<%BANNER%>

Laser Photofragmentation and Heterogeneous Chemiluminescence for Nitro-Based Explosive Detection

Permanent Link: http://ufdc.ufl.edu/UFE0021234/00001

Material Information

Title: Laser Photofragmentation and Heterogeneous Chemiluminescence for Nitro-Based Explosive Detection
Physical Description: 1 online resource (148 p.)
Language: english
Creator: Monterola, Maria Pamela Pineda
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: chemiluminescence, explosives, laser, luminol, petn, photodissociation, photofragmentation, photolysis, rdx, tnt
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The purpose of this research was to develop a simple, fast, reliable, sensitive and potentially portable explosive detection device employing laser photofragmentation (PF) followed by heterogeneous chemiluminescence (CL) detection. The PF process involves the release of NOx(x=1,2) moieties from explosive compounds such as TNT, RDX, and PETN through a stepwise excitation-dissociation process using a 193 nm ArF laser. The NOx(x=1,2) produced upon PF is subsequently detected by its CL reaction with basic luminol solution. The intensity of the CL signal was detected by thermoelectrically cooled photomultiplier tube with high quantum efficiency and negligible dark current counts. The research work was divided into four stages: (1) discerning the PF pathways of nitro-based explosives (2) luminol-CL method development and optimization for improving NO2 sensitivity (3) construction of CrO3 oxidizer for NO to NO2 conversion (4) evaluation of the device in detecting nitro-based explosives in various media including air and soil matrix. In the first stage of the research, the following PF pathways were verified using classical colorimetric analysis and a fast time resolved absorbance method. R--NO2 + hv goes to R + NO2; hv = 193 nm (1) NO2 + hv goes to NO + O*; hv = 193 nm (2) The second stage involved the development and optimization of NO2 detection using luminol CL. In this system, a stream of NO2 gas passed through a concentric nebulizer and was used to aspirate a spray of luminol solution. The aspiration process maximized NO2/luminol contact area enhancing the CL signal. The current system was able to improve the detection limit through optimization of all the feasible physical and chemical parameters involved in the CL reaction between luminol and NO2. Detection limit (LOD) of 19 ppt NO2 at (S/N) = 3 was reported. The optimal reagent solution from the viewpoint of sensitivity of the response to NO2 (maximum signal/signal noise ratio) is 5 x 10-3 M luminol + 0.01 M p-iodophenol + 0.2 M KOH. A Luminol/H2O2 CL set-up was also explored where a bundle of porous polypropylene fibers was used to bring the NO2 into contact with luminol solution. A LOD of 178 ppt NO2 at (S/N) = 3 was obtained. In the third stage, a CrO3 oxidizer was constructed for NO to NO2 conversion. Photolysis of nitro compounds produced NO (reaction 2) which is not detectable by luminol-CL system. To further increase the sensitivity of the device for NO2 fragments, NO must be converted back to NO2. Finally, the fourth stage was integration of PF and CL units into an explosive detection device via a CrO3 oxidizer. The system was able to detect energetic materials in real time at ambient conditions. Detection limits of 3.4 ppbv for PETN, 1.7 ppbv for RDX, and 34.5 ppbv for TNT were obtained. It was also demonstrated that the presence of PETN residue within the range of 61 to 186 ng/cm2 can be detected at a given signal to background ratio of 10 using a few micro joules of laser energy. The technique also demonstrated its potential for direct analysis of trace explosive in soil. LOD range of 0.5 to 4.3 ppm for PETN was established, an analytical capability comparable to currently available techniques.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Maria Pamela Pineda Monterola.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Winefordner, James D.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021234:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021234/00001

Material Information

Title: Laser Photofragmentation and Heterogeneous Chemiluminescence for Nitro-Based Explosive Detection
Physical Description: 1 online resource (148 p.)
Language: english
Creator: Monterola, Maria Pamela Pineda
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: chemiluminescence, explosives, laser, luminol, petn, photodissociation, photofragmentation, photolysis, rdx, tnt
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The purpose of this research was to develop a simple, fast, reliable, sensitive and potentially portable explosive detection device employing laser photofragmentation (PF) followed by heterogeneous chemiluminescence (CL) detection. The PF process involves the release of NOx(x=1,2) moieties from explosive compounds such as TNT, RDX, and PETN through a stepwise excitation-dissociation process using a 193 nm ArF laser. The NOx(x=1,2) produced upon PF is subsequently detected by its CL reaction with basic luminol solution. The intensity of the CL signal was detected by thermoelectrically cooled photomultiplier tube with high quantum efficiency and negligible dark current counts. The research work was divided into four stages: (1) discerning the PF pathways of nitro-based explosives (2) luminol-CL method development and optimization for improving NO2 sensitivity (3) construction of CrO3 oxidizer for NO to NO2 conversion (4) evaluation of the device in detecting nitro-based explosives in various media including air and soil matrix. In the first stage of the research, the following PF pathways were verified using classical colorimetric analysis and a fast time resolved absorbance method. R--NO2 + hv goes to R + NO2; hv = 193 nm (1) NO2 + hv goes to NO + O*; hv = 193 nm (2) The second stage involved the development and optimization of NO2 detection using luminol CL. In this system, a stream of NO2 gas passed through a concentric nebulizer and was used to aspirate a spray of luminol solution. The aspiration process maximized NO2/luminol contact area enhancing the CL signal. The current system was able to improve the detection limit through optimization of all the feasible physical and chemical parameters involved in the CL reaction between luminol and NO2. Detection limit (LOD) of 19 ppt NO2 at (S/N) = 3 was reported. The optimal reagent solution from the viewpoint of sensitivity of the response to NO2 (maximum signal/signal noise ratio) is 5 x 10-3 M luminol + 0.01 M p-iodophenol + 0.2 M KOH. A Luminol/H2O2 CL set-up was also explored where a bundle of porous polypropylene fibers was used to bring the NO2 into contact with luminol solution. A LOD of 178 ppt NO2 at (S/N) = 3 was obtained. In the third stage, a CrO3 oxidizer was constructed for NO to NO2 conversion. Photolysis of nitro compounds produced NO (reaction 2) which is not detectable by luminol-CL system. To further increase the sensitivity of the device for NO2 fragments, NO must be converted back to NO2. Finally, the fourth stage was integration of PF and CL units into an explosive detection device via a CrO3 oxidizer. The system was able to detect energetic materials in real time at ambient conditions. Detection limits of 3.4 ppbv for PETN, 1.7 ppbv for RDX, and 34.5 ppbv for TNT were obtained. It was also demonstrated that the presence of PETN residue within the range of 61 to 186 ng/cm2 can be detected at a given signal to background ratio of 10 using a few micro joules of laser energy. The technique also demonstrated its potential for direct analysis of trace explosive in soil. LOD range of 0.5 to 4.3 ppm for PETN was established, an analytical capability comparable to currently available techniques.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Maria Pamela Pineda Monterola.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Winefordner, James D.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021234:00001


This item has the following downloads:


Full Text





LASER PHOTOFRAGMENTATION AND HETEROGENEOUS CHEMILUMINESCENCE
FOR NITRO-BASED EXPLOSIVE DETECTION





















By

MARIA PAMELA PINEDA MONTEROLA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007
































2007 Maria Pamela Pineda Monterola































Gratefully dedicated to my family and to the loving memories of my father











ACKNOWLEDGMENTS

I would like to thank many people for all their help in completing this dissertation and in

my journey through graduate school. My main adviser, Prof. James D. Winefordner, has been a

wonderful mentor and inspiration. I greatly admire his dedication to education and research. I

would like to acknowledge Dr. Nicol6 Omenetto for his valuable research guidance. He never

ceases to amaze me with his creative ideas, and his passion for science is really contagious. I am

very grateful to Dr. Benjamin W. Smith, a.k.a. as the Mac Gyver of our lab, for all his advice in

instrument design and his cool ways in troubleshooting. I would like to extend my appreciation

to Dr. Igor Gornushkin for many intellectual conversations about spectroscopy and life.

I would like to thank my committee members, Dr. William Harrison, Dr. David Hahn and

Dr. Michael Scott for their valuable comments and suggestions in the completion of this research

dissertation.

I would like to recognize the US Army as a main source of research funding.

A special thanks to Dr. John Birks of University of Colorado for providing us the gas-

liquid exchange module and Dr. Mike Shepard of US Naval Surface Warfare Center for

providing us the with explosive samples. I would also like to acknowledge Duran and Tan's

research groups for allowing me to use their micro balance and spectrofluorometer. All the UV-

Vis measurements were conducted in the laboratory of Dr. Kathryn Williams.

I would like to express my deep gratitude to all support staff in chemistry, including the

machine and electronic shops for all of their contribution in instrument development. It has been

a blessing to have Ms. Lori Clark from the Graduate Student Coordinator's Office and Ms.

Jeanne Karably from the Analytical Chemistry Division with their unwavering patience and

assistance.









My five years at the University of Florida would not have been so enjoyable if it were not

for the wonderful people of Omenetto-Smith-Winefordner group. I would like thank Ron

Whiddon, Nick Taylor, Benoit Lauly and Jonathan Merten for being ready to help with anything

in the lab; with lasers, lifting, gas refilling and so on. I would like to recognize Jonathan Merten

again for his editorial assistance. A special thanks to Dr. Joy Guingab, Mariela Rodriguez, Dr.

Xi-hong Wu, Akua Oppong-Anane and Lydia Edwards for friendship. Those fun times and great

memories will always be treasured and cherished.

Heartfelt thanks goes to my friends who made my stay here at Gainesville pleasant these

five years. I am very grateful to Ms. Anne Moore for being so kind to me and for being my

second mom here in Gainesville. I would like to extend my gratitude to Merilla Stefan for her

camaraderie, Alexandr Oblezov for being so reliable at all times, especially during my second

year here at UF, Chen Liu and Ozlem Demir for their companionship and for setting examples of

hard work and perseverance. I am so blessed to have a very good support group from my fellow

Filipino graduate students. My heartfelt gratitude goes to Jhoanna Mendoza, Dr. Suzette Pabit,

Dr. Jemy Gutierrez, Machel Malay, Joey Orajay, Attorney Tesi Lou Guanzon, and Ai-ai

Cojungco. They helped make Gainesville a second home for me.

Most of all, I would like to thank my family for their incredible support and love. I would

like to acknowledge my brothers, Conrad, Chris and Carlo. My brothers and I shared wonderful

memories of playing basketball, chess, and play stations which instilled in me the art of being a
'warrior' in any game of life. I am forever grateful to my mother for her incredible strength and

unconditional love. I am also forever in debt with my father's legacy, of being so humble, kind,

and handsome.









TABLE OF CONTENTS



A C K N O W L E D G M E N T S ..............................................................................................................4

L IS T O F T A B L E S ................................................................................................. ..................... 9

LIST OF FIGURES ............................................. ........... ........................... 10

L IST O F A B B R E V IA T IO N S ......................................................................... ...... ............... 13

A B S T R A C T ............... ................................................................ .......................................... 1 5

CHAPTER

1 INTRODUCTION .................................... .. ........... ..................................... 17

B background and Significance ................................................. .............. ............... 17
Chem istry and H history of Explosives......................................................... ............... 17
Significance of Explosive Detection ......................................................................20
Review of Methods for Explosive Detection .............................................................22
C challenges on E explosive D election ............................................................ ................ 27
Scope of R research D issertation ..................................................................... ................ 27

2 PHOTOFRAGMENTATION PATHWAYS OF NITRO COMPOUNDS ............................38

In tro d u c tio n ............................................................................................................................. 3 8
B asic Theory of Photofragm entation.......................................................... ................ 38
Basic Theory of REMPI-TOFMS and LIF ......................................................................40
PF Mechanism of Explosives based on PF-REMPI-TOF and PF-LIF.........................42
E x p erim en tal S e ctio n .............................................................................................................. 4 6
R agents and C hem icals... ....................................................................... ................ 46
A pparatus and M ethodology .......................................... ......................... ................ 47
R results and D iscu ssion .................................................... ............................................... 49
D election of N O 2 Fragm ents ................................................................... ................ 49
Photofragm entation of N 2 ........................................................................ ..................... 51
C o n c lu sio n ................................................................................................... ..................... 5 3

3 NO2 DETECTION BY CHEMILUMINESCENCE .........................................................65

In tro d u c tio n ............................................................................................................................. 6 5
B asic Theory of Chem ilum inescence......................................................... ................ 65
Luminol-N02 Chemiluminescence ......................................................... 67
Review of CL Methods for NO2 detection.................................................................68
E x p erim mental S section .............................................................................................................. 7 0
R agents and C hem icals... ....................................................................... ................ 70
A pparatus and M ethodology .......................................... ......................... ................ 70


6









R results and D iscu ssion .................................................... ............................................... 73
C hem ical P aram eters .................................................. ............................................. 73
Physical Param eters ........................................................................... 75
Perform ance of Lum inol-CL D etector ....................................................... ................ 76
Perform ance of Lum inol/H202 CL D etector .............................................. ................ 77
Luminol CL Detector Compared to Other CL Set-ups .............................................77
C o n c lu sio n ................................................................................................... ..................... 7 8

4 CON VER SION OF N O TO N O 2 ................................................................... ................ 98

Introduction ...................................................... .................. 98
E x p erim mental S section .............................................................................................................. 9 9
R agents and C hem icals... ....................................................................... ................ 99
A pparatus and M ethodology .......................................... ......................... ................ 99
R results and D discussion ................ .. .................. .................. .......................... .. ............... 100
C o n c lu sio n ................................................................................................... ..................... 10 0

5 LASER PHOTOFRAGMENTATION AND CHEMILUMINESCENCE FOR NITRO-
BASED EXPLOSIVE DETECTION .................................. ...................... ................ 103

Introduction ............................................................................. 103
Experim ental Section ................................. .. ........... ............... ............... 105
R agents and C hem icals. .................................................................... ............... 105
A pparatus and M ethodology ................. ............................................................ 106
R results and D discussion ......................................................................... ...... .. .... .............. ............... 107
Relative PF Efficiency of Nitro-aromatic Compounds ...................... ...................107
NO2 Photolysis with Luminol-CL Detector........................................ 108
Analytical Capability of PF-CL Detector........................................... 108
C o n c lu sio n ........................................................................................................................... 1 1 1

6 DIRECT DETECTION OF EXPLOSIVE IN SOIL .......... ....................................126

In tro d u ctio n ... ................. .. .. .. .. .... ................................................................ 12 6
Environmental Hazards of Explosive Contaminated Soil.................. ...................126
Review of Explosive Detection in Soil ....... ... ...... ...................... 127
Experim ental Section ................................. .. ........... ............... ............... 128
C hem ical and R agents ..................... ................................................................ 128
A pparatus and M ethodology ................. ............................................................ 129
R results and D discussion ................ .. .................. .................. .......................... .. ............... 130
C o n c lu sio n ................................................................................................... ..................... 13 1

7 CONCLUSIONS AND FUTURE WORKS...... .... ...... ......................136

Sum m ary and C onclu sions ................................................. ............................................ 136
Future R research D irections... ..................................................................... ............... 137





7









L IST O F R E F E R E N C E S ....................................................... ................................................ 142

B IO G R A PH IC A L SK E T C H .................................................... ............................................. 148




















































8









LIST OF TABLES


Table page

1-1 Physical and chemical properties of nitrocompounds used in the study. ........................31

1-2 Physical and chemical properties of some nitro-based explosives...............................32

1-3 Summary of the performance of various explosive detectors reviewed in this study. ......34

2-1 Colorimetric analysis of NO2 fragments after photofragmentation of some nitro-
b a se d ex p lo siv e s................................................................................................................ 6 3

3-1 Summary of the LOD obtained from different luminol-N02 CL set-ups .......................97

5-1 Comparative table showing summary of the performance of various explosive
detectors based on PF-FD technique...... ............. ............ ..................... 113

5-2 Determination of PF efficiency of nitro-aromatic compounds................................. 118

5-3 Analytical figures of merit of PF-luminol CL detector for some nitro-based
ex p lo siv es...................................................................................................... ......... 12 5

6-1 Explosives concentration in the contaminated soil....... ... ..................................... 134









LIST OF FIGURES


Figure page

1-1 Classification of explosives based on composition and performance. .............................29

1-2 Structural formulas and names of nitrocompounds used in the study .............................30

1-3 Chart showing some of the technologies presently available and under development
for the detection of trace explosives ..................................... ..................... ................ 33

2-1 The fate of electronically excited species (AB*) formed after absorption of light,
(h o) ...................................................................................... ....................... .................. 54

2-2 Schematic representation of the photodissociation of a triatomic molecule (ABC)..........55

2-3 Various molecule ionization processes through photon absorption. ................................56

2-4 Schem atic of RE M PI apparatus ........................................ ........................ ................ 57

2-5 Schem atic of PF-L IF apparatus ......................................... ........................ ................ 58

2-6 Photofragmentation set-up. A) schematic of PF-NO2 fragment collection set-up B)
actual picture of the PF cell .. ..................................................................... ................ 59

