journal of the american society of trace evidence examiners · journal of the american society of...
TRANSCRIPT
Journal of the American Society of
Trace Evidence Examiners
Volume Number ThreeIssue Number One AUGUST 2012
www.asteetrace.org
www.asteetrace.org Volume Number Three, Issue Number One August 2012 ISSN: 2156-9797
JOURNAL OF AMERICAN SOCIETY OF TRACE EVIDENCE EXAMINERS
Christopher Bommarito, M.S. ASTEE Journal Editor c/o Forensic Science Consultants, Inc. P.O. Box 514 Williamston, MI 48895 (517) 202-5971 [email protected] JASTEE Editorial Board: Vincent Desiderio, New Jersey State Police Troy Ernst, Michigan State Police Amy Michaud, ATF National Laboratory Jeremy Morris, Johnson Cty. Crime Laboratory Scott Ryland, FDLE Orlando Laboratory Bill Schneck, Washington State Police Karl Suni, Michigan State Police Michael Trimpe, Hamilton Cty. Coroner’s Office Diana Wright, Federal Bureau of Investigation
The mission of ASTEE is to encourage the exchange and dissemination of ideas and information within the field of trace evidence through improved contacts between persons and laboratories engaged in trace evidence analysis. The journal of the American Society of Trace Evidence Examiners is a peer reviewed journal dedicated to the analysis of trace evidence. All original articles published in JASTEE have been subject to double-blind peer review.
JASTEE has established a working relationship with the Scientific Working Group on Materials Analysis (SWGMAT); whereby approved SWGMAT standards may
INSIDE THIS ISSUE Forensic Analysis and Discrimination of Duct Tapes 2 Andria H. Mehltretter, M.S. and Maureen J. Bradley, Ph.D. Analysis of Bicomponent Fibers Using Confocal 21 Raman Mapping Robyn Weimer, MS , Jennifer Clary, MS, Robert Heintz,Ph. D., and Mark Wall, Ph.D. Chemical Properties of Selected Plastic-Tipped Bullets 41 Forensic Paint Examinations* Melisa C. Thompson, Cady A. Lancaster, Michele G. Banta, Crystal N. Hart, Michael D. Scanlan, and Edgard O. Espinoza * Reprinted by permission of the AFTE Journal
be published in JASTEE. These standards have been peer reviewed and approved by the SWGMAT group as a whole and thus were not subject to peer review through JASTEE.
JASTEE has also established a working arrangement with The Microscope, the journal established and edited by the McCrone Institute. Under this arrangement, articles published in JASTEE may be selected for publication in The Microscope, and vice versa.
Page 1 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
Andria H. Mehltretter1, M.S. and Maureen J. Bradley2, Ph.D.
Forensic Analysis and Discrimination of Duct Tapes*
ABSTRACT
Duct tape is a common type of evidence submitted to forensic science laboratories, due to its potential for use in illicit activities. In this study, eighty-two commercially available duct tape samples were analyzed and compared to evaluate the significance of a failure-to-discriminate result. Samples were first evaluated through examination of their physical characteristics, including, but not limited to, backing color, backing surface features, fabric pattern, scrim count, and general description of the yarns. As a result of these examinations, 99.6% of the possible comparison pairs were discriminated. The chemical compositions of the backings and adhesives of the remaining indistinguishable samples were subsequently characterized through the use of Fourier transform infrared spectroscopy, X-ray diffractometry, and scanning electron microscopy with energy dispersive spectroscopy, with additional pairs discriminated at various stages. The overall discrimination power of this series of examinations was 99.8%. Each of the remaining pairs of indistinguishable samples likely shares a common manufacturing source. Keywords: forensic science, trace evidence, duct tape, discrimination, stereomicroscopy, Fourier transform infrared spectroscopy, X-ray diffractometry, scanning electron microscopy / energy dispersive spectroscopy
1 Corresponding author: Federal Bureau of Investigation, Laboratory Division, 2501 Investigation Parkway, Room 4220, Quantico, VA 22135 2 Federal Bureau of Investigation, Laboratory Division, Quantico, VA *This work has been presented in part at the American Academy of Forensic Sciences 58th Annual Meeting. Seattle, WA. 2006. This is the FBI Laboratory Division’s publication number 12-02. Names of commercial manufacturers are provided for identification only, and inclusion does not imply endorsement of the manufacturer, or its products or services, by the FBI. The views expressed are those of the authors and do not necessarily reflect the official policy or position of the FBI or the U.S. Government.
Page 2 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape INTRODUCTION
A wide variety of tapes are available to the consumer, and as a result, many different types can be used in the commission of a crime. In fact, tapes are routinely examined by forensic laboratories in investigations involving kidnappings and homicides and construction of improvised explosive devices. One of the most frequently encountered types of tape submitted to North American forensic laboratories is duct tape. Duct tapes are composed of three constituents: a polymeric backing, an adhesive, and fabric reinforcement (scrim) between the backing and adhesive. The backing of the tape provides the color and acts as a carrier for the adhesive, which in turn provides the tack to the tape. The fabric is included to add strength and bulk to the tape as well as to affect its tearing properties. The design and construction of a duct tape depends on its specifications, its commercial end use, the processes available at the manufacturing facility, and the raw materials available. Duct tape backings can be made in a variety of ways, which leads to observed differences between tapes. A duct tape manufacturer may purchase the polymeric backing from another company, which produces it via a blown film process. Such backings appear smooth on both surfaces. If the backing is made at the tape manufacturing facility, it is likely to have dimples or indentations on its surface(s), which can arise from the rollers (calenders) or the fabric when it is added. Depending on the tape specifications and procedures at the plant, the backing thickness and width can also be modified. Regarding their composition, backings are usually polyethylene with fillers. Silver is the most common color of duct tapes, and the silver color is provided by aluminum pigments, either throughout the entire thickness of the backing or in one or more layers in the backing. Some backings have different chemicals (e.g., acrylates) added to the adhesive side to aid in cohesion of the adhesive to the backing. This is called the “tie layer.” The primary observable differences for adhesives are color and chemical composition, which generally are related characteristics. The color is determined by the elastomer (i.e., different natural rubber sources have different colors) and/or the pigments/fillers (e.g., titanium dioxide will whiten adhesives). Natural rubber-based adhesives are typically made/mixed on-site at the tape manufacturing plant, and synthetic (e.g., styrene-isoprene-styrene, acrylic) adhesives can either be prepared on-site or purchased (1). The fabric portion has the greatest number of physical features that can be evaluated. Duct tape fabrics (scrim) tend to be loose weaves (plain weave) or knits (weft-insertion). In this fabric, yarns run along the length of the tape and across the width: the former are called warp or machine direction yarns and the latter are called weft or fill yarns. The scrim count (density of these yarns) is measured by counting the number of yarns per inch in each direction. Generally, a higher scrim count indicates a higher quality of tape. The warp and fill yarns can be constructed in several different ways: twisted, textured/crimped, or straight filament, and
Page 3 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape composed of synthetic (usually polyester) fibers, cotton fibers, or a blend of the two. The fibers of the yarns may also exhibit fluorescence if optical brighteners are present. As of 2005, there were well over 150 different reference numbers of duct tape found in the United States, produced by approximately four or five manufacturers (1). Because of all the variations possible, duct tape comparisons can be valuable evidence in criminal investigations, and forensic laboratories have been conducting these examinations for decades (2-17). Their value was documented in a 1998 study by Smith in which fifty-one duct tape samples were analyzed and compared (7). The study demonstrated variability in construction and composition between manufacturers and even within the same manufacturer. This study aimed to expand on Smith’s study and to evaluate the significance of failing to differentiate samples. The techniques used in this evaluation were physical examinations, Fourier transform infrared spectroscopy (FTIR), X-ray diffractometry (XRD), and scanning electron microscopy with energy dispersive spectroscopy (SEM/EDS). Although additional techniques can be applied to duct tape analysis, based on the discrimination obtained through the reported combination of techniques, this analytical scheme is routinely employed in the authors’ laboratory. MATERIALS AND METHODS Tape Collection The tape collection for this study consisted of 82 samples purchased by the FBI between 1993 and 2005 at common retail stores and marketed as general purpose or economy grade, and covered a range of manufacturers and distributors. The same tapes were reported on in a previous study (10). Table 1 provides the available manufacturer/product information. Once tapes are manufactured, many of them are sold to various distributors, who may resell them under different brand names or labels (1). As a result, different rolls of a single duct tape product may be labeled and packaged in more than one way. Physical examinations Physical characteristics of the tapes were recorded during visual and stereomicroscopical evaluations. The characteristics observed included backing and adhesive color, backing surface features and layer structure, width, and backing thickness. The fabric characteristics observed were weave/knit pattern, yarn description (e.g., twisted), yarn composition (e.g., synthetic or cotton), fluorescence, and scrim count. For the backing and adhesive color, the observations were conducted with the unaided eye. Only distinct color differences of the adhesives were considered significant, due to the typical
Page 4 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape condition of adhesives in casework samples (i.e., contamination by dirt or body fluids). Surface features of the backings were observed unaided and with a stereomicroscope and were described as smooth or dimpled (calendering marks). To determine the layer structure of the backings, thin cross-sections were taken and viewed with transmitted light (10). Width measurements were taken with a ruler to the nearest 0.5 mm. To measure thickness of the backing on each sample, the adhesive and fabric were removed with hexane or chloroform and the backing was placed between the two faces of a digital micrometer. A minimum of ten areas were measured, and the values were recorded to the nearest 0.05 mil (1 mil = 1/1000 inch) and averaged. A significant difference between pristine tapes (not stretched, deformed, or highly contaminated) is generally considered to be a width difference greater than 1.0 mm (personal communication with Mark Byrne, Technical Manager at Shurtape Technologies, on December 9, 2011) or a thickness difference greater than 10% (personal communication with Jerry Serra, consultant, on December 30, 2011). To best visualize the fabric, the adhesive was removed with hexane or chloroform. The weave/knit pattern and general yarn description were observed by stereomicroscopy. For the latter, the yarns were documented as twisted, textured, or straight filament. Using transmission microscopy, the yarns were classified as being synthetic, cotton, or a blend of synthetic and cotton fibers. Yarn fluorescence was observed under long wave UV light (λ = 366 nm). The scrim count was measured using a ruler, counted per square inch, and recorded as number of warp yarns / number of fill yarns. Scrim counts of +/-1 are generally acceptable in the manufacturing of duct tape products (personal communications with Jerry Serra and John Johnston, consultants, on March 27, 2012), so for this study, a count difference of +/-1 in either direction did not result in samples being discriminated. In other words, a significant difference was considered to be a count difference of two or more in either the warp or fill direction. Additional fabric/fiber examinations are generally conducted in the authors’ laboratory in casework, but for this study, examinations were limited to those described. FTIR Adhesive samples were smeared onto either one diamond window of a compression cell (Thermo Scientific, Waltham, MA) or a KBr disc and analyzed in the transmission mode using a Continuum microscope attached to a Nicolet Nexus 670 or 6700 FTIR E.S.P. spectrometer with a MCT/A detector (4000-650 cm-1) (Thermo Nicolet, Madison, WI). The resolution was 4 cm-1, the aperture was approximately 100x100 μm, and the number of scans was 128.
Backing samples were cleaned with hexane and each side of the backings was analyzed using a Dura SamplIR ATR (SensIR Technologies, Danbury, CT) attached to a Nicolet Magna 560 FTIR
Page 5 of 49
Tabl
e 1.
Sam
ple
Info
rmat
ion
and
Phys
ical
Cha
ract
eris
tics
Tape
N
umbe
r M
anuf
actu
rer /
Pro
duct
B
acki
ng
Col
or
Bac
king
Su
rfac
e Fi
lm T
hick
ness
(m
ils)
Wid
th
(mm
)
Bac
king
La
yer
Stru
ctur
e
(Mic
rosc
opy
only
)
Adh
esiv
e C
olor
Sc
rim T
ype
Yar
n D
escr
iptio
n
(w -
f)
Scrim
C
ount
(w
/f)
Fluo
resc
ence
Y
arn
Com
posi
tion
(w
- f)
1 PE
Tar
paul
in R
epai
r Tap
e si
lver
di
mpl
ed
5.3
48.0
si
ngle
be
ige
plai
n w
eave
tw
iste
d tw
iste
d 35
/ 29
no
ne
synt
hetic
-co
tton
synt
hetic
- co
tton
2 C
ante
ch
silv
er
smoo
th
3.3
48.0
si
ngle
of
f whi
te
wef
t in
serti
on
filam
ent
text
ured
19
/ 8
none
sy
nthe
tic
synt
hetic
3 3M
Hom
e an
d Sh
op
silv
er
smoo
th
2.6
48.0
cl
ear,
silv
er,
clea
r be
ige
plai
n w
eave
te
xtur
ed
text
ured
24
/ 7
none
sy
nthe
tic
sunt
hetic
4 3M
Pro
fess
iona
l HV
AC
si
lver
sm
ooth
2.
7 49
.0
clea
r, si
lver
, cl
ear
beig
e pl
ain
wea
ve
text
ured
te
xtur
ed
29 /
8 no
ne
synt
hetic
sy
nthe
tic
5 3M
Tar
tan
Util
ity
silv
er
smoo
th
2.0
48.5
cl
ear,
silv
er,
clea
r be
ige
plai
n w
eave
te
xtur
ed
text
ured
20
/ 7
none
sy
nthe
tic
synt
hetic
6 3M
All-
wea
ther
si
lver
sm
ooth
3.
1 48
.0
clea
r, si
lver
, cl
ear
beig
e pl
ain
wea
ve
text
ured
te
xtur
ed
29 /
10
none
sy
nthe
tic
synt
hetic
7 Fi
x-It
silv
er
dim
pled
5.
1 48
.5
sing
le
clea
r pl
ain
wea
ve
twis
ted
twis
ted
19 /
9 w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
8 D
egel
si
lver
di
mpl
ed
4.2
48.5
cl
ear a
nd
silv
er
med
ium
gr
ay
plai
n w
eave
tw
iste
d te
xtur
ed
19 /
7 w
sy
nthe
tic-
cotto
n sy
nthe
tic
9 Po
lar T
ape
911
silv
er
dim
pled
3.
