|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
CHARACTERIZATION OF COMMONLY ENCOUNTERED EXPLOSIVES USING
HIGH-FIELD ASYMMETRIC WAVEFORM ION MOBILITY SPECTROMETRY
COUPLED WITH MASS SPECTROMETRY
JARED J. BOOCK
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
O 2007 Jared J. Boock
To my family and friends.
I wish to thank my research director, Dr. Richard A. Yost, for teaching me a great deal
about the fundamentals of mass spectrometry and providing guidance throughout my research. I
would also like to thank Dr. Matthew Pollard at the Dr. Herbert H. Hill laboratory at Washington
State University for providing two opportunities for me to conduct research on their instruments,
and for providing invaluable assistance and conferring with me on the science. I thank lon
Metrics, Inc., of San Diego, CA, for funding these trips. I would like to recognize the members
of my committee: Dr. David H. Powell and Dr. John R. Eyler. I would like to thank the US Air
Force Institute of Technology Civilian Institutions program for giving me this opportunity.
Thanks also go to my colleagues in the Yost group for their support.
Lastly, I would like to thank my parents and my friends, without whom I would have gone
insane several times during this year and a half.
TABLE OF CONTENTS
ACKNOWLEDGMENT S ............ _...... ._ ...............4....
LI ST OF T ABLE S ............_...... ._ ...............7....
LIST OF FIGURES .............. ...............8.....
LIST OF ABBREVIATIONS ............_ ..... ..__ ...............11...
AB S TRAC T ............._. .......... ..............._ 12...
1 INTRODUCTION AND INSTRUMENTATION ................ ...............13................
1.1 Introducti on ............... ... ... ...... ........... ............1
1.2 Atmospheric Pressure Chemical lonization................ .. ...............1
1.3 High-field Asymmetric-Waveform lon Mobility Spectrometry............... ..............1
1.4 Ion Optics............... ....... .. .............2
1.5 Quadrupole lon Trap Mass Spectrometry............... ..............2
1.6 Multidimensional Mass Spectrometry ................. ...............26........... ...
1.7 Thesis Overview .............. ...............26....
2 PROPERTIES AND CHARACTERIZATION OF EXPLO SIVES .............. ...............27
2. 1 Explosives that were Characterized ............ ......__ ...............27.
2.2 Classification of Explosives .............. ...............27....
2.3 Chemical Properties of Explosives ................. ...............28........... ...
2.4 Instrument Settings .............. .. ...............29...
2.5 Compensation Voltage (CV) Scans .............. ...............30....
2.6 Characterized Compounds ................. ...............3.. 1......... ...
2.6. 1 Nitroaromatics ................... ............... .................3 1..
220.127.116.11 (2,4,6 -) Trinitrotoluene (TNT)............... ...............31.
2.6.2 Nitramines ................ ........... ...............4
18.104.22.168 Cyclotrimethylene trinitramine (RDX) ............. ...............43.....
22.214.171.124 Cyclotetramethylene tetranitramine (HMX) .............. ....................4
2.6.3 Nitrate Esters ................... ...............50..
126.96.36.199 Nitroglycerin (NG) ....................... ...............50
188.8.131.52 Pentaerythritol tetranitrate (PETN) .............. ...............51....
184.108.40.206 Ammonium Nitrate Fuel Oil (ANFO)............... ...............53.
3 FAIMS/MS AND IMS/FAIMS/MS OF EXPLOSIVES .............. ...............57....
3.1 Instrum entation ............... .. ........_ .. ...............57...
3.1.1 Electrospray Ionization (ESI) Source ................. ...............57........... .
3.1.2 lon Mobility Spectrometer (IMS).. ......___ ......_._......_. ..............59
3.1.3 Orthogonal Dome FAIMS Cell .............. ...............59....
3.2 FAIM S Limitations............... ......... ...............6
3.3 Effects of Addition of Helium to the Carrier Gas. ................ .............. ........ .....63
3.4 High-resolution FAIMS (HRFAIMS) .............. ...............63....
3.4.1 Separation of Mixtures of Explosives ................. ...............64...............
3.4.2 Resolving Isomers .............. ...............68....
3.5 Dispersion Voltage ................. ... ...............69.
3.6 Effects of Adjustment of Plate Gap ............ ....._ ...............70
3.7 Relationship of DV to CV .............. ...............80....
3.8 Sensitivity ............... ...............8 1..
3.9 IMS/FAIMS of Explosives ............... ...............8 1...
4 CONCLUSIONS AND FUTURE WORK ....._ .....___ .........__ ...........8
4.1 Conclusions............... .. ................8
4.1.1 Characterization of Explosive Compounds ............ ..... ................84
4.1.2 FAIMS Characteristics .............. ...............85....
4. 1.3 IMS/FAIMS/M S ............... ...............85....
4. 1.4 High-resolution FAIMS...................... ..............8
4.1.5 Mixture Separations with FAIMS/MS .............. ...............86....
4.1.6 Disadvantages of FAIMS .............. ...............86....
4.2 Recommendations............... .. ...........8
4.2.1 Further HRFAIMS Improvements .............. ...............87....
4.2.2 FAIMS Carrier Gas Experiments .............. ...............87....
4.2.3 Practical Separations and Analysis............... ...............88
4.2.4 IM S Gating .............. ...............88....
4.2.5 CV Scanning............... ...............88
4.2.6 Instrument Maintenance .............. ...............89....
4.3 Future Work............... ...............89..
4.3.1 Ionization Sources .............. ...............89..
4.3.2 Development of Field Instruments ................ ...............91..............
REFERENCE S .............. ...............92....
BIOGRAPHICAL SKETCH .............. ...............96....
LIST OF TABLES
2-1 Explosives to be Characterized. .............. ...............27....
3-1 Relative lon Intensities. ............. ...............79.....
3-2 Relative CV Values.............. ..............79..
LIST OF FIGURES
1-1 Finnigan LCQ APCI source ................. ...............15........... ...
1-2 FAIMS waveform and ion mobility behavior ................. ...............17........... ..
1-3 APCI source and FAIMS cell mounted on a Finnigan LCQ. ................... ...............1
1-4 Ionalytics line-of-sight FAIMS cell geometry. .............. ...............20....
1-5 LCQ quadrupole ion trap showing ion traj ectory. ................ ....._. ........._._....23
1-6 Example of oscillating ion motion according to the Mathieu equation (let a=1 and
1-7 Mathieu stability diagram. ............. ...............25.....
2-1 Compensation voltage (CV) scanning. ............. ...............31.....
2-3 Negative ion APCI mass spectrum of TNT. ............. ...............33.....
2-4 TNT total ion count over four 2 min CV scans. ............. ...............34.....
2-5 FAIMS/APCI mass spectrum of TNT. ............. ...............35.....
2-6 FAIMS/APCI mass spectrum of TNT: measured for 2 min at CV = 6.6V ......................36
2-7 FAIMS/MS/MS of TNT m/z 197.1 peak at 35% collision energy ................. ...............37
2-8 CV spectrum of 2,4-DNT. ............. ...............38.....
2-9 FAIMS/MS mass spectrum of 2,4-DNT. .............. ...............39....
2-10 2,4-DNT: 2 min at CV = 11.5V. .............. ...............40....
2-11 CV spectrum of 2,6-DNT. ............. ...............41.....
2-12 FAIMS/MS mass spectrum of 2,6-DNT. .............. ...............42....
2-13 Mass spectra of 2,6-DNT: 2 min at CV = 1 1.6V and 15.3V. ...........__.. ........__........43
2-14 CV spectrum of RDX ................. ...............44........... ...
2-15 FAIMS/MS mass spectrum of RDX. ................ ...................... ..................45
2-16 RDX: 2 min at CV = 5.1V. ............. ...............46.....
2-17 CV spectrum of HMX ................. ...............47........... ...
2-18 FAIMS/MS mass spectrum of HMX. .............. ...............48....
2-19 HMX: 2 min at CV = 1.2V. ............. ...............49.....
2-20 NG (MW 227) ................. ...............50................
2-21 FAIMS/MS mass spectrum of NG. ........._._ ...... .__ ...............51.
2-22 PETN (MW 3 16)............... ...............52..
2-23 FAIMS/MS mass spectrum of PETN. ............. ...............52.....
2-24 FAIMS/MS mass spectrum of ANFO made with hexane. ............. .....................5
2-25 FAIMS/MS mass spectrum of ANFO made with nitromethane. .............. .............. .56
3-1 WSU electrospray ionization source ......................... ...............58.....
3-2 WSU electrospray ionization source photograph. ............. ...............58.....
3-3 Drawing of lonalytics orthogonal dome FAIMS cell. ......____ ........_ ................60
3-4 Photograph of lonalytics orthogonal dome FAIMS cell. ......____ ........_ ...............61
3-5 W SU apparatus. ............. ...............62.....
3-6 CV spectrum of HMX/RDX mixture ................. ...............65...............
3-7 Mass spectrum of HMX/RDX mixture ................. ...............66...............
3-8 CV spectrum of TNT/RDX mixture.. ............ ...............67.....
3-9 Mass spectrum of TNT/RDX mixture. ............. ...............68.....
3-10 CV spectrum of both isomers of DNT. ................ ....__ ...............69 ...
3-11 RDX CV spectrum at 0.5 mm............... ...............71...
3-12 RDX mass spectra at 0.5 mm............... ...............72...
3-13 RDX CV spectrum at 0.75 mm ................. ...............73........... ..
3-14 RDX mass spectra at 0.5 mm............... ...............74...
3-15 RDX CV spectrum at 1.0 mm ................. ...............75..............
3-16 RDX mass spectra at 1.0 mm............... ...............76...
3-17 RDX CV spectrum at 2.0 mm ................. ...............77........... ..
3-18 RDX mass spectra at 2.0 mm............... ...............78...
3-19 Relationship between CV and DV. .............. ...............80....
4-1 Distributed plasma ionization source ................. ...............90........... ...
LIST OF ABBREVIATIONS
Air Force Research Laboratory
automatic gain control
atmospheric pressure chemical ionization
Defense Advanced Research Products Research Agency
high-field asymmetric-waveform ion mobility spectrometry
improvised explosive device
ion mobility spectrometry
limit of detection
quadrupole ion trap mass spectrometer
Research, Development, and Engineering Command
Sandia National Laboratory
volatile organic compounds
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
CHARACTERIZATION OF COMMONLY ENCOUNTERED EXPLOSIVES USING HIGH-
FIELD ASYMMETRIC WAVEFORM ION MOBILITY SPECTROMETRY COUPLED WITH
Jared J. Boock
Chair: Richard A. Yost
The goal of this research was to characterize explosive compounds using high-Hield
asymmetric waveform ion mobility spectrometry (FAIMS) as an ion separation device interface
to a mass spectrometer (MS). FAIMS is a relatively recently developed technology that is
promising, adding a dimension of separation. This method was employed in such a way as to be
conducive to possible future development of field instruments (e.g., use of atmospheric pressure
chemical ionization (APCI)). In addition, several experiments were conducted with the
application of this method, mainly dealing with optimization of the instrument parameters.
These experiments led to the development of a new method: high-resolution FAIMS. An ion
mobility spectrometer was used as a desolvation and ion-focusing device to drastically improve
FAIMS resolution. This allowed for the separation of mixtures that otherwise could not have
been separated. The resolution was high enough that isomers of the same explosive were
successfully resolved. This can be beneficial to forensics studies.
INTRODUCTION AND INSTRUMENTATION
A maj or threat to the national security of the United States is the continued use of
explosives abroad and possible use by terrorists on the domestic front.' As a result, there is a
great need for improved separation and detection technologies for trace amounts of explosive
compounds. There are many possible applications for such technologies, such as prevention of
terrorist events, military weapons/cache detection, interception of explosives trafficking,
improvised explosive device (IED)/landmine detection and identification, post-event forensics
analysis, and unexploded ordnance (UXO) detection. A significant challenge to the detection of
minute amounts of explosives is that oftentimes these analytes will be overwhelmed by
complicated background matrices.2 Another is that benchtop-size instruments are inadequate to
provide timely analyses to meet the needs of the user. For instance, a military unit searching for
landmines or IEDs in a combat zone cannot easily deploy a large and cumbersome instrument.
To meet these needs, an instrument is required that can both provide separation of analyte from
atmospheric background containing large amounts of volatile organic compounds (VOCs). The
required limit of detection (LOD) for such an instrument has been determined by many field
trials conducted by government and commercial agencies (US Air Force Research Laboratory
(AFRL), US Army Research, Development, and Engineering Command (RDECOM) Sandia
National Laboratories (SNL), and Defense Advanced Research Products Agency (DARPA) are a
few of these) to be in the parts-per-billion (ppb) range.3
One of the primary technologies in use for explosives detection today (as of 2006) is the
ion mobility spectrometer (IMS).4 An IMS consists of an ionization source, an atmospheric
pressure drift tube, and a detector. A sample is ionized and the ions are introduced to the drift
tube and are passed through a series of ring electrodes held at increasing voltages of polarity
opposite to that of the ions, which creates a strong electric field. Separation is based on mobility
of each ion in the field, which depends on size, shape, and mass of the ion. The IMS has
limitations, however. First and foremost is that it is easily susceptible to contamination from
background matrix. The IMS as an explosives detector is primarily used in indoor facilities such
as airports or railroad stations, and samples are usually introduced as a swipe rather than
sampling vapor from air.'
