MS for

17

separation by mass to charge ratio and subsequent detection. A diagram of an
ICP torch can be seen in Figure 2.2.
Sampling cone

Ar Ar
1 Quartz tube RF coil Plasma

.I...
.l .

  

 

 

 

   

 

Sample + Ar—i

—I—\

 

Figure 2.2 A schematic diagram of an ICP torch and its orientation with regards

to the sampling cone of the MS.

The ICP is composed of a pneumatic sample nebulizer where a stream of
argon (Ar) gas aerosolizes the injected solution-phase sample into a fine mist.
The finer droplets are swept into the center channel of the plasma and heated to
>8000 °C.38 The torch consists of three annular tubes made from quartz, or
another high-temperature resistant material, wrapped with an rf coil. The rf coil
generates a high power radio frequency which inductively transfers the energy to
the Ar, forming an electromagnetic field within the stream of gas. An electric
spark is delivered to initiate the plasma which is then self-sustained at
atmospheric pressure.“ 41

The tubes making up the torch carry Ar of different flow velocities through

the rf coil region. Ar through the inner and secondary tubes is a support gas (0-1

Umin), which enters the rf coil region and is seeded with free energy from the

18

Tesla discharge coil. The electrons interact with the magnetic field and gain
energy, which then ionizes Ar leading to a self-sustained plasma. The sample is
then introduced in a flow of Ar through inner tube (~1 Umin). The outermost
stream of Ar serves as a coolant with a high flow velocity (~15 Umin). The
coolant stream dissipates heat generated by the torch and centers and stabilizes
the plasma. The solution sample is introduced in a flow of Ar through the center
quartz tube, typically at a lower flow rate (0.13-1.5 lein).41 As the sample
aerosol passes through the hot core of the plasma, the sample is desolvated,
dissociated into individual atoms, and ionized through the interaction with the
plasma, all before entering the MS. The sample introduction efficiency is
reportedly between 1 - 15% for ICP and more than two thirds of the elements on
the periodic table are > 80% ionized in the torch.“ ‘2

Since the ICP is operated at atmospheric pressure and the MS must be
operated under vacuum, the ionized sample must pass through a series of
differentially pumped chambers of sequentially decreasing pressure before
entering the MS.” ‘2 The sample ions enter the first vacuum chamber through a
1 mm hole in a nickel, water-cooled cone, called the sampling cone. The
sampling cone is aligned axially with the ICP central channel to optimize ion
transport into the MS. A slight vacuum is maintained behind the sampling cone
(approximately 2 x 10'3 atm) and serves as the first chamber of sample
introduction.42

Upon entry into the expansion chamber, the ionized sample and Ar

expand rapidly due to the change in pressure and form a shock-wave structure

19

called a Mach disc.42

A skimmer cone, much like the sampling cone protrudes
into the Mach disc and has a centrally located hole of 0.7 mm diameter. The
skimmer cone samples ions into an intermediate vacuum stage (< 10'7 atm) in
the form of an ion beam. The ion beam is deflected along a narrow path via a
series of electrostatic lenses into the mass analyzer.42 Under the influence of the
electrostatic lenses, positive ions are focused and separated from electrons and
neutral particles.

The mass analyzer then separates the ions based on a mass-to-charge
ratio (mlz). The quadrupole mass analyzer, like that used in this research,
consists of four metal rods arranged to form a hyperbolic cross section.
Controlled RF and direct current (DC) voltages are then applied to the rods which
form an electric field within the area bounded by the rods and serve as an ion
filter. Dependent on the RF/DC ratio, ions of a narrow m/z are able to pass
through the mlz filter to the detector. All other m/z ions hit the rods, are
neutralized and hence, are not detected. The ions that pass through the
quadrupole mass analyzer then strike a dynode detector. The initial ion beam is
converted into electrons for detection.” ‘2 By scanning the RF/DC ratio
sequentially, ions of different mlz travel through the area bounded by the rods

unhindered and reach the detector. Hence, a mass spectrum of the full mass

range is obtained.

20

2.4 ICPMS for GSR Detection

Very few papers have discussed the analysis of GSR by ICPMS, possibly
due to the high price of the instrument and lack of availability. However, ICPMS
instrumentation is decreasing in price making this technique more accessible.43
Although little has been published on this particular application, the instrument
remains very popular for many other areas of trace elemental analysis such as

37 At the time of this writing, five papers were

geological and marine sciences.
located dealing specifically with GSR analysis by ICPMS, one from 1998, 2003,
2004, and two from 2007.5, 31.4445 Three reports detailed collection of GSR from
shooters hands through cotton swabs wetted with either 5% nitric acid or 2%
solutions of ethylenediaminetetraacetic acid (EDTA).““‘46 The swabs were
allowed to digest in a 10% nitric acid bath followed by sonication, dilution and
analysis by solution-phase ICPMS. Koons reported the limits of detection for
121Slo, 138Ba and 208% for his method involving 5% nitric acid were 0.05 ug/L,
0.02 pg/L and 0.1 ug/L, respectively, with relative standard deviations of <5% for
all three elements.44 Rels reported similar values for the 2% EDTA swabs of
GSR, with detection limits of 0.05 ug/L, 0.51 port, and 0.12 pg/L for "“55, 138Ba,
and 208Pb, respectively.45 The third paper, a follow-up to Reis’s work, used a
similar sampling procedure to analyze GSR from two types of pistols using two
different ammunition types with the 2% EDTA method.46 In this paper, Sarkis did
not report any novel observations or results.

Of the remaining two papers, one describes a method of extracting GSR

from larvae removed after feeding on decomposing beef into which 2 shots had

21

been fired.31 Roeterdink described a method of digesting larvae in concentrated
nitric acid on a hot plate, with several steps of drying and reconstituting the
sample before ICPMS analysis. The study detected significantly higher levels of
121Sb, 138Ba, and 208Pb in larvae feeding on beef that was shot than control
portions.

Santos and coworkers estimated firing distance based on ICPMS of bullet
entrance holes.5 The researchers fired onto cotton tissue paper at varying
distances and cut 1 cm2 squares from the targets at four radial distances from
the gunshot hole. The squares were then digested for 24 hours in 10% nitric
acid, sonicated, centrifuged, diluted, and then analyzed by ICPMS. The team
determined that ICPMS was applicable for detecting GSR fired from distances as
far as 80 cm and radial distances up to 5.5 cm from the bullet entrance hole.
However, missing from the literature are studies that use ICPMS to detect GSR
from tissue samples and decomposing tissue samples, which adds significance

to the research reported herein.

22

Chapter 3

Materials and Methods

3.1 Experimental Design

Two euthanized pigs (200 lbs for GSR and 70 lbs for control) were
obtained from the Michigan State University (MSU) Swine Facility. The animal
use in this study was granted exemption from the guidelines set forth by the
Michigan State University Institutional Animal Care and Use Committee (IACUC).

The larger pig was shot 10 times by a local firearms specialist using a
Glock 9 mm with American Eagle 115 grain full-metal jacketed cartridges. Each
shot was fired at a constant muzzle-to-target distance of 5 cm, spacing wounds
at least 10 cm apart to minimize cross-contamination of gunshot residue between
wounds (Figure 3.1). The gun was cleaned thoroughly between shots to prevent
build-up of residue within the barrel. The pigs were then placed in a MSU
research field and covered with a screen box to prevent attack by predators while
still allowing exposure to weather.

Gunshot wounds were excised with a surgical scalpel from the shot pig
daily for seven consecutive days. On day one, four wounds were excised from
the shot pig; three wounds were used to optimize sample preparation and
analysis procedures for both SEM/EDS and ICPMS, while the fourth wound was
actually analyzed. On days two through seven, wounds were excised from the
areas nearest the head first, working sequentially toward the rear of the pig.

Previous studies conducted by the MSU forensic biology program indicated that

23

decomposition and larval destruction of the tissue occurred most rapidly at the

head and progressed toward the rear of the carcass thereafter.

 

Figure 3.1 An illustration of the spacing and arrangement of the gunshot
wounds on the pig. The illustration is not drawn to scale.

Hence, removing the wounds nearest to the head first ensured less
potential for larval damage to the wound. Tissue samples were also excised
from the control pig, one for each consecutive day. The tissue samples collected
from the control pig were of approximately the same dimensions as the shot pig.
The excised wound tissue was then wrapped loosely in wax paper and stored in

a plastic bag in a freezer at -80 °C until analysis.