2-7 Nitrogen dioxide detection by colorimetric method. ....................................................60

2-8 Schematic of time-resolved absorbance experimental set-up.......................................61

2-10 Results obtained with fast time resolved experiments..................................................64

3-1 Schematic diagram for static Chemiluminescence (CL) device................ ................79

3-2 Schematic diagram for continuous Chemiluminescence (CL) device: ........................80

3-3 Lum inol oxidation m echanism ..................................................................... ................ 81

3-4 Lum inol-C L detector set-up ............................................................................. ................ 82

3-5 Schematic of spectrofluorometer used for acquiring luminol-N02 CL emission
sp e ctra ............................................................................................................. ....... .. 8 3

3-6 C hem ilum inescence Set-up. ....................................................................... ................ 84

3-7 L um inol/H 20 2-C L detector .............................................................................. ................ 85

3 -8 D ilu tio n set-u p ................................................................................................................ .. 8 6

3-9 E effect of K O H concentration ....................................................................... ................ 87









3-10 Effect of lum inol concentration on CL signal............................................... ................ 88

3-11 Effect of adding various type of alcohols on CL signal................................................89

3-12 Effect of p-iodophenol concentration on CL signal...................................... ................ 90

3-13 Lum inol-N O 2 E m mission Spectra ........................................ ........................ ................ 91

3-14 Effect of gas sam ple flow rate on CL signal................................................. ................ 92

3-15 Effect of the orientation of luminol-NO2 mist with respect to the PMT window on
sig n a l re sp o n se ................................................................................................................ ... 9 3

3-16 Effect of the storage time of luminol solution on CL signal.........................................94

3-17 Calibration curve of NO2 using Luminol-CL set-up..................................... ................ 95

3-18 Calibration curve of NO2 using Luminol/H202-CL set-up...........................................96

4-1 Experimental set-up used for evaluating the efficiency of CrO3 oxidation unit............101

4 -2 C L sp ectra. ..................................................................................................... .......... 102

5-1 Photofragmentation-Chemiluminescence detector. .....................................115

5-2 Photofragmentation set-up used for NO2 photolysis. ...................................116

5-3 Chemiluminescence signals obtained for nitro-aromatic compounds at varying laser
energy ........................................................................................... ........ .. 117

5-4 Chemiluminescence signals obtained before and after PF of NO2 ..............................119

5-5 Effect of varying laser energy on the CL signal of NOx(x 1,2) fragments obtained after
photofragmentation of A) solid and B) vapor phase of PETN. .................................120

5-6 Effect of varying total number of laser pulses on the CL signal of NOx(x 1,2)
fragments obtained after PF of solid and vapor phase of PETN................................121

5-7 Effect of varying total number of laser pulses on the mass of PETN ablated and on
the CL signal of NOx(x -1,2) fragments obtained after PF of solid PETN ......................122

5-8 Effect of varying laser energy on the mass of PETN ablated and on the CL signal of
NOx(x-1,2) fragments obtained after PF of solid PETN. .................................123

5-9 Calibration curves of PETN, RDX, and TNT............... .........................124

6-1 Cross section of the CL cell ................... .............................................................. 132

6-2 Calibration curve for PETN contaminated soil...... ........... ...................................... 133









6-3 CL Spectra after PF of A) empty sample holder B) pure NaNO2 C) pure NaNO3 ..........135

7-1 Alternative scheme for detecting peroxide-based explosives................ ...................141









LIST OF ABBREVIATIONS


Chemical Compounds

3-APA* 3-aminophthalate
BuOH butanol
DMNB 2,3 dimethyldinitrobutane
2,4 DNT 2,4 dinitrotoluene
DNB dinitrobenzene
EGDN ethylene glycol dinitrate
EtOH ethanol
HMX High Melting eXplosive
MeOH methanol
Luminol 5-amino-2,3-dihydro-1,4-phthalazine dione
NO2 nitrogen dioxide
NO nitrogen oxide
4 NT 4-nitrotoluene
NOx(x-1,2) NO + NO2
PETN Pentaerythridoltetranitrate
RDX Research Department eXplosive
TEA triethanolamine
TNT trinitrobenzene
TNT trinitrotoluene

Techniques

CE Capillary Electrophoresis
CI-MS Chemical Ionization-Mass Spectrometry
CL Chemiluminescence
CRS Cavity Ringdown Spectroscopy
DIAL Differential Absorption LIDAR
DIRL Differential Reflection LIDAR
ECD Electron Capture Detection
El Electron Ionization
FL Fluorescence
FNA Fast Neutron Analysis
GC Gas Chromatography
GC-TEA Gas Chromatography-Thermal Energy Analyzer
HPLC High Performance Liquid Chromatography
IMS Ion Mobilty Spectrometry
IR Infrared Spectroscopy
LIBS Laser Induced Breakdown Spectroscopy
LIDAR Laser Light Detection and Ranging
PD photodissociation
PF-FD Photofragmentation-Fragment Detection
PF photofragmentation
PF-LIF Photofragmentation-Laser Induced Fluorescence










Techniques (continued)


PF-REMPI PF-Resonance Enhanced Multi-photon Ionization
PL Photoluminescence
TNA Thermal Neutron Analysis
TOFMS Time-of-Flight Mass Spectrometry

Units of Measure

mC milli Coulomb
ppm parts in 10 6
ppb parts in 10 9
pptr parts in 10 12
ppmv parts in 10 6 vapor
ppbv parts in 10 9 vapor
pptv parts in 10 12 vapor
SLPM standard liter per minute

Miscellaneous

ArF Argon Fluoride laser
BDE Bond Dissociation Energy
EEG Electroencephalogram
HE high explosives
AHs heat of sublimation
LE low explosives
LOD Limit of Detection
MEMS micro electrochemical systems
NIST National Institute of Standards and Technology
NDF neutral density filter
PD photodiode
PMT photo multiplier tube
SAW surface acoustic wave
(S/N) signal-to-noise ratio









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

LASER PHOTOFRAGMENTATION AND HETEROGENEOUS CHEMILUMINESCENCE
FOR NITRO-BASED EXPLOSIVE DETECTION

By

Maria Pamela Pineda Monterola

August 2007

Chair: James D. Winefordner
Major: Chemistry

The purpose of this research was to develop a simple, fast, reliable, sensitive and

potentially portable explosive detection device employing laser photofragmentation (PF)

followed by heterogeneous chemiluminescence (CL) detection. The PF process involves the

release of NOx(x 1,2) moieties from explosive compounds such as TNT, RDX, and PETN through

a stepwise excitation-dissociation process using a 193 nm ArF laser. The NOx(x-1,2) produced

upon PF is subsequently detected by its CL reaction with basic luminol solution .The intensity of

the CL signal was detected by thermoelectrically cooled photomultiplier tube with high quantum

efficiency and negligible dark current counts.

The research work was divided into four stages: (1) discerning the PF pathways of nitro-

based explosives (2) luminol-CL method development and optimization for improving NO2

sensitivity (3) construction of CrO3 oxidizer for NO to NO2 conversion (4) evaluation of the

device in detecting nitro-based explosives in various media including air and soil matrix.

In the first stage of the research, the following PF pathways were verified using classical

colorimetric analysis and a fast time resolved absorbance method.

R-N02 + hv R + N02; hv =193 nm (1)

N02 + hv NO + O*; hv =193 nm (2)









The second stage involved the development and optimization of NO2 detection using

luminol CL. In this system, a stream of NO2 gas passed through a concentric nebulizer and was

used to aspirate a spray of luminol solution. The aspiration process maximized N02/luminol

contact area enhancing the CL signal. The current system was able to improve the detection

limit through optimization of all the feasible physical and chemical parameters involved in the

CL reaction between luminol and NO2. Detection limit (LOD) of 19 ppt NO2 at (S/N) = 3 was

reported. The optimal reagent solution from the viewpoint of sensitivity of the response to NO2

(maximum signal/signal noise ratio) is 5 x 10-3 M luminol + 0.01 M p-iodophenol + 0.2 M KOH.

A Luminol/H202 CL set-up was also explored where a bundle of porous polypropylene fibers

was used to bring the NO2 into contact with luminol solution. A LOD of 178 ppt NO2 at (S/N) =

3 was obtained.

In the third stage, a CrO3 oxidizer was constructed for NO to NO2 conversion. Photolysis

of nitro compounds produced NO (reaction 2) which is not detectable by luminol-CL system. To

further increase the sensitivity of the device for NO2 fragments, NO must be converted back to

NO2.

Finally, the fourth stage was integration of PF and CL units into an explosive detection

device via a CrO3 oxidizer. The system was able to detect energetic materials in real time at

ambient conditions. Detection limits of 3.4 ppbv for PETN, 1.7 ppbv for RDX, and 34.5 ppbv for

TNT were obtained. It was also demonstrated that the presence of PETN residue within the range

of 61 to 186 ng/cm2 can be detected at a given signal to background ratio of 10 using a few micro

joules of laser energy. The technique also demonstrated its potential for direct analysis of trace

explosive in soil. LOD range of 0.5 to 4.3 ppm for PETN was established, an analytical

capability comparable to currently available techniques.









CHAPTER 1
INTRODUCTION

Background and Significance

Chemistry and History of Explosives

A chemical explosive is a compound or a mixture of compounds which, when ignited by

heat, impact, friction, or shock, undergoes very rapid and self propagating decomposition that

results in formation of more stable material, the liberation of heat, development of a sudden

pressure effect, and production of loud noise and a flash which is called explosion.1

Explosives have been classified based on their composition and rates of decomposition

(Figure 1-1). Low explosives (LE) were prepared by mixing solid oxidizers and fuels, a process

that has remained virtually unchanged for centuries. Propellants and pyrotechnics are included in

this category. Propellants are used in guns and rockets. Most rocket propellants are composites

based on a rubber binder, ammonium perchlorate oxidizer, and a powdered aluminum fuel; or

composites based on a nitrate ester (e.g., nitrocellulose or nitramines) or nitroglycerine. On the

other hand, pyrotechnics include illuminating and signaling flares. Pyrotechnics are composed of

organic oxidizer and metal powder in a binder. Illuminating flares contain sodium nitrate

(NaNO3), magnesium (Mg), and a binder while signaling flares contain barium (Ba), strontium

(Sr), or other metal nitrates. High explosives (HE) were prepared using mixtures of mononuclear

energetic material such as trinitrotoluene (TNT), in which each molecule contain both the

oxidizing and the fuel components. Since, HE act as their own oxidant as well as fuel, their total

energy is much lower than that of LE. However, the rate at which the combustion energy is

released in HE is relatively faster when compared to the release rate of LE. The speed of HE

reactions gives them a greater power than LE. HE detonate at velocity of km s-1 while LE rapidly

burn or deflagrate at relatively low rate range of cm s-1.









HE are further classified as primary and secondary based on their susceptibility to

initiation. Primary explosives such as lead azide and lead stryphrate are highly susceptible to

initiation. Ther are also referred to as initiating explosives because they can be used to ignite

secondary explosives. Secondary explosives, which include nitroaromatics (e.g., TNT),

nitramines (e.g., RDX) and nitrate ester (e.g., PETN), are formulated to detonate only under

specific circumstances. Secondary explosives are often used as main charge or bolstering

explosives.2

PETN, RDX, TNT, NT, and 2,4 NT were the model compounds used in this study. Their

structural formulas as well as the summary of their physical and chemical properties are shown

in Figure 1-2 and Table 1-1, respectively. Other common explosives are also listed on Table 1-2.

Trinitrotoluene (TNT) was first synthesized in 1893 by a German chemist Joseph

Wilbrand.3 The synthesis is performed in two steps. First, toluene is nitrated with a mixture of

sulfuric and nitric acid to produce mono and dinitrotoluence. Next, the mixture of mono and

dinitrotoluence is further nitrated with a mixture of nitric acid and oleum, a more potent nitration

reagent.

TNT is a pale crystalline aromatic hydrocarbon that was originally used as a yellow dye.

Its potential as an explosive was not appreciated for several years mainly because of it does not

detonate easily. In order to detonate, TNT must be confined in a casting or shell and subjected to

severe pressures and/or temperatures (936F), such as from a blasting cap or detonator. In fact,

US army tests on pure TNT show that when struck by a rifle bullet, TNT failed to detonate 96%

of the time and when dropped from an altitude of 4000 feet onto concrete, a TNT filled bomb

failed to explode 92% of the time.3









TNT was first used on a wide scale during World War I. The importance of TNT as a

military explosive is based upon its safety in manufacturing, loading, transportation, and storage.

As of today, TNT is mainly used as military munitions, quarrying activities, and civilian mining.

TNT is a constituent of many explosives such as amatol, pentolite, tetrytol, torpex, tritonal,

piciatol, and edratol. 2

Research Department eXplosive (RDX), or also commonly known as cyclonite or hexogen

was first prepared in 1899 by German chemist Hans Henning.4 Its explosive properties were not

discovered until 1920. In 1940s, RDX was produced by the Bachmann process which reacts

examine with nitric acid, ammonium nitrate, glacial acetic acid, and acetic anhydride. The

crude product is filtered and recrystallized to form RDX.

RDX is a white crystalline nitramine compound. It deflagrates rather than explodes and

only detonates with a detonator. There are many interpretations of its acronyms including Royal

Demolition eXplosive, Research Department (Composition) X, and Research Department

eXplosive. The latter is mostly likely correct. In the UK, new military explosives were given an

identification number preceded by the letters 'RD' indicating 'Research Department Number'.

For some reason, no number was given to this explosive. Instead the letter 'X' was appended to

indicate 'unknown' with the intention of adding the a number letter. Although a number was

issued, the term 'RDX' stuck. 4

RDX was widely used during World War II as a main component of the first plastic

explosive. RDX has been used in several terrorist related explosions in India over the years such

as the serial bomb blast in Mumbai of March 8, 1993 in which more than 300 people were killed

and about 1500 injured. Again, on July 11 of 2006, where a series of powerful explosions took

place in seven suburban Mumbai's Western Railway train line, killing 209 people and injuring









over 700. RDX is a main constituent of many explosives such as Torpex, Tertrytol, Cyclotol,

Composition A, A5, B, C, H6 and C4.2

Pentaerythritol tetranitrate (PETN) was first synthesized in 1891 by Tollens and Wiegand.5

PETN's preparation involves the nitration of pentaerythritol with a mixture of concentrated nitric

and sulfuric acid.

PETN is a white crystalline compound belonging to the same chemical family as

nitroglycerin-i.e., the nitrate esters. RDX is less sensitive than nitroglycerin but is easily

detonated. It is the least stable of secondary explosive.

In 1912, after being patented by German government, the production of PETN started.

PETN was used by the German Army in World War I.5 It is used in detonating fuses (Primacord)

and in a mixture, called Pentolite, with an equal amount of TNT in grenades and projectiles.

During World War II, the Rocket Launcher (commonly known as a Bazooka) charged, with 8

ounce of pentolite, could penetrate up to five inches of armor. PETN is a main constituent of

many explosives such as Datasheet and Semtrex, a Czech-made explosive that has been used

commonly in many terrorist bombings.2 PETN is also used as vasodilator or a heart stimulant.

The medicine for heart disease, "Lentonitrat", is pure PETN.

Significance of Explosive Detection

The field detection of explosives is an extremely relevant analytical issue in law

enforcement and environmental applications.

With the increasing use of explosives by terrorist groups and individuals, law enforcement

and security agents are faced with the problem of detecting hidden explosives in luggage, mail,

vehicles, aircraft, on travelers and so on. Counter surveillance activities also include the

detection of trace explosives during investigation of bombing scenes. Post explosion analysis is a









great importance for such investigations, particularly for drawing connections between cases and

suspects.

The September 11th disaster on the twin towers was one of the most terrible attacks

worldwide. It was a temporary high point of numerous horrible attacks the recent years. Very

often, nitro-based explosives are involved in these attacks including the detonation of 250 g

plastic explosive to bring down Pan American Flight 103 in Scotland in 1998 and the bombings

at the World Trade Center in 1993, Oklahoma City in 1995, Moscow in 1999 and 2002 and

many more.6 To protect society against such attacks, early detection of explosive is needed.

Explosive analysis is also necessary in order to address environmental and/or health related

concerns. Explosive materials exhibit some toxicity, and ingestion or inhalation of explosive

materials may cause a significant health problems. Sites where analysis may be required include

military test ranges and industrial facilities where explosive materials are manufactured in order

to ensure the safety of the workers and nearby residents. In additions, sites may have become

contaminated by improper storage and disposal of explosive materials, and analysis may be

required as part of environmental clean-up.

The detection of landmines is an urgent worldwide problem that requires specific and

effective field detection methods. There are at least 125 million unexploded landmines buried in

70 countries around the world.7 Detection of landmines is essential to prevent the leaking of

undetonated explosives into water and soil. The inherent difficulty in mine clearance is

complicated by the great variety of materials in use. Most modern landmines contain only a

small amount of metal, so that traditional detection techniques based on metal detectors are less

useful.