Mass spectrometry could address many of the limitations of ion mobility. Research is
currently being conducted by many different organizations to design and construct a field-
portable time-of-flight (TOF) mass spectrometer (MS) to meet these needs.6,7 IOns are
accelerated through a vacuum drift tube by an electric field. The mass-to-charge (m/z) ratio is
determined by the amount of time each ion takes to pass through the drift tube. In most cases, a
TOF/MS is coupled with a gas chromatograph, which provides initial separation.8 Current
commercially available field-portable TOF/MS designs do not have a high enough resolving
power to meet the above needs, and they are not sensitive enough to detect levels of explosive
vapor that are practical to operational needs.9
A relatively recent (theorized in the late 1980s and developed in the 1990s and 2000s) and
promising technology is high-field asymmetric-waveform ion mobility spectrometry (FAIMS),1o
which will be discussed in detail. FAIMS is a novel, continuous chemical separation method
that has already proven successful in clinical, proteomics, and environmental analytical
applications.ll FAIMS does not have a high enough resolving power to function as a stand-alone
device;12 however, it is ideally suited to function as an ion separation device (in lieu of a
chromatography instrument) that can be coupled to a mass spectrometer.
1.2 Atmospheric Pressure Chemical Ionization
Atmospheric pressure chemical ionization (APCI) is a gas-phase ionization technique that
uses either a radioactive P-emitting source or a corona discharge at a pressure of I atm to
produce reagent ions and ultimately analyte ions. This method was chosen for this series of
experimentation because it most closely resembles an ionization source that is amenable to be
used for field instruments. Most other ionization techniques either require vacuum or are large,
complex systems containing lasers or high-voltage; these techniques are less suitable for a man-
portable field instrument. The APCI method is not identical to the ionization method to be
employed projected field instruments, as a difference arises from sample introduction. The
APCI ionization source used for this series of experiments (see figure 1-1) is designed for
AP Cl AP ORI ZE R
NOZZLE RHE R
INLET FI VAPORIZER
IN LET I SAMPLE
APCI PROBE CORONADIHAE
D ISCH ARGE C ''NEEDLE
P OWE R
Figure 1-1. Finnigan LCQ APCI source. [Reprinted from Thermo-Electron Corp. 2004.
Finnigan LCQ Series APPI APCI Combination Probe Operator 's Manual, Rev. A,
interfacing to liquid chromatography, and, therefore, introduces analyte samples as liquid. The
liquid sample is nebulized into droplets, which are passed pneumatically via nitrogen sheath gas
into a heated region where they are vaporized. The nitrogen and vaporized solvent then serve as
reagent gas as the vapor is passed over a corona discharge which produces reagent ions that
ionize the analyte via ion-molecular reactions.13 There are different mechanisms depending on
the type of analyte and whether or not ionization is positive or negative. Explosive compounds
tend to favor negative ionization (this behavior and ionization mechanisms will be discussed
further in a later section).
1.3 High-field Asymmetric-Waveform lon Mobility Spectrometry
As stated in the introduction, high-field asymmetric waveform ion mobility spectrometry
(FAIMS) is a relatively recently developed technology that in this series of experiments serves as
an ion separation device. FAIMS cells have many different configurations, though all basically
consist of two electrodes separated by a gap; the simplest configuration is two parallel metal
plates, which, consequently, was the first iteration of this design (Figure 1-2).15 An asymmetric
waveform is applied between these plates, which can either be the sum of two sine waves,12 Or a
square wave; in this case, a sum of sines was used with a high negative potential of -4000 V for a
short time, and a low positive potential for a slightly longer time. This peak voltage is known as
the dispersion voltage (DV), and creates a strong electric field. The rapid change in polarity of
the applied alternating potentials causes the ion to oscillate due to attraction/repulsion forces.
For this reason, a high negative potential is typically used for negative ions.
The basis for separation for a FAIMS device is a change in the mobility of an ion as this
field varies, unlike IMS which measures the mobility itself.16 IOns exhibit a variety of mobility
behaviors under high field, thus making FAIMS an ideal separation device. The changes in
lon Source k v Exit to Detector
Constant, or Asyrmetic Wavefrm;
decreei, : at ___---- High Field
High Electric Fields
Time t Hil? Fela, t CTime at Low Flr-Lu idld
Figure 1-2. FAIMS waveform and ion mobility behavior. [Adapted from Guevremont, Roger.
2004. High-Field Asymmetric Waveform lon Mobility Spectrometry (FAIMS). (p.
2, fig. 1-1). Ionalytics Corp., Ottawa, Canada.]
mobility are classified into three different categories: "A-type" behavior, or exponential
increase in mobility proportional to change in field strength, "B-type" behavior, or exponential
increase in mobility followed by exponential decay as field strength increases, or "C-type,"
which is exponential decay in mobility as field strength increases. The change in mobility
affects the direction the ion travels toward or away from the plates (Figure 1-2).
If these conditions are maintained, then all ions will drift toward the plates according to
each separate mobility. However, the drift of a particular ion can be halted by the application of
a small DC voltage to one of the plates. This compensation voltage (CV) allows a single type of
ion to pass through the plates. As mobility changes are greater, an increased compensation
voltage is needed to balance the drift and allow transmission of the ion.
The mobility of an ion influenced by an electric field is described by
Kh(E) = K[1 + f(E)]
where f(E) is the dependence of ion mobility (Kh) as a function of electric field.12
Field strength is described by units known as Townsend (Td), for which 1 Td =
1017 x (E/N) where N is the number density of the carrier gas (expressed in atoms/m3) and E is
the field strength (in V/cm). Change in mobility is described by
Kh/K = 1 + a(E/N)2 + P(E/N)4
where Kh is the ion mobility at high field and K is the mobility at low field.
The two constants a and P are dependent on the dimensions of the FAIMS cell. The
estimated field strength for this series of experiments was calculated to be about 80 Td
(20,000 V/cm) at 1 atm.
To analyze a mixture of analytes, with differing CVs, it is necessary to scan CVs over a
range. Therefore, each analyte in the mixture is transmitted through the cell separately and a CV
spectrum consisting of ion signal vs. CV can be acquired.
A simple FAIMS cell geometry of two parallel plates is not necessarily ideal. It is
understood that when FAIMS is interfaced between an ionization source and mass spectrometer,
there will be a degree of ion loss, though it should be noted that there is no loss of sensitivity.
This may not make much of a difference; however, for even though signal is reduced, noise may
be reduced even further. It has been proven that different types of cells such as the cylindrical
cell used in this study have an additional ion focusing capability.12
An lonalytics (Thermo) Selectra beta II prototype was used for the FAIMS work described
in this thesis, though there were two different types of cell geometries used. The line-of-sight
FAIMS cell was used for the characterization, and it consists of two concentric cylinders that are
separated by a 1 mm gap. In order to assemble the cell, a solid stainless steel inner electrode fits
into a hollow stainless steel outer electrode, and a PEEK endcap maintains uniform distance
between these electrodes. The ends of the inner and outer cylinders are machined to have
hemispherical dimensions toward the exit aperture. Unlike a flat-plate design, the electric field
in a cylindrical cell is not uniform; this configuration, along with the hemispherical exit area,
allows for the focusing of ions toward the exit aperture. A PEEK sleeve prevents arcing and
other types of electrical interference between the FAIMS cell and the environment. Figure 1-3
shows a photograph of the APCI source and FAIMS cell.
Curtain Plate lead
Figure 1-3. APCI source and FAIMS cell mounted on a Finnigan LCQ.
Prior to entering the FAIMS cell, ions pass through a hole in a curtain plate, which has a
potential of the same polarity as the ion mode of the source, in order to ensure that ions are
repelled into the FAIMS cell. For positive ion mode, the curtain plate voltage is kept constant at
+1000V, while negative ion mode curtain plate voltage is set to -1000V.
._00V1 To Waveform Generator
to mass spectromneter
I Sour ce 1m
Figure 1-4. Ionalytics line-of-sight FAIMS cell geometry.
A clean, dry curtain gas is added to the FAIMS cell via a side port at a rate of 3 L/min.
This curtain gas serves two purposes. First, a portion exits through the orifice in the curtain
plate, which assists with desolvation of the ions and droplets from the ionization source and
minimizes the accumulation of solvent neutrals from the APCI source into the FAIMS cell. The
remaining gas propels the ions through the FAIMS cell and into the heated capillary of the mass
spectrometer. This gas can consist of a pure gas or a gas mixture, typically nitrogen, helium or
carbon dioxide; however, nitrogen was chosen for these experiments due to its predominance in
the atmosphere. Any moisture or trace hydrocarbons introduced to the carrier gas will affect the
performance of the FAIMS device by reacting with the ions, so charcoal and/or molecular sieves
are used, to purify the gas.
A waveform generator serves as the power source for the asymmetric waveform. A cable
connects the generator directly to the FAIMS cell. There is also a connector on the waveform
generator that allows the operator to observe the waveform on an oscilloscope and ensure that
the waveform DV and shape are constant. The Selectra prototype FAIMS unit is extremely
sensitive, and any slight interference with the waveform cable can affect the waveform.
There are four different FAIMS "modes" depending on the polarity of the CV and DV
polarity required to transmit a particular ion.12 The determination as to which mode an ion falls
under at this juncture is made experimentally, as a library has not yet been fully constructed,
which justifies a portion of this thesis. An ion in the "Nl" quadrant (including many explosives)
passes through the FAIMS cell with a negative DV and positive CV; an ion in "N2" requires
positive DV and positive CV; an ion in "Pl" requires positive DV and negative CV; and "P2"
ions are transmitted with negative DV and CV. This rule applies only for "A" and "C" type ions;
for "B" type ion mobility behavior it is not well defined. The mobility behavior of an ion is
affected by many factors, including: size, rigidity, and interaction with carrier gas (according to
Stokes and Einstein).l7
Although use of FAIMS as a separation device decreases total ion transmission, when the
CV is set for a particular ion of interest, the relative signal to noise ratio of that ion increases
1.4 Ion Optics
The instrument used for this experimentation was a Finnigan (Thermo-Electron) LCQ,
which is a commercial, benchtop quadrupole ion trap mass spectrometer (QITMS). After the
analyte is vaporized, ionized by APCI, and separated via FAIMS, ions pass through the heated
capillary into a lower-pressure area, where the tube gate is located. The tube gate carries two
potentials of a polarity opposite to the analyte ions, one which focuses the ions into the skimmer,
which acts as a restriction between the higher pressure area (1 torr) of the tube gate and the lower
pressure (10-3 torr) of the ion optics. The second potential is used to deform ions and prevent
them from traveling further. The ion optics consist of a series of an octopole, lens, and a second
octopole. An octopole is a set of eight metal rods that carry an RF voltage (400 V) and DC
offset voltage (+10 V for negative ions) that induces an electric field that directs and focuses ions
into the mass analyzer. The lens between the octopoles serves both as a focusing medium and as
a pressure restriction, for the pressure once again decreases to 2x10-5 torr in the second octopole
and mass analyzer. As pressure is continually decreased, the mean free path (mfp) is increased,
and there are fewer neutral molecules and other ions present to interfere with the passage of the
focused analyte ion beam.18,19
1.5 Quadrupole lon Trap Mass Spectrometry
A quadrupole ion trap (Figure 1-5) consists of a hyperbolic ring electrode to which an RF
potential is applied, as well as two end caps with a DC potential.20 A DC offset voltage
(-10 V for negative ions) is applied to the end caps so ions will enter the trap. An RF waveform
of a particular drive frequency (02) is applied to the ring electrode which creates the quadrupolar
electric field, causing the ions to oscillate on a controlled trajectory. He gas at approximately
10-3 torr removes excess kinetic energy from the ions so that they can be trapped. The ion
traj ectory in the center of the trap is not to scale.
ENrTRANCE ENlDCAP SPRING
ELECTRODE POST WSASHER
Figure 1-5. LCQ quadrupole ion trap showing ion trajectory. [Adapted from Thermo-Electron
Corp. 2003. Finnigan2 LCQ Series Hardw~are Manual, Rev. A.]
The oscillating ion motion is described by the Mathieu equation:22
az = -2ar= q, = -2q =
The variables az and qz are functions of the dimensions of the trap and the potentials
applied, while z and r represent the axial and radial directions (between and perpendicular to the
endcaps, respectively); U is the DC amplitude applied to the ring electrode (if any), V is the RF
potential applied to the ring electrode, e is the charge on an ion, m is the mass of an ion, ro is the
inner radius of the ring electrode, zo is the axial distance from the center of the device to the
nearest point on one of the endcap electrodes, and 0Z is the angular drive frequency. The angular
drive frequency is equal to 2x1fRF where fRF is the frequency of the main RF voltage in Hertz.23 A
graphical example of the motion in arbitrary directions is shown in figure 1-6. The actual motion
is similar to this example.