3.2 SEM/EDS Sample Preparation and Instrumental Parameters

Sample tissue for SEM/EDS was removed from the storage freezer and
the adipose was removed. The sample was then prepared for analysis by either

air-drying or freeze drying. The air-dried samples were stretched and pinned to a

24

small piece of cork board. The samples were elevated from the cork board by
stretching the tissue over two disposable pipette barrels to prevent accumulation
of moisture between the board and the tissue. The tissue was then placed in a
glass petri dish and covered with a watch glass. The sample was allowed to dry
for several days in a hood until it became translucent and rigid. The tissue was
then removed from the cork backing and coated with a layer of carbon by a EFFA
Carbon Coater (Ernest F. Fullam, lnc., Latham, NY).

Freeze-dried samples were inserted into a Flexi-Dry freeze-dryer (FTS
Sytems, Stone Ridge, NY) held between the range of -54 °C and -84 °C and 25
mTorr for 24-48 hours until completely dried. The tissue was then coated in
carbon as described previously. Both air-dried and freeze-dried samples were
stored under vacuum in a dessicator until SEM/EDS analysis.

Tape lift of the gunshot wounds and control tissue was performed using
small aluminum stubs with one face covered with double-sided adhesive carbon
tape. The wounds were dabbed 100 times with the adhesive covered stub and
then placed in a plastic Petri dish until analysis. The number of dabs used for
sample collection was reported as the optimal amount in a review article which
discussed the effectiveness of tape-lifting GSR versus number of dabbings.16

SEM analyses were conducted on a JEOL 6400V SEM (Japan Electron
Optics Laboratories, Tokyo, Japan) equipped with a LaBs emitter and back-
scattered electron detector. EDS analysis was completed with an lNCAx-sight
analyzer (Oxford Instruments, Oxfordshire, UK). EDS quantitative optimization

was performed every 2 hrs through analysis of a pure copper grid affixed to the

25

brass sample housing. Dead-time was adjusted to lie within the range of 30-
40%. The accelerating voltage was set to 20 kV and the working distance to 15
mm. Imaging was performed by both secondary and back-scattered electron

detection.

3.3 ICPMS Sample Preparation

For ICPMS analysis of tissue, separate tissues were removed from the
freezer, cut in half, and the adipose was removed from the skin using a clean
surgical scalpel. One half of the tissue was used for ICPMS analysis, while the
remaining portion was placed back into the freezer until analysis by SEM/EDS.
The underlying flesh was discarded and the skin was used for analysis, as this
research was concerned about the detection of GSR from decomposing tissue.
Although there undoubtedly was some trace GSR in the bullet track, it was much
less than that deposited topically around the wound. The skin from the control
and wound tissue was then microwave digested.

Approximately 0.5 — 1.0 g of tissue was weighed out and placed in a 30
mL high-purity quartz vial (Milestone lnc., Shelton, CT). Along with the tissue
inside the quartz vial, 2 mL of high-purity nitric acid (70%, Optima, Fisher,
Pittsburgh, PA) and 1 mL of hydrogen peroxide were added. The quartz vial was
then placed into a 100 mL Teflon-lined reaction vessel (Milestone lnc., Shelton,
CT) that contained 10 mL deionized water and 2 mL of hydrogen peroxide. The
reaction vessel was sealed to a uniform pressure using a calibrated torque

wrench supplied with the digestion unit. The sample tissue was then digested

26

using a Milestone Ethos EZ microwave digestion system (Milestone lnc., Shelton,
CT).

The digestion procedure started at ambient temperature followed by a 15
min. ramp to 200 °C at which the temperature was held constant for 10 min.
before a final cool down back to ambient temperature. Temperature was digitally
monitored via a thermo—coupled probe which was inserted into an enclosed
thermowell located in one of the digestion vessels. The cool down was required
to minimize the internal pressure of the reaction vessel. Typically, digestion
vessels were opened once the internal temperature was below 100 °C. Of the
digested tissue solution, 430 pL was then diluted to 2% nitric acid for ICPMS
analysis. The remaining digest solution (approximately 2.6 mL) was stored under
refrigeration at approximately 10 °C.

For the 7-day sampling period, one gunshot wound and control tissue
sample was collected each day, resulting in a total of 14 tissues. The microwave
digestion unit was able to digest 5 samples simultaneously, each batch consisted
of 4 tissue samples and 1 blank in each batch. Along with each set of tissue
samples that were digested, a procedural blank was run in order to evaluate any
contamination. The procedural blanks were prepared by the above procedure
except without any tissue samples. A procedural blank was run along with each
batch of tissues digested, resulting in a total of 4 procedural blanks.

A cleaning step was performed to prevent cross-contamination between
sets of digests. Cleaning was carried out by preparing the quartz vials and

digestion vessels as described above except without any tissue samples. The

27

same microwave procedure was used as previously described and once the
procedure was completed, the vessels were removed from the microwave,
disassembled, and rinsed with deionized water. After the cleaning step, the next

set of tissues were digested.

3.4 ICPMS Operating Parameters

Calibration standards and samples were analyzed on a Micromass (now
Thermo Fisher Scientific Inc., Waltham, MA) Platform quadrupole ICPMS with a
quartz torch (1.5 mm diameter injector), hexapole collision cell, and a CECTAC
ASX-500 autosampler. Additional operational parameters are contained in Table
3.1 shown below. Tune conditions were optimized using a 10 pg/L solution of
9Be, 59Co, ”In, 140Ce, 2""Bi, and 238U. Selected ion monitoring (SIM) was used
for 1”Ba, 208Pb, and 121Sb with a sample scan time of 1.25 min. and a dwell time

of 0.1 5.

Table 3.1 ICPMS instrumental parameters.

 

 

 

 

 

 

 

Parameter Setting
Cool Gas Flow Rate 13.00 Umin
Auxiliary Gas Flow Rate 0.72 Umin
Sample Gas Flow Rate 0.72 Umin
RF Power 1350 W
Sampling Cone Ni with Cu core, 1.14 mm diameter orifice
Skimmer Cone Ni, 0.89 mm diameter orifice

 

 

 

 

28

Calibration standards of 138Ba, 208Pb, and 121Sb were prepared in 2% nitric
acid at concentrations of 0.5, 25, 250, 500, and 1000 pg/L by serial dilution. 11%
and 209Bi were added as internal standards to the external standards, digested
tissue, and procedural blanks. The calibration set was run at the beginning, half-
way through, and at the end of tissue sample analyses to compensate for any
ICPMS response drift or memory effects. Memory effects can be described as
the retention of analytes within the system, which attribute to a greater response
observed for subsequent sample injections.

Samples were injected via an autosampler and between each sample
analysis, the injector was rinsed for 3 minutes with 2% nitric acid. The gunshot
wound and control tissue digests as well as the procedural blanks were injected
in the order of expected lowest analyte concentrations; the procedural blanks,
followed by the control tissue samples, then the gunshot wounds starting with
day 7, and proceeding through the highest expected analyte concentrations on
day 1. This order was chosen in order to ensure that there would be minimal
memory effects on the target analyte concentrations.

The instrument response for analytes in the samples was normalized to
the internal standards. 115In was used for normalization of 121Sb and 138Ba, and
209Bi was used to normalize the response for 208Pb in the tissue samples. The
calibration curves were used for quantification of the concentration of analytes in
the tissue samples. Concentrations determined by ICPMS were in parts per

billion (ppb) which is equivalent to pg of analyte per liter of solution (pg/L). The

29

concentration was then multiplied by the dilution factor followed by the volume of
solution in which the tissue was digested. The previous calculations determined
pg of analyte in the stock digest, which was then divided by the sample tissue

weight to result in pg element per gram of tissue (pg/g).

30

Chapter 4

Results and Discussion

4.1 Introduction

GSR has been studied a great deal over the course of several decades,
however as revealed in the literature review in Chapter 2, there are very few
papers discussing direct sample analysis.16 In most reported analyses by
SEM/EDS, GSR samples were collected by means of tape lifts.7' 16' 19' 22' 27' ‘7
The researchers using bulk analytical methods such as FAAS and ICPMS mostly
report that cotton swabs are first used to collect GSR from samples before
analysis. The objective of this research was to determine the persistence of GSR
in decomposing tissue, using SEM/EDS and ICPMS and analyzing the tissue
directly rather than a swab of the tissue. In order to fulfill this objective, the
project had three goals. First, the optimum method of sample preparation for
SEM/EDS analysis of tissue samples containing GSR was determined and the
applicability of SEM/EDS for direct tissue analysis was assessed. Second,
ICPMS analysis of tissue samples containing GSR was investigated. Finally, the
GSR content on decomposing tissue as a function of time, using both SEM/EDS

and ICPMS, was investigated.