7 50
.0
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
twis
ted
30 /
14
none
sy
nthe
tic
synt
hetic
10
Kor
ea
silv
er
dim
pled
4.
8 48
.0
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
twis
ted
22 /
14
none
sy
nthe
tic
synt
hetic
11
Inte
rtape
, Vel
eur P
lus,
6945
si
lver
sm
ooth
2.
2 47
.5
sing
le
light
gra
y w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
8 w
& f
synt
hetic
sy
nthe
tic
12
Man
co, D
uck,
S10
si
lver
di
mpl
ed
4.1
50.0
si
ngle
of
f whi
te
plai
n w
eave
tw
iste
d fil
amen
t 25
/ 12
w
sy
nthe
tic-
cotto
n sy
nthe
tic
13
Shur
tape
PC
600
silv
er
dim
pled
5.
1 48
.5
thin
cle
ar
and
silv
er
off w
hite
w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
8 no
ne
synt
hetic
sy
nthe
tic
14
3M H
ighl
and
6969
si
lver
sm
ooth
3.
0 50
.5
sing
le
light
gra
y pl
ain
wea
ve
twis
ted
twis
ted
23 /
10
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
15
3M A
C a
nd V
entil
atin
g 13
3NA
si
lver
sm
ooth
2.
7 47
.5
sing
le
med
ium
gr
ay
wef
t in
serti
on
filam
ent
text
ured
19
/ 12
w
& f
synt
hetic
sy
nthe
tic
16
3M A
C a
nd V
entil
atin
g 13
1NA
si
lver
sm
ooth
2.
6 48
.5
sing
le
light
gra
y w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
12
w &
f sy
nthe
tic
synt
hetic
17
Tyco
Nas
hua
Patc
hing
and
M
endi
ng
silv
er
smoo
th
2.1
51.0
si
ngle
lig
ht g
ray
plai
n w
eave
tw
iste
d tw
iste
d 20
/ 9
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
18
Poly
ken
Irre
gula
r Mul
tipur
pose
si
lver
di
mpl
ed
5.5
50.5
lig
ht si
lver
an
d m
ediu
m
silv
er
med
ium
gr
ay
plai
n w
eave
tw
iste
d te
xtur
ed
26 /
13
w
synt
hetic
-co
tton
synt
hetic
19
Tyco
Gen
eral
Pur
pose
700
371
silv
er
dim
pled
3.
3 49
.5
clea
r and
si
lver
m
ediu
m
gray
pl
ain
wea
ve
twis
ted
text
ured
19
/ 8
w
synt
hetic
-co
tton
synt
hetic
20
Nas
hua
Gen
eral
Pur
pose
285
-4
silv
er
smoo
th
2.2
48.5
si
ngle
w
hite
pl
ain
wea
ve
twis
ted
text
ured
20
/ 8
w
synt
hetic
-co
tton
synt
hetic
Page 6 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
Tabl
e 1.
Sam
ple
Info
rmat
ion
and
Phys
ical
Cha
ract
eris
tics
(con
tinue
d)
Tape
N
umbe
r M
anuf
actu
rer /
Pro
duct
B
acki
ng
Col
or
Bac
king
Su
rfac
e Fi
lm T
hick
ness
(m
ils)
Wid
th
(mm
)
Bac
king
La
yer
Stru
ctur
e
(Mic
rosc
opy
only
)
Adh
esiv
e C
olor
Sc
rim T
ype
Yar
n D
escr
iptio
n
(w -
f)
Scrim
C
ount
(w
/f)
Fluo
resc
ence
Y
arn
Com
posi
tion
(w
- f)
21
Nat
iona
l 181
-6
silv
er
smoo
th
2.2
49.5
si
ngle
w
hite
pl
ain
wea
ve
twis
ted
text
ured
20
/ 8
w
synt
hetic
-co
tton
synt
hetic
22
Anc
hor P
rem
ium
Gra
de 9
602
silv
er
smoo
th
2.7
51.0
si
ngle
m
ediu
m
gray
w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
8 w
& f
synt
hetic
sy
nthe
tic
23
Tape
-It
silv
er
smoo
th
3.5
48.5
si
ngle
m
ediu
m
gray
pl
ain
wea
ve
twis
ted
text
ured
20
/ 8
w
synt
hetic
-co
tton
synt
hetic
24
Out
door
Res
earc
h G
ear -
REI
31
300
silv
er
smoo
th
4.7
19.0
si
ngle
m
ediu
m
gray
pl
ain
wea
ve
twis
ted
text
ured
20
/ 11
w
sy
nthe
tic-
cotto
n sy
nthe
tic
25
tesa
si
lver
sm
ooth
2.
7 52
.0
sing
le
light
gra
y pl
ain
wea
ve
twis
ted
twis
ted
22 /
10
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
26
tesa
Gen
eral
Pur
pose
214
10
silv
er
smoo
th
2.2
49.0
si
ngle
lig
ht g
ray
plai
n w
eave
tw
iste
d tw
iste
d 23
/ 10
w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
27
tesa
Ene
rgy
Savi
ng 1
241N
si
lver
sm
ooth
2.
2 51
.0
sing
le
beig
e w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
10
w &
f sy
nthe
tic
synt
hetic
28
Silv
er C
law
147
12
silv
er
smoo
th
2.1
38.0
si
ngle
lig
ht g
ray
wef
t in
serti
on
filam
ent
text
ured
19
/ 8
w &
f sy
nthe
tic
synt
hetic
29
Man
co In
dust
rial 3
158
silv
er
dim
pled
5.
1 48
.5
sing
le
beig
e pl
ain
wea
ve
twis
ted
text
ured
30
/ 15
w
sy
nthe
tic-
cotto
n sy
nthe
tic
30
Man
co In
dust
rial 3
157
silv
er
smoo
th
2.6
48.0
si
ngle
m
ediu
m
gray
pl
ain
wea
ve
twis
ted
filam
ent
24 /
12
w
synt
hetic
-co
tton
synt
hetic
31
Serv
iSta
r Pro
fess
iona
l 333
32
silv
er
dim
pled
4.
7 50
.5
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
filam
ent
25 /
12
w
synt
hetic
-co
tton
synt
hetic
32
Prid
e-C
hina
si
lver
di
mpl
ed
6.1
48.5
si
ngle
w
hite
pl
ain
wea
ve
twis
ted
twis
ted
34 /
30
none
sy
nthe
tic-
cotto
n sy
nthe
tic
33
Uni
ted
Util
ity F
D18
145
silv
er
smoo
th
2.1
46.0
si
ngle
lig
ht g
ray
wef
t in
serti
on
filam
ent
text
ured
19
/ 8
none
sy
nthe
tic
synt
hetic
34
Man
co E
xtre
me
Duc
k Ta
pe
fluor
esce
nt
oran
ge
smoo
th
not m
easu
red
- fa
bric
em
bedd
ed in
ba
ckin
g
46.5
si
ngle
be
ige
plai
n w
eave
tw
iste
d tw
iste
d 68
/ 44
n
one
cotto
n co
tton
35
Anc
hor C
ontin
enta
l Sta
ge T
ape
blac
k sm
ooth
7.
7 48
.5
clea
r and
bl
ack
whi
te
plai
n w
eave
tw
iste
d tw
iste
d 23
/ 16
w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
36
AB
C C
o.
blac
k sm
ooth
2.
9 76
.0
blac
k, g
ray,
cl
ear
blac
k pl
ain
wea
ve
twis
ted
text
ured
20
/ 9
w
synt
hetic
-co
tton
synt
hetic
Page 7 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
Tabl
e 1.
Sam
ple
Info
rmat
ion
and
Phys
ical
Cha
ract
eris
tics
(con
tinue
d)
Tape
N
umbe
r M
anuf
actu
rer /
Pro
duct
B
acki
ng
Col
or
Bac
king
Su
rfac
e Fi
lm T
hick
ness
(m
ils)
Wid
th
(mm
)
Bac
king
La
yer
Stru
ctur
e
(Mic
rosc
opy
only
)
Adh
esiv
e C
olor
Sc
rim T
ype
Yar
n D
escr
iptio
n
(w -
f)
Scrim
C
ount
(w
/f)
Fluo
resc
ence
Y
arn
Com
posi
tion
(w
- f)
37
Clin
g-A
us R
B45
si
lver
di
mpl
ed
4.8
48.0
si
ngle
of
f whi
te
plai
n w
eave
tw
iste
d te
xtur
ed
30 /
15
w
synt
hetic
-co
tton
synt
hetic
38
Clin
g-A
us R
B51
si
lver
di
mpl
ed
4.0
48.0
sing
le,
poss
ibly
de
lam
inat
ing
dow
n ce
nter
beig
e pl
ain
wea
ve
twis
ted
twis
ted
29 /
14
none
sy
nthe
tic
synt
hetic
39
Uni
ted
Dol
lar G
ener
al 2
33
silv
er
smoo
th
2.7
47.5
si
ngle
lig
ht g
ray
wef
t in
serti
on
filam
ent
text
ured
19
/ 12
no
ne
synt
hetic
sy
nthe
tic
40
Inte
rtape
691
0 si
lver
sm
ooth
2.
2 47
.5
sing
le
light
gra
y w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
8 no
ne
synt
hetic
sy
nthe
tic
41
Tyco
Nas
hua
FR33
3 si
lver
w/
writ
ing
smoo
th
3.3
48.5
si
ngle
, but
pr
inte
d lig
ht g
ray
plai
n w
eave
tw
iste
d tw
iste
d 23
/ 18
w
& f
synt
hetic
-co
tton
synt
hetic
42
Man
co S
kinn
y si
lver
di
mpl
ed
5.3
19.0
si
ngle
be
ige
wef
t in
serti
on
filam
ent
text
ured
26
/ 10
no
ne
synt
hetic
sy
nthe
tic
43
Wal
mar
t Mai
nsta
ys P
roje
ct
silv
er
smoo
th
1.9
48.0
si
ngle
ta
n pl
ain
wea
ve
twis
ted
text
ured
18
/ 9
w
synt
hetic
-co
tton
synt
hetic
44
Fros
t Kin
g si
lver
sm
ooth
2.
2 48
.5
sing
le
light
gra
y pl
ain
wea
ve
filam
ent
text
ured
23
/ 10
no
ne
synt
hetic
sy
nthe
tic
45
Tape
-It D
60
whi
te
smoo
th
1.9
47.5
si
ngle
be
ige
plai
n w
eave
fil
amen
t te
xtur
ed
23 /
10
none
sy
nthe
tic
synt
hetic
46
Anc
hor P
ocke
t Duc
t Tap
e si
lver
sm
ooth
2.
4 50
.5
sing
le
tan
wef
t in
serti
on
filam
ent
text
ured
19
/ 12
no
ne
synt
hetic
sy
nthe
tic
47
Inte
rtape
w
hite
sm
ooth
2.
2 49
.0
sing
le
light
gra
y w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
8 no
ne
synt
hetic
sy
nthe
tic
48
Inte
rtape
Pro
Gra
de 9
602
silv
er
smoo
th
2.6
47.0
si
ngle
lig
ht g
ray
wef
t in
serti
on
filam
ent
text
ured
19
/ 12
no
ne
synt
hetic
sy
nthe
tic
49
Pact
ape
A92
10
red
smoo
th
2.5
49.0
si
ngle
lig
ht g
ray
plai
n w
eave
tw
iste
d tw
iste
d 23
/ 17
w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
50
Ace
All
Purp
ose
4289
7 si
lver
di
mpl
ed
4.5
47.5
si
ngle
be
ige
wef
t in
serti
on
filam
ent
text
ured
19
/ 8
none
sy
nthe
tic
synt
hetic
51
Ace
Pro
Gra
de 4
2911
w
hite
sm
ooth
2.
6 49
.5
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
text
ured
25
/ 12
w
sy
nthe
tic-
cotto
n sy
nthe
tic
52
Hom
e H
ardw
are
- Can
ada
silv
er
smoo
th
2.0
49.5
si
ngle
ta
n pl
ain
wea
ve
twis
ted
filam
ent
18 /
9 w
sy
nthe
tic-
cotto
n sy
nthe
tic
53
Tago
- C
anad
a si
lver
sm
ooth
2.
2 48
.0
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
twis
ted
19 /
10
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
Page 8 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
Tabl
e 1.
Sam
ple
Info
rmat
ion
and
Phys
ical
Cha
ract
eris
tics
(con
tinue
d)
Tape
N
umbe
r M
anuf
actu
rer /
Pro
duct
B
acki
ng
Col
or
Bac
king
Su
rfac
e Fi
lm T
hick
ness
(m
ils)
Wid
th
(mm
)
Bac
king
La
yer
Stru
ctur
e
(Mic
rosc
opy
only
)
Adh
esiv
e C
olor
Sc
rim T
ype
Yar
n D
escr
iptio
n
(w -
f)
Scrim
C
ount
(w
/f)
Fluo
resc
ence
Y
arn
Com
posi
tion
(w
- f)
54
Ren
frew
- C
anad
a si
lver
sm
ooth
2.
2 48
.5
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
twis
ted
20 /
10
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
55
Hen
kel -
Duc
k si
lver
w/
whi
te o
n un
ders
ide
smoo
th
3.1
50.5
w
hite
, silv
er,
clea
r be
ige
wef
t in
serti
on
filam
ent
text
ured
19
/ 8
none
sy
nthe
tic
synt
hetic
56
Adv
ance
Tap
es
silv
er
dim
pled
4.
8 50
.0
sing
le
beig
e w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
15
none
sy
nthe
tic
synt
hetic
57
Prid
e D
olla
r Gen
eral
bl
ack
smoo
th
6.1
48.0
si
ngle
w
hite
pl
ain
wea
ve
twis
ted
twis
ted
34 /
30
none
sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
58
Pact
ape
A92
10
teal
sm
ooth
2.