-6 -4 -2 0 2 4 6
Figure 1-6. Example of oscillating ion motion according to the Mathieu equation (let a=1 and
q=1/5). [Adapted from McLachlan, N. W. 1947. Theory and Application of
Mathieu Functions, Dover].
For an LCQ, qz = 0. 1963 [RF frequency (number of charges)/ion mass].19 Solutions of
the Mathieu equation describe a stability diagram (figure 1-7) from which it can be determined
whether or not an ion will remain in the ion trap under a given set of conditions. If the qz of an
ion with known mass/charge falls within the boundaries of the stability diagram (is stable in both
axial and radial directions), then it can be trapped. If the qz falls outside the boundary, then the
ion will collide with the end cap electrodes and be lost. The factor P is based on the secular
frequency of the oscillation of an ion. When P equals one, the secular frequency equals half the
frequency of the RF field, and the magnitude of its oscillation increases such that the ion is lost.
Essentially, the P lines on the graph are voltage scans; in a quadrupole ion trap the amplitude of
the DC and RF voltages are ramped (while keeping a constant RF/DC ratio), to obtain the mass
spectrum over the required mass range. The sensitivity is a function of the scanned mass range,
san speed, and resolution.24
wS ~instability scan
Figure 1-7. Mathieu stability diagram. [Reprinted from McLachlan, N. W. 1947. Theory and
Application of Mathieu Functions, Dover].
The ion trap can only hold a certain number of ions at a time before repulsion forces (space
charge) cause ion displacement, resulting in a loss of resolving power and band broadening. To
counter this phenomenon, the LCQ employs a function known as automatic gain control (AGC)
where a prescan is conducted that monitors the generation of ions in the trap.25 The AGC scan
and the full MS scan form a microscan. In these experiments, three microscans were used for
each ion inj section period.
1.6 Multidimensional Mass Spectrometry
Another capability of the LCQ is multidimensional MS or MS/MS. This function can be
selected to determine the fragmentation patterns of an ion of a particular m/z in order to identify
the ion creating the peak or to examine the chemistry of that ion. A ion is selected, known as the
parent ion, and all other ions are ej ected from the trap. An excitation RF voltage is applied across
the end cap electrodes, causing the parent ion as kinetic energy to increase; collisions with the
He gas in the trap result in parent ion dissociation into fragment (daughter) ions. The
fragmentation pattern can be changed by altering the amplitude of the excitation RF voltage.19,26
1.7 Thesis Overview
This thesis is organized into four chapters. Chapter 1 is the introduction to the research
and an overview of the equipment used for the characterization portion. Chapter 2 presents data
relating to the characterization of explosives via FAIMS/MS. Chapter 3 introduces the second
instrument and compares it to the instrument used for characterization. There were several
experiments relating to FAIMS/MS and IMS/FAIMS/MS that were conducted throughout this
research; these are discussed in chapter 3. Chapter 4 discusses the conclusions of this research
and recommendations of other related efforts, as well as possible long-term future work.
PROPERTIES AND CHARACTERIZATION OF EXPLOSIVES
2.1 Explosives that were Characterized
A list of explosives was selected for characterization by FAIMS/MS based on threat
assessments by various government agencies (including the Department of Homeland Security
(DHS) and Federal Bureau of Investigation (FBI)). This list is contained in the table 2. 1. Each
of these compounds will be discussed in great detail below.
Table 2-1. Explosives to be Characterized.
Compound Formula llolecular Weight
TNT (2,4,6-trinitrotoluene) C7HSN306 227
HNIX (cyclotetramethylene tetranitramine) C4HsNsOs 296
RDX (cyclotrimethylene trinitramine) C3H6N606 222
2,4-DNT (dinitrotoluene) C7H6NZO4 182
2,6-DNT (dinitrotoluene) C7H6NZO4 182
NG (nitroglycerin) C3HSN30, 227
PETN I.~ ... lll. 1... tetranitrate) CSH8N4012 316
ANFO (ammonium nitrate fuel oil) *
*dependent on type of fuel used
2.2 Classification of Explosives
An explosive material is chemically unstable or becomes unstable under certain conditions,
and as a result of rapid exothermic decomposition, can be destructive. There are many types of
explosives; however, government threat assessments1,27 have determined that only a certain
number of them are significant threats to life and property, primarily due to factors such as
supply and cost. As a result of this determination, the compounds of the highest threat levels
were examined closely in these experiments, as well as their precursors and degradation
products, if possible, as the presence of these can also be used to detect manufacture or storage
of the respective explosive materials.
There are two forms of explosive decomposition. The more powerful of these, detonation,
is characterized by an extremely fast supersonic shock wave released by the energy produced in
the favorable chemical reaction. The sudden high pressure of the shock wave is the major cause
of destruction resulting from a detonation. The other form of decomposition, deflagration, is the
subsonic portion of an explosion, usually propogated by thermal conductivity of the substance in
contact with the outer edge of the plume. Explosives are classified as either "low" or "high"
depending on whether or not they detonate.28 Low explosives include gunpowder and
pyrotechnics. High explosives are those used for military applications, mining, or demolitions.
All of the explosive compounds included in this study are considered high explosives.
There are also subcategories of high explosives: primary, secondary, or tertiary. Primary
high explosives are extremely sensitive to shock, friction, and heat, and are difficult to store.
Secondary high explosives are stable enough to be stored for a period of time, yet can be
detonated due to a sudden shock or heat. Most of the compounds in these experiments are
secondary high explosives. Tertiary explosives (blasting agents) will not easily detonate without
the aid of a primary or secondary "booster." This list includes ammonium nitrate fuel oil
(ANFO), which was a subject in this study.
2.3 Chemical Properties of Explosives
The maj ority of secondary high explosives characterized in this study are nitroaromatics,
nitramines, or nitrate esters. In the nitroaromatics and nitramines, one or more nitro (NO2)
groups are attached to an aromatic ring structure. Nitrate esters also contain nitro groups. These
nitro groups are bulky and electron-withdrawing, causing the nitrogen in the nitro groups to be
partially positive. Under APCI, primarily negative ions ([M]- or [M-H]-) are formed. Non-
halogen groups with atoms that are more electronegative than carbon draw substantial electron
density from the aromatic system. If enough internal energy is applied, one or more of the nitro
groups may detach from the ring or central structure, producing fragments. In either case, the
ions produced are negative, and negative mode must be used in order to separate and detect
Nitroaromatics and nitramines are generally stable, but can be reactive; this property is
especially apparent in functional groups attached to the ring at a position meta to a nitro group.29
These compounds are thermally labile, so ionization techniques such as electron ionization (EI)
and chemical ionization (CI) will result in significant fragmentation of the molecule, which
makes detection difficult in a complicated background matrix. A "softer" ionization technique,
such as atmospheric pressure chemical ionization (APCI) or electrospray ionization (ESI) is
needed.13 Due to their low volatility under standard conditions, only a small amount of these
explosives will vaporize for detection in air, so a sensitive separation and detection techniques
are required. Because these compounds can condense on dust or other particles present in the
immediate environment, instruments designed to detect explosives in air should often sample for
explosives present on particles30,31 as well as the small amounts of vapor. Nitrate esters are
much less stable, and require lower instrument operating temperatures for analysis.
Explosive compounds are also highly toxic,28 creating an environmental hazard for areas
around disposal sites. Vapors can be introduced to the lungs, and liquids can be absorbed
through the skin, causing a variety symptoms depending on the relative concentration, even
leading to death.32
2.4 Instrument Settings
Both positive and negative ions can be formed in the APCI source, and the LCQ can
analyze ions of either polarity. As most of the explosives have large electronegative leaving
groups, they favor the formation of negative ions.33 It has also been shown that there is less
background noise in negative mode than positive mode. All of the mass spectra shown consist of
50 analytical scans, each consisting of three microscans.
The explosive solutions were diluted in a solvent containing 64.9% methanol and 35%
deionized water, as earlier studieS13 have yielded excellent results using this ratio. Solvents that
evaporate easily are best used for APCI, so that lower vaporizer temperatures can be used to
prevent thermal degradation of the analytes of interest. An extremely small amount,
approximately 0. 1%, of carbon tetrachloride was added to the solutions as some of the explosive
compounds favored the formation of a chloride adduct. This behavior will be discussed in turn.
All solutions were diluted to a concentration of about 10 ppm, as this is fairly realistic based on
calculated explosive detection applications.34
Solutions containing explosives were inj ected directly via the syringe pump of the LCQ
into the vaporizer at a flow rate of 20 CLL/min with a maximum ion inj section time of 50 ms for
AGC. The vaporizer was kept at a temperature of 3000C, while the heated capillary was set to
1500C. The LCQ software was used to tune the instrument as needed throughout the study in
order to maximize signal strength.
2.5 Compensation Voltage (CV) Scans
For all of the characterization in this section, CV was scanned from 0-20V four times over
2 min, for a total period of 8 min. The transition at the end of the scan from 20V back to OV is
extremely fast, covering a period of milliseconds. The ion of interest will only pass through the
FAIMS at its relative CV, therefore there will be a peak of ion flow at this voltage (ideally 100%
transmission); at other voltages the ion collides with the wall of the cell and is not transmitted.
(see Figure 2-1) This results in a peak at a particular time (and thus CV) in each of the four
CV CV CV CV
2 min 4 min 6 min 8 min
CV CV CV CV
Figure 2-1. Compensation voltage (CV) scanning.
2.6 Characterized Compounds
220.127.116.11 (2,4,6 -) Trinitrotoluene (TNT)
TNT is one of the most commonly encountered explosives due to its continued use for over
a century in both military and commercial applications. It is deemed a significant threat for use
by terrorists and other enemies of the United States. Like other nitroaromatics, it is stable until
ignition, other than slow decomposition over a period of years.28 Thus, there is the possibility of
low amounts of degradation products present in samples of TNT. Therefore, the most common
of these should therefore be characterized, as well as precursors used in the manufacture of TNT.
These compounds include the six dinitrotoluene isomers and trinitrobenzene (TNB). For the
structures of the nitroaromatics characterized in this study, refer to Figure 2-2. TNT has been
used on its own or as a component in mixtures, such as with ammonium nitrate (amatol),
aluminum powder, or other explosives.
1VW = 182
VI~W = 227
VI~W = 1 82
MIW = 213
Figure 2-2. Nitroaromatics and molecular weights. [Adapted from Reich, Richard.
2001. (Figure 1-11). PhD dissertation. University of Florida.]
When TNT detonates it (ideally) decomposes according to the following equation:28
2C7H5N306 4 7CO + 7C + 5H20 + 3N2
although complete decomposition is only reached after some oxygen from surrounding air is
added as an ignition reactant.
MIW = 182
979 113.212so 1~
TNT #1-105 RT:ODD-1 80 AV 105 NL: 1.34E5
T: p Full rns [50 DD-3 00 DO]i
60 80 100 120 140
160 180 200 220 240 260
Figure 2-3. Negative ion APCI mass spectrum of TNT. The first spectrum (Figure 2-3) shows
negative ion APCI of TNT without FAIMS separation. The [M]- ion at 227.0 and
[M-H]- ion at 226. 1 can be seen clearly, and a small amount of background noise
(ions that correspond to minor fragment ions of TNT or impurities) is visible.
MIW = 227
RT: 0.17-7.67 SM: 7B
100. 7 B2E5
1 O-i 74
OBi 02 231 362 _450 1 89 65 9
1 2 3 4 5 6 7
Figure 2-4. TNT total ion count over four 2 min CV scans. Figure 2-4 shows the CV spectrum
(total ion count) for TNT. As mentioned earlier, the ion of interest will only pass
through the FAIMS at its correct CV; therefore, there are ion peaks at these CV
values. There were four CV scans from 0-20 V, each with a duration of 2 min, for a
total of 8 min. It is from these spectra that the exact CV for each compound can be
calculated. From this spectrum, the CV can be calculated. For instance, the peak at
0.64 min: (0.64 min/2.00 min) x 20.0V = 6.4V.
TNT #34-660 RT: 0.38-7.58 AV: 627 SM: 7B NL: 2.81E3[M
T: p Full ms [ 50.0 0-40 0.0 0] M
Figure 2-5. FAIMS/APCI mass spectrum of TNT. Figure 2-5 shows the FAIMS/MS spectrum
for TNT. The background noise has been greatly reduced after the application of the
FAIMS. (see figure 2-2) The evident tradeoff for the improved signal/noise ratio is
the 50 x reduction in signal strength. In addition, two fragments of TNT also pass
through the FAIMS, as the fragment ion CVs are very close to the CV of the TNT
[M]- ion. The fragments show that TNT is a delicate ion that readily loses NO and
OH. The fragment peaks are in larger proportion to the molecular ion peak when
FAIMS is applied, most likely because the total ion count is lower, and there are few
to no impurities. Also, the [M-H]- no longer appears.