Review of Methods for Explosive Detection

Extensive efforts have been devoted to the development of effective and innovative

explosive detection systems. There are various techniques currently available and under

development. Explosive detection techniques can be broadly classified into two categories: bulk

detection and trace detection. In bulk detection, a macroscopic mass of explosive material is

detected directly, usually by viewing images by X-ray scanners or similar equipment. In trace

detection, the explosive is detected by chemical identification of microscopic residues of

explosive compound. In trace explosive detection, the explosive residue can be in either or both

of two forms: vapor and particulate. Vapor detection examines the vapor emanating from the

concealed explosive samples and particulate detection examines the microscopic residues of

explosives that would be present on individuals or contaminated materials.

X-ray detection is a well established conventional system where X-ray energy is measured

after passing through an object. This bulk explosive detection approach is capable of

differentiating substances according to their densities and atomic numbers. Commercial X-ray

detection systems used for airport baggage inspection have a throughput rate of up to 800 bags

per hour.7 No quantitative data on sensitivity and false alarm rate have been published for this

method. Other types of bulk explosive detectors are nuclear inducing techniques which used

neutrons instead of photons. Among of these methods are Thermal Neutron Analysis (TNA) and

Fast Neutron Analysis (FNA).

Figure 1-3 illustrates the classification of explosive trace detectors based on the operating

principle of the method. Commonly used explosive detectors can be broadly classified as: (1)

separation-based techniques, (2) mass sensors, (3) electrochemical sensors, (4) biosensors and

(5) optical sensors (6) photofragmentation-fragment detection approaches.









Separation-based techniques include High Performance liquid Chromatography (HPLC),

Capillary Electrophoresis (CE) 8; 9 and Gas Chromatography (GC) combined with Time-of-Flight

Mass Spectrometry (GC-TOFMS)1o, Electron Capture Detector (GC-ECD)11 or Thermal Energy

Analyzer (GC-TEA)12. GC can also be coupled with Ion Mobility Mass spectrometry (GC-IMS)
13-18

GC-ECD has an advantage in quantitative explosive analysis in soil due to its compatibility

with acetonitrile, a preferred solvent for extracting explosive residues from soil. ECD is based on

the high electron affinity of nitro moieties present in the molecules of explosive which readily

capture electrons produced by a radioactive 63Ni source. The decrease in a signal due to the

consumption of free electrons is proportional to the analyte concentration.

In IMS, explosive vapors are converted to ions at atmospheric pressure and these ions are

characterized by their gas phase mobility in a weak electric field. Currently, the most widely

deployed explosives screening technology at airports is IMS, including the Ionscan 400 and

Sabre 2000 which detect trace explosive collected from the contaminated surface of baggage or

paper tickets by either vacuum or wiping.14 Recently, a new IMS injection port inlet has been

developed which collects the explosive sample using solid phase micro extraction.1 Solid phase

micro extraction is a simple adsorption and desorption technique for concentrating volatile

compounds from air.

GC can also be coupled with chemiluminescence detector which is also known as TEA

analyzer. Pyrolysis of explosives produces NO fragments that are consequently reacted with

ozone (03) to produce electronically excited NO2 (N02*). The N02* decays back to its ground

state with emission of chemiluminescence light in the near-infrared region. The emitted light is









detected by a photomultiplier (PMT). The light intensity is proportional to the NO concentration

and hence, to the explosive concentration.

A mass sensor was used as the GC detector in the EST-4200 Ultimate Vapor Tracer, a

hand held instrument used in selected airports here in the U.S. It can also be an explosive

detector by itself when combined in an array, which is often referred to as "electronic noses".

Mass sensors include surface acoustic wave (SAW) detectors, normally coated with different

semi-selective polymeric layers and micro electromechanical systems (MEMS), including micro

cantilever detectors. Mass sensors typically adsorb the chemicals of interest into the surface and

the device detects the change in mass. The detection can be through changes in acoustic waves

propagating along the surface, as in the case of SAW, or by actual bending or a change in shape

of the device as mass is accumulated as in the case of MEMS.

Electrochemical sensors (ECD) detect signal changes in an electrical current being passed

through electrodes that interact with the target chemicals. The fact that nitro-based explosives

can be reduced into amines makes them an ideal candidate for electrochemical detection.

Electrochemical sensors can be classified as potentiometric (measure of voltage), amperometric

(measure of current) and conductometric (measure of conductivity).

Biological explosive detectors, including specially trained dogs are considered the most

successful and widespread system for explosive detection at airports and public areas. In addition

to canines, other animal species have been proposed as alternative methods of biological

explosive detection. A research project in Tanzania trains African Giant Pouched rats to detect

landmines. Reports indicate that rats maybe capable of detecting similarly low level of explosive

odor compared to dogs.19 Bees are also being studied as a biological explosive detection system.

It has been demonstrated that bees have a sense of smell comparable to dogs and are capable of









detecting explosives odors at concentrations lower than most instruments.20 The bees can be

imaged or traced to the source, or more commonly, used to survey areas by examining the

chemical residues brought back to hives. Advantages include that they can be trained quickly and

will not set off any mines. Limitations include the bees do not fly at night, in heavy rain, or in

cold weather (below 40F). 20

The immunosensor is a class of explosive biosensors and involves the use of antibodies as

the biosensing element. Reaction takes place between a target analyte and a specific antibody.

The basis of assay is the change in fluorescence emission intensity of a fluorescently labeled

TNT analogue pre-bound to an anti-TNT antibody. The change in intensity occurs due to

competitive displacement of the labeled TNT by the TNT from the samples.

Optical techniques under investigation include absorption based detection. This is the

simplest method among different spectrometric methods. Several kinds of color change chemical

sensors have developed for rapid onsite detection of explosives. On the other hand,

photoluminescence (PL) detection is based on monitoring the luminescence of an electron rich

semi-conducting polymeric film on exposure to the analyte in a flowing air stream. The PL is

quenched on exposure to nitroaromatic compounds by electron transfer.

In Laser Induced Breakdown Spectroscopy (LIBS), a short laser pulse is focused on the

sample. Laser energy heats, vaporizes, atomizes, and ionizes the sample material, generating a

small plasma plume. As the plasma decays or cools down (approximately 1 [is after the laser

pulse), excited atoms and ions in the plasma emit a secondary light, which is collected and

spectrally resolved by the spectrophotometer, then analyzed by a light detector. As a result, the

multi-elemental composition of the sample can be determined. For explosive detection, different









H, 0, and N emission intensity ratios were studied to identify the analyte as organic explosive,

organic non-explosive or non-organic sample.

The Terahertz (THz) explosive sensor is based on differential absorption. The given item

or region is illuminated by THz radiation containing at least two distinct frequencies. The

frequencies of radiation depend on the THz spectra of the targeted explosives and are chosen so

as to maximize the contrast between the presence and absence of explosives. THz spectroscopy

has not only the capability to detect the presence of explosive but can also determined the

specific type of explosive as different explosives have unique THz specific finger prints.

Raman spectroscopy can also be used for detecting explosive vapor by monitoring the NO2

stretching modes between 1300 and 1370 cm-1. This method is based on the inelastic light

scattering effect due to interaction of light with a sample. This technique is hampered by the

weak Raman signal.

Stand-off technologies under development include Laser Light Detection and Ranging

(LIDAR), differential absorption LIDAR (DIAL), and differential reflection LIDAR (DIRL) for

imaging and LIBS as well.

Cavity ringdown spectroscopy (CRS) is a suitable direct technique for vapor detection that

passes a high intensity pulse tunable laser beam through highly reflective mirrors, which form

the cavity. The intensity of the time dependent decay of the light leaking through the mirror is

measured.

The Photofragmentation-Fragment Detection (PF-FD) approach is based on detecting NO2

moieties after photofragmentation of explosive compounds. Photofragmentation is usually done

by pyrolysis or UV laser irradiation. NOx(x 1,2) fragments are then detected by TOFMS,

fluorescence, or TEA analyzer.









Table 1-3 summarizes the results obtained with different types of explosive detectors

except for PF-FD approach which will be discussed comprehensively in succeeding chapters.

Challenges on Explosive Detection

There are several obstacles to fast and easy vapor explosive detection. The vapor pressures

of explosives are low (Table 1-1 and 1-2). The equilibrium vapor pressure of most explosive at

ambient temperature ranges between 10-4 to 10-9 torr and corresponds to equilibrium vapor

concentrations in the region of 10 tg/L down to 10 ng/L.20 This vapor concentration is measured

for pure samples at equilibrium in a closed container. However, in the field, the environment will

not be a closed system and may contain possible interferents. In addition, the deliberate

concealment (e.g. hard top case, suicide bombers, etc.) of the explosive influences the detection.

As a rough 'rule of thumb', a method that is suitable for direct explosive vapor detection should

be able to detect concentrations down to less than 1 ng/L (ppb level).20

The limited sample size available for analysis is not only a critical problem with explosive

vapor but also with contaminated soil samples as well. Current methods used for detection of

explosive in soil such as GC-ECD and HPLC are inadequate due to their slow detection time and

tedious sample preparation.

Scope of Research Dissertation

There is no single explosive detector that promises speed, selectivity, and sensitivity all

together. Although each of the methods mentioned above exhibits one or more inherent

advantages and disadvantages, their common feature is a high degree of complexity which

hinders their potential to be miniature mass-deployable devices. IMS are the only

commercialized handheld trace explosive detectors usually employed in airports. IMS

instruments are quite expensive and use a radioactive source (63 Ni), which is not environmental









friendly. Because of these limitations, there is a pressing need for improved instrumental

technique for trace explosive detection.

The focus of this dissertation is on the development of a simple, fast, reliable, sensitive and

potentially portable explosive detection device employing laser photofragmentation followed by

chemiluminescence.

In Chapter 2, generation of NO2 fragments and its further dissociation into NO and O*

was confirmed after photolysis of nitro-containing explosives by a 193 nm ArF laser using

classical colorimetric analysis and a fast time-resolved absorption experiment, respectively. In

Chapter 3, NO2 sensitivity was improved using luminol-chemiluminescence technique. In

Chapter 4, an oxidation unit was constructed and evaluated for NO to NO2 conversion. Chapter 5

involved the integration and evaluation of the photofragmentation and chemiluminescence units

as a single device for detecting nitro-based explosives. In Chapter 6, the capability of the system

to directly detect an explosive in soil was evaluated. General conclusions and recommendations

for future directions are described in Chapter 7.









LOWExlosiv


HIGHEaoi


mU
^^^PriaryI^^
explosives^^


ma


Figure 1-1. Classification of explosives based on composition and performance.


Exlsie













02N 1 N02 N02



NO2 NO2 NO2

TNT 2,4 DNT 4 NT



/N02 NO2
0



/ 0 2N NN02
o PFTN RDX


Figure 1-2. Structural formulas and names of nitrocompounds used in the study.










Table 1-1. Physical and chemical properties of nitrocompounds used in the study.
Properties TNT RDX PETN


Chemical Formula

IUPAC name


C6H2(NO2)3CH3

2-methyl-1,3,5-
trinitrotoluence


C3H6N606

1,3,5-trinitro-1,3,5-
triazacyclohexane


C5HsN4012

1,3-dinitro-2,2-
bis(nitramethyl)
propane


Molar mass (g/mol)


Appearance


Density
(g/cm3 at 20C)

Melting point (C)

Boiling point (C)

Solubility in water
at 20C (mg/L)

Solubility in




Vapor pressure at
25C (torr)

Shock sensitivity

Friction sensitivity

Explosive velocity
(m/s)

RF Factor*


yellow needles


1.65


80.4


Decomposes at 295


ether,benzene,acetone,
pyridine



5.8 x 10 -6 (7.7 ppb)


insensitive

insensitive

6900


1.00


white crystalline solid


1.82


205.5


Ignites at 234


Acetone,
acetronitrile



4.6 x 10 -9 (6 ppt)


low

low


8750


1.60


white crystalline solid


1.77


141.3


Decomposes at 190


benzene, sparingly
solouble in
alcohol,ether and
acetone

1.4 x 10 -8 (18 ppt)


medium

medium


8400


1.66


* RF factor is a measurement of an explosive's power for military demolition purposes. TNT
equivalent is a measure of the energy released from the detonation of a nuclear weapon or a
given quantity of fissionable material, in terms of the amount of TNT which could release the
same amount of energy when exploded.


227.131


222.117


316.140










Table 1-2. Physical and chemical properties of some nitro-based explosives.
Explosive class Explosive/ Symbol Formula Molar mass Vapor pressure
(amu) at 25C (torr)


Acid salt


Aliphatic nitro


Aromatic nitro


Nitrate ester


Nitramine


Ammonium nitrate

nitromethane

2,3-Dimethyl-
dinitrobutane/
DMNB

2-Nitrotoluene/
o-MNT

4-Nitrotoluene/
p-MNT

2,4-Dinitrotoluene/
DNT

2,4,6Trinitrophenol/
Picric acid

Ethylene glycol
dinitrate/EGDN

Trinitroglycerin/
NG

Tetranitro-N-
methylamine/Tetryl

Tetranitro-
tetrazacyclooctane/
HMX

Hexanitro-
hexaai sowurzitane/
CL20


NH4N03

CH3NO2

C6H12N204


C7H7NO2


C7H7NO2


C7H6N204


C6H3N307


C2H4N204


C4H5N309


C7H5N508


C4H8N808


C6H6N12012


N/A not available.


80.04

61.04

176.17


137.14


137.14


182.14


229.11


152.06


227.09


287.15


296.16


5.0 x 10 -6

2.8 x 10 1

2.1 x 10 -3


1.5 x 10 -1


4.1 x 10 -2


2.1 x 10 -4


5.8 x 10 -9


2.8 x 10 -2


2.4 x 10 -5


5.7 x 10 -9


1.6 x 10 -13


438.19


N/A














CE GC-MS GC-ECD GC-TEA IMS


Absorption


DOGS


RATS


BEES


PF-REMPI-TOFMS

PF-REMPI-ION
PROBE

Pyrolysis-
Chemiluminescence


Immunosensors


Figure 1-3. Chart showing some of the technologies presently available and under development
for the detection of trace explosives.


HPLC









Table 1-3.Comparative table showing summary of the performance of various explosive detectors reviewed in this study.
Transducer Field of Explosive Detected Limit of detection Comments References
Application
HPLC-UV Aquifer samples TNT 0.10 ng/mL time consuming and 21


Absorption, 254 nm


RDX


0.10 ng/mL


tedious sample
preparation


GC-Laser Ionization
Time-of-Flight-MS


collected in the
vicinity of military
installation
Laboratory liquid
samples

laboratory vapor
sample


TNT, TNB

tetryl


NB

2,4-DNT


60, 160 [g/L

200 [ig/L


17-24 ppb; (S/N=2)

~40 ppb; (S/N=2)


time consuming and
tedious sample
preparation, costly


bulky


4-NT

2,4-NT,TNT

RDX

HMX


10 ig/kg

1 jg/kg

3 jg/kg

25 jig/kg


time consuming
and tedious sample
preparation
costly; not
environmental
friendly


PETN,NG 10-40 jg/kg


GC ECD


soil sample









Table 1-3. (Continued).
Transducer Field of Explosive Detected Limit of detection Comments References
Application
GC-TEA laboratory vapor 4-NT 137 tg/L Poor performance 12
sample for explosive vapor


2,4-NT


18.2 tg/L detection, bulky,


Electrospray-IMS


laboratory vapor
sample


TNT

TNT

RDX


HMX

EGDN


GC-IMS (multi-
capillary
chromatographic
column)


Corona discharge-
IMS


laboratory liquid
sample


PETN

TNT


2,4-DNT


laboratory solid
sample


PETN

TNT

RDX


22.7 [g/L


15 [g/L IMS in general has
a higher % of false
20 tg/L positive and
negative responses;
86 tg/L low mass
spectroscopic
190 [g/L resolution; uses
radio active source
which is not
environment
friendly
8 ng/L


14 ng/L

16 ng/L


8 x 10-11 g

7 x 10 11

3 x 10 -10
7x0lg









Table 1-3. (Continued).
Transducer Field of Explosive Limit of detection Comments References
Application Detected


MEMS


Electrochemical
Electrochemical
Electrochemical


Canines


Immunosensors


Absorption



PL based


laboratory vapor
sample


soil sample
marine water
soil extract and
ground water


airports, public
places
seawater


laboratory
samples of
contaminated
finger prints
air and sea water


PETN,RDX,TNT


RDX
TNT
RDX
TNT
2,4-DNT ,4-NT


almost all


TNT


DNT
DNB


TNT
picric acid


low femto gram level
520 ppt


0.12 ppm
25 ppb
60 ppb
0.11 ppm
0.15 ppm


approximately low
pptr level


250 ppt


N/A



4 ppb TNT in air, 1.5
ppt TNT in sea water,
6 ppb picric acid in
sea water


Limited applications
due to noisy chemical
background

In general, electro-
chemical sensors suffer
from limited sensitivity
and mobile electrolyte.
Also, electrodes can be
easily fouled

subject to fatigue and
behavioral variations


Limited
application,anti-bodies
are not re-usable,
unreliability


Applicable only for
qualitative analysis


bulky, relatively low
sensitivity


22,23


24,25,26


20,27









Table 1-3. (Continued).
Transducer Field of Explosive Detected Limit of detection Comments References
Application
FL based laboratory samples TNB,TNT,DNB, 1 ppm for all of 31


Water samples


LIBS


Tetrahertz


Raman


Optic probe Raman


Mid-IR CRS


Laboratory samples


Laboratory solid
samples


Laboratory vapor
sample


Laboratory samples
of contaminated
finger prints

Laboratory vapor
sample


Laboratory vapor
sample


tetryl,2,4-DNT
DNP
DNT,
NB


HMX,PETN,HMX,
TNT,C4,M43


RDX


2,4-DNT


RDX, PETN


TNT,RDX,PETN


TNT


these explosives
1 x 10 -6 mol/L
40 ppb
17-24 ppb


not available


not available


5 ppb


[ig range


not available


75 ng/L


qualitative analysis


qualitative analysis


34,35


qualitative analysis


qualitative analysis


qualitative analysis


bulky


FL based









CHAPTER 2
PHOTOFRAGMENTATION PATHWAYS OF NITRO COMPOUNDS

Introduction

This chapter begins with the discussion of the basic theory of photofragmentation (PF)

followed by a review of previous works elucidating the PF pathways of nitro compounds using

Resonance Enhanced Multi-photon Ionization Time-of-Flight Mass spectrometry (REMPI-

TOFMS) and Laser Induced Fluorescence (LIF). Generation of NO2 fragments and their further

dissociation into NO and O* was confirmed after photolysis of nitro-containing explosives by

193 nm ArF laser using classical colorimetric method and fast time-resolved absorption

experiment, respectively.

