m“ "2%,: V>‘1::2-~ ‘
”.3514: .1} : 11%

,3 .
., 51.34: {‘J
w'r'

’ ..--‘-v-’
,..‘ .

- fl .

' ‘ u‘r‘ Ha’

$52: C‘I'NLL-
w n

P y:
f . ,,. “"1755...

"P" L... 71" 1 ‘ "‘\ A. "'3,“
"9"}??sz
x; :"MI-‘ggiu,

r ' .,:‘-
..":.r.*1.' . ‘ . L32}: “5:31:
.1 m :51...“ ,,.‘., ' . 531:; ,3
..JM ; v- 1 ' .. b' *‘vw‘
.,' Myrna: A. v ‘ ”I... J“; :J '
.31.? "(-3 "‘v ‘.... 4-1 ‘. ‘ 3'. ,1".-.
3": Jar , ' ‘ w ' ‘3 ,3}:
‘C , .NW
6 I“ ' " 1'23”

.- ,m“

 

 

<u» .L'L'n, "A- J;

;‘:‘N‘:'¢‘J ‘ .-
4.7: . , "V3”!- .wf:v~;b;f‘
«xi-7,3"

 

L'mm’WQY 5 lbOL‘ 3050

Michigan mate
5 55555555 555 555 55555555555555555555

University 555
A 3 1293 00574 0000-—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled

A COVIPARISON OF THREE TECHNIQUES DEVELOPED FOR

SAMPLING AND ANALYSIS OF GUNSHOT RESIDUE BY

SCANNING ELECTRON MICROSCOPY AND ENERGY DISPERSIVE
X-RAY ANALYSIS

presented by

Douglas Hall DeGaetano

has been accepted towards fulfillment
of the requirements for

Master of_S;j._eng_e___degree in Criminal Justice

QOMCISWL

Uiajor professor

Date 10/35 A?

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES Mun on or More due due.

DATE DUE DATE DUE DATE DUE

 

 

 

W5

 

   

 

 

105L115 5.9.25.5

C"; 5

 

f firs
“cue/5
e ‘-'

-o—‘ 0-- .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L____

 

 

 

 

'5 .1
5

=_7

 

 

'_____—..,- —— »- -

MSU Is An Aflirmdlve Action/Equal Opportunity Institution

 

A COMPARISON OF THREE TECHNIQUES DEVELOPED FOR SAMPLING AND
ANALYSIS OF GUNSHOT RESIDUE BY SCANNING ELECTRON MICROSCOPY

AND ENERGY DISPERSIVE X-RAY ANALYSIS

BY
Douglas Hall DeGaetano

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

College of Social Science
School of Criminal Justice

1989

9005553

ABSTRACT

A COMPARISON OF THREE TECHNIQUES DEVELOPED FOR SAMPLING AND
ANALYSIS OF GUNSHOT RESIDUE BY SCANNING ELECTRON MICROSCOPY
AND ENERGY DISPERSIVE X-RAY ANALYSIS

BY
Douglas Hall DeGaetano

The purpose of this study was to compare three gunshot
residue (GSR) collection methods from hand samples by
SEM/EDX analysis. The methods were: the tape lift, glue
lift and concentration techniques. Efficiency of particle
collection was examined under various conditions including:
number of rounds fired, temperature and shelf life. The
tape lift surface demonstrated excellent particle collection
ability and remained stable for all conditions tested. The
glue lift was a relatively inefficient collection surface
under all conditions tested. Collection followed by
concentration gave highly variable results. An unusually
large amount of GSR was found on the hand from working the
action of a cleaned weapon. In addition, a survey of U.S.
Forensic Crime Laboratories was undertaken to determine
methods of GSR analysis. The two general types of GSR
analysis were compared and contrasted. Problems encountered

by analysts using SEM/EDX were discussed.

DEDICATION

To the little things - "Just because you can't see 'em don't
mean they ain't there"

Anonymous drunk, Downtown Detroit, 1973

ACKNOWLEDGEMENT

I gratefully acknowledge the cooperation of all
Forensic Science Laboratories participating in the survey.
This project was partially funded by a grant from Michigan
State University's All University Research Grant Initiation
Program and The MSU Center for Electron Optics. The author
gratefully acknowledges this support.

In addition I would like to acknowledge the patience,
guidance and generosity of Dr. Karen Klomparens, as well as,
the darkroom expertise of Chad Brinkerhuff. The
cooperation, patience and technical advise of the Michigan
State Police Crime Laboratory in East Lansing Michigan is
also gratefully acknowledged. My personal thanks to Lt.
Roger Bolhouse, Lt. Dave Townsend, Sgt. Millard Holton and
Sgt. Bob Cilwa for their patience and technical assistance.
Finally, my thanks to Dr. Jay Siegel for his unfailing

confidence and support.

ii

List of Tables............................................iv
List of Figures........................................... v
Introduction.............................................. 1
Literature review......................................... 7
Chapter one...............................................15
Introduction.........................................16
Materials and methods................................18
Results and discussion...............................19
Summary and conclusions..............................31
Chapter two...............................................32
Introduction.........................................33
Materials and methods................................34
Results..............................................45
Discussion...........................................53
Summary and conclusions..............................71
Chapter three.............................................77
Introduction.........................................78
Materials and methods................................80
Results and discussion...............................82
Summary and conclusions..............................84
List of references........................................86
Appendix A................................................91
Appendix B................................................98
Appendix C............. ..... ..............................99

iii

10.

11.

Ll§I_QE_IA§LE§

Percent of forensic laboratories analyzing gunshot
residue and method used for analysis................20

Frequency of challenge to GSR analysis in court.....24
Grounds for challenge in court......................25
Laboratory SEM equipment, age and dependability.....26

Statistical analysis of the ln (x+1) of particle per
hour means between collection methods ..... ..........46

Statistical analysis of the ln (x+1) of particle per
hour means for tape lifts by SNK procedure using
general linear models...................... ..... ....47

Statistical analysis of the ln (x+1) of particles per
hour means for concentration method by SNK
procedure using general linear models...............48

Statistical analysis of the ln (x+1) of particles per
hour means for glue lifts by SNK procedure using
general linear models...............................50

Statistical analysis of 1n (x+1) of particles per hour
means for glue lifts followed by tape lifts using
Ttests (LSD)OOOOOOOOOOOOOOOOOOOOO0.0.0.00000000000052

Observed advantages and disadvantages of GSR
collection methods tested...........................72

Comparison of GSR found on the hand between three live

rounds fired and "dry firing" an unloaded, cleaned
weapon three times..................................83

iv

10.

11.

12.

13.

LI§I_QE_EI§HBE§

Time required per analysis for labs analyzing
gunsnot reSidUBOOOOOOOOOOOOOOOOOOOOOOOOOOO0.00.0.00022

Analysis time in hours per week for labs
conducting GSR analysis.............................23

Variations in the minimum number of particles
analyzed by SEM/EDX to confirm GSR..................28

Gunshot residue collection from the hands of a
ShooterOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOCOO...0.00.35

Location and identification of GSR by SEM/EDX.. ..... 39
Identification of contaminating elements on a

filter surface used in the concentration
teChniqueeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee.055

Identification of contaminating elements in the
subfilter used in the concentration technique.......57

Varying amounts of debris found on filter surface
after the concentration technique...................59

Identification of contaminating tungsten particles
onacarbon planChetOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOGZ

Electron micrographs of small GSR particles on
various surfaces....................................66

Electron micrographs of large GSR particles on
various surfaceSOOOOOOOOOOOOOOOOOOOOOOCOOOOOOOOOO00.67

Effect of insufficient carbon coating on a tape
lift surfaceOOOOOOOOOOOOOOOOOOOOOO0.0.0.0.000000000068

Gun cleaning patches from the redesigned gun control
experimentOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.0.81

A jury's verdict of innocence or guilt in a homicide
case may depend on the accuracy and efficiency of gunshot
residue collection and analysis. Gunshot residue is a form
of transfer or trace evidence. This material can be thought
of as loosely complying with the famous Locard exchange
principle, derived by the French forensic scientist Edmond
Locard in 1910, which states that anytime there is contact
between two surfaces, there will be a mutual exchange of
matter across the contact boundary. Strictly speaking
gunshot residue particles are an example of transfer without
direct contact. The reason for collecting and analyzing
these particles is obvious and is the basis for the study of
trace evidence in forensic science. As was stated quite
well by Dr. Paul L. Kirk of the University of California,
"It is virtually impossible for a criminal to commit a crime
without leaving evidence behind and carrying evidence away
with him." (DeForest et al. 1983).

Gunshot residue can be defined as material composed
mainly of barium (Ba), antimony (Sb) and lead (Pb) combined
to form microscopic spheroid particles, left behind from the
primer after discharging a firearm that uses smokeless
powder. This material can be collected from the firing hand
of a shooter and analyzed by various techniques. The

identification of GSR is used frequently in homicide and

2
suicide cases to demonstrate if the suspect or victim had
recently discharged a firearm (Basu et al. 1984). To date,
the most definitive method for characterization of gunshot
residue (GSR) is by using scanning electron microscopy (SEM)
with energy dispersive X-ray analysis (EDX) capability.

The method using SEM-EDX, has several advantages over
other current techniques in the analysis of GSR. Unlike
neutron activation analysis (NAA) or flameless atomic
absorption spectroscopy (FAAS), which are destructive bulk
elemental techniques, SEM-EDX is a non-destructive technique
which can be applied to a specific particle of GSR. By
analyzing distinct particles of GSR, the technique is not
subject to spurious results from background or contaminating
levels of lead, barium, and antimony because these elements
do not combine to form spheroid particles in the environment
(Wolten et al. 1979b).

While extensive work has been done demonstrating the
ability of SEM-EDX to identify and characterize gunshot
residue, there is currently no literature describing how
frequently this application of SEM-EDX is being used in
Crime Labs throughout the United States. Therefore, a mail
survey of Crime Labs across the country was conducted. The
first chapter of this thesis addresses the survey results.

The objective of this survey was three-fold:

1) To determine what type of analysis is being performed

most frequently on GSR in the laboratory.

3
2) To establish the average age and dependability of the
scanning electron microscopes being used in Crime Labs
across the country.
3) To document collection and analysis techniques being used
for GSR by SEM-EDX, in the interest of identifying and
resolving common problems for the investigator using this
technique.

The second chapter of this thesis is a comparison of
three techniques used to collect and analyze gunshot residue
by SEM/EDX.

While SEM-EDX is conclusive in describing the
morphology and elemental composition of the particles in
question, it suffers from the excessive time (up to several
hours) required to perform the analysis. Several
investigators have studied this problem (Basu and Ferries
1980: Nesbitt et al. 1976: Ward 1982). An ever increasing
caseload in the laboratory demands a collection technique
that is rapid, dependable and efficient.

Analysis time can be reduced by:

A) Increasing sample concentration, Such a technique
was developed by D.C. Ward for concentrating GSR from a
Vistanex adhesive coated surface by repetitive centrifugal
concentration through a high density liquid.

B) ncreasin eff c enc o a ic e ft e
hang; The tape lift method is said to have excellent

particle lifting ability and has been employed by many

4
researchers (Andrasko and Maehly 1977: Goleb and Midkiff
1975: Matricardi and Kilty 1977: Nesbitt et a1. 1976; Tassa
et al. 1982b: Wolten et al. 1979b). This technique employs
the use of sticky tape (cellotape or 3M Scotch Brand 465)
either dabbed against the hand by a police officer wearing
rubber gloves or with tape mounted on an aluminum stub (25mm
or 5mm in diameter) which is dabbed on the shooter's hand.
The stub is sent to the lab where it is coated with a
conductive layer of carbon and examined with a scanning
electron microscope. An alternative to tape is the glue
lift method developed by S. Basu and S. Ferriss (Basu and
Ferriss 1980: Basu 1982: Basu et al. 1984). This method
uses rubber cement that has been diluted with toluene or
1,1,1,-trichloroethane. A glue layer is applied to a carbon
planchet (diameter 1/2") attached to an aluminum pin-type
mount for easy handling. Such glue-coated discs are kept in
storage boxes until a suspect's hand is to be sampled. The
hand is then dabbed three to five times with the sampling
disc. The sample is then ready to be examined in the
electron microscope, as this method does not require coating
the sample with carbon.

C) Decreasing ghe agea go be scanned, Ward scans a 2mm
area (Ward 1982), Basu and Ferriss scan several 1.5mm areas
(Basu and Ferriss 1980). A general reduction in stub size
from 25mm to 5mm has also been employed.

