'(‘Hiu

 

 

 

 

 

 

r "3

 

.“"_.,‘\I“V -—_- mu.‘ ... .

.

 

ms "

*fiuwéfl

>._ .
354.39..

{31

 

\ .
:uxthh...
5......

1:23.19...
r. X\.

§ x31
.3535. 9.. ‘5“...

d w , 1;
f e .. . . A 1....
.3... 1.. I .15. 00
2...
.2. Cassie!
3.. fl.

 

 

1113318

This is to certify that the
thesis entitled

Detection and Characterization of Multiply-Charged Ink Dyes
by Matrix-Assisted Laser Desorption/Ionization Mass
Spectrometry with 2-(4-Hydroxyphenylazo)-Benzoic Acid and
Diammonium Hydrogen Citrate

presented by

 

1

Jamie D. Dunn

LIBRARY
Michigan State
University

has been accepted towards fulfillment
of the requirements for the

 

 

 

MS. degree in Chemistry

 

 

OM: W

 

Major Professor’ 5 Signature

575.07

 

Date

MSU is an Affim'iative Action/Equal Opportunity Institution

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

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

 

DETECTION AND CHARACTERIZATION OF MULTIPLY—CHARGED
INK DYES BY MATRIX-ASSISTED LASER DESORPTION/IONIZATION MASS
SPECTROMETRY WITH 2-(4~HYDROXYPHENYLAZO)-BENZOIC ACID AND

DIAMMONIUM HYDROGEN CITRATE
By

Jamie D. Dunn

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirement
for the degree of
MASTER OF SCIENCE

Department of Chemistry

2007

ABTRACT

DETECTION AND CHARACTERIZATION OF MULTIPLY-CHARGED
INK DYES BY MATRIX-ASSISTED LASER DESORPTION/IONIZATION MASS
SPECTROMETRY WITH 2-(4-HYDROXYPHENYLAZO)-BENZOIC ACID AND

DIAMMONIUM HYDROGEN CITRATE
By

Jamie D. Dunn

Laser desorption mass spectrometry (LDMS) is a technique limited to detecting
singly-charged or neutral molecules. Several inks contain dyes that are multiply-charged,
hence, LDMS can not be employed directly for their identification. However, the
technique can be slightly modified by using a matrix. Matrix-assisted laser
desorption/ionization mass spectrometry allows for the detection of multiply-charged
species. Polyionic dyes are not sufficiently detected in the MALDI experiment, but can
be chemically modified through the use of an additive with a matrix, allowing cation-
exchange to occur. We have successfully detected 30 polyionic dyes by MALDI MS
with the matrix, HABA, and the additive, DAHC. A method was developed for the
detection and characterization of ink dyes that contain multiple charges, specifically,
polysulfonated dyes as well as other dyes containing sodium carboxylate and sodium
alkoxide groups. Inks containing multiply-charged dyes have been examined in this
research by MALDI MS, LDMS, and TLC. Although LDMS does not allow for the
direct detection of polyionic dyes, both mass spectrometry techniques contain structural
information which was used for characterization purposes. TLC was indirectly coupled

with MALDI MS to characterize the dyes in ink.

ACKNOWLEDGEMENTS

I would like to thank Dr. Jay Siegel, Dr. Gary Blanchard and Lisa Dillingham for
helping me survive Dr. Allison’s retirement, pushing me to complete two theses
Simultaneously, and allowing me to continue my future work in the department. I would
also like to thank Dr. Merlin Bruening and Yingda Xu for giving me the opportunity to
work on a new, exciting project. I thank Dr. John Allison for the MS knowledge that I
have acquired and can use in future research.

My family has continued to be very supportive of my choices and educational
goals and I would like to thank them for being understanding when I could not come
home for family gatherings. I also want to thank my friends, Stephanie and Amy Jo, for
staying in touch over the years and helping me to forget my life as a graduate student
when I would come home to St. Louis.

I would especially like to thank my friend/labmate, Anne Distler, for the great
times we shared in the lab and at your condo. Since your graduation, your emails have
entertained me as well and have kept me sane during my stay in MI. I also appreciated
your knowledge and your help with courses and research.

D.J., I couldn’t have survived these insane months without you by my side. You
gave me strength and love when I needed it the most. I thank you for not giving up! I
give thanks to God for watching over me and bringing us together.

Finally, I thank the National Institute of Justice for funding this research project
for two years. We couldn’t have successfully completed the project without the financial

support.

iii

TABLE OF CONTENTS

List of Tables
List of Figures
Chapter One: Introduction

Dye Classification
Dye Application
Pen Ink Composition
Ink Dye Analysis
Selection of an Analytical Method
Methods for Ink Dye Analysis
Mass Spectrometry
Field Desorption MS
Electrospray Ionization MS
Laser Desorption MS
Chromatography
Thin-layer Chromatography
High Performance Liquid Chromatography
LDMS and Forensic Science Applications
Questioned Document Examination
Trace Evidence
Research Overview
References

Chapter Two: Direct LDMS and TLC Analysis of Ink Dyes

LDMS Calibration and Copper Phthalocyanine
Mass Accuracy for Identification

LDMS Ink Dye Analysis
Molecular Information: Mass, Composition, and Charge State
Dye Impurities, Fragment Ions, and Cluster Ions

Coupling TLC with LDMS

References

Chapter Three: Direct LDMS Analysis of Ink Dyes: Limitations and
Solutions

Dye Characterization by Photodegradation and LDMS
Isomeric Dye Characterization

Rhodamine B

Rhodamine 6G
References

iv

vi

vii

43
43

45
55
63
68

7O

71
79
80
84
87

Chapter Four: Detection and Characterization of Multiply-Charged Ink
Dyes

Determining the Presence of Multiply-Charged Dyes in Pen Ink
LDMS
TLC
Ink Patents

Multiply-Charged Ink Dye Detection
Cation-Exchange Additives and Matrices
Commercial Dyes
Highly-Charged Dyes
Ink on Paper

Multiply-Charged Ink Dye Characterization
MALDI MS and TLC
Direct LDMS

References

Chapter Five: Conclusions

MALDI MS Limitations and Future Research
References

88

92
92
99
99
100
101
104
107
108
122
123
138
139

142

142
144

LIST OF TABLES

Chapter Two: Direct LDMS and TLC Analysis of Ink Dyes

Table 2.1: Isotopic masses and natural abundances for the atoms that
comprise Copper Phthalocyanine 51

Chapter Three: Direct LDMS Analysis of Ink Dyes: Limitations and
Solutions

Table 3.1 m/z values of the dye and corresponding deethylated degradation
products (n represents the number of ethyl groups bonded to the
amine nitrogen atoms) for a) Rhodamine B and b) Rhodamine 6G 78

Chapter Four: Detection and Characterization of Multiply-Charged Ink
Dyes

Table 4.1 Polyionic dyes examined by MALDI MS (HABA and DAHC) on
SS plate. Provides m/z values detected in the positive and/or
negative-ion mode and functional groups that may contribute to
the overall charge on the dye molecule 106

vi

Chapter One:
Figure 1.1:
Figure 1.2:
Figure 1.3:

Figure 1.4:

Figure 1.5:

Figure 1.6:

Figure 1.7:

Figure 1.8:

Figure 1.9:

Chapter Two:

Figure 2.1:

Figure 2.2:

LIST OF FIGURES

Introduction
Basic schematic of a mass spectrometer
Diagram of theVoyager-DE Mass Spectrometer
Schematic of an ESI source

Multiply-charged dye structures a) Acid Blue 9, b) Acid Blue
92, and c) Solvent Brown 20

Neutral dye structures a) Solvent Yellow 19, b) Solvent Orange
3, c) Copper Phthalocyanine, and d) Pigment Red 3.

LD mass Spectra of Pigment Red 3 on paper a) positive-ion and
b) negative-ion

Positive-ion LD mass Spectra of polyethylene glycol (PEG) on
paper a) complete spectrum from m/z 0-600 and b) expanded
region from m/z 250-600.

Polyethylene glycol a) m/z values that correspond to the
ionization of the polymer by sodium and potassium ion
adduction and b) basic polymer structure (It is the number of
repeating units)

Positive-ion LD mass spectra of black liquid from a Pilot®
rollerball ink a) complete spectrum from m/z 0-700 and b)
expanded region from m/z 250-600

Chapter Two: Direct LDMS and TLC Analysis of Ink
Dyes

Dye structures a) Copper Phthalocyanine and b) monosulfonated
Copper Phthalocyanine

LD mass spectra of blue gel ink containing Copper
Phthalocyanine a) positive-ion and b) negative-ion

vii

ll

13

l7

19

20

23

24

25

47

48

Figure 2.3:

Figure 2.4:

Figure 2.5:

Figure 2.6:

Figure 2.7:

Figure 2.8:

Figure 2.9:

Figure 2.10:

Figure 2.11:

Figure 2.12:

Figure 2.13:

The isotopic peak distribution of Copper Phthalocyanine a)
experimental isotopic distribution in negative-ion mode b)
theoretical isotopic distribution for positive and negative-ion
modes

Structure of the dyes used in several black ballpoint inks a)
Crystal Violet and b) Metanil Yellow

LD mass Spectra of black ink from Sanford® Supergrip ballpoint
pen a) positive-ion and b) negative-ion

Positive-ion LD mass spectrum of Methyl Violet ZB purchased
from the Aldrich Chemical Company

The direct analysis of Metanil Yellow from a metal plate by
negative-ion LDMS

Negative-ion LD mass spectrum of Metanil Yellow Shows peaks
due to fragment ions

Dye structure of Metanil Yellow with labeled values that
correspond to possible fragment ions or neutral molecules

Acid Orange 7 (to the left of the arrows) and fragment ions (m/z
80, 156, 171) formed when analyzed by ESI tandem MS and
MALDI-PSD MS”

Negative-ion LD mass spectra of Metanil Yellow on paper
shows peaks due to cluster ions

TLC analysis of black ballpoint ink from BIC® Cristal a) thin-
layer chromatogram and b) TLC bands and corresponding dyes

Direct coupling of TLC with LDMS: bands are analyzed by
LDMS directly from the aluminum TLC plate a) purple band
(Crystal Violet) and b) yellow band (Metanil Yellow)

Chapter Three: Direct LDMS Analysis of Ink Dyes: Limitations and

Figure 3.1:

Figure 3.2:

Solutions

Positive-ion LD mass spectra of isomeric, red ink dyes a)
Rhodamine B and b) Rhodamine 60

Crystal Violet a) dye structure and b) m/z values of the dye and

corresponding degradation products (11 represents the number of
methyl groups bonded to the amine nitrogen atoms)

viii

50

53

54

56

59

59

60

60

63

66

67

73

75

Figure 3.3:

Figure 3.4:

Figure 3.5:

Figure 3.6:

Figure 3.7:

Chapter Four:

Figure 4.1:

Figure 4.2:

Figure 4.3:

Figure 4.4:

Figure 4.5:

Figure 4.6:

Figure 4.7:

Figure 4.8:

Positive-ion LD mass spectra of blue ink from BIC® Cristal
ballpoint pen degraded with UV radiation a) complete spectrum
from m/z 0-500 and b) expanded region from m/z 250-400

Dye structures of a) Rhodamine B and b) Rhodamine 6G

Positive-ion LD mass spectra of Rhodamine B (a) no exposure
to incandescent light and (b) exposed to incandescent light for
12 hours

Positive-ion mass spectra of Rhodamine B analyzed by a)
LDMS and b) MALDI MS

Positive-ion LD mass Spectra of Rhodamine 6G (a) no exposure
to incandescent light and (b) exposed to incandescent light for
12 hours

Detection and Characterization of Multiply-Charged Ink
Dyes

Direct LDMS analysis of liquid, blue ink from a Pilot® Precise
V7 rollerball pen on paper a) positive-ion mass spectrum and b)
negative-ion mass spectrum

Structure of multiply-charged ink dyes a) Direct Red 23, b) Acid
Blue 93, c) Direct Red 80, (1) Acid Yellow 23, e) Solvent Blue
38, f) Acid Black 1, and g) Reactive Black 5

Negative-ion MALDI MS analysis (HABA and DAHC) of
Direct Red 80 on SS plate

Positive-ion LD mass spectra of blue, liquid ink from Pilot®
Precise V7 rollerball pen a) HABA, b) DAHC, and c) HABA
and DAHC
Negative-ion LD mass Spectra of blue, liquid ink from a Pilot®
Precise V7 rollerball pen a) HABA, b) DAHC, and c) HABA
and DAHC

MALDI MS analysis of Acid Violet 49 on SS plate a) positive-
ion mass spectrum and b) negative-ion mass spectrum

Dye structure of Acid Violet 49

MALDI MS analysis of Acid Violet 49 on paper a) positive-ion
mass spectrum and b) negative-ion mass spectrum

ix

76

78

82

83

86

94

97

108

110

111

114

115

117

Figure 4.9:

Figure 4.10:

Figure 4.11:

Figure 4.12:

Figure 4.13:

Figure 4.14:

Figure 4.15:

Figure 4.16:

Figure 4.17:

Figure 4.18:

Figure 4.19:

Figure 4.20:

Figure 4.21:

MALDI MS analysis of liquid, blue ink from Pilot® Precise V7
rollerball pen on paper a) positive-ion mass Spectrum and b)
negative-ion mass spectrum

Dye structure of Acid Blue 90

MALDI MS analysis of Acid Blue 90 on SS plate a) positive-ion
mass Spectrum and b) negative-ion mass Spectrum

MALDI MS analysis of liquid, blue ink from a Sanford® Uni-
ball® Roller pen on SS plate a) positive-ion mass spectrum and
b) negative-ion mass spectrum

Dye structures a) Acid Red 52 and b) Acid Blue 9

MALDI MS analysis of Acid Red 52 on SS plate a) positive-ion
mass spectrum and b) negative-ion mass spectrum

MALDI MS analysis of Acid Blue 9 on SS plate a) positive-ion
mass spectrum and b) negative-ion mass spectrum

MALDI MS analysis of red, liquid ink from Sanford® Uni—ball®
Roller pen on SS plate a) positive-ion mass spectrum and b)
negative-ion mass spectrum

Indirect coupling of TLC with MALDI MS: negative-ion mass
Spectra of extracted TLC bands a) yellow band, b) orange-red
band, and 0) red band

Dye structures of a) Food Yellow 3, b) Acid Red 87, and c) Acid
Red 92

Direct negative-ion LDMS analysis of liquid, red ink from
Sanford® Uni-ball® Roller pen a) complete Spectrum from m/z
0—100 and b) expanded spectrum from m/z 600—800

Isotopic peak distributions of Acid Red 87 a) experimental and
b) theoretical

Isotopic peak distributions of Acid Red 92 a) experimental and
b) theoretical

119

120

121

125

127

128

129

132

133

134

135

136

137

Chapter One: Introduction

Dye Classification

The classification of colorants (dyes and pigments) can be found in the Colour
Index (CI). The latest edition was revised in 1971 and classifies over 10,500 dyes.
Colorants are classified based on their chemical composition and industrial application”.
The chemical composition is the foundation for assigning each dye a unique five-digit CI
number. The color and application of a dye is denoted by a CI name. For example, the
CI number and name of Crystal Violet are CI 42555 and Basic Violet 3, respectively.
The Colour Index recognizes the following 15 application classes: acid dyes, anionic
dyes, azoic dyes, basic dyes, direct dyes, disperse dyes, fluorescent brighteners, food
dyes, ingrain dyes, mordant dyes, pigment dyes, reactive dyes, solvent dyes, sulfur dyes,
and vat dyes3. The colorants used in this thesis research encompass multiply-charged
(polyionic), specifically, acid and direct dyes. A particular metal complex pigment dye
was examined as well and was used as the mass spectrometric calibrant for analyzing the
dye used in this research.

Dyes are aromatic, organic compounds that absorb electromagnetic radiation
including visible light with wavelengths 350 -— 700 nm4 and UV radiation which allow for
easy detection in UV laser desorption mass spectrometry. Chromophores and
auxochromes that are contained in the dye structure are responsible for the color of the
compound and are the chemical constitutes that distinguish the dyes (20-30 various
groups)“. Typical Chromophores include -C=C, —C=N, -C=O, -N=N, and -N02, and

quinoid rings“. Principal, functional groups such as -NH3, -COOH, -SO3H, and —OH

comprise auxochromes'i’5 (the hydrogen atoms are frequently substituted). Some of the
way dye groups include azo (monoazo, diazo, triazo, etc.), anthraquinone,
phthalocyanine, triarylmethane, diarylmethane, indigoid, azine, oxazine, thiazine,
xanthene, nitro, nitroso, methane, thiazole, indamine, indophenol, lactone, aminoketone,
and hydroxyketone dyes”.

This research has focused on multiply-charged dyes, specifically, acid or direct
dyes that are polysulfonated containing azo groups, used in pen inks. In general, acid
dyes are anionic containing one or more hydrogen sulfonic acid (-SO3H) or sodium
sulfonic acid (-SOgNa) groups. Other acid dyes contain sodium carboxylate (RCOz' Na+)
and sodium alkoxide (RO' Na+) groups which were also studied in this research. Acid
dyes comprise the largest class of dyes. The CI lists about 2300 different acid dyes, but
roughly 40% of these compounds are used currently in industrys. Azo, anthraquinone,
and triarylmethane compounds comprise the majority of the acid dyes. Most often, direct
dyes are high molecular weight, polyazo compounds. About 30% of direct dyes (~1600)
are presently used in industrys. Copper Phthalocyanine was studied by LDMS and used
as the mass spectrometric calibrant in this research. Copper Phthalocyanine (Pigment
blue 15) is classified as a pigment dye. Specifically, the dye is a phthalocyanine metal
complex containing copper. Phthalocyanine metal complexes as well as azo compounds
represent the majority of pigment dyess. Pigment dyes are insoluble, non-ionic
compounds or insoluble salts making them suitable as colorants in gel ink. The
molecular properties of Copper Phthalocyanine were studied to demonstrate the basic
characterization information that can be acquired. Also, the dye served to be an excellent

LDMS and MALDI MS calibrant for this research. Both the non-ionic nature of the

complex which allows the dye to be detected in both positive and negative-ion modes and
the molecular weight of the complex make the species an ideal candidate for calibrating

the mass spectrometer.

Dye Application

The application in which the dye will be used depends on the molecular structure
and how the molecule interacts with the substrate. For example reactive dyes form
covalent bonds (strongest interactions) with -OH, -NH, or -SH groups in the substrate
materials, the acid and basic dyes interact with corresponding ionic species (acid dyes
interact with cations and basic dyes interact with anions) that are contained in the
substrate through ionic bonding, and pigment dyes have the weakest intermolecular
interactions with the substrate. Manufacturers take advantage of the intermolecular
interactions and chemical bonding for selecting the dyes used in the industry. Frequently,
ink used for the manufacture of ballpoint pens use basic and acid dyes where the

production of gel inks rely on the use of pigment dyes as the colorants.

Pen Ink Composition

The selection of analytical methodology greatly depends on the species being
examined. The more recent methods that have been used to examine ink have been
developed based on the colorants in the ink. Although, ink is a complex mixture and,
frequently, manufactures do not disclose the ingredients, a generalization can be made
about the compounds used in the production of ink. While the specific ink composition

depends on the pen type, ink is comprised of colorants (dyes) and a colorless vehicle.

The vehicle serves to distribute dye onto paper. Ink patent information is available to the
public and has been used to help characterize the general contents that may be used in the
various kinds of ink. US. Patent 5,769,931 gives insight into the complexity of aqueous-
based inks that may be used in a disposable, modern pen. The components include 1)
polysaccharide gums; 2) viscosity modifiers; 3) polar, non-aqueous solvents such as
alcohols or glycols; 4) water; 5) sequestering agents which improve Stability; 6)
preservatives; 7) surfactants; 8) corrosion inhibitors; 9) dyes and/or pigments; and 10) pH
control agentsé. LDMS and MALDI MS have shown to be highly selective for dye
detection and the other ink components do not interfere with the direct analysis and

generally are not detected in the experiment.

Ink Dye Analysis
Selection of an Analytical Method

The greatest challenge in analyzing dyes is obtaining the structures. Some dye
structures are purposely withheld by the manufacturer, yet others are unknown even to
the producer. There is not one specific source that reports all the dye structures. Most of
the structures collected for this work were found using two resources: The Sigma-Aldrich
Handbook of Stains, Dyes and Indicators and the CApluS database which is accessible
through SciFinder Scholar 2001, an electronic resource provided through Michigan State
University. In general, the first step to obtaining dye structures is to establish the names
of the possible dyes used in the manufacture of ink through patents. All ink patents were
acquired from an online database (http://ep.espacenet.com) though the European Patent

Office which contains over 30 million documents worldwide. Next, the names are used

to search for the Structures. Currently, over 250 dyes (CI names or numbers) that are
possibly used in ink have been identified for this research, but only 207 structures were
actually found. This poses the difficulty of identifying the dyes contained in a pen ink
under investigation. According to the patents, the dyes (Specifically anionic, azo dyes)
used to manufacture ink appear to be acid dyes (dye classification). The majority of the
dyes (~2300) listed in the Colour Index are classified as acid dyes (consisting of mostly
azo, anthraquione, or triarylamethane dyes), and approximately 40% of them are
currently used in productions.

Many analytical techniques need to be compatible with the volatility and
solubility nature of the dyes. Generally, dyes are manufactured as sodium or chloride
salts allowing them to be rather soluble in water and ethanol, but prohibit them from
being volatile. Techniques such as gas chromatography/mass Spectrometry (GC-MS)
require volatile samples. Also, some dyes are insoluble in aqueous media or organic
solvents making extraction-based methods implausible. Currently, many ink analyses
require the dye to be soluble. Instrumental methods such as high performance liquid
chromatography that require liquid samples are incompatible if the dyes are not able to
dissolve. Many gel inks contain pigment dyes that are insoluble compounds. LDMS is
compatible with insoluble dyes since no extraction step is required, however, the
technique is not cable of detecting multiply-charged dyes. Many dyes that are insoluble
are neutral molecules. Dye detection by MALDI MS (incorporates a matrix into the
LDMS experiment) does not discriminate against dyes that have multiple charges. ESI
MS is another mass spectrometry technique capable of detecting molecules with multiple

charges, but recent work showed that the detecting polysulfonated dyes using

electrospray was hindered by a charge state of three or higher. LDMS is used to analyze
samples directly from the paper substrate (written entries on a questioned document). In
this research, MALDI MS is also used in the same fashion as LDMS; however, an
additive and matrix are applied to directly to the ink-paper sample. The additive and
matrix are prepared as solutions and the additive is added and dried first, followed by the
matrix. Whenever ink is not analyzed directly there is a chance that the dye could be
altered chemically. Direct analysis of ink is the safest means for examination, however,
in this research, a simple, chemical modification of the dye allows the species to be
detected as singly-charged directly from the paper substrate.

