mm illMINI/HHIINIWUINN!!!”UIHNIUHHIINUIWI

‘ _ 301568 3687

LIBRARY

Michigan State
University

 

 

 

This is to certify that the

thesis entitled

THE DEVELOPMENT OF A FTIR SPECTRAL

LIBRARY OF COMMERCIAL ADHESIVES
presente y

Jeffrey P. Kindig

has been accepted towards fulfillment
of the requirements for

Master of SCIencedegree in Criminal Justice with
Specialization in
Forensic Science

 

 

 

Mn
U “or pgfessor
Date 11M (”13 [717

0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

PLACE II RETURN BOXto remove thie checkwtirom your record.
TO AVOID FINES return on or before dete due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

AUG 1 7 200
571-9-4—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU II An Affirmative Action/Emil Opponunity lnetltuion

 

THE DEVELOPMENT OF A FTIR SPECTRAL LIBRARY
OF COMMERCIAL ADHESIVES

By

Jeffrey P. Kindig

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

MASTER OF SCIENCE

School of Criminal Justice

1997

ABSTRACT

THE DEVELOPMENT OF A FTIR SPECTRAL LIBRARY
OF COMMERCIAL ADI-IPSIVFS

By

Jeffrey P. Kindig

Adhesives can be an important class of evidence in the forensic science
laboratory, particularly in the investigation of food or pharmaceutical product
tamperings. However, the forensic significance of adhesive analysis using Fourier
Transform Infrared (FTIR) specu‘oscopyhasbcenhinderedbythefactthatnospecn'al
library of general adhesives is commercially available. Therefore, the purpose of this
research was to analyze samples of commercial adhesive products to compile into a
FTIR spectral library. Based on spectral comparisons, the commercial adhesives
wereabletobecategorized bysimilartype. Atestofthespectmllibrarywas
performed to demonstrate its validity. It was concluded that a comprehensive spectral
library of commercial adhesives will assist in the classification and identification of

unknown adhesives in forensic samples.

ACKNOWLEDGMENTS

I would like to thank Jay Siegel, my professor and advisor at Michigan State
University, for being a great teacher of forensic science, for accepting me into and
guiding me through the Master’s program, and for reviewing this manuscript. I
wouldalsoliketorecognizeFred FrickeandDuaneSatzgeroftheU.S. Foodand
Drug Administration's Forensic Chemistry Center (FCC). This project would not
havebeenpossiblewithouttheirsupport. IcreditlarryEllisoftheFCC for
initiatingthepmjcct, andlthankhimespecially forhelpfuldiscussionsandfor
teaching me good technique. Finally, I want to aclmowledge Caroline Machal of the

FCC for her valuable technical assistance and suggestions.

iii

TABLE OF CONTENTS

LIST OF TABLES ....................................... v
LIST OF FIGURES ...................................... v1
INTRODUCTION ........................................ 1
REVIEW OF LITERATURE ................................. 5
MATERIALS AND METHODS ............................... 8
RESULTS AND DISCUSSION .............................. 13
Super Glues ...................................... 13
Poly(Vinyl Acetate) and Related Adhesives ................... 1"]
Glue Sticks ....................................... 21
Poly(Alkyl Acrylate) Adhesives .......................... 21
Contact Cements ................................... 24
Hide Glues ....................................... 24
Silicones ........................................ 24
Plastic and Model Cements ............................. 30
Miscellaneous ..................................... 30
The Spectral Library ................................. 30
Test of the Spectral Library ............................ 31
CONCLUSION ........................................ 34
APPENDIX ........................................... 37
LIST OF REFERENCES .................................. 38

iv

LIST OF TABLES

Table l - Commercial Adhesives ............................... 9
Table 2 - Adhesives Used to Test the Spectral Library ................. 12
Table 3 - OMNIC Spectral Libraries ........................... 37

LIST OF FIGURES

Figure 1 - Super Glue Type Adhesives (A) ........................ 14
Figure 2 - Super Glue Type Adhesives (B) ........................ 15
Figure 3 - Super Glue Gel Type Adhesives ....................... 16
Figure 4 - Poly(Vinyl Acetate) Type Adhesives (A) . .................. 18
Figure 5 - Poly(Vinyl Acetate) Type Adhesives (B) .................. 19
Figure 6 - Poly(Vinyl Acetate) Related Type Adhesives ................ 20
Figure 7 - Glue Stick Type Adhesives ........................... 22
Figure 8 - Poly(Alkyl Acrylate) Type Adhesives .................... 23
Figure9 - Contact CementTypeAdhesives ....................... 25
Figure 10 - I-Iide Glue Type Adhesives .......................... 26
Figure 11 - Silicone Type Adhesives ........................... 27
Figure 12 - Plastic and Model Cement Type Adhesives ................ 28
Figure 13 - Miscellaneous Adhesives ........................... 29

vi

INTRODUCTION

Forensic science laboratories are responsible for analyzing many different
types of evidence using avariety of analytical techniques. The techniques used by
forensic scientists can be categorized as either 'wet' chemistry or instrumental
analysis. Examplesofwetchendcaltestsarethespotteststhataremutinelyuscdin
drug analysis. While this type of test is useful, it generally only provides 'yeslno'
information: either a spotting solution’s color changes or it does not. In contrast to
this, many of the scientific instruments used in instrumental analysis have the ability
to be more informative by providing a chemical profile of the sample. Depending on
the particular instrument being used, it may even be possible for the forensic scientist
to determine the identity of the substance being analyzed.