4.2 Decomposition Observations During Sampling Period

Porcine skin has been demonstrated to possess similar characteristics to

human skin for gunshot residue studies as reported by Haag et al.48 Hence,

31

tissue samples from pigs were used in this study to simulate human skin as
closely as possible. Two pigs were used in the experiments discussed herein as
a preliminary effort to lay the groundwork for future studies. Figure 4.1 is a
picture of two gunshot wounds on the pig. Wounds had a noticeable blackened
ring of soot and GSR with a radius of approximately 3 cm. Two black markings

below the upper wounds show future wound sites, 10 cm apart.

 

Figure 4.1 A picture of the resultant wounds and spacing on the pig.

After the pig was shot, both the control pig and the shot pig were
transported from the firing range, placed about 2 meters apart in the MSU
entomology field, and covered with separate screen boxes. The pigs were
sampled daily during mid-July for a one-week period. The weather conditions,
shown in Table 4.1, illustrate that during this particular week the Lansing area

experienced a heat wave with high temperatures ranging from 87 - 96 °F.

32

Table 4.1 The weather, dates, and temperature for the course of the seven day
collection period.

 

 

 

 

 

 

 

 

Day Date Temperature (°F) Weather
1 July 13, 2005 ['33 :1 Hot and sunny
3 July 15, 2005 '33:): ,3; Hot and sunny
4 July 15, 2005 Efvt‘fi: Hot and sunny
5 July 17, 2005 'Efljé’; afie$:22x,uhrggi:t:.m.
raln
6 July 18, 2007 E33 :3: Hot and sunny
7 July 19, 2005 if: :77 Hot and Sunny

 

 

 

 

 

 

During this period, it was sunny for the majority of the time, however there
was a light rain shower over the night of the July 14'" and torrential rain through
the night of July 17’“. Weather is of particular importance in this study because
the elevated temperatures and rain cause decomposition to occur at an
accelerated rate.49' 5° The rain was of further concern because it was believed
that it might wash away GSR making its detection by SEM/EDS and ICPMS
difficult.

The pigs decomposed rapidly throughout the course of the week. On
days one and two the shot and control pigs were becoming bloated and there

was little insect activity (Figure 4.2a). The control pig was noticeably colonized

33

by insect larvae on the morning of day three and appeared to have similar larval
activity as the shot pig. However, the progression of decomposition and
desiccation that followed was much more rapid on the control pig than the shot
pig.

The control animal was reduced to dried skin and bone by the 5"1 day; the
larvae were much less active at this point and could be seen leaving the carcass.
The difference in the speed of decomposition between the two animals may be
due to the mass of the control pig versus the shot pig. The control carcass was
about 70 lbs. compared to the shot pig, which was about 200 lbs. The larger pig
accounted for much more tissue for insects to consume therefore taking a few
days longer. For the shot pig, reduced larval activity was not observed until the
7th day, during which the larvae were observed migrating from the body. By day
8, the shot pig was reduced to a state (Figure 4.2b) comparable to that observed

of the control pig on day 5.

34

a.)

b.)

   

tiff. /’

Figure 4.2 Pic ures collect

decomposition of the shot pig.

35

-~

a\

‘5

nd

 

   

4.3 Analysis of tissue for GSR by SEM/EDS

4.3.1 Optimizing sample preparation for SEM/EDS analysis

Particular attention was paid to the method of sample preparation of wound
tissue for SEM/EDS analysis of GSR. The methods of sample preparation had to
leave the wound tissue relatively undisturbed for too harsh of a method would
remove/dissolve GSR and ruin any chance of detection. For the previously
stated reasons, a common sample preparation method for SEM, critical-point
drying, could not be used. Critical-point drying requires that the sample be
submerged in a series of solutions of increasing ethanol content to remove water
before the final step of dehydration using liquid carbon dioxide.” The ethanol
solutions would have removed GSR from the wound tissue.

Three methods of sample preparation were determined feasible and
attempted; air-drying, freeze-drying and finally, tape-lifting GSR from the
collected wound tissue.16' 22' 33 For samples prepared by freeze-drying and air-
drying, the adipose layer of the tissue was first removed with a clean surgical
scalpel. Once the adipose was removed, samples that were air-dried were
stretched and fixed to cork board. Samples took approximately four days until
they were considered thoroughly dry, at which time the tissue was translucent
and stiff. Samples that were freeze-dried were placed in the Flexi-Dry freeze
dryer for 48 hours until the sample canister did not feel cold to the touch, as
described in the Flexi—Dry manual. For tape-lifting, each tissue was dabbed 100

times with an adhesive-covered stub.

36

All samples analyzed by SEM/EDS were coated with a thin layer of carbon
prior to analysis to ensure electrical conductivity. Vlfithout an adequate carbon
layer, samples can become charged and appear bright white or appear dark,
obscuring sample imaging and analysis. If the SEM images appeared either too
bright at spots or too dark, the samples were promptly removed from the SEM
chamber and coated with another layer of carbon.

Under the SEM, it was observed that the air-dried (AD) and freeze-dried
(FD) tissues had very different features as illustrated in Figure 4.3. The AD
tissue contracted to form a slightly ridged structure, whereas the FD tissue
appeared more wrinkled. The AD sample images appeared much clearer than
that of the FD sample as shown in the figures below. Despite the apparent clarity
of the image, when the area to be searched for GSR was magnified, it became
difficult to identify any GSR particles due to the debris and hair present on the
surface of the tissue. At high magnification (~15,000x), the tissue was difficult to

image without damaging the sample with the electron beam in the process.

37

 

Figure 4.3 SEM images of freeze-dried (top) and air-dried (bottom) wound

tissue prepared for analysis.

38

The FD samples appeared different from the AD sample; the FD sample
had much more interfering debris and suffered from its greater sensitivity to the
electron beam. FD samples, much like the AD tissue, had a greasy film coating
them after preparation. At high magnifications and accelerating voltages, the FD
samples could be observed to pucker and deform under the beam, more so than
the AD sample. The SEM images that were collected lacked clarity and suffered
still from poor electrical conductivity despite carbon coating.

All of the tissues directly analyzed by SEM, whether AD or FD, were difficult
to image consistently. Samples from the control pig were also prepared via AD
and FD and similar difficulties in sample preparation were encountered. Despite
difficulties in viewing the AD and FD wounds by SEM, it was possible to locate
GSR particles on wounds from days 1 and 2, as shown in Figure 4.4. The
blurring at the bottom of the image is believed to be due to some latent moisture
or other volatiles still present in the sample.

The GSR particles were located on the tissue by careful and tedious
examination at high magnifications between 15,000 and 75,000x. Puckering of
the tissue was observed at times from the electron beam interaction. The
accelerating voltage was set to 25 W in order to generate a detectable x-ray
spectrum; this high energy is also sufficient to destroy the tissue. Imaging for too
long a period on any one area led to the loss of focus of the image and at times
the GSR particle would disappear from view. It is believed that the GSR particle

would embed itself into the tissue when viewed at very high magnifications or

39

when imaged too long. Similar phenomena have been reported to happen on

tape adhesive.”

 

Figure 4.4 SEM secondary-electron image of a GSR particle from day 1 on an

AD wound tissue sample.

Although the goal was to determine GSR directly on the tissue, it was
apparent that direct analysis was much too difficult due to the changing surface
of the tissue with time, the obscured GSR particles, and the labor-intensive
sample preparation.

Tape-lifting provided a less labor-intensive approach for sample preparation
as well as achieving consistent, clear images in which GSR was more easily
located. Tape lifts of the gunshot wounds and control tissues were used for

subsequent analyses due to the aforementioned reasons. Figure 4.5 has two

40

images; the first illustrates the clarity of the surface morphology of the carbon
tape which was much more easily scanned for GSR than either the AD or FD
tissues. The second image shows a back-scattered electron (BSE) capture of a
tape-lifted sample of GSR from tissue collected on day 1.