Commercial dyes are frequently sold as impure mixturesl. The production of
impure dyes is common since dye purification, subsequent to their manufacture, is
arduous and expensive. If impurities do not pose a threat to the application, including the
manufacture of ink, to which the dye will be used, then dye purification is not highly
demanded. Some applications such as the use of laser dyes require dyes to be highly
pure. In terms of ink examination, the analytical technique selected for examining dyes
needs to be capable of analyzing multi—component systems. Methods such as infrared
spectroscopy are typically not suitable for examining more than one constituent at a time.
LDMS and MALDI MS can easily detect multiple dyes simultaneously.

Methodsfor Ink Dye Analysis

Early experiments in ink analysis relied heavily on the use of chemical spot tests7,
ion migration testss, electrophoresisg’lo, and a variety of extraction-based methodsl H3.
Extraction-based methods are Still the most widely used todayu'”. The analysis of ink

can be considered as two categories — methods for ink comparison and methods for

relative ink dating. A variety of chromatographic methods have been used for ink
comparison including paper chromatography”, densitometric thin-layer chromatography
(TLC)13"15 ’17, gas chromatography (GC)13, and high performance liquid chromatography
(HPLC)‘9'2'. Additional methods include fourier transform infrared spectroscopy
(FI‘IR)22'23, nlicrospectrophotometry24‘25, and GC/mass Spectrometry (GC-MS)26. Most
of the recent literature has focused on ink comparisons, which included the use of
capillary electrOphoresiS (CE) combined with UV/V is and laser induced fluorescence
(LIF) detection and particle induced x-ray emission (PIXE)27’28, field desorption mass
Spectrometry (FDMS)29, and electrospray ionization mass Spectrometry (ESI MS)”.
High performance liquid chromatography (HPLC)31 and laser desorption mass
spectrometry (LDMS)32 have been used to examine the dye degradation for the
possibility of determining the relative age of a document. The current (1999 — 2002)
research in ballpoint pen ink analysis has gradually moved toward methods that
encompass mass spectrometry and chromatography which will be evaluated and their

utility demonstrated.

Mass Spectrometry

Mass spectrometry (MS) is a highly sensitive and powerful analytical technique
that provides elemental and molecular level information such as molecular weight and in
some cases, elemental composition that can be used for identifying and characterizing an
extensive range of species with molecular weights exceeding 100,000 Daltons (Da).

Mass Spectrometers separate and allow for the detection of a species (in the form of gas-

phase ions) based on the corresponding mass-to-charge (m/z) ratio. The relative
abundance of the ions is plotted versus the m/z values yielding a mass spectrum.

Mass spectrometry is practical for a wide range of applications. There are a
several mass Spectrometry instruments that can be utilized, and the selection of the
instrument greatly depends on the analyte being examined. The two most important
parts of any MS instrument consist of the ionization source and mass analyzer. The basic

design of an MS instrument is shown in Figure 1.1.

 

 

 

 

,~;“‘“‘“"“~“‘"“““‘7 ,s.’ . “ . ~..-r' .4..- ~ ' , :7 gr: - 7 .4 ; ~. '9"
AM . (40— ,4-‘1- A . ‘

Sample l::> Ionization C> Mass Ci) l:>
Introduction Source Analyzer Detector Computer
Figure 1.1: Schematic of a basic mass spectrometer

The MS experiment requires a sample to be in the gas phase; however, samples
can be initially delivered or introduced to the instrument in a solid or liquid form. Ions
are generated when the ionization source converts the intact species. The mass
analyzer separates the ions based on their m/z values. The ions are collected and
counted by the detector. Ultimately, a mass spectrum is produced as the computer
records the data.

Mass spectrometry has been one of the analytical tools applied to the analysis of
inks. Early ink studies used gas chromatography/mass Spectrometry (GC-MS)26 to
examine the volatile components. Initially, mass spectrometry methods, including GC-

MS which utilizes the electron impact (E1) as the ionization method, were limited to the

analysis of volatile (high vapor pressure) compounds. In order to ionize molecules by El,
samples are required to be in the gas phase. The challenging issue of sample volatility
and thermal stability was overcome by the development of desorption/ionization (D/I)
techniques which yield gas phase ions from condensed phase analytes. For many years,
thermally labile/nonvolatile components were not amenable to analysis by MS, although
the possibility was often considered. Mass spectrometric D/I methods encompass field
desorption (FD), laser desorption (LD), matrix-assisted laser desorption/ionization
(MALDI), secondary-ion mass spectrometry (SIMS), and fast atom bombardment (FAB).

The early methods of field desorption and laser desorption had limited utility. FD
requires the analyte on the sample probe to be heated with a high electric field (107 — 108
V/cm) thus sample degradation is possible”. The early use of LD was limited to
compounds with molecular weights of approximately 1000 Daltons and below“. The
basic idea was to focus a pulsed laser such as a C02 laser onto a solid target. High laser
power results in rapid heating and a portion of the sample can be heated to several
hundred degrees in a few nanoseconds. At this point, desorption can occur at rates faster
than those for chemical degradation (although some fragmentation has been observed)
resulting in desorption of intact molecules. A portion of the sample is not only desorbed,
but also ionized. Both FD and LD have been used recently for ink dye analyse529‘32‘35'37.
Both methods will be discussed further and compared.

The relatively new method, MALDI, is a D/I method that is related to laser
desorption. MALDI is an LD experiment that incorporates the use of a solid matrix for
the detection of the analyte. MALDI is an experiment in which an analyte such as a

peptide is mixed with a large excess of matrix molecules. These matrix molecules are

small, aromatic, organic molecules that have been selected because they efficiently
absorb UV light at 337 nm. Analytes have molecular weights that exceed 100,000
Daltons. From an analyte/matrix solution, crystals are grown. When irradiated with light
from a pulsed UV laser, matrix and analyte ions are formed. The desorption/ionization
method requires a mass analyzer such as time-of-flight (T OF) that is capable of operating
with a pulsed ionization source. Significant commercial developments have occurred in
the past few years to make TOF MS one of the premiere MS methods.

Both the LDMS and MALDI MS experiments for this research were carried out
with an instrument that was designed for the MALDI experiment. In the current design
of modern MALDI TOF MS instruments, 3 100 to 400 well plate is used for sample
introduction. The computer controlled sample stage can be moved such that the laser
irradiates a selected sample. Thus, MALDI instruments allow for introduction of planar
targets with computer control of the area being sampled by the UV laser. For the forensic
applications here, the target will be, in most cases, a paper sample with dye on it (a
written or printed line). For the detection of multiply-charged dyes, an additive and
matrix were introduced to samples. The introduction of the additive and matrix to
samples directly on paper was employed to limit the number of analysis steps, prevent
sample loss, and reduce the risk of sample contamination. Commercial dyes were also
analyzed directly from paper as well as from a stainless steel plate to confirm the
identification of the dyes.

A PerSeptive Biosystems Voyager DE (delayed extraction) TOF mass
Spectrometer (Farrningham, MA) was used for the analysis of dyes and is Shown in

Figure 1.2”. The instrument utilizes a pulsed nitrogen laser (337 nm, 3 nanosecond

10

pulses at 3 Hz) and a linear time-of-flight mass spectrometer. The user-selected
parameters for the LDMS and MALDI MS experiments include an accelerating voltage
of 20,000 V for detection of positive ions and -15,000 V for the detection of negative
ions, an intermediate source grid voltage that is 94% of the accelerating voltage, a guide
wire voltage that is opposite in bias and 0.05% in magnitude of the accelerating voltage,

and extraction delay times ranged from 100 us to 200 ns.

 

   
  
    
 

 

 

 

 

 

 

 

 

Linear
detector
- _ _____
Ion path
...... >
Laser path

Beam

guide

wire
Laser Video
Laser _ '— Aperture (grounded)

attenuator .
. Ground grid
P . V
rlsm ‘ .
Variable-voltage ‘ ‘ - ~_..|.,,.,., Sample
grld i loading
. Chamber
Sample plate Mam
source chamber

 

Figure 1.2: Diagram of theVoyager-DE mass spectrometer38

11

 

FAB and SIMS are similar techniques in that they both use particle bombardment
to induce the desorption/ionization of a condensed phase sample. SIMS uses an ion beam
consisting of Cs“, Xe+, or Ar+ where FAB makes use of fast Xe atoms”. Like MALDI, a
FAB experiment utilizes a matrix, usually glycerol, a viscous liquid. Both of these D/I
methods have been used to analyze nonvolatile, organic dyes3941 and in 1994, the
analysis of inks, dyes, and pigments on paper by TOF—SIMS was employed“.

Electrospray ionization (ESI) MS is not a D/I technique, but is also capable of
detecting non-volatile compounds. ESI was developed about the same time as MALDI
for the analysis of high molecular weight, biomolecules, but has also been employed for
the detection and characterization of species including dyes”. ESI is an attractive
method over D/I techniques since ions can be formed and detected as multiply—charged
Species. Analytes are injected as liquids and pumped continuously through a needle with
a flow rate of roughly 1-10 [IL/min”. Droplets are sprayed from the end of the needle.
Using a combination of a drying gas and an applied potential field (3-6 kV) causes the
droplets to desolvate and become highly-charged. The coulombic explosion (charge
repulsion upon evaporation of the droplets cause them to explode to form smaller
droplets) results in the formation of ions that vary in m/z for the same Species”. The
number of charges that a species acquires is analyte dependent. Analytes that have
several ionization sites will be highly-charged. ESI is also capable of analyzing species
that lack sites of ionization, but usually required an acidic or basic additive in order for

the analyte to be detected. The schematic (Figure 1.3) below depicts the E81 apparatus”.

12

N2 (80°C)
+3 to 6 kV, ions + I

   
   

 

 

 

 

 

 

 

 

 

ll NOZZLE
Ie /
SKIMMER ANALYZER
ELECTROSPRAY
"J
—=
I F
METALLICCAPILLARY
/' '1 Hi
LENS PUMP
N2(80°C)

Figure 1.3: Schematic of an electrospray ionization source43

Recently (2002), E81 MS has been employed by N g et al. for the characterization
of ballpoint pen ink dyes”. However, the analysis of dyes by ESI has been used Since the
late 1990’s. Most of the investigative work has focused on sulfonated azo dyes and has
been carried out by Sullivan et al.44’45. Some of the dyes examined in this research were
the same dyes that were characterized by Ng et al., so their analyses will be discussed
further and compared to LDMS and MALDI MS.

Most of the mass spectrometry methods that have been used to analyzed dyes
have encompassed D/I methods, however, the limitation of these MS techniques is that
multiply-charged dyes are not detectable directly. This fact has made ESI an attractive
MS alternative for dye detection. FDMS and LDMS are the most recent D/I techniques

that have been utilized for characterizing ink dyes. Recently, the examination of ink dyes

13

has been carried out using ESI MS. The use FDMS, LDMS, and E81 MS for dye analysis
will be discussed briefly and compared against one another. In Chapter Two, extensive
details on dye characterization will be given for LDMS Since the technique is a precursor
of MALDI MS which was the mass spectrometric method employed for this thesis
research. The basic LDMS characterization methods can be applied to MALDI MS.
Discussion on the previous use of additives in MALDI MS for enhancing the detection of
multiply-charged species will follow in Chapter Four as well as the ink dye analyses
carried out for this thesis research.
Field Desorption Mass Spectrometry

Field desorption mass spectrometry was used by Sakayanagi et al. to identify
eight basic dyes in black, blue, and red ballpoint pens inks”. The dyes are cationic,
possess a single positive charge, hence, the molecular ion was detected in positive FDMS
mode. Once the dyes were determined, various ballpoint pen inks were compared in
order to differentiate between manufactures. The m/z ratios obtained from the ballpoint
pen inks used in the analysis were summarized. Some of the peaks could not be
identified. Basic Violet 3 (Crystal Violet) and Basic Violet 1 (Methyl Violet 2B) are
commonly found in both blue and black ballpoint pens inks. Generally, the mass spectra
of blue inks contain a peak at m/z 470, which corresponds to Basic Blue 26 (Victoria
Blue B). All of the red ballpoint pen inks tested gave a peak at m/z 443, which
corresponds to either Basic Red 1 (Rhodamine 6G) or Basic Violet (Rhodamine B) since
both of these dyes have the same elemental composition. Sakayanagi et al. Showed that
some of the ballpoint pen inks appear to be identical”. In order for the inks to be

analyzed, the samples are required to be extracted prior to using FDMS. This may be a

14

problem Since aged inks may be hard to extract and pigment dyes are generally insoluble
which eliminates inks containing insoluble pigment dyes such as those contained in gel
pens using FDMS. Another disadvantage is that FDMS is a destructive method; hence,
the samples are destroyed during the analysis. The dyes are extracted from paper
samples and deposited on the FD emitter which is used to desorb and ionize the
compounds using a high electric field. The heating of the emitter possesses a threat for
sample decomposition46 and dyes may degrade. Positive-ion mode was only reported
which does not favor several ink dyes that are anionic salts, however, FD is capable of
operating in negative-ion mode. LDMS and MALDI MS are also capable of operating in
both positive and negative-ion modes. The solubility of the dyes limits the utility of
FDMS where LDMS is capable of analyzing colorants that are not soluble such as
Copper Phthalocyanine.
Electrospray Ionization Mass Spectrometry

Electrospray ionization mass spectrometry (ESI MS) was recently used by Ng et
al. for the characterization of both positively and negatively-charged dyes in ballpoint
pen inks”. ESI was developed about the same time as MALDI and is also capable of
analyzing nonvolatile, thermally labile molecules, specifically biomolecules. Generally,
ESI is capable of detecting species that are multiply-charged which can be an advantage
when analyzing dyes compared to LDMS. Shown in Figure 1.4a is a polysulfonated dye,
Acid Blue 9, which contains three sulfonate groups and one irninium group and was
detected as both a singly (m/z 769) and doubly-charged (m/z 373) Species with negative-
ion ESI MS. However, triply and higher-charged dyes were not observed or gave a weak

Signal in negative-ion mode. Ng et al. found that polysulfonated dyes including Acid

15

Blue 92 (Figure 1.4b) and Solvent Brown 20 (Figure 1.40) which contain three and four

sulfonate groups, respectively, could not be successfully detected using E8130. Detection
in ESI appears to be highly sensitive to the nature of the species. MALDI MS employed
for the detection of highly-charged dyes. Direct Red 80 has a charge state of -6 (contains

6 sulfonic acid groups) and was detected using MALDI MS in this research.

16

 

N cu,
so;
so;
so;
i‘_°”2
CH3
b) '03s
'03s
C) HO OH
80; so;
N:N N=N

'o s -
3 so3

Figure 1.4: Multiply-charged dye structures a) Acid Blue 9, b) Acid Blue 92, and c)
Solvent Brown 20

17

The overall charge State of the two dyes Solvent Orange 3 (Figure 1.5a) and
Solvent Yellow 19 (Figure 1.5b) is zero. Solvent Orange 3 is an organic monoazo dye
and Solvent Yellow 19 is a chromium (III) complex. In general, the ability to detect dyes
depends on their structure. With the exception of the overall charge State, the structures
of the two dyes are unrelated, but difficultly in detecting the dyes using ESI MS were
apparent. Solvent Orange 3 was protonated and gave rise to a weak signal in the
positive-ion ESI mass spectrum where Solvent Yellow 19 can not easily accept a proton
under the E81 experimental conditions and a signal was not observed. LDMS can easily
detect neutral molecules in both positive and negative-ion modes. In order to make a
direct MS comparison of the ability of LD and E81 to detect the two dyes, Solvent
Orange 3 and Solvent Yellow 19 should be analyzed using LDMS. However, these dyes
were not readily available for purchasing, so Copper (H) Phthalocyanine and Pigment
Red 3 were analyzed to make an indirect comparison of the ability of LDMS to detect
neutral dyes. Copper (11) Phthalocyanine (Figure 1.5c) is a neutral metal complex dye
that is detected without difficulty in both LD modes. Neutral aromatic compounds are
detected with ease in LDMS. Pigment Red 3 (Figure 1.5d) is a monoazo dye that gives
rise to peaks with m/z values of 308 and 306 in the positive and negative ion LDMS
Spectra, respectively, as shown in Figures 1.6a and 1.6b. Protonation in LDMS should be

structurally favored for Solvent Orange 3 with respect to the amine groups.

18

 

Figure 1.5: Neutral dye Structures a) Solvent Yellow 19, b) Solvent Orange 3, c) Copper
Phthalocyanine, and CI) Pigment Red 3

 

 

 

 

 

 

 

 

 

 

a) 308
E
W
9‘.»
s
.‘z’
I"!
0
C
. l I . - ii 11
o 100 200 300 400
rnlz
b) 306
E
O
E»
s
S
E
O
C
_ I T r A
o 100 200 300 400
mlz

Figure 1.6: LD mass spectra of Pigment Red 3 on paper a) positive-ion and‘b) negative-
ion

20

In order for ink dyes to be analyzed by ESI MS, the written entries are required to
be extracted from the paper (samples in the liquid form is a prerequisite for injection).
The extractions can usually be directly delivered to the source without additional sample
preparation. The necessity to extract the ink dyes may pose a problem if the dyes are
insoluble. Also, paper contains several components such as whiting agents that may
extract into the solvent and can be detected using ESI. Dyes may not be the only
components that get extracted from the ink or paper. N g et al. reported the analysis of
the same ink from six different brands of white paper”. The difference in paper
components did not have a significant effect, but the observed peaks due to the paper
were not identified nor were any m/z values reported. Other constituents, Specifically a
corrosion inhibitor and an antioxidant, in the ink were identified from the negative ion
ESI mass spectrum. In general, LDMS is highly selective for dyes. Mass spectral peaks
due to paper are not observed when analyzing the ink directly from paper since the spot
from the laser is focused within the width of a pen stroke. The size of the laser spot and
width of a pen stroke is discussed later. The detection of non-dye ink components are
seldom observed in the LDMS experiment since the dyes are the only species that
efficiently absorb the lasers radiation. However, polyethylene glycol (PEG) is a polymer
used in inks as a humectant and has been observed in the mass Spectra of gel and liquid
inks. Typically, neat polyethylene glycol is not detected in LDMS. The polymer is not
capable of being desorbed/ionized without the assistance of another compound such as a
matrix as described in matrix-assisted laser desorption/ionization (MALDI). The gel and
liquid ink dyes may serve a similar role as a matrix, which aids in the

desorption/ionization of an UV non-absorbing species, allowing the polymer to be

21

detected. The paper may also influence the detection of the polymer. Neat polyethylene
glycol with an average molecular weight of 300 was analyzed directly from paper.
Surprisingly, the polymer was detected in LDMS directly from the paper substrate and
the positive-ion mass spectrum is Shown Figure 1.7a. The region of the mass Spectrum
that contained representative polymer peaks was enhanced as shown in Figure 1.7b.
Figure 1.8a illustrates the m/z values that represent the ionization of PEG by the
formation of sodium and potassium adducts. The polymer is a mixture of PEG oligomers
that differ in the number of repeating units, 11. The repeating unit of PEG is —OCH2CH2—
as seen in Figure 1.8b and has a molecular mass of 44 Da. The distribution of the peaks
observed in the mass spectrum is typical of a synthetic polymer. Mass spectral peaks that
correspond to PEG are separated by 44 amu. The tetramer of ethylene glycol that is
observed in the mass spectrum is the smallest of the PEG polymers and has a molecular
weight of 194 Da. The tetramer gives rise to two peaks in the positive ion mass spectrum
at m/z values 217 and 233 that correspond to sodium ion (Na+) and potassium ion (K+)
adducts, respectively. Paper contains several components and some brighteners used in
paper are actually dyes. In fact, some paper contains Copper Phthalocyanine which is a
colorant associated with blue gel ink pens. Figure 1.9 demonstrates the detection of PEG
in black, liquid ink from a Pilot® V-Ball rollerball pen. Either the paper or the ink
components are assisting in the detection of the polymer. Neat PEG is not detected when
analyzed directly from the standard, stainless steel sample plate. The DH of the polymer
is not by LDMS alone since PEG requires the assistance of another species in order to be

desorbed/ionized.

22

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a)
E'
a
.8
E
3
E
O
a:
L l .1 I. is i L. I I I
0 100 200 300 400 500 600
mlz
b)
305 349
E: 393
g 321 365
f 261
32 409 437
g 277
453 481
497 525
ll . I PS 569
250 350 450 550
mlz

Figure 1.7: Positive-ion LD mass spectra of polyethylene glycol (PEG) on paper a)
complete spectrum from m/z 0-600 and b) expanded region from m/z 250-600

23

 

 

 

 

 

 

 

 

 

 

a) n M (M + Na+) (M + Kt)
4 194 217 233
5 238 261 277
6 282 305 321
7 326 349 365
8 370 393 409
9 41 4 437 453
10 458 481 497
11 502 525 541
12 546 569 585

 

 

 

 

 

 

b) H-(-O—CH2—CH2-);OH

Figure 1.8: Polyethylene glycol a) m/z values that correspond to the ionization of the
polymer (M) by sodium (M + Na”) and potassium ion (M + K+) adduction and b)
polymer structure (11 is the number of repeating units)

24

Relative Intensity

b)

Relative Intensity

 

23

 

 

 

 

PEG

300 400 500 600

 

700

 

 

 

 

 

 

 

 

 

 

0 100 200
mlz
393 437
481
349 525
305 569
I
261 '
I
I . l -l ..L.. .u...... “Jim-n.1,; I ‘1‘.
250 350 450 550
mlz

Figure 1.9: Positive-ion LD mass Spectra of black, liquid ink from a Pilot® V-Ball
rollerball pen a) complete spectrum from m/z 0-700 and b) expanded region from mlz

250-600

25

A major drawback to using ESI for the detection of dyes is sample carryover.
Samples that are analyzed by ESI leave residue on the walls of the capillary (sample
introduced through a capillary) which lead to the contamination of following samples.
Carryover is inevitable, but washing of the capillary between samples helps to alleviate
the problem. Sample contamination is not a concern in LD since the ink is analyzed
directly from the paper. Sample preparation is minimal as described in the experimental
section. Extractions can be rather arduous and dilute the sample. Ng et al. found that
five times more sample was needed to be extracted for dyes for detection in negative-ion
versus positive-ion modes”.