Chapter two of this thesis is a comparative study of

5
the £2D2§D§IQ§123_£§QDDIQQ§. described by Ward (1982) with
modifications by Sugarman (Appendix A): the glae_li;§
:eehnigae, described by Basu and Ferriss (1980): and the
Sé2§_11£L_L§QDDng§. as described in the paper by Nesbitt,
et al. (1976).

While several methods have been developed for
collection of GSR, no thorough investigation has ever been
made comparing several of the most common methods for GSR
collection and analysis by SEM-EDX. The results of such an
investigation should allow crime labs to select a collection
technique for GSR based on that technique's relative
strengths and weaknesses. Currently, each laboratory must
decide somewhat arbitrarily which collection method it will
use for GSR analysis by SEM-EDX.

The objectives of this portion of the thesis are to
determine which of the three collection techniques tested
has:

1) The highest efficiency for lifting particles from the
hand.

2) The minimum analysis time.

3) The most stable adhesive surface with respect to
temperature and deterioration over time.

Chapter three of this thesis describes the finding of
gunshot residue on the hand after working the trigger
mechanism of an unloaded, freshly cleaned handgun.

Experimental conditions and implications of this observation

are discussed.

The oldest method for determining "gunshot" residue on
the hand was introduced in the U.S. in 1933 by Teodoro
Gonzales (Cowan and Purdon 1967). It is known by various
names including: the Gonzales test, the paraffin test and
the dermal nitrate test. It was used to detect nitrates and
nitrites left behind on the hand after firing a weapon. A
molten wax cast was made of the suspect's hand. This
material was then reacted with diphenylamine and
diphenylbenzidine which gave pinpoint color reactions with
nitrates and nitrites. The technique suffered from several
problems including false positives with substances other
than nitrates and nitrites, and the presence of nitrates and
nitrites on the hands of people who had not fired a weapon.
The FBI advised against the use of this test in 1935 (FBI
Law Enforcement Bulletin 1935), and again in 1940 (FBI Law
Enforcement Bulletin 1940). Despite these warnings, the use
of the "paraffin test" continued through the mid 1960's. In
1967, the paraffin test was further discredited in a
controlled study which showed no significant difference from
paraffin tests done on people that had fired weapons and a
"control" group that had not presumably fired a weapon
(Cowan and Purdon 1967).

It is interesting to note that Professor Ralph Turner,

the former director of the Forensic Science program at MSU

8
worked on the specificity of the paraffin test as a graduate
student (personal communication). He had often wondered if
the GSR particles composed of lead, barium and antimony also
had nitrates or nitrites associated with their surface,
giving rise to the pinpoint color reactions in the paraffin
test. In regards to this question, an investigator using
Auger electron spectroscopy detected no nitrogen associated
with GSR particles (Hellmiss et al. 1987). Those results
were corroborated on GSR analyzed using a Cameca Microprobe
unit at the University of Michigan, by the author.

Other chemical tests developed to detect gunshot
residue were reported in 1959 (Harrison and Gilroy 1959).
These methods used sodium rhodizonate for detection of lead
and barium and a solution of triphenylmethylarsonium iodide
for the detection of antimony. These tests were specific
for elements associated with cartridge primers but lacked
the sensitivity required to detect small amounts of GSR on
the hand. The sodium rhodizonate test is still used quite
effectively today to demonstrate lead particles on clothing
for firing distance determinations.

Sophisticated instruments capable of detecting
microgram quantities of specific elements were employed in
an attempt to detect GSR. This type of analysis may be
broadly termed "bulk" elemental analysis. A great deal of
work was done using neutron activation analysis (NAA) for

detection of barium and antimony (Kilty 1975: Kinard and

9
Lundy 1975; McFarland and McLain 1973: Pillay et al. 1974:
Rudzitis and Wahlgren 1975: Rudzitis et al. 1973: Rudzitis
1980). NAA has been used quite frequently in the detection
of GSR since the early 1970's. While NAA is a very
sensitive technique, it suffers from a variety of problems.
Among these are: 1) lengthy analysis and processing times
2) high cost 3) inability to detect lead 4) requires a
nuclear reactor 5) interpretation - must know "threshold"
values of barium and antimony on the hands of the general
population, otherwise results in false positives 6)
specificity - high levels of barium and antimony may be from
other sources other than GSR and 7) must handle
radioactive samples.

In the late 1970's, flameless atomic absorption
spectroscopy (FAAS), another type of bulk elemental
analysis, was used to detect gunshot residue (Goleb and
Midkiff 1975: Kinard and Midkiff 1978: Koons et al. 1988:
Newton 1981; Portis and Tilley 1981). The advantage of this
technique over NAA was that trace amounts of lead on the
hand could also be measured. Additional advantages were
shorter processing and analysis times of nonradioactive
samples and a relatively low cost of equipment. However,
this technique still suffered from the same threshold
interpretation and specificity problems as in NAA. A
variety of sources exist other than GSR that could cause

deposition of elements in question on the hand. Common

10
sources of lead include (Wolten et al. 1979b): leaded
gasoline, old paint and printing inks. Sources of barium
are: greases, lubrication oil additives, and extenders for
paint and rubber. Sources of antimony include: children's
cap guns, and paint (Krishnan 1976).

Other techniques, which have occasionally been used for
GSR analysis, include: photoluminescence (Loper et al.
1981), inductively coupled plasma-atomic emission
spectroscopy (ICP-AES) (Koons et al. 1988), Auger
spectroscopy (Hellmiss 1987), X-ray diffraction (Tassa et
al. 1982a) and anodic stripping voltammetry (ASV) (Brihaye
et al. 1982).

In the mid 1970's, it was discovered that gunshot
residue particles could be visualized by the scanning
electron microscope (SEM) (Andrasko and Maehly 1977: Nesbitt
et al. 1976: Wolten et al. 1977). GSR particles were found
to be generally spheroidal in shape. The use of energy
dispersive X-ray analysis allowed the elemental composition
of the particles to be determined.

Gunshot residue particles are composed of a variety of
elements whose sources are consistent with the primer,
bullet, cartridge case and barrel of the gun (Basu 1982:
Bergman et al. 1988: Wolten and Nesbitt 1980). An extensive
report was made by the Aerospace Corporation (Wolten et al.
1977), that described, in essentially a tutorial fashion,

how to optimize the operational variables of the SEM in

11
order to obtain a good backscatter image of GSR particles.
The report goes on to examine various types of ammunition
used and weapons fired in an attempt to characterize the
type of GSR found in test firings.

Other researchers investigated the formation of GSR and
the location of the elements associated with GSR in the
spheroid particles (Basu 1982: Burnett 1989: Tassa and
Zeldes 1979: Tassa et al. 1982a). The spheroidal shape of
GSR is assumed to be caused by the condensation of elements
in the primer from a molten state when a gun is fired. In
general, particles of ten microns or larger are composed
mainly of barium. In the smaller particles, the elements
lead, barium and antimony are mixed throughout the particle
unless the particle was nodulated. In the case of a
particle with nodules, the nodules are typically composed of
lead and sometimes antimony. The body of the particle is
composed of barium with occasional antimony present. The
data were obtained by X-ray mapping individual particles
either whole or in cross section.

The major advantage of using SEM/EDX to analyze GSR is
the ability to locate single particles with unique elemental
composition (Ba, Pb, and Sb) indicative of GSR. Studies
were made that demonstrated the unique elemental composition
of GSR compared to particles in the environment or on the
hands of people in occupations that could potentially

produce particles similar to GSR (Nesbitt et al. 1976:

12
Wolten et al. 1979b).

A great deal of the literature on the analysis of GSR
by SEM/EDX describes various types of ammunition used and
weapons fired to determine the amount and type of GSR
produced (Andrasko and Maehly 1977: Matricardi and Kilty
1977: Taylor et al. 1979: Wolten et al. 1977: Wolten et al.
1979a). Ideally, one would like to be able to match the
type of ammunition used to the type of GSR particles found
on the hand of a shooter. Unfortunately, this is not
usually possible, at least with ammunition found in the U.S.

There have been a number of papers published describing
the use of SEM/EDX in analysis of GSR in actual casework
(Basu et a1. 1984: Krishnan 1976: Bergman et al. 1988).
While in most instances the type of GSR left on the hand can
not be associated with the type of ammunition used,
particles of the bullet left behind along the "wound track"
can sometimes be used to identify the type of ammunition
used and, thereby, the particular weapon used (Taylor et al.
1979).

Over time a variety of collection and/or concentration
techniques have been developed for analysis of GSR by
SEM/EDX (Andrasko and Maehly 1977: Basu and Ferriss 1980:
Gansau and Becker 1982: Matricardi and Kilty 1977: Portis
and Tilley 1981: Sild and Pausak 1979: Tassa et al. 1982b:
Wallace and Keeley 1979: Ward 1982). There are three

general types of collection devices: 1) tape lift

l3
(Andrasko and Maehly 1977: Gansau and Becker 1982:
Matricardi and Kilty 1977: Tassa et al. 1982b) 2) glue
lift (Basu and Ferriss 1984) and 3) concentration (Portis
and Tilley 1981: Sild and Pausak 1979: Wallace and Keeley
1979: Ward 1982).

Among the tape lift techniques, a variety of different
types of tape have been tested (Andrasko and Maehly 1977).
The type of tape lift cited most often in the literature is
3M type 465 adhesive transfer tape.

The glue lift collector consists of a thin layer of
diluted rubber cement on a polished carbon planchet.

Concentration methods vary: one employs suction of
particles directly onto an adhesive coated disc (Sild and
Pausak 1979), another allows examination of GSR by both FAA
and SEM/EDX (Portis and Tilley 1981). In this case, the
sample is collected from the hand using isopropanol swabs
and then concentrated onto a filter using a 5 m1 syringe.
The other two concentration methods mentioned used a diluted
Vistanex (glue) coated stub to lift the particles off the
hand. Particles were then dissolved out of the Vistanex and
passed through the concentration device. In the one case,
concentration was achieved by suction filtration (Wallace
and Keeley 1979): in the other case by centrifugation (Ward
1982). The centrifugation procedure also allowed for
addition of high density solvents to float off contaminating

debris (Epidermal cells, hair, fibers etc.).

14

A major disadvantage of SEM/EDX analysis is lengthy
analysis time and operator fatigue. In an effort to
overcome this problem, various automated GSR analysis
systems for SEM/EDX systems have been developed (DeForest et
al. 1983: Tillman 1987: White and Owens 1987). Currently,
there are a number of energy dispersive X-ray systems on the
market that are designed to handle automated particle search
and recognition of GSR. The manufacturers include: Link

Analytical Systems, Tracor Northern and Kevex.

CHA ONE

SURVEY OF GUNSHOT RESIDUE ANALYSIS IN FORENSIC SCIENCE

LABORATORIES ACROSS THE UNITED STATES.

15

16

There are two general types of methods currently used
for analysis of gunshot residue from the hands of a shooter.
One type, may be termed "Bulk Elemental Analysis
Techniques", which includes: flameless atomic absorption
(FAA) (Koons et al. 1989), neutron activation analysis (NAA)
(Hoffman 1975), inductively coupled plasma-atomic emission
spectrometry (ICP-AES) (Koons et al. 1989), and anodic
stripping voltammetry (ASV) (Brihaye et al. 1982). The
other common type of GSR analysis is by scanning electron
microscopy with energy dispersive X-ray capability (SEM/EDX)
(Wolten et al. 1977).

Gunshot residue is encountered frequently as evidence
in homicide and suicide cases. However, not all forensic
science laboratories choose to analyze this evidence. Those
that do, have a variety of analysis methods to choose from.
To determine who is analyzing GSR and by what means, a
nationwide survey of forensic science laboratories was
undertaken.

The purpose of this survey was three-fold:

1) To determine the methods of analysis being used
nationwide on gunshot residue samples in forensic science
laboratories.

2) To compare and contrast the two general types of

methods being used to analyze GSR.

l7
3) To document the procedures and types of equipment
being used in GSR analysis by SEM/EDX in the interest of
identifying and resolving common problems for the'

investigators using these techniques.

18

The survey was mailed to two hundred Forensic Science
Labs in every state in the U.S.. The data were then
collected and tabulated. A report summarizing the data and
conclusions drawn from the information collected was
available upon request to all participating laboratories. A
cover letter, to help elicit cooperation by the Labs, was
also sent with this survey. Results of the survey were
submitted for publication to the Journal of Forensic

Sciences in July, 1989.