The white dots in the bottom image in Figure 4.5 are images of GSR
particles resulting from the electron interaction with the metals of the particulates.
BSE detection is sensitive to increasing atomic number and the image produced
shows differences in contrast proportional to atomic number; areas containing
elements with higher atomic number appear brighter. This feature is related to
the size of the atomic nucleus; a larger nucleus results in greater electron beam
interaction and therefore more detectable signal is produced and detected. BSE
images allow one to qualitatively differentiate GSR (higher atomic number) from

the background (carbon tape).

41

 

Figure 4.5 Images taken by SEM. A secondary electron image of unused
carbon-coated carbon tape (top) and a back-scattered electron image of a tape-
Iift from a gunshot wound collected on day one (bottom).

42

4.3.2 SEM/EDS Analysis of Tape-lifts

The tape-lifts were taken for each of the gunshot wounds and
corresponding control tissues (n = 14). The GSR was dense on tape lifts from
wounds collected on days 1 and 2 and particles were easily located in under two
minutes of searching. Example BSE and SE images are provided in Figure 4.6.
The images illustrate the observable differences between the two detectors; BSE
provides a high-contrast image capable of distinguishing differences in elemental
atomic number, whereas the SE image provides a higher resolution image for
observing morphological features. BSE imaging quickly revealed deposits of
GSR at low magnification. GSR particles were then singled out and imaged at a
much higher magnification until they took up most of the viewing screen.

The first feature that was observed was the size, shape, and morphology
of the potential GSR particle. Several papers have described in detail the
particle morphology encountered when analyzing GSR.19'27 The most common
consensus is that GSR particulates are spherical in shape, some may be perfect
spheres while others are irregular and distorted, which is most often the case
observed in the research discussed herein.19 The surfaces of the particles have
been reported as being anywhere from smooth to resembling ‘peeled oranges’.27
The resolution of the particles imaged in this study was not adequate to
definitively observe the topology of the particle surfaces. Most of the GSR
particles that were detected throughout these analyses were in the 0.5 pm - 2

pm diameter size range and existed as individual spherical particles with a slight

distortion on one side of the particle as seen in the images in Figure 4.6. Papers

43

have reported much larger particles in the area of 50 pm however, the largest

that was discovered herein was approximately 9 pm.27

 

 

Full Scale 10.154

 

 

 

 

    

o 8 keV

Figure 4.6 BSE (left) and SE (right) images for a single GSR particle on the
tape-lift from a gunshot wound of day 1. The EDS spectrum below the images

shows the elemental composition of the particle.

The elemental composition of a number of these particles was determined

using EDS after initial imaging of the particle was performed. EDS analysis

44

indicated the presence of Sb, Ba, and Pb at detectable levels in most of the GSR
particles located and analyzed. Figure 4.6 shows an x—ray spectrum resulting
from the EDS analysis of the particle shown above in the same figure. As one
can see, significant levels of Sb, Ba, and Pb were detected. The spectrum in
Figure 4.6 reveals multiple peaks for each element as was discussed in Chapter
2. Elemental mapping (Figure 4.7) was also conducted and showed the
distribution of each element detected by the analyzer. The mapping showed that
the Sb, Ba, and Pb that was detected by EDS was actually within the particle and

not the background.

 

Sh L81 Ba Lat Pb L51

Figure 4.7 A SE electron image of a GSR particle from the tape lift of the
gunshot wound from day 1 (above) and three elemental maps acquired by EDS
detection showing the relative intensity of Sb, Ba, and Pb (L to R).

45

The combination of these three elements is considered a diagnostic
criterion of GSR.16' 27 Several studies have attempted to determine the elemental
ratios of Sb, Ba, and Pb as well as other detected elements such as Cu and Al,
but have concluded that each particle has large elemental variability.19'27 Each
individual GSR particle may contain the same elements but not at the same
levels.16 In this study, the levels of Sb, Ba, and Pb were found to vary by several

weight percent, as shown in Table 4.2 for the tape lifts of wounds from day 1.

Table 4.2 Average weight percent of Sb, Ba, and Pb in GSR particles collected
from gunshot wound tissue sample of day 1.

 

 

 

 

 

Weight % Standa(ll"d=Dse)viation
Sb 1.44 0.83
Ba 2.60 1.99
Pb 12.60 9.87
Cu 0.34 0.17

 

 

 

 

 

As element abundances are highly variable, their presence is more
diagnostic than the relative proportions of each. Due to such variation, it may not
be possible to differentiate ammunition according to manufacturer based on
elemental ratios. However, such differentiation may be possible based on trace
elements present in the primer.16' 5‘

In the EDS spectra collected from the wounds on day 1 and 2, Cu was

also routinely detected. The Cu that was detected was typically at much lower

levels than the Sb, Ba, and Pb as shown in the x—ray spectrum in Figure 4.8. A

46

background spectrum of the carbon coated tape revealed the presence of trace
Cu, which may contribute to the Cu observed in the sample spectrum. However,
due to the levels of Cu observed in the sample, it is possible that the copper
jacket of the bullet was also making a significant contribution to the elemental
profile observed.

Aluminum was also observed in a few GSR samples, however, this
element was observed very rarely. Figure 4.8 is an x-ray spectrum collected
from one particular GSR particle located on the tape-lift of the gunshot wound

tissue from day 2.

Full Scale 17892 Cts.

 

 

. CU CU

 

7 8 Ire

Figure 4.8 An x-ray spectrum for GSR located on the tape-lift from the gunshot
wound sample collected on day 2. Of note is the aluminum peak in the

spectrum.

According to the Federal Cartridge Company datasheet on the
ammunition composition, Al is present in the 9 mm American Eagle cartridges

which would explain its presence in the spectrum. The peak corresponds to 2.41

47

weight percent of the particle, which compared with the data in Table 4.2 falls
between the average weight percents of Sb and Ba. No GSR particles
containing Sb, Ba, and Pb were visible in wounds collected on days three
through seven or on any of the control tissues. The fact that no GSR detected on
control tissue indicates there was no contamination. However, the lack of GSR
in wounds from days 3-7 was more surprising given the dense and easily visible
GSR on day 2.

From Table 4.1, it can be seen that during the evening of day two, the
area experienced a rainstorm. The rain may account for the abrupt loss of any
detectable GSR on the following days. The GSR that was observed on days one
and two consisted of very small particles, typically less than 1 pm in diameter and
they were easily removed via tape-lift, leading to the conclusion that rain could
have easily washed the loose GSR from the surface area around the wounds.

When analyzed by SEM, spherical particles were observed in the tape lifts
from wounds collected on days 5 and 6. Initially, when imaged using BSE
detection, these particles were thought to be GSR due to their bright appearance,
indicating elements of high atomic number. However, analysis by EDS indicated
that, while the spherical particles were in fact metal, the metal was a mix of Fe,
Cr, and Ni (Figure 4.9) with no Sb, Ba, or Pb present. This combination of Fe-Cr-
Ni is consistent with stainless steel alloys, which may be possibly derived from
the barrel of the handgun. The particles detected on days five and six appeared
different than the characteristic GSR found on days one and two. The Fe-Cr-Ni

particles found on the day 5 and 6 gunshot wound tape lifts could be imaged with

48

better resolution than the GSR found on the tape lifts from the gunshot wounds of
days 1 and 2. The Fe—Cr-Ni particles also appeared almost perfectly spherical

whereas the GSR was spherical with slight deformation such as elongation.

 

 

Full Scale 10,154 cts

 

 

 

Figure 4.9 An example SEM image of the Fe-Cr-Ni metal particle discovered on
a tape lift from day 5. The x-ray spectrum below shows the highest elemental

abundance due to Fe.