Laser Desorption Mass Spectrometry

This project began with a demonstration that UV laser desorption mass (LDMS)
spectrometry has the sensitivity necessary to analyze dyes that are present on paper in a
written pen line. Demonstrations of impressive sensitivity were made when LD was
being developed two decades ago, but at that time IR lasers such as the C02 laser were
frequently used, and sensitivity is extremely dependent on the ability of the analyte to
efficiently absorb the light. Consider a typical ballpoint pen cartridge that contains about
0.6 grams of ink“. Written pen lines are typically 0.3-0.4 mm wide. If the ink is 20%
dye by weight, and the typical molecular weight of a dye molecule is 400 grams/mole,
the ink cartridge would contain approximately 0.3 millimoles of dye. If the ink in a
cartridge can write a line approximately 3000 meters long, then there is roughly 0.03
millimoles per square meter, as a molar surface coverage per unit area. In the instrument
that will be used, the laser spot size is approximately 0.03 cmz. In this area, there is

roughly 10'10 moles — 0.1 nanomoles of dye. With such a small amount of material being

26

irradiated by the laser, detectable signals for mass spectrometry will only be generated if
there is efficient coupling of the energy available at the surface to efficiently desorb and
ionize analyte dye molecules.

Even if such small amounts can be studied, the dye needs to be analyzed from
paper which may be thought of as potential problem for the DH of the analyte. The metal
suface is an important parameter in the D/I experiment. However, analyses directly for
paper have not been problematic. In time-of-flight mass spectrometry, ions are
accelerated by a well-defined electric field to a specific kinetic energy. In this particular
instrument, the ions are accelerated to kinetic energies of 15,000-20,000 electron volts.
The time that it takes for these ions to move through a distance of 1 meter (distance of the
flight tube) is very accurately measured, and from this, ionic mlz values are determined.
In the MALDI experiment, analytes are introduced on a metallic surface on which a
potential is applied once the laser has fired. In this proposed research, ions desorb from a
non-conducting surface, paper. However, severe deviations in the measured ions’ time-
of-flight have not been observed from direct paper analyses. UV LDMS is sufficiently
sensitive to generate ions from ink on paper since the time-of-flight mass spectrometer

adequately resolves/separates the ions.

Chromatography

In general, mechanistic concept of chromatography based methods is that
components in a mixture are separated while equilibrating between a stationary phase
(solid material) and a mobile phase (gas or liquid material). Simple and inexpensive

solid-liquid separation methods such as paper chromatography and thin-layer

27

chromatography (TLC) have been applied to the separation of dyes as early as 1950’s“.
A typical TLC experiment consists of a developing tank with a cover, an applicator, a
solvent system, and a TLC plate. More advanced methods for separation of ink dyes
have been recently employed such as high performance liquid chromatography (HPLC).
The appearance of the resulting chromatogram is Similar to an LD mass Spectrum. The
analysis of Crystal Violet and corresponding dye homologues have been analyzed by
both methods. TLC was sufficient for the analyses carried out in this research. Pertinent
information including the number of dyes and the color of the dyes present in each ink
was quickly acquired by TLC. Frequently, TLC was indirectly coupled to MS to
associate molecular mass with dye color. This involves extracting dyes from the TLC
bands (extractions of the individual/separated ink dyes) with subsequent LDMS or
MALDI MS analysis. Instruments capable of performing HPLC-MS are commercially
available, but this coupling method was not used in this research.
Thin-Layer Chromtgraphy

Thin-layer chromatography (TLC) is a simple, planer chromatography method
that utilizes two phases (1) a liquid, mobile phase (typically, more than one organic
solvent is used) and (2) a solid, stationary (such as porous Silica) phase for performing
separations. The molecules separate based on their interactions with (their affinity for)
the stationary and mobile phases. Glass TLC plates coated with silica are inexpensive
and are the most commonly used plates among all types of separation analyses.

TLC has been previously applied to the separation of ink dyes. Several solvent
systems (comprises the mobile phase) have been examined to improve the quality of the

separations and have been determined Specifically for a particular pen type. The ink

28

dyes analyzed here were separated using a three—part mobile phase consisting of 70:35:30
(v:v:v) ethyl acetatezethanolzwater. TLC has several advantages that make the technique
valuable for dye separations. The ink is placed directly on the TLC plate and once the
sample spot dries the plate is placed inside the TLC chamber in which the solvent system
was given time to equilibrate between the liquid and gas phases. The direct application
of the ink does not appear to hinder the separation of the dyes from the ink. The
separation of the ink dyes are compared to a Standard for comparison. A solution
containing the dyes Crystal Violet, Methyl Violet 2B, and tetramethylpararosaniline was
used as the standard (8) and their corresponding TLC bands are labeled 8 1, S2, and S3.
The ink dyes were designated using the letter “D” and are consecutively numbered
beginning from the bottom of the TLC chromatogram.
High Performance Liquid Chromatography

The examination of blue ballpoint pen inks stored under various light conditions
was carried out by high performance liquid chromatography by Andrasko et al.3 1.
Various blue inks were exposed to daylight, extracted from the document, and analyzed
by HPLC. In the fresh ink sample, basic dyes, Crystal Violet (CV), Methyl Violet 2B
(MV), and tetramethylpararosaniline (TPR), were detected. The only difference between
these dye homologues is the number of methyl groups bonded to the three nitrogen
atoms. Andrasko and co-workers reported that as Crystal Violet is exposed to light, the
dye decomposes to form CV homologues by the gradual loss of methyl groups“. The
decomposition products elute before Crystal Violet since the polarity increases with

increasing substitution of methyl groups by hydrogen. Andrasko et al. demonstrated that

29

Crystal Violet decomposed after three and six days exposed to light3 1. This experiment

demonstrated that dyes and decomposition products can be detected using HPLC.

LDMS and Forensic Science Applications
Questioned Document Examination

A questioned document is any material object that contains suspicious markings”.
More Specifically, questioned document analysis encompasses many types of evidence
including typewriting, laser printing, photocopies, facsimiles, forgeries and frauds,
indented and obliterated writing. Most markings are in handwritten or printed form and
the type of analysis which examiners are called upon the most often is handwriting
analysis. Traditionally, this type of work has involved determining the author or
authenticity of handwriting by careful comparison of known and unknown handwriting
samples. More recently, other evidence besides the characteristic nature of the
handwriting have become the focus of questioned documents involving written entries on
paper using an ink pen. Expertise continues to be developed in diverse fields including
the chemical characterization of components of the document (paper, ink, watermarks).

The use of laser desorption mass Spectrometry (LDMS) as a tool for the analysis
of pen ink dyes on paper has been established. Previous work indicated that the
technique has great potential in detecting singly-charged or neutral dye molecules
directly on paper as well as other substrates. Direct LDMS ink dye analysis provides
basic molecular information (molecular mass, composition, and charge state) that can be
used for characterizing dyes; however, the differentiation of isomeric dyes requires more

advanced characterization methods. The use of dye photodegradation by incandescent

30

light and UV radiation with subsequent LDMS analysis proved to be successful. The
thorough study of ink dyes using LDMS, TLC, and patent searches lead to the conclusion
that LDMS is restricted to singly-charged or neutral dye molecules. Alternative detection
methods for analyzing multiply-charged ink dyes (from stainless steel and paper
substrates) by LDMS were investigated which is the focus of this thesis research. The
use of cation—exchange resins was the first analytical method employed for charge state
reduction. The method was simplified by the direct addition of a cation-exchange reagent
to the sample. The use of several matrices and cation-exchange additives were explored.
A general method that encompasses one matrix and one additive for the analysis of all
dyes would be ideal. MALDI MS (with an additive) can be used to characterize ink dyes
which can be used as a tool to distinguish between two inks that are macroscopically
Similar (in terms of the same ink color and same kind of pen ink). Assaying inks via
coupling thin-layer chromatography with MALDI MS was used to simplify dye
identification. The prior use of coupling TLC with LDMS proved to be an excellent
method for determining the number of dyes present in one ink and associating dye color
(color obtained from thin-layer chromatogram) with molecular mass (mass obtained from
the LD mass Spectrum) to make dye identification easier.
Trace Evidence

The dye characterization method developed in this research is not limited to the
examination of questioned documents. The method can be applied to other applications
in forensic science that encompass detecting colorants. Such areas of examination

include trace evidence. Colored materials frequently analyzed as trace evidence include

31

automobile paint chips and natural and synthetic fibers. Colorants used in fibers have
been examined previously by LDMS by Balkoso.

Instrumental techniques such as microspectrophotometry and fourier transform
infrared (FT IR) spectroscopy are frequently used in forensic science to examine paint
chips and fibers, however, extracting colorant information from an FTIR spectrum is
cumbersome. Knowledge of analyzing these kinds of trace evidence was sustained
through course work. Paint chips are analyzed as individual layers as well as cross-
sections. FT IR analysis of a complex (more than one component) sample results in a
fingerprint Spectrum and the possibility of using the data to characterize and identify a
Specific colorant is unlikely. The main peaks that appear in FTIR spectra obtained from
analyzing paint chips are due to the resins or binders in the paint and their corresponding
peaks generally overlap and dominant peaks that represent organic pigments, thus,
making colorant identification unattainable. Fibers are analyzed by FTIR as well, but the
technique is used to determine the type of fiber polymer, not to acquire information on
the colorants. Microscopy is the most common method used for analyzing fiber color as
well as other physical characteristics of the fiber and microspectrophotometry is an
instrumental technique frequently employed strictly to examine the color of the fiber.
The technique is also used as a means to compare the color of paint chips. Analysis by
microspectrophotometry is problematic when dealing with fibers that contain more than
one dye, comparing two fibers that are inherently the same color, and fibers that are
colored black. The visible Spectra that are acquired from analyzing fibers and paint chips
are fringerprints and can not be use to identify a specific colorant. A visible spectrum

can be thought of as a color spectrum. Two separate fibers that are microscopically

32

similar in color and give rise to similar spectra can be misleading if the fibers in fact
contain different dyes or pigments. Microspectrophotometry is not a selective technique
for distinguishing dyes or pigments of the same color. Fibers containing more than one
colorant may yield spectra that have overlapping spectral features which do not allow for
identifying individual colorants and may be similar to the problem with analyzing black
fibers using microspectrophotometry. Black fibers are difficult to analyze using
microspectrophotometry since the resulting Spectrum will be one broad peak that spans
the entire range of the visible wavelengths; hence, no distinctions can be made between
unknown and known fiber samples that are colored black.

Difficulties arose when Balko tried to detect multiply-charged dyes, Specifically,
reactive dyes that contained multiple sulfonic acid groups, from the fibers using LDMS
directly”. Balko stated that the dyes (Reactive Red 11, Reactive Blue 4, and Reactive
Yellow 1) were not able to be detected Since they are covalently bonded to the molecules
that comprise the cotton fabric”, however, direct LDMS analysis is not capable of
detecting multiply-charged species. Most likely the multiple charge distribution
prevented the dyes from being detected rather than the interactions between the dye and
the fibers. MALDI MS can be applied to the analysis of fiber dyes to ascertain if

multiply-charged dyes can be detected from cotton material or other types of fibers.

33

Research Overview

This thesis research has focused on methods for detecting and characterizing
multiply-charged ink dyes as a forensic tool for analyzing a questioned document. The
following determinations can be drawn when examining the characteristics of ink dyes:

1. Inks on the same or different documents can be differentiated.

2. An ink source of a specific entry can be identified or eliminated from an

exemplar.

Advancements in ink dye detection and characterization methods were carried out
using mass spectrometry. This program of research provides new data to demonstrate the
above deductions. The major approach taken in this project:

1. A novel method of analysis was employed to help detect, identify, and
characterize multiply-charged dyes present in writing inks based on chemical
and structural information.

Specifically, this research thesis discusses:

1. The use of laser desorption mass spectrometry to demonstrate the molecular
information that can be attained for the characterization of singly-charged and
neutral dye molecules which parallels dye characterization in the MALDI MS
analysis.

2. The use of laser desorption mass spectrometry and thin-layer chromatography
for determining the presence of multiply-charged dye in an ink.

3. The use of matrix-assisted laser desorption/ionization mass spectrometry with
a cation-exchange additive for the analysis of multiply-charged dyes (from

inks) on paper directly

34

4. The basic MS characterization methods that can be employed when

examining the MALDI and LD mass Spectra of polysulfonated ink dyes.

35

References

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

Venkataraman K, editor. The Analytical Chemistry of Synthetic Dyes. John
Wiley & Sons, New York, 1977

Venkataraman K. The Chemistry of Synthetic Dyes. Vol. 1, Academic Press Inc.,
New York, 1952

Kiernan JA. Classification and Naming of Dyes, Stains and Fluorochromes.
Biotechnic & Histachemistry. 2001 (76) 261-277

Llewellyn B. Stains File: Structure and Colour in Dyes. Nov 2002
<http://members.pgonline.com/~bryand/StainsFile/dyeS/dyecolor.htm>

Zee, FP van der. “Anaerobic Azo Dye Reduction.” Ph.D. Dissertation.
Wageningen University, 2002

Wang A, Chadwick BW, inventors. BIC Corporation, assignee. Ink Composition.
US Patent 5,769,931 Jun 1996

Crown DA, Conway JA, Kirk PL. Differentiation of Blue Ballpoint Pen Inks
Using Chemical Spot Test. J Crim Law Criminal Police Sci. 1961 (52) 338-343

Mezger O, Rall M, Heess W. New Methods for Examining Inks. Z. Angew. Chem.
1931 (44) 645—651.

Brown C, Kirk PL. Comparison of Writing Inks Using Electrophoresis. J Crim
Law Criminal Police Sci. 1954 (45) 334-339

MacDonnell HL. Characterization of Fountain Pen Inks by Porous Glass
Chromatography and Electrophoresis. J Crim Law Criminal Police Sci. 1962 (53)
507

Brunelle RL, Lee H. Determining the Relative Age of Ballpoint Ink Using a
Single-Solvent Extraction, Mass-Independent Approach. J of Forensic Sci. 1989
(34) 1166-1182

Brunelle RL. A Sequential Multiple Approach to Determining the Relative Age of
Writing Inks. Int J Forensic Doc Examiners. 1995 (1) 94-98

Cantu AA, Prough RS. On the Relative Aging of Ink — The Solvent Extraction
Technique. J Forensic Sci. 1987 (32) 1151-1174

36

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

Somerford AW, Souder JL. Examination of Fluid Writing Ink by Paper
Chromatography. J Crim law. 1952 (43) 124-127

Aginsky VN. Determination of the age of ballpoint pen ink by gas and
densitometric thin-layer chromatography. J Chromatogr. 1994 (678) 119-125

Aginsky VN. Comparative Examination of Inks by Using Instrumental Thin-
Layer Chromatography and Microspectrophotometry. J F arensic Sci. 1993 (38)
1 l 1 1-1 130

Brunelle RL, Negri JF, Cantu AA, Lyter AH. Comparison of Typewriter Ribbon
Inks by Thin-Layer Chromatography. J Forensic Sci. 1977 (22) 807-814

Stewart JL. Ballpoint Ink Age Determination by Volatile Component Comparison
— A Preliminary Study. J Forensic Sci. 1985 30 405-411

Colwell LF, Karger BL. Ball Point Pen Examination by High Pressure Liquid
Chromatography. J Assa 017 Anal Chemists. 1977 (60) 613-618

Griffin RME, Kee TG, Adams R. High Performance Liquid Chromagraphic
System for the Separation of Basic Dyes. J Chromatogr. 1988 (445) 441-448

Lyter AH. Examination of Ball Pen Ink by High Pressure Liquid
Chromatography. J Forensic Sci. 1982 (27) 154-160

Becker JL, Brunelle RL. Determination of Age of Ballpoint Pen Inks by Fourier
Transform Infrared Spectroscopy. Program 36th Annual Meeting of the American
Academy of Forensic Sciences, Anaheim, CA, Feb 1984

Humecki H. Experiments in Ballpoint Ink Aging Using Infrared Spectroscopy.
Proceedings of International Symposium of Non-Handwriting Aspects of
Questioned Document Examination, US Government Printing Office,
Washington, DC, 1985 131-135

Zeichner A., et al. Transmission and Reflectance Microspectrophotometry of
Inks. J Forensic Sci. 1988 (33) 1171-1184

Aginsky, VN. A Microspectrophotometric Method for Dating Ballpoint Inks — A
Feasible Study. J Forensic Sci. 1995 (40) 475-478

Aginksy VN. Dating and Characterizing Writing, Stamp Pad, and Jet Printer Inks

by Gas Chromatography/Mass Spectrometry. Int J F arensic Doc Examiners.
1996 (2) 103-110

37

(27)

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

(37)

(38)

Vogt C, Vogt J, Becker H, Rohde E. Separation, Comparison, and Identification
of Fountain Pen Inks by Capillary Electrophoresis with UV-Visible and
Fluorescence Detection by Proton-Induced X-Ray Emission. J Chromatogr. 1997
(781) 391-405

Vogt C, Becker A, Vogt, J. Investigation of Ball Point Pen Inks by Capillary
Electrophoresis (CE) with UV/V is Absorbance and Laser Induced Fluorescence
Detection and Particle Induced X-Ray Emission (PIXE). J Forensic Sci. 1999
(44) 8 19-831

Sakayanagi M, Komuro J, Konda Y, Watanabe K, Harigaya Y. Analysis of
Ballpoint Pen Inks by Field Desorption Mass Spectrometry. J Forensic Sci. 1999
(44) 1204-1214

N g LK, Lafontaine P, Brazeau L. Ballpoint pen inks: Characterization by positive
and Negative Ion-Electrospray Ionization Mass Spectrometry for the Forensic
Examination of Writing Inks. J Forensic Sci. 2002 (47) 1238-1247

Andrasko J. High Pressure Liquid Chromatography Analysis of Ballpoint Pen
Inks Stored at Different Light Conditions. J Forensic Sci. 2001 (46) 21-30

Grim DM, Siegel JA, Allison J. Evaluation of Laser Desorption Mass
Spectrometry and UV Accelerated Aging of Dye on Paper as Tools for the
Evaluation of a Questioned Document. J F arensic Sci. 2002( 47) 1265-1273

Watson JT. Introduction to Mass Spectrometry. Lippencott-Raven Publishing,
Philadelphia, 1997

Denoyer E, Van Grienken R, Adams F, Natusch DFS. Laser Microprobe Mass
Spectrometry 1: Basic Principles and Performance Characteristics. Anal Chem.
1982 (54) 26A-41A

Grim DM, Siegel J A, Allison J. Evaluation of Desorption/Ionization Mass
Spectrometric Methods in the Forensic Applications of the Analysis of Inks on
Paper. J Forensic Sci. 2001 (46) 1411-1420

Grim DM, Siegel JA, Allison J. Does ink age inside of a pen cartridge? J Forensic
Sci. 2002 (47) 1294-1297

Dunn JD, Siegel JA, Allison J. Photodegradation and Laser Desorption Mass
Spectrometry for the Characterization of Dyes Used in Red Pen Inks. J Forensic
Sci. 2003 (48) 652-657

Voyager Workstation User’s Guide version 5, PerSpective Biosystems, 1999

38

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

(49)

(50)

Monaghan JJ, Barber M, Bordoli RS, Sedgewick RD, Tyler AN. Fast Atom
Bombardment Mass Spectra of Involatile Sulphonated Azo Dyestuffs. Org Mass
Spectrom. 1982 (17) 569-574

Scheifers SM, Verma S, Cooks RG. Characterization of Organic Dyes by
Secondary Ion Mass Spectrometry. Anal Chem. 1983 (55) 2260-2265

Detter-Hoskin LD, Busch KL. SIMS: Secondary Ion Mass Spectrometry. In:
Conners TE, Banerjee, editors. Surface Analysis of Paper. CRC Press, New York
1995 206-234

Pachuta SJ, Staral JS. Nondestructive Analysis of Colorants on Paper by Time-of-
Flight Secondary Ion Mass Spectrometry. Anal Chem. 1994 (66) 276-284

De Hoffinann E, Charette J, Stroobant V. Mass Spectrometry Principles and
Applications. John Wiley & Sons, Chichester, 1996

Sullivan AG, Gaskell SJ. The Analysis of Polysulfonated Azo Dyestuffs by
Matrix-Assisted Laser Desorption/Ionization and Electrospray Mass
Spectrometry. Rapid Commun Mass Spectrom. 1997 (11) 803-809

Sullivan AG, Garner R, Gaskell SJ. Structural Analysis of Sulfonated Momoazo
Dyestuff Intermediates by Electrospray Tandem Mass Spectrometry and Matrix-
Assisted Laser Desorption/Ionization Post-Source Decay Mass Spectrometry.
Rapid Commun Mass Spectrom. 1998 (12) 1207-1215

Skoog DA, Holler FJ, Nieman TA. Principles of Instrumental Analysis. 5‘h ed.
Saunders College Publishing, Philadelphia, 1998

Kunjappu JT. Molecular Thinking in Ink Chemistry. Ink World 1999 (5) 50-53

Sherma J, Fried B, editors. Handbook of Thin-Layer Chromatography. 3rd ed.
Marcel Dekker, Inc., New York, 2003

Gialamas DM. Criminalistics. Encyclopedia of Forensic Sciences. Academic
Press, 2000 471-477

Balko LG. “Laser Desorption/Ionization Mass Spectrometric (LD/MS) Analysis

of Samples of Forensic Interest: Chemical Characterization of Colorant and
Dyes.” MS. Thesis. Michigan State University, 2002

39

Chapter 2: Direct LDMS and TLC Analysis of Ink Dyes

This chapter discusses the direct methods that were used to examine LD mass
Spectra for dye characterization. Direct characterization methods involve the
examination of an unmodified sample. Analyzing ink dyes by LDMS is accomplished
directly from written entries on the paper without any chemical or physical modifications
(including dye extractions) to the sample. In a typical GC/MS experiment, a sample
mixture is separated using chromatography prior to the analyte detection by mass
spectrometry, hence, each mass spectrum acquired from the analysis will contain peaks
(peaks due to the molecular and fragment ions) that are associated with one particular
analyte. Unlike a mass spectrum obtained from a GC/MS analysis, a mass spectrum
collected using LDMS contains peaks that are associated with multiple Species that are in
the sample mixture and many of the components in the mixture are not detected.
However, both MS experiments allow molecular information such as mass and
composition (elemental information maybe obtained from isotOpic distributions). Dyes
are typically manufactured as salts. A salt is an ionic compound that is comprised of a
cation and an anion, excluding the hydronium (H1), hydroxide (OH'), and oxide (02‘)
ions 1. Many colorants used in ink are dye salts and some are neutral dyes such as
pigments. The overall charge State varies among the ionic dyes and can be deduced from
the LD mass spectra to characterize and identify neutral or singly-charged colorants.
Both positive and negative-ion LD mass spectra were collected for each sample to
discern the charge state of dyes prior to the MS analysis. Determining the charge State of

the dye will be discussed through the LDMS examination of Copper Phthalocyanine,

40

Crystal Violet, and Metanil Yellow. Additional information that maybe used for
characterization is to examine the LD mass spectra for the presence of ions due to
impurities, clusters, and fragments of the analyte. The mass spectral peaks were divided
into the following two categories for purposes of characterization: (1) peaks (includes
isotopic peaks) that denote the detection of the intact analyte and (2) peaks that are
indicative of the analyte indirectly. Acquiring molecular weights, compositions, and
charge states can be deduced from examining the mass Spectral peaks that are associated
with the intact dye. Determining the molecular level information will be demonstrated
through the direct LDMS analysis of Copper Phthalocyanine (commonly used in blue gel
ink), Crystal Violet and Metanil Yellow (frequent dyes contained in black ballpoint ink).
Indirectly characterizing ink dyes indirectly based on impurities, cluster ions, and
fragment ions which give rise to peaks in the LD mass spectra will be elucidated by

examining the dyes mentioned above.