DESCRIPTION OF SURVEY INSTRUMENT

A mail survey was conducted on two hundred forensic
science laboratories distributed to all fifty states in the
U.S. in November of 1988. The response rate to the first
mailing of the survey was 51.0%. A second mailing of the
survey in December of 1988 brought the response rate up to
71.5%. A copy of the survey instrument appears as Appendix

C.

19

BEEHLI§_AND_DI§£Q§§IQN

Table 1 lists the percentage of laboratories analyzing
gunshot residue and the method used for GSR analysis. A
total of 57% of the labs responding don't analyze gunshot
residue themselves: 52% of those labs send GSR samples
either to the FBI or a state/regional lab for analysis. Of
the labs analyzing GSR, 57% use a bulk elemental analysis
technique, 34% employ SEM/EDX alone or combined with FAA.

It is of interest to note that while X-ray Fluorescence and
Photoluminescence have been used in the past to analyze GSR,
no labs indicated the use of those techniques in this
survey.

The results of the survey indicate that thirty-one
states have at least one lab conducting GSR by one of the
methods listed above. Fourteen states had no labs
conducting GSR analysis by the above methods. In five
states, no labs responded.

Some interesting results are obtained when comparing
GSR analysis techniques. One of the main contentions for
using bulk elemental analysis techniques over SEM/EDX in the
past has been the shorter analysis times of the former
technique. The survey data indicate however, that on the
average, the amount of time spent per analysis using either
technique is about the same. In fact, the mean time

required to analyze a sample, as well as the mean time spent

20

Table 1. Percent of forensic laboratories analyzing gunshot
residue and method used for analysis.

 

 

 

% of Labs Responding N
W
GSR Analyzed in Lab 43% 144
Sent to FBI 15.3%
Sent to State/Regional Lab _ 14.6%
Not Done 27.1%
NQLDQd ef GSR Analysis
FAA 48.4% _ 62
SEM/EDX Alone 21.0%
SEM/EDX Combined with FAA 12.9%
NAA 1.6%
ASV 4.8%

Microchemical Tests 11.2%

 

21
on GSR analysis per week, was fairly similar for both bulk
analysis and SEM/EDX techniques: 3.0 hrs per analysis vs
3.1 hrs per analysis, and 13.3 hrs per week vs 15.9 hrs per
week, respectively (Figures 1 and 2).

Survey participants were asked to respond to the
question of how frequently and on what grounds GSR analysis
was challenged in court. The response rate for this
particular question was relatively low but still bears
examination. Table 2 lists how frequently GSR analysis is
challenged in court and Table 3 lists the grounds for
challenge. Bulk analysis techniques are challenged slightly
more often, mainly on the grounds of specificity. This may
be a significant concern with respect to the potential for
false positives. SEM/EDX analysis, when challenged, is
usually challenged on the examiner's interpretation of the
data. Since there are well defined, accepted,
characteristic criteria for defining gunshot residue by this
technique, it is less likely to lead to false positives.

Currently, 54% of the labs with SEM/EDX capability use
their instrument for GSR analysis. Table 4 lists the type
of SEM equipment being used by forensic science labs and its
age and dependability. The scanning electron microscopes
used most frequently throughout the country are listed in
Table 4. Tracor Northern was the most frequently used EDX
system (35%), with EDAX following with 27%, Princeton Gamma

Tech 24% and Kevex with 15%, (n=34). The mean age of

22

Figure 1. Time required per analysis for labs analyzing
gunshot residue.

 

 

 

whéudd mod-2(9). ANN-2V X0w>2mw
OWN-dd No.0uz<m_2 vauzv w_w>._<z< v33m

m_w5<z< 692% § m_w>._<z< xsam I

Am_w>._<z<\.wmzv msE. w_m>._<z<
a; To o-m mé Ya TN «.1 To

 

r _ fl. 0

 

 

.\.\

tow

 

T2

 

 

i; i ON

1mm

 

 

 

 

row

 

GZEZOawwm wm<._ “.0 1x.

_.

23

Figure 2. Analysis time in hours per week for labs
conducting GSR analysis.

 

mdP-dfi 0.072(5). uw_w>._<z< xom\_2mw N
ofiidd 0.072(5). nw_w>.._<z< viam \

w_w>._<z< xamzzmm § w_w>._<z< v33 I

xmmgxmmDOI z_ NEE. w_w>._<z<

om. omumN mmuow owner mwuow 070 0:0

 

 

 

 

or

 

ON

 

 

om

{III-I’ll}... :il:

 

 

0v

 

 

 

om

 

85-2. 02.020mwmm wm<._ ..._O .8

24

Table 2. Frequency of challenge to GSR analysis in court.

 

Frequency of Challenge
(Number of Labs Responding)

0% Rarely (1-10%) 10-30% >50% N

 

Method of Analysis

Bulk Analysis‘ 10 12 5 2 29
SEM/EDX Alone 3 5 1 9
SEM/EDX with FAA 2 4 6

 

1Includes-Flameless atomic absorption, neutron activation,
inductively coupled plasma-atomic emission, and anodic
stripping voltammetry

Table 3. Grounds for challenge in court.

25

 

METHOD OF ANALYSIS (# of labs responding)
SEM/EDX

SEM/EDX & FAA

 

Bulk Analysis Combined
Basis for challenge
Not Specific for GSR 10 1(FAA)
Did Defendant Fire 4
a Gun?
Interpretation of 4
Threshold Level
Interpretation in 3
General
Evidence Consumed 2

Operator Proficiency
EDX Sensitivity

Collection Technique

 

26

Table 4. Laboratory SEM equipment, age and dependability.

 

 

% of Labs Standard
Responding Mean Deviation N
Mek§_91_§flu_flfigd
181 28.6% 35
AMRAY 22.9%
CAMBRIDGE 14.3%
HITACHI 11.4%
CAMSCAN 8.6%
ETEC 5.7%
JEOL 5.7%
BAUSCH & LOMB 2.9%
X-Ra Ana zer
TRACOR NORTHERN 35.3% 34
EDAX 26.5%
PRINCETON GAMMA TECH 23.5%
KEVEX 14.7%
SEM
1-5yrs. 55.6% 6.57yrs. 4.77yrs. 36
6-10yrs. 30.6%
11-15yrs. 8.3% (5.0yrs.=Median)
16-20yrs. 5.6%
i 9; Weeks SEMzEDX
as o
0-1 26.7% 2.8wks. 6.8wks. 30
1-2 26.7%
2-3 24.3% (1.5wks.=Median)
3-4 13.3%
4-5 6.7%
4

3.3%

 

27

scanning electron microscopes being used was 6.57 years with
a standard deviation of 4.77 years and a range of 1 to 20
years old, (n=36). As an estimate of instrument
dependability, the mean number of weeks the SEM/EDX system
was "down" in 1988 was 2.81 weeks with a standard deviation
of 6.81 weeks and a range of 0 to 40 weeks, (n=30). As far
as the type of collection technique being used most
frequently by labs analyzing GSR by SEM/EDX, a tape lift
technique was used by 48% of the labs responding; followed
by a concentration technique (16%), glue lift (12%), swab
for FAA & SEM/EDX, (12%), vacuum suction (4%) and other
techniques (8%), (n=25).

A potential problem that exists in using SEM/EDX for
GSR analysis is the variation between labs in determining
the minimum number of particles analyzed to confirm gunshot
residue (Figure 3). A total of 41% of the labs responding,
reported that finding one particle which meets the shape and
elemental characteristics of GSR is enough to confirm GSR on
the hand (n=17). However, the range of responses to this
question ran from 1 to 10 particles. Some responses gave
options such as, "1-2 unique" or "8-10 characteristic
particles", " depends on the type of particle", or "none
set; based on particles 8 FAA”. What should be the accepted
standard is an important question which needs attention by
experts in the field of GSR analysis using SEM/EDX.

Other problems cited by laboratories using GSR analysis

28

Figure 3. Variations in the minimum number of particles
analyzed by SEM/EDX to confirm GSR.

<<m a mm._0_._.m<n_ 20 omw<m hww mZOz
m..0:.m<n_ “.0 ma>._. ZO mazwamn— Mu ‘
0_._.w_mm._.0<m<IO 07w m0 mac—23 Nu_. u.n.m_.._._.0

Emmy SEEZOO O... QMN>._<Z< wmmofimgi m0 .fi

NF 2. or m m h m m V m N w
_ _ h F _ j o

 

 

 

 

 

 

 

 

 

 

 

 

..... l .Ifltl... Ilvxl»! 1:9.{11'11

 

 

 

|I.Illl|.l..lllllllll‘| :l....ll.. , .11 III)... . .x. :: l... 1;... \v‘lll‘vf. K

 

 

 

 

OZEZOmem wm<.._ ..._O h.

29

by SEM/EDX were as follows:

1) w b o d r
QiIE¥_hénQ§l
A possible solution to this problem would be to use a
stickier medium such as 3M adhesive transfer tape; although
dirty hands will remain a potential problem.

”WM:

cencentratien teennigne, This observation is in agreement
with the findings of Zeichner et al. (1989). Their

recommendation was to eliminate the concentration step and
observe the glue or tape lift directly. Dennis Ward at
the FBI laboratory in Washington D.C., as well as Loren
Sugarman at the Orange County Sheriff-Coroner Department,
Santa Ana, CA. suggest the centrifugal force used in
concentration can be critical. Too high of a g force may
pellet debris on the filter (personal communication).

3) Len th anal sis t' es as w 11 s t an 1 s s

neing fatiguing to the opetatet, especially on negative
samples. An observation was elso mage thet it is difflcnlt
te find a method condneive to netn SEMZEQK eng FAA, A

possible solution to this problem, which is currently being
used in some labs, is to collect samples for SEM/EDX from
the web area of the hand and possibly the face. In
addition, swabs from the back and palm of the hands are
collected. FAA is then used as a screening technique and

only potential positive samples are analyzed by SEM/EDX.

30
Dr. Robin Keeley from the Metropolitan Police Forensic
Science Laboratory (Scotland Yard) points out however, that
FAA is a relatively insensitive technique with respect to
GSR. One may actually have over a hundred particles of GSR
(assuming an average particle size of 3 micrometers) and
still fall below the threshold level for Pb, Ba and Sb as
detected by FAA (personal communication).

4) GSR is collected tee leng eftet the ineigent
eeentet This is an inherent problem. Stressing the need
for collecting samples as soon as possible would be helpful.

5) Ci aret e i te i a ic es 1 GS 'n
ngrpnology and increase enalyeis tine. This is an
interesting observation for which there is no proposed

solution at this time.

31

C N U S

To determine who is analyzing GSR and by what means, a
nationwide survey of forensic science laboratories was
undertaken. Data on gunshot residue analysis were obtained
from a mail survey of two hundred forensic laboratories in
the U.S. with a response rate of 71.5%. Over half of the
labs responding don't analyze gunshot residue themselves;
52% of those labs, send GSR samples either to the FBI or a
regional lab for analysis. Of the labs analyzing GSR, 57%
use a bulk elemental analysis technique, 34% employ SEM/EDX
alone or combined with FAA. Interestingly, the mean time
required to analyze a sample, as well as the mean time spent
on GSR analysis/week, was fairly similar for both bulk
analysis and SEM/EDX techniques. Currently, about half of
the labs with a SEM/EDX use it for GSR analysis. Bulk
analysis techniques are challenged slightly more often in
court, mainly on the grounds of specificity. This may be a
significant concern, with respect to the potential of
reporting false positives. A number of problems and
potential solutions encountered by investigators using the
SEM/EDX technique for GSR analysis were discussed. With the
commercial availability of automated gunshot residue
programs for SEM/EDX equipment a shift towards this type of

analysis may appear in the future.

CHA E O

A COMPARISON OF THREE TECHNIQUES DEVELOPED FOR SAMPLING

AND ANALYSIS OF GSR BY SEM/EDX ANALYSIS

32

33

The objectives of this portion of the thesis are to
determine which of the three collection techniques tested
has:

1) The highest efficiency for lifting particles from the
hand.

2) The minimum analysis time.

3) The most stable adhesive surface with respect to

temperature and deterioration over time.