One technique of instrumental analysis with the power to potentially identify a
substance based on its chemical profile is Fourier Transform Infrared (FTIR)
spectroscopy. The instrument used for FTIR spectroscopy is called a spectrometer,
and it is operated using a computer. Inside the spectrometer, an infrared source gives
off light which passes through a sample compartment and then reaches a detector.
Many spectrometers are also equipped with an infrared microscope for the analysis of
very small samples. The sample of evidence to be analyzed (or 'measured') is placed
in the sample compartment or under the microscope. When the analysis is
performed, some of the infrared radiation from the source will be absorbed by the
sample. This absorbance is registered by the detector and represents the characteristic
chemical profile or I'spectrurn" of the sample being analyzed. The computer

1

2
interfaced with the instrument records and displays the spectrum of the sample,
allowing further data analysis to be performed. The importance of this is that it may
be possible for the forensic scientist to identify the sample based on its spectrum.
This is most commonly accomplished through the use of computerized "spectral
libraries. "

A spectral library is a searchable compilation of spectra, collected under
carefully controlled, optimized conditions. Spectral libraries are commercially
available for FTIR spectroscopy and other instrumental techniques, and may contain
several thousand spectra of pure compounds. There are also spectral libraries
available with specific forensic applications, such as those containing the spectra of
drugs, fibers, arson aceelerants, or explosives. Virtually all spectral libraries have
been computerized, so that an unknown sample spectrum can be rapidly searched
against the spectral libraries by the computer. Thus, a FTIR spectrum is analogous to
a fingerprint: just as the Automated Fingerprint Identification System (AFIS) can be
uscdtomatchanunknownprintwith oneinthecomputerdatabase, socanan
unlmownFl‘IRspecu'umbematchedwithaspecu'uminaspectrallibrary.

One class of evidence for which no FI'IR spectral library is currently
commercially available is general adhesives. The term “adhesive” as used in this
paperistobeinterpretedbroadly, andreferstoany materialusedtofastenonething
to another. When considering the different types of evidence commonly processed by
forensic science laboratories, adhesives may not immediately come to mind.
However, there are a number of situations in which adhesive analysis could play an

important role. In almost any offense involving the postal service, from mail fraud to

3
letter bombs, adhesives may be valuable evidence. Stamps, envelopes, and packaging
tapes all utilize adhesives, and laboratory analysis could demonstrate, for instance,
thatastamphadbeengluedtoanenvelopewithanadhesiveotherthantheoriginal
one on the stamp backing. In another case, a questioned document analysis may
involve a ransom note that was 'written' by cutting out words from a magazine and
gluingthem toapieceofpaper. Comparisonoftheadhesives fromtheransomnote
andfromabottleofgluesubsequently foundinthepossession ofasuspectcould
provide important corroborative evidence. Finally, adhesive analysis can play an
important role in cases involving food and pharmaceutical products.

Aside from fresh foods such as produce, the vast majority of food and
pharmaceutical products are sold in some outer container, and most of these
containers have adhesives present on them. Among other things, adhesives are used
toholdpaperlabelstometalsoupcans,toaffixfoilsafetywalstothemouthsof
medicinebottles,andtosealtheendflapsofcardboardcerealcartons. Ina
oomerfeitingcase, namebrandsouplabelsmaybeadheredtocanscontaining
generic soup. Inatampen‘ngcase, asealedcerealcarton maybeopened, thecereal
inside altered in some way, and the carton then rescaled. These examples of
counterfeiting and tampering cases both involve the adulteration of adhesives on
product containers. In both instances, proper processing of the evidence by the
forensic science laboratory would require a comparison of the counterfeit/tampered
product adhesive with the adhesive on an unadulterated product of the same type (a

”control" sample).

4

This comparison of the adhesives from the suspect and control samples can be
readily accomplished using FTIR spectroscopy with an infrared microscope. This is
an ideal technique for the analysis ofadhesives due to the ease of sample preparation,
thesrnallamountofsarnplercquired foranalysis, andthefactthatFITRspectroscopy
is non-destructive to the sample. Assuming that the suspect and control samples are
in fact different (i.e., the suspect sample has an additional adhesive), it would be
desirable to identify the suspect adhesive’s potential source. However, without a
spectral library of adhesives, the classification and identification of samples is difficult
to accomplish.

Therefore, thepurposeofthisresearchwastodevelopaFl‘IRspectrallibrary
of commercially available adhesives. FTIR spectra were collected of numerous
adhesive samples, and the spectra were then compiled into a spectral library. This
development should allow the classification and/or identification of suspect adhesives
in the types of cases described above. Commercial adhesives were chosen for the
spectral library due to their availability and their case of use. Whereas product
manufacturers typically use 'hot-melt' type adhesives to heat-seal their packages,
most commercial adhesives are designed to be applied as liquids and then solidify by
drying in air. This makes most commercial adhesives more practical for use by
offenders. Furthermore, if an offender applies an adhesive to a stamp, a piece of
paper, or a product container, chances are good that the adhesive was one the
offender had in his or her residence or office. Therefore, in all likelihood, the
adhesive was purchased commercially rather than being obtained directly from a

manufacturer.

REVIEW OF LITERATURE

A review of the literature reveals that many authors have discussed the topics
of spectral libraries, adhesive analysis, FTIR spectroscopy, and forensic science in
various combinations with one another. However, there have been very few papers
that bring together all four of these subjects, as is being done in this research.

Some novel uses of spectral libraries in the context of forensic science have
recently been described, though techniques other than FTIR spectroscopy were being
utilized. In two papers (1,2), high-performance liquid chromatography with diode
array detection (I-IPLCIDAD) was used to create spectral profiles of numerous drugs
of abuse. Then, a library was used to identify drugs extracted from blood in forensic
toxicology. Another paper described the use of a spectral library in conjunction with
proton nuclear magnetic resonance spectroscopy (NMR) to identify 'nitro-compounds'
from explosives (3). In a fourth article, gas chromatography (GC) profiles of some
common arson aecelerants were matched against the profiles of suspect samples (4).
These four papers demonstrate the potential utility of spectral libraries in the analysis
of forensic samples.

Techniques besides FTIR have been used in the forensic analysis of adhesives.
Two separate authors described the use of pyrolysis-gas chromatography (PGC) in
forensic science analyses, with adhesives being among the samples discussed (5,6).
In another paper (7), size-exclusion chromatography with electrochemical detection
was used to detect the compound nitrocellulose. This compound may be found in
trace amounts in many items, including some adhesives. While these three papers

5

6
point out some forensic aspects of adhesive analysis, they do not focus on adhesives
as their primary subject.