Experimental
Laser Desorption Mass Spectrometry

LDMS analyses were carried out using a PerSeptive Biosystems Voyager DE
(Farmingham, MA) mass spectrometer. The instrument utilizes a pulsed nitrogen laser
(337 nm, 3 nanosecond pulses at 3 Hz) and a linear time-of-flight mass spectrometer.
The user-selected parameters for the LDMS experiments include an accelerating voltage
of 20,000 V for detection of positive ions and -15,000 V for the detection of negative
ions, an intermediate source grid voltage that is 94% of the accelerating voltage, a guide

wire voltage that is opposite in bias and 0.05 % in magnitude of the accelerating voltage,

41

and an extraction delay time of 100 ns. In some cases, the analysis on the paper substrate
required an increased delay time of 200 ns for optimum resolution. Blue ink from a gel
pen (Pentel® Hybrid® Gel Grip RXT) containing Copper Phthalocyanine was used as the
calibrant for both positive and negative-ion modes.

Single written lines on MSU letterhead paper were analyzed directly. The spot
from the nitrogen laser beam was focused on the line within the width of the pen stroke
which is approximately 0.3—0.4 mm. Eighty—five pen inks have been analyzed using
LDMS; however, the results presented here focus on a few of the analyses. All of the
analyses can be found on the world wide web at the following URL:
http://poohbah.cem.msu.edu/peninkslpens_main.htm

In order to simulate the analysis of ink on paper, 10,000 pmol/uL of known dye
solutions were prepared using 1:1 (v:v) methanolzwater, spotted (5 [.lL) on paper and
allowed to dry prior to the LDMS analysis. Squares approximately 2 x 2 mm2 were cut
from the paper and mounted on a disposable, gold MALDI plate using Scotch® double
coated tape (3M, St. Paul, MN).

Thin-Layer Chromatography

TLC was carried out using K5F silica gel 150 A TLC plates (Whatman, Ann
Arbor, MI) with dimensions of 5 x 10 cm and a stationary phase thickness of 250 um.
The solvent system used for the separation of the dyes in the ink consisted of 70:35:30
(v:v:v) ethyl acetatezethanolzwater. Methyl Violet 2B (Aldrich Chemical Co.,
Milwaukee, WI), purchased as an impure mixture of dyes, included Crystal Violet and
tetramethylpararosaniline and was used as the comparison standard. Dyes extracted from

TLC bands were analyzed by LDMS. Individual bands were scrapped from the TLC

42

plate, placed in a centrifudge vial, and extracted with 5-10 uL of 1:1 (v:v) ethanolzwater.
The vials were vortexed, and the Silica was separated from the dye solutions by
centrifugation. The dye extracts were dried on paper or the metal plate and subsequently

analyzed by LDMS.

LDMS Calibration and Copper Phthalocyanine
Mass Accuracy for Identification

Proper calibration (mass-to-charge ratio calibration) in laser desorption mass
spectrometry (time-of-flight) is required to ensure accurate mass measurements for
identification purposes. The mass accuracy for small mass molecules (< 1500 Da) in
conventional TOF instruments is reliable. Mass accuracy for a given ion is detemined by
taking the difference between the true mass and measured mass and dividing the
difference by the true mass. Typically, mass accuracy is expressed in parts per million
(ppm), so the value is multiplied by 106. Low resolution mass Spectrometry was used for
these dye analyses. Deviations in m/z values are small and were observed in the first
decimal place. Isotopically-resolved peaks are achieved in the TOP experiment for low
molecular weight molecules which allows accurate assignment of mass-to-charge (m/z)
values (the mlz values are presented as whole numbers).

The instrument should be calibrated fi'equently as the sampling position changes.
The sample plate including the paper substrate is not uniform, hence, the distance from
the sample to the detector changes as the sampling position moves across the horizontal
plane (the sample position varies with the x and y-coordinates). Even small differences

(irregularities of the paper’s thickness) in the length of the ions’ flight path influence the

43

accuracy of the masses of the ions detected; hence, the calibrant needs to be placed in
close proximity to the sample. Also, the calibrant and sample need to be on the same
type of substrate. For example, if the written lines on paper are being analyzed, then the
calibrant needs to be on paper as well. The thickness of the paper will cause the
calibration to be inaccurate.

In order for the mass spectrometer to be properly calibrated, the appropriate
calibration standard should be used. When selecting a calibrant, the molecular weight of
the standard compound needs to be considered. The mass of the analyte needs to be
within the mass range of the calibration. For example, if an unknown dye has a mass of
450 Da, then the calibrant should have a mass higher or near that value. Confirming that
a calibration is valid can be accomplished by analyzing samples with known masses.
Generally, analytical techniques require more than one standard for a calibration, but
using multiple calibrants is not necessary for LDMS. A one point calibration allowed for
sufficient mass accuracy for the analyses presented here Since the dyes have low
molecular weights. The calibrant used for the experiments for this thesis research is

Copper Phthalocyanine.

LDMS Ink Dye Analysis

UV laser desorption mass spectrometry (LDMS) has proven to be a powerful
technique for the detection of several ink dyes. The detection of Crystal Violet, an
organic dye frequently found in black ballpoint pen inks, using LDMS was first
demonstrated in 20012. The instrumental method is attractive Since a solvent extraction

step is not required; hence, the dyes are detected directly from written lines on paper.

Success has been reported with black, blue, and red ballpoint pen ink dyesz‘s. The direct
characterization information that can be extracted from the LDMS mass spectra will be
discussed next.
Molecular Information: Mass, Composition, and Charge State

Copper Phthalocyanine is a pigment dye that is frequently encountered in gel pens
with blue ink. Copper Phthalocyanine is a stable, neutral dye molecule (M), shown in
Figure 2.1a, and has a monoisotopic, molecular mass of 575 Daltons (exact molecular
mass of l2C321H16MN363CU is 575.0793] Daltons). The molecular masses that are
presented here will be rounded to the whole numbers for Simplicity since decimal places
can be neglected for these experiments. The pigment is easily detected in both positive
and negative-ion LDMS modes. The formation of the positive and negative ions of
Copper Phthalocyanine has been examined extensively using LDMS. Copper
Phthalocyanine is ionized by the addition or loss of an electron giving rise to molecular
ions in the negative and positive-ion LD mass spectra, respectively. Shown in Figure
2.2a and 2.2b are the positive and negative-ion LD mass spectra, respectively, of the blue
gel ink. In both spectra, the peaks at m/z 575 correspond to the intact
desorption/ionization of Copper Phthalocyanine. The mass difference due to the addition
or loss of an electron (an electron has a mass of9.11 x 10'31 kg) is neglected since the
change in the mass of the molecule is not significant. AS mentioned previously, an
electron is responsible for the ionization of the neutral molecule. An electron is donated
to a neutral molecule (M) of Copper Phthalocyanine causing the pigment to become
negatively-charged (M") and can be detected in negative-ion LDMS. In positive-ion

LDMS, Copper Phthalocyanine becomes positively-charged (M1) when an electron is

45

removed from the neutral molecule. The loss of an electron is the favored mechanism for
the positive ion formation of Copper Phthalocyanine. The most intense isotopic peak for
Copper Phthalocyanine has an m/z value of 575 which denotes that electron transfer is

the preferred D/I mechanism for the dye in both positive and negative-ion LD modes.

46

 

Figure 2.1: Dye structures a) Copper Phthalocyanine and b) monosulfonated Copper
Phthalocyanine

47

Relative Intensity

b)

Relative Intensity

 

 

 

 

 

 

 

 

 

 

 

575

520

I I l L AI 1

200 400 600 800 1000
mlz
575
80
654
1111. lzLTJhL Juli- 1. I11 .L 1 .f 1 I

200 400 600 800 1000

mlz

Figure 2.2: LD mass spectra of blue gel ink containing Copper Phthalocyanine a)
positive-ion and b) negative-ion

48

The distribution of the isotopic peaks observed in the LD mass Spectra contains
information that can be used to characterize Copper Phthalocyanine as well. The isotopic
distribution of the peaks depends on the natural abundance of the isotopes (atoms that
have the same number of protons but differ in the number of neutrons) that make up the
composition of the species. The mass of a neutron is 1.67 x 10'27 kg (~1 Da), so the mass
difference between isotopic peaks is due to the number of neutrons of the isotopes.
Several atoms have characteristic isotope distributions. The abundance and distribution
of isotopes of a dye molecule are specific and can be used to characterize the colorant.
Molecular mass and isotopic information are specific for a single molecular formula and
can be used to identify a species with the exception of isomers. Capper Phthalocyanine
has a unique isotopic distribution as seen in Figure 2.3. The isotopically-resolved peaks
of Copper Phthalocyanine obtained experimentally using LDMS and theoretically using
an isotopic distribution calculator accessed through the website of Scientific Instrument
Services6 at http://207.97. 159.7/cgi-bin/mass10.pl are compared in Figures 2.3a and 2.3b,
respectively. The Spectrum in Figure 2.3b was recreated by obtaining the theoretical
relative peak intensities from the website and plotted versus mlz values of Copper
Phthalocyanine. The atoms (C, H, N, and Cu) contained in Copper Phthalocyanine along
with the isotopic masses and percent abundances are listed in Table 2.1. The peak at m/z
575 represents 12C321H1614N863Cu, the most abundant combination of atoms. The 65 Cu
isotope provides the largest contribution to the peak observed at m/z 577. The isotope
peak pattern along with the molecular mass are two pieces of information obtained from

the mass spectra that are characteristic of ink dyes in general.

49

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a) fl

5

tn

f.»

s

i

E

0

I:

570 575 580 585
mlz
b)
575
5'
t0
.5.
s.
g 577
E
e 576
578
579
t 580
570 575 580 585
Ml

Figure 2.3: The isotopic peak distribution of Copper Phthalocyanine 3) experimental
isotopic distribution in negative-ion mode b) theoretical isotopic distribution for positive
and negative-ion modes

50

 

 

 

 

 

 

 

 

 

 

 

 

Atom Isotopic Mass Percent
(u) Abundance

1H 1 .007825 99.9885
2H 2.014102 0.115
120 12.000000 98.93
13C 13.003355 1 .07
14N 14.003074 99.757
15N 15.000109 0.368

530a 62.929601 69.17

65Cu 64.927794 30.83

 

 

Table 2.1: Isotopic masses and natural abundances for the atoms that comprise Copper
Phthalocyanine

Determining the ionization mechanism of a dye in LD is also useful information
for characterization. The ionization of the dye is influenced by the charge state of the
structure. Charged and neutral dye molecules are ionized differently and can be
predicted. The charge state of a dye can be deduced from obtaining the negative and
positive-ion LD mass Spectra. As Stated previously, Copper Phthalocyanine is a neutral
molecule. As a general ionization rule in LDMS, neutral dye molecules form ions by
electron transfer or proton transfer. The mass (1.67 x 10'27 kg or ~1 Da) of a proton is
significant to influence the mass of the protonated or deprotonated molecule. Neutral
molecules undergoing proton transfer reactions will have a mass difference of 2 Da
between the peaks in the positive and negative-ion LD mass Spectra that correspond to
the intact analyte. For example, if the neutral molecule (M) has a molecular mass of 450,
then the protonated molecule, (M+H)+, will give rise to a peak at m/z 451 and the

deprotonated molecule, (M-H)‘, will give rise to a molecular ion at m/z 449, hence, there

51

is a two mass unit difference between the positive and negative ions of the analyte. The
ions formed and detected in LDMS provide important clues concerning the charge
distribution of the absorbing dye molecule. LDMS can be used to deduce the charge State
(cationic, anionic, or neutral) of the dye. The characteristic charge associated with an
unknown dye structure promotes identification. LDMS detects singly-charged or neutral
dye molecules and allows a distinction to be made between dyes that are cationic,
anionic, and neutral.

Several black ballpoint pen inks from various manufacturers have been analyzed
by LDMS. All of the inks contain the same two dyes, Crystal Violet and Metanil Yellow,
which have the structures shown in Figures 2.4a and 2.4b, respectively. Analyses of
Crystal Violet and Metanil Yellow by LDMS demonstrate the ability of the technique to
distinguish between cationic and anionic dyes. Figures 2.5a and 2.5b are representative
examples of the positive and negative-ion mass spectra, respectively, that are acquired
when black ballpoint ink is examined directly from paper by LDMS. The most abundant
peaks in the LD mass Spectra correspond to the D/I of the intact dyes that are contained in
the ink since the other components in the ink are incapable of efficiently absorbing the
UV pulsed laser energy. Energy absorption is necessary to ensure desorption/ionization
of the Species. The positive and negative-ion LD mass spectra of the black ink each
contain one peak that is relatively intense. The positive-ion mass spectrum contains a
peak at m/z 372 representing the intact desorption of the cationic component of the dye
Crystal Violet [C25H30N3]+. A peak at m/z 352 in the negative-ion Spectrum denotes the
D/I of the anionic component of the dye Metanil Yellow [C13H14N3O3S]'. Unlike singly-

charged colorants, neutral molecules of Copper Phthalocyanine (M) yields both positive

52

(M1) and negative (M°') ions when ink is ablated with the UV laser as discussed
previously. In the LD experiment, singly-charged colorants are detected in their native
charge state. The negative-ion mass spectrum clearly indicates that Crystal Violet does
not form an intact negative ion and Metanil Yellow is undetected in the positive-ion mass
spectrum. Desorption/Ionization of multiply-charged dyes directly by LDMS is typically
not observed. The charge state needs to be reduced prior to detection and the multiply-
charged dyes are detected as singly-charged ion in matrix-assisted laser
desorption/ionization MS. The charges associated with the polyionic dye Still influence if

the species is detected in the positive or negative-ion modes or both.

 

Figure 2.4: Structure of the dyes used in several black ballpoint inks a) Crystal Violet
and b) Metanil Yellow

53

 

 

 

 

 

 

 

 

 

 

a) 372
E

3

2
E
3
E

e
a:
0 1 00 200 300 400 500
mlz
b)
352

E

“c"

3

E

o

.?..

E

o

I

1 56
30 [171 l
0 1 00 200 300 400 500
mlz

Figure 2.5: LD mass Spectra of black ink from Sanford® Supergrip ballpoint pen a)
positive-ion and b) negative-ion

54

Dye Impurities, Fragment Ions, and Cluster Ions

Singly-charged dyes are detected intact by LDMS, however, there are some
instances when additional peaks appear in the mass spectrum of a single commercial dye.
The appearance of peaks lower in mass than the peak corresponding to the intact dye may
be from two sources, dye impurities or fragment ions, from the analyte. The formation of
clusters or presence of impurities may be detected with m/z values higher than that of the
intact dye.

Commercial dyes are frequently sold as impure mixtures7 and Methyl Violet 2B
purchased from the Aldrich Chemical Company is an example. The dye mixture is
comprised of Crystal Violet, Methyl Violet 2B and tetramethylpararosaniline. In fact, the
peak associated with the dye impurity, Crystal Violet, dominates the mass spectrum. AS
shown in Figure 2.6. The three dyes are cationic and are detected solely in the positive-
ion LD mass spectrum. The m/z values of Crystal Violet and Methyl Violet 2B are 372
and 358, respectively. Surprisingly, Crystal Violet is more concentrated than Methyl
Violet 2B according to the LD mass spectrum and this observation has also been made by
Fales et al. 9. The detection of dye impurities is not uncommon when analyzing ink
samples. The peak at m/z 654 in the negative-ion LD mass spectrum of the blue gel ink
corresponds to a monosulfonated Copper Phthalocyanine, Shown in Figure 2. lb. This
dye may be present as an impurity as well. To make certain that the dye is an impurity in
the gel ink, the phthalocyanine pigments Should be purchased and analyzed directly.
Frequently, multiple dyes are used in the manufacture of ink, so both dyes could have

been intentionally used.

55

 

372

358

Relative Intensity

344

330 1,1

300 325 350 375 400
mlz

 

 

 

 

 

Figure 2.6: Positive-ion LD mass spectrum of Methyl Violet ZB purchased from the
Aldrich Chemical Company

Although a dye may be detected intact, peaks can also arise lower in mass
compared to the dye that corresponds to prompt fragment ions. Fragment ions are
structural pieces that are associated with the intact dye. Fragment ions are formed when
the dye molecule breaks apart due to the fact that fragmentation is energetically favored,
partially or completely, for that particular species. Fragment ions were identified for the
monosulfonated Copper Phthalocyanine complex and Metanil Yellow.

Although, Copper Phthalocyanine (CP) is stable upon D/I of the dye by LDMS,
the monosulfonated CP derivative experiences some fragmentation. When a hydrogen
atom is replaced by a sulfonic acid (-SO3H) group a mass increase of 80 Daltons results,
and the monosulfonated CP derivative is formed as shown in Figure 2. lb. In Figure 2.2b,

the peak at m/z 80 corresponds to sulfite ion (803’) and is evidence that a sulfonated dye

56

is present in the ink. Another ion that has an m/z 80 is HPOg", however, this functional
group has not been observed in dye structures. Examinations of the dye Structures in the
literature7‘9 lead to this conclusion. Upon laser D/I of the copper phthalocyanine dyes,
the monosulfonated metal complex fragments losing the sulfonic acid. The loss of this
functionality has also been observed in ESI MS and MALDI MS‘O‘”. MALDI and ESI
are softer ionization techniques than LD; however, fragment ions can be formed when
tandem MS methods are applied. Bruins et al. state that m/z 80 denotes 803" and can be
used as an indicator that polysulfonated azo dyes are present in the sample”. The
formation of fragment ions in ESI MS was carried out by implementing collision-induced
dissociation (CID)“‘12. The loss of the metal center has not been observed using MALDI
MS nor ESI MS12 as well as in the experiments performed here with LDMS. Copper
appears to be strongly bound to the phthalocyanine. Conneely et al. found the metals Cu,
Ni and Zn in phthalocyanine complexes are stable, but the analysis of the aluminum
phthalocyanine derivative gave rise to a peak at m/z 27 which corresponds to the loss of
Al”. The peak at m/z 520 in the positive-ion LD mass spectrum of the Copper
Phthalocyanine has not been identified; however, the peak is always observed when the
pigment is analyzed and has a similar isotope distribution as copper complex.

The low mass peaks (< 300 Da) in the negative-ion LD mass Spectrum (Figure
2.5b) of the black ballpoint ink provide some useful information about the structure of
Metanil Yellow. When analyzing an ink a complete examination of the LD mass
spectrum should be carried out as a general rule. An intriguing aspect of Metanil Yellow
is that, although, the dye was detected intact, the colorant is not entirely resistant to

prompt fragmentation which can be seen at m/z values below that of the intact dye. In

57

order to directly examine Metanil Yellow to avoid possible interference from other ink
constituents that in the black ink, the commercial dye was purchased from the Aldrich
Chemical Company and was analyzed. Figure 2.4 Shows the negative-ion LD mass
Spectrum of Metanil Yellow analyzed directly from a metal plate. The region of the low
mass peaks was enhanced as seen in Figure 2.8. Possible fragment ions are illustrated in
Figure 2.9. A Metanil Yellow fragment ion was observed at m/z 156. Other analytical
methods including MALDI, TLC, and post source decay (PSD) may be used to make
certain that this was a fragment ion of Metanil Yellow rather than an impurity. The peaks
at m/z 80, 156, and 171 are associated with fi'agment ions from the D/I of Metanil
Yellow. All three fragment ions (m/z 80, 156, and 171) have been previously observed
for the dyestuff, Acid Orange 7'0, as seen in Figure 2.10. The dyestuff was analyzed
using ESI tandem MS and MALDI-PSD MS. The structural similarities Shared between
the Acid Orange 7 and Metanil Yellow result in the formation of the same fragment ions.
The peak at m/z 171 corresponds to ions formed from the cleavage of the ketohydrazone
single N-N bond and m/z 156 results as ions are formed the homolytical fission of the azo
C-N bond“). The peak at mlz 80 was previously discussed for the monosulfonated
Copper Phthalocyanine complex and corresponds to the loss of sulfonic acid form the

structure of Metanil Yellow.