Basic Theory of Photofragmentation

Photochemical processes can be classified based on the fate of electronically excited

species formed after absorption of light. As shown in Figure 2-1, when a molecule (AB) has

absorbed a quantum of radiation, (hv), it becomes 'energy rich or excited' in the process. There

are several pathways by which an electronically species (AB*) may lose its energy. Some of

these pathways are through chemical reaction of the reactive fragments formed, isomerization or

ionization processes.40 The process whereby the excited state species is split into simpler

fragments is called photofragmentation (PF) or photodissociation (PD).

There are two important PF rules that may be stated as follows:

(1) Only the light absorbed by a system is effective in producing a PF in a molecule.
This is also known as the Grottus-Draper Law. 41
(2) Each photon absorbed activates one molecule in the primary excitation step of a
photochemical sequence.41

Over the past four decades, various theoretical models and experimental techniques have

been developed to advance the understanding of molecular dissociation induced by photon

excitation. For instance, theoretically constructed potential energy curves or surfaces for PF of









diatomic or polyatomic molecules have been widely used to predict or explain the experimental

phenomena.42 As an example, consider the following PF for a hypothetical molecule ABC.

ABC + hv AB + C (2-1)

In Figure 2-2, the ground and excited state potential energy surface are termed G and E,

respectively while R is the internuclear distance between atom C and the AB moiety and hv is

the energy of the laser photon. The ground state potential energy surface is bound and evolves at

long AB-C internuclear distances to an asymptotic energy which corresponds to the energy

difference of the separated AB and C fragments. The energy difference between the ground state

of the ABC molecule and the ground state energy of the two AB and C fragments is the bond

dissociation energy (DE). Following the absorption of a photon (of energy hu), the ground state

molecule is, in this case, excited to a repulsive excited state, (E). Depending on the repulsive

character of the excited state, the molecule will dissociate promptly or more slowly, into the

fragments AB and C. The excess energy, which is the difference between the energy of the

photon and the bond dissociation energy, may be distributed among all available degrees of

freedom of the nascent fragments translationall and internal). The release of kinetic energy

causes the AB and C fragments to move away from the point in space where they were

generated. The proportion of the excess energy released as kinetic energy and the masses of the

AB and C fragments determine their recoil speeds. The residual excess energy is divided over the

other degrees of freedom of the fragments and may, for instance, lead to rotational, vibrational,

and/or electronic excitation of the fragments. By considering the law of conservation of energy,

an expression for the partitioning of energy can be written as follows:

AE ex = hI) + EPint Do(AB-C) = EE + ET + Ev + ER (2-2)









where available or residual energy (AE ex) consists of the photon energy (hu) plus the internal

energy of the parent molecule (EPint) minus the energy Do(AB-C) required to break the B-C bond.

This residual energy is partitioned into the electronic (EE), translational (ET), vibrational (Ev) and

rotational (ER) degrees of freedom of the fragments.42

Basic Theory of REMPI-TOFMS and LIF

Propellants and explosives contain molecules with characteristic functional groups, namely

the NO2 moiety bonded to carbon, (-C-NO2 for nitrocompounds), oxygen, (-C-O-N02 for

nitric ester) or nitrogen ((-C-N-NO2 for nitramines). This functional group is weakly bound to

the main skeletal portion of the parent molecule by approximately 40 to 60 kcal/mole depending

upon the specific molecule.43 It is this moiety that is responsible for the weak and structureless

absorption observed in the UV region near 230 nm.44

Resonance Enhanced Multi-photon Ionization (REMPI) technique typically involves a

resonant single or multiple photon absorption to an electronically excited intermediate state

followed by another photon which ionizes the atom or molecule. An ion and a free electron will

result if the photons have imparted enough energy to exceed the ionization threshold energy of

the system. Typically, small molecules composed of light atoms will have ionization energies

around 10-15 eV, corresponding to the absorption of three UV photons from the ground state of

the neutral molecule.45 The cross section of the ionization process is greatly enhanced if there is

a real excited state resonant at the energy of one or two absorbed photons (Figure 2-3).45

Multi-photon fragmentation and ionization can occur by two distinct mechanisms. First,

molecules absorb photons, exciting them to a dissociative state below the ionization limit. If the

laser pulse duration is much shorter than the lifetime of dissociative state, then the up pumping

rate may be so high that ionization occurs before dissociation. The ions may absorb additional

photons before fragmenting. This process is ionization followed by dissociation (ID) or ladder









climbing.45 However, if the laser pulse duration is longer than the dissociative state lifetime, then

the laser intensity is sufficiently high, these fragments may absorb additional photons to further

dissociate or ionize. This process is dissociation followed by ionization (DI) or ladder switching.

Thermally labile compounds, which include many nitro compounds, fragment primarily by the

DI route while many aromatic compounds fragment by ID route alone.45

A schematic of a typical REMPI apparatus is presented in Figure 2-4. A gas phase sample

is introduced into the apparatus, typically in the form of a molecular beam. A high power, pulsed

laser is used to induce multi-photon absorption, resulting in fragmentation and ionization.46 The

resulting ions are then detected by Time-of -Flight-Mass Spectrometry (TOF-MS).

Laser Induced Fluorescence (LIF) is simply the optical emission from molecules that have

been excited to higher levels by absorption of laser radiation. LIF has been used to study the

electronic structures of the molecules and to make quantitative measurements of their

concentrations.46

LIF for explosive detection is usually combined with photofragmentation in order to

dissociate a target molecule and subsequently detect the fluorescence of the generated fragments.

This is done because the fluorescence of most explosives is weak while the fluorescence

intensity of NO(x 1, 2) fragments produced after photolysis is strong.47

Figure 2-5 shows a schematic of a PF-LIF experimental system. A UV laser is used to

dissociate and induce fluorescence in the sample. A photomultiplier (PMT) detects the LIF signal

which can be averaged and recorded on a digital oscilloscope. The laser wavelength can also be

scanned to collect LIF spectra. These spectra can be used to identify the wavelength of the

highest intensity LIF signal.









PF Mechanism of Explosives based on PF-REMPI-TOF and PF-LIF

In 1983, Butler, et al. 48 studied the dissociation of nitromethane (CH3NO2) at 193 nm

using LIF and molecular beam photofragment translational energy spectroscopy which is similar

to REMPI-TOFMS except for the addition of an electron ionization (EI) source at the entrance of

TOF analyzer. The primary photodissociation process is shown to be cleavage of the C-N bond

to yield CH3 and NO2 radicals with quantum yield almost equal to unity. The translational energy

distribution for this chemical process indicates that there are two distinct mechanisms by which

CH3 and NO2 radicals are produced. The dominant (major) mechanism produces CH3 and NO2 in

its electronically excited (A 2B2) state. The electronically excited NO2 has sufficient internal

energy to further dissociate to NO and O* as follows:

NO2 (A 2B2) -- NO (X2HI) + 0 (1D) (2-3)

In the minor mechanism, the NO2 fragment is produced in the excited (B 2B2) state, which has a

large cross section for absorbing another photon. The subsequent absorption of an additional

193 nm photon results in further photodissociation of NO2 into NO and O* as shown below.

NO2 (B 2B2) + hv "- NO (A2+); hv = 193 nm (2-4)

Almost at the same time, Blais 49 obtained the same results as Butler et al. and determined

that about 30% of the NO2 produced from primary photodissociation participates in the second

absorption. Fragments from photodissociation were observed at m/z values of 15, 16, and 30

corresponding to CH3, 0, and NO. A very small signal at m/z = 46 (NO2) was measured. The

NO2+ signal was significantly lower than that of NO+ signal, presumably due to extensive

fragmentation of vibrationally excited NO2 molecules in the ionizer.

In 1984, Renlund 50 reported the decomposition of nitromethane (CH3NO2), nitrobenzene

(C6H5NO2), and n-propylnitrate (CH3(CH2)20NO2). The emission of C*, CH*, and N02* was

monitored from 380 to 650 nm. Of particular note is the behavior of the NO2 yield as a function









of laser fluence. In the case of nitromethane and n-propylnitrate, the NO2 yield varies linearly

with fluence until it saturates; however, saturation was not observed with nitrobenzene, but

instead the NO2 yield increases more sharply with higher fluence conditions. It was concluded

then that by increasing the rate of energy deposition, the yield of NO2 can be accelerated

suggesting that the primary dissociation of nitrobenzene into C6H5 and NO2 may change as a

function of laser intensity.

In 1989, Capellos et al. 51 used PF-LIF technique to study the photodecomposition of RDX

vapor at 428K using KrF laser with 248 nm and 15 ns pulse radiation. They observed prompt

emission at 305-320 nm and at 380-420 nm following photolysis of RDX. Based on the spectra

recorded, they attributed the UV emission to electronically excited OH (2 +) and the

structureless visible emission centered at 608 nm to electronically excited NO2 (2B2). Monitoring

the signal intensity as a function of laser fluence revealed that NO2 (2B2) was formed by a one-

photon process while OH (2 +) was formed by a two-photon process. The production of NO2

(2B2) was claimed to result primarily from the N-N bond scission of electronically excited RDX

and OH (2 +) possibly via dissociation of vibrationally excited nitrous acid (HONO) which is

generated by a H atom rearrangement in RDX, followed by five member ring formation and

HONO elimination.

In 1993, Galloway et al. 52 used PF-REMPI-TOFMS in studying the pathways and kinetic

energy disposal of the photodissociation of nitrobenzene. At several photolysis wavelengths

between 220 and 320 nm, the following photodissociation pathways for nitrobenzene were

observed:

C6H5NO2 C6H5 + NO2 (2-5)

C6H5NO2 C6H5NO + 0 (2-6)









C6H5NO2 C6H50 + NO (2-7)

C6H5NO2 C5H5 + CO + NO (2-8)

The relative yield of the pathways producing NO2 and NO varies strongly with the photolysis

wavelength. The production of NO2 exceeds that of NO by about 50% for the 280 nm photolysis,

but increases to almost six fold excess in 222 nm dissociation.

Lemire et al.53; 54(1993), Marshall et al. 55(1994), and Ledingham et al. 56(1995) used the

same technique and all confirmed the characteristic fingerprint at m/z = 30 (NO ) for

nitrobenzene; 2,4 DNB; TNT; EGDN; PETN; RDX; and Semtrex sample at X= 226 nm.

Simeonsson et al. 57(1994) repeat the same experiment using a 193 nm laser instead. The

same mass spectra were obtained except for a higher signal intensity at m/z =30 (NO), which can

be attributed to the higher pulse energy of an argon fluoride (ArF) laser (X = 193 nm) compared

to a dye laser (X = 266 nm).

In 1996, Moyang et al. studied the photodissociation of nitrocellulose by means of

electron impact ionization ion trap mass spectrometer (EI-MS). In this article, nitrocellulose was

irradiated at 532, 355, and 266 nm and with a tunable dye laser operating in the range of 310-330

nm. The first mass spectrum obtained with 532 nm laser light only showed mass peaks

corresponding to background gases: water, nitrogen and oxygen. However, irradiation with 355

nm light produced three major peaks at m/z = 19, 30, and 48, corresponding to protonated water

(H30 ), nitric oxide (NO ), and hydrated nitric oxide complex (NO-H20 ). When 266 nm laser

light was used, the intensities of the mass peaks occurring at m/z 30 and 48 increased

approximately four times in comparison with the same peaks observed using 355 nm light. The

wavelength dependence of removing the NO2 moieties from its parent molecule (or denitration)

in the 310-330 nm range has been measured, and the NO yield normalized by laser intensity









showed a linear decrease of NO yield with increasing laser wavelength. All of these results

suggest the simple trend that shorter wavelength UV light induces more denitration. The weight

loss rate of sample due to denitration induced by 266 nm laser light was measured to be 4.00.4

mg J-1. This value represents a 5.8% photon-to-nitro group conversion efficiency. A laser

intensity threshold about 6 mJ cm-2 was observed; above which thermal decomposition occurs

(detonation of sample occurs).

Research on photofragmentation of nitro-containing compounds within the last 20 years

did not experimentally confirm the detection of NO2 fragments which is fundamentally essential

in supporting the primary photodissociation of nitro compounds as shown below:

R-NO2 R + NO2 (2-9)

PF-REMPI-EI-TOFMS during 1980's claimed that NO2 was not detected due to extensive

fragmentation of NO2 molecules in the ionizer while PF-REMPI-TOFMS during the 1990's

attributed the absence of NO2 signal due to the absence of a real excited state resonant with the

energy of two absorbed photons to ionize NO2 to NO2 However, NO2 fragments were detected

by its prompt fluorescence emission. Detecting NO2 using this method is not that reliable since

the NO2* emission is irradiated over a large spectral region (from visible to near IR) where other

excited fragments are also expected to emit.

The work presented in this dissertation was able to detect directly NO2 fragments from

explosives after their photolysis with an ArF laser by a classical colorimetric method.

Furthermore, the photodissociation of NO2 fragments into NO and 0* was confirmed using a

time resolved absorption experiment.









Experimental Section


Reagents and Chemicals

Explosive samples. Five percent (v/v) solutions of PETN, RDX, and TNT in acetone

were provided by Naval Surface Warfare Center (MD, USA). Acetone was allowed to evaporate

at room temperature to obtain the crystalline samples. Solid reagent grade chemicals (99%

purity) of 4-NT and 2,4 NT were used and purchased from Sigma Aldrich (St. Louis, MO).

Colorimetric experiment. All chemicals used were analytical reagents from Fischer

Scientific Inc. (Hampton, NJ) unless otherwise stated. Deionized water from Millipore Milli-Q

deionizing system was used for preparing all the reagent solutions.

Triethanolamine (TEA) solution was prepared by dissolving 15 g of TEA in 1 L of water.

Standard Nitrite solution was prepared by dissolving 0.2 g dried NaNO2 (105C for 1

hour) in 100 mL of water. Exactly one mL of this solution was diluted using TEA solution to get

4 [tg/mL of nitrite which serves as a stock solution.

Neutral red solution was prepared by dissolving 0.1 g of neutral red (Basic red 5 C.I.

50040) in 1 L of water containing 2.5 mL of 4.25 M H2SO4. The dye solution was stable for

thirty days.

1 % (wt./v) Sodium hypophosphite(NaH2PO2) was prepared by dissolving 1 g of

NaH2PO2 in 100 mL of water.

1% (wt./v) Potassium bromide (KBr) was prepared by dissolving 1 g of KBr in 100 mL

of water.

Time-resolved absorption experiment. Copper powder (325 mesh) was purchased

from Alfa Products (Danvers, MA) and reacted with 98 %(v/v) nitric acid (HNO3).









Apparatus and Methodology

Photofragmentation-N02 fragment collection. Figure 2-6A shows the

photofragmentation (PF) set-up used for NO2 fragment collection. The set-up consists of the PF

cell containing the explosive sample, an argon fluoride (ArF) laser to photolyze the sample, and

a delivery gas system to purge all the photofragmented products into the glass bubbler containing

TEA solution. The PF cell (Figure 2-6B) is made of cylindrical stainless steel with an O.D. of

1.93 cm and a length of 4.85 cm. Both ends of the PF cell were enclosed with quartz window (D

= 1.80 cm) to allow passage of UV radiation into the cell. The delivery gas system was coupled

into the PF cell through solenoid valves (Asco Inc., NJ) that were attached to both sides of the

cell. A tank containing ninety nine percent pure N2 gas was attached to the entrance of solenoid

valve I while the exhaust end of solenoid valve II connects the PF cell to the gas bubbler. The

gas flow rate was controlled by a mass flow controller manufactured by Alicat Instrument Inc.