Several articles have described the use of FTIR spectroscopy in the analysis of
specific adhesive types, including anaerobic sealants and adhesives (8) and the
adhesives in multi-layer polymer laminates in packaging materials (9). In an analysis
of water-based adhesives (10), a FTIR spectral library was used to identify specific
components within the adhesive matrix. Among those papers that apply directly to
forensic science using FTIR spectroscopy, two described the analysis of the adhesives
on stamps (11,12). The materials used to contain illegal drugs, including adhesives,
were also analyzed by FTIR (13). Finally, another paper described the use of both
PGC and FTIR for adhesive analysis (14). All of these articles are important relative
to this research because they demonstrate the practicality of using FTIR spectroscopy
to analyze various adhesives.

Twopreviouspapersdescribedprojects similartotheresearchpresented here.
In 1966, in one of the first attempts at the forensic identification of adhesives,
researchers in Canada analyzed commercially available adhesives using a series of
tests that included paper chromatography, spot tests, PGC, and IR spectroscopy (the
precursor to FTIR). The authors were able to distinguish between different classes of
adhesives (15). The second paper, in 1974, described the analysis of commercially
available adhesives in Great Britain using PGC and IR spectroscopy. Once again, the
researchers demonstrated the ability to differentiate among the products analyzed (16).
In these two studies, the authors ended up with several infrared spectra of adhesives

that could be considered spectral libraries. However, without the ability to do

7
computer library searching, the researchers had to accomplish the comparison and
identification of their adhesive samples by hand-searching hard copies of the spectra.
Furthermore, because the authors were using conventional IR spectroscopy rather than
EIIR, they could not achieve the greatly improved signal-to-noise ratio provided by
Fourier transform instruments. Nevertheless, these two papers were the first

examples of spectral libraries of adhesives for forensic science purposes.

MATERIAIS AND METHODS

Atotalof56cornmercialadhesives werepurchasedtoconstitutetheF'I‘IR
spectral library. An alphabetical list ofthe products is shown in Table 1. These
products were purchased from six different locations (one hardware store, two
department stores, and three office supply stores) and represent almost all of the
adhesive products available at these six stores. An attempt was made to select
adhesives designed for a wide variety of purposes, so that several different adhesive
types would be represented in the spectral library. However, a premium was placed
onthosetypesofadhesives mostlikelytobeusedinoffensessuchastamperingwith
food or pharmaceutical products. Therefore, a number of 'school' or 'white' glues
and "super“ or 'krazy' glues were selected because they were considered very
common household adhesives. Also, most of the products purchased were name-
brand adhesives (e.g., Elmer’s, 3M, Instant Krazy Glue) because it was expected that
these would be the most widely available at different stores throughout the country.

Prior to analysis, the adhesive samples were applied to glass microscope slides
and allowed to dry in air. This was done for a number of reasons. First and
foremost, the majority of adhesive evidence that is brought into the forensic science
laboratory is likely to be in the dried form. However, even when this is not the case,
it is easy enough to allow an adhesive to dry before analysis. The reverse is not true:
in most instances it would be impossible to convert a dried adhesive back into its
original, undried form. A second reason for drying the adhesives is that sample
preparation is much less complicated with a dried solid than with a viscous, sticky

Table 1 - Commercial Adhesives

op

 

10
liquid. Finally, the solvent(s) present in an undried product may dominate the FTIR
spectrum and mask those spectral features due to the actual adhesive constituents.
Therefore, drying the adhesive samples was considered the most practical procedure.

For the majority of the adhesives, the sample drying time ranged fiorn 1 day
to l9days. 'l‘hisvariationindryingtimewasmostlydependentuponhowsoon, after
applyingthesamplewdreghssdidemdry,moppormmtywasfoundmconductdre
analysis. However, five of the samples were re-measured at later dates and have
drying times ranging from 34 days to 227 days. This occurred in the course of
double-checking the spectra of the adhesive samples to ensure their accuracy before
creating the spectral library.

The instrument used for the FTIR analysis of the adhesive samples was a
Nicolet Magna-IR"l 550 spectrometer equipped with a Nic-Plan"I infrared microscope
and a mercury-cadmium-telluride (MCT) detector. The samples were measured from .
4000 cm" to 650 cm" at a resolution of 4 cm" using a minimum of 64 scans. The
software used for data collection and analysis was OMNIC" version 3.1a (the
OMNIC software was upgraded twice—from version 2.1 to 3.0 to 3.1a—during the
course ofthe project, but these upgrades did not affect the data in any way). The
software was loaded with spectral libraries containing approximately 25,000 spectra.
A list of these spectral libraries is in the appendix.

Each adhesive sample was prepared for FTIR analysis by transferring a small
portion ofthe dried adhesive using a scalpel blade to one window ofa miniature
diamond anvil cell (High Pressure Diamond Optics, Inc.). The second window of the

diamond anvil cell was placed on top and the screws tightened to compress the

11
adhesivesarnpletoathinfilm. Thediamondanvilcellwasthendismantlcdanda
singlediamondwindowcontainingthethin filmofsamplewastransferredtothe
infrared microscope. The infrared spectrum of the adhesive sample was measured in
% transnrittance using a blank portion of the diamond window as the background.
Approximately fourspectraweremeasured ofeachsampletoensurethata
representative spectrumwascollected, withthebestspectrumbeing kept. Thefinal
specflumofeachadhesivewaswnverwdmabsorbanceandbasefinewnectcdusing
theOMNIC software. Fachadhesivespectrumwassearchedagainstthespectral
libraries loaded onto the OMNIC software. This assisted in the classification of the
adhesive samples by allowing comparison with the library spectra. Additionally, the
major components of some of the adhesives were identified based on the library
searches.