ThiT #1-173 RT: 0 00-1 98 AV: 173 NL' 1.71E4
T: p Full ms [50.00-400.00]
S40- [M-(2NO+CH3 1
-63.3 78.3 151.9 166 8
200 250 300 350 400
Figure 2-6. FAIMS/APCI mass spectrum of TNT: measured for 2 min at CV = 6.6V Figure 2-6
shows a 2 min spectrum taken at a CV of 6.6V, which is when the most TNT [M]- ion
passes through the FAIMS. This spectrum is similar to that in the previous figure; the
only difference is that instead of averaging one of the CV peaks, data points were
taken for a stationary CV. As a result, this spectrum is improved further from that in
TNT #816-344 RT: 0.23-4.90 AV. 329 NL. 2.55E2
T. p Full ms2 email@example.com [50.00-400.00]
MW = 227
II I I.I
Figure 2-7. FAIMS/MS/MS of TNT m/z 197. 1 peak at 35% collision energy. Figure 2-7
demonstrates that FAIMS/MS/MS of even weak ions from explosives is feasible.
The FAIMS CV was set to 6.2V to allow the [M-NO]- fragment of TNT to pass
through the cell. This ion was collided with 35% dissociation energy, causing the
loss of a second nitro group. In other words, the predominant fragment ion peak
shown here is the same elemental composition as the [M]- ion for mononitrotoluene.
RT: 0.00 -8.01 SM: 7B
Figure 2-8. CV spectrum of 2,4-DNT. Figure 2-8 shows the CV spectrum for 2,4-dinitrotoluene
(DNT), which is a precursor of TNT, and is often an impurity in explosives
containing TNT. This CV spectrum is different from that of TNT (figure 2-4) in that
the CV is 11.5V, compared to 6.6V. In addition, there is a small amount of solvent
ions that transmitted at slightly lower CV than the maj or DNT ion through the
24DNT #399-419 RT: 500-5 25 AV 21 NL: 553E3 [M-H]~
T: pms[50 00-400 00]
2 so- 1VW = 182
35 1[M-2NO]- [M]~
5~ 5.6 877
100 150 200 250 300 350 400
Figure 2-9. FAIMS/MS mass spectrum of 2,4-DNT. Figure 2-9 shows the mass spectrum of 2,4-
DNT after FAIMS separation (averaged over the peak at 6.6 min in the CV spectrum
in figure 2-8). As with TNT, this ion fragments readily, and experiments have shown
that many of these fragments have CV' s that are close to the primary ion.
24DNT3 #2-159 RT 0 02-1 99 AV 158 NL 5 35E3
T: p ms [50.00-400.00]
so M0 IW = 182
30 [M2O~\ [M-NO]~
100 150 200 250 300 350 400
Figure 2-10. 2,4-DNT: 2 min at CV = 11.5V. A mass spectrum of 2,4-DNT at a fixed CV of
11.5V is displayed in figure 2-10. There is little to no noise.
RT 0 00 801 SM 7B
05 071513141 9 1 6513
m/z 181 m/z 1
3 50 3
3 36 35
3 3 3 76 6
25 30 3 0
Tim ( in
7 44 7 4 4
n6 59 65
485 5 15
5 33 1 5 71
45 50 55 60
M8S 4-182 4
15S 5152 5
80_ m/z 152 1419
40 1 41 1 6 179
C 11 .1 1.1 102191
00 05 10 15 20
65 70 75 80
Figure 2-11i. CV spectrum of 2,6-DNT. 2,6-dinitrotoluene is an isomer of 2,4-DNT and another
precursor of TNT. Both nitro groups are in ortho positions to the toluene methyl
group. 2,6-DNT is more likely than 2,4-DNT to lose an NO group, and as such, there
are two predominant ions formed in APCI, causing the CV spectrum to display
different behavior than 2,4-DNT. The two ion peaks have close CVs and are not
completely resolved from each other. These can be seen in figure 2-11.
Experimentation in improvement of FAIMS resolving power was conducted and will
be discussed in chapter 3.
26DNT#B12-636 RT 0 14-7 98 AV 625 NL 3 62E2
T p ms [50 00-400 00]
Figure 2-12. FAIMS/MS mass spectrum of 2,6-DNT. Figure 2-12 is a mass spectrum averaged
over the entire 8 min run. The two predominant ions are visible at m/z 181 ([M-H]~)
and m/z 152 ([M-NO]-), as described at figure 2-11.
181 1 26DNT#234-262
NL 4 43E2
[M-NO] C 16 2 O
I CV =11.6V2,6-DNT
30 1 1 0
20)c~ M= 182
92 152 O 26DNT#12-636
625 NL: 3 62E3 T
0-[M-N O]- -pms
50 00-400 00]
so- CV = 15.3V
40-- ~[M-H]- M
53 8 eg 01 1 1 8 1
100 150 200 250 300 350 400
Figure 2-13. Mass spectra of 2,6-DNT: 2 min at CV = 1 1.6V and 15.3V. The fact that 2,6-DNT
more readily fragments under APCI than its isomer does not affect the separation
when the CV is fixed, as each ion is transmitted through the FAIMS cell at a separate
CV. This behavior can be seen in figure 2-13. Note that the optimal CV values for
the [M-H]- ions of the two isomers are too close (1 1.6V vs. 1 1.7V) for them to be
separated. However, only 2,6-DNT produces a second FAIMS peak (at CV = 15.3V)
for the [M-NO]- fragment ion.
18.104.22.168 Cyclotrimethylene trinitramine (RDX)
The nitroamine RDX (research department composition X) was developed as an explosive
during the 1930s and was used widely during World War II. This explosive is found in many
mixtures (such as Torpex, RDX mixed with TNT and aluminum powder), though it is usually
encountered as the base for several types of plastic explosives, the most common of which is
Composition C-4 (RDX with polyisobutylene and di(2-ethylhexyl)sebacate as the binder and
plasticizer). RDX is stable at room temperature will not detonate without a detonator.35 Another
type of plastique that contains RDX is Semtex, which is a commercial explosive that has been
used by terrorists many times in the past (such as Pan Am Flight 103 in 1988). The other
component in the Semtex mixture is PETN (pentaerythritol tetranitrate), which will be discussed
in its own right.
RDX decomposes completely according to:
C3H6N606 4 3CO + 3H20 + 3N2
and also reacts with oxygen in the air to complete the reaction.
RT: 0.00 -15.99 SM: 7B
5.50 9.52 NL:
Figure 2-14. CV spectrum of RDX. Figure 2-14 shows the CV spectrum for RDX. The peaks
are fairly well reproducible. There are two small peaks before the ion of interest in
all four scans, which are identified in figure 2-15.
RDX #193-205 RT:012-15.97AV 13 SM 7~B NL 4.20E3
T~ p ms [ 50 00-400 00]
Figure 2-15. FAIMS/MS mass spectrum of RDX. The [M+Cl]- adduct is formed with the small
amount of carbon tetrachloride in the solvent. (see Figure 2-15) One property of both
RDX and HMX is the formation of little to no molecular ion. Interestingly, the
adduct formed of RDX with the 37Cl isotope of chloride can also be seen in the mass
spectrum. The first additional peak shown in the CV spectrum (figure 2-14) is m/z 62
which is an [NO3]- fragment peak; it is seen in many mass spectra of explosives,
especially at higher temperatures. The second is a small amount of m/z 286, which is
another adduct formed with RDX: [M+NO3 -.
RDX (MIW 222)
RDX (MIW 222)
RDX20 #~2-167~ RT 0.01-1.99 AV: 166 NL 2 26E4
T: p Full ms [50.00-400.00]
Figure 2-16. RDX: 2 min at CV = 5.1V. Figure 2-16 demonstrates that FAIMS/MS is an
effective method of separating and detecting chloride adducts of RDX, as there are no
22.214.171.124 Cyclotetramethylene tetranitramine (HMX)
HMX, derived from High Molecular Weight RDX (sometimes known as High Melting
Point Explosive), is a nitroamine discovered as a byproduct of RDX that consists of an eight-
membered ring structure rather than a six-membered ring. It is slightly more powerful than RDX
because molecules of greater mass generate greater energy when they decompose. Another
difference between HMX and RDX is that HMX has a greater temperature at which the molecule
begins to fragment (hence the name).35 HMX is used exclusively for military applications, and is
most often included in a mixture, such as in some types of plastic bonded explosives, or octol,
formed with TNT.
RT: 0.00 16.OO SM: 7B
12.29 j13 54 1431
4 2 _
II 2 4
5.43 6 20 7.95 8.28
l I l ll
Figure 2-17. CV spectrum of HMX. HMX behaves similarly to its nitroamine counterpart RDX.
The FAIMS CV spectrum shown in figure 2-17 reflects this. There are some
differences between the two, though. HMX has a different CV (thus it can be
resolved, as will be discussed in chapter 3), and there is no evidence of the [NO3]
peak or formation of the [M+NO3]- adduct.
HMX #r22-796 RT. 0.43-16.00 A 775 NL. 5.05E3
T: p Full ms [50.00-400.00]
65- -g I\
see ON .O
35~ HMX (MIW 296)
[M +37 I-
Figure 2-18. FAIMS/MS mass spectrum of HMX. Similar to RDX, the dominant ion for HMX
was [M+Cl]-. This can be seen in figure 2-18. Once again, adducts of HMX were
formed with both dominant isotopes of chloride. There was also evidence of a small
amount of [M] .
HMX #1-109 RT O 01-1 98 W*/ 109 NL 7 73E4
T p Full ms [50 00-400.00]
35] HMX (MIW 296)
Figure 2-19. HMX: 2 min at CV = 1.2V. Figure 2-19 shows the HMX [M+Cl]- peak as the CV
is fixed at 1.2V for a period of 2 minutes. The mass spectrum contains no
interference, and there is a small amount of the molecular ion peak, as there was
during the CV scans (figure 2-18).
2.6.3 Nitrate Esters
126.96.36.199 Nitroglycerin (NG)
There were two nitrate esters that underwent characterization via FAIMS/MS. The first of
these is nitroglycerin. Nitrate esters contain nitro groups like the previously mentioned
explosives, though they are bonded to oxygen atoms, rather than nitrogen or carbon. There is
also no ring structure, unlike the nitroaromatics and nitramines (Figure 2-20). Because of this
structure, nitrate esters are extremely unstable, and fragment easily, requiring only a light shock
to explode. Nitroglycerin is one of the oldest known explosives, discovered in the mid-
nineteenth century.28 It is mainly used in the manufacture of dynamite, gunpowder, and rocket
Figure 2-20. NG (MW 227) [Adapted from Paul W. 1996. Explosives Engineering, (Figure 3-3
p. 76), Wiley].
NG #1-538 RT 0.01-6.75S AV 538 NL 9.19E2
T -pms[50 00-40000]
10- 1.9[N 3O_ O N O
eJ s o* '"o-
ss-~ NG (MW 227)
500 .5 0 5 0 5 0
45ur -2 FISM a setu fG iial t ntomns hr o
40-ua o o G I loddntfr adcs vntog tehae ailr
35-eAC orewa erae ote iiu f10C alo h irt
30-hsoN eefamntd hs irt os(mz6)wr eaae n
25-tdwt AM/Ma hw nFgre22.Bcuete asrnewl o
20-dblw5 m o hs ntuet ti o nwnhwtecro tutr
Figerure 2-21. FAIMoS/MeS. maseeFgr -2 Es spcrmofN.Smlalote i aniitramnes thereis drno
WWIIandof thea APIsource wiiasy decresedton thres miniu deoaof 10C alle fof thenditraes
artiler branchn oe yes of NGl weefagmbentd Thesenitra. te ios (ee m/z 62) were separated an
relatve detecilted with FlAIMS/MSd as hw m intrt Figure -21. Because mpasst rangewl o
Figure 2-22. PETN (MW 316). [Adapted from Paul W. 1996. Explosives Engineering, (Figure
3-3 p. 76), Wiley].
PETN #81-158 RT 0 00-1 98 AV 158 NL 7 93E2
T p ms [50 00-400 00]
PETN (MW 316)
100 150 200 250 300 350 400
Figure 2-23. FAIMS/MS mass spectrum of PETN. Like the nitroamines and NG, PETN does not
yield a detectable molecular ion. It did form both chloride and nitrate adducts;
however, these were in extremely low intensity. As a result, PETN did not form any
discernable peaks of high intensity during this experimentation. However, due to its
extreme instability, (also like NG) PETN did fragment into nitrate ions which were
separated and detected by FAIMS/MS (Figure 2-23). As this figure shows, the total
ion count is relatively low.