34

MAIEBIAL§_AED_MEIEQD§

un' 'o - Federal .38 special, 125 grain, jacketed
soft point, Lot# 12A. fleenen - Smith and Wesson .38 special
caliber, model 10-8, 6 shot revolver with a two and one half
inch barrel. Eleetten__nieneeeene - JEOL 35C SEM with a
solid state backscatter detector and equipped with a Tracor
Northern 5500 energy dispersive X-ray analyzer (EDX).
Carbon Coating - Ladd Vacuum Evaporator, Ladd Research
Industries Inc., P.O. Box 1005, Burlington, VT 05402.
getbon Elanchets - Specially smoothed surface, carbon
specimen mounts (disc thickness 5mm, diam. 15mm), Ladd
Research Industries Inc., P.O. Box 1005, Burlington, VT
05402. Gun Clean n o e - Brite Bore, Mill Run
Products CO., Cleveland, OH 44104. n e ' s -
Hoppe's #1204 .38 - .45 cal., Airport Industrial Mall,
Coatsville, PA 19320. Ceneent:etien_peyieee - See Appendix
A for materials and methods according to Sugarman.
Centrifuge - Precision Vari-Hi-Speed Centricone, Precision

Scientific Company, Chicago, IL 60647.

Firing and Collection - (Figure 4) The shooter washed

and dried the hands prior to firing. The firing distance
was 32 inches from barrel to target face ("cotton box").
The shooter used a weapon that had been cleaned prior to

firing. The same weapon was used for all firings. After

35

Figure 4. Gunshot residue collection from the hands of a
shooter. 4a) Shooter holding Smith & Wesson model 10-8
prior to firing. 4b) Muzzle blast from firing one round of
Federal .38 special ammunition. 4c) Firing hand sampled
with a collection device used in the concentration
technique.

 

36
firing, the shooter left the "firing room" and his hand was
sampled by the researcher. The researcher wore latex gloves
and sampled the thumb, web and index finger of the firing
hand by successive dabbings of the hand with the collection
stub (Figure 4c). The collection stub was then numbered and
covered with its protective cap. The collection stub was
removed from its support piece (rubber stopper) and
carbon-coated in the vacuum evaporator or processed through
the concentration technique (Appendices A & B).

Hand and gun conttols - Before analysis on the electron
microscope, all samples were given a random number using a
table of random numbers. The analyst knew what type of
collection device was being analyzed by the surface
characteristics of the device but had no idea what treatment
the collection device had received or whether it was a
control collector. For every treatment, five repetitions
were made and two controls were run. The controls consisted
of a "hand control" where the subject who had been firing
the revolver washed his hands and was then sampled, and a
"gun control" where the revolver was cleaned immediately
after firing by running a solvent (Brite bore) soaked patch
on a .38 caliber copper brush down the barrel and cylinder
chambers several times. This procedure was repeated with a
second solvent soaked patch which was also used to clean off
debris on all metal surfaces of the revolver with the

cylinder removed (open). A third "dry" patch was then run

37
through the bore and chambers as well as cleaning all metal
surfaces as described above. The revolver grips and
exterior were then blown off with compressed air. At this
point, the shooter washed his hands and picked up the
cleaned revolver working the action three times by pulling
the trigger on empty chambers. The hand was immediately
sampled.

Collection devices - Tape lift and glue lift collection
devices consisted of a number 3 inverted stopper that had a
15mm diameter aluminum stub mounted on top of it using 3M
brand 465 adhesive transfer tape. The surface of the
aluminum stub was coated with one of the following adhesive
surfaces to be tested:

1) 3M brand 465 adhesive transfer tape (Nesbitt et a1.
1979).

2) 3M brand 465 adhesive transfer tape fixed to a 15mm
diameter polished carbon planchet coated with rubber cement
diluted 1:4 with toluene (Basu and Ferriss 1980: Basu 1982:
Basu et al. 1984).

Each collection device was covered by a protective cap
consisting of a number 20 plastic test tube closure.
Devices were stored at room temperature and used within 24
hours after preparation, unless otherwise noted.

Collection devices for the concentration technique
consisted of 3M brand double stick tape bound to a 15mm

diameter mylar surface that is coated with Vistanex adhesive

38

diluted to 15% with hexane (Appendix A) (Figure 4c).

Eleetten_uieteeeeny - The JEOL 35C was operated at an
accelerating voltage of 25kV. The lithium-drifted silicon
crystal of the X-ray detector was kept in liquid nitrogen at
a distance of 55mm from the center of the column. The
working distance for the specimen was 39mm. A brass
specimen holder with a 25mm diameter was lined with an
aluminum adapter to accommodate the 15mm diameter stub being
analyzed. An objective aperture setting of 600 micrometers
was used to obtain increased signal for backscatter
detection. A known sample of gunshot residue collected on
3M brand 465 adhesive transfer tape (3 rounds fired) was
inserted into the microscope as a standard for fine
adjustment of the backscatter image (Figure 5). The image
was focused at 300x in the secondary electron image (SEI)
mode. The backscatter image, which is sensitive to
increasing atomic number, was collected using the slow
scanning option while adjusting the gain and contrast to
give a dark background with GSR appearing as bright circular
white spots with a circular halo around them. A particle
was selected and the magnification was increased so that the
particle image filled the majority of the screen. The image
was refocused in the SEI mode and a X-ray spectrum was
accumulated for 70 seconds at a beam current of 550
picoamps. The particle was confirmed as gunshot residue if

it fell into one of the following four categories:

39

Figure 5. Location and identification of GSR by SEM/EDX.
5a) Secondary electron image of a tape lift surface used to
collect GSR from hand after firing three rounds of Remington
.38 special ammunition. Arrowheads point to potential
gunshot residue particles. Magnification - 30X, bar = 1000
micrometers. 5b) Backscatter electron image of the same
field as 5a. Arrowheads point to potential GSR particles.
Large particle in inset is enlarged in micrograph 5c.
Magnification - 30X, bar = 1000 micrometers. 5c) Secondary
electron image of large nodulated GSR particle from inset in
5b. Magnification - 5500X, bar = 1 micrometer. 5d) Energy
dispersive X-ray spectrum from GSR particle in SC.

 

VFS = 2048

 

d

 

 

 

0.000

 

40
1) Lead (Pb), Barium (Ba), and Antimony (Sb).
2) Ba and Sb.
3) Ba, Calcium (Ca), Silicon (Si) with a trace of Sulfur
(S).
4) Ba, Ca, Si with a trace of Pb provided that no zinc (Zn)
was present (residue from stud guns has been found to

contain Ba, Ca, Si, Pb, Cu and Zn) (Wolten et al. 1979b).

OPERATIONALIZING THE VARIABLES

Dependent Vatiaples

1) Efficiencx_2f_cgllectien - Efficiency of collection
is defined as: The number of GSR particles found in one hour

searching at 300x magnification in the backscatter mode in
the SEM. If five GSR particles were found in less than one
hour, the time taken to find the particles was recorded and
the number of particles found in one hour was extrapolated

according the equation:

60 minutes/search time (minutes) X # of particles found =

number particles found per hour

This analysis assumes that GSR was distributed randomly on
the stub (Wolten et al. 1977).

2) Analysis time - Analysis time is defined as the time
required from insertion of the sample into the electron

microscope (time zero) to the moment that five particles

41
whose spectra and morphology characterize them as GSR, have
been found and saved to disk. An upper limit of one hour
was set for the analysis time per sample.

3) Cententretien_tine - This refers only to the
collection technique using the Vistanex adhesive.
Concentration time is defined as: the time from removal of
the mylar surface to the moment that the 0.45 um filter was

dry and ready to be carbon coated.

Independent Vetianles
1) Numbet e: rounds fiteg - Either one or three rounds

were fired. Sampling was conducted using a collection
device prepared 24 hours in advance and stored at room
temperature.

2) Ienpetetnte - Holding the number of rounds fired
constant at three, each collection device was prepared 24
hours in advance keeping the devices at a temperature of 56
degrees Celsius or -4 degrees Celsius for 12 hours prior to
sampling.

3) line - Holding the number of rounds fired constant
at three, and the temperature constant at 22 degrees
Centigrade (room temperature) collection devices were
prepared and stored for three weeks or six weeks prior to

sampling.

42
§IAII§IIQ§

Three different methods of collecting GSR were
examined. They were: 1) The tape lift 2) The glue lift
3) A Vistanex glue lift followed by concentration via
centrifugation.

For each collection method six different treatments
were examined. The treatments were: 1) three rounds fired
2) one round fired 3) collection device stored twelve
hours at -4 degrees Centigrade 4) collection devices
stored twelve hours at 56 degrees Centigrade 5) collection
devices stored for three weeks prior to use 6) collection
devices stored for six weeks prior to use. Unless otherwise
stated, collection devices were prepared 24 hours in advance
and residue was collected from three rounds being fired.

For each treatment there were five repetitions. The
gun was cleaned between each repetition and both the shooter
and the investigator collecting GSR washed their hands
between repetitions.

In addition to the five repetitions for each treatment
a "hand control" and a "gun control" sample was collected
from the band (see Materials and Methods section for firing
and hand sampling procedure). Washing of the hands is
usually sufficient to remove gunshot residue, however,
depending on how thoroughly the hands are washed, some GSR
may remain (Harrison and Gilroy 1959).

Unfortunately, the data from each method were not

43
collected in a totally random fashion. Reasons for this
were two-fold. 1) The concentration method required a
special size 0.45 micron Nucleopore filter, which had to be
special ordered. This item was backordered for several
months, at which point it was decided to begin collecting
GSR using the tape lift method. 2) The treatments of
collectors prepared three and six weeks in advance made a
totally random design impractical.

GSR collections were made one method at a time
beginning with the tape lift method. A single treatment
consisting of five repetitions and a "hand " and "gun"
control was completed on a given day. After all GSR
collections were made for the tape lift method and the glue
lift method, the collection devices were renumbered using a
table of random numbers and analyzed in a totally random
fashion on the electron microscope. When the 0.45 micron
filters arrived for the Vistanex glue lift concentration
method, samples were collected and random numbers assigned
for analysis as described above.

Since the experimental design was not random between
methods, the significance of statistical analysis between
methods should be interpreted with caution. However, within
a given method, the collection devices were analyzed in an
independent and random fashion allowing an analysis of
variance to be made between treatments within a single

method.

44

Assuming a random distribution of particles on the
collection surface (Wolten et al. 1977), particle count data
were normalized to time. For ease of statistical
manipulation, the number of particles per one hour search
time was determined.

Search time was equal to the time taken to find five
GSR particles, or 60 minutes if less than five GSR particles
were found.

The ln (x+1) transformation was required to fulfill the
homogeneity of variance assumption of ANOVA. Homogeneity of
variance within method and between treatments was examined
with Bartlett's test (Steel and Torrie 1980). Means from In
(x+1) transformed data were separated via the SNK (Student
Newman Keuls) procedure or Student's t test following ANOVA

(Proc. GLM, SAS, 1983) (SAS Institute Inc.).

45

COMPARISON BETWEEN COLLECTION METHODS

An indication of the collection device particle lifting
efficiency was reflected by the mean number of particles
found per hour. Obviously, the more particles found per
unit time the greater the efficiency of collection. The
data in Table 5 indicated that the ln(x+1) particle per hour
means were significantly different from each other at the
.05 alpha level for the three collection methods tested.

If one examines the particles/hr. mean for the tape
lift method the mean particles/hr. was greater than five.
In other words, on the average, it took less than one hour
to find five particles of GSR on the tape lift surfaces
(Table 5).

The particle per hour means for both the concentration
technique and the glue lift were much lower than in the
tape lift method. However, since the experimental design
was not random, comparisons between methods must be

considered carefully.

COMPARISONS WITHIN COLLECTION METHOD
The data in Table 6 suggest that the collection
efficiency of the 3M type 465 adhesive transfer tape was

stable under all conditions tested. The SNK test showed a

46

Table 5. Statistical analysis of the ln (x+1) of particle
per hour means between collection methods.

 

 

Method SNK2 Mean N Mean
Grouping (ln (x+1) (particles per

of particles hour)

per hour)
Tape lift A 1.973 30 6.192
Concentration B 1.081 30 1.948
technique
Glue lift C 0.538 30 0.713

 

fCalculated by SNK procedure using analysis of variance.

2 Means sharing the same letter are not significantly

different from each other at alpha = .05 level.

47

Table 6. Statistical analysis of the ln (x+1) of particle
per hour means for tape lifts.

 

 

Treatment SNK2 Mean N Mean
Grouping (1n (x+1) (particles per
of particles hour)
per hour)
1 Round fired A 2.233 9.380
3 Rounds fired A 2.188 8.100
-4 Degrees C. A 2.105 8.320
3 Weeks old A 1.938 7.540
Gun control A 1.705 5.517
56 Degrees C. A 1.694 5.240
6 Weeks old A 1.678 6.128
Hand control B 0.384 2.567

A

T

Calculated by SNK procedure using general linear models.