2 51
.0
sing
le
light
gra
y pl
ain
wea
ve
twis
ted
text
ured
20
/ 9
w
synt
hetic
-co
tton
synt
hetic
59
Inte
rtape
AC
50
blac
k sm
ooth
6.
4 48
.0
clea
r, bl
ack,
cl
ear
light
gra
y pl
ain
wea
ve
twis
ted
twis
ted
38 /
20
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
60
Fros
t Kin
g T9
03
silv
er
smoo
th
1.8
51.0
po
ssib
le th
in
clea
r lay
er
light
gra
y pl
ain
wea
ve
text
ured
tw
iste
d 17
/ 9
none
sy
nthe
tic
synt
hetic
-co
tton
61
Shur
tape
si
lver
sm
ooth
2.
6 48
.5
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
text
ured
28
/ 12
w
sy
nthe
tic-
cotto
n sy
nthe
tic
62
Pana
cea
6003
1 gr
een
smoo
th
5.0
6.0
sing
le
off w
hite
pl
ain
wea
ve
twis
ted
text
ured
(2
0) /
11
w
synt
hetic
-co
tton
synt
hetic
63
Tape
-It D
A10
si
lver
sm
ooth
2.
7 48
.0
sing
le
light
gra
y pl
ain
wea
ve
twis
ted
twis
ted
18 /
7 w
& f
synt
hetic
-co
tton
synt
hetic
64
Tape
-It D
A10
w
hite
sm
ooth
2.
2 48
.5
sing
le
off w
hite
pl
ain
wea
ve
filam
ent
text
ured
22
/ 9
none
sy
nthe
tic
synt
hetic
65
tesa
Gen
eral
Pur
pose
214
02
silv
er
smoo
th
1.9
48.5
si
ngle
lig
ht g
ray
plai
n w
eave
tw
iste
d tw
iste
d 22
/ 10
w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
66
tesa
Gen
eral
Pur
pose
214
02
silv
er
smoo
th
2.0
48.5
si
ngle
lig
ht g
ray
plai
n w
eave
tw
iste
d tw
iste
d 22
/ 10
w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
67
Duc
tTite
, Tite
Seal
si
lver
sm
ooth
3.
0 50
.5
sing
le
light
gra
y pl
ain
wea
ve
twis
ted
twis
ted
23
/ 12
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
68
Tuck
ST4
4 ca
mou
flage
sm
ooth
3.
3 49
.5
beig
e, c
amo
prin
t lig
ht g
ray
plai
n w
eave
tw
iste
d tw
iste
d 3
6 / 2
4 w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
69
Tyco
960
763
cam
oufla
ge
smoo
th
2.4
48.5
ca
mo
film
, cl
ear
med
ium
gr
ay
plai
n w
eave
tw
iste
d te
xtur
ed
20 /
14
w
synt
hetic
-co
tton
synt
hetic
70
Cam
o D
uct T
ape
8220
ca
mou
flage
sm
ooth
2.
4 48
.5
cam
o fil
m,
clea
r m
ediu
m
gray
pl
ain
wea
ve
twis
ted
text
ured
2
0 / 1
4 w
sy
nthe
tic-
cotto
n sy
nthe
tic
Page 9 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
Tabl
e 1.
Sam
ple
Info
rmat
ion
and
Phys
ical
Cha
ract
eris
tics
(con
tinue
d)
Tape
N
umbe
r M
anuf
actu
rer /
Pro
duct
B
acki
ng
Col
or
Bac
king
Su
rfac
e Fi
lm T
hick
ness
(m
ils)
Wid
th
(mm
)
Bac
king
La
yer
Stru
ctur
e
(Mic
rosc
opy
only
)
Adh
esiv
e C
olor
Sc
rim T
ype
Yar
n D
escr
iptio
n
(w -
f)
Scrim
C
ount
(w
/f)
Fluo
resc
ence
Y
arn
Com
posi
tion
(w
- f)
71
Uni
ted
3958
57
silv
er
smoo
th
1.8
51.0
si
ngle
m
ediu
m
gray
pl
ain
wea
ve
text
ured
tw
iste
d 18
/ 9
non
e sy
nthe
tic
synt
hetic
-co
tton
72
Uni
ted
10s2
sh
silv
er
smoo
th
4.2
51.5
cl
ear a
nd
silv
er
med
ium
gr
ay
plai
n w
eave
tw
iste
d te
xtur
ed
20 /
10
w
synt
hetic
-co
tton
synt
hetic
73
Vic
tor v
302
blac
k sm
ooth
1.
8 49
.5
sing
le
light
gra
y pl
ain
wea
ve
twis
ted
twis
ted
19 /
9 w
& f
synt
hetic
-co
tton
synt
hetic
-co
tton
74
True
Val
ue D
uck
Mul
ti U
se A
ll Pu
rpos
e si
lver
di
mpl
ed
4.8
47.5
th
in c
lear
an
d si
lver
of
f whi
te
wef
t in
serti
on
filam
ent
text
ured
19
/ 8
none
sy
nthe
tic
synt
hetic
75
Hen
kel -
Duc
k dx
660
silv
er
dim
pled
3.
9 48
.0
sing
le
off w
hite
pl
ain
wea
ve
filam
ent
text
ured
20
/ 7
none
sy
nthe
tic
synt
hetic
76
Gen
eral
Pur
pose
- C
hina
si
lver
sm
ooth
6.
0 48
.0
sing
le
whi
te
pla
in
wea
ve
twis
ted
twis
ted
36 /
32
w &
f sy
nthe
tic-
cotto
n sy
nthe
tic-
cotto
n
77
Inte
rtape
All-
wea
ther
si
lver
sm
ooth
2.
2 48
.5
sing
le
beig
e w
eft
inse
rtion
fil
amen
t te
xtur
ed
19 /
9 no
ne
synt
hetic
sy
nthe
tic
78
Inte
rtape
Util
ity
silv
er
smoo
th
2.0
48.5
si
ngle
of
f whi
te
wef
t in
serti
on
filam
ent
text
ured
19
/ 9
none
sy
nthe
tic
synt
hetic
79
Mad
e in
Pol
and
silv
er
dim
pled
5.
5 46
.5
sing
le
whi
te
pla
in
wea
ve
text
ured
te
xtur
ed
27 /
11
none
sy
nthe
tic
synt
hetic
80
Man
co U
tility
dx6
60
silv
er
dim
pled
3.
0 48
.0
sing
le
beig
e pl
ain
wea
ve
filam
ent
text
ured
2
0 / 7
no
ne
synt
hetic
sy
nthe
tic
81
3M H
ome
Off
ice
silv
er
smoo
th
2.7
48.5
cl
ear,
silv
er,
clea
r of
f whi
te
plai
n w
eave
te
xtur
ed
text
ured
25
/ 7
none
sy
nthe
tic
synt
hetic
82
3M H
ome
and
Shop
si
lver
sm
ooth
2.
5 48
.0
clea
r and
si
lver
of
f whi
te
plai
n w
eave
te
xtur
ed
text
ured
25
/ 7
none
su
nthe
tic
sunt
hetic
Page 10 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape (Thermo Nicolet, Madison, WI) with a DTGS KBr detector (4000-650 cm-1). The resolution was 4 cm-1, and the number of scans was 32.
XRD
Samples were prepared in two ways: intact and backings only (following removal of adhesive and scrim). Each was mounted on a silicon wafer in a sample holder. Analysis was performed on a PANalytical X’Pert Pro MPD X-ray diffractometer (Westborough, MA), with Cu Kα radiation, operated at 45kV and 40 mA, scanning continuously between 8 and 80° 2Θ with a step size of 0.0170° 2Θ, and using a 10 mm beam mask. Total analysis time was approximately 8 minutes.
SEM/EDS
Backing samples were attached to a pyrolytic carbon planchet using their own adhesives, grounded with carbon paint, and carbon coated by vacuum evaporation. Adhesive samples were smeared onto a pyrolytic carbon planchet and carbon coated by vacuum evaporation. Analysis was performed using a tungsten filament source on either a JEOL JSM-6300 (JEOL, Peabody, MA) SEM with an Oxford ISIS L300 EDS (Oxfordshire, United Kingdom) or a Camscan MV2300 (Tescan, Cranberry Township, PA) with a 4pi Analysis EDS (Durham, NC). SEM conditions were as follows: a magnification of approximately 50X, working distance of approximately 15 mm, take-off angle of approximately 30°, and accelerating voltage of 25 kV. Both EDS systems were operated with a dead time of approximately 30% and live counting time of 200 s.
Data Evaluation
For the physical characteristics, the data was entered into Microsoft® Excel and sorted according to the various characteristics. Samples that remained indistinguishable were analyzed and compared by FTIR, with data review by two examiners. Samples that continued to be indistinguishable were analyzed and evaluated by both XRD and SEM/EDS.
Discrimination Calculations
The total number of comparison pairs possible from a population of 82 samples is 3321,
calculated with the formula 2)1( −nn
, where n is the number of samples (18). Following the physical examinations and the entire analytical scheme, the number of comparison pairs for each indistinguishable group was calculated using the same formula and subsequently summed across the groups to provide the total number of indistinguishable pairs. The percentage of pairs that were discriminated, which is equivalent to the discrimination power (DP), was then calculated as follows:
DP = % of pairs discriminated =
Page 11 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
100% x ⎟⎟⎠
⎞⎜⎜⎝
⎛−
pairscomparisonofnumbertotalpairsishableindistinguofnumber1
An example follows in the results of the physical examinations section.
RESULTS
Microscopical examinations and physical examinations
The physical characteristics observed/measured for these samples are detailed in Table 1. Most of the tape samples had backings that were silver in color, but black, white, and camouflage were also observed for several tapes. The adhesives covered a range of shades of white, gray, and beige. Roughly three-quarters of the tape backings had smooth surfaces and one-quarter had dimpled surfaces. Figure 1 depicts one example of each. Most of the tape widths were between 47.5 and 51.0 mm; the narrowest and widest widths were 6.0 mm and 76.0 mm, respectively. The range of backing thicknesses was from 1.8 to 7.7 mils. Nearly three-quarters of the tape backings appeared single-layered when a cross-section was viewed (10).
Figure 1: Sample 11 (left) has a smooth backing surface, and Sample 12 (right) has a dimpled backing surface.
A variety of fabrics was also observed. About 75% of the tapes had a plain weave pattern, with the remaining being weft-insertion. For the former, an over-under pattern is observed, whereas for the latter, a chain-stitch pattern can be seen. In the tapes with a plain weave pattern, a variety of different combinations of twisted, textured, and straight filament yarns was apparent. Figure 2 shows the difference between plain weave and weft-insertion, with three examples of plain weave construction. Yarn composition and fluorescence both varied with no obvious correlation between them. The fabric characteristic that varied the most was scrim count, ranging from 17/9 or 18/7 to 68/44, with many possible combinations in between. Half
Page 12 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape of the warp values were between 18 and 20, and half of the fill values were between 8 and 10; the warp count was always higher than the fill count.
Figure 2: Four different scrim patterns, oriented so that the warp yarns run left to right. Clockwise from top left: plain weave with twisted yarns in both directions (Sample 67), plain weave with twisted yarns in the warp direction and straight filament yarns in the fill direction (Sample 52), plain weave with textured yarns in both directions (Sample 4), and weft‐insertion (Sample 40).
Due to the variety of combinations of the aforementioned characteristics, there were only eight groups of samples that remained indistinguishable following comparisons of the physical features of these tapes. The eight groups were as follows: 3 and 81; 12 and 31; 13 and 74; 26, 65, and 66; 39 and 48; 40, 77, and 78; 45 and 64; and 53 and 54. This resulted in 12 total pairs of indistinguishable samples, with a discrimination power (DP) of 99.6%, calculated as follows:
Page 13 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢
⎣
⎡
⎟⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜⎜
⎝
⎛−
+−
+−
+−
+−
+−
+−
+−
−×3321
2)12(2
2)12(2
2)13(3
2)12(2
2)13(3
2)12(2
2)12(2
2)12(2
1%100
FTIR
The adhesives and backings of any samples that remained indistinguishable in observed and measured physical properties were compared by FTIR. As a result, Samples 12 and 31 were differentiated from each other, and Sample 78 differed from Samples 40 and 77. As seen in Figures 3 and 4, respectively, the differences could be attributed to the presence of kaolin in the adhesive of Sample 12, and the presence of dolomite versus calcite in the adhesives of 78 and 40/77. The rest of the adhesive samples remained indistinguishable after FTIR spectroscopy. None of the samples in these groups were discriminated by ATR analysis of the backing.
Figure 3: FTIR spectral overlay of two adhesives that differ. The peaks present in Sample 12 (3700‐3600 and 1100‐1000 cm‐1 ranges) that are absent in Sample 31 are due to kaolin.
Page 14 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
Figure 4: FTIR spectral display of two adhesives that are indistinguishable and one that differs. Calcite peaks are observed at 875 and 711/712 cm‐1 in Samples 40 and 77, and dolomite peaks are observed at 881 and 729 cm‐1 in Sample 78.
XRD
The remaining indistinguishable samples were compared by XRD, which successfully discriminated two additional pairs: Samples 39 and 48 and Samples 53 and 54. XRD analysis indicated the presence of talc in the backings of Samples 48 and 54, but not in Samples 39 and 53. Figure 5 demonstrates this difference for Samples 39 and 48.
Page 15 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape
Figure 5: XRD diffraction patterns of two backings that differ. The peaks present in Sample 48 (top) that are absent in Sample 39 (bottom) are due to talc (T). Polyethylene (P) and calcite (C) are observed in both samples.
SEM/EDS
The same samples that were compared by XRD were also compared by SEM/EDS, but no discriminating characteristics were readily observed.
Figure 6: SEM/EDS spectral overlay (displayed in square root scale) of the backings of Samples 39 and 48, demonstrating that magnesium (Mg) is not readily apparent in Sample 48, despite talc being a major component of the XRD pattern. Mg would be observed to the left of aluminum (Al).