The preparation described above is also used for adhesives from actual case
samples. The surface with the adhesive residue present is observed using a
stereonricroscope, andascalpelbladeisusedtoremoveasmallportionofthe
adhesive. Caremustbetakennottotransferany fibers fromapaperorcardboard
surface to the diamond window along with the adhesive sample. To demonstrate this
sample preparation, and to test the spectral library of commercial adhesives upon its
completion, 11 of the adhesive samples were applied to a cardboard pharmaceutical
product carton and allowed to dry. These 11 'test' samples were arbitrarily selected
from within each of the adhesive categories described in the "Results and
Discussion,” and they are listed in Table 2. The dried test adhesives were prepared

and measured as described above.

12

Table 2 - Adhesives Used to Test the Spectral Library

Wm W“
Instant Krazy Glue - Wood Formula Super Glue

Power Poxy One-N-Done Super Glue Gel Super Glue Gel

Elmer’s School Glue ‘ Poly(Vinyl Acetate)

Glue Par Poly(Vinyl Acetate) Related
Elmer’s Glue Stick Glue Stick

3M Counter Top Cement Poly(Alkyl Acrylate)
DAP Weldwood Contact Cement Contact Cement

Titebond Liquid Hide Glue Hide Glue

Duro Clear Silicone Sealer Silicone

3M Plastic & Model Cement Plastic and Model Cement
Duco Cement Miscellaneous

"' Adhesive categories described in the 'RFSULTS AND DISCUSSION' section.

RESULTS AND DISCUSSION

Once the FTIR spectra of all 56 samples were collected, analysis of the data
revealed that many of the spectra could be categorized based on similarities to other
spectra. Those that could not be grouped were placed in the last category,
'miscellaneous.' For the most part, the categories of spectra were highly
characteristic, so that, in a forensic case, a suspect adhesive could at least be
classified if not identified. The results are arranged and discussed by category.
Super Glues

Fifteen 'super" glue type adhesives were analyzed, and all were observed to
havesimilarspectralfeatures. 'I'hiswastobeexpected,asallfifteenproductswere
labeled to contain a 'cyanoacrylate' compound. The spectra of the nine regular super
glues are shown in Figures 1 and 2. Among these nine spectra, the only one that
differs substantially from the others is the spectrum of Instant Krazy Glue -
All-Purpose / Skin Guard Formula. Since normal super glue rapidly bonds skin, the
Skin Guard Formula must contain an additional ingredient that alters this behavior and
gives rise to the additional spectral features observed. The spectra of the six super
glue gels are shown in Figure 3. The spectrum of Instant Krazy Glue - No Run Gel
appears very similar to the spectra of the regular super glues, with only variations in
some band intensities. However, the spectra of the other super glue gels obviously
exhibit additional strong bands. These spectral variations are due to the additional

compound(s) that give the gels their different consistency.

13

14

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«Duro Super Glue
1.0 ~
g 0.5-
2 .
0.0 ‘ fix a J
0 6 mer's Wonder Bond Plus Super Glue
g 0.4
s 02
0.0 /\~. - J L
uicktite High Performance Super Glue Pen
g 0.6
5 0.4
8
2 0.2
0.0 A k a J L
0.5 g oss Super Glue
g 0.4-:
2 ”i M
0.0 L k - J L
Super Duper Super Glue
g 0.6;
g 0.4-:
Q 02-:
0'0 -%\K_F=—_—#FL__
r e I—r—r
4000 3000 2000 1000
Wavenumbers (cm-1)

 

 

Figure 1 - Super Glue Type Adhesives (A)

15

 

1.04

Absorbence

0.5 1

0.0 ‘

d

d

06-
0

q

1

0.4-:

Absorbance

02{

0.0 3

 

’Power Foxy One-N-Done Super Glue

 

4_ JR. ca,__~l

 

 

 

 

 

0.8 jnstant Krazy Glue - Ail-Purpose

L and

 

 

 

 

 

1
1
0.5 1

0.4{

Absorbance

02-:

0.0 3

‘—

0.8 Jnstant Krazy Glue - Wood Formula

k a

 

 

 

J

g g %

 

Absorbance

0.2

0.0 L

 

4000

 

3000 2000
Wavenumbers (cm-1)

nstant Mane - All Purpose I Skin Guard Fonnulfl

 

 

 

 

;%

1000

 

Figure 2 - Super Glue Type Adhesives (B)

 

16

 

0.5 -

Absorbence

0.0

fnstant Krazy Glue - No Run Gel

A J

 

1...

2 fewer Poxy One-N-Done Super Glue Gel

 

 

k - it

 

0.5 a

Absorbance

0.0

ichbond 31;» Glue Gelg

 

 

A Jt

 

4
1a

Absorbance

0.5 ‘

Absorbance

0.0

13M Super Glue Gel

 

lOuicktite Super Glue Gel

 

A A

 

0.5 a

Absorbance

 

0.0 1

 

Puro Quick Gel

 

 

4000 3000 2000 1000

Wavenumbers (cm-1)

 

Figure 3 - Super Glue Gel Type Adhesives

 

1'7
1’0in inyl Acetate) and Related Adhesives

The common link between the various general purpose home, school, office,
and wood glues in this category is their similar chemical composition. These sixteen
products were all determined to contain poly(vinyl acetate), a poly(vinyl acetate)
copolymer, or poly(vinyl alcohol). Because none of these compounds were listed on
theproductlabels, theidentificationsweremadebasedonsearchesoftheOth
spectral libraries. Figure 4 displays the spectra ofsix adhesives that were determined
to contain poly(vinyl acetate). The spectrum of Franklin White Glue differs
somewhat in its band intensities, and the spectrum of Ross Office Glue Plus exhibits
threeadditionalweakbands. 'I‘hespectrainFigureSareallofwoodglues, andall
five products also were determined to contain poly(vinyl acetate). Among these, the
spectra of Titebond Wood Molding Glue and 3M Wood Glue exhibit additional weak
bands at the same frequencies as the additional bands in the Ross Office Glue Plus
spectrum. The spectrum of 3M Wood Glue also exhibits additional bands at
3000-2900 cm“. In all of the spectra in Figures 4 and 5, the slight fluctuation in
intensity of the broad band centered at about 3350 cm" may be due to the degree of
dryness 0f the sample.