58

 

352

 

 

 

 

 

 

i?
to
.8
.5
S
:4
a?
1 56
727
80 1 l
l 171 I i- v.1 I 1 . 1
o 200 400 600 800

Figure 2.7: The direct analysis of Metanil Yellow from a metal plate by negative-ion
LDMS

 

 

 

 

 

156

i?
a
if»
.5
.‘z’
E
0
C

80

l 171

50 100 150 200

mlz

Figure 2.8: Negative-ion LD mass spectrum of Metanil Yellow Shows peaks due to
fragment ions

59

'03s

80

156

 

Figure 2.9: Dye structure of Metanil Yellow with labeled values that correspond to
possible fragment ions or neutral molecules

_|| mlz 80

Ho /
Q —’
'038‘.’N'\ m/z171
H

Figure 2.10: Acid Orange 7 (to the left of the arrows) and fragment ions (m/z 80, 156,
171) formed when analyzed by ESI tandem MS and MALDI-PSD MS11

60

f

To identify the source of low mass peaks, a method that distinguishes between
fragment ions and ions representing dye impurities should be applied. Three methods
that may be used in combination with LDMS to determine if the peaks are due to
fragment ions or dye impurities include TLC, post source decay (PSD), and
implementation of a matrix (MALDI). TLC indirectly coupled with LDMS is a simple
experiment that can be used to determine the source of the ions. TLC separates species in
a mixture and is suitable for the separation of multiply-charged dyes in an ink. A dye
present in an ink as an impurity can be separated from other colorants. The individual
dyes, including the impurity, can be extracted from the TLC plate and subsequently
analyzed by LDMS to confirm that the low mass peaks dyes and not fragment ions.
Prompt fragment ions are formed during the desorption/ionization process. Prompt
fragments are encountered in other ionization techniques, notably, electron impact (E1)
and are generally more extensive in El than UV LD. Another method to determine if
fragment ions were produced is to employ a matrix. Matrices are typically used in
LDMS for the detection of analytes that are prone to undergo fragmentation. The method
is recognized as matrix-assisted laser desorption/ionization and is used for the analysis of
biomolecules. The matrix assists in the D/I of the analyte while reducing prompt
fragmentation. A matrix can be added to the ink and the fragment ions should be
eliminated or reduced. The addition of the matrix should eliminate or reduce the
formation of fragment ions. If the low mass peaks are absent when the ink is analyzed by
MALDI, then low mass peaks observed previously in the LD experiment likely represent

fragment ions.

61

The formation of clusters (dimers, trimer, etc.) result from electrostatic
interactions between ions and has been observed for ionic dyes examined by LDMS.
Desorption/Ionization of Metanil Yellow resulted in the formation of at least one cluster
and can be seen in the negative-ion LD mass spectrum in Figure 2.11. The peak at m/z
727 corresponds to a negatively-charged cluster. The peak represents a dimer and
consists of two Metanil Yellow ions and one sodium ion [(2 x 352 + 23) = 727]. Multiple
clusters may from one dye are easily identified by relative peak intensities. As the
number of dye monomers increase, the peak intensity of the cluster decreases. In other
words, the peak intensity of a trirner is smaller than the intensity of a dimer, in general,
for a given dye. Cluster recognition may be determined by analyzying the commercial
dye using TLC. When Metanil Yellow was subjected to TLC, the analysis indicated that
no impurity dyes were present in the commercial dye since only one band was present.
Based on this additional information, the peak at m/z 727 was confidently assigned to a

Metanil Yellow dimer.

62

 

Relative Intensity

727

Ii I1, ,1.

400 500 600 700 800
mlz

.—

 

 

 

 

 

Figure 2.11: Negative-ion LD mass spectra of Metanil Yellow on paper Shows peaks due
to cluster ions
Coupling TLC with LDMS

Inks are complex mixtures and frequently, contain multiple dyes making the
identification by LDMS difficult, at times. TLC can be coupled with LDMS to ease the
identification process. The color and the number of dyes present in an ink are determined
by TLC. The CI name of the dye is based on its color and the separation of the dyes from
an ink by reveals their true color. The ink color can be misleading for determining the
color of the dyes in the ink. Frequently, the ink color is not the same color as the dyes
and depends on the combination of dyes used. For example, the combination of Crystal
Violet and Metanil Yellow are surprisingly used to make black ink for several ballpoint
pens as previously stated. When a concentrated, aqueous Crystal Violet solution is dried

on paper the spot appears to be black. The addition of Metanil Yellow did not appear to

63

influence the color of the dye mixture. Despite the use of violet and yellow dyes, the
color of the ink is black which can be misleading when trying to identify a dye based on
the color of the ink. TLC data, including the number and color of the bands, is acquired
to prevent the possibility of misconstruing the data. Knowing the dye color lessens the
search for the identification purposes. Indirect coupling of TLC with LDMS first used
for the analysis of black ballpoint ink allowed Metanil Yellow to be identified (the
identity of Crystal Violet was known). Figure 2.12 shows three distinctly colored bands
of a TLC chromatogram upon analyzing black ballpoint ink. The two purple bands
correspond to Crystal Violet (D1) and Methyl Violet 23 (D2) and the yellow band
represents Metanil Yellow (D3). The bands were scraped, solvent extracted and analyzed
by LDMS. The analysis of the extracted dye confirmed that the yellow TLC band
denoted the peak at m/z 352 in negative-ion LD mass Spectrum, the purple bands, D1 and
D2, were correlated with m/z 372 and 358, respectively, in the positive-ion LD mass
spectrum. Preliminary experiments of coupling TLC with LDMS directly showed
potential. Direct coupling of TLC eliminates the extraction step prior to the LDMS
analysis reducing the analysis time. Instead of using a glass TLC support, a plate backed
with aluminum was used for the direct analysis. Glass TLC plates are difficult to couple
directly with LDMS for two reasons (1) glass is difficult to cut and mount to the sample
plate and (2) glass is an insulating material which passes a threat when trying to obtain
analyte signal. Figure 2.13 demonstrates that adequate signal can be acquired when the
aluminum TLC plate is directly coupled with LDMS. The dyes in black ballpoint ink
from a BIC® Cristal pen were separated by TLC and a single purple band which

corresponded to Crystal Violet was analyzed directly with LDMS. The positive-ion

64

spectrum shows the dye was easily detected showing strong Signals with isotopic
resolution (results are comparable to detecting Crystal Violet directly from paper). The
presence of the lower mass peaks were most likely due to the silica from the TLC plate.
Occasionally, a peak at m/z 27 was observed in the positive-ion LD mass Spectra. Most
likely, this peak is due to the DH of aluminum metal from the TLC plate. Paper
chromatography was also used to separate the ink dyes in black ballpoint ink, but the

separation was poorly resolved so further LDMS coupling was not employed.

65

 

 

‘9 ”1 TLC Band Dye

 

 

 

 

 

 

 

 

 

S1 Crystal Violet
—— D3
53 .
82 __ D2 82 Methyl Vlolet 2B
81 .3—4— D1 83
Tetramethylpararosaniline
D1 Crystal Violet
DZ Methyl Violet 2B
D3 Metanil Yellow

 

 

 

 

 

 

 

Figure 2.12: TLC analysis of black ballpoint ink from BIC® Cristal a) thin-layer
chromato gram and b) TLC bands and corresponding dyes

66

 

Relative Intensity

 

372

[L- .ALLL. A - I 4.. 1

0 1 00 200 300 400 500

 

 

 

 

 

 

 

 

b)

Relative Intensity

 

 

 

352

J hi IIHJ. I ILL I I
I T

0 100 200 300 400 500
mlz
Figure 2.13: Direct coupling of TLC with LDMS: bands are analyzed by LDMS directly

from the aluminum TLC plate a) purple band (Crystal Violet) and b) yellow band
(Metanil Yellow)

 

 

 

 

 

 

 

 

 

 

 

67

References

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

Whitten KW, Davis RE, Peck ML. General Chemistry. 5‘h ed. Saunders College
Publishing, Fort Worth, 1996

Grim DM, Siegel J A, Allison J. Evaluation of Desorption/Ionization Mass
Spectrometric Methods in the Forensic Applications of the Analysis of Inks on
Paper. J Forensic Sci. 2001 (46) 1411-1420

Grim DM, Siegel JA, Allison J. Does Ink Age Inside of a Pen Cartridge? J
Forensic Sci. 2002 (47) 1294-1297

Grim DM, Siegel JA, Allison J. Evaluation of Laser Desorption Mass
Spectrometry and UV Accelerated Aging of Dye on Paper as Tools for the
Evaluation of a Questioned Document. J Forensic Sci. 2002 (47) 1265-1273

Dunn JD, Siegel J A, Allison J. Photodegradation and Laser Desorption Mass
Spectrometry for the Characterization of Dyes Used in Red Pen Inks. J Forensic
Sci. 2003 (48) 652-657

Manura JJ, Manura DJ. Isotope Distribution Calculator anfd Mass Spec Plotter.
Scientific Instrument Services. 1996-2003 <http://207.97.159.7/cgi-
bin/mass]0.pl>

Venkataraman K. The Chemistry of Synthetic Dyes. Vol. 1, Academic Press Inc.,
New York, 1952

Fales HM, Pannell LK, Sokoloski EA, Carmeci P. Separation of Methyl Violet 2B
by High-Speed Countercurrent Chromatography and Identification by CF-252
Plasma Desorption Mass-Spectrometry. Anal Chem. 1985 (57) 376-378

Green PI. The Sigma-Aldrich Handbook of Stains, Dyes, and Indicators. Aldrich
Chemical Company, Inc., Milwaukee, 1990

Sullivan AG, Garner R, Gaskell SJ. Structural Analysis of Sulfonated Monoazo
Dyestuff Intermediates by Electrospray Tandem Mass Spectrometry and Matrix-
Assisted Laser Desorption/Ionization Post-Source Decay Mass Spectrometry.
Rapid Commun Mass Spectrom. 1998 (12) 1207-1215

Conneely A, McClean S, Smyth WF, McMullan G. Study of the Mass
Spectrometric Behavior of Phthalocyanine and Azo Dyes Using Electrospray
Ionization and Matrix-Assisted Laser Desorption/Ionization. Rapid Commun
Mass Spectrom. 2001 (15) 2076-2084

68

(12) Bruins AP, Weidolf LO, Henion JD. Determination of Sulfonated Azo Dyes by
Liquid Chromatography/Atmospheric Pressure Ionization Mass Spectrometry.
Anal Chem. 1987 (59) 2647-2652

69

Chapter Three: Direct LDMS Analysis of Ink Dyes: Limitations and Solutions

Two limitations were encountered when multiply-charged ink dyes were
examined by direct LDMS. One disadvantage is that the mass Spectra lack structural
information. Unlike an E1 MS experiment, little fragmentation, if any, has been observed
in the LDMS experiments. This limits the amount of Structural information that can be
acquired from the LD mass Spectra, thus making dye characterization arduous. A second
limitation is the ability of multiply-charged dyes to be detected by LDMS. The energy of
the laser is insufficient to desorb dyes that contain more than one charge. The latter
challenge was overcome by chemically modifying the dye prior to the MS analysis and
the results are presented in Chapter Four. Previous research carried out in our laboratory
implemented photodegradation to characterize ink dyes'. Using photochemistry to
characterize multiply-charged dyes is potential research for the filture, so the

characterization method will be discussed.

Experimental
Laser Desorption Mass Spectrametg

The inks and dyes on paper were analyzed using a PE Biosystems Voyager DE
(Farmingham MA) mass spectrometer. The instrument utilizes a pulsed nitrogen laser
(337 nm, 3 nanosecond pulses, 3 Hz) and a linear time-of-flight mass Spectrometer. The
user-selected parameters for the LDMS experiments include an accelerating voltage of
20,000 V for detection of positive ions and -15,000 V for detection of negative ions, an

intermediate source grid voltage that is 94% of the accelerating voltage, a guide wire

70

voltage that is opposite in bias, and 0.05% in magnitude of the accelerating voltage, and
an extraction delay time of 100 ns. To parallel inks on paper, aqueous dye solutions were
prepared, spotted on paper, and allowed to dry for subsequent LDMS analysis. The dyes
analyzed were Rhodamine B (Sigma, St. Louis, MO) and Rhodamine 6G (Aldrich,
Milwaukee, WI). Photodegradation was accomplished by using incandescent light (75

W, 120 V light bulb).

Dye Characterization by Photodegradation and IDMS

In the laser desorption experiment, energy is deposited into the ink in 2-3
nanoseconds, quickly raising the temperature of a very small portion of the sample to a
point where desorption is favored over degradation. Thus, ionic species desorb intact,
giving a Single mass spectrometric peak (plus isotopic peaks), and providing molecular
weight information and possibly elemental composition. One dye yields one peak, so
structural information is limited. In order to obtain structural information from the
LDMS experiment dyes on the paper substrate are degraded photochemically. LDMS
analysis following photodegradation yields a set of new peaks, each representing a
degradation product. Structural insights can be obtained from the degradation products
and understanding the reactions that occur. For example, if Crystal Violet is irradiated
with UV radiation, N-demethylation reactions occur and the dye degrades to form Methyl
Violet ZB and extensive degradation leads to the formation of other homologues.
Replacement of a methyl group from Crystal Violet by a hydrogen atom results in a net
loss of 14 atomic mass units. Crystal Violet can lose up to six methyl groups in this

manner. Previous work carried out by Grim et al. utilized the Crystal Violet degradation

71

and proposed an approach for determining the age of a documentz. From an analytical
standpoint, the degradation products formed naturally or artificially are related to the dye
structure. The fact that Crystal Violet forms a maximum of six degradation products due
to demethylation provides an important structural context in which the dye can be

identified.

The use of LDMS to detect dyes in red ballpoint inks marked the discovery of two
isomeric dyes which instigated the development of a method for characterizing ink dyes
by photodegrdation. The combination of photodegradation and LDMS for dye
identification was demonstrated through a challenging example, the differentiation
between two isomeric xanthene dyes, Rhodamine B and Rhodamine 66. Both dyes are
used in the manufacturing of red ballpoint inks3 and an ink may contain one or both of
the dyes. The xanthene dyes are cationic and have the molecular formula
[C23H31N402]+[C1]'. Shown in Figure 3.1 are the positive-ion LD mass spectra of the
isomers. The intact detection of each dye is denoted by the peak at mlz 443. Although,
the mass spectrum of Rhodamine B contains an additional peak at m/z 399 that can be
used to characterize the dye structure, Rhodamine 6G could not be characterized. If an
unknown ink sample was analyzed by direct LDMS and produced a spectrum similar to
that of Rhodamine B (Figure 3.1a), then the presence of Rhodamine 6G is uncertain. In
order to properly characterize an ink that contains two dyes, both need to be identified.
The challenge to differentiate the rhodamine dyes was overcome by using
photochemistry to artificially degrade the dyes and subsequently analyze the degradation

by LDMS‘.

72

 

 

 

 

 

 

 

 

 

 

 

a)
443
E
m
5
s
g
3:
O
a:
399
415 465
I . I- 1 LI II
300 350 400 450 500
mlz
b) 443
E
w
8
g
0
..>.
5
O
a:
415 w
, I I l ,
300 350 400 450 500
mlz

Figure 3.1: Positive-ion LD mass Spectra of isomeric, red ink dyes a) Rhodamine B and
b) Rhodamine 6G

73

Structural similarities shared between the isomeric rhodamine dyes and Crystal
Violet were used to develop an analytical method that involved degrading dyes using
incandescent light. The detection of Crystal Violet by direct LDMS was briefly
discussed in Chapter Two. Grim et al. analyzed naturally aged ink samples and
formulated ink dating curves based on an accelerated aging study that incorporated the
use of ultraviolet radiationz. Ultraviolet radiation was found to mimic the natural
degradation process of Crystal Violet. The dye was found to naturally and artificially
degrade by a process called oxidative demethylationz. The dye structure contains Six
methyl groups that are chemically modified when degradation occurs. The dye Shown in
Figure 3.2a can form a maximum of Six degradation products as depicted in Figure 3.2b,
which can be simultaneously detected by LDMS. N-demethylation occurs by the
substitution of a methyl group for a hydrogen atom. One methyl group substitution
results in a mass difference of 14 Da. Shown in Figure 3.3 is the positive-ion LD mass
Spectrum of blue ballpoint ink containing Crystal Violet which has been degraded using
UV radiation. In the mass spectrum, the peaks due to Crystal Violet (m/z 372) and the
corresponding dye homologues (m/z 358, 344, 330, 316, and 302) are separated by 14

atomic mass units.

74

 

 

m/z
372
358

\ +’CH3 b)

 

 

 

 

330
31 6
302

288

 

 

41000-5010):

 

Hac—Ilv N—CH3
CH3 H36

 

 

 

O

 

Figure 3.2: Crystal Violet a) dye structure and b) m/z values of the dye and
corresponding degradation products (11 represents the number of methyl groups bonded to
the amine nitrogen atoms)

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a)
E

2

2

E

0

.2

E

C!

a:

L LII IEIIIAI. I 1 1 L IMAM II I I [All LIL
0 100 200 300 400 500
mlz
b) 344 358
330 372

E:

tn

5

E

3 316

.2

0

I:

302
288 i
250 300 350 400
mlz

Figure 3.3: Positive-ion LD mass spectra of blue ink from BIC Cristal ballpoint pen
degraded with UV radiation a) complete spectrum from m/z 0-500 and b) expanded

region from m/z 250-400

76

The structures of Rhodamine B and Rhodamine 6G are shown in Figures 3.4a and
3.4b, respectively. When the dyes are subjected to UV radiation or incandescent light
they degrade and the deethylated products are detectable by LDMS. As the rhodamine
dyes degrade, hydrogen atoms replace the ethyl groups bonded to the nitrogen atoms.
Rhodamine B forms four N-deethylated products when chemically induced by
incandescent light where two N-deethylated products are form from Rhodamine 6G when
photodegraded. Tables 3.1a and 3.1b illustrate the m/z values of the deethylated products
of Rhodamine B and Rhodamine 6G, respectively. The degradation products result from
a Similar degradagtion mechanism as Crystal Violet. The m/z values of the dye and the
deethylated degradation products are separated by 28 Da. The deethylated products can
be used characterize the dyes individually, however, the deethyalted products can not be
used to discern Rhodamine 6G apart from Rhodamine B if an ink contains both dyes.
Fortunately, another structural characteristic of Rhodamine 6G can be used to

differentiate the dyes which will be discussed in the next section.

77

9

Il_,cl-l,,c\N

 

 

Figure 3.4: Isomeric dye structures a) Rhodamine B and b) Rhodamine 6G

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a) n m/z b) n m/z
4 443 2 443
3 41 5 1 41 5
2 387 0 387
1 359
0 331

 

 

 

 

Table 3.1: m/z values of the dye and corresponding deethylated degradation products (It
represents the number of ethyl groups bonded to the amine nitrogen atoms) for a)
Rhodamine B and b) Rhodamine 6G

78

Isomeric Dye Characterization

According to US Patent 5,993,0983, two xanthene dyes, Rhodamine B and
Rhodamine 6G, can both be used to manufacture red pen ink. The Sigma-Aldrich
Handbook of Stains; Dyesgnd Ind—icators4 revealed that the two rhodamine dyes are
cationic isomers with a molecular mass of 443 Daltons (excludes the mass of the
counterion, Cl'), hence, both dyes yield a peak at m/z 443 in the positive-ion LD mass
spectrum. The dyes are distinguished by their functional groups. AS previously stated,
Rhodamine B contains four ethyl groups, two attached to each nitrogen atom, while
Rhodamine 6G contains two ethyl groups, one attached to each nitrogen atom, but the
number of ethyl groups can be used to differentiate the isomeric dyes as a mixture. The
appearance of the direct LD mass spectrum would lead one to believe that only
Rhodamine B was contained in the ink. In this case, the deethylated degradation products
overlap in m/z values and can not be used for dye differentiation. However, other
structural differences may help to undoubtingly distinguish the dyes in LDMS.
Fortunately, the xanthene dyes have additional differences in functional groups.
Rhodamine B contains a carboxylic acid group where Rhodamine 66 contains an ester
functionality, however these groups can not be used as characteristics to identify
Rhodamine 66. Another difference between the dye structures is Rhodamine 6G
contains methyl groups ortho to the amine groups. The methyl group was used to
characterize Rhodamine 66 as well as to determine the presence of the dye in an ink
containing both xanthene dyes. The characterization of Rhodamine B and Rhodamine

6G will be discussed in the following two sections.

79

Rhodamine B

The direct LDMS spectra of the two rhodamine dyes are different in that a low
intensity peak at m/z 399 is consistently present in the mass Spectrum of Rhodamine B
(Figure 3.5a). The peak represents the decarbonylated form (RB-C02)+ of Rhodamine B
(RB)+. The mass difference between the intact dye and this species is 44 amu, which
corresponds to carbon dioxide. Rhodamine B, a carboxylic acid, can decarbonylate while
the analogous peak would not be expected to be seen in the mass spectrum of Rhodamine
66, which does not contain a free acid group. Initially, the peak at m/z 399 was assigned
to be a dye impurity. Rhodamine 6G is commercially available through the Aldrich
Chemical Company at a purity of 99% where Rhodamine B available through Sigma is
95% pure. The purity of the dyes appeared to correlate with the presence of the
additional peak at mlz 399 in the mass Spectrum of Rhodamine B and the absence of
additional peaks in the mass spectrum of Rhodamine 6G. This information and LDMS
results lead us to believe that the peak was due to a dye impurity. However, more recent
data suggests that the species is the result of prompt fragmentation. TLC and MALDI
MS experiments showed that the peak at m/z 399 corresponds to a fragment ion. The use
of TLC and MALDI MS to distinguish between ions due to fragmentation and impure
dye mixtures were discussed in Chapter One. The chromatographic separation of
Rhodamine B yielded one TLC band and upon using indirect TLC coupled to LDMS the
band was identified as the intact dye. Dye impurities were not observed from the TLC
chromatogram. Figure 3.6a and 3.6b Shows the positive-ion mass spectra of Rhodamine
B analyzed by LDMS and MALDI MS, respectively. Upon using a matrix, the soft

ionization reduced the peak intensity which signifies that the m/z 399 ion is due to a

80

fragment ion produced during the D/I of Rhodamine B. In any instance, the LDMS
spectra of a red dye found in a pen can immediately suggest that the dye is Rhodamine B,
if the peaks at m/z 443 and 399 are present.