49

The Fe—Cr—Ni particles have a similar morphology as D’Uffizi et al.
described in their GSR research, however, the composition does not have any
consistency with ‘unique’ GSR.19 The Fe-Cr—Ni particles were also found to be
slightly larger than the GSR with diameters between 3 — 5 pm. One paper briefly
mentions the presence of Fe—Cr-Ni particles related to GSR but considers such
particles of little evidential value.27

Another interesting observation was that the particles were found only on
days five and six. Although no literature could be located which discusses this
observation, it is supposed that it may be due to the physical properties of the
particles themselves. The Fe-Cr—Ni particles were larger in diameter than the
characteristic GSR that may have allowed for the particles to embed within the
skin around the gunshot wound. As the tissue decomposed, the skin around the
particles may have receded and exposed the Fe-Cr-Ni particles, which then on
days five and six were easily tape-lifted from the wound. During the evening of
day five, the area experienced heavy rain. Expectedly, the Fe-Cr-Ni particles
would have been washed away as the GSR had been during the rain on the
evening of day 2. However, on day five, six Fe-Cr—Ni particles were located and
after the rain on day six, one was found, providing evidence that the Fe-Cr—Ni
particles were more tightly bound than the GSR. On day seven no particles of

GSR or Fe-Cr-Ni were located.

50

4.4 Summary of SEM/EDS Results

The results of this study are multidimensional. First and foremost, it was
found that tape lifting GSR from tissue was more consistent and less labor-
intensive than direct tissue analysis prepared either by air-drying or freeze-drying
tissue samples. Tape lifts provided a more stable substrate for imaging the
particles. With the tape lifts, high magnification and high accelerating voltages
could be used in order to acquire clearer images and spectra whereas this was
impossible on tissue since the substrate is so sensitive to the electron beam. It
was also corroborated that elemental ratios for GSR were highly variable. The
level of variability for each of the elements detected was large, indicating that
elemental ratios may not be diagnostic for a cartridge manufacturer. GSR was
easily found on wounds from days 1 and 2, however after a brief rainstorm, it
appeared that the GSR may have been washed from the wounds. Weather
plays an important role in the detection of GSR particles from gunshot wounds in
fissue.

The Fe-Cr-Ni particles although interesting are not conclusive GSR
evidence. It the particles were either found in conjunction with Sb, Ba, and Pb,
they would be of greater significance, however since they lack the ‘unique’
elements, they are of little worth. Although it may be possible to determine if the
particles were from the barrel of the gun the casings were not collected and

investigation of the barrel composition is beyond the scope of this project.

51

4.5 Analysis of Tissue for GSR by ICPMS

4.5.1 ICPMS Calibration

Before the tissue digests were analyzed, the instrument was calibrated
using standard solutions of 121Sb, 138Ba, and 208Pb with concentrations of 0.5, 25,
250, 500, and 1000 ppb that were prepared in 2% nitric acid by serial dilution. In
addition to the elements of interest, two internal standards were also used;
indium ("5In) and bismuth (2°9Bi). These elements were chosen as internal
standards since their atomic masses are above and below the atomic masses of
the target analytes, and they are not commonly encountered elements in high
concentrations.

Internal and external standards were used in this study. The internal
standards, 115In and 209Bi, served as a means to normalize the instrument
response, and the external standards were used to quantify the concentration of
target analyte based on the response. Figure 4.10 shows the resultant
calibration curves generated for 121Sb, 138Ba, and 208Pb. Each data point on the
graph is the average of the detector response for three injections. The standard
devation for each point was less than the data point itself so the error bars were
omitted from the plot for clarity. The plot illustrates linearity for all three analytes
in the range of 0 - 1000 ppb with r2 values greater than or equal to 0.999 and

intercepts at or near 0.

52

ICPMS Calibration
50

 

 

0135 Ba
45 " 1.12155

40 l 5205 Pb /
3. /

30 / /¢
25 / /

.5 //

10 // /'

 

 

 

 

 

 

 

 

 

 

 

Average Normalized Instrument
Response (n = 3)

 

 

 

 

0 200 400 600 800 1000 1200
Concentration (ppb)

Figure 4.10 Calibration curves for the target analytes, 121Sb, 138Ba, and 2”Pb.

A 30 detection limit was calculated from the calibration curve data as
described by Koons.44 The detection limit was determined by multiplying the
standard deviation of the blank (concentration of 0 ppb) by three (30), then
dividing by the sensitivity, which is the slope of the standard response curve for
each element.44 The detection limits for 121Sb, 138Ba, and 208Pb were determined
using the formula above and were 0.002, 0.033, and 0.009 ppb, respectively.
These values are comparable to those Koons determined for the same analytes
in GSR using ICPMS. Koons collected GSR from shooters hands with cotton
swabs then digested the cotton tips in a vial of nitric acid. Koons” limits of

detection for Sb and Pb were higher than those determined in this study by

53

approximately 1-2 orders of magnitude with values of 0.05 and 0.1 ppb,
respectively. However, Koons reported a lower limit of detection for Ba (0.02
ppb) than the value reported herein (0.033).44 The calibration curves illustrated
that the ICPMS has a wide linear dynamic range and good sensitivity, which

makes this instrument a valuable tool for trace elemental analysis.

4.5.2 ICPMS Precision

Ten separately prepared dilutions of the digest of a single gunshot wound
sample were analyzed in order to evaluate the precision of the instrument and
homogeneity of the sample. The data from these analyses can be seen in Table
4.3. The relative standard deviation (RSD) for the 10 injections was less than 3%
for all three analytes, which is evidence attesting to the precision of the

instrument and that the sample digests were homogeneous.

Table 4.3 ICPMS data for 10 individually prepared gunshot wound dilutions from
day 3.

 

 

 

 

 

Relative Standard
Average (”g/9) Deviation (RSD)
Antimony (1218b) 3.35 2.00%
Barium (5533) 3.95 2.51%
Lead (mph) 5.29 1.50%

 

 

 

 

 

4.5.3 ICPMS Analysis of Control Tissue Digests
Figure 4.11 shows example ICPMS spectra, with mlz along the x-axis and
the relative percent intensity on the y-axis. The spectra for a gunshot wound and

control tissue collected on day 3 illustrate the difference in the intensity of the

54

signal for the target analytes, 121Sb and 138Ba. The spectra are also provided to
show that there are not many interfering peaks (i.e. low backgrounds) even
though the analytes are in a complex matrix of digested tissue. The isotopes of
Sb are present at m/z 121 and 123 and are observed in the wound tissue but not
in the control.

The most abundant isotope of Ba is observed at mlz 138 in the gunshot
tissue but not in the control. 138Ba has a natural abundance of 72% and the next
highest Ba isotope is 1373a with a natural abundance of 11%, approximately a 7:1
ratio between these peaks is expected and is seen in the spectrum peak
intensities. 137Ba should be the second most abundant isotope of Ba present in
the spectrum. The peaks present at the other mlz corresponding to Ba isotopes
(i.e 130, 132, 134, 135, 135) should therefore be of lesser intensity than 1375a if
they truly corresponded to the isotopes. Although several other isotopes of Ba
are observed in the spectra, they are likely interfering peaks from the matrix

since they would not be present in such high concentrations naturally.

55

138

10°] Gunshot wound, day 3

   
    

134 136

Ba137
~15: ..

    

   
 

  
   

961 132

  
 

134 133
Ba Ba

   

130
Ba

    
   
 

01......

120 122 124 ‘ 1'25 120 ‘ 130 132 134 135 135

 

Figure 4.11 ICPMS spectra of the digests of a gunshot wound and control tissue
from day 3. The spectra show the mlz range of 120 to 140 amu.

In Figure 4.12, the large peak appearing at m/z 209 is of 209Bi, an element
used as an internal standard throughout the study. 2°3Pb is seen also in the
wound tissue, but at much lesser intensity than the internal standard. No 208Pb

was observed in the control tissue.

56

209
Bl

100, Gunshot wound, day 3

  
   

O/o..

208
206 207 Pb

  

 

o I I v r

100 209
,Control, day 3

J

 

 

1

 

 

0 e . . r . . s . - . . . . . . 5 fi
200 202 204 206 208 210 212 214 216 218

 

t

Figure 4.12 ICPMS spectra of the digests of a gunshot wound and control tissue
from day 3. The spectra show the mlz range of 200 to 220 amu.

The control tissues collected from the seven consecutive days, were
analyzed. Figure 4.13 shows the plots of the concentrations of the target
analytes versus the day of collection. A spike in the concentrations of 138Ba and
2°"Pb was observed, first occurring on day 5 and then fluctuating through day 7.
The levels of 138Ba increased from 0 pg/g on day 4 to 0.30 pg/g on day 5

followed by a decrease to 0.17 pg/g on day 6 and rising slightly on day 7 to reach

0.19 pg/g.