(Tucson, AZ).

In these experiments, solid explosive sample was deposited and tightly packed into a small

aluminum cylindrical cup (O.D. = 1.9 mm; Depth = 4.5 mm). The sample was then securely

placed into the circular quartz disk having its center drilled and hollowed to exactly fit the

sample holder. This circular disk serves as the backside window panel of the PF cell.

An argon fluoride (ArF) excimer laser (GAM Laser Inc., Orlando, FL) was used to

generate the photolysis wavelength of 193 nm. Each experimental run used 80 000 laser pulses

with an average energy of 8 mJ/pulse. The pulse length and the repetition rate are 15 ns and 50

Hz, respectively. The beam diameter is 3.1 mm. It should be noted that PF is distinct from Laser-

Induced Breakdown Spectroscopy (LIBS). The laser flux employed in PF is always sufficiently

low to avoid laser induced breakdown of the sample medium or the creation of plasma.









The ArF laser ablated and photolyzed the explosive sample while N2 gas (0.5 SLPM

flowrate) which serves as the carrier gas is continuously introduced to purge the

photofragmented products into the bubbler containing 10 mL of TEA solution. The TEA solution

trapped the NO2 fragments into the bulk solution by converting them to nitrite ions (NO2-) as

shown by the following equation:

N(HOCH2CH2)3 + 2NO2 + 20H- 2NO2- + -0N(OHCH2CH2)3 + H20 (2-10)

Colorimetric analysis. The nitrite ions in TEA solution were then analyzed using

colorimetric method by mixing exactly 2 mL of aliquot solution of TEA into 25 mL standard

flask containing 2 mL of 0.01 % neutral red, 1 mL of 4.25 M H2SO4, and 1 mL KBr. This was

followed by addition of 1 mL of 1 % NaH2PO2 and mixed thoroughly. The mixed solution was

then diluted to mark and the absorbance was measured at 530 nm against distilled water.

Preparation of the Calibration Graph. Aliquots of the standard stock solution of NO2

containing 0 to 85 tg of nitrite were added to a series of 25 mL standard flask containing the

same reagents used for preparing the samples. Absorbance was measured at 530 nm against

distilled water. The calibration graph obtained was linear but with a negative slope.

Nature of Species Responsible for Color. A known excess amount of neutral red on

treatment with nitrite solution undergoes diazotization in acidic medium (Figure 2-7A). KBr acts

as a catalyst for this reaction On treatment with NaH2PO2, the diazonium salt undergoes

deamination and causes a decrease in absorbance proportional to the concentration of added

nitrite ions. The absorption spectrum of unreacted neutral red is shown on Figure 2-7B.

PF-time resolved absorption experiment. Figure 2-8 shows the PF set-up used for

confirming the further dissociation of NO2 fragments into NO and O* upon absorption of a

photon with 193 nm wavelength. The set-up consists of the PF cell containing the NO2 gas









sample, an ArF laser to photolyze the sample, and a detection system to monitor the change in

the profile of the laser pulse before and after photolysis. Neutral density filters were added to

attenuate the laser pulse energy and to prevent the saturation of the ET-2030 photodiode detector

(Electro-Optics Technology, Traverse City, MI). The set-up is capable of monitoring the

intensity versus time profile of the laser pulse after photofragmentation of NO2 gas.

Nitrogen dioxide gas was produced by reacting pulverized Cu with concentrated nitric acid

in a sealed 250 mL Erlenmeyer flask (reaction 11). The NO2 gas generated from this reaction

was collected using a gas syringe and injected into a quartz cell (Diameter = 2.0 cm; Length=

10.0 cm) through its rubber septum cover.

Cu(s) + 4HN03 (aq.) Cu(N03)2(s) + 2 H20 (1) + 2NO2(g) (2-11)

The ArF laser beam was focused into the cell and used to photolyze the NO2 gas sample.

The photodiode detector is triggered by the PMT system to collect the signal from the laser. The

intensity versus time profile of the laser pulse was monitored and plotted by fast oscilloscope

(HP54542C Oscilloscope, Hewlett Packard, Pab Ator, California). The ArF laser was calibrated

using an MPE 2500 power meter from Spiricon Power Products Inc., Logan, Utah.

Results and Discussion

Detection of NO2 Fragments

The calibration graph (Figure 2-9) for NO2 fragment detection was obtained by plotting

absorbance values at 530 nm against the mass of nitrite in [g. Linear dynamic response was

found within the range of 0 to 85 tg NO2-. The calibration plot is a straight line; with correlation

coefficient(R) value of -0.98506 and obeying the following equation:

A (dye) = -0.00753(m) + 0.65734 (2-12)

where A is the absorbance and m is the amount of nitrite in rig. The limit of detection (LOD) was

0.78 tg (0.02 mmol) nitrite. LOD is defined by the following equation:









LOD = (k-sblk) / m (2-13)

where k is the confidence factor which is equal to 3, sblk is the standard deviation of the blank

and m is the slope of the line or the calibration sensitivity.

The precision of the method has been established at 19 tg nitrite which gave a relative

standard deviation (RSD) of 5.66% where n = 10.

The amount of NO2 fragments detected for each solid explosive sample is shown in Table

2-1, which experimentally shows the primary photodissociation of explosives after photolysis at

193 nm. The amount of the sample ablated was determined from another experimental run where

there is no flow of carrier gas. The aluminum cup sample holder was weighed before and after

photofragmentation using a microbalance. The fraction of NO2 detected was calculated by the

following equation:

Fraction of NO2 detected = moles of NO2 detected in TEA solution
NO2 moles ablated (2-14)


The fraction of NO2 detected for all the explosive samples was approximately the same

with the exception of RDX. RDX' significantly higher fraction of detected NO2 might be due to

its PF mechanism as elucidated by Capellos et al.51 Nitramines have a characteristic dissociation

pathway which involve HONO elimination from a parent molecule. HONO produced during

photolysis can be readily trapped and absorbed in a basic TEA solution, subsequently increasing

the fraction of NO2 detected.

The validity of the colorimetric analysis has been established using NO2 gas standards

having concentrations of 5 and 10 ppm. The NO2 gas was purged into the same bubbler

containing 10 mL TEA solution at flow rate of 0.3 SLPM. Exactly two mL of this solution was









analyzed for nitrite content following the procedure described under the calibration graph. The

following equation was used to convert the pg NO2 determined into ppm concentration.

NO2 (ppm) = (mD)/(40.9VMAS) (2-15)

where m is the mass of nitrite ions in rg, D is the dilution factor, V is the volume of the gas in L

collected at standard temperature and pressure, M is the molar mass of NO2 in g/mol, A is the

fraction of NO2(g) absorbed by TEA solution, and S is the fraction of NO2 gas converted to

NO2 ions. The numerical values of A and S are 0.95 and 0.85, respectively.59 Based on this

equation, NO2 concentrations of 6.1+0.0 and 11.90.4 ppm were obtained after analyzing 5 and

10 ppm NO2 standards, respectively.

From these sets of data, the direct confirmation of NO2 fragments was established after

photolysis of explosives at 193 nm. Relatively low fraction of NO2 detected for each sample

indicates the further photodissociation of NO2 into NO and 0* upon absorption of an additional

photon.

Photofragmentation of NO2

Nitrogen dioxide exposed to tropospheric UV radiation photodissociate at X 5 420 nm to

give NO and 0*.60

NO2 + hu NO + O*; hu < 420 nm (2-16)

A fast time resolved absorption experiment has been used to verify the photodissociation

of NO2 at 193 nm. Light from ArF excimer laser was used to decompose NO2 at ambient

conditions. The laser pulse has a duration of 10 ns and delivered output energy of 10 mJ/pulse at

a repetition rate of 20 Hz. The oscilloscope was set to obtain the average of 32 pulses for each

run of the experiment under closed batch system. The NO2 concentration diluted in air was

approximately around 120 ppm or about 3% in terms of the partial pressure of NO2 gas with









respect to the atmospheric pressure. The concentration of NO2 was calculated using the

following equation:

% NO2 (PN2 /Patm) = ART C x 100% (2-17)
0.434cLPN

where A is the initial absorbance (0.7), R is the universal gas constant (0.08206 L-atm/mol-K), T

is the room temperature (298K), C is the conversion factor (1000 cm3/L), c is the absorption

cross section of NO2 at 193 nm (2.0 x 10-19 cm2)61, L is the length of the PF cell (10 cm), N is

Avogadro's number (6.02 x 10 23 molecules/mole).

Figure 2-10A shows the profile of the ArF laser pulse before and after the photolysis of

NO2 gas sample. Plot A represents the actual profile of the laser pulse. Plot B represents the

profile of the laser pulse after passing it through an empty PF cell and an added neutral density

filter. Plot C represents the profile of the laser pulse after passing it through the PF cell

containing NO2 gas sample. From these sets of data, the time-resolved absorbance plots were

obtained (Figure 2-10B). If the absorption of light is uniform during the entire duration of the

pulse and the detector responds linearly to light, then the plot of time-resolved absorbance versus

time will show a constant value for the entire duration of the pulse. This is exactly the case after

adding a neutral density filter which is shown by the Ai(t) plot. On the other hand, if the first part

of the pulse induces a dissociation process and the fragments produced do not absorb, the shape

of the pulse will be distorted since now the second part of the pulse will show no absorption. The

overall result would be a non-constant behavior of the time-resolved absorbance plot during the

entire duration of the pulse as shown by the A2(t) plot.











Conclusion

The following PF pathways were confirmed using classical colorimetric analysis and fast

time resolved absorbance.

R-N02 + hv R + N02; h = 193 nm (2-18)

NO2 + ho NO + 0*; h = 193 nm (2-19)

Confirmation of NOx(x 1,2) generation after photolysis of nitro-based explosives was the basis for

developing an explosive detector based on PF-Fragment detection scheme












AE+B


AB + hi) AB*


Direct reaction




BA
-J Isomerization




AB+ + e-
Ionization



A+B
Dissociation


Figure 2-1. The fate of electronically excited species (AB*) formed after absorption of light,
(ho).



















AAEx .
V(R) V __ -
V F AB + C

hv



O 1 ABC (G)


R

Figure 2-2. Schematic representation of the photodissociation of a triatomic molecule (ABC). R
is the distance between C and the AB diatomic fragment, DE is the bond dissociation
energy and hlu) is the energy of the photon. Through absorption of a photon, the
molecule is excited to a repulsive state via which it dissociates. The excess energy
(AEex) is the energy difference between the photon energy and bond dissociation
energy.




















AX'

AX*







AX


Figure 2-3. Various molecule ionization processes through photon absorption. A) Ionization via
direct one-photon abortion B) non-resonant multi-photon ionization and (c)
resonance enhanced multi-photon ionization (REMPI). Process C) is an example (2 +
1) REMPI scheme; two photons are required to excite the molecule AX to a resonant
excited intermediate state AX* and one more to ionize AX*.

























sample I
container


Data acquisition and
analysis system


laser power meter


Figure 2-4. Schematic of REMPI apparatus.











tunable
UV laser


data
analysis
computer


sample
cell


filter


PMT







Digital
oscilloscope


Figure 2-5. Schematic of PF-LIF apparatus.














I Mass flow controller I


PF.......... cell
PF cell


Legend :
SV = solenoid valve

B


I


TEA sol'n


Figure 2-6. Photofragmentation set-up. A) schematic of PF-NO2 fragment collection set-up B)
actual picture of the PF cell


I N2 gas














NO2 + f Nr Br T N 2
2H I + 2H20
N2 2N N N N2+ Br

neutral red dye
A max= 530 nm [H+l

N

SN2Br-

colorless




B

0.75


a 0.50 -
C-

0.25
0


0.00
150 300 450 600 750 900

wavelength (nm)


Figure 2-7. Nitrogen dioxide detection by colorimetric method. A) species responsible for color
B) absorption spectrum of neutral red dye.













ArF Beam
laser splitter Focusing lens


NDF = neutral density filter


Figure 2-8. Schematic of time-resolved absorbance experimental set-up.


* Channel 1 -triggering output
4 Channel 2 data output














0.7- BLANK


0.6- Adye =-0.00753(pg NO2-) + 0.65734
R -0.98506

E 0.5-


s 0.4-
II
x
E 0.3-


0.2-


o 0.1
.~0

0.0--
0.0 20.0 40.0 60.0 80.0 100.0
pg NO2
Figure 2-9. NO2- calibration curve.









Table 2-1. Colorimetric analysis of NO2 fragments after photofragmentation of some nitro-based
explosives.
Explosive sample Actual NO2 NO2 detected % NO2 detected
ablated(pfmol)a (plmol)b
4 nitrotoluene 1.040.22 0.1400.008 13.51.4
2,4 nitrotoluene 1.240.05 0.1900.011 15.33.8
trinitrotoluene 0.780.06 0.1500.008 19.25.8
PETN 7.960.10 1.0000.056 12.60.7
RDX 2.460.22 1.690.096 68.77.3

a based on the ablated mass of the explosive sample obtained by weighing the sample before
and after photofragmentation using microbalance
b amount of NO2 detected using colorimetric method













0.10


0.08


0.06


0.04


0.02


0.00

-30



B 0.7-


0.6-

0.-

0.5-
0


-20 -10 0 10 20 30
time (ns)


V \ Legend:
IA = initial intensity
IB = intensity with NDF
Ic = intensity with NO2


A2(t) = -log (Ic/IA)


A,(t) = -log (IB/IA)


-4 -2 0 2 4 6 8 10
time (ns)


Figure 2-10. Results obtained with fast time resolved experiments. A) profile of ArF laser pulse
before and after NO2 photofragmentation B) time resolved absorbance plots for NO2
photofragmentation.









CHAPTER 3
NO2 DETECTION BY CHEMILUMINESCENCE

Introduction

Highly sensitive detection of nitrogen dioxide gas (NO2) is essential not only for air

monitoring purposes, but also for developing explosive detection. Common explosives can be

detected based on their characteristic functional group, specifically the NO2 moieties that can be

photofragmented from the parent molecule using a UV laser. This chapter begins with a

discussion of the theory of chemiluminescence, followed by a review of previous NO2 detection

systems, and finally, an evaluation and assessment of a newly developed instrument based on

luminol chemiluminescenece (CL) for the analysis of NO2 fragments from explosives.

Basic Theory of Chemiluminescence

Chemiluminescence (CL) is defined as the production of electromagnetic radiation (UV,

Vis, or IR) as a result of a chemical reaction.62 A significant fraction of reaction intermediates or

products are produced in excited electronic states and emit light on returning to the ground state.

The following reactions are three common sequences used to produce CL:

A+B I +I*; I* I+ hv P (3-1)

A+B P+P*; P* P+hv (3-2)

I*(orP*)+F I (orP)+F F + hv (3-3)

where A, B are reactants, I is an intermediate, P is the product, and F is a fluorescence acceptor.

Reactions 1 and 2 illustrate that the luminescing species can be an intermediate or a product in a

reaction. Reaction 3 represents sensitized CL in which the excited species produced transfers its

energy to an efficient fluorophore acceptor which then fluoresces.63

The CL quantum efficiency (OcL) is defined as:

(OCL) = (number of photons emitted/number of photons reacted) (3-4)









It can be represented as the product of two efficiencies or

(OcL) = Oex OL (3-5)

where 0ex is the efficiency of the production of excited species (the fraction of reacting

molecules that produces an excited molecule) and OL is the luminescence efficiency (or the

fluorescence quantum efficiency) of the luminescing species.63

For CL to occur, three general conditions must be met. First, there must be sufficient

energy to produce an excited state. Thus, the reaction must be exothermic such that

-AG > (hc/Xex) > (2.86 x 104) / Xex (3-6)

where AG is the free energy change (kcal mol-1) of the reaction and ),x is the upper wavelength

limit in nm for excitation of the luminescing species. For CL emission in the visible region, -AG

must be 40 to 70 kcal mol-1.62 Second, there must be a favorable pathway to produce the excited

state. Third, photon emission must be a favorable deactivation process for the excited product in

relation to other competitive non-radiative process.62

CL has the potential for being the basis of chemical analysis. Chemiluminescence

detection is based on measuring the emission of photons from the excited molecules. Often, it is

possible to arrange reaction conditions such that the luminescence signal is related to the analyte

concentration.

The analytical performance of CL techniques is governed primarily by the chemistry of

the reaction utilized and the luminescence characteristics of the emitting species. Careful control

of the reaction conditions is mandatory. By chemical parameter optimization, the analytical

advantages of CL including high sensitivity, low detection limits, and wide linear ranges can be

maximized and achieved with very simple equipment. These attributes also result from the fact

that no light source is required, significantly eliminating the background noise from the source.