Although, Rhodamine B can be characterized by direct LDMS, photodegradation
can be used to characterize the dye based on the resulting degradation products. The LD
mass spectrum in Figure 3.5b was obtained upon photodegrading Rhodamine B. AS
previously discussed, upon being exposed to incandescent light, Rhodamine B forms four
deethylated products. These degradation products of Rhodamine B also Show the loss of

44 amu as seen with the D/I of the intact dye.

81

 

 

 

 

 

 

 

443
>5
.1: (a)
(I)
= 399
d)
H
3:: 415
H I
Q 443
>
’5 415
- (b)
v 387
a:
359
331 399
L @433 371
m / z

Figure 3.5: Positive-ion LD mass spectra of Rhodamine B (a) no exposure to
incandescent light and (b) exposed to incandescent light for 12 hours

82

 

 

 

 

 

 

 

 

 

 

 

a) 443
E
m
5
E
3
g
2 399
41 5
370 395 420 445 470
mlz

b) 443

E

to

5

E

3

E

0

a:

399 4L5
370 395 420 445 470
mlz

Figure 3.6: Positive-ion mass spectra of Rhodamine B analyzed by a) LDMS and b)
MALDI MS

83

Rhodamine 60

When Rhodamine 6G is directly analyzed by LDMS a peak at m/z 429 is barely,
but frequently observed as shown in Figure 3.7a. The intensity of the peak is relatively
small and is likely to be overlooked in the LD mass Spectrum of the dye (ink). However,
the intensity of the peak increases as the sample was degraded by incandescent light as
shown in Figure 3.7b. The structure of Rhodamine 6G contains two methyl groups on
the xanthene structure which dissociate from the dye upon photodegrading the sample. A
previous study was implemented to determine if the functional group yielded the peak at
m/z 429 by examining the photodegradation of two dyes, Rhodamine 123 hydrate and
New Fuchsinl. The analysis confirmed the loss of the methyl group from thdamine 6G
gave rise to m/z 429 ion. The formation of the degradation product was proposed to be
due to a bimolecular disproportionation reaction]. The presence of the peak is
characteristic of Rhodamine 6G which helps to discriminate against Rhodamine B which
may otherwise have been difficult to discern without. If an ink contained both xanthene
dyes, then the initial examination of the LD mass spectrum might suggest that
Rhodamine B was the only dye contained in the ink. Without the induced demethylation
of Rhodamine 6G there would be no way to differentiate the dyes by LDMS alone. The
photodegradation of Rhodamine 66 results, preferentially, in the loss of two ethyl groups
where Rhodamine B Shows the loss of four ethyl groups consuming the mass spectrum.
However, the peaks due to deethylation overlap, hence, N-deethyalation can not be used
directly to differentiate between the two dyes as a mixture. Photodegradation with
subsequent LDMS analysis was proven to be highly valuable when distinguishing

between isomeric dyes.

84

LDMS is a versatile and sensitive tool for detecting dyes in a variety of inks.
Spectra acquired from LDMS alone may not provide sufficient information for dye
identification. Dyes can be photodegraded directly on paper using incandescent light,
and the photoproducts can be analyzed by LDMS, providing structural information.
While Rhodamine 6G and Rhodamine B are isomers, the photodegradation/LDMS
combination can be used to distinguish between the two dyes. The structural differences

between the isomers allowed different photoproducts to be formed and detected.

85

 

443

 

 

 

 

 

>:

H

7,; (a)

Q

G)

H

t: 415

~ AI

cu 443
>
0:

c:

T, (b) 415

M 387

429 457
3Z3 401
m / z

Figure 3.7: Positive-ion LD mass spectra of Rhodamine 6G (a) no exposure to
incandescent light and (b) exposed to incandescent light for 12 hours

86

 

References

(1)

(2)

(3)

(4)

Dunn JD, Siegel JA, Allison J. Photodegradation and Laser Desorption Mass
Spectrometry for the Characterization of Dyes Used in Red Pen Inks. J Forensic
Sci. 2003 (48) 652-657

Grim DM, Siegel JA, Allison. Evaluation of Laser Desorption Mass Spectrometry
and UV Accelerated Aging of Dyes on Paper as Tools for the Evaluation of a
Questioned Document. J Forensic Sci. 2002 (47) 1265-1273

Osada T, inventors. Mitsubishi Pencil Kabushiki Kaisha, assignee. Aqueous Gel
Ink-filled Ball Point Pen. US Patent 5,993,098 Nov 1999

Green P]. The Sigma-Aldrich Handbook of Stains, Dyes, and Indicators.
Milwaukee: Aldrich Chemical Company, 1990

87

Chapter Four: Detection and Characterization of Multiply-Charged
Dyes from Pen Ink

We have recently learned that LDMS can not be used to detect several pen ink
dyes. Frequently, dyes analyzed directly from ink on paper did not yield substantial
peaks in the LDMS experiment. The assumption that we had encountered multiply-
charged ink dyes was made certain in this research by MALDI MS. A method was
developed for the detection and characterization of ink dyes that contain multiple
charges, Specifically, polysulfonated dyes as well as other dyes containing sodium
carboxylates (RCOz' Na“) and sodium alkoxides (RO' Na“). The method was the focus of
this thesis research. The results and discussion presented in this chapter entails (1) the
data used to determine the presence of multiply-charged dyes contained in pen inks, (2)
the experiments performed to detect multiply-charged dyes, and (3) the molecular
information acquired from MALDI and LD mass Spectra for characterizing polyionic ink
dyes.

Polyionic dyes containing sodium sulfonic acid, sodium carboxylate, sodium
alkoxide and irninium chloride groups were detected in this research by MALDI MS.
The detection of polysulfonated dyes by MALDI MS has been reported in the
literature”. We have Shown that MALDI MS can be used to detect multiply-charged
dyes that contain functional groups, besides sodium sulfonic acid (RSOg' Na+), that
contribute to the overall charge. This research demonstrated that UV MALDI MS can be
used to successfully desorb a variety of polyionic dyes, directly fiom a paper surface, as
singly-charged ions, with the appropriate selection of MALDI matrix and additive. The

experiment demonstrates the same sensitivity (pmolar amounts of analyte detected) and

88

the same Simplicity as LDMS with the exception of additional sample preparation. The
detection of multiply-charged ink dyes by MALDI MS requires only a minor
modification, addition of a few rnicroliters of a matrix/analyte solution to the ink-on-
paper sample which is allowed to dry, then analyzed as in the LDMS experiments. While
LDMS spectra do not Show desorption of intact polyionic dyes, fragment ions indicate

that multiply-charged dyes are present and allows the dye structures to be predicted.

Experimental
Mass Spectrometry

LDMS and MALDI MS analyses were carried out using a PerSeptive Biosystems
Voyager DE (Farmingham, MA) mass spectrometer. The instrument utilizes a pulsed
nitrogen laser (337 nm, 3 nanosecond pulses at 3 Hz) and a linear time-of-flight mass
Spectrometer. The user-selected parameters for the LDMS experiments include an
accelerating voltage of 20,000 V for detection of positive ions and _-15,000 V for the
detection of negative ions, an intermediate source grid voltage that is 94% of the
accelerating voltage, a guide wire voltage that is opposite in bias and 0.05% in magnitude
of the accelerating voltage, and an extraction delay time of 100 ns. MALDI MS
experiments utilized the same parameters except, in some cases, a delay time of 200 ns
for optimum resolution was necessary when dyes or ink was analyzed from paper. A gel
ink containing Copper Phthalocyanine was used as the calibrant for both positive and

negative-ion modes.

89

Laser Desorption Mass Spectrometry

Single written lines on MSU letterhead paper were analyzed directly. The laser
was focused on the ink line within the width of the pen stroke which is approximately
0.3-0.4 mm. Eighty-five pen inks were analyzed by LDMS, however, the results
presented here focus on the analysis of three inks. The LD mass spectra of liquid, blue
ink from a Pilot® Precise V7 and liquid, red ink from a Sanford® Uni-ball® Roller are
Shown. Results for liquid, blue ink from a Sanford® Uni-ball® Roller are discussed;
however, no LD mass spectra are presented. These particular pens were chosen since the
dyes in the inks were undetected by LDMS and were suspected to be polyionic.

In order to simulate the analysis of ink on paper, known dye solutions were

prepared with a concentration of 10,000 pmol/uL in 1:1 (v/v) methanol:water, spotted (5
uL) on paper and allowed to dry prior to the LDMS analysis. Several polyionic dyes
were purchased from either the Aldrich Chemical Company or the Sigma Chemical
Company and are listed in Table 4.1.
Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry

A smooth surface, lOO-well stainless steel (SS) VoyagerTM MALDI plate was
used for the analyses. Diammonium hydrogen citrate, DAHC, (J .T. Baker & Co.,
Phillipsburg, NJ) was used as an additive and 2-(4—hydroxyphenylazo)benzoic acid,
HABA, (Aldrich Chemical Co., Milwaukee, WI) was used as the matrix for the MALDI
MS experiments shown here. The concentrations of the reagents used were 100 mM
DAHC in water, 24 mg/mL of HABA (6 mg in 2:2:1 (v/v/v) acetonitrile:waterzmethanol),
and 100 pmol/uL dye solution in 1:1 (v/v) methanol:water, 1 [IL aliquot of each solution

was spotted in a well on the SS plate in the following order: HABA, DAHC, dye.

90

For the MALDI MS dye analyses on paper, 5 [1L aliquot of 10,000 pmol/uL dye
solution in 1:1 (v/v) methanol:water was spotted on MSU letterhead paper and allowed to
dry. Squares approximately 2x2 mm2 were cut from the paper and mounted on a
modified MALDI plate using Scotch double coated tape (3M, St. Paul, MN). DAHC was
applied first to the paper samples, followed by HABA. T WO 1 uL aliquots of DAHC
were Spotted on the paper samples and each droplet was allowed to dry between aliquots.
Three 2 uL aliquots of HABA were placed on top of the paper and each droplet was
allowed to dry between aliquots.

For the MALDI MS analyses of ink samples on paper, a single pen stroke was
made on MSU letterhead paper and squares approximately 2x2 mm2 were cut from the
paper and mounted on a disposable, gold VoyagerTM MALDI plate using Scotch® double
coated tape (3M, St. Paul, MN). DAHC was first applied to the paper samples, followed
by HABA. Two 1 ltL aliquots of DAHC were spotted on the paper samples and were
allowed to dry between aliquots. Three 2 11L aliquots of HABA were placed on top of
the paper samples and the each droplet was allowed to dry between aliquots. The ink
calibrant on paper was mounted adjacent to the samples.

Thin-Layer Chromatography

TLC was carried out using K5F Silica gel 150 A TLC plates (Whatman, Ann
Arbor, MI) with dimensions of 5 x 10 cm2 and a stationary phase thickness of 250 um.
The solvent system used for the separation of the dyes in the ink consisted of 70:35:30
(v/v) ethyl acetatezethanolzwater. Methyl Violet 2B (Aldrich Chemical Co., Milwaukee,
WI) contained Crystal Violet and tetrapararosaniline and a saturated solution was used as

a comparison standard. To indirectly couple TLC with MS, individual bands were

91

scraped from the TLC plate, placed in a centrifuge vial, and 5-10 [ILL aliquot of 1:1 (v/v)
ethanolzwater was added to extract the dyes. The vials were vortexed and the Silica was

separated from the dye solutions by centrifugation.

Determining the Presence of Multiply-Charged Dyes in Pen Ink

When LDMS was applied to 85 pen inks (varied in manufacturer, type and color)
the identification through the interpretation of LDMS spectra has been arduous due to the
complexity of the mass spectra or the lack of information that can be acquired from the
mass Spectra. Several inks, available to the public, likely contain dyes that are multiply-
charged. The assumption is based on LDMS data, TLC data, and patent information that
were accumulated for this research. The presence of polyionic dyes in ink is discussed in

the next three sections.
laser Desorption Mass Spectrometry

The mass Spectra in Figure 4.13 and 4.1b Show the positive and negative-ion LD
mass spectra, respectively, of a Pilot® Percise V7 rollerball pen that contains liquid, blue
ink. The molecular information that can be acquired from the mass spectra for dye
characterization is minimal. Peaks representing the detection of molecular ions are
absent, hence, molecular mass and elemental information of the dye can not be
deteremined. However, few peaks that are present correspond to fragment ions from the
dye molecule. Most of the dyes (neutral or singly-charged dye molecules) that we have
analyzed by LDMS have had m/z values between 300 and 1500. An LD mass Spectrum
is initially examined for dominant peaks that have m/z values in the range specified

above. The majority of the peaks in both LD mass spectra (Figure 4.1) are present at low

92

mass. In general, the complex mass spectra that have been encountered in this research
consist of peaks low in mass (<300 Da). This observation may suggest that an ink
contains polyionic dyes. If a dye is tetra-anionic, [Na]4+ [Dye]4' and there is no chemical
mechanism through which the dye could be desorbed and ionized with a single charge,
then the dye will not be detected intact. Dye fragmentation is a likely to occur and may
be induced when analyzing a species that is multiply-charged. Each fragment ion formed
gives rise to a peak in the mass spectrum. The formation of several fragment ions from
one dye may complicate the mass spectrum. The amount of fragmentation varies
depending on the species being analyzed. Typically, the mass spectra of multiply-
charged dyed contain few peaks, if any. The absence of peaks in a mass spectrum also
suggests that an ink contains polyionic dyes. If a multiply-charged dye is structurally
stable upon D/I, then the resulting mass spectrum will contain negligible peaks. LDMS
and other MS literature (except for electrospray ionization) have certainly shown that
ionization mechanisms yield predominantly singly-charged ions. In order to elucidate the
information that may be acquired from the LD mass Spectra for multiply-charged dye
characterization, a variety of commercial dyes containing multiple charges were obtained

and analyzed directly by LDMS.

93

 

 

 

 

 

 

 

 

 

a) 23
39
74
11. . 1.72
b) 79
170
42 120
M14185 393- -
0 200 400 600 800 1 000

mlz

Figure 4.1: Direct LDMS analysis of liquid, blue ink from a Pilot® Precise V7 rollerball
pen on paper a) positive-ion mass spectrum and b) negative-ion mass spectrum

94

Significant peaks due to fragmentation were not observed in the positive-ion LD
mass spectrum of the blue rollerball ink. The most abundant peaks in the positive-ion LD
mass spectrum (Figure 4.1a) have mlz value of 23 and 39 which correspond to sodium
and potassium ions. These alkali metal ions are typical artifacts present in the mass
spectra of ink. However, when the dyes are detected intact these peaks are not the most
abundant in the mass Spectrum. The negative-ion LD mass Spectrum (Figure 4.1b) shows
an intense peak at m/z value 79 which is frequently seen upon analyzing ink. Once again,
the peak intensity of the low mass ion is minimal compared to that of the dye if detected.
Currently, the peak has not been identified. Prompt fragmentation is useful information
that enables the structure of the dye to be predicted. For example, polysulfonated dyes
are known to readily loose sulfonic acid when analyzed by LDMS. The fragment ion that
results gives rise to a peak in the negative-ion mass spectrum at mlz 80 which is
indicative of a S03". The presence of this peak is an indication that an undetected dye by
LDMS contains multiple sulfonic acid groups. Further discussion about the
fragmentation of polysulfonated dyes can be found in Chapter Two. Other fragment ions
can be used to determine the presence of multiply-charged dyes. For instance, the peak at
m/z 170 in the negative—ion LD mass Spectrum (Figure 4.1b) appears to correspond to a
fragment ion, but may be overlooked since the intensity of the peak is relatively small.
An association has been proposed for the fragment ion and is discussed further detail
when results are presented for the dye structure. Since fragment ions denote structural
pieces of the intact species they aid in dye identification. The LD mass Spectra of the

liquid, blue ink demonstrates that direct LDMS can not be used to acquire molecular

95

information for multiply-charged dyes, although, low mass peaks are helpful for

determining that a polyionic dye in the ink contains specific functional groups.

In order to make certain that LDMS is incapable of detecting multiply-charge
dyes intact, several commercial dyes were purchased and analyzed directly. Dye
solutions were prepared with a concentration of 10,000 pmol/ltL and 100 pmol/uL in 1:1
(v/v) methanol:water. The dyes were analyzed both on paper and the SS plate directly
using LDMS in the positive and negative-ion modes. Analyses were performed with
Direct Red 23, Acid Blue 93, Direct Red 80, Acid Yellow 23, Solvent Blue 38, Acid
Black 1, and Reactive Black 5 whose structures are shown in Figure 4.2 (structures are
illustrated without their counterion, Na+). The Structural characteristic Shared among
these dyes is the presence of multiple sulfonic acid groups. Each sulfonic acid group
contributes -1 to the overall charge of the dye molecule. The number of charges varies
among the dyes. The dyes contain other functional groups such has hydroxyl groups that
may influence the overall charge state and detection in LDMS. Solvent Blue 38 was the
only multiply-charged dye out of the set that was detected intact by LDMS and some
fragmentation was observed. The direct analysis of the dye from paper and metal plate
were similar. The LD mass spectra of Solvent Blue 38 on the different substrates
contained peaks with the same mlz values with similar distributions of relative intensities.
Solvent Blue 38 is a Copper Phthalocyanine derivative which differentiates the dye from
the other Structures. The other dyes were not detected intact on paper, however, Acid
Blue 93 (100 meI/[ILL concentration only) was detected from the metal substrate. The
difference between Acid Blue 93 and the other dyes is that the structure contains a

positive charge due to the presence of an irninium group.

96

as owed 8 82583 n 652m 9:93. a. e5 a 662m e6< c .8 cam 23.. 3
.mm 2.5 Eczem e .8 ao=o> e6< a .3 Be seen 3 .8 com sense a as as 835-2338 E oeeoesm ”3. came

N1 1.

 

 

 

 

 

n 0 :2
- om __
m o
z 010 - mac .
19— H n
:o\ /zuz moo . z m o.
o
A 1 0: tag 1 AD
:o
100% Z“:
exclolzz Hz :zlol: moo-
__ :o __ Am.
0 o

97

so once ace 82583 n 662m 388m 9. can ._ 562m 23.. c .8 cam 23.. 3
.mm 2.5 Eczom e .8 4.6:; 25.4 6 .8 com coca 3 .2 com cacao a one as Ewaeoéaeea .6 238% ”S. 23$

.. now 2.14112 0 _
:0

2...:
«oz zflz‘ H
o:
m -o . moo.
-aom z”z mluzolazololwao . a

O

Inzolazololmao -

(D

i:

c

__
0

no

:2 _ - :2

’6?
' n

 

98

Thin-Layer Chromatography

TLC data collected has also provided insight into the charge state of the dyes used
in inks. Generally, the number and color of the dyes present in the ink are determined by
TLC. When an ink is analyzed by TLC each colored band on the chromatogram
represents one dye. Each relatively intense peak in the LD mass spectrum represents a
Single dye as well. The chromatogram and mass spectra, positive and negative-ion, are
compared to account for the number of dyes that are contained in an ink. LDMS is
incapable of detecting polyionic dyes and if the number of dyes detected in TLC and
LDMS are not the same, then the ink most likely contains multiply-charged dyes. Dyes
were extracted from the TLC bands and analyzed directly by LDMS. The absence of
peaks in the LD mass spectra of the extractions concluded that multiply-charged dyes are
present in the inks. TLC data supported the assumption that the liquid, blue rollerball ink
contained multiply-charged dyes. The TLC analysis suggested that the ink contained two
dyes and the observed color of the bands were blue and purple which were not detected
by LDMS. TLC data is acquired and compared to LD mass spectra to account for the
presence of polyionic dyes. A discrepancy between the number of TLC bands and the
number of peaks observed in the mass spectra signifies that multiply-charged dyes are

contained in an ink.
Ink Patents

A review of the patent literature resulted in the identifying more than 250 dyes
used for manufacturing pen inks. The dyes used in ballpoint, gel, and liquid ink have
been catalogued for this research. US. Patent 5,993,098 demonstrates the various dyes

that can be used in the manufacture of gel ink3. Many of the dyes share a common

99

feature of containing multiply-charged groups. Thus, the dyes are polyionic. Most of the
dyes encountered in the patents are polysulfonated azo dyes, and can be used in various
classes of pen ink such as ballpoint, gel, and liquid inks. This finding supported the TLC
and LDMS data that multiply-charged dyes are used in the manufacture of inks. The use
of polyionic dyes was unexpected, and introduces several complications to the LDMS
analysis. Fortunately, structures for the majority of the cataloged dyes have been found,
so m/z values have been predicted. The dye catalogue served as a reference for

interpreting the peaks present in the LD mass Spectra.