57

Element Concentrations from Control 'l'lssue

 

 

 

 

 

 

0.345 - ——~—_~-~s~:~—~;-m__-_- _-_u:--..__. L, - f
—a—- 121 so i

3 0.295 ' T '138 Ba 5 l
8 —-— 205 Pb ; g
'f'r. - . I

0.245 - . .. . . r, - «
E ; ,
E .' X l
.2 0.195 ' , , , .7 . ‘ ' . . l
u . “ . ' 1
g . . ‘ }
§ 0.145 5 . ,
r: ' '
8 . ,
.. 0.095 .
C r
w 0
g 0 045 .'
m - '. ”k

. / \ \ ‘— -‘.
-0305 —- - l——-.l——I——l£___fi____£'==a-_;__l

 

 

 

Day of Collection

Figure 4.13 Plot of concentration of 121Sb, 138Ba, and 2°"Pb versus the day of

collection.

The level of 208Pb increased from 0 pg/g on day 4 to 0.03 pg/g on day 5
followed by a decrease to 0.003 pg/g on day 6 and a rise to 0.01 pg/g on day 7.
The levels of 2“Pb were much less than the 138Be during days 5, 6, and 7. The
concentrations of 121Sb remained at 0 pg/g for the entire 7-day sampling period.
The observed spikes of 138Ba and 203Pb are of unknown origin although possible
factors include; insects, rain, or other environmental factors such as fertilizer
overspray since the entomology field is located adjacent to agricultural land. This
observation provided evidence that GSR contamination was not the origin of the
spiking since one would have expected to see elevations in all three elemental

levels if that were the case. The differences in the levels of 121Sb, 138Ba, and

58

208Pb were very small, their origin may be due to the different tissue samples

collected on each day.

4.5.4 ICPMS Analysis of Gunshot Wound Tissue

Figure 4.14 is the concentration of 121Sb, 138Ba, and 208Pb plotted versus
the day of collection for the GSR tissue. The top plot is the full scale version of
the graph and the plot below is the same plot except with the scale expanded to
include only 0 - 1 pg/g. The plots show that with the passage of time, the
concentration of elemental GSR decreased. The decreases from day 1 to day 7
were 9.19 pg/g to 0.15 pg/g for 121Sb, from 30.67 pg/g to 0.06 pg/g for 138Ba, and
26.71 pg/g to 0.10 pg/g for 208%. The most significant decrease in concentration
occurred for 121Sb and 208Pb between days 1 and 3. 138Ba showed its greatest

decrease between days 1 and 2, decreasing from 31 pg/g to 0.06 pg/g.

59

Element Concentrations from Gunshot Wounds

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

__ --___-__._,
A -5-121 Sb 1
0
g 30 +138 Ba
2; 25 +208 Pb
g 20
C
3 15
c
o
3 10
C
E
2 5
m

0 .
0 1 2 3 4 5 6 7 8
Day of Collection
1 Element Concentratlons from Gunshot Wounds

A -0r-121 Sb
5
a +138 Ba

_ 0.8 - — 1 -——~ —-

*- -l-208 Pb

.2

3

5 0.6

t
8 0.4
C
o

o

E
E 0.2

2

tn

0 . , .
0 1 2 3 4 5 6 7 8

Day of Collection

Figure 4.14 Plots of GSR concentration versus the day of collection. The upper
plot is the full-scale version showing all of the data points and below is a
magnified scale to show the values of GSR on later days of collection.

60

The drastic decrease in GSR within the first three days for all three
elements is likely due, in part, to the weather. The rain on the evening of day 2
cannot account for all of the decrease of GSR, which was observed over the
course of day 1 — 2, but it is believed to have some role in the decreases. It does
not appear by the trends in Figure 4.14 plots that the hard rainfall on the night of
day 5 had much of an effect on the concentrations of GSR on the wounds. The
GSR on the tissue at these later stages of decomposition may have been
embedded in the skin making the rain less effective at washing away the
residues.

Although the overall trend for the GSR analytes was to decrease, spikes in
the concentration of the analytes were seen on day 4 for 121Sb, and day 5 for
138Ba and 208Pb. 121Sb increased from 0.47 pg/g to 0.61 pg/g, 138Ba increased
from 0.05 pg/g to 0.10 pg/g, and 208% increased from 0.39 pglg to 0.50 pg/g.
The increases that were observed were all less than 150 ng/g of each respective
analyte and hence, were considered to be of little concern. The small increases
could be due to differences in the weathering and/or decomposition of each
particular wound. The fact that each sample was a different, separately collected
tissue from a different area of the animal may be enough to account for these

small variations of element concentrations.
4.5.5 Concentrations of the Analytes in Tissue Digests

Table 4.4 contains the concentrations of each analyte in the gunshot

wounds and control tissues analyzed for each day. For the control tissue, Sb

61

remains below the instrument’s limit of detection throughout the 7-day sampling
period. Pb was below the LOD until day 5 when the concentration increased to
30 ng/g. The levels of Pb were very low during days 5, 6, and 7 and are less
than the concentrations observed in the gunshot wounds.

The elevation of 138Ba in the gunshot wound tissue that occurred from
days 4 to 5 was also seen in the control tissue, although at much greater levels.
The gunshot wound tissue 138Ba increased very slightly from 0.06 pg/g to 0.10
pg/g, a total of 40 ng, whereas the control tissue increased from 0 p919 to 0.30
pg/g, a total of 300 ng. Although the increases occur over the same time period,
it cannot be concluded that they are related due to the discrepancy in the
magnitude of elevation between them. Had the increases been more systematic,

it may have been possible to relate the increases to some common source.

Table 4.4 ICPMS data for the gunshot wounds and control tissue collected
during the course of the 7-day sampling period.

 

 

 

 

 

 

 

 

 

 

 

 

Tlssue Elements Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7
so - - - - - - -
“ml”; Ba - - - - 0.30 0.17 0.19
(”9 9 Pb - - - - 0.03 0.003 0.01
Gunshot so 9.19 2,75 0,47 0.51 0.25 0.15 0.15
Wound Ba 30.7 0.60 0.55 0.05 0.10 0.09 0.07
(11919) Pb 25.7 4.07 0.57 0.39 0.50 0.20 0.10

 

 

4.5.6 ICPMS Analysis of Procedural Blanks
The procedural blanks that were analyzed along with the control and GSR

samples did not contain any significant levels of the target analytes. The

62

concentrations contained in the procedural blanks were 0.5 i 1.0, 0.1 :l: 0.2, and
0.2 :l: 0.5 ppb for me, 138Ba, and 208Pb respectively. These concentrations
were much lower than any of the levels encountered in the GSR samples which
for day 7, the lowest values, were 29 ppb for 121Sb, 13 ppb for 1388a, and 20 ppb
for 208Pb. Thus, it can be concluded that the levels of GSR analytes within the
tissue were not significantly affected by the solutions or preparation. Hence,
there was no significant contamination contributing to the GSR, l.e. the levels

observed were truly GSR.

4.6 Summary of ICPMS Results

The results presented in this chapter are the first known application of
ICPMS for analysis of GSR directly from decomposing tissue. There have been
several papers that have discussed the use of ICPMS for the trace metal
analysis of GSR from cotton-tipped swabs, cotton paper, and blowfly larvae as
discussed in Chapter 2 but, according to a recent survey of literature, never for
tissue containing GSR. The most important outcome of this study is that the
results were successful. Preliminary findings illustrated that ICPMS does have
the capabilities to easily detect GSR through the complex digested tissue matrix.
The ICPMS used in these studies had low detection limits, two of which, me
and 208Pb, were lower than previously reported values.“ The precision of the
ICPMS was demonstrated with the gunshot wound samples. A tissue digest was
prepared for 10 separate analyses and a RSD of less than 3% for the 10 sample

injections for each element were obtained.

63

GSR was determined in gunshot wounds throughout the period of 7 days
despite unfavorable weather conditions. The concentration of the initial digest
solution from the gunshot wound of day 7 was 29, 13, and 20 ppb for me,
138Ba, and 208Pb respectively, all of which were greater than the detection limits
of the instrument. The solutions analyzed, however, were diluted from the initial
tissue digest by a factor of 43.3. If no GSR was detected during the later stages
of the experiment, then a less dilute sample could easily be prepared for
analysis.