2 Means sharing the same letter are not significantly
different from each other at alpha = .05 level.

48

significant difference in means (1n (x+1) of particles/hr.)
at an alpha level of .05 for the hand control data compared
to the other treatments, as would be expected. A result
which was not expected was that the gun control data showed
no significant difference in mean (ln (x+1) of
particles/hr.) when compared to the other treatments. This
observation will be discussed in detail later in chapter 3.

The concentration technique gave highly variable
results between treatments tested. A comparison of the 1n
(x+1) of particle/hr. means for the treatments in the
concentration method by SNK, resulted in means which were
significantly different at the .05 alpha level (Table 7).
Namely, one round fired samples had the highest mean and was
significantly different from the six week old collections,
gun controls and three rounds fired. The potential source
of the decreased particle counts found in the concentration
method will be examined in the discussion section later in
this chapter.

In the present study, statistical analysis by SNK of ln
(x+1) of particles/hr. means for the glue lift method showed
no significant differences between hand controls and any of
the treatments examined (Table 8). Too few particles per
stub surface were found. This is indicative of an
inefficient particle lifting surface. Note the mean number
of particles found per hour for the glue lift technique in

Table 8.

49

Table 7. Statistical analysis of the ln (x+1) of particles
per hour means for concentration method.1

 

 

Treatment SNK2 Mean N Mean
Grouping (1n (x+1) (particles per
of particles hour)
per hour)
1 Round fired A 1.828 6.080
3 Weeks old A B 1.520 4.160
56 Degrees C A B 1.430 4.320
Hand control A B 0.855 1.600
-4 Degrees C A B 0.748 1.800
6 Weeks old B 0.599 1.200
Gun control B 0.462 0.833
3 Rounds fired B 0.358 0.600

A

‘ Calculated by SNK procedure using general linear models.

2 Means sharing the same letter are not significantly
different from each other at alpha = .05 level.

50

Table 8. Statistical analysis1of the ln (x+1) of particles
per hour means for glue lifts.

 

 

Treatment SNK2 Mean N Mean
Grouping (ln (x+1) (particles per
of particles hour)
per hour)
1 Round fired A 1.011 2.540
3 Rounds fired A 0.832 1.800
3 Weeks old A 0.555 0.800
6 Weeks old A 0.416 0.600
Gun control A 0.366 0.667
56 Degrees C A 0.277 0.400
-4 Degrees C A 0.139 0.200
Hand control A 0.000 0.000

 

i

Calculated by SNK procedure using general linear models.

2 Means sharing the same letter are not significantly
different from each other at alpha = .05 level.

51

In order to test the efficiency of the glue lift
technique, an experiment was designed as described in the
procedure for three rounds fired, where the hand was sampled
by twelve dabs with the glue lift, the same area of the hand
was then sampled with a tape lift collector. The procedure
was repeated five times with gun cleaning and hand sampling
as described previously. Tape lift stubs were carbon coated
and all stubs were assigned random numbers and analyzed on
the electron microscope.

An ANOVA by Student's t test indicated a significant
difference in the ln (x+1) of particle/hr. means for the
glue lift and tape lift collectors, alpha = .05.

These results demonstrated the less efficient particle
lifting surface of the glue lift devices (Table 9).
Decreased collection efficiency was reflected not only by
the particle/hr. means being much lower in the glue lift vs
the tape lift, but also in the fact that particles collected
on the tape lift surface represent GSR, which was left on
the hand after initially sampling the hand with a glue lift

device.

52

Table 9. Statistical analysis of 1n (x+1) of particles per
hour means for glue lifts followed by tape lifts.

 

 

Method T2 Mean N Mean
Grouping (ln (x+1) (particles per
of particles hour)
per hour)
Tape lift A 1.788 5 6.040
Glue lift B 0.8832 5 1.400

 

iCalculated by Students t test (LSD).

2 Means sharing the same letter are not significantly
different from each other at alpha = .05 level.

53
DI§§H§§IQH

The tape lift collection devices in this study proved
to be the most efficient particle lifting devices examined.
The concentration technique gave highly variable results
between treatments but still had a higher ln (x+1) of
particle per hour mean than the glue lift technique (Table
5).

The tape lift surfaces were found to be stable under
all treatments tested. They have a shelf life of at least
six weeks and are not effected by twelve hour exposure to
temperatures which might be encountered by collection
devices stored in a crime scene vehicle.

The concentration technique on the other hand, gave
highly variable results between treatments. Some possible
explanations which may contribute to the highly variable
results (Table 7) obtained using the concentration method
are as follows: 1) The concentration technique is actually
a combination of two techniques; collection and
concentration. Whether a decreased number of particles
found per hour was due to the Vistanex surface being less
efficient in collecting the particles or due to loss of
particles during the concentration procedure can not
be determined from this experiment.

There are at least three possible areas where GSR may
be lost in the concentration method: 1) particles lost

from nonadhesive 0.45 micrometer Nucleopore filter when it

54
was teased away from subfilter. 2) actual particles not
counted due to Ba, Ca and Si contamination of subfilter. 3)
particles trapped in debris and either aspirated out of
concentrator or pelleted onto a filter surface.

The data, in fact, suggest that a problem may exist in
the concentration procedure. If one examines the
particles/hr. means of three rounds fired vs one round fired
in the concentration method, the means are 0.60 and 6.1,
respectively (Table 7). This was exactly opposite to the
results one might expect. Going back to the actual
concentration procedure employed, it was noted at the time
that in six out of the seven concentrators the 0.45
micrometer Nucleopore filters could not be peeled away from
the underlying Nucleopore D-79 subfilter. The filters were
mounted together on an aluminum stub, carbon coated and
viewed. Analysis in the electron microscope gave high
background levels of Ba, Ca, Si, and K for the 0.45 um
Nucleopore filters adhered to subfilters (Figure 6). The
surface itself tended to pucker and charge to a degree that
a reliable backscatter image was unattainable. At this
point, the 0.45 um filter was dissected away from the
subfilter with a razor blade and remounted on an aluminum
stub. This required manipulation of the 0.45 um filter,
which has no adhesive nature of its own, and it is quite
likely that GSR particles were lost during this

manipulation.

55

Figure 6. Identification of contaminating elements on a
filter surface used in the concentration technique. 6a)
Secondary electron image of a 0.45 um Nucleopore filter
adhered to a D-79 subfilter after carbon coating.
Magnification - 16,000X, bar = 1 micrometer. 6b) Energy
dispersive X-ray spectrum of the filter sandwich in 6a.
Vertical full scale = 512 counts, X axis from 0.0 - 15.0
KeV.

 

VFS: 512

bSi

 

 

Cu
l Zn

 

 

 

 

15.0

56

In contrast, with the concentrators used in the one
round fired experiment, five out of seven Nucleopore filters
were easily removable from the subfilter surface with the
remaining two picking up only slight subfilter
contamination.

After barium and calcium contamination was observed, a
subfilter was mounted on an aluminum stub and carbon coated
to determine what elements were present in the subfilter.
Results of this analysis showed the presence of silicon,
potassium, zinc, calcium and barium (Figure 7). This
contamination compounds the problem of GSR analysis. One of
the forms of GSR recognized in the Aerospace Corp. study
(Wolten et al. 1977), were spheroid particles containing the
elements: lead or sulfur, silicon, calcium and barium. On
a 0.45 um Nucleopore filter adhering to a subfilter, such a
particle would be difficult to distinguish from a pure lead
or sulfur particle. Therefore, all GSR reported in the
concentration method consisted of particles composed of
lead, barium and antimony, unless the filter surface showed
no background element contamination from adhering subfilter
material.

Other investigators, using a different type of
concentrator, noticed lead and barium contamination of the
fifty micrometer porous polythene filter in their
concentration device. Washing the filter with 20%

hydrochloric acid was found to remove the contamination in

57

Figure 7. Identification of contaminating elements in the
subfilter used in the concentration technique. 7a)
Secondary electron image of carbon coated D-79 subfilter.
Magnification - 300x, bar = 100 micrometers. 7b) Energy
dispersive X-ray spectrum of subfilter surface in 7a.
Vertical full scale = 4096 counts, X axis from 0.0 - 15.0

KeV.

 

VFS = 4096

 

b

 

 

Si

 

 

 

15.0

58
their case (Wallace and Keeley 1979).

A third factor, which could influence the number of
particles found after concentration, is the amount of debris
deposited on the 0.45 um Nucleopore filter. One of the main
reasons originally proposed for the concentration method was
to reduce the amount of epidermal cells and other debris
picked up by the collection device. Such debris may cover
GSR particles present making them undetectable in the
electron microscope (Ward 1982). Mr. Dennis Ward at the FBI
crime lab has suggested that centrifugal force may be a
critical factor depending on the amount and type of debris
present on the Vistanex surface. Too low of a g-force
results in material floating on the bromoform surface, which
may have trapped GSR in it. Too high of a g-force may
pellet debris onto the filter obscuring GSR particles
(personal communication).

A varying amount of debris was found on the filter
surface in the concentration technique (Fig. 8). While the
present study was being conducted, an experiment comparing
tape lifts to the concentration technique was performed by
Zeichner et al. (1989). These researchers concluded that
the build up of debris on the filter was such a problem,
that direct observation of a tape or glue lift surface was
preferable to concentration.

The concentration technique used in the current study

was a modification of Ward's technique (Ward 1982),

59

Figure 8. Varying amounts of debris found on filter surface
after the concentration technique. 8a) Secondary electron
image of the surface of a 0.45 um Nucleopore filter after
GSR concentration. A small amount of debris can be observed
scattered across the center of the filter. 8b) Secondary
electron image of the surface of a 0.45 um Nucleopore filter
after GSR concentration. Many pieces of debris can be seen
across the center of the filter. 8c) Secondary electron
image of the surface of a 0.45 um Nucleopore filter after
GSR concentration. The filter surface is almost totally
obstructed by debris. All micrographs are at a
magnification of 10X, bar = 1000 micrometers.

 

60
developed by Loren Sugarman at the Forensic Science
Laboratory of the Orange County Sheriff-Coroner Department
in Santa Ana, CA. (Appendix A). Mr. Sugarman has been able
to circumvent some of the problems in the present study
(personal communication). He has not noticed any
contamination of the D-79 subfilters obtained from
Nucleopore, which suggests the contamination of the
subfilters in the current study may be a batch defect. To
reduce adhesion between the 0.45 um filter and subfilter,
Mr. Sugarman recommended placing the subfilter with the
cross hatched surface facing up and removing the 0.45 um
Nucleopore filter immediately after centrifugation. He also
advised washing the filter through methanol thoroughly after
the bromoform step to remove any traces of bromine on the
filter surface which would give interfering backscatter
signals.

The glue lift technique was found to have an
inefficient particle lifting surface (Table 5). These
findings were not in agreement with the observations
published by the developers of the glue lift technique (Basu
and Ferriss 1980).

The glue lift technique was developed by Dr. Samarendra
Basu and Dr. Stark Ferriss (Basu and Ferriss 1980). It was
designed to be less sticky than the tape lift surface. The
reasoning was that the decreased stickiness of the glue lift

surface would not collect so much interfering epidermal

61
cells and other debris.

In their original paper on the development of the glue
lift technique, several advantages of the glue lift surface
were demonstrated compared to the tape lift surface (Basu
and Ferriss 1980). Advantages included: 10 or more
particles found per area searched (1.5mm diam. circle) 2)
no electron beam damage to glue lift surface vs melting of
the tape lift surface 3) smoother surface of carbon
planchet 4) no carbon coating required.

At this point it is worth examining some differences
between the two studies.

The carbon planchets used in the study by the
developers of the glue lift were obtained from Ernest F.
Fullam Inc. (Schenectady, NY 12301.) They were described as
the "clean, polished carbon planchets (disc thickness 1/8",
diameter 1/2"). In the present study carbon planchets were
obtained from Ladd Research Industries Inc., specially
smoothed surface carbon specimen mounts (disc thickness 5mm,
diameter 15mm). At high magnification, the surface of the
carbon planchets obtained from Ladd Research Industries
appeared somewhat irregular (Fig. 9). Dr. Basu suggests
that lack of a smooth regular surface may impair particle
lifting ability (personal communication). In addition, the
carbon planchets obtained from Ladd Research Industries were
contaminated with tungsten particles generally of about 0.5-

2.0 microns in diameter (Fig. 9). Spraying the planchets

62

Figure 9. Identification of contaminating tungsten
particles on a carbon planchet. 9a) Secondary electron
image of the surface of a carbon planchet obtained from Ladd
Research Industries Inc. Magnification - 300X, bar a 100
micrometers. 9b) Backscatter electron image of the same
field as in 9b. Arrowheads point to contaminating tungsten
particles. Magnification - 300X, bar = 100 micrometers.