CL is up to 100 000 times more sensitive than absorption spectroscopy and at often 1000 times

more sensitive than fluorometry.64

One attractive feature of CL technique is the simplicity of the instrumentation. All that is

required is a reaction cell, a detector, and a signal processor-readout system. The emitted light

intensity is measured by means of a photomultiplier (PMT). Because only one emitter is

normally present in the reaction medium, no monochromator is required to discriminate among

different wavelengths.

There are two ways of measuring CL, namely: (1) static mode and (2) continuous flowing

stream mode.63 In the static mode, (also known as batch or discrete sampling), the CL reagents

and the analyte are mixed rapidly and the CL intensity versus time profile is measured (Figure 3-

1). The initial increase in intensity is due to the finite time for mixing of reagents or an induction

period in the reaction. The decay of the signal is caused by the consumption of the reagents and

changes in the CL quantum efficiency with time. The analytical signal can be taken as the peak

signal or the signal after some fixed delay time from the point of mixing, (the integral of the

signal over some period) or the signal (area) of the entire peak.

In the continuous flow mode, the CL reagents and the analyte are continually pumped

and combined at tees and sent to a flow cell or mixed and observed in an integral reactor flow

cell as shown in Figure 3-2. When the cell is totally filled with the reaction mixture, a maximum

CL signal is obtained. Then, the system will equilibrate reaching a steady state signal. The signal

represents the output integrated over the residence time of the reaction mixture in the cell.

Luminol-N02 Chemiluminescence

While there are many chemical reactions that are chemiluminescent, the oxidation of

luminol (CsH7N302, which is also known as 5-amino-2,3-dihydro-1,4-phthalazinedione) is one of

the most commonly known. Luminol was first synthesized in 1893, and its ability to yield a









chemiluminescent reaction in basic solution in the presence of an oxidant such as hydrogen

peroxide was first reported in 1928 by H.O. Albrecht.65

The luminol CL system consists of water, base (sodium or potassium hydroxide), oxidant,

catalyst, and luminol. A variety of oxidants including permanganate, hypochlorite, hydrogen

peroxide, and nitrogen dioxide have been used in the luminol reaction. Typical catalysts include

peroxidase, hemin, and transition metal ions (Co 2+, Cu 2+, Fe 3+ etc.).66

The CL reaction used for NO2 detection is based on the same principle. NO2 is detected by

an increase in CL intensity that results when NO2 reacts with an alkaline solution of luminol. The

CL reaction is as follows:

2C8H7N302 + 40H- + 2NO2 2C8H504 + 3N2(g) + 4H20 + hv (3-7)

The CL spectrum has maximum emission intensity at 425 nm with an approximately 100 nm

bandwidth.67 Despite intensive studies, the mechanism of luminol oxidation in water has not

been fully elucidated. However, the general outline of the reaction has been agreed upon and is

shown in Figure 3-3. Base oxidizes the luminol leaving negative charges on nitrogen atoms

which move onto the carbonyl oxygen to form what is known as enolate anion. The oxidizing

agent next performs a cyclic addition to the two carbonyl carbons which leads to formation of 3-

aminophthalate (3-APA*).68 White et al. have shown that the CL spectrum of luminol matches

the fluorescence spectrum of the 3-aminophthalate anion and, hence, identifying it as the

emitting species.69 Since then, several other investigators have confirmed the identity of the

emitter. 70-72

Review of CL Methods for NO2 detection

The detection of nitrogen dioxide gas (NO2) by means of chemiluminescent reaction with

luminol in aqueous solution has been the subject of considerable interest and development .The

first prototype detector employing this approach was developed in 1980's by Maeda et al.73 by









directly reacting the NO2(g) at the surface of continuously flowing luminol solution. The intensity

of luminescent radiation emitted by the reaction was monitored by photomultiplier (PMT)

detector. This system was very sensitive to movements of the compartment and had relatively

slow time response. An alternate design for chemiluminescent detection of NO2(g) was

subsequently presented by Wendel et al.66;74, who allowed a stream of NO2(g) to pass over filter

paper drenched with basic luminol solution. The CL reaction occurs on the surface of the filter

paper which is adjacent to the PMT window. A fabric wick has also been used instead of filter

paper to make a compact commercial NO2 detector (Luminox LMA-3, by Scintrex/Unisearch

Associates Inc.); a detailed description of this instrument was given by Schiff et al.75 and Kelly et

al.76 A further variation on the luminol chemiluminescent detector for N02(g) was introduced by

Mikuska and Vecera;77 this configuration employed a continuous spray of luminol solution,

immediately below the PMT. CL was generated from the stream of analyzed gas. Another

detection scheme modification was made by Collins and Pehrsson78 wherein N02(g) reacted on

the surface of a glass containing immobilized luminol molecules within a hydrogel or polymeric

thin film positioned in front of a PMT tube. Yet another variation was developed by Robinson

et al.79 where a bundle of porous polypropylene hollow fiber membranes was used to bring the

NO2(g) into contact with luminol solution. Chemiluminescence occurring within the translucent

hollow fibers was detected using a miniature PMT tube.

The present work was adapted from the chemiluminescent aerosol detector of Mikuska

and Vecera77 for its simplicity, efficiency and high sensitivity. The aim of this work is to

evaluate, assess and further improve the analytical capability of detection of NO2(g) with the

luminol reaction.









Experimental Section


Reagents and Chemicals

Luminol powder of 98% purity (5-amino-2,3-dihydro-1,4-phthalazinedione) was obtained

from Acros Organics (New Jersey, USA) and used without further purification. Potassium

hydroxide, methanol, ethanol, tert-butanol, hydrogen peroxide, sodium bicarbonate, and sodium

carbonate were all obtained from Fischer Science Company (Pittsburg, PA). Para-iodophenol

was obtained from Sigma Aldrich Chemical Company (St. Louis, MO). Deionized water was

obtained by passing distilled water through a Millipore Milli-Q deionizing system

All gases purchased were of the highest purity. A tank of 99% nitrogen gas was purchased

from BOC Gases Inc. (Murray Hill, NJ). Nitrogen dioxide gas buffered with nitrogen gas was

purchased from Safety Products Inc. (Lakeland, FL).

Apparatus and Methodology

Luminol-CL detector. Figure 3-4 is the instrumental set-up used for measuring and

optimizing the chemiluminescence signal arising from the basic luminol solution and nitrogen

dioxide gas. The three main sections are a gas delivery system, reaction chamber and measuring

electronics. The gas delivery system consisted of the gas samples and mass flow controllers

(Alicat Instrument Inc.,Tuczon.AZ) used for dilution. The chemiluminescence reaction set-up

was adapted from the NO2 atmospheric monitoring system developed by Mikuska and Vecera.77

A stream of NO2 gas passed through a concentric nebulizer (Type C-Q nebulizer manufactured

by Precision Glass Blowing, Centennial, Colorado) allowing the basic luminol solution to be

aspirated into the reaction chamber and dispersed to form a fine mist with a large interface area

for NO2 and luminol reaction. The aerosolized luminol solution is directed toward the bottom

wall of the reaction cell with PMT orthogonal to the direction of the mist's propagation. The

reaction was done in a batch mode which last for twenty seconds. The luminescent radiation









resulting from the reaction between the luminol alkaline solution aerosol and NO2 gas is detected

by a cooled R3896 photomultiplier tube (PMT).The PMT has a spectral response range from 185

to 900 nm with maximum sensitivity at 450 nm. The PMT is operated at -850 V and a

temperature of 10 to 15C. The photocurrent signal from the PMT is amplified by Kiethley

Model 6514 electrometer which is connected to a computer by a GPIB-USB-HS digital-analog

converter from the National Instrument Laboratory. EXCELINK software is used to control all

the data acquisition. All data are then processed using Origin 6.1 and Excel software.

The reaction cell is made of a Teflon tube with an inner diameter of 2.10 cm and a length

of 4.57 cm. It is encased in an aluminum housing to completely seal the reaction from external

light sources. One end of the reaction cell is open to the PMT tube, separated only by a fused

silica window with the diameter of a 1.50 cm.

The concentric nebulizer is connected to the side of the reaction cell and oriented parallel

with respect to the PMT detector.

CL emission spectrum set-up. A Fluorolog-Tau-3 spectrofluorometer (Jovin Yuon

Inc., Edison, NJ) in chemiluminescence mode (excitation source turned off) is used to obtain the

emission spectra. A block diagram of the instrument is depicted in Figure 3-5. In the first set-up,

a 25 ppm NO2 standard gas was purged into the quartz cell containing 0.01 M luminol in 1 M

KOH. In the second set-up, a peristaltic pump (Rainin, Model Rabbit) was set at desired flow

rate to deliver H202 solution into the quartz cell containing luminol solution.

Gas-liquid exchange module set-up. A diagram of the gas liquid exchange module is

shown in Figure 3-6A. This device was provided by Dr. John W. Birks from University of

Colorado. The device is made up of microporous membrane allowing the NO2 gas to react with

alkaline luminol-H202 solution. The membrane consists of a bundle of approximately fifty









polypropylene capillary fibers having an outer diameter of 380 rim, a wall thickness of 50 inm,

and a pore size of 0.2 [im (Akzo Nobel, Scottsboro, AL). The fibers are encased in a polymer

housing and the entire unit will be referred as the gas-liquid exchange module. The actual picture

of the set-up is shown in Figure 3-6B.

The luminol solution flows through the interior of the fibers, while the NO2 gas flows

around the exterior of the fibers. The aqueous solution remains in the interior of the fibers due to

the hydrophobic nature of the polypropylene. Gases diffuse through the pores of the membrane

and into solution, resulting in the chemiluminescent reaction. NO2 was detected above the

background signal due to the reaction of H202 and luminol. The fibers and the polymer are

translucent, allowing the transmitted light to be detected by a PMT.

Luminol/H202-CL detector. Figure 3-7 is a schematic diagram for a luminol/H202-CL

detector. The gas-liquid exchange module is used as the CL reaction cell, and a miniature head-

on R647 photomultiplier tube (PMT) tube is used to detect the light emitted. The PMT has a

spectral response range from 300 to 650 nm with a maximum sensitivity at 420 nm. The PMT is

operated at -850 V at ambient temperature. The photocurrent signal from the PMT is amplified

by a Kiethley Model 6514 electrometer which is connected to a computer by GPIB-USB-HS

digital-analog converter from National Instruments. EXCELINK software is used to control all

the data acquisition. All data are then processed using Origin 6.1 and Excel software.

A series of peristaltic pumps is used to pump the luminol and H202 solutions into the gas-

liquid exchange module. The two solutions are mixed in a tee just prior to the exchange module.

The flow rate was set at 1.0 L/min for each solution. The gas outlet is connected to the exhaust

fume hood.









Dilution set-up. NO2 standards in ppb range were prepared by dynamic dilution of 5 to 25

ppm standard NO2/N2 mixtures. Pure N2 gas was used as the diluent. A block diagram of the set-

up is shown in Figure 3-8. Gas flow rates were controlled by mass flow controllers. An in-line

gas mixer filled with 4 mm glass beads housed in a polystyrene casing was used to ensure

efficient gas mixing. Mass flow controllers were calibrated electronically using a Humonics

Optiflow 650 Digital flow meter and manually using a GC bubble flow meter.

Results and Discussion

In order to have a highly sensitive method for NO2 determination based on its

chemiluminescence reaction with luminol, both the physical and chemical parameters of the

system were optimized. Physical parameters are related to overall design of the reaction

cell/detector configuration, whereas chemical parameters are affected by the preparation of the

chemical components of the basic luminol solution.

Chemical Parameters

The optimal solution composition for NO2 detection was determined by systematically

varying the concentration of potassium hydroxide (KOH) and chemiluminescent enhancer

reagents such as p-iodophenol, methanol, ethanol and t-butanol. The effect of the concentration

of these reagents on chemiluminescence intensity was studied by varying the concentration of

one component over a range while holding the other components constant.

The effect of KOH concentration on the emission intensity is shown in Figure 3-9A. As

the concentration of KOH increased, the decay rate of the reaction becomes faster, which

consequently decreased the integrated chemiluminescence signal. By obtaining the integral or

area of the entire peak for each corresponding signal from Figure 3-9A and plotting it against

the pOH, the optimum KOH concentration of 0.2 M can be inferred which corresponds to









pH=13.1 as seen from Figure 3-9B. The result is consistent with previous studies 777 verifying

that excess OH- quenches the chemiluminescence radiation.

Maeda et al.73 obtained a maximum CL signal for luminol concentrations of 3 x 10-4 M to 3

x 10-3 M with no variation within that range. Mikuska and Vecera77 as well as Wendel 74 agreed

with this observation. Kok et al.80 obtained peak chemiluminescence intensity at 2.4 x 10-4 M

luminol and a 50% decrease for an order of magnitude change about this value. The result of our

experiment is presented in Figure 3-10 where a peak occurs for 5 x 10-3 M luminol which is in

agreement with previous results 73;77

Based on the experimental results acquired by Wendel et al., the addition of water soluble

alcohols such as methanol, ethanol or t-butanol to the alkaline luminol solution resulted in a

two-fold increase in the chemiluminescence signal with N02(g). Based on our results, addition of

alcohols had negligible effect on the reaction, as can be seen in Figure 3-11.

Thorpe and et al.81 used phenol derivatives such as p-iodophenol and p-phenyl phenol to

enhance light emission for the reaction between hydrogen peroxide (H202) and luminol. They

were able to observe that the light emitted in the reaction decayed slowly and its intensity was

about 1000 fold greater than the unenhanced reaction. Addition of p-iodophenol to the basic

luminol solution was investigated and it was found out that the presence of 0.01 M p-iodophenol

significantly enhanced the integrated chemiluminescence signal up to 10 times, as shown in

Figure 3-12.

The mechanism of the p-iodophenol enhancement of chemiluminescence signal is still

unresolved. Here, it was shown that the emission spectra of phenol-enhanced and unenhanced

reactions are remarkably similar (Figure 3-13), suggesting two plausible hypotheses that (1) p-

iodophenol as a chemiluminescent enhancer reagent is neither a more efficient emitter nor









fluorescent acceptor, but exerts its action earlier in the complex reaction between luminol and the

oxidant and (2) p-iodophenol enhanced the chemiluminescence signal of aerosolized luminol-

NO2(g) system by significantly decreasing the dielectric constant of the aerosolized luminol

solution, which enhances the desolvation of luminol enolate anions on the surface of the aerosol

and thus, increasing the frequency of interaction of the reagents involved. This is supported by

the fact that the signal enhancement using p-iodophenol can only be observed with the

chemiluminescence aerosol detector and not by simply bubbling the N02(g) into the bulk of the

basic luminol solution as seen on Figure 3-13B and 3-13C,respectively.

Physical Parameters

Gas and liquid flow rates. The gas flow rate was varied with a mass flow controller

and it was noted that above 0.7 SLPM (-24 psi), the chemiluminescence signal saturates (Figure

3-14). The luminol uptake at this gas flow is 1.10 mL/min. Since the response was independent

of gas flow rate, the instrument is a concentration detector rather than a mass detector. A flow

rate of 0.7 SLPM was used for further studies.

Cell design. The change in chemiluminescence intensity as a function of the orientation

of the concentric nebulizer with respect to the detection window was investigated. It was

observed that if the N02-luminol mist entered parallel to the PMT window, a signal enhancement

of about 8 fold occurred compared to introduction of the N02-luminol mist perpendicular to the

detection window (Figure 3-15).

Two different sets of reaction cell were used for this experiment. The set-up described under

the methodology section was used for the experiment where the orientation N02-luminol mist

was parallel with respect to the PMT window which will be referred as Set-up A. For the set-up

where the N02-luminol mist was oriented orthogonal to the PMT window, a completely different

reaction cell of the same design and configuration as Set-up A was used except that the distance









between the nebulizer and the detection window was approximately four times farther than Set-

up A and a stream of N2 gas was allowed to blow continuously on the PMT window as the

reaction was in progress. In this way, it could be argued that the significant difference in the

signal was due to impaction of aerosolized solution on the surface of the PMT window.

Shelf life of the alkaline luminol solution. It was observed from previous studies 74; 80 that

luminol-NO2 chemiluminescence intensity increased with the age of the luminol solution. Hence,

a controlled study of the chemiluminescence intensity as a function of time from the preparation

of the solution was made. An optimized luminol solution was prepared (5 x 10-3 M luminol, 0.01

M p-iodophenol in 0.2 M KOH) and the chemiluminescence reaction from a 5 ppm standard

source of NO2 was measured. This measurement was done as soon as the solution was made and

repeated periodically for at least 2 months. According to Wendel et al.74, the chemiluminescence

intensity increased by as much as 20-fold over the course of days, and after this initial "aging"

period, it was found out that the signal did not change appreciably over longer storage period.

However, our results (Figure 3-16) did not observe this trend. Instead, it was observed that the

basic luminol solution gave a slight change in chemiluminescence readings from day 0 up to day

71.

Performance of Luminol-CL Detector

All the optimized parameters obtained were then applied to construct the calibration curve

for NO2 (Figure 3-17). A calibration working curve was generated by making three replicate

measurements of the NO2 standards. The linear range of the detector was measured using known

concentrations of NO2 from 33 ppbv to 25 ppmv. The calculated detection limit with a signal to

noise ratio of 3 was 19 ppt NO2. A calibration sensitivity of 7.08 x 10 -8 C/ppb NO2 was

obtained. The non-zero intercept is a reflection of the background of the instrument.