The five spectra in Figure 6 represent compounds with structural similarity to
poly(vinyl acetate). Because of this, each of these spectra has certain similarities to
those in Figures 4 and 5. The spectrum of Seal-All, when searched against the
OMNIC spectral libraries, was observed to have spectral features similar to those in a
poly(vinyl chloride):poly(vinyl acetate) copolymer. It can be seen that the spectra of

Duro Household Cement and 3M Super Strength Adhesive also have similar spectral

18

 

Elmer's Glue-All
0.5 .

1

Absorbence

 

i
0.0

A J L M A A
Elmer‘s School Glue

s Washable School Glue

1.0:
0.5% A]
0.0 ‘ N J L

 

 

Absorbence

 

Absorbence

 

 

 

 

 

 

 

 

1.0 'lFranItlit‘l White Glue H
8 i
g 0.5 ‘
0.0 ‘ M J D“ W
1 0 . Itebond Home/School Glue
g 0.53
i 1 L Me.
0.0 d A J
Ross Office Glue Plus
g 1 d
2 .

 

 

 

 

 

1
0 # 5F$#_w_ m
I I I
4000 3000 2000 1000

Wavenumbers (cm-1)

 

 

 

Figure 4 - Poly(Vlnyl Acetate) Type Adhesives (A)

19

 

 

 

 

J[Elmer's Carpenters Wood Glue
g 1 .0 -
g 0.5-3
2 .
:Tltebond Original Wood GIue
g 1.03 l

 

 

 

g 0.5-:
0.0; A A lL/‘A-J

bond ll Premium Wood Glue

 

 

 

 

 

: Itebond Wood Molding Glue

 

0.03 M JLJU W
W.

Abeorbence
p
f

 

23M Wood Glue

Absorbanoe

 

 

 

 

 

 

4000 3000 2000 1000
Wavenumbers (cm-1)

 

 

 

Figure 5 - PolyMnyl Acetate) Type Adhesives (B)

20

 

 

 

 

 

 

 

1.0 4saga-Au
0.5 -
0.0 ‘ A J W W
puno Household Cement
0.6 -:
§ 3
g 0.4 1
W
0.0 3 k : J
0.8 13M Super Strength Adhesive
g 0.6-j
- 3
‘g 0.4 1
" :
0.0 i ‘ 4* J

 

 

 

02-:

Absorbence

0.1 {

0.0 3

0.3 :Elmer’s Sch lue Gel

_ MW

 

0.5 -

Absorbence

«Glue Pen

 

0.0

 

 

 

3000

2000 1000
Wavenumbers (cm-1)

 

 

Figure 6 - PolyMnyl Acetate) Related Adhesives

2 1
features, but have additional bands present. Based on the library searches, the spectra
of Elmer’s School Glue Gel and the Glue Pen were found to be consistent with
poly(vinyl alcohol). This compound is formed by the hydrolysis of poly(vinyl
acetate) (17), and thus the two substances are structurally related.
Glue Sticks ‘

The glue stick type adhesives analyzed were not labeled to contain any
particular compound(s), and the library search of the six spectra shown in Figure 7
did not yield any spectral matches. Five of these glue stick spectra exhibit many
similar spectral features, with some differences observed. The sixth spectrum, that of
Avery Removable Glue Stic, is very different from the other glue stick spectra but
was placed in this category for two reasons: to contrast it with the spectrum of Avery
Permanent Glue Stic and because it did not fit into any of the other categories
(except, of course, the miscellaneous category). The spectrum of Scotch Restickable
Adhesive Glue Stick, though of a glue stick, fit into the next category.

Poiy(Alkyl. Acrylate) Adhesives

The spectra of Scotch Restickable Adhesive Glue Stick and 3M Counter Top
Cement are not identical, but they exhibit many spectral similarities (see Figure 8).
Both of these spectra, when searched against the spectlal libraries, were observed to
exhibit spectral features similar to those in the spectra of poly(2-ethylhexyl acrylate)
and poly(butyl acrylate). However, neither adhesive spectrum exactly matched either
library spectrum. Therefore, it was determined that either the adhesives contain one

of the above compounds plus an additional ingredient(s), or else the adhesives contain

some other poly(alkyl acryiate).

22

 

:Elmer's Glue Stick

Absorbance
O
‘1'

0.0 ‘ ML

 

 

 

Elmer's School Glue Stick

05 ‘ /JL
0.0 V

 

 

 

 

 

UHU Stic

Absorbance

0.5-
o.o‘ A + J

 

 

 

 

 

Ross Envelop Stilt
0.5 -

00 NL

 

Absorbance

j

 

 

 

0 4 cAvery Permanent Glue S c

 

 

 

 

 

 

 

 

 

Wavenumbers (cm-1)

s “1
2 . N
0.0 ‘
«Avery Removable Glue S' c
g J
g 0.1 -
2
0.0 1......—
‘—+ r l P—r—i 1 r
4000 3000 2000 1000

 

 

Figure 7 - Glue Stick Type Adhesives

23

 

f5cotch Resticltable Adhesive Glue Stick
0.35 -

1

0.303
025%

0205

l

0.15-
,

q

0.105

0.05%
0001—...A] L__ _‘_J

Absorbance

 

 

 

 

 

 

 

 

 

0,45 - 3M Counter Top Cement

0.40-
0.35 +
A 0.30J
025«

0.20 -

Absorbance

0.15~

 

0.10-

 

0.05 -

0.00 - fl Avg)

 

 

 

 

 

 

4000 3000 2000
Wavenumbers (cm-1)

 

 

 