9c) Energy dispersive X-ray spectrum of a single tungsten
particle. Vertical full scale 2048 counts, X axis from 0.0
- 15.0 KeV.

 

VFS = 2048

 

 

 

0.000

63
with compressed air prior to applying the glue surface
proved insufficient to remove all of the tungsten particles.
The source of the tungsten was presumably material left
behind during the cutting process at the factory. An
average of nine contaminating tungsten particles per glue
lift were encountered. These particles mimic GSR in the
backscatter mode and lengthen analysis time.

The authors of the glue lift technique examined a
minimum of four 1.5mm circles on the glue lift surface,
finding an average of 58 particles per circle (Basu and
Ferriss 1980).

Typically, one hour of search time, in the present study,
at 300x resulted in searching approximately 15% of the total
surface area of the 15mm diameter carbon planchet. This was
equivalent to examining 2.85 of the 1.5mm circles described
by the developers of the glue lift (Basu and Ferriss 1980).

Ammunition used in the two studies was also different.
The ammunition used in the present study was Federal .38
special cal., jacketed soft point lead, 125 grain for law
enforcement use. Ammunition from the same lot number (12 A)
was used throughout the study. This ammunition was chosen
because it gave consistently few gunshot residue particles.
It was felt that this resembled actual case work conditions
in a more realistic fashion than an ammunition which
produces hundreds to thousands of particles. The developers

of the glue lift used either "standard Winchester or

64

Remington ammunition" (Basu and Ferriss 1980) for the pistol
loads. In the present study, Remington .38 cal., 158 grain
lead ammunition was test fired from the same revolver used
in this work and found to produce hundreds of particles,
mainly lead in composition. The glue lift developers
findings of "30-116 residues per 1.5mm diameter circle"
(Basu and Ferriss 1980), on the glue lift surface was
consistent with the Remington ammunition tested. However,
the finding that "a typical 1/2" diameter tape-lift disc may
contain from 2-10 observable GSR." (Basu and Ferriss 1980),
with one round fired was not consistent with the Remington
ammunition tested. In support of the finding of few GSR on
a tape lift surface, the authors quote a table in the work
by Sild and Pausak (1979), where it was mentioned that
"firing two shots with a .38 cal revolver, they recovered
slightly more than 8 GSR and 20 lead particles from a one
inch diameter "tape-lift" disc (Basu and Ferriss 1980). If
one examines Table I. in Sild and Pausak (1979), one can
also find a test firing of 2 shots with a .38 cal. revolver
where a single sweep (magnification 1000X) shows more than
20 particles of lead and greater than 20 particles of GSR
(Pb+Sb+Ba) are found on the whole stub.

In the original study of the glue lift technique, Basu
and Ferriss cite the beam damage that occurs on the tape
lift surface. They demonstrated this with a micrograph

(Basu and Ferriss 1980 Fig. 2-f) depicting "a lead particle

65
disappearing into a cavity on transfer tape, created by the
bombarding electrons." . The authors went on to discuss
particles imbedding themselves and disappearing into the
melted tape surface.

In the present study, where hundreds of particles of
various composition were observed and spectra obtained, no
particle was ever seen to "disappear into the melted surface
of the tape". Electron beam damage to the tape surface did
occur and usually appeared as a crater with surrounding
folds around the particle (Figures 10a & 11a). The only
time the tape lift surface was seen to crack or melt
severely was if it was not coated with enough carbon
initially (Fig. 12). This problem was easily remedied by
applying another carbon coat. The thickness of a single
carbon coat was typically in the range of 35-40nm. (Appendix
B).

Another way to induce electron beam damage is by using
excessive beam current. In the work by Basu and Ferriss
(1980), the beam or specimen current used was not mentioned.
The emission current was listed as 100 microamps but this
gave no information as to the current which the specimen is
encountering. In the present study, a beam current of 550
picoamps was used. This was measured using a Faraday cup
inserted after the final aperture. The tape lift surface
was found to be stable under these conditions.

Older energy dispersive x-ray analysis systems may

66

Figure 10. Electron micrographs of small GSR particles on
various surfaces. 10a) Secondary electron image of GSR on
a tape lift surface. 10b) Backscatter electron image of
GSR particle in 10a. 10c) Enlargement of GSR particle in
10a. 10d) Secondary electron image of GSR on a 0.45 um
Nucleopore filter surface. 10e) Backscatter electron image
of GSR particle in 10d. 10f) Enlargement of GSR particle
in 10f. 10g) Secondary electron image of GSR on a glue
lift surface. 10h) Backscatter electron image of GSR
particle in 10g. 10i) Enlargement of GSR particle in 109.
Magnification for micrographs a,b,d,e,g and h - 300X, bar =
100 micrometers. Magnification c,f and i - 12,000X, bar = 1
micrometer.

4

 

 

 

 

67

Figure 11. Electron micrographs of large GSR particles on
various surfaces. 11a) Secondary electron image of GSR on
a tape lift surface. 11b) Backscatter electron image of
GSR particle in 11a. 11c) Enlargement of GSR particle in
11a. 11d) Secondary electron image of GSR on a 0.45 um
Nucleopore filter surface. 11e) Backscatter electron image
of GSR particle in 11d. 11f) Enlargement of GSR particle
in 11f. 119) Secondary electron image of GSR on a glue
lift surface. 11h) Backscatter electron image of GSR
particle in 11g. 11i) Enlargement of GSR particle in llg.
Magnification for micrographs a,b,d,e,g and h - 300x, bar =
100 micrometers. Magnification in c - 2,400X, bar = 10
micrometers. Magnification in f - 2,700X, bar = 10
micrometers.

Magnification in i - 1,200X, bar = 10 micrometers.

 

 

 

 

 

 

68

Figure 12. Effect of insufficient carbon coating on a tape
lift surface. 12a) Secondary electron image of tape lift
surface cracking and charging due to beam damage from
insufficient carbon coating. 12b) Secondary electron
image of the same tape lift stub after a second carbon
coating was applied. Magnification of both micrographs -
30X, bar = 1000 micrometers.

 

69
require higher beam currents when accumulating EDX spectra.
An example of this was observed during the present study at
the Center for Electron Optics. The Tracor Northern TN2000
EDX system required a beam current of approximately 1000
picoamps to accumulate a spectrum with a 30% dead time for a
given GSR particle. Keeping the same detector but changing
the hardware to a newer TN5500 EDX system resulted in the
accumulation of an EDX spectrum at 500 picoamps with a 30%
dead time on the same GSR particle.

Perhaps the developers of the glue lift technique were
using an older instrument requiring a high beam current or
had insufficient carbon coating on their tape lift samples.
This would explain the observed melting of the tape surface
and the overall diminished particle counts, as particles
"disappeared" from view into the melted surface.

The procedure for sampling the hand with the glue lift
disc was different than the hand sampling procedure using
the tape lift disc. The developers of the glue lift
maintain that the hand should be touched only 5 times along
the thumb, web and forefinger for sampling the "back" of the
shooters hand (Basu and Ferriss 1980). Whereas, authors
using the tape lift method recommend touching the entire
area of thumb, web and forefinger (about 12 touches) or
until the stickiness of the tape is lost (Matricardi and
Kilty 1977, Nesbitt et al. 1976, Wolten et al. 1977, Wolten

et al. 1979b). In the present study, perhaps the tape lift

70
picked up more particles because a greater surface area of
the hand was sampled.

This possibility was examined by collecting GSR from
the hand using a glue lift surface first and dabbing the
hand twelve times along the thumb, web and forefinger. This
collection was then followed by a tape lift collection along
the same area. The data in Table 9 indicated that when the
surface area sampled was held constant the glue lift
remained a less efficient particle lifting device compared

to the tape lift.

71

OBSERVED ADVANTAGES AND DISADVANTAGES OF EACH COLLECTION
METHOD

Besides examining collector particle lifting efficiency
for a variety of conditions such as number of rounds fired,
temperature and storage time, a table of observed advantages
and disadvantages for each collection method was developed
(Table 10).

The tape lift method for GSR collection has the primary
advantage of having an efficient particle lifting surface,
as previously discussed. The tape itself is inexpensive and
the collection devices are simple to construct. The
adhesive surface was found to be stable, i.e. particle
lifting ability was not decreased significantly, under all
conditions tested. The tape lift surface gave a good
secondary electron image, which is important for
photographing particles (Figs 4a, 4c & 5a, 5c), especially
when using an instrument that has a backscatter detector
that does not operate at the normal TV scanning rate (not
recommended).

Disadvantages of the tape lift method are few. The
surface requires a carbon coat. Depending on the size stub
used, there is a relatively large surface area to be

scanned. Debris collected from the hand may hide GSR

72

Table 10. Observed advantages and disadvantages of GSR
collection methods tested.

 

 

 

 

Method Advantages Disadvantages
Tape Lift Efficient particle Requires carbon coat
lifting surface
Inexpensive Large surface area
Simple to prepare Skin debris may hide
particles
Temperature stable
Stable shelf life
at least 6 weeks
Good secondary image
Glue Lift Requires no carbon Inefficient particle
coat lifting surface
Easy to prepare Contaminated with
tungsten particles
Picks up less Carbon planchets
debris expensive
Fair secondary Large surface area
image
Skin debris may hide
particles
Concentration Separate debris Requires carbon coat

from GSR
Small surface area
Pre-made collectors

and concentrators
can be purchased

Collection efficiency
variable

expensive

2hr. Processing time
Subfilter contamination
Filters stick together

Poor secondary image

 

73
particles beneath it.

The chief advantage of the glue lift is that the carbon
coating step may be skipped. The devices are also quite
simple to construct. The secondary image was not optimal in
the present case due to the roughness of the carbon
planchets obtained from Ladd Research Industries (Figs 10g,
10i & 11g, 11i). Their theoretical advantage is that the
surface is less sticky and therefore picks up less debris
from the hand.

Unfortunately, the glue lift surface did not pick up
much GSR either. It was found to be an inefficient particle
lifting surface. The stability of the glue lift surface to
temperature and storage could not be determined due to the
minimal number of GSR found on the glue lifts for all
treatments tested. The carbon planchets themselves were
moderately expensive and were found to be contaminated with
interfering tungsten particles. The problem of picking up
debris from the hand was reduced with the glue lift surface
but not entirely eliminated. The surface area to be
searched is the same as in the tape lift which is relatively
large. Typically, one hour of search time at 300X resulted
in searching approximately 15% of the total surface area of
the 15mm diameter carbon planchet.

The concentration technique has the potential of
separating GSR from debris. Pre-made collection devices may

be purchased from Kinderprint Co. Inc. or made by the

74
investigator at minimal cost. The total surface area to be
searched is reduced to the point where a manual search of
the entire filter is possible in less than one hour,
provided that there are not a lot of interfering particles
of high atomic number.

In the present study, the concentration method was
found to give highly variable results. Particle lifting
efficiency and stability of collection surface could not be
determined due to several factors contributing to particle
loss. As discussed previously and as listed in Table 10.
those factors were: 1) contamination of subfilter with
barium, calcium and silicon 2) particle loss due to
manipulation of filter surface 3) particle loss due to
aspiration of particles trapped in debris or particles
trapped in debris on filter surface.

In addition to above factors, the concentrators
themselves were relatively expensive items (Appendix A). In
order to achieve a decent secondary electron image, the 0.45
um Nucleopore filter had to be carbon coated rather heavily.
Even under these conditions, discriminating GSR particles
from the background in order to obtain a photograph was
difficult at best (Figs. 10d, 10f & 11d, 11f). An extensive
methanol wash must also be used to remove bromoform from the
filter to reduce interference in the backscatter mode.
Finally, one must consider the additional time required to

concentrate the samples. This was approximately two hours

75

to prepare six samples.

0 C O O V

All of the methods examined had their advantages and
disadvantages (Table 10). Perhaps the optimal collection
device would be one that combines advantages of all three
collection techniques. One would like to have a surface
that does not require carbon coating, as in the glue lift
technique. That surface would ideally be polished smooth
and flat, with no contaminating elements of high atomic
number. The surface should be coated with a substance that
has the stickiness of 3M 465 adhesive transfer tape. If the
collector was viewed directly in the electron microscope and
found to have too much debris on its surface, then one would
like to be able to take that same surface and apply the
concentration technique to it. The subfilter of the
concentrator should be free of contaminating elements of
high atomic number. The final filter surface should give a
better secondary image than is currently obtained on 0.45 um
Nucleopore filters. The collection device should be
relatively inexpensive and stable to conditions of
temperature and time (i.e. long shelf life).