[M-NO2]- [M+Cl]- +NO3]
261 3498A,57 1 r377 ,
188.8.131.52 Ammonium Nitrate Fuel Oil (ANFO)
ANFO is the most commonly encountered explosive today, used mostly for mining
operations. It is simple and inexpensive to produce, and the components are not difficult to
acquire. It is one of the most used explosives by terrorists, having been used by organizations
such as: Fuerzas Armadas Revolucionarias de Colombia (FARC), Provisional Irish Republican
Army (IRA), Palestinian extremists, and some American extremists. Ammonium nitrate (used in
fertilizers) is an oxidizer and can explode inefficiently on its own if shocked. A hydrocarbon
mixture provides fuel for the reaction. Usually kerosene or diesel fuels serve this purpose
industrially; although, terrorists have used other fuels (such as nitromethane to make the
"kinepak" mixture used for the Oklahoma City bombing). Since there are so many different
production methods and types of fuels used in different proportions, no two batches are identical.
This, coupled with the fact that ammonium nitrate and hydrocarbon fuels are present in
background air, suggest that ANFO will be extremely difficult to detect. On the other hand, a
destructive bombing requires large amounts of it (eg., over 5,000 pounds at Oklahoma City)
Ammonium nitrate decomposes according to: 2NH4NO3 & 4H20 + 2N2 + 02
the most important product being oxygen. The hydrocarbon fuel added burns rapidly to include
CO2 aS a product.28
ANFO is extremely stable, but is hygroscopic, which hinders long-term storage. ANFO
will not explode without a secondary booster (blasting cap); the power of the explosion can be
controlled by altering the mixture.
Five different types of ANFO were examined in this research. The hydrocarbon fuels used
were: n-pentane, n-hexane, toluene, nitromethane, and diesel fuel. All were made using about
94% (by mass) ammonium nitrate and the remainder of fuel. The pentane and hexane were used
to determine the effects of using a simple straight-chain hydrocarbon fuel to simplify (and thus
identify peaks in) the mass spectra. Nitromethane was used to produce the "kinepak" mixture
that was discussed above; toluene was used to provide a large structure with a methyl group to
compare and contrast results to kinepak ANFO. Characterization was done on the FAIMS/MS in
both positive mode and negative mode. Dilute solutions (about 10-15 ppm) were heated slightly
to instigate the oxidation reaction without causing detonation.
Characterization of ANFO using this method is difficult. There are no other mass spectra
with which to compare the results of this effort. This is likely due to the fact that no two batches
of ANFO are alike, and the mass spectra are difficult to reproduce. A limitation of the LCQ is
that the mass range cannot be set below m/z 50; therefore, fragment ions from ammonium nitrate
and the fuels cannot be seen.
Nearly all of the negative ion APCI data sets collected show an ion base peak in common:
[NO3]- (m/z 62). This may result from ammonium nitrate that has not yet reacted. For those
fuels of m/z greater than 50; though, the fuel itself was not seen neither in positive mode MS, nor
in negative mode MS (though straight chain hydrocarbons are almost never seen in negative
mode MS), the exception to this being ANFO made with diesel. In every case for ANFO, a
single 16 min scan was conducted comprising an area from CV -40V-40V. In each of the
following figures (2-24 and 2-25), a single peak was detected.
ANFOHex#82-639 RT~ 0 02-8 02 AV~ 638 NL~ 2 04E5
T p Full ms [5O 00-400 00]
Figure 2-24. FAIMS/MS mass spectrum of ANFO made with hexane. Figure 2-24 shows a
FAIMS/MS mass spectrum of ANFO made with hexane. The base peak is [NO3 ~,
which passed through the FAIMS cell at a CV of 9.7V. There is also a small peak at
I _I I In II I
ANFOk~inepakt_0701 04150533#~81-1294 RT~ 0 01-15 55 AV~ 1294 NL~ 9 48E2
T p Full ms [50 00-400 00]
9r 0- 10
Figure 2-25. FAIMS/MS mass spectrum of ANFO made with nitromethane. ANFO made with
nitromethane (figure 2-25) as the fuel has a much lower ion signal than that made
with hexane; this behavior was also seen with ANFO made with toluene. It also has a
[NO3]- base peak, though there are some other ions as well. These ions were in
common with some of the other runs as well: m/z 125, m/z 141, and m/z 284.
Detection of ANFO with a field instrument would prove to be difficult, as ammonium
nitrate exists in the background (mostly in fertilizers) and the most common fuel, diesel, is used
to power engines everywhere. The most likely type of ANFO to be detected would be
"kinepak," for nitromethane is used for little other than as a fuel additive for race cars.
304 2350.7 38O392 7
FAIMS/MS AND IMS/FAIMS/MS OF EXPLOSIVES
The experiments described in this chapter were conducted using a different instrument
configuration than that in chapter 2. As in the other experiments, the mass spectrometer was a
3.1.1 Electrospray Ionization (ESI) Source
Electrospray ionization (ESI), employed in this study, is similar to APCI. Instead of
relying on a corona discharge to ionize molecules in the gas phase, ESI uses electrical charge to
ionize molecules in solution. A liquid solution containing the analyte of interest in solvent is
passed through a capillary held at high potential. Droplets leaving the capillary become highly
charged. As the solvent in the droplets evaporates, ionized molecules are ejected into the gas
p ase. 36
The unique non-commercial ionization source used in the studies in this chapter was
developed by the Herbert H. Hill laboratory at Washington State University (WSU). It is a fairly
simple apparatus, consisting of a fused silica capillary held in place by a ferrule, and mounted in
a nonconductive teflon casing. Sample is introduced via syringe pump into this capillary, to
which approximately -13.5 kV of ionization potential is applied to a stainless steel ferrule
through which the capillary is threaded. Due to the relatively small (about 0.5 mm) internal
diameter of the capillary, the flow rate of the syringe pump can be low (as low as 1 mL/min).
Figure 3-1. WSU electrospray ionization source. Figure 3-1 shows the WSU ESI source. The
darkly shaded areas are constructed of Teflon, and the capillary is fused silica. The
ferrule used is conductive stainless steel.
Figure 3-2. WSU electrospray ionization source photograph. Figure 3-2 is a photograph of the
WSU ESI source. Note that the tip of the capillary extends through the screen of the
IMS. This was done so that the ions would not be deflected backwards by the electric
3.1.2 lon Mobility Spectrometer (IMS)
An ion mobility spectrometer (IMS), as briefly discussed in Chapter 1, is similar to
FAIMS, in that it employs an electric Hield to separate ions based on mobility. An IMS drift tube
was designed and patented37 by the Hill laboratory at WSU. This tube has a total length of 34
cm, containing a series of flat ring-shaped electrodes composed of conductive stainless steel with
dimensions of 50 mm outer diameter x 48 mm inner diameter x 3 mm, separated by ceramic
rings. A voltage is applied to each of these ring electrodes that generates the electric field that
propels ions through the drift tube. For negative ions, the first electrode carries a potential of
about -10.5 kV, and this is incrementally decreased on successive electrodes such that the last
electrode carries about -2 kV. The curtain plate of the FAIMS (set to -1 kV) acts as the Einal
electrode of the IMS to attract negative ions into the FAIMS cell. This IMS operates at
atmospheric pressure and employs heating to help desolvate the sample. The heater was adjusted
to be as high as possible without fragmenting the sample ions (temperature ranging from about
1500C to 2500C). Countercurrent gas flow also aids in desolvation.
The IMS contains a Bradbury-Nielsen gate ring electrode38 which divides the tube into a
desolvation region and IMS drift region. A reference potential is used to "open" the gate, thus
allowing ion transmission from the desolvation region into the drift region, or a different voltage
can be applied to "close" the gate by creating an orthogonal electric Hield that prevents ion
transmission. In these experiments, the IMS was first operated in total ion transmission mode,
and then it was set to gate ions for separation.
3.1.3 Orthogonal Dome FAIMS Cell
This instrument configuration also utilized an lonalytics Selectra beta prototype; however,
the cell geometry was not the same as that used in the first series of experiments (chapter 2). The
cell used in the characterization experiments consisted of a line-of-sight cylindrical geometry; in
contrast, the cell used in this chapter (figure 3.3) employs an orthogonal dome-shaped geometry.
The waveform generator, application, and waveform shape are identical. Two major differences
between the two configurations are that the dome-shaped cell permits the distance between the
electrodes to be adjusted, and provides slightly higher resolution.39
STo Waveform Generator
la Dispersion voltage
to mass spectrometer
NOT TO SCALE
Figure 3-3. Drawing of lonalytics orthogonal dome FAIMS cell. Figure 3-3 contains a drawing
of the orthogonal dome FAIMS cell. The red and black dots describe a simple
scenario in which the black dots represent the analyte of interest and the red dots are
noise (interferent ions). Both types of ion enter the FAIMS cell through the curtain
plate and are separated as the red ions collide with the walls of both electrodes when
the proper CV is set. A percentage of the analyte ion is also lost, but this is the only
ion type that enters the mass spectrometer.
Figure 3-4. Photograph of lonalytics orthogonal dome FAIMS cell. Figure 3-4 is a photograph
of the disassembled orthogonal dome cell used for these experiments. The plate gap
adjuster rotates to change the distance between the inner and outer electrodes.
Figure 3-5. WSU apparatus. A photograph of the entire WSU apparatus including ESI source,
IMS drift tube, and FAIMS cell can be seen in Figure 3-5. The waveform generator
is not connected to the FAIMS cell so that it is easily visible.
3.2 FAIMS Limitations
As discussed briefly in Chapter 1, one of the primary limitations of FAIMS is fairly low
resolution compared to chromatographic separation methods, hence its use here in conjunction
with mass spectrometry as a separation device rather than a stand-alone instrument. There are
several factors that can affect resolving power of a FAIMS cell.40 The factors that have been
addressed in the following experiments include dispersion voltage, distance between the inner
and outer electrodes in the domed FAIMS cell, carrier gas consistency, and addition of a
desolvation region prior to the FAIMS cell. In addition, another limitation is band broadening
over time as buildup of solvent interferes with ion separation within the FAIMS cell.
3.3 Effects of Addition of Helium to the Carrier Gas
For the entire first set of experiments (chapter 2), the carrier gas in the FAIMS cell was
100% pure nitrogen. However, experimentation has shown that the addition of a percentage of
helium can increase transmission and sensitivity. This effect slightly narrows CV peaks and
therefore increases resolving power. The reason for this behavior is not clear; although, it is
suspected that because helium increases mean free path within the cell, ion dispersion and
desolvation are also increased. It is also possible that the high-Hield is effectively strengthened.
It is not desirable to use too high a percentage of helium carrier gas into the cell, as the
high voltage can induce discharge. It should be noted that because the amount of helium in the
atmosphere is trace, Hield instruments developed with FAIMS technology would most likely not
employ helium in the carrier gas.
3.4 High-resolution FAIMS (HRFAIMS)
High-resolution FAIMS (HRFAIMS) experiments were conducted on the WSU apparatus
described in this chapter. Three maj or differences between this instrument and that used for
characterization in chapter 2 in terms of FAIMS resolution are: FAIMS cell geometry, addition
of an IMS (set to total ion transmission mode) in front of the FAIMS cell, and use of a
percentage of helium (25%) in the nitrogen carrier gas.
As discussed above, the FAIMS cell used for this set of experiments was an orthogonal
dome cell, rather than a line-of-sight cylindrical cell. The dome cell can provide higher
resolution, due in part to the longer path length through the cell, and due in part to the use of two
Hield areas in series; there is a cylindrical area and a spherical area.
The IMS drift cell was set in full ion transmission mode (ie., there was no gating);
although, the heater and electric Hield were still engaged. There is also gas flowing
countercurrent through the IMS, which can dramatically reduce the amount of solvent reaching
the FAIMS cell. The temperature of the heater was set to be as high as possible without
fragmenting the explosive molecules. The IMS had the most significant effect on overall
resolution compared to the other factors. Transmission was increased, and CV peaks were
narrower, demonstrating increased resolution. This is probably due to the fact that the heater
desolvates the sample, and the electric field focuses ions into a tight beam for more efficient
introduction into the FAIMS cell.
3.4.1 Separation of Mixtures of Explosives
Although use of FAIMS as a separation device may decrease total ion transmission, when
the CV is set for a particular ion of interest, the signal-to-noise ratio of that ion can increase
The first mixture that was examined consisted of about 20 ppm of HMX and about 15 ppm
of RDX. The DV was set to -2500V, which was the optimal setting when the IMS was installed.
The carrier gas was made up of 75% nitrogen and 25% helium to improve ion count and
resolution. Figure 3-6 is the CV spectrum for the HMX/RDX mixture. CV was scanned from 0-
20V once over a period of 8 minutes. Each peak is distinct and fully resolved. The RDX peak is
significantly smaller because there was not only a slightly smaller concentration in the sample,
but also the boiling point is much less than HMX.
-800 SM 7B
200 2 12127,
10 15 20 25 30 35 40 45
50 55 BO BS 70 75
Figure 3-6. CV spectrum of HMX/RDX mixture.