The levels that were detected during the later stages (days 5-7) of
decomposition were very low ranging from 0.61 pg/g to 0.15 pg/g for me, 0.06
pg/g to 0.10 pg/g for 138Ba, and from 0.39 pg/g to 0.10 pg/g for 208Pb in tissue.
The control tissues from the 7-day period had undetectable levels of 121Sb, 138Ba,
and 2”Pb for the first four days. On day 5, there was a spike in the levels of
2°an and 138Ba, and they remained at elevated fluctuating levels for the
remainder of the study. Currently, there is no known explanation for this
elevation in analyte concentrations. The 121Sb, however did not become
elevated throughout the course of this study. Additional control tissue would
need to be analyzed to determine the origin of the 138Ba and 208Pb levels. The
ICPMS has been shown as a technique with great potential for analysis of GSR

even at trace levels from complex sources.

64

Chapter 5

Chapter 5
Conclusions and Further Work

5.1 Conclusions

This preliminary project was designed to lay the foundation for future work
in GSR research. The research addresses a particular issue within forensic
analysis; that is the identification or detection of a gunshot wound or gunshot
residue on decomposing tissue. The project aimed to detect and analyze GSR
from decomposing tissue within two different domains; morphology and
elemental composition of individual particles with SEM/EDS and also elemental
composition of bulk samples with ICPMS. Both avenues of this research project
had successful results.

Air-drying and freeze-drying gunshot wounds for direct SEM/EDS analysis
of GSR was attempted. SEM/EDS was found to be of limited use for GSR
analysis directly from tissue samples. Tissues imaged under the SEM were
easily damaged by the electron beam and were of poor quality, making the
location and analysis of GSR particles difficult. GSR was also obscured by the
large amount of hair and debris on the samples. The preparation of samples was
also labor intensive and time-consuming, beginning with the removal of adipose
tissue with a scalpel to that of drying the sample by either of the methods. Both

air- and freeze-drying required greater than 48 hours.

65

Tape lifting was determined to be the most effective method of sample
preparation for GSR analysis in terms of both time and ease of preparation.
Tape lifts from the gunshot wounds and control tissues were analyzed by
SEM/EDS and GSR was found to be dense on wounds from days 1 and 2. The
GSR particles that were located and analyzed had a slightly deformed spherical
shape with a diameter ranging from 0.5 — 1 pm and had elemental compositions
of Sb, Ba, and Pb. On days 3—7, no distinct GSR containing significant levels of
Sb, Ba, and Pb were located. However, on days 5 and 6, spherical particles with
no observed deformation were found. The particles were analyzed by EDS and
had an elemental composition of Fe-Cr—Nl, which is consistent with stainless
steel possibly from the barrel of the gun.

For ICPMS, the samples were fully digested in nitric acid, diluted, and
analyzed by ICPMS. The ICPMS was shown to have a linear response between
0 and 1000 ppb for 121Sb, 138Ba, and 208Pb. Precision of the instrument was
demonstrated with less than a 3% RSD for all three elements over the course of
10 sequential replicate injections. The limits of detection for the target elements
were 0.002 ppb for Sb, 0.033 ppb for Ba, and 0.009 ppb for Pb. The limits of
detection for Sb and Pb were 1-2 orders of magnitude lower than those reported
for GSR detection by Koons.44

GSR was detected in the gunshot wound tissues collected throughout the
7-day period and was determined as the elemental components, me, 138Ba,
and 208Pb. 121Sb and 208Pb showed a significant decrease in concentration over

the course of days 1-3; 121Sb dropped from an initial concentration of 9.2 pg/g of

66

tissue to 0.5 pg/g and 208Pb dropped from an initial concentration of 26.7 pg/g to
0.9 pg/g. The GSR levels found in the gunshot wounds, even at the end of the 7-
day period, were well above the limits of detection for each of the elements. A
more abrupt decrease in concentration was observed for 138Ba. Over the course
of days 1-2, the concentration of 138Ba dropped from 30.7 pg/g to 0.7 pg/g.

Control tissues were also analyzed by ICPMS and for days 14, none of
the GSR elements were detected. On day 5, there was a spike in the
concentrations of 138Ba and 208Pb in the control tissues. 138Ba increased to 0.3
pg/g and 208Pb increased to 0.03 pg/g, their concentrations then fluctuated for the
following 2 days. There was no explanation for these observed increases at the
current time.

Lastly, the weather was seen to have an effect on the levels of GSR
present on the wounds by SEM/EDS. SEM/EDS analysis showed that after the
first rainstorm, during the night of day 2, there were no detectable GSR particles.
ICPMS showed significant decreases in GSR but not anything directly
attributable to the weather experienced throughout the course of the week. The
rapid decrease of GSR observed by SEM/EDS between days 2 and 3 was also
seen by ICPMS in the form of an abrupt decrease in the GSR elements.
Although no GSR was found by SEM/EDS after day 2, ICPMS was able to

determine the presence of GSR throughout the entire 7-day sampling period.

67

5.2 Further Work

There are many opportunities within the framework of this research for
new projects. The first and most obvious project would be expanding the ICPMS
analyses for a larger sample size to allow for statistical conclusions to be drawn.
Expanding the sampling period would also be interesting since after 7 days, GSR
was still determined in gunshot wounds collected throughout this study. The
tissue digests analyzed in this study were dilute and the analytes could still be
detected, analysis of undiluted samples should allow for the extension of the
sampling period. Another study assessing the effect of GSR deposition when the
barrel of the firearm is not cleaned between shots could be completed. During
this study, the firearm was cleaned between shots. Hence, a comparison of the
GSR concentrations determined herein with the concentrations from a handgun
that was not cleaned could be made. The study would determine if GSR levels
increase linearly with each shot and whether the firing order could be determined
by the relative levels of GSR and/or if one or more elements increases more so
than another.

Comparison of different firearms and ammunition as well as varying the
distance of firing would also be extremely interesting studies. The use of
different firearms and ammunition would provide an assessment of whether trace
elements originating from the gun barrel, cartridge, or other components could be
detected, possibly allowing for firearm and ammunition differentiation or
elimination in a crime. Varying the distance from which the firearm was

discharged from the target would provide elaboration of Santos’ work with ICPMS

68

and GSR.5 Santos’ study marked the start of the development of a statistical
GSR versus distance model, but certainly not the end. In addition to Sb, Ba, and
Pb, other trace elements present in the GSR could also be determined, such as
Al, Zn, Si, or others described in literature.7' 27 The use of trace elements could
provide another dimension to GSR analysis, such as determining the
manufacturer of the ammunition or even the propellant type or whether the bullet
was jacketed or not. All of these projects would highlight a particular area in GSR
analysis, which up until this point, had not been researched thoroughly.

The SEM/EDS analysis produced similar conclusions to previous
publications in which tape lifting was demonstrated as a superior method of
sample collection for this type of analysis.22 This research has demonstrated that
ICPMS is a valuable tool for GSR analysis, it is sensitive, has a wide linear
dynamic range and good response stability. This study is the first known
application of ICPMS for the determination of GSR in decomposing tissue. The
findings from this project provide a foundation from which improvements can be

made.

69

1. Justice, U. S. D. 0. Bureau of Justice Statistics, Firearms and Crime
Statistics. http://wwwoip.usdoi.gov/bis/quns.htm (March 07),

2. Violence, T. C. t. S. G. CSGV Briefing Book. (March 17),

3. Collins, K. A.; Lantz, P. E., Interpretation of Fatal, Multiple and Exiting
Gunshot Wounds by Trauma Specialists. Joumal of Forensic Sciences 1994, 39,
(1), 94-99.

4. Denton, J. S.; Segovia, A.; Filkins, J. A., Practical Pathology of Gunshot
Wounds. Archives of Pathology Laboratory Medicine 2006, 130, 1283-1289.

5. Santos, A. S.; Magalhaes, T.; Nuno Vieira, D.; Almelda, A. A.; Sousa, A.
V., Firing Distance Estimation Through the Analysis of the Gunshot Residue
Deposit Pattern Around the Bullet Entrance Hole by ICPMS. The American
Journal of Forensic Medicine and Pathology 2007, 28, (1 ), 24-30.

6. Hussain, G.; Rees, G. J., Combustion of Black Powder. Part 1.
Propel/ants, Explosives, Pyrotechnics 1990, 15, (2), 43-47.