In actuality, the proposed collection device could use
a polished graphite circular 15mm wafer as the sample
surface. This wafer could be coated with diluted Vistanex

adhesive which is stickier than rubber cement. If, after

76
viewing the sample initially, concentration was deemed
appropriate a concentration device with a larger bore
diameter to accommodate the 15mm wafer could be used. The
entire concentration device would then be sonicated to
remove and solubilize the Vistanex surface at which point
the wafer could be removed and concentration could proceed
as described in Appendix A. A final filter of the same
diameter but of a wider pore spacing which is more conducive
to carbon coating could be used as suggested by Wallace and
Keeley (1979).

At the moment, cost is the initial stumbling block.
Highly polished graphite planchets are available but are
extraordinarily expensive. It would be interesting to
obtain some of these highly polished planchets and see to
what degree particle lifting efficiency could be improved

over the current collection devices.

THE "GUN CONTROL" EXPERIMENT

77

78

The purpose of the hand and gun control collections was
to determine whether there was a certain amount of
"background" GSR present after washing of the hand or
working the action of a freshly fired, but cleaned gun. The
expectation of the researcher was that these collection
devices would show low levels of GSR. After the randomly
numbered devices collected as part of the experiments in
Chapter Two of this thesis had been analyzed in the electron
microscope, the numbers were decoded and the hand and gun
control data were compiled. The hand controls gave a
characteristically low GSR particle/hr. mean (Table 6).

The gun controls however, resulted in a consistently
high number of GSR particles being found on the hand (Table
6). This result was rather unexpected, as finding GSR on
the hand after pulling the trigger on a cleaned unloaded
weapon had never been previously reported in the literature.

Since the gun control finding was unexpected and rather
startling, a more rigorous experiment was designed in an
attempt to reproduce the gun control observations. In the
original gun control collections, the revolver was cleaned
in the same room where the test firing had taken place.

This room was equipped with an exhaust fan that was kept

running while shooting and cleaning took place. It is

79
possible that GSR was settling out of the air onto the
revolver during the cleaning process. It has been
demonstrated that a collection device left exposed in a
firing range behind the shooter will collect GSR (White and
Owens 1987). In addition, in the original gun control
observations, the person doing the repetitive firing was
also the person who dry fired the weapon for the gun control
collection. In the redesigned gun control experiment these

potential factors were controlled.

80

In the redesigned gun control experiment, one
investigator fired three rounds of Federal ammunition (.38
special cal. copper jacketed 158 grain bullet). The
revolver (S&W Model 10-8) was immediately removed from the
firing room and taken to a separate room where it was
cleaned, as described previously. Swabs from the cleaning
operation were kept and photographed to demonstrate the
diminished powder residue by the third swabbing (Fig. 13).
A second investigator, not having fired the weapon or been
in the firing room, washed his hands and dry fired the
revolver three times. His hand was then sampled as
described previously with collection devices of the 3M 465

adhesive transfer tape type. This procedure was repeated

five times along with one hand control sample taken from the

second investigator's freshly cleaned hands. Collection

devices were then assigned random numbers, carbon coated and

analyzed in the electron microscope.

81

Figure 13. Gun cleaning patches from the modified gun
control experiment. 13a) Solvent soaked patch used in the
initial cleaning of the barrel, chambers and interior of the
revolver after three rounds had been fired. 13b) Second
solvent soaked patch used to clean barrel, chambers,
interior and exterior of the revolver. 13c) Final dry

patch used to clean barrel, chambers, interior and exterior
of the revolver.

 

82

As in the original experiment a high amount of GSR was
found on the investigator's hand after dry firing the
cleaned weapon (Table 11). In fact, there was no
significant difference with a 95% confidence interval, in a
comparison of the mean number of particles found per hour
between three live rounds fired and working the trigger
mechanism on the unloaded, cleaned weapon three times (Table
11).

The finding that significant amounts of GSR remain on
or in the weapon after cleaning may help explain why no
significant difference was observed between three rounds and
one round fired for the tape lift method (Table 6). If
cleaning the weapon results in high amounts of GSR still
associated with the cleaned weapon, then differences between
one or three rounds fired may be less pronounced than they
would be if all GSR had been removed in the cleaning
process.

Other researchers described cleaning the firearm
between firings (Basu et al. 1984, Brihaye et al. 1982,
Gansau and Becker 1982, Wolten et al. 1979a), but made no
mention of collecting a sample from the hand off a cleaned
"dry fired" weapon.

There was one case where the possibility of finding GSR

on a cleaned weapon was mentioned (Basu et al. 1984). This

83

Table 11. Comparison of GSR found on the hand between three
live rounds fired and "dry firing" an unloaded, cleaned
weapon three times.

 

Number of GSR particles found on the hand

 

Treatment Lower 95% C.L. Mean Upper 95% C.L.
3 rounds fired 10.8 7.9 5.7
"dry fired" 14.3 8.4 4.8

3 times

 

84

appeared in a study of suicide reconstructions where the
authors attempted to determine whether a suicide victim used
a cleaned weapon or an unclean weapon in the shooting. The
authors proposed that gun cleaning oils and grease may
retain residues and, therefore, "dry cleaned" the gun with
patches soaked in methanol to remove the contaminating
particles of GSR before test firing for reconstruction
purposes. The authors then concluded: "one is able to
estimate by a comparison of the GSR densities if the suicide
victim's hands have had excess residues over the amounts
obtained from the test shooter's hands. The presence of
these excess residues on the victim's hands was the
indication that he used an unclean gun." (Basu et al. 1984).

The observation that gun cleaning oils and grease
retain GSR particles was significant and pertinent to the
gun control findings in the present work. The conclusion
that excess residue on the suicide victim's hands compared
to the test shooter's hands was indicative of the victim
having used an unclean gun was questionable however, since
lay-people do not normally clean their guns with methanol
soaked patches to remove gun cleaning oils and grease that

may retain GSR.

85
EQHMABX_AHD_QQE§LH§IQH§

The finding of large amounts of GSR on the hand from
dry firing a cleaned weapon has serious implications in the
interpretation of finding GSR on a suspect's hand. Use of
the SEM equipped with EDX allows the investigator to
conclude that certain particles removed from the hand are
gunshot residue and nothing else, but the question still
remains as to how the GSR ended up on the hand.

The answer to the question: "How did it get there?"
has always been the Achilles heel of GSR analysis. One
would like to say that the only way for GSR to appear on a
suspect's hand is if the suspect had fired a gun.
Unfortunately, as the present study demonstrates, that may
not always be the case. The results of the present study
indicate that one must proceed with extreme caution in
making inferences from a positive GSR finding on the hand.

Future experiments should examine the following
questions: 1) How well does the cleaning of the weapon
with methanol soaked patches as described by Basu et a1.
(1984) work? 2) If GSR can be removed from a cleaned
weapon using the above procedure can significant differences
between one round and three rounds fired be determined for
the tape lift method? 3) Is GSR picked up by simply
handling a cleaned weapon without operating the trigger

mechanism?

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

BEEEBEEQEfi

Andrasko, J., and Maehly, A. C. (1977). Detection of
gunshot residue on hands by scanning electron

microscopy- I9nrnal_gf_£2ren§ic_§212nce§. 22. 279-287.

Basu, S., and Ferriss, S. (1980). A refined collection
technique for rapid search of gunshot residue particles

in the SEM. Scanning_Elesfr9n_Microsceex. 1. 375-384
and 392.

Basu, S. (1982). Formation of gunshot residues.
goutnal of Forensic Scienees, 21, 72-91.

Basu, S., Ferriss, 8., and Horn, R. (1984). Suicide
reconstruction by glue-lift of gunshot residue.

Qggrngl Qt Fotensic §01§DQ§§p 22, 843-864.

Bergman, P., Enzel, P., and Springer, E. (1988). The
detection of gunshot residue (GSR) particles on the

bottom of discharged bullets. 19n;nel_efi_Feteneie
Seiences, en, 960-968.

Brihaye, C., Machiroux, R. and Gillain, G. (1982).
Gunpowder residues detection by anodic stripping

voltammetry. E2rensic_§cien2e_lnternational. 29. 269-
276.

Burnett, B. (1989). The form of gunshot residue is

modified by target impact. iournal_2f_£orensig
Seiences, 11, 808-822.

Cowan, M. E. and Purdon, P. L. (1967). A study of the

"Paraffin Test". geurnal e: Fenensie Scienees, 12. 19-
36.

DeForest, P. R., Gaensslen, R. E., and Lee, H. C.
(1983). Forensic Science An Introduction to
Criminalistics, McGraw-Hill, Series in Criminology and
Criminal Justice, 149.

The Dermal Nitrate Test (1935). Ffil_Ley_Fnjezeenent
Bulletin. 1. 5-

86

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(13)

(19)

(20)

(21)

87

Further Observations on the Diphenylamine Test for Gun-

powder Residue (1940). FBI Law anereement Bulletin,
2' 10.

Gansau, H. and Becker, U. (1982). Semi-automatic
detection of gunshot residue (GSR) by scanning electron
microscopy and energy dispersive X-ray analysis
(SEM/EDX). anni on M' osco , 1, 107-114.

Goleb, J. A., and Midkiff, C. R. (1975). Firearms
discharge residue sample collection techniques.

{gunnel o; Forensie Seiences, 29, 701-707.

Harrison, H. C., and Gilroy, R. (1959). Firearms

discharge residues. Journal of Forensic Sciencee, A,
184-190.

Hellmiss, G., Lichtenberg, W., and Weiss, M. (1987).
Investigation of gunshot residues by means of Auger

Electron Spectroscopy. £99126; of Fotensic Sciences,
3;, 747-760.

Hoffman, C. N., (1975). Neutron activation analysis
for the detection of firearm discharge residue
collected on cotton swabs. J a o t e ssoc

ef foicial Analytieel gnenists, 5Q, 1388.

Kilty, J. W. (1975). Activity after shooting and its
effect on the retention of primer residue. Journal e:
Forensic Sciences, 29, 219-230.

Kinard, W. D., and Lundy, D. R. (1975) IN Forensic
Science, 2nd edition (Davies, G. editor), American
Chemical Society, ACS Symposium Series 13, Chapter 11.

Kinard, W. D., and Midkiff, C. R. (1978). The
application of Oxygen Plasma Ashing to gunshot residue

analysis. Journal 9: Foteneic Sciences, A;, 368-374.
Koons, R. D., Havekost, D. G., and Peters, C. A.

(1988). Determination of barium in gunshot residue
collection swabs using Inductively Coupled Plasma-
Atomic Emission Spectrometry. Qournal of Forensic
Sciences, lg, 35-41.

Krishnan, S. S. (1976). Examination of paints by trace

element analysis. goutnal of Forensic Sciences, Al,
908-916.

(22)

(23)

(24)

(25)

(25)

(27)

(28)

(29)

(30)

(31)

(32)

88

Loper, G. L., Calloway, A. R., Stamps, M. A. Wolten, G.

M., and Jones, P. F. (1981). Use of Photoluminescence

to investigate apparent suicides by firearms. IQBIDQL
, 25, 263-287.

Matricardi, V. R., and Kilty, J. W. (1977).
of gunshot residue particles from the hands

shooter. Icnrne1_cf_£creneic_§ciencee. 22.

McFarland, R. C., and McLain, M. E. (1973). Rapid
Neutron Activation Analysis for gunshot residue

determination. 1curnal_cf_£creneic_ficiencee. 18. 226-
231.

Detection
of a
725-738.

Nesbitt, R. S., Wessel, J.E., and Jones, P. F. (1976).
Detection of gunshot residue by use of the scanning

electron microscope- lcnIncl_cf_£creneic_§ciencee. 21.
595-610.

Newton, J. T. (1981). Rapid determination of antimony,
barium, and lead in gunshot residue via Automated
Atomic Absorption Spectrophotometry. Jen;nel_efi

Ecreneic. Sciences. 25. 302-312-

Pillay, K. K. S., Jester, W. A., and Fox, H. A. (1974).
New method for the collection and analysis of gunshot
residues as forensic evidence. lentnel_efi_Feteneie

Sciences. 12. 768-783-

Portis, L. C., and Tilley, R. K. (1981). A filtration
technique for gunshot residue analysis by SEM and/or

FAAS. AEIE Jonpnal, 1;, 93-99.

Rudzitis, E., Kopina, M., and Wahlgren, M. (1973).
Optimization of firearm residue detection by Neutron

Activation AnalysiS- Jcnrnel_cf_Ecreneic_§ciencee. 18.
93-100.