SI I I 'li l l i I l i l l i l l'I '
100 1 50 200 250 300 350 400
Figure 3-7. Mass spectrum of HMX/RDX mixture. Figure 3-7 contains the mass spectra of the
RDX/HMX mixture, showing the [M+Cl]~ adducts of both RDX (m/z 257) and HMX
(m/z 331) taken from the relative CV peaks (which are seen in Figure 3-6). The
HMX peak at m/z 341 is an [M+NO3]- adduct
RD~XH MXMidurel 6#
R DXH MXMMIurel 6#
40-54 RT' 0 80-1 08
AV 15N L1 18E 3
1746 2027 i
RT 0a00-1589 SM 7B
100- 8 49E3
95: TIC MS
so- 1 RDX
451" I Solvent CV 7.2V
40- ~CV 5.0
30 ~5.76 ,
15~ 11.20 5 47
10- 0.67 9 54 6 9 8 42 9 62 10 96 119
23467810 1112 1314 15
Figure 3-8. CV spectrum of TNT/RDX mixture. A second mixture examined was torpex, which
consists of TNT and RDX in approximately equal portions. The only difference
between the mixture that was examined here and the military explosive is that in this
case there is no aluminum powder. In this case, one CV scan from 0-20V was done
over 16 minutes. Once again, both components (see Figure 3-8) are fully resolved
from each other and the solvent background in the CV spectrum.
225 9 RDXTNTMlxturel 2#Q8
100 -4-92 RT 1 69-1 85
90- AV 9 NL 5 20E2 T -
TNT p ms i
68- 256 8 RDXTNTMlxturel 2#2
5.70-5.78 AV. 5 NL.
60-- 3 55E2 T p ms [
50RDX 50 00-400 00]
100 150 200 250 300 350 400
Figure 3-9. Mass spectrum of TNT/RDX mixture. Figure 3-9 shows the mass spectra of the
respective CV peaks of TNT and RDX in the mixture (Figure 3-8). Both components
(the TNT [M]- ion at m/z 226 and the RDX [M+Cl]- ion at m/z 257) are shown. It
should be noted that because the same DV was used for both of these mixtures
(-2500V), the CV values of RDX and the solvent is identical in the CV spectra of
both mixtures. (Compare figs. 3-6 and 3-8)
3.4.2 Resolving Isomers
The next step in separation is the resolving of two isomers of the same explosive. The
explosive chosen was dinitrotoluene (DNT), because it has two isomers that are precursors to
TNT: 2,4-DNT and 2,6-DNT. The analysis of these isomers can potentially be important to
explosives forensics. Few, if any, technologies deployed today can detect and resolve these
isomers from each other. Because they are isomers, they have identical molecular weight. Both
have similar mobility behavior, but due to the fact that the shapes of the molecules are slightly
different, each behaves uniquely under high field.
RT' 000-1589 SM' 7B
95: TIC MS 24
9012,4-DNT and 26
85. \CV 110.8V
"i CV 12.2V
40 0 7 95
II 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Figure 3-10. CV spectrum of both isomers of DNT. A mixture was created using about 20 ppm
of both isomers in equal proportion. This is a relatively small amount, hence the low
ion transmission. The CV scan was conducted slowly from only 10-15V over 16
minutes. Both isomers are clearly resolved from each other, as shown in Figure 3-10.
It should be noted that there was a [2,6-DNT-NO]- fragment peak that appeared
during the characterization (as discussed in Chapter 2, section 2.6). This fragment did
not appear while using the high-resolution setup in the 10-15V CV range.
3.5 Dispersion Voltage
Another effect that the IMS added was that the optimal DV was no longer the highest
possible (-4000V); the ion signal was nearly zero above a DV of -3000V. The FAIMS carrier
gas flow rate also had to be reduced to about 0.5 L/min from 4.0 L/min for optimum
performance. This is in contrast to the situation with FAIMS by itself (as in chapter 2), where
optimum sensitivity is observed at maximum DV and high carrier gas flow rate. There are
several possible reasons for this unusual behavior. For example, higher flow leaving the curtain
plate may cause ion scattering at the IMS/FAIMS interface. This may be overcome by closing
the system with a nonconductive sleeve that encloses the space between the IMS and FAIMS.
This might also make directing the ion beam into the FAIMS easier. Another possible
explanation is that the small difference in potential between the last electrode of the IMS (-2kV)
and the curtain plate (-1kV) required carrier gas flow to be decreased to allow the potential to
transport ions efficiently into the FAIMS cell.
3.6 Effects of Adjustment of Plate Gap
As discussed above, the distance between the inner and outer electrodes of a FAIMS cell
affects the electric field generated. Since the electrode gap of the dome cell can be adjusted
(figure 3.3), data were collected to determine the optimal electrode gap for explosive detection.
The cylindrical cell electrode gap is fixed at 1 mm; whereas, the gap in the cylindrical
portion of the dome cell is fixed, the gap in the dome portion is adjustable up to 2.4 mm in
increments of 0. 1 mm. Four data sets were taken with a fixed DV at -3000V, each consisting of
two CV scans from 0-20V over a period of 8 min each. The sample selected was a solution of
about 15 ppm RDX, which was chosen because it has a relatively low fragmentation temperature
(about 1700C) compared to the other explosives. The FAIMS/MS behavior of both the [M+Cl]~
ion and the fragment ions were analyzed.
The expected relationship between resolution and electrode gap is that the lowest gap
should provide the highest resolution, as the field is stronger. As expected, the ion count was the
lowest, as there is a lower mean free path; most ions collide with the electrodes.
. as CV = 11.0V
RT 000-1600 SM 7B
CV = 7.8V
10 97 1~131
1or 0 8
CV = 16.0V
CV = 16.0V
5 67 8 9
10 11 12 13 14 15
Figure 3-11. RDX CV spectrum at 0.5 mm. The first trial was at a relatively small gap of 0.5
mm (Figure 3-11). The CV was scanned from 0-20V over 8 minutes twice. On the
figures, ion count is seen after "NL." This figure shows three peaks. One of these
(m/z 310) is probably a solvent cluster. A second (m/z 283) is the nitrate adduct of
RDX. The third (m/z 257) is the RDX chloride adduct. In this case, the RDX analyte
of interest is the smallest peak compared to the solvent. Probability of transmission
of RDX molecules relative to solvent indicates that when the electrode gap is
smallest, most RDX molecules collide with the electrodes and are lost. The
maximum ion signal is 2.46 x 104 COunts.
100 191 200 250 3131 35i0 4110
~ hM 1.16f~ T:-~
4dl Ne I.1BB T.-~
RT. 4.15-4.54 CWF
X NL ~,3r a ;r, p~
ssr.~ C PF;3 3557
2509 271.7 /7 310.1
Figure 3-12. RDX mass spectra at 0.5 mm. Figure 3-12 shows the mass spectra for each of the
three CV peaks at an electrode gap of 0.2 mm. Relative to the previous figure, the
top mass spectrum is the right-most CV peak. The RDX mass spectrum has an
additional peak at m/z 267, which is an adduct ion [RDX+NO2]- that co-elutes
through the FAIMS cell with the chloride adduct. This second adduct is not unusual,
though it is usually found in extremely low intensity compared to the chloride adduct.
RT:0 00-12 86 SM 7B
100- m?/Z 310 4.36 4 43 RDX+CI- 2E4
asCV = 10.7V bn CV = 18.0V TIC MS
851i 4 58
70- 1 6 69 RDX+ NOs-
esRDX+ NOs~ CV = 7.3V
CV I 477' 26.24
40.32 2 BB4 1 5 89
35- 3 4 4 7.6010 410hl 8 711.5 1.82
30. ~~~2.81771 771 2
3 34 I 7.95 10
25-27 712 2
20-7 265 570
11 2 1 2 2 9
II 1 2 3 4 5 6 7 8 9 10 11 12
Figure 3-13. RDX CV spectrum at 0.75 mm. The second trial done was at a gap of 0.75 mm,
which is 1.5 times larger. Again, the CV was scanned twice from OV to 20V over a
total period of 16 minutes. As expected the ion count is slightly higher. With this
larger gap, the resolution is lower (CV peaks are wider); furthermore, the relative
intensities of the different ions appear different (in particular, the RDX [M+Cl]- ion is
significantly more abundant). The maximum ion count is 2.92x104 COunts. Due to an
error, only the first 12 minutes are shown.
IS* -~jl I
CV = 10.7V
1BS 1326 I10.B 233.0I 287.1 21167 13223 354.I2 3M% 2 391.B
100 150 205 250 300 351) 400
RaT 504~7,2 AYI
id8 NL~1 20E3T -
CV = 18.0V
RT 3 9519 4 N A
83 NL 24~6E T: 0
CV = 7.3V
218 7 224t 91 277
328.1 341 8 32 9
Figure 3-14. RDX mass spectra at 0.5 mm. Figure 3-14 displays the mass spectra corresponding
to the three CV peaks in the previous figure. These are similar to those with the
smaller electrode distance, as each ion is well resolved.
PT 00 -1600 SM 7B
S45 m/Z 310 (
aso0 CV = 8.2V
10~ 273 27301 617
0518261 359 42594
RDX+CI~ TIC MS
CV = 18.5V
CV = 18.5V
7-83 CV = 8.2V 1470
O 3 1419
111 234 5 67 8
9 10 11
12 13 14
Figure 3-15. RDX CV spectrum at 1.0 mm. Next, the electrode gap was increased by a factor of
1.3, to 1.0 mm (Figure 3-15). The ion count was higher, yet the transmission
behavior is completely different from the smaller gaps. This CV spectrum appears
much like those in previous chapters, with a well-resolved peak for the analyte ion
and only small background. The RDX [M+Cl]- ion is much more intense, the peak at
m/z 310 is much less intense, and the peak at m/z 283 is hardly visible. The
maximum ion intensity has increased to 6.84x104 COunts.
Figure 3-16. RDX mass spectra at 1.0 mm. The mass spectra for RDX taken at a FAIMS
electrode gap of 1.0 mm can be seen in Figure 3-16. Thus far, both ion count and
resolution are superior to those at smaller electrode gaps. At this gap, resolution and
ion count are well-balanced. Likely, the electric field is at a strength optimal for ion
focusing. Note that the peak at m/z 267 is barely visible.
II ~_ I_ ____ I__I_ I _
RT 6 7 4.31 AVr
49 NL 5 6'iEJ 3 T p
CV = 18.5V
rF9: lsrs :~0? !I;O
RT 2514 8 A
148 N 21.51.1EJ
CV = 8.2V
15r 9 2T5.
_ta~i.9 3~97 3~2:8 j~7 g
10D 5401 300
RT 0 00 -1600 BM glJ/Z 310*
100, CV = 10.
Figure 3-17. RDX CV spectrum at 2.0 mm. The last trial was done with an electrode gap of 2.0
mm (Figure 3-17). If the distance is set too high, as in this case, then the electrodes
can no longer carry a waveform. This ion count was the highest of all these trials,
presenting three orders of magnitude greater transmission than the smallest setting at
0.5 mm. The RDX peak was robust and well-resolved; however, the peaks at m/z 283
and m/z 310 were again visible.
"iCV = 10.2V
CV = 8.7V
CV = 19.6V
WpSlI 4 g9E4Tl
I~il IM :Ot! ~W rO
Figure 3-18. RDX mass spectra at 2.0 mm. The mass spectra in Figure 3-18 show that the ion
counts are the highest compared to the spectra taken at other electrode distances.
These four trials confirmed that a distance between the inner and outer FAIMS
electrodes in the dome region set at about 1.0 mm yields the optimal transmission and
resolution for explosives. It was observed that the CV for the chloride adduct of
RDX increased as the electrode gap was increased, while the other peak CVs
decreased (allowing the peaks to be resolved), but when the gap was at 2.0mm, the
m/z 283 and m/z 310 peak CVs increased.
Table 3-1 shows the intensities of each of the major ions discussed above. As the electrode
gap was increased, the intensity of the chloride adduct of RDX increased by almost two orders
of magnitude, while the m/z 3 10 and m/z 283 ion counts decreased until the electrode gap was
set to 2.0 mm, where the trend reverses.
Table 3-1. Relative lon Intensities.
Electrode gap (mm) [RDX+ NOJ]" m/z 310 [RDX+CI]~
(counts) (counts) (counts)
0.5 1.5x10 2.5x10 2.5x10
0.75 1.2x10 2.9x10 2.5x10
1.0 5.7x10 6.0 x10 6.8x10
2.0 6.0 x10 2.0 x10 2.0 x10
Table 3-2 lists the CV values for the same ions. The behavior mentioned above occurs for
the CV values as well as the ion intensities. The chloride adduct of RDX increases slightly as
the electrode gap increases, and the other ions decrease until the electrode gap is set to 2.0 mm,
where the CVs increase.
Table 3-2. Relative CV Values.