Page 16 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape Overall Discrimination
Following all examinations, seven pairs (five groups) of samples remained indistinguishable: 3 and 81; 13 and 74; 26, 65, and 66; 40, 77; and 45 and 64. The resulting overall discrimination was calculated as 99.8%.
Of the samples that were not discriminated, one pair was from the same manufacturer: Samples 40 and 77 were both Intertape products.
Samples 45 and 64 were both distributed by Tape-It, which reportedly does not manufacture the duct tape it distributes (personal communication with Arnold Rabinowitz, President of Tape-It, Inc., on April 20, 2011). It is quite possible that these tapes share a common manufacturer, but this information is not known.
Regarding Samples 13 and 74, one was made by Shurtape and the other was labeled as True Value, which does not make its own tape products; Shurtape is known to have manufactured tapes for True Value (personal communication with Mark Byrne, Technical Manager at Shurtape Technologies, Inc., on October 20, 2010).
One pair of samples (3 and 81) was labeled as 3M products (from Canada), and 3M is known to be both a duct tape manufacturer and distributor. Likewise, tesa [sic] is a known manufacturer and distributor, and Samples 26, 65, and 66 are all tesa-labeled tapes. Therefore, the samples within each of these latter two groups could have come from the same source.
DISCUSSION
The comparison of the physical characteristics of duct tape yielded an impressive discrimination power of 99.6%. Therefore, the vast majority of unrelated tapes were discriminated at the physical examination stage. Most of these characteristics can be evaluated with simple laboratory tools: a microscope, scalpel, tweezers, ruler, and solvents.
Tape products change over time due to market demands, supplier sources, and manufacturing trends. Therefore, it should be noted that while the characteristics observed for this sample set are quite diverse, the trends observed for these samples might differ for another sample set or another time frame.
Although more complicated than a simple physical characteristic evaluation, FTIR is a widely available technique that is relatively easy to use. For this sample set, FTIR yielded additional discrimination following physical examinations. As a stand-alone technique, it is expected that FTIR would have a relatively high discrimination power. Due to the amount of information available on a very small amount of sample, FTIR would be quite valuable if a tape’s condition limited the physical features available for evaluation.
Page 17 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape Few laboratories use XRD in duct tape examinations, primarily due to availability, but the technique has proven to be simple and reliable in the authors’ laboratory. The presence of talc in the backing was the only way two pairs of samples were discriminated using this suite of analytical techniques. XRD can distinguish between the rutile and anatase forms of titanium dioxide, and between calcite and dolomite, all of which are common duct tape pigments/fillers. Further, as demonstrated in this study, XRD can detect talc (a magnesium silicate) in instances where Mg is not readily observed by SEM/EDS; refer to Figures 5 and 6. In these ways, XRD has clearly differentiated samples where other techniques have not.
Although SEM/EDS did not add additional discrimination to this particular study, the authors believe that this result says more about the sources of the samples than the discrimination ability of SEM. The authors’ experience suggests that taken alone, SEM/EDS is very discriminating. In fact, the authors’ laboratory uses SEM/EDS in conjunction with the fabric characteristics to initially narrow down potential manufacturers in duct tape sourcing examinations.
Regarding sourcing, manufacturing and distribution channels make describing a potential source roll to investigators nearly impossible, even when a single manufacturer has been determined. In most instances, a particular manufacturer can be identified, but the source roll could have a number of different brand names or labels, and would be widely available in the marketplace. Occasionally, an atypical tape is examined and the list of potential suppliers is more limited.
Since duct tape products are mass-produced, there could be many other rolls (hundreds of thousands) that have the same physical and chemical properties as the two samples being compared. Despite this, a duct tape comparison in which samples remain undifferentiated still has probative value due to the large number of possible combinations of characteristics and compositions available: the number of duct tape rolls that would differ from those in question is far greater than the number of rolls that would be indistinguishable.
In this study, most of the samples were successfully discriminated, demonstrating that common laboratory techniques are capable of a high degree of discrimination. Since 99.8% of samples were ultimately discriminated, 99.6% by physical characteristics alone, samples that remain indistinguishable following all examinations likely share a manufacturing source.
Acknowledgments
The authors wish to thank Jennifer Gauntt, Roger Keagy, Preston Lowe, and Dennis Ward for their assistance in analyzing these samples and Diana Wright for her insightful comments on the manuscript.
Page 18 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape REFERENCES
1. Johnston J, Serra J. The examination of pressure sensitive adhesive tapes. International Association for Microanalysis Newsletter 2005;5:19-31.
2. Benson JD. Forensic examination of duct tape. Proceedings of the International Symposium on the Analysis and Identification of Polymers; 1984 Jul 31- Aug 2; Quantico, VA: FBI Academy:145-146.
3. Jenkins Jr. TL. Elemental examination of silver duct tape using energy dispersive X-ray spectrometry. Proceedings of the International Symposium on the Analysis and Identification of Polymers; 1984 Jul 31- Aug 2; Quantico, VA: FBI Academy:147-149.
4. Blackledge RD. Tapes with adhesive backings: Their characterization in the forensic science laboratory. In: Mitchell J, editor. Applied polymer analysis and characterization: Recent developments in techniques, instrumentation, problem solving. Munich: Hanser, 1987:413–21.
5. Snodgrass, H. Duct tape analysis as trace evidence. Proceedings of the International Symposium on the Forensic Aspects of Trace Evidence; 1991 Jun 24-28; Quantico, VA: FBI Academy:69-73.
6. Courtney M. Evidential examinations of duct tape. Southwestern Association of Forensic Scientists Journal 1994;16:10-16.
7. Smith J. The forensic value of duct tape comparisons. Midwestern Association of Forensic Scientists Newsletter 1998;27(1):28-33.
8. Merrill RA, Bartick EG. Analysis of pressure sensitive adhesive tape: I. Evaluation of infrared ATR accessory advances, J Forensic Sci 2000;45:93-98.
9. Bradley MJ, Keagy RL, Lowe PC, Rickenbach MP, Wright DM, and LeBeau MA. A validation study for duct tape end matches. J Forensic Sci 2006;51:504-508.
10. Hobbs AL, Gauntt JM, Keagy RL, Lowe PC, and Ward DC. A new approach for the analysis of duct tape backings, Forensic Sci Comm 2007;9(1).
11. Smith JM. Forensic examination of pressure sensitive tape. In: Blackledge RD, ed, Forensic analysis on the cutting edge: New methods for trace evidence analysis. Hoboken, NJ: Wiley and Sons, Inc., 2007;291–332.
12. Tulleners FA, Braun JV. The statistical evaluation of torn and cut duct tape physical end matching. National Criminal Justice Reference Service 2011, document number 235287.
Page 19 of 49
JASTEE, Vol. 3, Issue 1 Mehltretter & Bradley: Duct Tape 13. Scientific Working Group on Materials Analysis (SWGMAT). Guideline for the forensic
examination of pressure-sensitive tapes. JASTEE 2011;2(1):88-97.
14. SWGMAT. Guideline for assessing physical characteristics in forensic tape examinations. JASTEE 2011;2(1):98-105.
15. SWGMAT. Guideline for using light microscopy in forensic examinations of tape components. JASTEE 2011;2(1):106-111.
16. SWGMAT. Guideline for using Fourier transform infrared spectroscopy in forensic tape examinations. JASTEE 2011;2(1):112-121.
17. SWGMAT. Guideline for using scanning electron microscopy/energy dispersive X-ray spectroscopy in forensic tape examinations. JASTEE 2011;2(1):122-132.
18. Smalldon KW, Moffat AC. The calculation of discriminating power for a series of correlated attributes, J For Sci Soc 1973;13:291-295.
Additional information and reprint requests: Andria Mehltretter, M.S., F-ABC Forensic Examiner/Chemist Federal Bureau of Investigation Laboratory Division 2501 Investigation Parkway, Room 4220 Quantico, VA 22135
Page 20 of 49
Robyn Weimer,1 MS , Jennifer Clary,1 MS, Robert Heintz,2Ph. D., and Mark Wall,2Ph.D.
Analysis of Bicomponent Fibers Using Confocal Raman Mapping ABSTRACT Raman microscopy is an analytical tool which nondestructively provides chemical information on samples, often without the need for any preparation. Bicomponent fibers, those composed of at least two polymer types, were examined using Confocal Raman mapping to determine chemical composition and cross-sectional shape. Cross-sections were prepared for the bicomponent fibers of known composition and compared to the Raman results. Confocal Raman mapping provided chemical compositions and indications of cross-sectional shape for bicomponent fibers without any sample preparation. For an accurate shape determination and/or comparison, however, preparation of a cross-section is still recommended. INTRODUCTION Raman spectroscopy, a long recognized technique in research, is an instrumental technique forensic chemists may use to characterize a number of types of evidentiary materials. These materials include, but are not limited to, inks, drugs, paints, explosives, minerals, and cosmetics. Researchers have also applied Raman spectroscopy to a variety of types of textile fibers. Raman has been found to distinguish chemical compositions of generic fiber classes (1,2,3,4), including vegetable fibers (2,4,5), and sometimes can distinguish sub-generic classes of fibers (2,3,4). Pigmented and dyed synthetic fibers (6,7,8), dyed cotton (7,8,9), and wool fibers (8) have also been examined. For colored fibers, Raman has provided information on the colorant, often in addition to the generic fiber class, without any required extraction from the fiber. Raman spectroscopy is useful for forensic analysis because it requires limited sample preparation and is typically nondestructive. Sample preparation of fibers simply requires placing them on a Raman-inactive substrate. The coupling of a Raman instrument with a microscope permits focusing and analyzing samples with extremely small diameters. The sample must be at least the size of the focal spot, which can vary between one to five microns, depending on the objective power and laser wavelength (8,10).
1 Virginia Department of Forensic Science, 700 North Fifth St, Richmond, VA 23219 2 Thermo Fisher Scientific, 5225 Verona Rd, Madison, WI 53711
Page 21 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping Mapping is also possible with Raman spectroscopy to include area mapping (x- and y-axes) and confocal mapping (x- and z-axes, depth profiling). Confocal Raman Microscopy involves the collection of Raman scattering from thin slices taken along the vertical (z) axis, producing a spectral cross-section of the sample. Confocal microscopes have the ability to capture sharp optical sections making it possible to build three-dimensional renditions of a specimen. For fiber samples, this means coupling chemical composition data with optical cross-sectioning. This application is particularly suited for bicomponent fibers, or fibers which are composed of two chemically and/or physically different polymers. Most commercially available bicomponent fibers have a sheath/core, side-by-side, or eccentric sheath/core configuration (11). The market, however, has expanded in more recent years to include a variety of other configurations, such as pie wedge, sea/islands, and tipped trilobal arrangements (12). When submitted for forensic analysis, bicomponent fibers may be difficult to recognize. The more obvious indicators include differences in color, delustrant concentration, or birefringence between the polymer types (13). If these indicators are not present, other commonly used instrumental techniques may not be able to assist in this determination (13). Furthermore, characterization of each component’s chemical composition may be difficult as well. Traditionally, bicomponent fibers have been analyzed using Fourier Transform Infrared Spectroscopy (FTIR). Sample preparation is required for FTIR analysis of fiber samples involving component isolation or sample flattening. If the spatial configuration is known, spectra can be collected for each component. As previously mentioned, bicomponent fibers may not be recognized as such or their configuration may not be revealed prior to FTIR analysis. This lack of information may lead to an incomplete characterization or worse yet a misidentification of the fiber evidence as well as a loss of evidential value recognition. Bicomponent fibers have inherently stronger evidential value due to their rareness in society and casework. (13) For the purpose of this study, Confocal Raman mapping was employed to identify fiber chemical composition(s) of bicomponent fibers and to determine how the components were arranged. As Confocal Raman mapping produces an image of a vertical section of the fiber, width measurements and cross-sectional shapes were also compared from this image to manually prepared cross-sections. MATERIALS AND METHODS Fiber Samples The sample set consisted of seven fiber samples. One sample, previously acquired from a fiber manufacturer, was available within the forensic laboratory. Five samples were obtained by contacting fiber manufacturers specifically requesting bicomponent fibers with sheath/core configurations. The final sample was obtained and analyzed by Thermo Fisher Scientific. The
Page 22 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping bicomponent fiber samples include Nylons, Olefins, and Polyesters. All samples are listed in Table 1. The fiber samples used for this study were of indefinite length. For ease of handling, long sections were used and secured to a potassium bromide block that was placed upon a microscope slide. The fiber was taped on each end, ensuring straight portions from which data could easily be collected. Potassium bromide was used as the sampling substrate because it does not produce a Raman signal. As the x, z-map is not length dependent, much shorter fiber lengths can be used for examination. Sample Number
Composition (identified by
source)
Helpful Information Source
1 Nylon 6/Nylon 6,6 Thermo Fisher Scientific
2 PE/PP PE: Polyethylene; PP: Polypropylene Fiber Manufacturer 3 PE/PP Found to be PE/PET during analysis Previously obtained –
Fiber Manufacturer 4 PP/PP Fiber Manufacturer 5 Hytrel/Hytrel Hytrel: Thermoplastic polyester
elastomer Fiber Manufacturer
6 PETG/PET PETG: Glycol-modified polyethylene terephthalate (polyester); PET: Polyethylene terephthalate (polyester)
Fiber Manufacturer
7 PBT/PET PBT: Polybutylene terephthalate (polyester); PET: Polyethylene terephthalate (polyester)
Fiber Manufacturer
Table 1: List of Bicomponent Fiber Samples
Raman Spectroscopy Raman spectra, in the range 3389-50cm-1 of Raman shift, were collected using a dispersive Thermo Fisher Scientific DXR Raman Microscope. Spectra were collected using two lasers, with excitation wavelengths of 532nm and 780nm, and their respective optical filters and diffraction gratings. The 532nm laser has a 10mW maximum power while the 780nm laser has a 14mW maximum power at the sample. The 532nm laser provided superior data with greater abundance; therefore only data obtained with this laser will be presented. The microscope is equipped with two objectives, 10x and 50x, with the 50x used due to the small fiber size, smaller focal spot size and increased signal strength. A motorized stage was used for the mapping experiments.