Figure 8 - PolylAlkyl Acrylate) Type Adhesives

24

Contact Cements

The spectra of the three contact cement type adhesives in Figure 9 exhibit
many similar spectral features. However, much like the glue stick type adhesives
above, the compound(s) in the contact cements could not be identified from either the
product containers or the library searches. Based on the complexity of the spectra in
Figure9andthespectralvariationsobserved fromonetothenext,itislikelythatthe
contact cement adhesives are in fact mixtures of more than one compound, which
would complicate identification. In any case, it is still clear that contact cements
represent a characteristic class of adhesives.
Hide Glues

Like the contact cements above, the two ”hide glues” in this category exhibit
very characteristic, if not readily identifiable, spectra (see Figure 10).. In this
instance, the spectra are nearly identical. Titebond Liquid Hide Glue is labeled to
contain the compounds ammonium rhodanate and dicyanodiarnide, and, based on the
spectra, Franklin Hide Glue (which is manufactured by the same company but has no
ingredients listed) contains a similar formulation.
Silicones

The spectra of the silicones shown in Figure 11 represent another category of
adhesives with highly characteristic spectra. One of the products, GE Silicone II, had
ingredients listed on its label, which were polydimethylsiloxane and
methoxypolydimethylsiloxane. All four of the silicones in this category exhibit very
similar spectra, and all four spectra matched well with a spectrum of

polydimethylsiloxane from the OMNIC spectral libraries.

25

 

0.51)» Weldwood Contact Cement

0.4:

0.35

Absorbance

 

 

 

«Duro Fast Drying Contact Cement

4

0.4— H
I
0.3~

 

 

Absorbance

——~—r~

025

q

 

0.13

0.05 N

 

 

023
0.1 : /\/J
0.0 ‘ __

 

3Elmer's Heavy Grip Cement
0.6 1

0.5~j

0.4-j

 

0.3-j ‘

Absorbance

025

 

0.1-j

 

0.0 3

-r

 

 

 

I T I r I I I

4000 3000 2000
Wavenumbers (cm-1)

 

 

Figure 9 - Contact Cement Type Adhesives

 

 

26

 

1.6 .Titebond liquid Hide Glue I)

d

1.4-:

1

1.21
1.0;

1

d

4

08-
o

q

d

Absorbance

 

1
i

1

0.45

02-:
0.0 i J

 

 

0.63
l

 

 

1.0 -lFranklin Hide Glue n
0.9 -
0.8 -
0.7 -
0.6 d

0.5 -

Absorbance

0.4 -

 

0.3 -

 

 

 

 

 

 

 

l

4000 3000 2000
Wavenumbers (cm-1)

 

1000

T

 

Figure 10 - Hide Glue Type Adhesives

 

27

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Elmer's Stix-All Hi-Tech Adhesive
0.6 a
E 0.4-:
2 02-:
lDuro Clear Silicone Sealer
1 .0 -
8 ..
S
e
3 0.5l
2 .
IDuro Stick With it Fix-All Adhesive
0.6 ~
g l
a 0.4 "‘
*3 l
8
2 023
0.0 ‘
.GE Silicone ll
3 0.43
c <
a
E
O
Q 4
2 03 -
0.0 ‘ i fiMAJ U E
1 1 1 r I 1 r r r —1_F l r r I .
4000 3000 2000 1000
Wavenumbers (cm-1)

 

 

 

Figure 1 1 - Silicone Type Adhesives

28

 

*3M Plastic 8. Model Cement

Absorbance
«'5

 

 

 

 

 

 

WKL - Mellllmwwlt

Duco Plastic 8. Model Cement

 

0.6~
0.5-
0.4-

i

g 0.3-

2

 

 

 

 

 

 

 

 

 

 

 

' ' l ' I ' f ' ' l
4000 3000 2000 1000

Wavenumbers (cm-1)

 

 

 

 

Figure 12 - Plastic and Model Cement Type Adhesives

29

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

oss Rubber Cement
g 1.0
G
'E
8 0.5
2 J)
0.0
JIDuco Cement
.
g 1 d n
2 j J
0 A A
.3M Water-Resistant Adhesive
g 1.0 «
fl
‘3 0.5-
2 l
0.0 * ¥ A
:Duro Fixture Adhesive
g 1.0 r
a .
E .
8 0.5 -
2 i
0.0 ‘ )«L N J
‘Pliobond General Purpose Adhesive
8 e
E 0.5 «
E .
3 .
2 . M W
0.0 _jl l
:Devcon 2-Ton Crystal Clear Epoxy
8 1.0d
c .
fl
2
8
2 W
I i T I l I
4000 3000 2000 1000
Wavenumbers (cm-1)

 

 

 

 

Figure 13 - Miscellaneous Adhesives

3 0

Plastic and Model Cements

Plastic and model cements are a final category of adhesives which are easily
identifiable. The similar spectra of 3M Plastic & Model Cement and Duco Plastic &
Model Cement are shown in Figure 12. Both of these products are labeled to contain
polystyrene, and both spectra were identified as such based on the library searches.
Miscellaneous I

Thespchaofmesixremainingadhesiveswerenotsinfilarmeachofllernor
were they similar to any of the spectra in the other categories (see Figure 13). This
fact would actually make them the most readily identifiable adhesives if present in a
forensic sample. However, some of these products are also ' unusual and not very
likely to appear in offenses like product tampering. The only spectrum in this
category that matched reasonably well with any of the spectra in the OMNIC spectral
libraries was that of Ross Rubber Cement, which showed some similarity to the
spectrum of poly(isoprene).
The Spectral Library

The FTIR spectra of the 56 adhesives were finally compiled into a new
“Commercial Adhesives" spectral library, created using the OMNIC software. When
a spectral library is created, information regarding the spectra that will comprise the
new library must be specified. This information includes the library frequency limits
(in cm"), the spectral data format, and the data spacing. For the Commercial
Adhesives spectral library, the specifications were set to match those of the baseline-
corrected adhesive spectra: the frequency limits were 4000 cm'1 to 650 cm", the

spectral data format was designated as 'Absorbance,‘ and the data spacing was set to