Electrode gap (mm) [RDX+ NOJ]" m/z 310 [RDX+CI]~
(V) (V) (V)
0.5 7.8 11.0 16.0
0.75 7.3 10.7 18.0
1.0 5.9 8.2 18.5
2.0 8.7 10.2 19.6
To examine the ultimate resolution of the FAIMS system, an attempt was made to resolve
[M]- from [M-H]- ions for individual explosive compounds. Unfortunately, none of the
explosives have both of these ions in approximately equal proportion; this makes it more difficult
to allow for the separation of the ions. Nevertheless, for other types of ions, resolution is
certainly feasible. To maximize resolution, the CV scan would have to be extremely slow or
CV and DV created using data sets taken using the FAIMS/MS setup described in
chapter 2. The curves indicate that the CV of each ion increases at more negative DV
(hence greater field strength). TNT and other nitroaromatics do not produce ions that
are transmitted at a positive DV, although the chloride adduct ions of RDX and HMX
respond to both positive and negative DVs. The lonalytics instrument will not allow
the DV to be set less than & 2000V or greater than & 4000V; this is indicated by the
vertical red lines. Because of this DV/CV relationship, most of the explosives appear
to exhibit "B" type behavior when high field is applied.
over a narrow CV range, and the gas (and thus ions) entering the FAIMS cell would need to be
completely desolvated. Different adducts of the same explosive, however, have been separated,
for instance, the [M+Cl]- and [M+NO3]~ ions of RDX (see section 3.5).
3.7 Relationship of DV to CV
As discussed earlier, there is no single set of conditions that optimizes separation of every
mixture for every FAIMS-based instrument. In order to characterize a FAIMS-based separation
instrument, CVs must be found based on field strength, which is mainly a function of DV and
cell geometry. There is, however, the same relationship between DV and CV for all FAIMS
instruments, as this relationship is based on the change in mobility for an ion between low and
Data sets collected on the FAIMS instrument used for characterization (as described in
chapter 2) show that the DV for the optimal separation (highest resolution with lowest
background) for the nitroaromatic explosives is -4000V, and +4000V for the nitramines. The
IMS/FAIMS instrument (as described in this chapter) did not display the same behavior. The
peak DV was at about -2500V, and at any DV set higher than -3000V the ion signal decreased
rapidly to zero. In addition, the flow rate of the FAIMS carrier gas for this instrument could not
be set higher than 1 L/min, or ions were lost. There are several possibilities for this behavior; the
desolvation and focusing attributes of the IMS most likely require that a lower amount of carrier
gas and voltage are needed to achieve the same separation results.
One method to measure sensitivity and resolving power of a FAIMS instrument is to
examine the CV spectrum for both peak width and ion intensity. As discussed earlier, there is a
degree of ion loss through a FAIMS (data gathered during this research shows about two orders
of magnitude compared to no FAIMS cell in the system). Most FAIMS instruments (including
the instrument used for chapter 2) demonstrate an average CV peak width of about 2-3V;40 in
contrast, the best HRFAIMS system CV peaks were as narrow as about 0.75V (for examples, see
Eigs. 3-6 and 3-8). This has both benefits and disadvantages; narrower peaks require slower scan
speed to define the peak, and necessitate precise CV values when Eixed, but the narrower peaks
are better resolved from each other and from background ions.
3.9 IMS/FAIMS of Explosives
In another set of experiments, the IMS was operated in gating mode in order to attempt to
add an additional dimension of separation. This involved manually setting the times at which the
gate would pulse open with a delay to allow ions of a fixed drift time (and, thus, ion mobility) to
pass through the drift tube.
All other ions would be excluded from transmission. Hence, a percentage of all ions were
discarded by the IMS, in addition to those discarded by the FAIMS portion of the instrument.
An attempt was made to separate explosive mixtures by gating the IMS; however, ion counts
were not high enough to discern individual components of the mixture. This is largely due to the
loss in sensitivity due to the low duty cycle of the IMS/FAIMS system. For instance, if a 500 ms
gate width is used, and the gate is only open for 25 ms, then the duty cycle is only 5%, which
results in a significant loss of signal. Furthermore, the lack of synchronization between the IMS
gate pulses and the fill times of the LCQ quadrupole ion trap may result in even greater loss in
Gating the IMS was not critical in these studies, as data have indicated that individual
explosive peaks as part of a mixture are fully resolved while the IMS is functioning simply as a
desolvation and focusing apparatus for the FAIMS. Any future experiments should include
efforts to increase ion signal when the IMS is gating. (For full discussion see Chapter 4)
CONCLUSIONS AND FUTURE WORK
The United States is under constant threat from the use of explosives by her enemies.
There is a great need for the development of technologies that can eventually be applied in
instruments that will be used to detect improvised explosive devices. These analytical
instruments mainly will consist of laboratory and intermediate-sized instruments or man-portable
Hield instruments, and all will be utilized both overseas and domestically.
Instruments designed for the detection of explosive compounds can fall into two
categories, depending on function and/or mission. These are either qualitative threat detectors,
or quantitative threat characterizers.41 Factors that affect instrument use are portability, resolving
power, sensitivity, and selectivity.
Mass spectrometry is a widely used analytical methodology. When a separation device is
coupled to a mass spectrometer, all of the analytical figures of merit are drastically improved.
FAIMS is a promising technology that functions well as a separation device and is
compatible with a mass spectrometer. There are many advantages to employing FAIMS as such.
Even though the use of FAIMS may lower overall ion transmission, the signal-to-noise ratio of
the overall instrument may be much higher than a lone mass spectrometer. This can be
extremely important when an instrument is used in a Hield environment, away from the
laboratory, when there is a great deal of background matrix with which to contend.
The goal of this research was to determine the degree to which a FAIMS/MS device can be
employed to detect explosive compounds that are most likely to demonstrate a threat, and to
attempt to ascertain methodologies to eventually do so in a field environment. The experiments
discussed herein consisted primarily of three parts. The first characterized explosive compounds
using a FAIMS/MS, while the second included additional FAIMS/MS experiments conducted to
analyze and improve the analytical factors of the instrument. The third section described a novel
high-resolution FAIMS/MS method, employing an IMS drift tube to desolvate and focus ions
into the FAIMS cell. The HRFAIMS method is sensitive, and the resolution is high enough that
valuable forensic data can be collected.
4.1.1 Characterization of Explosive Compounds
A list of explosive compounds was compiled based on both the frequency of encounter
during terrorist events and the presence in common military explosives. The feasibility of
detection of these compounds was first evaluated; once accomplished, the characteristics of each
compound were analyzed and noted. All of the explosives in this study contained nitro leaving
groups. This list includes three nitroaromatics: TNT, which is one of the oldest known and
commonly encountered explosives, and two isomers of DNT that are precursors to the
manufacture of TNT, and, therefore, are forensically important. TNT is also found in many
military explosive mixtures. All of the nitroaromatics were easily detectable and were readily
characterized using the FAIMS/MS instrument.
There were two nitramines evaluated, RDX and HMX, which are encountered along with
TNT in military explosives and plastic explosives mixtures. The most common explosives
include mixtures of TNT and RDX. Both RDX and HMX are similar molecules that have
similar characteristics. The formation of a molecular-type ion requires adduct formation to be
well detected. The best adduct for this purpose is formed with the addition of a chloride (Cl-) ion
(improved by doping the solvent with a small amount of a source of chloride, such as CCl4),
though all will form adducts with nitrate (NO3 ) aS Well.
There were two nitrate esters evaluated, NG and PETN. Both are extremely unstable. The
molecular ion cannot be detected by MS; PETN will only form adducts in low concentrations.
NG will not form adducts at all. Because of this instability, though, NG and PETN are rarely, if
ever, used without another explosive in a mixture.
The last explosive evaluated in this study was ANFO, a mixture of ammonium nitrate and
a fuel, which is a commonly used commercial explosive. It is also favored by terrorists due to
the ease of its manufacture and acquisition of its components. The nature of these components
creates difficulties for detection, as ammonium nitrate and many of the fuels used (usually
diesel) may be found in high levels in the background. Because there is no single structure,
ANFO has not been well characterized for detection by MS.
4.1.2 FAIMS Characteristics
FAIMS separates ions using a high electric field to affect the change in ion mobility. One
of most important FAIMS characteristics that affect the strength of this field is the dispersion
voltage. Once the optimum DV was determined, then other analytical parameters could be
ascertained for each explosive compound. No two types of FAIMS instruments can be
characterized identically, though there is a basic relationship that applies in all cases between DV
Other FAIMS factors that were examined were the addition of a small amount of helium to
the carrier gas, which improves resolution and ion transmission, and the distance between inner
and outer electrodes of a FAIMS cell. The optimal electrode gap for the detection of explosives
on the instrument used corresponds to other FAIMS studies that have been done (a distance of
During these experiments, an IMS was added to the instrument to add another dimension
of separation. When the IMS was gated, only ions with a specific ion mobility were transmitted
on to the FAIMS cell; therefore, only a fraction of the ions were transmitted. The ion counts
during this experiment were not high enough to distinguish individual components of a mixture.
Changes could be made to the instrument to attempt to improve transmission and sensitivity.
This is discussed further in the "recommendations" section of this chapter.
4.1.4 High-resolution FAIMS
The addition of the IMS drift cell was proven to be beneficial to the instrument, however.
When the IMS was in total ion mode, there was no ion gating, and all ions passed through the
drift tube. Because the tube is heated and carries an electric field, the analyte was desolvated,
and the ion beam was focused into the FAIMS. When this method was used along with the
optimal FAIMS parameters, resolving power of the FAIMS, and ion transmission increased
drastically. At certain junctures, the CV peaks were so narrow that the CV scan speed had to be
slowed to allow the entire peak to pass through before the CV changed.
4.1.5 Mixture Separations with FAIMS/MS
Once characterization was completed, then the value of the FAIMS to separate mixtures
was analyzed. If an analytical instrument is unable to resolve mixtures, then its value as a field
instrument is limited. This phase of experiments was greatly enhanced using the HRFAIMS
method. First, two mixtures containing different explosives were separated successfully and the
components were well-resolved. Resolution was high enough that mixtures containing two
isomers of the same compound (DNT) were also successfully separated from each other. This
ability can be valuable for forensics analysis of either post-blast or manufacturing of TNT.
4.1.6 Disadvantages of FAIMS
FAIMS instruments still have limitations, especially pertaining to field use. The resolution
and transmission of the FAIMS cell are strongly affected by water and other solvent, and the cell
must be as dry as possible. After continued use, if the FAIMS cell is not periodically dried out,
the sensitivity and resolving power drop as either the electrodes are hindered from carrying a
high electric field, or the saturation of solvent causes ion-molecule reactions. As such, the
carrier gas must not contain impurities; furthermore, carrier gas is required in large volumes
(between 3-5 L/min). This can provide issues in terms of instrument period of use, maintenance,
and design, for it is undesirable to add large cylinders of gases to an apparatus if the obj ective is
The lonalytics prototype instruments are extremely delicate, and are not compatible with
field use. For instance, the DV is transported from the waveform generator to the FAIMS cell
via an insulated cable. Any environmental changes in the area around the cable can significantly
affect the waveform and field generated.
Analysis of data sets collected during this series of experiments has determined that
FAIMS/MS is a feasible method to use for the detection of explosive compounds. A high-
resolution methodology for the detection of explosives with FAIMS/MS has been identified
during this research study, and should be further explored.
4.2.1 Further HRFAIMS Improvements
The use of the IMS as an ion focusing/desolvation device could be further improved,
because its use during this study was not intended to be so. The interface between the IMS and
FAIMS consisted of an empty space between the two components. The IMS drift tube must be
lined up with the FAIMS cell manually, and both the electric fields as well as carrier gas may
cause ions to leave the system. A nonconductive sleeve could be used to enclose the system, and
may decrease ion loss, and may improve transfer into the FAIMS cell.
4.2.2 FAIMS Carrier Gas Experiments
Further, experimentation could be conducted into the composition of the FAIMS carrier
gas. Future field instruments will probably not include carrier gas cylinders. Other gas mixtures
could be tested, and the effects determined. A desirable approach would be, first, to use
laboratory-provided pure, dry "zero air," and if this is effective as a FAIMS carrier gas, then to
use ambient air run through a scrubber/dessicant just prior to introduction to the FAIMS cell.
4.2.3 Practical Separations and Analysis
Because the separation of mixtures was done successfully, the next logical step is to
separate unknown mixtures by, for example, having another individual prepare the samples.
Also, efforts should be made to separate explosive compounds from various background
matrices, thus determining the lowest possible concentrations at which explosives can be
detected. After this is done, then a study should be done to identify matrices (if any) that
strongly interfere with detection, and attempt to ameliorate them.
4.2.4 IMS Gating
The IMS/FAIMS/MS method requires more efforts to be made in terms of synchronizing
times when the gates open to the FAIMS CV scanning, and the inj section times of the quadrupole
ion trap. This will further ensure that the lowest number of ions possible is lost.