Page 23 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping Confocal depth profile mapping along the fiber width was used to accomplish cross-sectional analysis. Mapping experiments were collected using Thermo Fisher Scientific’s OMNIC Atlµs software package. One to two micron step sizes were used resulting in mapping measurement times in the range of three and one half to five hours. The parameters, which affect the measurement time, varied and were based on the size of the fiber, the time requirement, and signal abundance. While step sizes of less than one micron may provide a more accurate chemical map and better component resolution, step size is a function of the motorized stage capability. Components were determined using a combination of methods. The components were often determined manually by comparing spectral peaks which were present in one component and not in another. Once these peaks were located, the components could be differentiated and identified using spectral libraries, either generated internally or commercially available (i.e. Thermo Scientific High Resolution Polymer Library). Occasionally, it was more useful to use the Multivariate Curve Resolution (MCR) option in the Omnic Software package. MCR analyzed the data and attempted to produce spectra that represented the user defined number of pure components. MCR software also provided spatial distributions of each component. Width measurements were made using the Confocal Raman maps. The ruler tool, available in the Atlμs software, allowed a user defined line to be measured. Ten measurements were collected for each component of samples 2, 3, 6, and 7 and then respectively averaged.
Confirmation Techniques Cross-sections were manually prepared for comparison to the resulting Confocal Raman mapping cross-sectional shape. Fibers were inserted into micropipettes. Norland Optical Adhesive 60 was allowed to fill the pipette via capillary action. The filled pipettes were cured under UV light to harden the adhesive and then cross-sections cut using a razor blade. These cross-sections were mounted on a microscope slide and the sheath and core component widths were measured using an Olympus BX40 Compound Microscope (400x). Multiple measurements were taken using the calibrated eyepiece reticule. When appropriate, a width range was necessary to fully represent the variability observed. When needed, a Thermo Nicolet 6700 FTIR with Continuµm Microscope attachment was used to confirm a fiber component. This was accomplished using a microcompression cell with diamond windows. Spectra were collected from 4000cm-1 to 650cm-1.
Page 24 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping RESULTS AND DISCUSSION Nylons In the data provided by Thermo Fisher Scientific, sample 1 appears to have a sheath/core bicomponent configuration with a Nylon 6 sheath. Figure 1 shows a correlation image based on the distribution of Nylon 6 within sample 1. The areas colored red indicate a better correlation, or better match, to the spectrum of Nylon 6 and the blue the lowest correlation to the spectrum of Nylon 6. Figure 2 shows a comparison of the Raman spectrum from the sample’s sheath with a corresponding library spectrum of Nylon 6, drawing attention to the doublet which is present ~1300 cm-1 in Nylon 6.
Figure 1: Correlation image of the sheath component based on the distribution of Nylon 6 within sample 1 (Red = better match to the spectrum of Nylon 6; Blue = lowest correlation to the spectrum of Nylon 6.)
Figure 2: Raman spectrum of the sheath component of sample 1 (top spectrum) versus a corresponding library spectrum of Nylon 6 (bottom spectrum)
Page 25 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping The fiber core appears to be Nylon 6,6. Figures 3 and 4 show the correlation image for this component and the comparison of the Raman spectrum from the sample with a library spectrum of Nylon 6,6. Where a doublet appears for Nylon 6, a single peak is present for Nylon 6,6. Although the sheath and core material are very similar polyamide materials, there are clear spectral characteristics which allow for their differentiation.
Figure 3: Correlation image of the core component based on the distribution of Nylon 6,6 within sample 1 (Red = better match to the spectrum of Nylon 6,6; Blue = lowest correlation to the spectrum of Nylon 6,6.)
Figure 4: Raman spectrum of the core component of sample 1 (top spectrum) versus a corresponding library spectrum of Nylon 6,6 (bottom spectrum)
A third component, Titanium Dioxide (TiO2), was detected in this cross-sectional analysis. Titanium dioxide may be added during the manufacturing process of synthetic fibers as delustrant to reduce sheen or luster. Of interest, two different structural forms of TiO2 were detected within this fiber sample: Anatase and Rutile. The two forms of TiO2 differ in number of
Page 26 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping peaks as well as peak shift locations. The spatial distribution of the Rutile and Anatase forms are shown in Figure 5. Both forms were confirmed by comparison to known library spectra, Figures 6 and 7. Figure 5: Correlation images showing the distribution of TiO2 Rutile Form (Left) and TiO2 Anatase Form (Right) in the sample 1 cross‐section
Figure 6: Raman spectrum from sample 1 core (top spectrum) versus a corresponding library spectrum of Rutile TiO2
(bottom spectrum) The extra peaks above 1000 cm‐1 in the top spectrum are due to the polymer matrix of the fiber.
Page 27 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping
Figure 7: Raman spectrum from sample 1 core (top spectrum) versus a corresponding library spectrum of Anatase TiO2 (bottom spectrum) The extra peaks above 1000 cm‐1 in the top spectrum are due to the polymer matrix of the fiber. Polyethylenes and Polypropylenes Sample 2 was identified as having a sheath/core configuration, based upon manual component analysis, with a PE sheath and a PP core. The chemical images in Figures 8 and 9 illustrate the relative arrangement of each component. The Raman spectrum originating from the component is included as well as the corresponding library spectrum.
Figure 8: Chemical image of sample 2’s sheath component (using 1129 cm‐1 shift) as well as the Raman spectrum of this component (red) versus the corresponding library PE spectrum (blue)
Polyethylene sheath
Page 28 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping
Figure 9: Chemical image of sample 2’s core component (using 808 cm‐1 shift) as well as the Raman spectrum of this component (red) versus the corresponding library PP spectrum (blue)
Both chemical images and correlation images allow for a sample to be visualized in terms of how the chemical components are distributed spatially within the sample. However, chemical images differ from correlation images. While a correlation map looks for a specific reference spectrum within the sample data, a chemical image is a color representation of a specific wavenumber shift. When the particular peak or wavenumber shift is selected, the intensity of the wavenumber shift is indicated by a color gradient. Red indicates the areas in the sample in which the selected wavenumber shift is most intense; blue indicates the areas in the sample in which the selected wavenumber shift is least intense. By manually identifying peaks that are present in only one sample component, separate chemical images can be created that depict the sheath and core. The sample 2 components are very similar, both being olefins by generic class, however, differing by subgeneric class. The Raman spectra may be easily differentiated. The relative arrangement of the components can be compared with the manually prepared, more symmetrical appearing, cross-section in Figure 10.
Figure 10: Image of manually prepared cross‐section of sample 2
Polypropylene core
Page 29 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping Sample 3, was found to also have a sheath/core configuration with a PE sheath and a PET core. Interestingly, the core was identified by the manufacturer as PP, not PET. The core material was analyzed on the FTIR to verify the identification of the core component. The FTIR results confirmed the Raman component identification as PET. Confocal Raman mapping correctly identified the sheath/core components and was able to graph their relative spatial configuration, Figures 11 and 12. Comparison of the Raman spatial distributions versus the manually prepared cross-sections in Figure 13 illustrates how important this practice can be.
Figure 11: Chemical image of sample 3’s sheath component (using 2846 cm‐1 shift) as well as the Raman spectrum of this component (red) versus the corresponding library PE spectrum (blue)
Figure 12: Chemical image of sample 3’s core component (using 858 cm‐1 shift) as well as the Raman spectrum of this component (red) versus the corresponding library PET spectrum (blue). The black arrow indicates a peak consistent with Anatase TiO2.
Polyethylene sheath
PET core
Page 30 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping
Figure 13: Images of manually prepared cross‐sections of sample 3
While a general idea of cross-sectional shape can be obtained with Confocal Raman mapping, it is not able to provide the entire circumference of the cross-sectional shape. As illustrated by this sample, one cannot assume the fiber cross-sectional shape is symmetrical. The sheath is not evenly surrounding the core in this sample. Also, the manually prepared cross-section shows TiO2 in the core, which is only indicated by a single peak ~140 cm-1 in the spectrum of the core component (see black arrow). This peak location indicates the presence of Anatase TiO2 as opposed to Rutile. Multiple components could not be identified within Sample 4. This sample was identified by the manufacturer as having a sheath/core configuration with both components consisting of PP, differing only in melt flows. Confocal Raman mapping was able to show only one PP component, as seen in Figure 14, however, indicated a different cross-sectional shape than had been identified by the manufacturer. The manually prepared cross-section illustrates the accurate bow-tie shape, which is also represented by Confocal Raman mapping. The only indication of differing components is provided in both the manually prepared and the optical cross-sections seen as a difference between the lobe sizes. Chemical lobe differentiation would not be expected by Raman or FTIR given the melt flow difference, but Raman was able to give an indication of the shape that FTIR could not have provided.
Figure 14: Chemical image of sample 4 (top left ‐ using 2958 cm‐1 shift), single component, and the image of the manually prepared cross‐section (top right). Raman spectrum included of this sample (red) versus the corresponding library PP spectrum (blue).
Page 31 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping Polyesters Similarly, sample 5 was also identified as containing a single component. This fiber was identified by the manufacturer as having a sheath/core configuration which differs only by grade of Hytrel. The manually prepared cross-section as seen in Figure 15 indicated the presence of two components and confirmed a sheath/core configuration. Initially, not expecting high-quality Raman results due to the black fiber color, the Raman surprisingly was able to identify the composition as polyester without any difficulty, Figure 16. Therefore, the guidance that Raman does not analyze dark or black materials well is sample specific and should not be taken as a rule of thumb. For this specific sample, the recognition of its bicomponent nature can be obtained visually due to the color difference between the colorless sheath and dark core. No peaks could be identified, manually or with the use of MCR, as possibly originating from the dye or pigment used to color the core. As expected, Raman was not able to differentiate the sheath and core components of sample 5 based on a material grade difference. Raman was also not able to distinguish the components based on peaks associated with the colorant; however, this could be due to a lack of a reference pigment library. Consistent with the Raman results, FTIR was able to identify the sheath and core components as polyester, but could not distinguish them from one another.
Figure 15: Image of manually prepared cross‐section of sample 5
Figure 16: Chemical image of sample 5 (using 1607 cm‐1 shift), single component, as well as the Raman spectrum of this sample (red) versus the corresponding best match to library Polyester PBT spectrum (blue).
Page 32 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping Samples 6 and 7, however, are each comprised of two different types of polyester. Confocal Raman mapping was able to differentiate the two polyester components as well as the third component located only in their cores, TiO2. Sample 6’s PETG sheath and PET core could be distinguished from one another spectrally with the use of the 2850cm-1 peak, which is present only in PETG. The relative spatial distribution of the core was visible with the help of peaks originating from TiO2, which is present only in the PET core. PETG was the more difficult of the components to identify, mostly attributable to the limitations of the spectral searching library. PETG, glycol modified, and PET are very similar but small differences can be seen around 2850cm-1 and 800cm-1 in Figures 17 and 18. PET and PETG are structurally similar; however, PETG’s structure includes a non-aromatic, six carbon ring (14). Peaks around 2850cm-1 and 800cm-1 in the PETG spectrum are most likely attributable to the in-phase stretching of the ring and the ring’s methylene groups (15). The manually prepared cross-section of sample 6, as seen in Figure 19, corresponds with the Confocal Raman mapping results for component spatial distribution.
Figure 17: Chemical image of sample 6’s sheath component (using 2853 cm‐1 shift) as well as the Raman spectrum of this component (red) versus the corresponding library PETG spectrum (blue). Black arrows indicate the differences between PETG and PET: peak at ~2850cm‐1 present in PETG and absent in PET, and doublet ~800cm‐1 present in PETG where single peak present in PET.
PETG sheath
Page 33 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping
Figure 18: Chemical image of sample 6’s core component (using 143 cm‐1 shift) as well as the Raman spectrum of this component (red) versus the corresponding library PET spectrum (blue). Black arrows indicate the differences between PETG and PET: peak at ~2850cm‐1 present in PETG and absent in PET, and doublet ~800cm‐1 present in PETG where single peak present in PET. Black box surrounds 4 peaks in the red core component spectrum indicating the presence of Anatase TiO2.
Figure 19: Image of manually prepared cross‐section of sample 6
Sample 7’s PBT sheath and PET core were distinguished from one another only with the help of MCR coupled with the RGB (red, green, blue) display. MCR will attempt to seek out the number of pure components present in the sample that is input by the user. Using the component information obtained, the RGB display will assign a red, green or blue color to each of the components the software identifies. The components, a maximum of 3, are presented individually as well as in a compilation, Figure 20. This particular way of displaying the data was the most useful for this sample given the similarity between the core and sheath chemical components. PBT and PET are very similar structurally, with PBT having four methylene groups where PET has only two. The additional methylene group vibrations may account for the slight peak shifts seen at ~1700cm-1, 1300cm-1, and 1100cm-1 which allow for their differentiation. Titanium dioxide, Anatase form, was also present in the core of sample 7 indicated by the presence of three minor peaks between 650-350cm-1 and a dominant peak ~140cm-1. Figure 21 compares Sample 7’s sheath component to corresponding PBT and PET library spectra. The PET library spectrum is shifted to the left of PBT and the sheath component at ~1700cm-1, 1300cm-1, and 1100cm-1 which is consistent with the spectra in the RGB display. The sheath was therefore identified as PBT. The core component of sample 7, seen in Figure 22, was identified as PET. Keeping in mind the very thin sheath seen in the manually prepared cross-sectional images in Figure 23 and the fact that PET is such a good Raman scatterer, it is not
PET core
Page 34 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping surprising that there is influence from the PET core affecting the spectrum of the PBT sheath, as seen in the areas of ~2920-2960cm-1 and ~826cm-1.