7. Meng, H.; Caddy, B., Gunshot Residue Analysis. Journal of Forensic
Sciences 1997, 42, (4), 553-570.

8. Romolo, F. S.; Margot, P., Identification of Gunshot Residue: A Critical
Review. Forensic Science lntemational 2000, 1 19, 195—21 1.

9. Basu, 8., Formation of GSR. Journal of Forensic Sciences 1982, 27, (1),
72-91.

10. Di Maio, V. J. M., Gunshot Wounds: Practical Aspects of Firearms,
Ballistics, and Forensic Techniques. CRC Press: Boca Raton, 1993; p 331.

11. Burrard, G., The Identification of Firearms and Forensic Ballistics. 2nd ed.;
Wolfe Publishing Company: London, 1951; p 217.

12. Doyle, J. S. Gunshot Residue.
http://wwwfirearmlecom/A DistanceGSRhtm (March 07),

13. Lebiedzik, J.; Johnson, D. L., Handguns and Ammunitions Indicators
Extracted from the GSR Analysis. Joumal of Forensic Sciences 2002, 47, (3),
483-493.

14. Wolten, G. M.; Nesbitt, R. 8., On the Mechanism of GSR Particle
Formation. Joumal of Forensic Sciences 1980, 25, (3), 533-545.

70

15. Torre, C.; Mattutino, G.; Vasino, V.; Robino, 0., Brake Linings: A Source
of Non-GSR Particles Containing Lead, Barium and Antimony. Journal of
Forensic Sciences 2002, 47, (3), 494-504.

16. Zeichner, A., Recent Developments in Methods of Chemical Analysis in
Investigations of F irearm-related Events. Analytical and Bioanalytical Chemistry
2003, 376, 1178-1191.

17. Morales, E. B.; Revila Vazquez, A. L., Simultaneous Determination of
Inorganic and Organic GSR by CE. Journal of Chromatography A 2004, 1061,
225-233.

18. Tschirhart, D. L.; Noguchi, T. T.; Klatt, E. C., A Simple Hitochemical
Technique for the Identification of GSR. Journal of Forensic Sciences 1991, 36,
(2), 543-547.

19. D'Uffizi, M.; Falso, G.; lngo, G. M.; Padeletti, G., Microchemical and
Micromorphological Features of GSR Observed by Combined Use of AFM, SA-
XPS, and SEM/EDS. Surface and Interface Analysis 2002, 34, 502-506.

20. Woolever, C. A.; Starkey, D. E.; Dewald, H. D., DP-ASV of Lead and
Antimony in GSR. Forensic Science lntemational 1999, 102, 45-50.

21. Stein, K. M.; Bahner, M. L.; Merkel, J.; Ain, 8., Detection of GSR in
Routine CTs. lntemational Journal of Legal Medicine 2000, 114, 15-18.

22. DeGaetano, D.; Siegel, J. A.; Klomparens, K. L., A Comparison of Three
Techniques for Sampling and Analysis of GSR by SEM/EDX Analysis. Joumal of
Forensic Sciences 1992, 37, (1), 281-300.

23. Zeichner, A.; Ehrlich, S.; Shoshani, E.; Halicz, L., Application of Lead
Isotope Analysis in Shooting Incident Investigations. Forensic Science
lntemational 2005.

24. Ravreby, M., Analysis of Long-Range Bullet Entrance Holes by AAS with
SEM. Journal of Forensic Sciences 1982, 27, (1), 92-112.

25. Ruch, R. R.; Buchanan, J. D.; Guinn, V. P.; Bellanca, S. C.; Pinker, R. H.,
NAA in Scientific Crime Detection. Journal of Forensic Sciences 1964, 9, 119-
132.

26. Koons, R. D.; Havekost, D. G.; Peters, C. A., Analysis of GSR Primer
Residue Using FAAS: A Reexamination of Extractions and Instrument
Procedures. Journal of Forensic Sciences 1987, 32, (4), 846-865.

27. Romolo, F. S.; Margot, P., Identification of Gunshot Residue: A Critical
Review. Forensic Science lntemational 2001 , 119, 195-21 1.

71

28. JEOL, L. SEM Products Info. http://www.jeol.com/Default.aspx?tabid=102
(March 21),

29. Tillman, W. L., Automated GSR Particle Search and Characterization.
Journal of Forensic Sciences 1987, 32, 62-71.

30. White, R. S.; Owens, A. D., Automation of GSR Detection and Analysis by
SEM/EDS. Joumal of Forensic Sciences 1987, 32, 1595-1603.

31. Roeterdink, E. M.; Dadour, I. R.; Watling, R. J., Extraction of GSR from the
Larvae of the Forensically Important Blowfly. lntemational Journal of Legal
Medicine 2004, 118, 63-70.

32. Breton, B. C., The Early History and Development of the SEM. In
Cambridge, 2006; Vol. 2007.

33. Flegler, S. L.; Heckman, J. W.; Klomparens, K. L., Scanning and
Transmission Electron Microscopy: An Introduction. Oxford University Press:
New York, 1993; p 225.

34. American Society for Testing and Materials, I., Standard Guide for GSR
Analysis by SEM/EDS. In ASTM E1588-95, ASTM: West Conshohocken, 2000;
VOL 14.02: pp 657-659.

35. Boehm, A., Application of the SEM in Forensic Medicine. Scanning
Electron Microscopy 1971, 62, 553-560.

36. Kage, 8.; Kudo, K.; Kaizoji, A.; Ryumoto, J.; Ikeda, H.; Ikeda, M, A Simple
Method for Detection of GSR Particles from Hands, Hair, Face ad Clothing Using
SEMNVDX. Joumal of Forensic Sciences 2001, 46, (4), 830-834.

37. Beauchemin, D., ICPMS. Analytical Chemistry 2006, J Pages, J.

38. McCurdy, E.; Potter, D., ICPMS. In Atomic Spectroscopy in Elemental
Analysis, Cullen, M., Ed. Blackwell Publishing Ltd.: Boca Raton, 2004.

39. Skoog, D. A.; Holler, F. J.; Nieman, T. A., Principles of Instrumental
Analysis. 5 ed.; Books/Cole Thomson Learning: Crawfordsville, 1998.

40. Olesik, J. W., Elemental Analysis Using ICP-DES and ICP/MS. Analytical
Chemistry 1991, 63, (1), 12A-21A.

41. Meyer, G. A., ICP. Analytical Chemistry 1987, 59, (23), 1345A-1354A.

42. Ebdon, L.; Evans, E. H.; Flsher, A.; Hill, S. J., Analytical Atomic
Spectrometry. John VWey and Sons, Ltd.: West Sussex, 1998.

43. Cottingham, K., Analytical Chemistry 2004, 35A-38A.

72

44. Koons, R. 0., Analysis of Gunshot Primer Residue Collection Swabs by
ICPMS. Journal of Forensic Sciences 1998, 43, (4), 748-754.

45. Rels, E. L.; Sousa Sarkis, J. E.; Neto, O. N.; Rodriguez, C.; Kakazu, M. H.;
Viebig, S., A New Method for Collection and Identification of GSR from the
Hands of Shooters. Journal of Forensic Sciences 2003, 48, (6), 1269-1274.

46. Souza Sarkis, J. E.; Neto, O. N.; Viebig, S.; Durrant, S. F., Measurements
of GSR by Sector Field ICPMS - Further Studies with Pistols. Forensic Science
lntemational 2007, In press.

47. Singer, R. L.; Davis, D.; Houck, M., A Survey of GSR Analysis. Journal of
Forensic Sciences 1996, 41, (2), 195-198.

48. Haag, M.; Wolberg, G., Scientific Examination and Comparison of Skin
Simulants for Distance Determinations. The Journal of the Association of Firearm
and Toolmark Examiners 2000, 32, (2), 136-142.

49. Megyesi, M. The Effects of Temperature on the Decomposition Rate of
Human Remains. University of Indianapolis, Indianapolis, 2001.

50. Archer, M., Rainfall and temperature effects on the decomposition rate of
exposed neonatal remains. Science and Justice: Joumal of the Forensic Science
Society 2004, 44, (1), 35-41.

51. Miyauchi, H.; Kumihashi, M.; Shibayama, T., The Contribution of Trace
Elements from Smokeless Powder to Post Firing Residues. Journal of Forensic
Sciences 1998, 43, 90-96.

73

   

lllllllllllllllllllllllll
3 1293 02956 1606