4.2.5 CV Scanning
A practical instrument would require scan speeds that are as fast as possible, so that
possible threats can be detected and warnings can be administered. CV scan speed can be
adjusted to any length of time, though the higher resolution the instrument, the slower the scan
speed that is necessary. For practical use, an optimal length of time should be set such that
mixtures are fully resolved, but the instrument can scan as quickly as possible. The CV scan
range can also be adjusted so that all of the explosives of interest are covered, yet it is set to the
minimum amount of time needed. If a qualitative threat warning is all that is required, then the
scan time can be minimized to allow just enough of the analyte ions to pass through the system
to be detectable above the noise level, thus further reducing the time required for a scan.
4.2.6 Instrument Maintenance
Both IMS/FAIMS instruments used in this research introduced samples into the system as
liquids. As such, the sensitivity and resolution of the FAIMS degraded over time, and regular
cleaning of the cell with a volatile solvent that did not leave a residue, such as methanol,
followed by drying of the cell were needed. Sampling explosives in a vapor or particulate
sample, rather than in liquid solution, may reduce this problem. Practical instruments should
include a small amount of solvent for regular cleaning; and note that carrier gas should always be
run through the system when in standby.
4.3 Future Work
Now that feasibility has been established, and the first step toward method development
for explosives detection via FAIMS/MS has been taken, the next step includes working toward
developing an instrument employing this method. A great deal of effort is required, however, to
design and create an apparatus that can be employed for use in a Hield environment. The best
way to approach this would be to Birst design and test an intermediate instrument that would be
used specifically for explosives and other related compounds. This work would likely include
further research in interfaces between each component of the instrument, FAIMS cell
geometries, and other ways to allow the instrument to serve its purpose. Also, a method of
practical sample introduction to the system is needed, as explosives have low vapor pressures
and are not conducive to collection via gas collection.
4.3.1 Ionization Sources
One possible way to move toward development of a Hield instrument was actually
attempted during this research study: the use of a more portable ionization source. ESI is
designed for ionizing liquid solutions, and APCI requires a heater and corona discharge; both of
these limit the capabilities of an instrument.
A novel type of ionization source was evaluated during this research for use with an
IMS/MS for the detection of explosives. Developed by Blanchard & Co., the distributed plasma
ion source (DPIS) is designed to provide ionization at atmospheric pressure using a simple,
portable, low-power device. The DPIS is constructed with two overlapping electrodes of
different sizes that are separated with a dielectric. A potential applied across the electrodes
creates a distributed plasma around the edge of the small electrode, which ionizes the
surrounding air and sample vapor. This type of source can be useful because sample can be
introduced as a vapor. In this case, liquid sample was kept in a bottle, across which a nitrogen
carrier gas was blown, and the headspace was directed over the DPIS source, into one of WSU' s
homebuilt IMS tubes.
Figure 4-1. Distributed plasma ionization source. Figure 4-1 is a photograph of the DPIS
extended through the ion screen of the IMS drift tube. The plasma discharge is a
distinct blue color. The sample line is taped to the insulated wire that powers the
The DPIS circuit has limitations, however. There is no diode to prevent overload of the
circuit, and the maximum potential cannot exceed -1kV. Data were collected with several
explosives. The DPIS source did ionize the samples, but the signal-to-noise ratio was extremely
low. There may be many reasons for this. First, the source voltage is low compared to the high
voltage of the IMS, which may limit transmission of ions into the IMS drift tube. Also, the only
sample directed into the instrument is headspace, which contains significantly less analyte than a
liquid which is vaporized.
This source is not compatible with laboratory-scale IMS/MS instruments, as the ionization
voltage is not scaled with the other voltages of the system. Nevertheless, the DPIS did ionize the
explosive samples, and can potentially be used with low-power, portable Hield instruments.
4.3.2 Development of Field Instruments
The first step in this research has been to determine the feasibility of the IMS/FAIMS/MS
method for detecting explosives. Now that this has been accomplished, improvements can be
made to the overall method (as discussed above). Once this is done, then work can proceed to
reduce size and power consumption of each component, and then durable packaging can be
developed for a Hieldable instrument.
1. Country Reports on Terrorism, US State Department, 2005.
2. Opportunities to Improve Airport Passenger So I emling~ \ ithl Mass Spectrometry,
Committee on Assessment of Security Technologies for Transportation, National
Materials Advisory Board, 2004.
3. Shea, Dana A.; Morgan, Daniel, Detection of Explosives on Airline Passengers:
Recommendation of the 9/11 Commission and Related Issues, Congressional Research
Service, Library of Congress, 2006.
4. Eiceman G. A.; Borsdorf H., lon M~obility Spectrometry: Principles and Applications,
Applied Spectrometry Reviews, 2006, vol. 41, no. 5, pp. 323-375.
5. Hill, Herbert H.; Simpson, Greg. Capabilities and Limitations oflon M~obility
Spectrometry for Field Screening Applications, Field Analytical Chemistry, 1997, vol.1,
no. 3, pp. 119-134.
6. Lee, Harold G.; Lee, Edgar D.; Lee, Milton L., Atmospheric Pressure Ionization Time-
of-flight Mass Spectrometer for Real-time Explosives Vapor Detection, Proceedings of
the First International Symposium on Explosive Detection Technology, pp. 619-63 3,
7. Osorio, Celia; Gomez, Lewis M.; Hernandez, Samuel P.; Castro, Miguel E., Time-of-
flight Ma~ss Spectroscopy M~easurements of TNT and RDX on Soil Surfaces, Detection and
Remediation Technologies for Mines and Minelike Targets, 2005, v ol. 5794, pp. 803-
8. B ottrill, A.R., High-energy Collision-induced Dissociation of2acromolecules using
Tandem Double-focusing/Time-of-f light Ma~ss Spectrometry, PhD thesis, University of
Warwick, Coventry, UK., 2000.
9. Gruber, Ludwig. Characteristics of Different Ma~ss Analyzers, Fraunhofer-Institute
for Process Engineering and Packaging, http://www. ivy.fraunhofer. de/ms/ms-
10. Purves, Randy W.; Guevremont, Roger; Day, Stephen; Pipich, Charles W.;
Matyj aszczyk, Matthew S., Ma~ss Spectrometric Characterization ofa High-field
Asymmetric Waveform lon M~obility Spectrometer, Review of Scientific Instruments,
1998, v ol. 69, no. 12, pp. 4094-4106.
11. Kolakowski, Beata M.; McCooeye, Margaret A.; Mester, Zoltan; D'Agostino, Paul
A., Rapid Separation and Identif ication of Chemical and Biological Warfare Agents in
Food and Consumer Matrices Using FAIM~S-M~S Technology, National Research Council
of Canada, 2006.
12. Guevremont, Roger, High-Field Asynametric Waveform lon M~obility Spectronsetry
(FAIM~S), lonalytics Corp., Ottawa Canada, 2004.
13. Reich, Richard, Fundamentalsd~~~dd~~~ddd~~ and Applications ofAtnzospheric Pressure Chemical
ionization Quadrupole lon Trap M~ass Spectronzetry for the Analysis ofExplosives, MS
thesis, University of Florida, 2001.
14. Finnigan2 LCQ Series APPL APCI Combination Probe Operator 's Manual, Rev. A,
Thermo-Electron Corp., 2004.
15. Buryakov, I.; Krylov, E.; Nazarov, E.; Rasulev, U., A New M~ethod ofSeparation of
M~ulti-atontic Ions by M~obility at Atmospheric Pressure Using a High-Frequency
Anaplitude-a;synametric Strong Electric Field, International Journal of Mass Spectrometry,
1993, vol. 128, pp. 143-148.
16. Kanu, Abu B.; Hill, Herbert H., lon M~obility Spectronzetry: Recent Developments
andNovel Applications, Washington State University, LabPlus International, 2004.
17. Atkins, Peter; dePaula, Julio, Physical Chentistly, 7th Ed., Chap. 24, Oxford Univ.
Press, Oxford, UK, 2001.
18. Welling, M.; Schuessler, H.A.; Thompson, R.I.; Walther, H.; lon M~olecule
Reactions, Ma~ss Spectronzetry and Optical Spectroscopy in a Linear lon Trap,
International Journal of Mass Spectrometry and lon Processes, 1998, vol. 172, no. 1, pp.
19. De Hoffman, Edmond and Stroobant, Vincent, Ma~ss Spectronzetry Principles and
Applications, 2nd Ed., 2005
20. Paul, W., Electromagnetic Traps for Charged and Neutral Particles,~~PPP~~~PP~~~PP Modern
Physics, 1992, vol. 62, no. 3, pp. 531-540.
21. Finnigan2 LCQ Series Hardware Manual, Rev. A, Thermo-Electron Corp., 2003.
22. Mathieu, E., 2Menoire sur Le M~ouvenzent Vibratoire d'une M~embrane de fornze
Elliptique, Journal of Pure Mathematics, 1868, vol. 13, pp. 137-203.
23. McLachlan, N. W., Theory and Application of2athieu Functions, Dover, 1947.
24. March, Raymond E.; Todd, John F. J., Quadrupole lon Trap Ma~ss Spectronsetry,
25. Murray, K. K.; Boyd, Robert K.; et. al., Standar~dd dddD efinitionsd~~d~ of Terms Related to
Ma~ss Spectronzetry, IUPAC, 2006.
26 Sparkman, O.D.; Klink, F. E., LC M~S: Fundamentalsd~~~dd~~~ddd~~ and Applications, ACS, 2001.
27. Annual Threat Assessment of the Director ofNational Intelligence for the Senate
Armed Services Committee, Statement by the Director of National Intelligence, John D.
Negroponte to the Senate Armed Services Committee, transcript available on
www.globalsecurity. org, Feb. 2006.
28. Davis, Tenney L., The Chemistry of Powder and Explosives, Ch. 1, AngriffPress,
Hollywood, CA, 1972.
29. Wade, L. G., Organic Chemistry, 6th Ed., Ch. 15-16, Prentice Hall, 2005.
30. Hannum, David; et. al., M~iniaturized Explosives Preconcentrators for Use in Man-
portable Explosives Detection Systems, Sandia National Laboratory, 1999.
31. Committee on the Review of Existing and Potential Standoff Explosives Detection
Techniques, Existing and Potential StandoffExplosives Detection Techniques, National
Research Council, National Academies Press, 2004.
32. Yinon, Jehuda, Toxicity and Metabolism ofExplosives, CRC Press, Boca Raton, FL,
33. Cooks, R. Graham; et. al., Direct, Trace Level Detection of Explosives on Ambient
Surfaces by Desorption Electrospray ionization Ma~ss Spectrometry, Royal Society of
Chemistry, 2005, vol. 15, pp. 1950-1952.
34. Federal Facilities Assessment Branch, Public Health Assessment, US Army Umatilla
Depot Activity, Centers for Disease Control and Prevention, 1997.
35. Cooper, Paul W., Explosives Engmneering, Ch. 3, Wiley, New York, NY, 1996.
36. Fenn, John B., Electrospray Ionization, Principles and Practices, Mass Spectrometry
Reviews, 1990, vol. 9, no. 1, pp. 37-70.
37. United States patent 7071465: lon mobility spectrometry method and apparatus,
inventors: Hill, Herbert H. et. al., Washington State University research foundation,
38. Bradbury, Norris E., and Nielsen, Russell A., Absolute Values of the Electron
Mobility in Hydrogen, Physical Review, 1935, vol. 49, no. 5, pp. 388-393.
39. Guevremont, R.; Thekkadath, G.; Hilton C. K., Compensation Voltage (CV) Peak
Shapes Using a Domed FAIM~S 0 ithr the Inner Electrode Translated to Various
Longitudinal Positions, Journal of the American Society of Mass Spectrometry, 2005,
vol. 16, no. 6, pp. 948-956.
40. Schvartsburg, A.; et. al., High-Resolution FAIM~S Using New Planar Geometry
Analyzers, Analytical Chemistry, 2006, vol. 78, no. 11, pp. 3706-3714.
41. Transportation Security: Systematic Plan2ning Needed to Optimize Resources,
statement of Cathleen A. Berrick to the US Senate Committee on Commerce, Science,
and Transportation, USGAO, 2005.
Jared J. Boock was born in Hazleton, PA, in 1980. He grew up in Torrington, CT,
and Panama City, FL. He completed his B.S. in chemistry from Florida State University
in 2002 and received a commission in the US Air Force.
His first duty station was the Molecular Sciences Division, Materials Technology
Directorate, Air Force Technical Applications Center, Patrick AFB, FL, where he served
as the division Test Director. His primary duty was to plan and execute field tests, and
evaluate data for the test and evaluation of chemical weapons detection equipment. He
wrote several classified test and technical reports.
In 2005, he was selected for the Air Force Institute of Technology Civilian
Institutions program, to study for his M. S. degree. He chose to pursue his M. S. degree in
analytical chemistry at the University of Florida. Upon completion of his M.S. degree he
will be stationed at the High Explosives Research and Development Facility at Eglin