Figure 20: RGB display of sample 7, with spectra and images of each component and their compilation image to the bottom right. Note the four peaks outlined with the black box in the blue spectrum indicating the Anatase form of TiO2. Illustrating components 1 (core) and 2 (sheath) are very similar with slight differences, the black arrows note peak shifts at ~1700cm‐1, 1300cm‐1, and 1100cm‐1 (red PET core is shifted to the left)
Core Sheath Delustrant
Compilation Image
Page 35 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping
Figure 21: Comparing sample 7’s sheath component (red Raman spectrum) to corresponding PBT (blue) and PET (purple) library spectra. Noted by black arrows, the PET library spectrum is shifted to the left at ~1700cm‐1, 1300cm‐
1, and 1100cm‐1 (consistent with RGB display). Also, correlation image of sample 7’s sheath component based on the distribution of PBT (Red = better match to the spectrum of PBT; Blue = lowest correlation to the spectrum of PBT.).
Figure 22: Chemical image of sample 7’s core component (using 857 cm‐1 shift) as well as the Raman spectrum of this component (red) versus the corresponding library PET spectrum (blue). Note only a weak indication is present ~150cm‐1 of TiO2 in the red spectrum.
PBT PET
PET
Page 36 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping
Figure 23: Images of manually prepared cross‐sections of sample 7
Though the components could be identified and differentiated for both samples 6 and 7, the Confocal Raman maps are not equally useful. As can be seen in the manually prepared cross-section of samples 6 and 7 in Figures 19 and 23 respectively, sample 6 is symmetrical while sample 7 has a more sporadic sheath covering. For asymmetrical samples, the quality of the Confocal Raman map will most likely depend on the orientation of the fiber at the point of measurement, and may not provide an entirely accurate representation of the sheath/core configuration. Width Measurements Samples 2 and 6 are good examples for the potential of simultaneously measuring fiber diameter, identifying fiber composition, and determining cross-sectional shape with a single Confocal Raman map analysis. Both of these samples are symmetrical and the measurements taken manually are similar to those measured using the Altµs software, see Table 2. With this additional possibility, Confocal Raman analysis can accomplish a number of tasks in one mapping analysis that would normally require multiple steps using various other pieces of instrumentation. The map experiment may be set up and allowed to run without any supervision, requiring minimal data interpretation once complete. While measurements are possible, these would be approximate width measurements only. Their accuracy relies heavily on the beginning and ending of the software measuring tool, which is user defined, Figure 24. As can be seen in the component chemical images, where one component begins and ends is subjective. The color gradient from red to blue also makes the measurement points unclear. Width measurements can also be incorrectly interpreted in the case of non-symmetrical cross-sections. This can be seen with samples 3 and 7 in Table 2 where the sheath is not of a uniform width surrounding the core. Multiple Confocal Raman maps could be collected and these measurements averaged to get a better idea of the ranges that might be present. However, the addition of more map experiments will greatly increase the time required to obtain this information. It would be a better use of time to use other methods if multiple mapping experiments were required. Confocal Raman mapping may be best used as
Page 37 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping an initial screening method supplemented with manually prepared cross-sections or other confirmatory techniques as needed. The incomplete cross-sectional image provided by Confocal Raman mapping is also a limitation to this type of analysis. It is particularly problematic when the fibers are not symmetrical to achieve an accurate depiction of the cross-sectional shape and accurate width measurements. As can be seen in the Confocal Raman maps, one surface is sharp while the other is indistinct and hazy. What is displayed as the bottom surface in the Confocal Raman map is actually the uppermost surface of the fiber. The blurred surface is not a function of fiber width, as much thinner fibers were analyzed with the same results. Thermo Fisher Scientific proposes the indistinct surface may be due to the fiber acting as a lens, producing lensing effects.
Sample Number
Component Measured
Width Measurement using Raman (μm)
Width Measurement using Compound Microscope
(μm) 2 Sheath 5.65 5
Core 20.55 20 6 Sheath 10.98 11.25
Core 55.11 55 3 Sheath 4.07 < 2.5 – 5
Core 13.96 12.5 7 Sheath 3.28 < 2.5 – 5
Core 20.30 17.5 – 20 Table 2: Comparing Approximate Width Measurements of Fiber Sample Components
Figure 24: Chemical image of sample 2’s core with the software measuring tool overlaid displaying the distance measured to the right. The measuring tool’s beginning and ending points are user defined.
Distance: 20.19μm
Page 38 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping CONCLUSIONS Recognition of a bicomponent fiber can be difficult. Confocal Raman mapping can alert the examiner to the presence of a bicomponent fiber, as long as the components vary in sub-generic class. Simultaneously, Raman mapping can provide the chemical composition of each component and offer their relative spatial arrangements. Lengthy analysis times are required to obtain the map using small micron steps, but the analysis does not demand the examiner’s time to monitor the collection. The Confocal x, z-map presents an image of the general cross-sectional shape of the fiber sample; however, due to a possible lensing effect, a complete image is often not available. Minimal sample preparation is required and presentation into the instrument is simple. For comparisons, accurate shape determinations, and/or accurate width measurements, preparation of a cross-section is still recommended as Confocal Raman mapping is limited to the general shape present.
ACKNOWLEDGEMENTS The authors would like to gratefully acknowledge Thermo Fisher Scientific for the instrumentation. The authors would also like to thank Michael Hodge and Scott Hammitt from Fiber Innovation Technology, Inc., for providing fiber samples.
REFERENCES
1. Agbenyega JK, Ellis G, Hendra PJ, Maddams CP, Willis HA, Chalmers C. Applications of
Fourier Transform Raman spectroscopy in the synthetic polymer field. Spectrochim Acta A 1990;46(2):197-216.
2. Keen IP, White GW, Fredericks PM. Characterization of fibers by Raman microprobe
spectroscopy. J Forensic Sci 1998;43(1):82-89. 3. Miller JV, Bartick EG. Forensic analysis of single fibers by Raman spectroscopy. Applied
Spectr 2001;55(12):1729-32. 4. Cho L. Identification of textile fiber by Raman microspectroscopy. For Sci Journal
2007;6(1):55-62. 5. Edwards HGM, Farwell DW, Webster D. FT Raman microscopy of untreated natural plant
fibres. Spectrochim Acta A 1997;53(13):2383-2392. 6. Jochem G, Lehnert RJ. On the potential of Raman microspectroscopy for the forensic
analysis of coloured textile fibers. Sc & Justice 2002;42(4):215-221.
Page 39 of 49
JASTEE, Vol. 3, Issue 1 Weimer, et. al.: Confocal Raman Mapping 7. Lepot L, De Wael K, Gason F, Gilbert B. Application of Raman spectroscopy to forensic
fibre cases. Sc & Justice 2008;48(3):109-117. 8. Massonnet G, Buzzini P, Jochem G, Stauber M, Coyle T, Roux C, Thomas J, Leijenhorst H,
Van Zanten Z, Wiggins K, Russell C, Chabli S, Rosengarten A. Evaluation of Raman Spectroscopy for the analysis of colored fibers: a collaborative study. J Forensic Sci 2005;50(5):1028-1038.
9. Thomas J, Buzzini P, Massonnet G, Reedy B, Roux C. Raman spectroscopy and the forensic
analysis of black/grey and blue cotton fibres: Part 1. Investigation of the effects of varying laser wavelength. For Sci Int 2005;152(2-3):189-197.
10. Martin K, Shearer G. Raman Microspectroscopy training course. College of Microscopy.
Wesmont, IL. 6-8 April 2010. 11. FiberSource (n.d.) Bicomponent Fiber. [online] Available at:
http://www.fibersource.com/f-tutor/bicomponent.htm>. [Accessed: 3/16/11]. 12. Dugan JS. Specialty Markets: Bicomponent fibers. Textile World 2010;160(4):22-24. 13. Flynn K, O’Leary Robyn, Roux C, Reedy B. Forensic analysis of bicomponent fibers using
Infrared Chemical Imaging. J Forensic Sci 2006;51(3):586-596.
14. Baranowski, M. et al. (n.d.) MOLECULAR DYNAMICS IN POLY(ETHYLENE TEREPHTHALATE). [online] Available at: http://www.ifj.edu.pl/conf/mrj/mrj06/abstrakty/baranowski_m.pdf
[Accessed: 05/03/12].
15. Larkin, P. (2011) IR and Raman Spectroscopy: Principles and Spectral Interpretation.
[online] Available at:
http://www.elsevier.com/authored_subject_sections/P04/IYC_chapters/IR_and_Raman_Spectros
copy_9780123869845.pdf [Accessed: 05/04/12].
Page 40 of 49
38 AFTE Journal--Volume 44 Number 1--Winter 2012
Introduction
The National Fish and Wildlife Forensic Laboratory is charged with assisting law enforcement in prosecuting wildlife crime. Typical victims seen in the Lab include bears, eagles, wolves, etc. In recent years, veterinary pathologists have observed and collected colorful plastic fragments associated with the gunshot wounds. Sometimes the color and shape of the polymer pieces are sufficient to give an indication of the caliber and manufacturer of a plastic-tipped projectile. Such is the case with some of the Nosler Ballistic Tip® bullets, due to a color code assigned to different calibers, however, white tips are used for all the calibers of Nosler Accubond® bullets and olive-drab green is used by Nosler in all calibers of their E Tip® bullets. Hornady uses red-colored tips for all of the bullets sold under the Hornady brand. Hornady also furnishes bullets to other ammunition manufacturers with different colored tips. Hornady V-Max® bullet tips loaded in 17 HMR cartridges are black for CCI and Federal Premium® labeled boxes, gold in Remington boxes, silver in Winchester boxes, and red in Hornady boxes [1, 2, 3]. A previous study considered the physical properties of selected polymer-tipped bullets [1].
Plastic-tipped bullets are designed to have excellent aerodynamics and offer increased terminal velocity and a flatter trajectory. A plastic tipped bullet is in essence a hollow-point bullet that has a pointed plastic tip. Upon impact, the polymer tip assists in expansion of the bullet nose [3, 4].
Rigid polymer tips were originally used to replace exposed lead tips in pointed (spitzer) style bullets. The purpose was to prevent deformation of the bullet tip, which may affect accuracy and trajectory. Damage to the bullet tip could occur
Chemical Properties of Selected Plastic-Tipped Bullets By: Melisa C. Thompson, Cady A. Lancaster, Michele G. Banta, Crystal N. Hart, Michael D. Scanlan, and Edgard O. Espinoza U.S. Fish and Wildlife National Forensics Laboratory 1490 East Main Street, Ashland, Oregon 97520-1310
Keywords: plastic-tipped bullets, ballistic-tip ammo, FTIR, XRF, discriminate analysis
ABSTRACT
Fragments of plastic-tipped bullets are often found in wound tracts of unlawfully killed wildlife. Using color, Fourier transform infrared spectroscopy analysis coupled to the statistical power of discriminate analysis and x-ray fluorescence spectrometry, we were able to characterize the polymers found in common commercially available plastic-tipped bullets. The data is surprising because the high quality control used by the manufacturers provides an opportunity for forensic class character determination.
to exposed lead tips during handling, loading into the firearm, or from recoil battering the bullet tips in the magazine. Additionally, the polymer tip helped to expand the nose of the bullet upon impact [2]. The hard plastic-tipped bullets could not be used in rifles with tubular magazines because of the possibility of the hard tip detonating the primer of another cartridge in the magazine during recoil. Conventional bullets for tubular magazine rifles were flat tipped or rounded to prevent detonation of primers due to recoil. The cartridges were classed as short range because of the poor aerodynamic shape causing velocity loss and rainbow trajectories. More recently, Hornady added soft polymer tips in their LEVERevolution® Flex Tip® bullets to cartridges primarily used in tubular magazines [3]. The soft-tipped bullets allow more aerodynamic, pointed bullets to be used in tubular magazines, without the possibility of detonating a primer that the bullet tip is resting on in the magazine. The results have been flatter trajectories and higher terminal velocities. The 2010 Hornady catalog lists the 30-30 Winchester, 308 Marlin Express, 338 Marlin Express, 32 Winchester Special, 357 Remington Magnum, 35 Remington, 44 Remington Magnum, 444 Marlin, 45 Colt, 45-70 Government, and 450 Marlin cartridges loaded with Flex Tip® bullets. Hornady also uses the Flex Tip (FTX)® bullets in their SST® Shotgun Slug and muzzle-loading bullets [5].
The purpose of this study was to determine if it is possible to associate the remnants of evidentiary plastic fragments, suspected of being from plastic-tipped bullets, to a manufacturer through the analysis of the chemical properties of the polymers. The plastic component of these bullets was analyzed using three analytical instruments. A video spectral comparator (VSC) was used to determine color and ultraviolet characteristics, a Fourier transform infrared spectrometer (FTIR) was used to characterize the type of polymer used and a statistical tool (discriminate analysis [DA]) was used to distinguish FTIR spectral differences. Lastly,
Date Received: September 12, 2011Peer Review Completed: December 1, 2011
Page 41 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 39
Color Tip Manufacture Type of Tip Polymer Identification
Red Hornady Not Available Polyoxymethylene
Red Nosler 7mm BT Polyoxymethylene
Red Nosler Win OEM AB
Polyoxymethylene
Red Nosler 7mm AccuBond
Polyoxymethylene
Red Winchester 30 Cal XP3 Polyetherimide
Red Winchester 12GA XP3 Polycarbonate
Red/Gold Nosler 22 BT Polyoxymethylene
Orange Nosler Norma AB Polyoxymethylene
Orange Noser 22 BT Polyoxymethylene
Yellow Nosler 270 BT Polyoxymethylene
Green Nosler 30 BT Polyoxymethylene
Green Nosler 27 Cal E-Tip Polyoxymethylene
Green Nosler E-Tip Polyoxymethylene
Blue Nosler 25 BT Polyoxymethylene
Blue Nosler 8mm BT Polyoxymethylene
Blue Hornady Not Available Polyoxymethylene
Purple Nosler 243 BT Polyoxymethylene
Brown Nosler 264 BT Polyoxymethylene
Brown Hornady Not Available Polyoxymethylene
Grey Nosler 30 Cal BST Polyoxymethylene
Grey Nosler CTBST Polyoxymethylene
White Nosler AccuBond Polyoxymethylene
X-ray fluorescence (XRF) was used to analyze elemental composition.