Multiply-Charged Ink Dye Detection

The following sections provide the procedure implemented for detecting multiply-
charged dyes by MALDI MS. A matrix and additive are both required for enhanced
detection of the dye which will be demonstrated. The method was applied to inks that
contained dyes that were not detected by LDMS and the analyses were performed from
paper. The focus of this research was to find a general MS method that can be applied to
the detection of multiply-charged ink dyes directly from paper. Matrix-assisted laser
desorption/ionization mass spectrometry (MALDI MS) has been used previously for the
detection of polysulfonated azo dyesl, and we have found that using matrix-assisted laser
desorption mass spectrometry (MALDI MS) can be very beneficial when trying to detect
intact multiply-charged dyes as previously discussed. We have adapted a method used by
Sullivan, et al. 1 for the analysis of polyionic dyes on a paper substrate. The MALDI

matrix, 2-(4-hydroxyphenylazo)benzoic acid (HABA), is well suited for the analysis of

100

multiply-charged dyes from ink on paper and diammonium hydrogen citrate (DAHC) is
used to further enhance dye detection.
Cation-Exchange Additives and Matrices

Sources were considered for eliminating or reducing the presence of sodium ions
which inhibit the desorption/ionization of the multiply-charged dye molecules. DAHC,
other amine bases, and cation exchange resins have been used by mass spectrometrists to
detect and enhance the mass spectra of multiply charged speciesl'4'20. The detection of
multiply-charged dyes first began by using cation-exchange resins. Two types of
resinous beads (Bio-Rad AG 50W) were used, one type contained exchangeable protons
and the other contained exchangeable ammonium ions. The beads as a solution were
mixed with a dilute, aqueous dye solution. The slurry was filtered through a column and
the filtrate was collected and subsequently analyzed by MALDI MS. The mass Spectra of
the analyzed dyes, Direct Red 23 and Acid Blue 93, however, showed only a slight
increase in sensitivity from using cation-exchange resins. Resins work by displacing the
sodium ions from the dye salt. The resins appeared to be ineffective for the removal of
the sodium ions. The order of cation relative selectivity for AG 50W-X8 resin is NHII+ >
Na+ > H1, Na+ Should be displaced from the analyte by H+ since Na" has a higher affinity
for the resin than the dyes. The poor results attained with the resins lead to examining
additives that have been used to isolate the alkali ions from the analyte. When the matrix,
a—CHCA (contained 0.1% TFA), was introduced to the sample the sensitivity increased
significantly for the dye samples exposed to the resinous bead that contained NHIt+
exchange resin. 2,5-DHB was also used as a matrix, but the detection of the dyes was not

enhanced. However, there was concern that a-CHCA (with 0.1% TFA) alone was

101

involved in the detection of the dyes and had nothing to do with the cation-exchange
resin. At this point, the LDMS experiment turned into an MALDI MS experiment by the
introduction of a matrix to the sample.

In the MALDI experiment, picomole amounts of analyte are mixed with an excess
of organic matrix molecules which facilitate the absorption of the UV laser light at 337
nm (the wavelength of light emitted from a pulsed N2 laser). A few microliters of an
analyte/matrix solution are deposited on a metal sample plate. Upon rapid solvent
evaporation, crystals are formed containing the analyte at a low concentration. When the
laser is fired the matrix absorbs the UV radiation assisting in the desorption/ionization of
the analyte. Both matrix and analyte ions are detected.

Various MALDI matrices were considered to enhance the detection of the
multiply-charged commercial dyes. Matrices that were evaluated included alpha-cyano-
4-hydroxycinnamic acid (a-CHCA), 2-5-dihydroxybenzoic acid (2,5-DHB), 3-
hydroxypicolinic acid (3-HPA), Sinapic acid (SA), 5-methoxysalicylic acid (MSA), and
6-aza-2-thiothymine (ATT). The concentration of the dye solutions was varied to
achieve the best sensitivity. Dye concentrations of 5, 10, 100, and 10,000 pmol/uL were
examined. Initial results indicated that analyzing10,000 pmol/uL dye solution with a-
CHCA on the SS plate provided the best positive and negative-ion MALDI mass Spectra,
however, peaks in the mass Spectra were poorly resolved. Additives or co-matrices have
been known to enhance (peak resolution and sensitivity both increased) analyte detection
in MALDI MS.

In order to increase resolution, additives were examined with the matrices

previously examined. Additives or co-matrices have been used previously and are known

102

to enhance the performance of the MALDI experiment4'20. Enhancements include
increased analyte sensitivity, improved peak resolution (increased mass accuracy),
reduced analyte fragmentation, and increased sample homogeneity. The basis of
introducing an additive is to enhance the performance of the MALDI MS experiment by
eliminating or reducing the effects of cations such as sodium or potassium ions. In
general, polyanionic dyes are manufactured in the form of sodium salts; therefore, there
is a need to suppress the influence of the sodium ions by using an additive. In this
research, the additives that were examined included ammonium acetate (AA),
diammonium hydrogen citrate (DAHC), and trifluoroacetic acid (TFA).

The common procedure for preparing a solution a—CHCA requires a final solution
concentration of 0.1% TFA to increase sample solubility”. An increase in sensitivity and
resolution was observed in the mass spectra of Direct Red 23 and Acid Blue 93 when
TFA was used with a-CHCA. When the concentration of the dye solutions were reduced
from 10,000 pmoVuL to 100 pmol/uL the sensitivity increased once again. AA and
DAHC have been used previously as additives for reducing the charge state of the
phosphate groups on phosphorylated peptides by displacing alkali cations". When an
aqueous solution of 1 mM DAHC was used with a-CHCA instead of TFA peaks due to
sodium adduction were reduced or eliminated providing a less complicated mass
spectrum, hence, chemical noise was reduced. A disadvantage to using a—CHCA is that
the matrix yields several peaks in the mass spectrum which may interfere with the
detection of the dyes. Matrices are typically used with high molecular weight
biomolecules, so the peaks that correspond to the matrix are not a concern in those

experiments. However, dyes have lower molecular weights, typically, so a matrix that

103

yields substantially less peaks was considered for these MS experiments. The MALDI
matrix 2-(4-hydroxyphenylazo)-benzoic acid (HABA) with DAHC is well suited for this
application and was used as a general method for detecting multiply-charged dyes in this
research. The matrix alone does not sufficiently enhance dye detection.
Commercial lives

In this research, over 30 multiply-charged dyes (100 pmol/uL was used for all
dye concentrations) including the dyes (Direct Red 23, Acid Blue 93, Direct Red 80,
Acid Yellow 23, Solvent Blue 38, Acid Black 1, and Reactive Black 5) that were initially
examined by direct LDMS were further analyzed by using HABA (24 mg/mL in 2:2:1
ACN:H20:Me0H) and DAHC (100 mM) from the SS plate. The dyes undetected by
LDMS were detected once the matrix and additive were applied, except for Reactive
Black 5. However, a negative fragment ion was detected at m/z 742 and corresponds to
the loss of two sulfonic acid groups from the dye. The intact dye was expected to give
rise to a negative ion with an m/z 902. This peak and the peak observed in the MALDI
mass spectrum have a mass difference of 160 amu. Most likely during the D/I process,
the dye lost two sulfonic acid groups. The structure of Reactive Black 5 contains two
functional groups (—02SCH2CH2OSO3' Na") that the other dyes to not have which may
prevent the intact desorption of the dye. Listed in Table 4.1 are the multiply-charged
dyes examined in this research by using HABA and DAHC. The m/z values for each dye
detected experimentally are listed as well as functional groups that may contribute to the
overall charge on the dye molecule in the table. Functional groups that appear to
influence the detection of the dye in LDMS include sodium sulfonic acid, sodium

alkoxide, and sodium carboxylate. However, the pH of the ink may not necessarily

104

influence the desorption/ionization process. Hydroxyl goups may also contribute to the
charge and most likely depends on the pH of the solution. Some MALDI matrices, such
as 2,5-DHB, have multiple hydroxyl groups which do not prohibit detection in LD.
Additional studies should be carried out to understand the role of hydroxyl groups on the
detection of dyes in LDMS as well as MALDI MS. Although, we have successfully
detected several polyionic dyes from the metal plate, the ultimate challenge was to detect

the dyes in ink from paper.

105

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Cl Name (+) mlz (-) mlz SO3Na cozua ONa OH imlnium
Acid Black 1 571 2 0 0 1 0
Acid Black 24 688 686 2 0 0 0 0
Acid Black 26 668 666 2 0 0 1 0
Acid Blue 1 545 543 2 0 0 0 1
Acid Blue 7 669 667 2 0 0 0 1
Acid Blue 9 749 747 3 0 0 0 1
Acid Blue 90 832 830 2 0 0 0 1
Acid Blue 93 756 754 3 0 0 0 1
Acid Green 3 669 667 2 0 0 0 1
Acid Green 16 595 593 2 0 0 0 1
Acid Green 25 579 577 2 0 0 0 0
Acid Red 14 457 2 0 0 1 0
Acid Red 18 537 3 0 0 1 0
Acid Red 37 479 2 0 0 1 0
Acid Red 52 559 557 2 0 0 0 1
Acid Red 87 649 647 0 0 2 0 0
Acid Red 92 785 0 1 1 0 0
Acid Violet 49 712 710 2 0 0 0 1
Acid Yellow 1 313 1 0 0 1 0
Acid Yellow 17 505 2 0 0 1 0
Acid Yellow 42 713 2 0 0 2 0
Direct Black 19 796 794 2 0 0 1 0
Direct Black 22 1016 3 0 0 2 0
Direct Black 38 738 736 2 0 0 1 0
Direct Black 51 556 1 1 0 3 0
Direct Blue 71 942 940 4 0 0 1 0
Direct Blue 106 697 695 2 0 0 0 0
Direct Red 1 582 1 1 0 2 0
Direct Red 4 841 3 0 0 1 0
Direct Red 23 768 2 0 0 2 0
Direct Red 28 653 651 2 0 0 0 0
Direct Red 75 901 4 0 0 2 0
Direct Red 80 1239 6 0 0 2 0
Direct Violet 51 676 674 2 0 0 1 0
Direct Yellow 4 579 2 0 0 2 0
Food Yellow 3 409 407 2 0 0 1 0
Reactive Black 5 742 4 0 0 1 0

 

Table 4.1: Polyionic dyes examined by MALDI MS (HABA and DAHC) on SS plate.
Provides m/z values detected in the positive and/or negative-ion mode and functional

groups that may contribute to the overall charge on the dye molecule

106

 

Highly-Charged Dyes

An important aspect of the technique of using HABA and DAHC that Should be
noted is that the method does not appear to be limited based on the number of charges
that a dye contains. Most of the colorants that can be used in ink are polysulfonated azo
dyes. In this research, over 30 polyionic dyes have been detected successfully directly
from paper using HABA and DAHC. The dyes analyzed by the MALDI method were
selected from the ink patents and they vary in Size or molecular weight and overall
charge, but all contain multiple sulfonic acid groups. Dyes containing 2 to 6 sulfonic
acid groups have been detected by MALDI MS. We are most certain that dyes
containing more than 6 sulfonic acid groups can be detected as well. Investigating dyes
that have more than Six sulfonic acid groups would be intriguing. AS the number of
sulfate groups increase the task of detecting the dye becomes more challenging in ESI”23
and would be interesting to see if the same phenomena occurs by MALDI mass
spectrometry. Mechref et al. noticed this phenomenon when trying to detect
oligosaccharidess. Preliminary results suggested that multiple sulfonic acid groups
influence dye detection in MALDI. One particular dye, Direct Red 80, shown in Figure
4.2b, contains six sulfonic acid groups, each contributing -1 to the overall charge on the
dye molecule. When the dye was analyzed by MALDI MS the dye was detected at mlz
1239 (Figure 4.3), but the sensitivity was low compared to most of the other dyes that
contained fewer multiple sulfonic acid groups. The reduction in sensitivity may be
associated with the number of sulfonic acid groups present. However, additional dyes
need to be studied to be able to understand the effect the number of charges has on the

sensitivity of the experiment. Maintaining isotopic resolution during the analysis of

107

Direct Red 80 was more difficult than for the other dyes, however, by attenuating the

MALDI TOF parameters isotopic resolution can be obtained.

 

 

 

 

 

100

I?

0)

g.

r_:

0

.2

E

g 1239

_I__IL. 1.21 - . 11.591229

 

 

0 500 1000 1500
mlz

Figure 4.3: Negative-ion MALDI MS analysis of Direct Red 80 (HABA and DAHC) on
SS plate

Ink on Paper

We focused on using the matrices a—CHCA and HABA with DAHC for analyzing
dyes directly from paper since the two combinations yielded the best results for analyses
from the SS plate. Growing matrix crystals of a-CHCA on paper was difficult and dye
detection was insufficient. Fortunately, we had great success with HABA. The
procedure that was used involved, first, adding a droplet (1 uL) of DAHC to the dye
sample on paper, followed by adding increments of the matrix until the appearance of

orange crystals were visible on the surface of the dye-paper sample. Typically, three 1

108

[1L aliquots of HABA were used. The soluble dyes are extracted into the aqueous DAHC
droplet. The matrix was not added until the sample was dried. Although, the ink sample
frequently spreads outside the width of a written line when the additive and matrix
solutions are added to the paper, there are no peaks in the positive and negative-ion
MALDI mass spectra that arise due to the paper. Paper contains components that could
be detected, however, interference from the paper is not a problem since the content of
the dye in the ink outweigh the paper components. DAHC and HABA were also
examined separately for the detection of dyes on paper and found the reagents must be
used simultaneously for the dyes to be sufficiently detected. First, several commercial
dyes were examined and then the method was applied to inks. When LDMS failed to
detect dyes in an ink, the sample was subjected to analysis by MALDI MS using the

method described above.

Dye detection using a matrix and additive yields molecular information, such as
the molecular mass of the dye, ultimately, leading to the identification of the dye. The
importance of using a matrix and an additive in conjunction for adequate detection will
be demonstrated through the examination of the liquid, blue ink from the Pilot® Precise
V7 rollerball pen that was previously examined by direct LDMS from paper. Figures 4.4
and 4.5 show the positive and negative-ion LD mass spectra, respectively, of the blue ink

using HABA (a), DAHC (b), and HABA and DAHC (c).

109

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a)
832 854
876

E
i
g 734
a? 712\‘56

0 500 1000 1500 2WD
13)
i
5 712
5
5
IE

0 500 1000 1500 2000
c) 712
i
5
it
E 698

\ 832
l
542
.2111 I- ,1 -...I I . ,
0 500 1000 1500 2WD

mlz

Figure 4.4: Positive-ion LD mass spectra of blue, liquid ink from Pilot® Precise V7
rollerball pen on paper a) HABA, b) DAHC, and c) HABA and DAHC

110

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a)
E:
2
E 652
g 830
i \
a. 762
V
710\
I 11 II
0 500 1000 1500 2000
b) 191
170 Z
.3:
I
.5
E
3;
E
8
III g 1
0 500 1000 1500 2000
C)
830
E
2
g 710
g 852
696
“ 880
I II. III I ‘ I .
0 500 1000 1500 2000
mlz

Figure 4.5: Negative-ion LD mass spectra of blue, liquid ink from a Pilot® Precise V7
rollerball pen on paper a) HABA, b) DAHC, and c) HABA and DAHC

111

The multiply-charged dyes contained in the liquid, blue ink were identified and
will be discussed later, however, the m/z values of the dyes are significant for this
discussion. In order to understand the importance of using the matrix and additive
Simultaneously, the mlz value of the intact dyes will be given. Two multiply-charged
dyes are contained in the ink. The molecular ions of the first dye have mlz values 712
and 710 in the positive and negative-ion modes, respectively. The positive and negative
molecular ions of the second dye are denoted by m/z 832 and 830, respectively. Using
HABA alone did detect both dyes, however, the positive-ion mass Spectrum contains
several peaks, most of which represent sodium adducts (m/z 734, 756, 854, and 876).
The presence of these peaks makes identifying the dyes unlikely. When this mass
spectrum is compared to the mass Spectrum of the experiment that utilized both reagents
the peaks due to adducts are absent and peaks that represent the intact dyes are easily
identified at m/z 712 and 832. Also, the resolution of the peaks greatly increases when
the combination of matrix and additive compared to the using them alone. Isotopic
resolution was not observed in the positive and negative-ion mass spectra of the inks
analyzed with HABA alone. Accurately, determining the mass of the molecular ion is
difficult with poor peak resolution. Although, the use of DAHC alone yielded isotopic
resolution, the dyes were not able to be detected in the negative-ion mode. The use of
both reagents generated mass spectra with isotopically resolved peaks, enhanced
sensitivity, and reduced chemical noise. Additional peaks (m/z 698 and 542 in the
positive-ion and m/z 696 in the negative-ion) in the MALDI (with DAHC) mass

Spectrum are associated with the dyes.

112

Figure 4.6 demonstrates how using a matrix and additive for the detection of a
polyionic dye, Acid Violet 49, directly from the SS plate. Acid Violet 49, shown in
Figure 4.7, contains two sulfonic acid groups and one iminium group and has a molecular
weight of 734 Da. The loss of the sodium results in the detection of the dye and can be
seen in the positive and negative-ion MALDI mass spectra. The replacement of the
sodium ion for a proton (protons come from the ammonium ions from DAHC) enables
the dye to be detected as shown in Figure 4.6a, the positive-ion MALDI mass spectrum.
The D/I of the dye in positive ion mode has a m/z value of 712. The loss of the sodium
with no substitution occurs as well and the dye can be detected in the negative-ion mass
spectrum shown in Figure 4.6b. The nature of the charge state allows the dye to be

detected in both ion modes.

113

 

 

 

 

 

 

 

 

 

a) 243
71 2
225 372
3581 542 693
i L l l
b) 241
1 91
71 0
“fl
0 200 400 600 800
mlz

1 000

Figure 4.6: MALDI MS analysis of Acid Violet 49 on SS plate a) positive-ion mass
spectrum and b) negative-ion mass spectrum

114

H3C
\N VIP—CH2

I
CH CH
CH3 2 3 SO3Na

Figure 4.7: Dye structure of Acid Violet 49

As a side note, the peaks with m/z value 243 and 241 in the positive (Figure 4.6a)
and negative (Figure 4.6b) ion mass spectra, respectively, correspond to the HABA
matrix. HABA is a solid, neutral, organic molecule and has a molecular weight of 242.
The matrix molecules are protonated (gains a hydrogen ion) or deprotonated (looses a
hydrogen ion) during the desorption/ionization process and is observed in the positive
and negative-ion mass spectra, respectively. As previously stated, the matrix is in excess
compared to the analyte in the MALDI MS experiment, so the peaks that correspond to

the matrix are more abundant than the analyte.

Detecting dyes in a written line directly from a paper substrate is important in the
field of question document examination, however, paper is not an ideal sample for the
MALDI experiment. MALDI requires the incorporation of the analyte into the matrix
crystals. Growing crystals on paper in order to detect the dyes has been a challenge. We

have found that HABA with DAHC works well for detecting multiply-charged dyes from

115

paper. Figure 4.8, demonstrates the success of using HABA and DAHC for the detection
of Acid Violet 49 on paper. The mass spectra of the dye on paper and SS are similar with
the exception that the peaks that pertain to the dye are more intense in the second set of
spectra. Commercial dyes are frequently sold as impure mixtures24 and this is such the
case with Methyl Violet 2B purchased from the the Aldrich Chemical Company as well
as Acid Violet 49. The mass spectra of Acid Violet 49 contain additional peaks at m/z
value 542 and 372 that correspond to dyes that are similar in structure to Acid Violet 49.
These dyes were probably a result of the incomplete synthesis of Acid Violet 49.
Actually, the peak at m/z value 372 corresponds to Crystal Violet and the peak at mlz
value 542 represents a compound that has a molecular formula C32H35N3033 and is
similar to the dye structure in Figure 4.7. The alkyl groups bonded to the nitrogen could
be either methyl or ethyl groups, and light, ultraviolet25 or incandescent”, induced dye
degradation (oxidative dealkylation) can be used to determine which functional groups
are present. The three sets of peaks that are present in Figure 4.8a. The subset of peaks
are separated by 14 mass units and have been associated with oxidative demethylation25 .
These peaks are characteristics of the dye and these clues can be useful when identifying

an unknown dye.

116

 

 

 

 

 

 

 

 

 

6)
372
358 712
542 698
l 1 .l .52?
b)
710
696
O 200 400 600 800 1 000
mlz

Figure 4.8: MALDI MS analysis of Acid Violet 49 on paper a) positive—ion mass
spectrum and b) negative-ion mass spectrum

The Pilot® Precise V7 rollerball pen containing liquid, blue ink was selected to be
analyzed by MALDI MS since LDMS was ineffective for detecting the dyes present in
the ink. LDMS results were shown in Figure 4.1. The TLC analysis indicated that two
dyes (one blue band and one purple band) were present in the ink. Shown in Figure 4.9
are the MALDI mass spectra of the liquid, blue ink on paper. The MALDI analysis of
the ink on paper successfully detected two dyes. The mass spectra contain peaks that

appear to be associated to dye structures that contain both positive and negative charges.

117

The peaks at mlz values 712 and 710 in the positive (Figure 4.9a) and negative (Figure
4%) ion mass spectra, respectively, correspond to Acid Violet 49. The peaks at m/z
values 832 and 830 in the positive and negative ion mass spectra, respectively, have been
identified as corresponding to Acid Blue 90, shown in Figure 4.10. MALDI analysis of
Acid Blue 90 was carried out directly to ensure the correct peak assignment. Figure 4.11
shows the positive (a) and negative (b) ion MALDI mass spectra of Acid Blue 90 and
confirms the dye association. According to the mass spectra (Figure 4.9),, another dye
might be contained in ink. The peaks at m/z value 882 and 880 in the positive and
negative-ion mass spectra respectively, may correspond to a third dye that has a similar

structure to the other two dyes, but no correlation has been made currently.