Rudzitis, E., and Wahlgren, M. A. (1975) IN Forensic
Science, 2nd edition (Davies, G. editor), American
Chemical Society, ACS Symposium Series 13, Chapter 10.

Rudzitis, E. (1980). Analysis of the results of
gunshot residue detection in case work. Jenznel_efi

Fonensie Seiencee, 2;, 839-646.

SAS Institute Inc. (1985). SASG User's Guide:
Statistics, Version 5 Edition. Cary, NC : SAS
Institute Inc.,.

(33)

(34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

(42)

(43)

89

Sild, E. H. and Pausak, S. (1979). Forensic

applications of SEM/EDX- Scanninc_Electrcn_nicrceccnx.
2, 185-192.

Steel, R. G., and Torrie, J. H. (1980). Principles and
procedures of statistics a biomedical approach, 2nd
edition, McGraw-Hill (1980).

Tassa, M., and Zeldes, N. (1979). Characterization of
gunshot residue particles by localizations of their
chemical constituents. acanninc_Electrcn_nicrceccnx.
2, 175-178.

Tassa, M., Leist, Y., and Steinberg, M. (1982a).
Characterization of gunshot residues by X-ray

diffraction. Jcnrnc1_cf_£creneic_§ciencee. 21.
677-683.

Tassa, M., Adan, N. Zeldes, N., and Leist, Y. (1982b).
A field kit for sampling gunshot residue particles.

Qournal of Forensie Scieneee, 21, 671-676.

Taylor, R. L., Taylor, M. S., and Noguchi, T. T.
(1979). Firearm identification by examination of

bullet fragments: an SEM/EDS study. fieenning_Fleetten
Micrceccnx. 2. 167-174-

Tillman, W. L. (1987). Automated gunshot residue
particle search and characterization. Qen;nel_efi

Fotensic Sciences, Ag, 62-71.

Wallace, J. S. and Keeley, R. H. (1979). A method for
preparing firearms residue samples for scanning

electron microsconv §canninc_Eleccrcn_Micrc_ccnx. 2.
179- -184.

Ward, D. C. (1982). Gunshot residue collection for

scanning electron microscopy. §eenning_§leetnen
Mictoscopy. ;, 1031-1036.

White, R. S. and Owens, A. D. (1987). Automation of
gunshot residue detection and analysis by scanning
electron microscopy/energy dispersive X-ray analysis

(SEM/EDX). gournal e: Fotensic Scienees, 12, 1595-
1603.

Wolten, G. M., Nesbitt, R. S., Calloway, A. R., Loper,
G. L., and Jones, P. F. (1977). Final report on
particle analysis for gunshot residues. ATR-77
(7915)-3. The Aerospace Corporation, El Segundo, CA.

(44)

(45)

(45)

(47)

(48)

9O

Wolten, G. M., Nesbitt, R. S., Calloway, A. R., Loper,
G. L., and Jones, P. F. (1979a). Particle analysis for
the detection of gunshot residue. : scanning electron
microscopy/energy dispersive X-ray characterization of

hand deposits from firing. lentnel_et_Feneneie
$21902§§. 21. 409-429.

Wolten, G. M., Nesbitt, R. S., Calloway, A. R., and
Loper, G. L. (1979b). Particle analysis for the
detection of gunshot residue. II: Occupational and
environmental particles. 2cnrne1_cf_£creneic_§cience§.
2A, 423-430.

Wolten, G. M., Nesbitt, R. S., and Calloway, A. R.
(1979c). Particle analysis for the detection of
gunshot residue. III: The case record. Jentnel_e§

Fereneic_§ciencee. 24. 864-869.

Wolten, G. M., and Nesbitt, R. S. (1980). On the
mechanism of gunshot residue particle formation.

Icnrnel_cf_Ecreneic_§ciencee_. 25. 533-545-

Zeichner, A., Foner, H. A., Dvorachek, M., Bergman, P.,
and Levin, N. (1989). Concentration techniques for the
detection of gunshot residues by scanning electron
microscopy/energy dispersive X-ray Analysis (SEM/EDX).
1cnrnel_cf_£creneic_§ciencee. 24. 312- -320-

The following material in Appendix A was distributed to
Crime Labs throughout the country by Kinderprint Co., Inc.
P.O. Box 16, Martinez, CA. in the Spring of 1988.

91

92

AEEENDIX_A

THE CONCENTRATION AND ISOLATION OF GUNSHOT RESIDUES FOR
PARTICLE ANALYSIS

BY

LOREN A. SUGARMAN, SENIOR CRIMINALIST
ORANGE COUNTY SHERIFF-CORONER DEPARTMENT
SANTA ANA, CALIFORNIA

An efficient, cost effective method of concentrating GSR
particles for SEM/EDX analysis has been developed, using
commercially available, disposable materials.

A 25mm diameter adhesive sample is concentrated to a 5mm
area on a polyester membrane filter having a 0.4 micron pore
size. Particles are released from an adhesive coated disc
of polypropylene by dissolving the adhesive with
trichloroethylene in a microcentrifuge filter tube.
Following centrifugation, the light-particle fraction is
aspirated from the top of the solvent. The density
separation is repeated using bromoform. High density
particles are deposited onto the filter by centrifuging the
unaspirated solvent through the membrane. Sample
preparation requires 15-20 minutes.

Recovery efficiency, evaluated by FAAS, was determined
relative to the total Pb and Sb from the aspirate and
filtrate fractions.

The concentration procedure reduces the searching area to 4%
of the original surface area and the density separation
removes the organics and low density debris, minimizing the
burying of GSR under other particles. Approximately 90% of
the total GSR is recovered on the membrane. The remainder
of the GSR is found in the aspirated fraction.

 

M) I 0 I I0)

93

Protective i
Shell

Vistanex I
Polypropylene ?
3" #485 I

:22“: I

Shell Cap

BAS MF-1 MICRO-FILTER

=6!-

 

 

HHHE?

 

 

 

 

Sample
Compertment

F Ilter

Subfilter (preflltel)

Filter
Support

Receiving i
Tube A

INDIVIDUAL I
COMPONENTS '-

1

94

SAMPLE PREPARATION:

1a)

b)

C)

2)

3a)

b)

4)

5a)

13)

6a)

b)

7)

Remove the polypropylene disc (with Vistanex) from the
collection stub.

The disc should be curled, adhesive inward, and
inserted into the sample compartment of the filtration
tube.

The cork should be firmly seated in the base of the
filter section.

Add 2 ml Hexane to nearly fill tube. Allow 5 minutes
for the adhesive to dissolve. Cap the filter tube and
shake vigorously to assist in dissolution.

Remove the cap and place the filter tube into a
disposable test tube so that the system is resting on
the cork.

Centrifuge at moderate rate for 5 minutes to speed the
separation. The centrifuge used for this work has a
radius of 12 cm to the bottom of the tube and is spun
at 1,100 rpm.

Aspirate off the top 1.5 ml of solvent (approximately
80% of the depth due to conical shape of the tube) to
remove the floating debris.

Refill the tube with bromoform (density 2.89)
approximately 1.5 ml.

Recap and shake vigorously to mix bromoform and Hexane.

Return the filter to test tube and centrifuge for 10
minutes at 1,100 rpm.

Aspirate off the top 1.0 ml of solvent mixture with the
floating debris.

Pull the cork, replace the receiving tube on the filter
device and centrifuge the remaining solvent through the
filter until dry (approximately 5 minutes at 1,100

rpm).

Two milliliters of OmniSolve1 grade MeOH are added to
the sample compartment and centrifuged through the
filter. (approximately 5 minutes at 1,100 rpm).

 

1Filtered for particulates (0.2 microns).

95

8) The membrane is removed with clean forceps, mounted
onto an SEM stub and carbon coated for examination.

The entire preparation procedure requires about 40-60
minutes. Samples are generally prepared in groups of four,
so preparation time averages between 10-15 minutes per
sample.

Material costs run approximately $3.00 per sample excluding
chemical costs.

96

Material Requirements for
Gunshot Residue Concentration

Aluminum Mounts: (.48 each)

Ted Pella Co. Offices and Warehouse:
Box 510 16812 Millikin Ave.
Tustin, CA 92681 Irvine, CA 92714
(714) 863-0666 (714) 557-9434
Cat. #16279-Specimen Mounts, 15/16" X 3/8", aluminum.
$480.00/1000

Covers and Caps: (.102 each)
Order: Supplier:
Riekes Container Co. Brockway
6270 Caballero Blvd. Flex Products
Buena Park, CA 90620 445 Industrial Rd.
(714) 522-8740 Carlstadt, NJ 07072

(201) 933-3030

1" X 1-1/2" 01 Shell Containers and 1" PE Plugs,
natural color. $102.00/1000

Centrifuge Tube: (2.085 each)
Bioanalytical Systems Inc.
2701 Kent Ave.
Purdue Research Park
W. Lafayette, IN 47906 (317) 463-4527

MF-l Microfilter. $29.00/pk. of 12

Membrane Filters: (.64 each)

Subfilter Supports: (.13 each) prices vary with order

size
Nucleopore Corp.

7035 Commerce Circle
Pleasanton, CA 94566 (415) 463-2530

Non-stock item: Polyester Membrane, 8.3mm dia., 0.4um pore.
$64.00/100

Non-stock item: Subfilter, 8.3mm dia., D79 type.
$13.00/100 in bulk

Culture Tubes: (.056 each)
16X 100mm disposable. Typically priced at $14.00/250

Corks: (.052 each)

XXXX Quality or better. $26.00/500

5.5mm-8.0mm size 00 from Thomas Scientific (6-8mm std. size
is too big) 4mm-6mm size 000 may be sufficient.

Teflon Sealing tape: Insignificant price)

97
Available at hardware stores.

Centrifuge:
Clay Adams Dynac II with timer and tachometer.
Horizontal head with 6" radius to bottom of tube.

Total Price: $3.54 per sample excluding chemical costs.

Chemicals: **Prices vary with supplier and size of order**
Hexane
Purified Bromoform
Methanol. OmniSolv grade from EM Science. Prefiltered
for particulates.

98

Calculation of thickness of Carbon Coating1

Carbon thickness based upon the premise that a given mass of

carbon evaporates uniformly to cover a spherical target

(e.g. the bell jar). The basic equation for this assumption

is:

weight in g/cmZ on the surface

W = M Where: W

 

mass of the evaporant

4piR2 M

R = radius or evaporant/specimen
distance

W may also be expressed = t (thickness) x p (density in
)

g/Cm

t may be considered = W
P

Typical carbon coat:

t=n=me
P P
M = 4 mg
P = 9 g/cm3
R = 10 cm

4.0 1: 10" g

(4911(10 em)2 = . o" c2

9 g/cm 9 g/cmL

= 3.5 x 10'7 cm

= 35 nm

 

1Taken from: Exercises in Electron Microscopy, a
laboratory manual for biological and medical sciences.
Hooper, G R., Baker, K. K., and Flegler, S. L. (1979). Center
for Electron Optics, Michigan State University, E. Lansing,
MI 48824.

99

1988 GUNSHOT RESIDUE ANALYSIS SURVEY
1. Please circle the type/types of analysis currently
being used in your laboratory on gunshot residue (GSR)
samples.

A) NAA B) FAAS C) SEM-EDX D) Other (Please describe)

 

 

2. Approximately how much time per week is devoted to
GSR analysis ?

3. Approximately how much time is required per
analysis

 

 

4. How frequently is your current method for analysis
of GSR challenged in court and on what grounds?

 

 

 

 

 

The following questions need only be answered by
laboratories equipped with an SEM.

5. Please indicate the make and model of your SEM.

 

6. If your microscope is equipped with X-ray analysis
capability, please indicate make and model.

 

7. If possible, please indicate how old the above
instrument is.

 

8. In the past year please estimate the number of
weeks the SEM-EDX was "down" due to mechanical or
electrical failure

 

 

9. If examining GSR by SEM-EDX, please indicate the
sample collection and processing technique used most
frequently.

100

A) The Vistanex adhesive lift, followed by sample
concentration.

 

B) A tape lift method.
C) A glue lift method.

D) Other (Please explain)

 

 

 

10. Please indicate the minimum number of particles
analyzed to confirm gunshot residue in your lab.

 

11. For GSR samples being analyzed by SEM-EDX:

A) Approximately how much time per week is devoted to
this type of analysis ?

 

B) Approximately how much time is required for a
positive (evidence of GSR) analysis ?

 

12. Please describe any recurring problems and/or
solutions you have found in either collection or analysis
of GSR by the SEM-EDX technique.

 

 

 

 

 

"Iiiil‘fliilfiiiflilfliflifS