Video Spectral Comparator
A video spectral comparator (VSC) is an imaging system typically used for the analysis of documents. This instrument is a multi-spectral imaging system that uses a camera, light sources, viewing filters, and spectrometer (9-nm resolution) to allow for visualization under alternate light sources or spectral acquisition (Vis-NIR, 400-1000 nm). This spectral data can describe and measure color. While developed as a tool for the analysis of documents, we have used it for determining color in fibers, paint, ivory, pelts, skin and many other evidentiary items. In this research, the VSC ( Foster and Freeman VSC®6000 [6]) was used to visualize the plastic tips of bullets under various light conditions and to determine if ultra-violet (UV) illumination induced a color change in the polymer [6].
Infrared Spectroscopy
Fourier transform infrared spectroscopy (FTIR) is a non-destructive analytical tool that, when used in examining polymers, stands out for its robustness, ease of sample preparation, simplicity of operation, and the ability to make structural elucidations [7]. The resolving power of FTIR has been applied in such diverse fields as forensic fiber identification [8, 9] and natural polymers such as hair [10] and sea turtle keratin [11].
Discriminant analysis of vibrational spectra (Raman or infrared) has been successfully used to extend the limitations inherent in vibrational data. Examples include confirmation of edible oils and fats [12], sub-typing of nylon polymers [13], geographic sourcing of medicinal plants [14], forensic identification of fingernails vs. toenails [15], the forensic identification of fiber blends [9] and the identification of sea turtle keratin from bovid horn [11]. This method has also proven to be useful in distinguishing polymer bullet tips that may commonly be encountered in wildlife crime investigations.
X-ray Fluorescence Spectrometry
X-ray fluorescence spectrometry (XRF) is a non-destructive analytical tool that can provide rapid, multi-element measurements of solid or liquid samples. It has been used in a wide variety of analytical disciplines including cadmium and lead detection in hazardous waste sites [16], copper contamination of agricultural soils [17], as well as a method for distinguishing keratins from casein-based plastics [11].
High-energy x-ray photons excite the atoms in a sample, and the energy difference between the excited state and ground state are reflected in the emission. By operating under a vacuum there is an increased sensitivity to elements lighter than argon.
Methods
Reference polymer samples
Winchester (n=2), Nosler (n=17), and Hornady (n=3) provided reference polymer tips used in their respective bullets (Table 1). Dupont [18] provided a reference sample of Delrin® 500P, a polyoxymethylene (POM) acetal homopolymer, and Ticona [19] provided copolymer samples of Celcon®POM (M25, M90, and M270 natural).
Reference plastic-tipped bullets
Twenty-nine boxes of plastic-tipped bullets of various calibers and one box of plastic-tipped shotgun slugs were purchased locally (See Figure 1 and Table 2). Some of the plastic tips
Table 1
Page 42 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 40
Figure 1
were sampled from loaded cartridges, and others were selected randomly from boxes of bullets, such that each box provided a sample size of 20 polymer tips. The only exception was Hornady SST® FTX® which had five plastic-tipped slugs. The boxes of bullets were from the following manufacturers: Nosler (n=11), Hornady (n=7), Federal Premium® (n=6), Winchester (n=2), FNH USA (n=1), CCI (n=1), Remington (n=1), and Sierra (n=1). Some of the brands of ammunition utilize plastic-tipped bullets of their own fabrication, others use plastic-tipped bullets manufactured by other companies (See Table 2). The sampled populations were not a complete sweep of all manufacturers that distribute plastic-tipped bullets or the colors used.
Video Spectral Comparator (VSC)
A Foster and Freeman VSC®6000 (v. 6.5 with HS updates [6]) was used to determine color variation within each box of ammunition. Each box was viewed using the auto exam function of the software, which varies illumination sources and filters to produce various viewing conditions to determine consistency. The viewing conditions included visible, absorbance, transmission, spot fluorescence, near-infrared, and ultraviolet (312 nm and 254 nm).
Fourier Transform Infrared Spectroscopy (FTIR)
Shavings of the plastic bullets were used for FTIR analysis. A Nicolet 6700 FTIR (Omnic™ 8.2 software) equipped with a Smart iTR™ accessory was used to study the spectral properties of each polymer tip (n=585). Collection parameters for the iTR™ analysis consisted of 80 scans at a resolution of 4 nm, which resulted in data spacing of 1.928 cm-1 under autogain control. The final format was recorded in log (1/R) vs. wavenumber (cm-1) with a spectral range of 4000-840 cm-1. No correction was performed on the resulting spectra. A background spectrum was collected every 25 minutes [20].
Discriminant Analysis
The Omnic™ Specta™ (1.0.0) HR Specta Polymers and Plasticizers by ATR library was used to identify the material of the tips. TQ Analyst™ (v.8.3) software package (Nicolet) [21] was used to perform discriminant analysis (DA) on samples of the same color. Discriminant analysis was performed over various spectral regions to determine if separation was achieved based on the performance index.
Discriminant analysis is a multivariate statistical method that assists in the classification of spectral data into distinct groups.
Page 43 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 41C
olor
Bul
let D
istr
ibut
orPo
lym
er T
ip
Man
ufac
ture
r
Lot
#C
alib
erG
rain
Type
of T
ipn
Col
or c
hang
e un
der
UV
ligh
t at 3
12nm
Poly
mer
Iden
tifica
tion
Ele
men
ts
Pres
ent
Sem
i-Qua
ntifi
catio
n (%
W±
Vari
ance
)
Bla
ckC
CI®
H
orna
dyM
13R
1917
HM
R17
V-M
ax™
9-
POM
Ca,
Cu,
Fe,
S,
Si, T
i
Ca:
9 ±
12
Cu:
12
± 12
Fe:
4 ±
1
S:
9 ±
2 S
i: 4
± 1
Ti:
61 ±
8
Bla
ckC
CI®
H
orna
dyM
13R
1917
HM
R17
V-M
ax™
11-
POM
-
Bla
ckFe
dera
l Pre
miu
m®
H
orna
dyB
01S1
617
HM
R17
V-M
ax™
20-
POM
-
Bla
ckR
emin
gton
®Sw
ift™
L24
SA29
L30
0 R
em. U
ltra
Mag
180
Sciro
cco™
Bon
ded
20-
POM
-
Blu
eN
osle
rN
osle
r11
4113
1213
42 IH
1F71
2510
0B
allis
tic T
ip®
20G
reen
PO
MC
u,Si
, Ti
-
Blu
e F
NH
USA
®
Hor
nady
11
3096
9032
5.7
x 28
mm
40V-
Max
™20
Ora
nge
POM
-
Bro
wn
Nos
ler
Nos
ler
1342
8618
61 K
H12
F72
6.5
mm
100
Bal
listic
Tip
®20
-PO
MFe
, Ti,
Zn-
Gre
enFe
dera
l Pre
miu
m®
N
osle
r N
V15
X54
030
0 W
SM15
0B
allis
tic T
ip®
20-
POM
Cu,
Ni,
Si, T
iC
u:10
± 2
Ni:
14 ±
1
S
i: 11
±
2 T
i: 65
± 4
Gre
enSi
erra
Sier
ra92
7630
1440
2240
Plas
tic T
ip20
-PO
MC
a, C
r, Ti
-
Gre
enN
osle
r®
Nos
ler
4164
1636
7030
Cal
iber
N/A
Plas
tic T
ip20
-PO
MC
a, C
u, T
iC
a: 4
± 1
C
u: 1
5 ±
3 T
i: 81
± 2
Gre
yW
inch
este
r®H
orna
dyF0
6R03
17 H
MR
17V-
Max
™20
-PO
MA
l, Ti
-
Silv
er/G
rey
Win
ches
ter®
N
osle
r16
WL5
27
mm
WSM
140
Bal
listic
Silv
ertip
®20
-PO
MFe
, S,T
i-
Ora
nge
Nos
ler C
usto
m®
N
osle
r13
59 6
0003
020
1122
-250
Rem
55B
allis
tic T
ip®
20-
POM
Fe,
Si,
ZnFe
: 29
± 6
Si: 3
9 ±
22 Z
n: 3
1
± 19
Ora
nge
Nos
ler C
usto
m®
N
osle
r60
001
0201
622
3 R
em40
Bal
listic
Tip
®20
-PO
MFe
, Si,Z
nFe
: 28
± 7
Si:
39 ±
14
Zn:
33
± 11
Ora
nge
Fede
ral P
rem
ium
®
Nos
ler
NO
U47
Z562
223
Rem
55B
allis
tic T
ip®
20-
POM
Fe, S
i,Zn
Fe: 2
8 ±
7 S
i: 38
± 1
2 Z
n: 3
5 ±
15
Ora
nge
Nos
ler
Nos
ler
1181
953
1165
AE0
4F6-
1
22 C
alib
er55
Bal
listic
Tip
®20
-PO
MFe
, Si,Z
nFe
: 28
± 6
Si: 3
6 ±
16 Z
n: 3
6
± 13
Purp
leFe
dera
l Pre
miu
m®
N
osle
rN
OU
5X
724
243
Win
.70
Bal
listic
Tip
®20
-PO
MC
a, S
i, Ti
-
Purp
leFe
dera
l Pre
miu
m®
N
osle
r03
3X50
4 20
4 R
uger
32B
allis
tic T
ip®
20Ye
llow
POM
-
Red
Hor
nady
Hor
nady
3102
025
7mm
-08
Rem
139
SST®
20
-PO
M S
,Ti
S: 3
6 ±
6 T
i: 64
± 3
Red
Fede
ral P
rem
ium
®
Nos
ler
U1T
442
7mm
-08
Rem
140
Bal
listic
Tip
®20
-PO
MC
a, F
e, S
i, Ti
-
Red
Hor
nady
Hor
nady
E12S
0617
HM
R17
V-M
ax™
20-
POM
S, T
iS:
35
± 4
Ti:
65 ±
4
Red
Nos
ler
Nos
ler
3323
1036
707m
m15
0B
allis
tic T
ip®
20-
POM
Fe, Z
nFe
: 23
± 1
Zn:
77
± 1
Red
Hor
nady
Hor
nady
3111
056
243
Win
58V-
Max
™20
-PO
MS,
Ti
S: 3
5 ±
4 T
i: 65
± 4
Red
Hor
nady
Hor
nady
3060
330
204
Rug
er32
V-M
ax™
20-
POM
S, T
i, Zn
S: 3
3 ±
3
Ti
: 37
± 4
Zn:
30
± 9
Red
Hor
nady
Hor
nady
J03K
0917
Mac
h 2
17V-
Max
™20
-PO
MS,
Ti,
ZnS:
33
± 4
Ti: 3
7 ±
4 Z
n: 3
1 ±
5
Red
Hor
nady
Hor
nady
3101
826
308
MA
R E
XP
160
FTX
®20
-Po
lyes
ter u
reth
ane
-
bis(
phen
ylis
ocya
nate
)
copo
lym
er
Ca,
S, T
i-
Red
Hor
nady
Hor
nady
3110
280
12 G
A S
hotg
un30
0FT
X®
5
-Po
lyes
ter u
reth
ane
-
bis(
phen
ylis
ocya
nate
)
copo
lym
er
Ca,
S, T
i-
Silv
erW
inch
este
r®
Nos
ler
16W
L52
7 m
m W
SM14
0B
allis
tic S
ilver
tip®
20-
POM
Fe, S
,Ti
-
Whi
teN
osle
r Cus
tom
®
Nos
ler
6006
3 07
008
300
WSM
180
Acc
uBon
d®
20-
POM
Ca,
Si, T
i-
Whi
teN
osle
r Cus
tom
®
Nos
ler
1326
600
42 0
9017
7mm
-08
Rem
140
Acc
uBon
d®
20-
POM
Si, T
i-
Whi
teN
osle
r Cus
tom
®
Nos
ler
8145
51 6
0049
059
0630
8 W
in16
5A
ccuB
ond®
20
-PO
MSi
, Ti
-
Yello
wN
osle
rN
osle
rN
ot A
vaila
ble
270
130
Bal
listic
Tip
®20
-PO
MTi
-
Tabl
e 2
Page 44 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 42
The rationale of discriminant analysis in the present situation was to establish discriminant functions to characterize the forensic samples of polymer tips. The software (TQ Analyst™) compiles an average spectrum from the known samples, and then each is assigned a numerical score based on the deviation from the calculated spectrum. These numerical scores are then plotted to provide a graphical representation [21].
Lastly, each known sample is validated by determining the Mahalanobis distance of the sample from the average spectrum. Therefore, each polymer tip sample is assigned to the nearest group centroid based on its calculated Mahalanobis distance. The closer a specimen is to a particular centroid class, the higher the likelihood that it will be classified with that particular sample set.
An example of the discriminant analysis experiment is to look at two of the red plastic-tipped bullet samples from Hornady and Nosler. This set consisted of 20 Hornady V-Max® and 20 Nosler polymer tips that were characterized by the FTIR as being polyoxymetheylene (POM), a thermoplastic polymer commonly referred to as acetal or polyacetal. These polymer samples were used as a reference population to calculate the discriminant function of each type and to establish the performance index. The performance index is a measure of how well a discriminant analysis method can categorize spectra from calibration samples. The performance index of the discriminant analysis was 95.0 percent, which indicates that DA can differentiate between Hornady and Nosler POM polymers. Reliable categorizations occur when the
performance index exceeds 90% [21]. As demonstrated in Figure 2, Hornady and Nosler POM polymers are segregated within their corresponding populations, thus FTIR and discriminant analysis can distinguish between the red Hornady and the red Nosler POM polymers, despite the fact that both have similar FTIR spectra.