3 1

2 cm" (which is equivalent to a resolution of 4 cm“). Additionally, the software
provides for up to five ”fields“ of information to be entered with each spectrum when
added to the library. The first field is required to be the compound name, which in
thiscasewasactuallytheproductname. Thesecond fieldwaschosentobethedate
of the data collection for each sample. This was done to facilitate retrieving the
original data, if necessary, at a later time. Because the drying times of the adhesives
varied from sample to sample, this information was entered into the third field. The
fourth field was designated for the sampling technique used for data collection. In
this project, all the samples were prepared using a diamond anvil cell. However, the
spectral libraries are designed to be built upon, and it is possible that future
commercial adhesives may be analyzed using a different sampling technique and their
spectra added to the library. For this reason, a fifth field was also created for other
additional information, if necessary. This field was left blank for the 56 adhesives
sampled in this study, but future samples may require further details.
Test of the Spectral Library

The spectra of the 11 adhesive samples from the cardboard pharmaceutical
product carton were searched against the OMNIC spectral libraries, including the new
Commercial Adhesives library. This was a test of the new spectral library to
determine if an adhesive from a general category could be distinguished from other
adhesives with similar spectra.

When a search of the OMNIC spectral libraries is performed, the computer
essentially compares every data point in the spectrum being searched (the unknown

spectrum) with every data point in each of the library spectra. The computer assigns

3 2
a 'match quality“ (a number from 0 to 100) to each library spectrum based on how
closely it resembles the unknown spectrum. Then, a list of the best several matches
isreported, withthetopmatchbeingthelibrary spectrumthathasthehighestrnatch
quality, and therefore is mathematically the most similar to the unknown spectrum. It
should be noted that in identifying an unknown spectrum, the final determination of
the 'best' match is always made based on visual comparison with the library spectra.

When the 11 test adhesives were searched against the OMNIC spectral
libraries, the Commercial Adhesives library produced very good results. The test
adhesive was correctly identified by the top match in the library search in 9 of the 11
cases and was identified by the second match in the other two cases. For each of the
test adhesives, the general adhesive category could easily be determined from the top
several matches of the library search.

The 9 test adhesives identified by the top library match were: Power Poxy
One-N-Done Super Glue Gel, Elmer’s School Glue, Glue Pen, 3M Counter Top
Cement, DAP Weldwood Contact Cement, Titebond Liquid Hide Glue, Duro Clear
Silicone Sealer, 3M Plastic & Model Cement, and Duco Cement. Among these,
some have very characteristic spectra and there was little doubt that they could be
identified by the library. This was especially true for Duco Cement, since it is from
the miscellaneous category, and thus has a unique spectrum in the library. Similarly,
although Power Poxy One-N-Done Super Glue Gel, 3M Counter Top Cement, and
DAP Weldwood Contact Cement fit into categories, their spectra have distinct
differences from the other adhesives in their categories. The other five test adhesives

identified by top matches were much more telling about the utility of the Commercial

3 3
Adhesives spectral library. In particular, Elmer’s School Glue was distinguished from
several very similar poly(vinyl acetate) type adhesives, and the spectrum of Titebond
liquid Hide Glue was differentiated from the nearly identical spectrum of the other
hide glue.

The two adhesives identified by the second matches from the library searches
were Instant Krazy Glue - Wood Formula and Elmer’s Glue Stick. The top match for
the Instant Krazy Glue - Wood Formula test sample was Quicktite High Performance
Super Glue Pen, which is one of several super glue type adhesives with very similar
spectra. The top match for the Elmer’s Glue Stick test sample was UHU Stic; these
two glue sticks and Elmer’s School Glue Stick all have very similar spectra. Possible
explanations for the 'second—place finishes” of the two test adhesives are slight
variations in the sample preparation (e. g. , sample film thickness) or slight
heterogeneity in the adhesives. Nevertheless, the general adhesive category was
readily identifiable in both instances, which would still be important information in an

actual forensic case.

CONCLUSION

The field of forensic science evolves constantly as new technology regularly
becomes available. Each new development provides the forensic scientist with
another tool in the evaluation of evidence. Along these lines, the development of the
FTIR spectral library of commercial adhesives will be very beneficial in the forensic
science laboratory. In all future cases, adhesives from forensic samples can be
analyzed by FTIR and their spectra searched against all available spectral libraries,
including the Commercial Adhesives library. If a commercially available adhesive
has been used in a product tampering or other offense, the spectral library search may
allow the suspect sample to be classified into a general adhesive category. If the
suspect adhesive spectrum matches particularly well with a single adhesive spectrum
from the library, then a preliminary identification of the suspect adhesive may be
made. Either of these instances supplies the forensic scientist with important
information.

The value of the scientific information obtained from adhesive evidence
extends from the forensic science laboratory to the investigation process. In a
particular case involving adhesives, the forensic scientist can potentially furnish the
investigator with a specific product name to search for in the possession of a suspect.
Additionally, the increased value of adhesive evidence provided by the Commercial
Adhesives spectral library should encourage the investigator to collect adhesives for

forensic analysis in all applicable cases.

34

3 5

The development of the FTIR spectral library of commercial adhesives also
has the potential to save a great deal of analysis time in the forensic science
laboratory. Before the library existed, attempting to classify or identify a suspect
adhesive was a difficult task. Several different products might need to be analyzed
before one was found that resembled or matched the suspect sample. However, with
the Commercial Adhesives library, classification or identification of a suspect sample
can potentially be accomplished by simply conducting a library search. The time
saved in this process gives the forensic scientist the opportunity to pursue other
analyses.

'l‘hereareanumberofareasoffutureresearch thatcouldbepursuedtobuild
upon the work presented in this paper. First of all, it is acknowledged that there are
many more products in the entire world of adhesives, and that the Commercial
Adhesives library may never be totally complete. Therefore, as other commercially
available adhesives are encountered in the future, their spectra should be added to the
spectral library. Additionally, it may be desirable to obtain the adhesives used by
product manufacturers for inclusion in a sepalate spectral library.