Performance of Luminol/H202 CL Detector

The luminol/ H202 CL detector was tested for direct detection of NO2. NO2 reacted with

H202 to form a potent oxidizer peroxynitrite (ONOO-) as shown by the following reaction:

2N02 + H202 + OH 20N02 + H20 (3-8)

Peroxynitrite is capable of oxidizing luminol to an excited state of 3-aminophthalate, which

relaxes by emitting light as can be seen in reaction 3-9.82

20N02 + luminol -- 3-aminophthalate + N2 + hv (3-9)

The NO2 signal resulting from peroxynitrite formation is detected on top of the background

signal resulting from the slow oxidation of luminol by H202. Optimal conditions obtained

previously were used to generate the calibration curve for NO2. The concentrations of H202 and

luminol in a bicarbonate/carbonate buffer (with pH=10.25) were 2 mM and 4 mM, respectively.

Responses of the detector to NO2 standards in the range of 33 ppb to 25 ppm are depicted in

Figure 3-18. The working curve was generated by making three replicate measurements of the

NO2 standards. The calculated detection limit and the calibration sensitivity were 178 pptr NO2

and 2.43 x 108 C/ppb, respectively.

Luminol CL Detector Compared to Other CL Set-ups

The LOD ratio of luminol CL detector to luminol/H202 CL detector is 9 which can be

attributed to the following reasons: (1) in the first set-up surface interaction between the NO2 gas

and luminol solution was maximized through nebulization (2) the first set-up also had a

significantly lower background signal due to the absence of H202 and the peristaltic pumps (3)

the side-on PMT used for the first set-up was more sensitive compared to the head-on PMT used

in the second set-up.









In general, the luminol-CL detector gave a better LOD compared to other existing CL set-

ups as shown on Table 3-1.

Conclusion

The luminol-CL detector for quantifying NO2 is based on the NO2 CL reaction with

aerosolized basic luminol solution. The aerosol is formed by continuous nebulization of reagent

solutions by the stream of analyzed gas in a sealed reaction chamber. The intensity of the CL

signal was detected by a PMT. The current system improved the detection limit obtained

previously by others optimization of all the feasible physical and chemical parameters involved

in the chemiluminescence reaction between luminol and NO2. A detection limit of 19 ppt NO2 at

(S/N) = 3 was obtained. The optimal reagent solution from the viewpoint of sensitivity of the

response to NO2 (maximum signal/signal noise ratio) was 5 x 10 -3 M luminol + 0.01 M p-

iodophenol + 0.2 M KOH. A Luminol/H202 CL set-up was also explored where a bundle of

porous polypropylene fibers was used to bring the NO2 into contact with luminol solution. A

LOD of 178 ppt NO2 at (S/N) = 3 was obtained.

The newly developed CL instrument benefits from superior sensitivity, simplicity, and

reliability, which are very attractive in chemical sensor. It is anticipated that the CL detector

developed can be used as a diagnostic tool and aid in developing a better explosive detector.
























B 4




'I CL


reaction
starts







reaction time -




Figure 3-1. Schematic diagram for static CL device. A) the sample and the CL reagent(s) are
introduced in the reaction cell and the final reagent is injected to initiate the CL
emission, then light is monitored by the PMT detector B) curve showing CL intensity
as a function of time after the reagents are mixed to initiate the reaction.











A sample
gas


reagent
gas


reaction time .


Figure 3-2. Schematic diagram for continuous CL device: A) the sample and the CL reagent(s)
are introduced simultaneously and continuously into the flow cell B) a steady state
signal is achieved when the cell is totally filled with the reaction mixture.











NH, 0 NH, (o
A' H C) ^ \
-OH
H0
Luminol Enolate anion
NH 0 NH20 *
S+hv+N2 : ]
O O
0 0
3-Aminophthalate* (3-APA*)


Figure 3-3. Luminol oxidation mechanism.
































Figure 3-4. Luminol-CL detector set-up.


PMT


I-


U!
reaction:
chamber


NO2(g) buffer
with N2 (9)




aluminum
housing













Excitation
source (turn-off)


monochromator


Parameters
Scanning wavelength : 200-650 nm
Slit width : 10 nm
Excitation source : turn-off
HV (PMT): 850 V
Integration time : 0.3 s


data
cell monochromator PMT pr c r
reaction processor


Figure 3-5. Schematic of spectrofluorometer used for acquiring luminol-N02 CL emission
spectra.















A gas inlet











solution in

O.D. : 380 pm[
Wall thickness: 50 pm
Pore size : 0.2 pm


gas outlet


solution waste



hollow fibers


cross sectional view of gas exchanger


Figure 3-6. Chemiluminescence Set-up II.A) schematic of gas-liquid exchange module set-up
B) the actual picture of the set-up.

























gas
inlet


gas
outlet


- - - - - I -


Peristaltic
pump


gas-liquid exchange
module


Peristaltic
pump


electrometer

4
AID


computer


power supply


Figure 3-7. Luminol/H202-CL detector.


I
solution waste























gas
mixer
reaction

N2gas MFC II cell


Figure 3-8. Dilution set-up.


































0 50 100 150 200 250 300


-0.6


pH = 13.1
0.2 M KOH


-0.3


Figure 3-9. Effect of KOH concentration on
CL signal.


0.3 0.6 0.9 1.2 1.5

-(log [O H ])

A) decay rate of luminol-N02 reaction and B) on

















#5.0 x 103 M luminol








U


2.5


-log([Luminol])
Figure 3-10. Effect of luminol concentration on CL signal.


0.4-



0.3-


0.2-



0.1-


0.0


-4


1.0


1.5


2.0


3.0











0.4-


- 0.3-


0.2-


0.1-


0.0-


1.~ Z ~h~SSSSS~SSS 1 .1 SSSS~S~S~h~ I ~SS ~~SSS~ I I~ ~SS ~SSWS~ 1.


I I
vio alcohol Me(OH)


type of alcohol

Figure 3-11. Effect of adding various type of alcohols on CL signal.


0.5-


'Bu(OH)


/


Ow

o


m


Et(OH)









3.0-


2.5- 1

2.0-

S 1.5-

1.0-
0.

0.5-

0.0 .
0.00 0.02 0.04 0.06 0.08 0.10
p-iodophenol concentration (M)
Figure 3-12. Effect of p-iodophenol concentration on CL signal.









32000 -


28000-


24000-


o. 20000-
0

-16000-


12000- b

-I
0 8000-


4000-



200 300 400 500 600 700

emission wavelength(nm)


Figure 3-13. Luminol-N02 Emission Spectra. A) chemiluminescence spectrum obtained from
reaction of luminol with H202 B) chemiluminescence spectrum obtained from
reaction of luminol with NO2 C) chemiluminescence spectrum obtained from
enhanced reaction of luminol with N02 using p-iodophenol.








0.20-

0.18-

0.16-

0.14-

0.12-

0.10-

0.08-

0.06-


I................ .. .


0.4


0.8


1.2


1.6


NO2 flowrate (SLPM)

Figure 3-14. Effect of gas sample flow rate on CL signal.


2.0
















7.6E-5C9.0E-6

0.0


-0.5-





-J
-1.5 I
PM

-2.0
5.8E-4C8.0E-6

0 25 50 75 100 125 150
time (s)


Figure 3-15. Effect of the orientation of luminol-NO2 mist with respect to the PMT window on
signal response.











0.50 -

0.45 -

0.40 -

0.35 -

0.30-

S0.25- T

( 0.20-

Q 0.15-

0.10-

0.05

0.00 ,
0 10 20 30 40 50 60 70 80
duration of storage (days)

Figure 3-16. Effect of the storage time of luminol solution on CL signal.













-28


-3.2
0




S -3.6

0



-4.0


I I I I I I '
1.5 20 25 3.0 3.5 4.0 4.5

ogFigure 3-17. Calibration curve of N ing Luminol-CL set-up.

Figure 3-17. Calibration curve of NO2 using Luminol-CL set-up.












-3.0-


-3.5-


-4.0-



-4.5 -


-5.0-



-5.5-


2.0


2.5


3 I
3.0


3. I
3.5


Og NO ppb



Figure 3-18. Calibration curve of NO2 using Luminol/H202-CL set-up.


4.0


4.5










Table 3-1 Summary of the LOD obtained from different luminol-NO2 CL set-ups.
Researchers LMN-N02 reaction set-up Detection limit


Maeda et al. (1980)73


Wendell et al.
(1983)74


Schiff et al.
(1986)75


Mikuska et al.
(1992)77


Collins et al.
(1995)78

Monterola et al.
(2007)


Passing NO2 gas on the surface of
continuous flowing luminol solution

Blowing NO2 gas on the surface of
filter paper on which luminol
solution was allowed to flow

Blowing NO2 gas on the surface of
fabric wick on which luminol
solution was allowed to flow

Reacting NO2 gas with aerosol of
luminol solution


Blowing NO2 gas on the surface of
polymeric thin film of luminol

Reacting NO2 gas with aerosol of
luminol solution


50 pptr


30 pptr



5 pptr


0.24 ppm
(lowest concentration
detected)


0.46 ppbv


19 pptr









CHAPTER 4
CONVERSION OF NO TO NO2

Introduction

Photofragmentation of nitro-based explosives using a UV laser produced both nitrogen

dioxide (NO2) and nitrogen oxide (NO) fragments. Detection of these fragments indicates the

presence of explosives. However, NO cannot be detected by luminol chemiluminescence and

hence must be converted back to NO2 to enhance the sensitivity for explosive detection. Several

methods for NO to NO2 conversion has been reported for atmospheric monitoring purposes. This

chapter begins with a review of those methods and is followed by a description of the oxidation

unit used to integrate the photofragmentation and chemiluminescence set-ups into a single device

for explosive detection.

Review of Methods for NO oxidation to NO2

Several methods have been used for oxidizing NO to NO2. Among these are potassium

permanganate and sulfuric acid (KMnO4/ H2SO4), iodine pentoxide (1205), acidified MnO2,

persulfate, ozone (03), heated chlorine dioxide (C102), oxygen (02), and chromium oxide

(CrO3).82 Each method has certain disadvantages especially during real time analysis. Persulfate

and KMnO4/ H2SO4 are not quantitative oxidizers while acidified MnO2 is not stable.82 An

excess of 03 will convert NO all the way to nitric acid (HNO3) in the presence of moisture.83

Also, 03 cannot be used as an oxidant because it produces an interfering signal with luminol

solution. A disadvantage in the use of I205 is that it is sensitive to many other reducing agents.

C102 has been demonstrated to yield good conversions, but must be generated in situ which is

quite inconvenient. Gaseous 02 oxidizes NO slowly and can only be used for concentrations

above 100 ppm.83









CrO3 coated on an inert substance has been shown to give the best results for NO to NO2

conversion. Oxidation efficiency close to 100 % was achieved in the range of 30 to 80 % relative

humidity.82-84

The work presented here described the fabrication of an oxidation unit for conversion of

NO to NO2 conversion that was used to integrate the photofragmentation and chemiluminescence

units into an explosive detector device.

Experimental Section

Reagents and Chemicals

Reagent grade CrO3 (-98%) was purchased for Sigma Aldrich (St. Louis, MO). Glass

beads (D = 4 mm) were purchased from Fischer Scientific (Hampton, NJ). A tank of 10 ppm NO

and 10 ppm NO2 both buffered with N2 were purchased from Airgas South Inc. (Gainesville,

FL) and Safety Products Inc.(Lakeland, FL), respectively.

Apparatus and Methodology

Cr03 Oxidation unit. The CrO3 oxidation unit was prepared by coating CrO3 on glass

beads. Glass beads were chosen since it is an inert substance and beads were used to provide

high surface area. The first step for preparing the CrO3 converter was soaking the glass beads in

a 20 %(w/v) aqueous solution of CrO3 for about 30 minutes. After soaking, the glass beads were

filtered and dried in an oven at 105C for 30 to 45 minutes. The glass beads were then packed in

a 1" diameter by 8" length stainless steel tubing.

Cr03 Oxidation Unit/Luminol-CL Detector Tandem. Figure 4-1 is the experimental

set-up used for evaluating the efficiency of the CrO3 oxidation unit. A 10 ppm NO in N2 was

used as reference sample. The gas sample flow rate was controlled by the mass flow controller

which was then connected to the CrO3 oxidation unit. Oxidized products were then introduced









into the CL unit through the side arm of the concentric nebulizer. The operating conditions of the

instrument were essentially the same as those described in Chapter 3.


Results and Discussion

The CrO3 oxidation unit was tested for efficiency in converting NO to NO2 by passing a

10 ppm NO/N2 sample through the converter. The sample flow rate was set to 100 mL/minute.

The NO2 produced upon oxidation was detected by the luminol-CL detector. Figure 4-2A shows

the CL signal acquired with 10 ppm NO in N2 without CrO3 oxidation unit whereas Figure 4-2B

shows the signal obtained after passing the same gas sample through the CrO3 converter. Figure

4-2C shows the signal observed after passing 10 ppm NO2 in N2 through the same converter.

Based on the three replicate measurements, average peak areas for 10 ppm NO and 10 ppm NO2

after passing through the same converter were 857.1127.5 and 1488.5225.3, respectively.

Conversion efficiency of 58% was obtained. The lower conversion efficiency can be attributed to

relative humidity of the NO sample. It is well established that the efficiency of NO oxidation by

CrO3 varies with sample water content, being greatest at mid-range relative humidity (30 to 80

%) but much poorer near both extremes. 82-84

Conclusion

CrO3 oxidation unit has approximately 58% NO to NO2 conversion efficiency which was

anticipated to significantly enhance the sensitivity of the Photofragmentation-Chemilumines-

cence apparatus for explosive detection.





















I P MFC


Luminol solution


S Optical window
oxidation --
unit


Legend :
MFC = mass flow controller


Figure 4-1. Experimental set-up used for evaluating the efficiency of CrO3 oxidation unit.















-1 1 0 -

-120

-130

-140-

-150

-160 -

-170
-2


5-

0
-5
-5 -
-l
-10

-15

-20

-25

-30

-35 -


0 0 20 40 60 80 100 120 140
time (s)


-20 0 20 40 60 80 100 120 140 160 180
time (s)


10 -

0 -

-10-

-20-

-30-

-40-

-50-

-60-

-70

-20 0 20 40 60 80 100 120
time (s)


Figure 4-2. CL spectra of A) 10 ppm NO/N2
oxidation unit C) 10 ppm N02/N2


140 160 180


B) 10 ppm NO/N2 after passing through CrO3
after passing through CrO3 oxidation unit.









CHAPTER 5
LASER PHOTOFRAGMENTATION AND CHEMILUMINESCENCE FOR NITRO-BASED
EXPLOSIVE DETECTION

Introduction

Common explosives contain NO2 functional group and since there are relatively few

naturally occurring sources of nitro compounds, its presence can be used as a signature for

explosive detection. Explosive detection using Photofragmentation-Fragment Detection (PF-FD)

has been explored. This chapter begins with a review of those methods and is followed by a

description of a new instrument based on laser photofragmentation and luminol

chemiluminescence.

Review of PF-FD Methods for Explosive Detection

The PF-FD approach most often utilized when the analyte molecule does not lend itself to

direct spectroscopic detection. In general, atoms and small molecules (composed of 2 to 3 atoms)

can be detected directly by absorption, fluorescence, or ionization techniques.44 This is due to a

favorable combination of usually strong optical transitions and sharp, well resolved spectral

features that provide effective optical activity. However, it is often the case for larger molecules

such as the explosive compounds that the transitions are weaker and the spectral features are

broad and poorly defined. In these cases, direct detection of the molecule by any of the above

techniques is not analytically practical. While direct detection may not be feasible, the PF

products of explosive such as NOx (x 1, 2) can be readily detected. Since NOx (x 1, 2) fragments are

characteristic of the chemical composition of the explosive compounds, they also contribute to

the selectivity of the method.

A patented instrument based on PF-FD technique for explosive compounds was reported

by Nguyen et al.85;86 The instrument is based on coupling pyrolyis with luminol

chemiluminescence detector. Explosive samples were collected and pyrolyzed under









electronically heated Pt-Rh wire at 700C. The NO2 fragments that were generated after

pyrolysis were swept into a reaction cell containing basic luminol solution. The luminol solution

was separated from the remainder of the reaction cell by a semi-permeable hydrophobic PTFE

membrane. The membrane was permeable to NO2, thereby allowing diffusion to bring the NO2

into contact with luminol solution. Light emitted from the reaction was detected by a

photomultiplier tube (PMT) coupled to the reaction cell. The method claimed a detection limit of

100 ppt for DMBA.