Therefore, if an unknown red plastic was associated with wildlife mortality, and FTIR analysis indicates the plastic is a POM polymer, then the spectrum will be compared against the Hornady and Nosler sample set. The specimen will be classified as similar to Hornady if it falls within the centroid class made by the Hornady samples, or similar to Nosler if it falls within the centroid class made by the these polymer tip samples.
X-ray Fluorescence (XRF)
The plastic-tipped bullets were sampled by a transverse cut three-quarters through the polymer tip with a soldering iron equipped with a blade. The tip was then snapped off to achieve melt and break zones on the cross-section. The melt and break zones were analyzed with an Edax Orbis X-ray Fluorescence Spectrometer (XRF) equipped with an Apollo Silicon Drift Detector. The X-ray source was set at 30 keV and 300 μA with a 30-μm spot size and collection lifetime of 30 seconds. The Hornady FTX® polymer tips were shaved with a utility knife, and the shavings were analyzed under equivalent parameters. All samples were analyzed under vacuum to determine the elemental content. The analysis considered elements with higher atomic weights than sodium on The Periodic Table
Figure 2
Page 45 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 43
of the Elements. Semi-quantitative analysis was used to determine weight percent of select samples [22].
Results and Discussion
The results from commercial sources indicate that POM is the most common polymer being used in plastic-tipped bullets (See Table 1). POM polymers have been described and characterized by Chanda and Roy [23]. Additionally, we encountered a polyester urethane - methylene bis (phenylisocyanate) copolymer from Hornady. The analysis of the reference polymers obtained directly from manufacturers also identified POM as the most common polymer, although we also encountered a polyetherimide and a polycarbonate (See Table 1), which were not seen in the commercial ammunition. Although we are aware that POM is manufactured as a homopolymer [18] and as a copolymer [19], our analysis could not distinguish between these two polymer types. The results of all the analysis are summarized in Table 1 and Table 2. Table 2 is sorted based on color, because the recovery of evidence and the subsequent analysis is guided by the original color of the polymer tips.
Black polymer tips
Within this study, there were three sources of black plastic-tipped bullets: CCI (17 HMR), Remington (Swift Scirocco™ Bonded 300 Rem. Ultra Mag), and Federal Premium® (17 HMR). Black plastic-tipped V-Max® bullets are loaded into 17 HMR cases under the labels of CCI and Federal. The V-Max® was identified as POM. Swift Bullet Company uses black polymer tips in all of their Scirocco™ line of bullets, which are also molded from POM. There was no appreciable difference observed with the VSC. Analysis by XRF showed that the CCI box contained two different batches of POM. Nine of the twenty polymer tips sampled contained silicone, sulfur, calcium, titanium, iron, and copper, while the remaining 11 exhibited none. XRF analysis of the Swift Scirocco™ plastic-tipped bullets also did not reveal elements present. Nevertheless, FTIR-DA was able to differentiate the two polymers that did not have elements present, i.e., CCI vs. Swift Scirocco™. Blue polymer tips
The two sources of blue plastic-tipped bullets encountered were Nosler (25 cal) and Hornady (5.7 x 28 mm). Both tips were molded from POM. When viewed with the VSC at 312 nm and 254 nm, the Nosler tips had a green color, and the Hornady exhibited an orange color (See Figure 3). XRF analyses, as well as FTIR-DA, were also useful in
differentiating these two brands.
Brown polymer tips
Nosler 0.264-inch/6.5-mm bullets were the only encountered source of brown plastic tips. These tips are manufactured from a POM polymer.
Green polymer tips
We encountered three sources of green, plastic-tipped bullets: Sierra (22 cal 40 grain BlitzKing), Federal Premium® (300 WSM Nosler Ballistic Tip®), and Nosler (30 caliber Trophy Grade Bullets™). All three polymer tips are manufactured with POM. XRF was able to distinguish the three classes based on their elemental composition (See Table 2). FTIR-DA also distinguished the polymer tips of the Sierra bullets from the tips of both the Nosler and Federal Premium® bullets. However, FTIR-DA could not distinguish between the polymer tips of the Federal Premium® and the Nosler bullets. It was interesting to note that the tips could be distinguished with XRF analysis but not FTIR-DA even though they are both from Nosler.
Grey polymer tips
Winchester ammunition and Nosler Combined Technology® Ballistic Silvertip® bullets are the only known sources of silver/grey plastic-tipped bullets [1]. Winchester 17 HMR cartridges are loaded with Hornady V-Max® bullets with a silver/grey tip and the Ballistic Silvertip® line of cartridges utilize bullets manufactured by Nosler with a silver/grey polymer tip. Bullets from Winchester 17 HMR V-Max®
Figure 3
Page 46 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 44
cartridges and Ballistic Silvertip® 7-mm bullets were sampled. Both tips were molded from POM. The VSC gave equivocal results, but the analysis with XRF and FTIR-DA was able to distinguish the polymer tips. While both polymer tips contained titanium, the Ballistic Silvertip® also contained aluminum and the V-MAX® tips contained iron and sulfur.
Orange polymer tips
There are two sources of orange polymer tips: Nosler Ballistic Tip® (22 caliber, 22-250 Remington, and 223 Remington) and Federal Premium® (223 Remington). All of the orange polymer tips are produced using POM. Neither VSC, XRF, nor FTIR-DA analyses could differentiate the different sources. This result is not surprising since all of the orange tips are made by Nosler.
Purple polymer tips
Two boxes of ammunition containing purple plastic-tipped bullets were analyzed, which are sold by Federal Premium® (243 Winchester, 204 Ruger) with the Nosler Ballistic Tip®. Both tips were molded from POM. Visually, the 204 Ruger polymer tips appeared to be more of a reddish-purple than
the 243 Winchester polymer tips. The two sources of purple plastic tips could be distinguished by XRF and/or FTIR-DA.
Red polymer tips
Three sources of red plastic-tipped bullets were analyzed: Hornady, Federal Premium®, and Nosler. Analysis indicates that Hornady uses two different types of polymer tips: POM and a polyester urethane - methylene bis(phenylisocyanate) copolymer. Federal Premium® and Nosler use tips molded from POM. Of the nine red polymer tip populations, only two could not be distinguished from one another with combined analysis, leaving eight different classes.
The VSC could not distinguish any color difference between the red tips. XRF analysis distinguished five major categories of plastics based on elemental composition (See Figure 4). Discriminant analysis was able to elucidate differences between each population with the exception of the two polyurethane tips (See Figure 4).
The polyester urethane - methylene bis(phenylisocyanate) copolymer tips by Hornady (308 cal. FTX® Marlin Express and 12 gauge 300 gr. FTX®, SST® shotgun slugs) were easily
Figure 4
Page 47 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 45
Nosler Inc., Dupont, Ticona, Paul Szabo from Winchester Ammunition©, and Doug Derner from Hornady Manufacturing Co. We also thank Zack Waterman for the Nosler Inc. factory tour in Bend, Oregon. We would also like to thank Bonnie Yates for her editorial comments and Dr. John Thornton for his thorough review.
References
[1] Mann, M.J. and Reinholz, A.D., “Physical Characteristics of Selected Polymer Tipped Rifle Bullets,” AFTE Journal, Vol. 29, No. 2, Spring 1997, pp. 165-169.[2] Nosler 2010 catalog p.7, Nosler, Inc. Bend, Oregon.[3] “LEVERevolution.” Hornady Manufacturing Company. <http://www.hornady.com/store/leverevolution>. Accessed 2011 Aug 16.[4]“The Polymer Tip Advantage.” Precision Rifle. <http://www.prbullet.comptq.htm>. Accessed 2011 Aug 16.[5] Hornady Product Catalog 2010, Hornady Manufacturing Company, Grand Island, NE, 2010. [6] Foster and Freeman VSC 6000 Video Spectral Comparator Software Manual, May 2010“VSC®6000”. Foster and Freeman Product Site and Brochure. <http://www.fosterfreeman.com>. Accessed 2010 Oct 8.[7] Rintoul, L., Panayiotou, H., Kokot, S., George, G., Cash, G., Frost, R., Bui, T., and Fredericks, P., “Fourier transform infrared spectrometry: a versatile technique for real world samples,” Analyst, Vol. 123, No. 4, 1998, pp. 571-577.[8] Kirkbride, K. P. and Tungol, M. W., “Infrared microspectroscopy of fibers,” in Forensic Examination of Fibers, 2nd edition, Robertson, J. and Grieve, M., Eds., Taylor and Francis, Philadelphia, PA., 1999, pp. 179-222. [9] Espinoza, E., Przybyla, J., and Cox, R., “Analysis of fiber blend using horizontal attenuated total reflection Fourier transform infrared and discriminant analysis,” Applied Spectroscopy, Vol. 60, No. 4, 2006, pp. 386-391.[10] Espinoza, E. O., Baker, B.W., Moores, T.D., and Voin, D., “Forensic identification of elephant and giraffe hair artifacts using HATR FTIR spectroscopy and discriminant analysis,” Endangered Species Research, Vol. 9, No. 3, 2008, pp. 239-246. [Special journal issue titled “Forensic Methods in Conservation Research”] http://www.int-res.com/articles/esr2009/9/n009p239.pdf[11] Espinoza, E.O, Baker, B. W., and Berry, C. A., “The analysis of sea turtle and bovid keratin artifacts using DRIFT spectroscopy and discriminant analysis,” Archaeometry, Vol. 49, No. 4, 2007, pp. 685-698.[12] Baeten, V. and Aparicio, R., “Edible oils and fat authentication by Fourier transform Raman spectrometry,” Biotechnology, Agronomy, Society and Environment, Vol. 4, No. 4, 2000, pp.196-203.[13] Enlow, E. M., Kennedy, J. L., Nieuwland, A. A., Hendrix,
identified by FTIR. The XRF analysis of both boxes showed similar elemental composition (Table 2), and FTIR-DA could not differentiate them. This is expected given that they are both Hornady FTX® molded tips [5].
White polymer tips
The three boxes of ammunition containing white plastic-tipped bullets were analyzed, which are distributed by Nosler Custom® (300 WSM, 7mm-08 Rem, and 308 Win.). Each of the boxes contained Nosler polymer tips (Accubond®) molded from POM. The VSC, XRF, and FTIR – DA could not distinguish between these samples. This was to be expected as each of the samples contained Nosler Accubond® tips [2].
Yellow polymer tip bullets
One source of yellow plastic-tipped bullets was analyzed: Nosler (270 cal). These yellow tips are manufactured from a POM polymer.
Conclusions
Plastic-tipped bullets seem to be here to stay. Plastic tips have successfully replaced exposed lead points in rifle bullets and have been incorporated into some of the popular handgun cartridges as well as shotgun slugs and muzzle loader projectiles. Plastic tips have also been added to lead free bullets [24].
Using color, FTIR coupled to the statistical power of discriminate analysis, and XRF, we were able to characterize the polymers found in common commercially available plastic-tipped bullets. The data is surprising because the high quality control used by the manufacturers provides an opportunity for forensic class character determination. In only one instance, a box of ammunition indicated that the polymer originated from two different batches of POM (CCI® 17 HMR). The analysis of the black plastic-tipped CCI® 17 HMR is also a good reminder that the analysis of bullet polymers will continue to evolve, and that future batches may be composed of different polymers or the elemental profiles may reveal elements or ratios not encountered in this study.
Disclaimer
The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the U.S. Fish and Wildlife Service.Acknowledgements
We thank the following for providing reference standards:
Page 48 of 49
AFTE Journal--Volume 44 Number 1--Winter 2012
Thompson et al. -- Chemical Properties of Selected Plastic-Tipped Bullets 46
J. E., and Morgan, S. L., “Discrimination of nylon polymers using attenuated total reflection mid-infrared spectra and multivariate statistical techniques,” Applied Spectroscopy, Vol. 59, No. 8, 2005, pp. 986-992.[14] Dharmaraj, S., Jamaludin, A.S., Razak, H.M., Valliappan, R., Ahmad, N.A., Harn, G.L., Ismail, Z., “The classification of Phyllanthus niruri Linn. according to location by infrared spectroscopy,” Vibrational Spectroscopy, Vol. 41, No. 1, 2006, pp. 68-72.[15] Widjaja, E. and Seah, R.K.H., “Use of Raman spectroscopy and multivariate classification techniques for the differentiation of fingernails and toenails,” Applied Spectroscopy, Vol. 60, No. 3, 2006, pp. 343-345.[16] Pyle, S.M., Nocerino, J.M., Deming, S.N, Palasota, J.A., Palasota, J.M., Miller, E.L, Hillman, D.C, Kuharic, C.A., Cole, W.H., Fitzpatrick, P.M., Watson, M.A., Nichols, K.D., “Comparison of AAS, ICP-AES, PSA, and XRF in determining lead and cadmium in soil,” Environmental Science & Technology, Vol. 30, No. 1, Dec. 1995, pp. 204-213.
[17] Strawn, D.G., Baker, L.L, “Speciation of Cu in a Contaminated Agricultural Soil Measured by XAFS, μ-XAFS, and μ-XRF,” Environmental Science & Technology, Vol. 42, No. 1, Nov. 2007, pp. 37-42[18] Dupont <http://plastics.dupont.com/plastics/pdflit/americas/delrin/DELRIN_Prod_Prop_11_06.pdf>. Accessed 2011 Sept 6.[19] Ticona, <http://www.ticona.com/products/celcon>, 8040 Dixie Highway, Florence, KY 41042. Accessed 2011 Sept 6.[20] Thermo Nicolet, Custom Software OMNIC v. 6.0, Nicolet Instrument Corporation, Madison, WI, 2003.[21] TQ Analyst, User Guide, Thermo Electron Corporation, Madison, WI, 1992.[22] Edax Orbis Vision user’s manual, August 20, 2009.[23] Chanda, M, and Roy, S.K. “Industrial polymers, specialty polymers and their applications.” Plastic Engineering, Vol. 73, CRC Press Boca Raton, Florida, 2008.[24] “Tipped TSX Bullet.” Barnes Bullets.<http://barnesbullets.com/products/rifle/tipped -tsx-bullet/>. Accessed 2011 Sept 6.
Page 49 of 49