Another topic worthy of future study is the effect of drying time on adhesive
samples. Dried adhesives could be studied to observe if significant spectral
differences are noted over time. Most liquid adhesives utilize some solvent which
suspends the adhesive components, and the adhesive dries as the solvent evaporates.
Therefore, one way in which an adhesive’s spectrum may change over time is by the
disappearance of solvent bands. If the subject of drying time variations was

thoroughly explored, it may theoretically be possible in the future to not only identify

3 6
an adhesive based on its FTIR spectrum, but to determine approximately how long
ago the adhesive was applied. Though this would require a great deal of future work,
the ability to estimate the application time of an adhesive would be very valuable in
the context of forensic science.

A final area of future research involves the use of different sampling
techniques or instrumentsin the analysis of adhesives. In FTIR spectroscopy, a
technique known as Attenuated Total Reflectance (ATR) could be used to measure the
spectrum of an adhesive without removing it from the surface to which it is bound.
This would have practical application to very thin adhesive films. Also, some past
research used the instrumental technique of pyrolysis-gas chromatography (PGC) in
adhesive analysis. Like FTIR, this technique has the ability to generate a
characteristic profile of an adhesive sample, and therefore it also offers the potential
to create a library of adhesives. Using both PGC and FI'IR would provide two pieces
of information about a sample, and may therefore allow similar samples to be more

easily differentiated.

APPENDIX

APPENDIX

Table 3 - OMNIC Spectral libraries

We

Aldrich Condensed Phase

Aldrich Condensed Phase Supplement
Aldrich Vapor Phase Sample Library
*Black Plastic Tape Adhesive - 8 cm"
‘Black Plastic Tape Plasticizers

‘Black Plastic Tape-Back Side
*Extracted Inks (Limited)

*FBI Fibers (Version 3.1)

Georgia State Crime Lab Sample Library
*Glues and Adhesives

Hazardous Chemicals - Condensed Phase Spectra
Hummel Polymer Sample Library
‘Photocopy Toners by R-A, 4 cm" Resolution
I"Raw Inks in Pyridine (Limited)

Rubber Compounding Materials

Sigma Biological Sample Library

Sigma Steroids

Sprouse Polymer Additives

*8me

Synthetic Fibers by Microscope

Toronto Forensic

U.S. Geological Survey Minerals

‘User-created iibrariesobtainedfromtheFedenl Bureau ofInvestigstion (FBI)
'User-creatcd library ofadhesives fiomcaaeworkaampies

’User-created libraryofhousestandards
AllotherlibrariespurchasedfmmNicoletInstrumcntCorporation.

37

411
53
510

350

3,01 l
325
20
376
3,549

25,293

LIST OF REFERENCE

(1)

(2)

(3)

(4)

(5)

(6)

(8)

C9)

(10)

(11)

LIST OF REFERENCES

Koves EM. Use of high-performance liquid chromatography-diode array
detection in forensic toxicology. J Chromatogr A 1995 Feb lO;692(l-2):103-
l9.

Tracqui A, Kintz P, Mangin P. Systematic toxicological analysis using
HPLC/DAD. J Forensic Sci 1995 Mar;40(2):254-62.

Burns DT, Lewis RJ. Analysis and characterisation of nitroglycerine—based
explosives by proton magnetic resonance spectrometry. Anal Chim Acta 1995
Jan 20;300(l-3):221-5.

Koussiafes P, Bertsch W. Profile matching for the analysis ofaccelerants in
suspected arson cases. J Chromatogr Sci 1993 Apr;31(4):137-44.

Wheels BB. Analytical pyrolysis techniques in forensic science. J Anal Appl
Pyrolysis l981;2(4):277-92.

Challinor 1M. Forensic applications of pyrolysis capillary gas
chromatography. Forensic Sci Int l983;2l(3):269«85.

Lloyd JBF. Detection and differentiation of nitrocellulose traces of forensic
science interest with reductive-mode electrochemical detection at a pendant
mercury-drop electrode coupled with size-exclusion chromatography. Anal
Chem 1984 Sep;56(ll):l907-12.

Leonard RG, Brennan M, Quigley K, Heatley D. Analysis of anaerobic
sealants and adhesives. Anal Proc (London) 1992 Sep;29(9):393-6.

Mandanis A. Identification of packaging material constituents. Analusis 1993
Jun;21(5):M22-M26.

Fenner RA. Analysis of water-based adhesives using internal reflectance and
curve-fitting. Appl Spectrosc 1984 Jan-Feb;38(l):84-6.

Poslusny M, Daugherty KB. Nondestructive adhesive analysis on stamps by
Fourier transform infrared spectroscopy. Appl Spectrosc 1988 Nov-
Dec;42(8): 1466-9.

38

(12)

(13)

(14)

(15)

(16)

(17)

39

Wang JH. A versatile analytical method of identifying adhesive on stamps by
specular-reflectance Fourier transform infrared spectroscopy. Appl Spectrosc
1990 Mar-Apr;44(3):447-50.

Sachdev HS, Reed GD. Spectroscopic analysis of illicit drugs and their
containment materials. Spectrosc World 1990 Sep-Oct;2(5):27—8,30-2.

Lennard CJ, Mazulla WD. Simple combined technique for the analysis of
toners and adhesives. J Forensic Sci Soc 1991 Jul-Sep;31(3):365-7l.

Caldde BB, Smith M. The comparison and identification of adhesives on
questioned documents. J Forensic Sci 1966 Jan;ll(1):28-42.

Noble W, Wheals BB, Whitehouse, MI. The characterisation of adhesives by
pyrolysis gas chromatography and infrared spectroscopy. Forensic Sci
1974;3:163-74.

Jaffe HL, Rosenblum FM. Poly(viny1 alcohol) for adhesives. In: Skeist I,
editor. Handbook of adhesives. 3rd ed. New York: Van Nostrand Reinhold,
1990;401-7.

"‘llllllllllllllll“