118

 

 

 

 

 

 

 

 

 

 

a) 712
832
542 698
l 1 1 5291 Z34 882
b)
830
710
862
696 E
1 $810
. l, . - f 1
O 200 400 600 800 1 000
mlz

Figure 4.9: MALDI MS analysis of liquid, blue ink from Pilot® Precise V7 rollerball pen
on paper a) positive-ion mass spectrum and b) negative-ion mass spectrum

119

H3CH zc-OQ

121

 

 

Figure 4.10: Dye structure of Acid Blue 90

120

 

 

 

 

 

 

 

 

 

 

 

 

 

a)

E

m

C

.2

E

.‘z’ 832
£11

a

a:

662
s . l . l .
0 200 400 600 800 1 000
mlz

b)

E

a

c

8

E

.‘z‘

E

8 830

0 200 400 600 800 1000
mlz

Figure 4.11: MALDI MS analysis of Acid Blue 90 on SS a) positive-ion mass spectrum
and b) negative-ion mass spectrum

121

Multiply-Charged Ink Dye Characterization

The third part of this chapter encompasses the basic characterization information
that can be acquired from MALDI and LD mass spectra and thin-layer chromatograms of
multiply-charged ink dyes. The previously two chapters described the characterization
information that can be acquired from examining ink dyes (neutral and singly-charged
dye molecules) directly and indirectly (photodegradation of the sample prior to the MS
analysis) by LDMS. MALDI and LD are related ionization techniques, hence, the
information that can be deduced fi'om both types of experiments is similar. Molecular
information such as mass, composition, and charge state (the reduced charge state in
MALDI) can be obtained from MALDI mass spectra in the same manner as LD. The
information can be pieced together to identify and characterize the species. Although,
MALDI is a softer ionization technique than LD, some fragmentation does occur. The
formation of clusters also occurs in MALDI as seen in LD. As discussed previously for
LDMS, fragment ions, cluster ions, and dye impurities can be used to characterize ink
dyes. These species can be used in to make associations as well from the MALDI
experiment. The charge state of a dye molecule, prior to the MS analysis, is more
complicated to characterize by MALDI than by LD. In MALDI, the charge state of the
dye is reduced by using a cation-exchange additive and a matrix prior to the analysis in
order to be able to detect the dye. The multiply-charged dyes are detected as singly-
charged species in the MALDI experiment. Acquiring information from the mass
spectrum is limited; however, clues do present themselves which can be used to
determine the presence of a multiply-charged dye, specifically, a polysulfonated dye. A

dye catalogue has been assembled from ink patents for this thesis research which includes

122

dye structures, molecular compositions, charge states, reduced charged states, mlz values,
and molecular masses. The catalogue is used to establish the identities of unknown dyes
using the information that is attained from the corresponding MALDI mass spectra.
Identifying and characterizing multiply-charged dyes from ink will be demonstrated
through the examination of two inks from Sanford® Uni-ball® Roller pens by MALDI

MS and TLC.
MALDI MS and TLC

Inks are initially examined by MALDI MS from the SS plate and by TLC to
acquire preliminary results. Once dyes have been identified, they are further analyzed
from paper. This procedure limits the amount of sample preparation allowing for rapid
results and reduces the risk of analyzing ink dyes that are not contained in the catalogue
of ink dyes developed for this research. Two pen ink examples will be discussed to
demonstrate the use of MALDI MS as well as using TLC indirectly to help identify
multiply-charged dyes. The inks examined were Sanford® Uni-ball® Roller pens that

contained liquid, blue and red inks.

MALDI MS analysis of ink from the SS plate is a rapid method for acquiring
preliminary dye information. Like all of the other MS analyses carried out for this
research, a positive and negative-ion mass spectrum is obtained for each ink sample.
Figure 4.12 shows the MALDI mass spectra of the liquid, blue ink from a Sanford® Uni-
ball® Roller pen. Both spectra contain two relatively intense peaks that are separated by
two atomic mass units; hence, the ink appears to contain two dyes. The thin—layer
chromatogram also suggested that the ink contained two dyes, one blue and the other red.

One might assume that both dyes are neutral molecules; however, the dyes were not

123

easily detected in LDMS. The dye that is represented by the peaks at m/z 559 and 557 in
the positive and negative-ion MALDI mass spectra, respectively, appears to contain an
ethylated amine group. The detection of dyes containing alkylated amines such as
Crystal Violet and the rhodamine dyes by LDMS were previously discussed. Like the
LD mass spectra of Rhodamine B and Rhodamine 66, the MALDI mass spectra of
Sanford® Uni—ball® Roller pen show a peak that is 28 atomic mass units lower than that
of the peak of the intact dye. One can assume that the particular dye that is associated
with m/z 557 and 559 contains an irninium group or an amine group with at least one
ethyl group bonded to the nitrogen. The difference between the mass spectra of this dye
and the spectra of Crystal Violet and the rhodamine dyes is that this dye is detected in
both D/I modes. Crystal Violet, Rhodamine B, and Rhodamine 66 are cationic dyes that
are not detected in the negative-ion mode. Also, this dye was not easily detected in

LDMS, unlike the cationic dyes, which suggests that the dye has multiple charges.

124

 

 

 

 

 

 

 

 

 

 

 

 

 

8)
559
E
(n
.5
E
g
E
0
a:
747 1117
1 531 L L
. 1 . . -
o 300 600 900 1200

mlz

b)
745
E
U)
.5:
E
g 557
E
Q
a:
A l . 529 719 we
0 200 400 600 800 1000 1200
mlz

Figure 4.12: MALDI MS analysis of liquid, blue ink from a Sanford® Uni-ball® Roller
pen on SS plate) positive-ion mass spectrum and b) negative-ion mass spectrum

125

One might mistake the peak at m/z 1117, in the positive-ion MALDI mass
spectrum, to be another dye present in the ink; however, this peak a dimer. The cluster
ion is comprised of two dye molecules plus a proton [2 (558) + 1]. The dye that is
represented by the peaks at m/z 747 and 745 in the positive and negative-ion MALDI
mass spectra, respectively, also appears to be a multiply charged dye that contains an
ethylated nitrogen atom. The peak at m/z 769 in the negative-ion mass spectrum is the
result of cation adduction by a sodium ion. Generally, sodium ion adduction for dye
analysis suggests the presence of a multiply-charged dye. The loss of an ethyl group
from the dye was observed in the negative-ion mass spectrum as well. Coupling TLC
indirectly to MALDI MS determined that the red TLC band corresponded to mlz 559 and
557 and the blue TLC band was related to Hill 747 and 745. The MS and TLC data were
used to search the dye catalogue and the two dyes were tentatively identified as
Sulforhodamine B (Acid Red 52) and Acid Blue 9. The dye structures of the red and blue
dyes (the Na+ counterions were not shown) are illustrated in Figures 4.13a and 4.13b,
respectively. The dyes were purchased and analyzed directly to confirm that the correct
identifications were made. Figures 4.14 and 4.15 show the MALDI mass spectra of the
ink dyes directly. The dyes were accurately identified. Multiply-charged dyes such as
Acid Red 52 and Acid Blue 9 which contain both positive and negative charges are
detected in positive and negative-ion MALDI MS. Polyionic dyes containing strictly
negative charges are detected only in negative-ion MALDI MS, except if the dye is
multiply-charged exclusively with sodium alkoxide groups. The means ionization of
multiply-detection of polyionic dyes in MALDI MS can be used to help characterize and

identify the dye.

126

a)
so;
\

H CH C ,CH CH
3 2 \T 0 'ir 2 3
CHZCH3 H3Cl-lzc
CHZCH3

b)
"LC“:
803'
803'
N—CH2
cHam's

Figure 4.13: Dye structures a) Acid Red 52 and b) Acid Blue 9

127

Relative Intensity

b)

Relative Intensity

 

 

559

 

 

 

 

 

 

 

 

 

 

1 1 17
531 l
I l 1
300 600 900 1200 1500
mlz
557
529
200 400 600 800 1000
mlz

Figure 4.14: MALDI MS analysis of Acid Red 52 on SS plate a) positive-ion mass
spectrum and b) negative—ion mass spectrum

128

Relative Intensity

b)

Relative Intensity

 

 

 

 

 

 

 

 

 

 

 

747
1 771
200 400 600 800 1 000
mlz
745
‘llk l 1 A _L I
200 400 600 800 1 000
mlz

Figure 4.15: MALDI MS analysis of Acid Blue 9 on SS plate a) positive-ion mass
spectrum and b) negative-ion mass spectrum

129

A liquid, red ink was also examined by MALDI MS and the mass spectra
obtained can be seen in Figure 4.16. The mass spectra are quite different that the
previous blue ink examined. There are several peaks in the negative-mass spectrum
(Figure 4.16b) and only one predominant peak in the positive-ion mass spectrum (Figure
4.16a). From an initial examination of the negative-ion mass spectrum, there appears to
be at least five dyes that are contained in the ink. However, TLC ink analysis isolated
three dyes (yellow, red, red-orange TLC bands). TLC was indirectly coupled to MALDI
MS and Figure 3.17 shows the resulting mass spectra. The peak at m/z 407 represents a
yellow dye, the peak at 647 denotes the detection of an orange-red dye, and the
combination of peaks at m/z 741 and 785 correspond to a red dye. Again using the dye
catalogue the dyes were determined to be Food Yellow 3, Acid Red 87, and Acid Red 92
as shown in Figure 4.18a, 4.18b, and 4.18c, respectively. The three dyes are polyanionic;
however, the functional groups responsible for the charge distribution are different
among the dyes. Sodium sulfonic acid, alkoxide, and carboxylate groups are responsible
for the charge state of the dyes. Food Yellow 3 is a disulfonated azo dye.. The peak at
m/z 429 is due to the sodium adduction of the dye and demonstrates that the dye is
multiply-charged. Food Yellow 3 was detected exclusively in the negative-ion mode as
expected since the dye does not contain any positive charges. Acid Red 87 contains two
alkoxide groups. Acid Red 92 contains one alkoxide group and one carboxylate group.
Neither of these red dyes contains functional groups that carry a positive charge,
however, they were detected in positive-ion MALDI MS. Perhaps the alkoxide and
carboxylate groups can acquire a positive charge upon D/I. where a dye containing

strictly sulfonic acid groups will only be detected in the negative-ion mode. In negative-

130

ion LDMS, multiple sulfonic acid groups completely inhibit the DH of Food Yellow 3
where Acid Red 87 and Acid Red 92 are weakly detected. Acid Red 92 is not detected
intact. The dye fragments as a result of containing carboxylic acid and is similar to the
fragmentation that results when Rhodamine B is examined by LDMS. The decarboxylic
acid form of the dye is detected at m/z 741. Even though the laser power was
significantly increased for detection in LDMS as compared to the MALDI experiment,
the number of ions detected in the LDMS experiment was considerably less. For a
typical MS experiment the ion count is about 12,000 and the ion count for this particular
LD experiment was less than 800. MALDI MS significantly enhanced the detection of
the Food Yellow 3, Acid Red 87, and Acid Red 92. The LD mass spectrum of the red,
liquid ink is shown in Figure 4.19. Peaks of low intensity encompass the m/z range of
the ink dyes making it difficult to distinguish the dyes from other chemical noise. Some
information can be acquired from the negative-ion MALDI mass spectrum to characterize
the ink dyes. The acid red dyes give rise to distinct isotope distributions as seen in
Figures 4.20 and 4.21. Isotope distributions have been discussed for Copper
Phthalocyanine in Chapter Two. The theoretical isotopes of Acid Red 87 and Acid Red
92 were compared with experimental isotopes from the mass spectra to characterize and
identify the ink dyes. Another characteristic of Acid Red 92 that can be used to
characterize the dye is that the dye contains a carboxylic acid group. As shown
previously, Rhodamine B decarbonylated when the xanthene dye was analyzed by
positive-ion LDMS. The loss of C02 (44 amu) can be observed in the negative-ion

MALDI mass spectrum of Acid Red 92.

131

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6)
E
2
2
S.
o
.2
g
o
a:
649
.ixiiu..-1Li 1..le -1. A A £70
0 200 400 600 800 1000
mlz

b) 407

E

on

5

5

5

s

g 429

647
741 837
785 1
J- L . L 1 i L 1' 1 .
0 200 400 600 800 1000
mlz

Figure 4.16: MALDI MS analysis of red, liquid ink from Sanford® Uni-ball® Roller pen
on SS plate a) positive-ion mass spectrum and b) negative-ion mass spectrum

132

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a) 407
5
5
3c:
3
§
I
. J , . 11 1

o 200 400 600 800 1000
b)
E
5 741
s \ 785
3
E
a /

J ll . 11 1 i

o 200 400 600 800 1000
c) 647
b
i
8
"5'
2

o 260 400 600 860 1000

mlz

Figure 4.17: Indirect coupling of TLC with MALDI MS: negative-ion mass spectra of
extracted TLC bands a) yellow band, b) orange-red band, and c) red band

133

 

Figure 4.18: Dye structures a) Food Yellow 3, b) Acid Red 87, and c) Acid Red 92

134

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6)
E
0
C
2
E
..'>.’.
E
o
a:
1.11 ml“; .1 ”-1111
0 200 400 600 800 1000
mlz
b)
647
741
E
2
2
5
E
a
3'
a:
600 650 700 750 800
mlz

Figure 4.19: Direct negative-ion LDMS analysis of liquid, red ink from Sanford® Uni-
ball® Roller pen a) complete spectrum from m/z O — 100 and b) expanded spectrum from

m/z600-800

135

Relative Intensity

1))

Relative intensity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

647
643 W W 651
630 640 650 660
mlz
, I. I I H- ,
630 640 650 660
mlz

Figure 4.20: Isotopic peak distributions of Acid Red 87 a) experimental and b)

theoretical

136

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8‘) 785
g 783
E 787
0
>
3 l
O
C
781 w 789
779 ’1 1 MA 791

770 780 790 800

mlz
b)
E
a
.5
E
.3
g
0
I

12. I 1 I I ll 1 .
770 780 790 800
mlz

Figure 4.21: Isotopic peak distributions of Acid Red 92 a) experimental and b)

theoretical

137

Direct LDMS Analysis

The peak at m/z 170 in the negative-ion LD mass spectrum (Figure 4. lb) of the
liquid, blue ink was classified as corresponding to a fragment ion from Acid Violet 49
and Acid Blue 90. The fragmentation of the structures is illustrated in Figures 4.7 and
4.10. The fragment ion corresponds to [C7H603S]". The dissociation of the ON bond
that lead to the fragment ion is depicted in the dye structures. The dyes were analyzed
directly and the negative-ion mass spectra showed the presence of the fragment ion.
Without having previous knowledge of the dye structures one would be able to use this
fragment ion observed in the LD experiment to assume that the ink contains a dye with
the specific functional group (-CH2C6H4803' Na+). Fragmentation demonstrated that
direct LD analyses of ink can be used to make such associations about the multiply-

charged dye structures.

138

References

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

Sullivan AG, Gaskell SJ. The Analysis of Polysulfonated Azo Dyestuffs by
Matrix-Assisted Laser Eesorption/Ionization and Electrospray Mass
Spectrometry. Rapid Commun Mass Spectrom. 1997 (l 1) 803-9

Conneely A, McClean S, Smyth WF, McMullan G. Study of the Mass
Spectrometric Behavior of Phthalocyanine and Azo Dyes Using Electrospray
Ionization and Matrix-Assisted Laser Desorption/Ionization. Rapid Cam in
Mass Spectrom. 2001 (15) 2076-2084

Osada T, inventors. Mitsubishi Pencil Kabushiki Kaisha, assignee. Aqueous Gel
Ink-filled Ball Point Pen. US Patent 5,993,098 Nov 1999

Asara JM, Allison J. Enhanced Detection of Phosphopeptides in Matrix-Assisted

Laser Desorption/Ionization Mass Spectrometry Using Ammonium Salts. J Am
Soc Mass Spectrom. 1998 1035-44

Mechref Y, Novotny MV. Matrix-Assisted Laser Desorption/Ionization Mass
Spectrometry of Acidic Glycoconjugates Facilitated by the Use of Spermine as a
Co-Matrix. J Am Soc Mass Spectrom. 1998 91293-1302

Nordhoff E, Ingendoh A, Cramer R, Overberg A, Stahl B, Karas M, Hillenkamp
F. Matrix—Assisted Laser Desorption/Ionization Mass Spectrometry of Nucleic
Acids with Wavelengths in the Ultraviolet and Infrared. Rapid Comm in Mass
Spectrom. 1992 (6) 771-776

Pieles U, Zurcher W, Schar M, Moser HE. Matrix-Assisted Laser Desorption
Ionization Time-of-Flight Mass Spectrometry: A Powerful Tool for the Mass and
Sequence Analysis of Natural and Modified Oligonucleotides. Nucleic Acids Res.

1993 (21) 3191-3196

Currie GJ, Yates JR III. Analysis of Oligodeoxynucleotides by Negative-Ion
Matrix-Assisted Laser Desorption Mass Spectrometry. J Am Soc Mass Spectom.
1993 (4) 995-963

Zhu YF, Chung CN, Taranenko NI, Allman SL, Martin SA, Haff L, Chen CH.
The Study of 2,3,4-Trihydroxyacetophenone and 2,4,6-Trihydroxyacetonphenone
as Matrices for DNA Detection in Matrix-Assisted Laser Desorption/Ionization
Time-of-Flight Mass Spectrometry. Rapid Comm Mass Sepctrom 1996 (10) 383-
388

139

(10)

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

Green-Chruch K, Simmons TA, Limbach PA. Chemistry Effects on the Analysis
of DNA by Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry. In
Ultrasensitive Biochemical Diagnostics II sponsored by SPIE-the International
Society for Optical Engineering. 1997 (2985) 88-99

Cheng S, Chan TWD. Use of Ammonium Halides as Co-Matrices for Matrix-
Assisted Laser Desorption/Ionization Studies of Oligonucleotides. Rapid Comm
Mass Spectrom. 1998 (10) 907-910

Zhu YF, Taranenko NI, Allman SL, Martin SA, Haff L, Chen CH. The Effect of
Ammonium Salt and Matrix in the Detection of DNA by Matrix-Assisted Laser

Desorption/Ionization Time-of-Flight Mass Spectrometry. Rapid Comm Mass
Spectrom. 1996 (10) 1591-1596

Simmons TA, Limbach PA. The use of a Co-Matrix for Improved Analysis of
Oligonucleotides by Matrix-Assisted Laser Desorption/Ionization Time-of—Flight
Mass Spectrometry. Rapid Comm Mass Spectrom. 1997 (1 1) 567-72

Simmons TA, Limbach PA. Influence of Co-Matrix Proton Affinity on
Oligonucleotide Ion Stability in Matrix-Assisted Laser Desorption/Ionization
Time-of-Flight Mass Spectrometry. J Am Soc Mass Spectrom. 1998 (9) 668-75

Leo YC, Cheng SW, Chan TWD. Evaluation of Ammonium Salts as Co-Matrices
for Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry of
Oligonucleotides. Rapid Comm Mass Spectrom. 1998 12993-8

Vandell VE, Lambach PA. Polyamine Co-Matrices for Matrix-Assisted Laser
Desorption/Ionization Mass Spectrometry of Oligonucleotides. Rapid Comm
Mass Spectrom. 1999 (13) 2014-2021

Distler AM, Allison J. 5-Methoxysalicylic Acid and Spermine: A New Matrix for
the Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry of
Oligonucleotides. J Am Soc Mass Spectrom. 2000 (12) 456-462

Distler AM, Allison J. Improved MALDI-MS Analysis of Oligonucleotides
through the Use of Fucose as a Matrix Additive. Anal Chem. 2001 (73) 5000-3

Distler AM, Allison J. Additives for the Stabilization of Double-Stranded DNA in
UV-MALDI MS. J Am Soc Mass Spectrom. 2002 (l3)1129-1137

Distler AM, Allison J. Stabilization of Duplex Oligonucleotides for Analysis by
Matrix-Assisted laser Desorption/Ionization Mass Spectrometry Using
Crystallographic Condensing AgentC (III) Hexamine. Anal Biochem 2003
(319)332-334.

Voyager Workstation User’s Guide version 5, PerSpective Biosystems, 1999

140

(22)

(23)

(24)

(25)

(26)

Ballantine J A, Games DE, Slater P. The Use of Amine Base to Enhance the
Sensitivity of Electrospray Mass Spectrometry Towards Complex Polysulfonated
Azo Dyes. Rapid Comm Mass Spectrom. 1995 (9) 1403-1410

N g LK, Lafontaine P, Brazeau L. Ballpoint Pen Inks: Characterization by Positive
and Negative Ion-Electrospray Ionization Mass Spectrometry for the Forensic
Examination of Writing Inks. J Forensic Sci. 2002 (47) 1238-1247

Venkataraman K. The Chemistry of Synthetic Dyes. Vol. 1, Academic Press Inc.,
New York, 1952

Grim DM, Siegel JA, Allison. Evaluation of Laser Desorption Mass Spectrometry
and UV accelerated aging of dyes on paper as tools for the evaluation of a
questioned document. J Forensic Sci. 2002 (47)]265-1273

Dunn JD, Siegel J A, Allison J. Photodegradation and Laser Desorption Mass

Spectrometry for the Characterization of Dyes Used in Red Pen Inks. J Forensic
Sci. 2003 (48) 652-657

141

Chapter Five: Conclusions

MALDI MS Limitations and Future Research

Limitations are apparent in analyzing ink dyes by MALDI MS. MALDI, with
and without an additive, was used to detect singly-charged and neutral dye molecules as
well. Preliminary results indicated that MALDI may suppress the ion signal for dyes that
are neutral molecules. Perhaps the matrix and the dye molecules compete for the DH
energy. Additional dyes should be tested to see the influence of the matrix on detection
of neutral molecules. Neither the positive nor negative ions of Pigment Red 3 were
detected in the MALDI experiment. The dye was detected in the LDMS experiment
directly from paper and metal substrates. Both MS techniques should be used when
examining ink dyes. The D/I of Copper Phthalocyanine, a neutral pigment dye, was also
seen to be influenced by the MALDI experiment. A significant increase in laser power
was required for a signal that was comparable to the detection of the neutral dye by
LDMS. Another limitation of using MALDI is that some dyes are not soluble in the
matrix or additive solution which ultimately inhibits the detection of the dye. The
solubility of the dye does not necessarily influence detection in MALDI. Some of the
black dyes were barely or not detected, although, most were soluble in the solutions.
These dyes should be further examined. Structural similarities among the dyes were
limited, so no conclusions could be drawn based on the structures of the molecules.
Several black ink from gel or liquid pen were undetectable by MALDI MS. Perhaps,
these inks contain carbon black. US. Patent describes the composition of a black gel

inkl. According to the patent, carbon black is used as the colorant.

142

Characterizing multiply-charged dyes using advanced methods such as
photodegradation which was previously used as a characterization method with LDMS
can be future research. Some of the multiply-charged dyes contain alkylated amine
groups and azo bonds which have been known to photodegrade or thermally degrade,
respectively. Once the dyes have been degraded, then MALDI MS can be used to
characterize the ink dyes based from the photoproducts. Analyzing multiply-charged ink
dyes from other substrates such as fibers can be investigated using the method presented
here can also be done in the future. Previously work by Balko showed that
polysulfonated dyes in fabric were undetectable by direct LDMSZ. Applying the method

to other forensic evidence would be a great advancement.

143

References

(1) Takahiro O, inventor. Mitsubishi Pencil Kabushiki Kaisha, assignee. US Patent
5,993,098 Dec 1997

(2) Balko LG. “Laser Desorption/Ionization Mass Spectrometric (LD/MS) Analysis
of Samples of Forensic Interest: Chemical Characterization of Colorant and
Dyes.” MS. Thesis. Michigan State University, 2002

144

 

 

 

 

. . . . .. s“
. . . .