There are several disadvantages regarding the pyrolysis method. First of all, pyrolysis

produced many decomposition pathways, one of which is the detonation reaction of the

explosive samples. TNT, for instance, is known to detonate at 502C and upon detonation, TNT

decomposes as follows: 3

2C7H5N306 + A 3N2 + 5H20 + 7CO + 7C (5-1)

Reaction 5-1 significantly diminished NO2 fragments detectable. Secondly, pyrolysis is a time

consuming method which is definitely a drawback for a fast real time analysis.

Due to limitations of the pyrolysis method, research on PF-FD techniques was pursued

using laser photofragmentation. Laser photofragmentation followed by fragment detection using

Resonance Enhanced Multi-photon Ionization Time-of-Flight Mass Spectrometer (REMPI-

TOFMS) or Laser Induced Fluorescence (LIF) was not only performed for elucidating the

photodissociation pathways of explosive compounds but for real time explosive monitoring as

well.

Lemire et al.53 and Simeonsson et al.54 used PF-REMPI-TOFMS with a laser operating at

226 nm to photofragment the parent explosive compound and photoionize the resulting NO

fragments. The NO+ ions were subsequently detected using a TOFMS. The technique was









demonstrated using nitrocompounds such as nitromethane, DMNA, RDX, NB, o-NT, m-NT and

TNT. Limit of detection for each analyte is tabulated in Table 5-1.

Wu et al.87, Boudreax et al.88, Parpar et al.89, and Swayambunathan et al.90 used PF-LIF in

detecting explosives in various media including soil. Analytical figures of merit for these works

are also tabulated in Table 5-1.

In 1999 and onwards, Sausa et al.91-94 replaced TOFMS with an ion probe made up of

miniature square electrodes to collect the resulting electrons and ions from PF of explosive

compounds. The modification makes the device simpler and portable without trading its

sensitivity for NO(x -1, 2). Pertinent results regarding this technique are noted at Table 5-1.

Finally, a new prototype instrument was developed in this study by coupling the accuracy

and convenience of laser photofragmentation over pyrolysis and the sensitivity for NO(x 1,2)

detection of luminol-chemiluminescence (CL) detector over TOFMS and LIF.

Experimental Section

Reagents and Chemicals

All of the reagents and chemicals were essentially the same as those described in Chapters

II, III, and IV. Basic luminol solution was prepared by dissolving luminol and p-iodophenol in

0.2 M KOH to make 5 x 10 -3 M luminol and 0.01 M p-iodophenol, respectively. Para-

iodophenol served as the chemiluminescent enhancer reagent and was added after dissolving the

luminol into the KOH solution.

Standard solutions of aromatic compounds such as NT, TNT, and 2,4 NT were purchased

from Accustandards (New Haven, CT).









Apparatus and Methodology

Photofragmenation-Chemiluminescence Detector (PF-CL). Figure 5-1 is a schematic

of the experimental set-up for explosive detection using PF followed by CL. It consists of three

main sections: (1) PF unit (2) Oxidation unit (3) CL unit.

The PF unit is essentially the same as described in Chapter II except for an additional

heating system allowing the generation of the explosive vapor. After PF, the products were swept

into the oxidation unit where NO fragments were converted back to NO2. The NO2 fragments

were then subsequently detected by the CL unit. The complete description of the CL and

oxidation units was discussed in Chapter III and IV, respectively.

Varying explosive vapor concentration can be generated as a function of temperature based

on Clausius-Clapeyron Equation. The following equations were used to calculate the explosive

vapor concentration: 91; 95

log [PETN] ppt = -7243/T(K) + 25.56 (5-2)

log [RDX] ppt = -6473/T(K) + 22.50 (5-3)

log [TNT] ppt = -5481/T(K) + 19.37 (5-4)

The temperature of the vapor was assumed to be the same as the temperature of the PF cell. The

temperature was monitored using a thermocouple thermometer.

The sample preparation for solid explosive was the same as described in Chapter II. The

total mass ablated for each experimental run was determined using Sartorious micro balance.

PF efficiency of Nitro-aromatic compounds. Uniform thin films of NT, TNT, and 2-4

NT were prepared by depositing 40 [L of solution of the target compound dissolved in acetone

in a small aluminum cylindrical cup container (O.D. = 1.9 mm; Depth = 4.5 mm).The solvent

was then allowed to evaporate at room temperature overnight. For each measurement, five

different containers containing uniform thin film of selected explosive were used.









PF of NO2 with luminol-CL detector. Figure 5-2 shows the PF set-up for verifying the

further photodissociation of NO2 to NO and O* upon absorption of a photon in 193 nm

wavelength. The set-up consists of the analyte-transport system and luminol-CL detector.

The experiment was performed by generating the NO2 gas through reacting Cu and concentrated

HNO3 in a reaction vessel A. The NO2 generated from the reaction was swept into the quartz

reaction cell B by opening solenoid valves 1 and 2 simultaneously. This step allowed the reaction

cell B to be filled up by NO2 gas. After quite sometime, solenoid valves 1 and 2 were closed and

the laser was fired into reaction cell B. A total of 1000 pulses were used for each experimental

run with pulse energy of 5 mJ/pulse and a repetition rate of 50 Hz. After PF, solenoid valves

number 2 and 3 were opened to sweep all the PF products into the CL unit. Then, the luminol-CL

detector measured the remaining NO2 left in the cell. Then, a reference signal was acquired by

repeating the whole process without turning the laser on.

Results and Discussion

Relative PF Efficiency of Nitro-aromatic Compounds

PF of solid explosive samples can be described in the following stepwise manner. The first

step involved the absorption of sufficient laser radiation by the solid explosive sample. Upon

absorption, a plume of explosive will form and expand away from the sample surface as a result

of pressure difference between the rapidly vaporized sample and the surrounding atmosphere.

The subsequent photon absorption of the explosive vapor will produce PF products, such as NO

and NO2.

Photolysis of solid structurally similar compounds such as NT, 2,4 NT, and TNT were

performed. The parent molecules of these compounds contained one, two, and three NO2

moieties, respectively. Figure 5-3 shows the luminol-CL signal obtained from the NOx(x 1,2)

fragments of these compounds at varying laser energy. The slopes acquired from these sets of









data (Table 5-2) suggest non-linearity in the PF efficiency of these compounds where PF

efficiency is defined as follows:

% PF efficiency = (moles of NO2 detected by CL)/(moles of NO2 available) x 100% (5-5)

The non-linear behavior in the PF efficiency can be attributed to the difference in the

latent heat of sublimation (AHs) for each compound. As AHs increases, the more energy is

required for a substance to vaporize. In the PF process for a solid explosive sample, this will

consequently result in less explosive vapor for photolysis. Actual numerical values of AHs for

these nitrocompounds were not available, however, their vapor pressure at ambient condition

have been well studied and are given in Table 5-2. The vapor pressure is linearly related to AHs.

From Table 5-2, NT has the lowest molecular mass implying that it has the weakest

intermolecular force of attraction resulting in the highest vapor pressure at ambient condition. As

a consequence, the laser irradiated spot of NT produced a plume with the highest concentration

of explosive vapor favoring the most efficient photolysis. The higher slope value of TNT

compared to 2,4 NT can be explained by a single photon absorption capable of simultaneous

multi-NO2 scission. This process defies the higher vapor pressure of 2,4 NT compared to TNT.

NO2 Photolysis with Luminol-CL Detector

Results obtained from the time resolved absorption experiment in Chapter 2 on photolysis

of NO2 into NO and O* using ArF laser was further confirmed by using the luminol-CL detector.

Figure 5-4 shows the CL signal of the NO2 gas before and after photodissociation. Significantly

lower CL signal was observed after the PF of NO2 gas sample.

Analytical Capability of PF-CL Detector

The interaction of laser photons with both solid and vapor phase PETN was investigated

by the quantity of NOx (x 1,2) fragments generated after photolysis. Figure 5-5 presented the

effect of varying laser pulse energy on CL signal obtained from the NOx (x 1,2) generated. In the









absence of CrO3 converter, it can be assumed that the CL signal is due to NO2 fragments alone.

The amount of NO2 released for both solid and vapor phase of PETN increased linearly within

lower fluence range (e.g., a range of 2.5 to 5 mJ/pulse for solid PETN and 2.5 to 7.5 mJ/pulse for

PETN vapor). However, at higher laser fluence, the quantity of NO2 fragments produced does

not significantly changed. The pattern signifies that higher laser fluences yield faster photon

deposition on the sample which favorably induced the sequential two-step PF process (Reactions

5-4 and 5-5) to take place.

C(CH2)404(NO2)4 + hv "- C(CH2)404(NO2)3 + NO2; hv = 193 nm (5-4)

NO2+hv -NO+O*; hv= 193nm (5-5)

The presence of CrO3 oxidizer for NO to NO2 conversion yielded an increasing plot of

NO2 fragments within the entire range of laser fluence for both solid and vapor phase of PETN

indicating that higher photon intensity per pulse consequently produced more NOx (x 1,2)

fragments.

Figure 5-6 shows the effect of varying total number of laser pulses at constant pulse

energy for both solid and vapor phase of PETN. The trend shows that the detected NOx (x 1,2)

fragments increase with increasing pulse number as long as there is an available sample to

photolyze.

The analytical capability of the method for trace explosive detection was obtained. In trace

explosive detection, the explosive residue can be present in two forms: vapor and particulate.

Vapor detection examines the vapor emanating from concealed explosive sample and particulate

detection refers to the microscopic residues of explosives that would be present on individuals or

material which have been through contamination.









Figure 5-7 and 5-8 show the effect of laser parameters such as the total number of pulses

and energy on the amount of solid PETN ablated which directly affects the corresponding CL

signal obtained. The dependence of the CL signal on the amount of PETN ablated was projected

on x,y axis for both graphs (solid points). Both sets of data best fitted into exponential equations

of y = (4x10-6) m15021 and y = (2x10-5) m 1.6266 for the total number of pulses and laser energy

dependent graphs, respectively. The y and m variables correspond to the CL signal and the

amount of the ablated PETN, respectively. Both of these equations were used to estimate the

amount of PETN residue that can be positively confirmed by the method. A positive

confirmation is defined as the amount of PETN residue that provides a signal to background ratio

of 10. A surface concentration range of 61 to 186 ng/cm2 of PETN was calculated.

By using identical laser parameter conditions, calibration curves for the vapor of RDX,

PETN, and TNT were obtained (Figure 5-9). Rough limits of detection (LOD) as well as the

calibration sensitivity for each analyte were determined. The LOD is defined as (3 *Blk)/m

where 5Blk is the standard deviation of the background signal and m is the calibration sensitivity

or the slope of the CL peak area versus the explosive concentration (in ppb) plot. Table 5-3 gives

the LOD values of 3.45 ppb for PETN, 1.73 ppb for RDX and 34.5 ppb for TNT.

Ranking the compounds by LOD yields RDX
is dependent on a number of parameters including (1) the PF cross section or absorption

coefficient of the parent molecule at 193 nm, and (2) the PF efficiency of the parent molecule at

193 nm to yield NOx (x =1,2) fragments. Ordering the compounds by absorbance at 193 nm

yields TNT>RDXzPETN. A priori, the anticipated result should be that the order of compound's

LOD is parallel with their absorbance order. Surprisingly, this is not the case: The TNT's









absorption coefficient is higher than that of RDX and PETN, yet its LOD is poorer. Clearly, the

molecule's absorption coefficient plays less of a role in its LOD.

The LOD ratio of TNT to PETN is ~ 10 which is similar to the value reported previously92.

A plausible argument for TNT's LOD being higher than PETN and RDX might be due to TNT's

several alternative decomposition pathways that compete with R-N02 bond scission. These

include the oxidation of -CH3 to form anthranil, 96; 97 nitro-nitrite isomerization98 and possibly

rearrangement of the ring substituents.99 These pathways decrease the initial production of

NOx(x- 1,2) and contribute to TNT's lower sensitivity relative to RDX and PETN.

PETN has a LOD that is a factor of nearly 2 times greater than that of RDX. This suggest

that the process of NO2 release in RDX molecules is more complicated than the simple cleavage

of a single nitro functional group and may involve the loss of more than one nitro group for each

molecule. The energy for the ring's N-N bond cleavage is lowered after the removal of a nitro

group, and further decomposition generating additional NO2 is feasible even without photon

absorption.51

Conclusion

PF efficiency of nitro compounds is not linearly related to the quantity of NO2 moieties of

the parent molecule which was proven by photo dissociation of NT, TNT, and 2,4 NT.

Photolysis of NO2 fragments into NO and O* was verified using luminol-CL detector.

The interaction of laser photons with solid and vapor phase of PETN was investigated. At

lower laser fluence, a linear increase on the quantity of NO2 fragments produced was observed

for both solid and vapor phase of PETN. However, at higher laser fluences, saturation of NO2

production occurred which implies that at higher photon intensities, the proposed sequential two-

step PF mechanism as shown below took place.









R-NO2 + hi) R + NO2; hv = 193 nm (5-6)

NO2 + h) -- NO + 0*; hv = 193 nm (5-7)

The presence of CrO3 oxidizer for NO to NO2 conversion significantly enhanced the

sensitivity of the detector in analyzing both the solid and vapor phase of PETN.

By using laser PF followed by NO2 detection with luminol-CL technique, it is feasible to

detect energetic materials in real time at ambient conditions. Detection limits of 3.4 ppb for

PETN, 1.7 ppb for RDX, and 34.5 ppb for TNT were obtained. It was also demonstrated that the

presence of PETN residue within the range of 61 ng/cm2 to 186 ng/cm2 can be detected at a

signal to background ratio of 10 using a few micro joules of laser energy.









Table 5-1. Summary of the performance of various explosive detectors based on PF-FD
technique.


Research group/
Year
GW Lemire,
JB Simeonsson,
RC Sausa (1993)53


JB Simeonsson,
GW Lemire,
RC Sausa (1993)54


D Wu, J Singh,
F. Yuch, D. Monts
(1996)87

G Boudreax, T.
Miller, A. Kurefke,
J Singh, F. Yuch, D
Monts (1999)88

T. Arusi-Parpar, D.
Heflinger, R. Laui
(2001)89

V. Swayambunathan,
G. Singh, RC Sausa
(1999)90



V. Swayambunathan,
G Singh, RC Sausa
(1999)91


Method/Comments

PF-REMPI-TOFMS


PF-REMPI-TOFMS


PF-LIF
TNT in soil


PF-LIF


PF-LIF at latm,
24C


PF-LIF
both CH3NO2 and
TNT are not
detected at 454 nm
and (355 + 450)nm

PF-REMPI with ion
probe detector
PF-LIF


LOD


CH3NO2 1000 ppb
DMNA 450 ppb
RDX 8 ppb
TNT 24 ppb
NB 2400 ppb

LOD (ppm) at
X=193nm
CH3NO2 0.18
DMNA 0.51
NB 0.49
o-NT 0.12
m-NT 0.10
TNT 0.21

TNT = 40 ppb at
373 K


TNT H20 500 ppm
TNT soil 100 ppm


LOD (ppm) at
X=226nm
1.0
0.45
2.4
15
36
1.7


TNT 8 ppb at
(S/N) = 10


LOD at 227 nm
CH3NO2 4.3 ppm
TNT 37 ppm



LOD with REMPI
TNT 70 ppb
RDX 7 ppb
PETN 2 ppb


LOD at (355+227)
3.3 ppm
2.6 ppm



LOD with LIF
37 ppm
NOT DETECTED
NOT DETECTED









Table 5-1. (Continued)
Research group/ Method/Comments LOD
Year


V. Swayambunathan,
G Singh, RC Sausa
(1999)92


J. Cabalo, R Sausa,
(2003)93




J. Cabalo, R Sausa,
(2005)94


Monterola, M.P.P.,
Smith,B.W.,Omenetto,
N.,Wineforner,J.W.
(2007)


PF-REMPI with ion
probe detector
PF-LIF


Laser Surface PF-
Fragment Detection
Spectroscopy
(REMPI-ion probe
detector)

Laser Surface PF-
Fragment Detection
Spectroscopy
(REMPI-ion probe
detector)

Photofragmentation-
Luminol-N02
Chemiluminescence
Method


LOD with PF-
REMPI at 227 and
454 nm (ppm),
respectively
PETN 0.5;20.4
RDX 0.4;not
detected
TNT 4.4 ppm;not
detected

RDX 1.4 ng/cm2






RDX 1.4 ng/cm2
HMX 2.0 ng/cm2
CL20 7.1 ng/cm2
TNT 15.4 ng/cm2


RDX 1.7 ppbv
PETN 3.4 ppbv
TNT 34.5 ppbv
(S/N) = 3


Research group/
Year
LOD with PF-LIF at
227 and 454 nm
(ppm), respectively

2.2;140
1.6 ppm;not
detected
3.8 ppm;not
detected


PETN
61 to 186 ng/cm2















N2 gas

"" Lse mc| PFcell

:Laser

hot late
" Photofragmentation unit


Oxidation
unit
... . . . . .


PF products


Sluminol Power supply


electrometer |


: .... D/A converter "


........... .---.--- 1 "
computer :


Chemiluminescence unit
.................................................*


Figure 5-1. Photofragmentation-Chemiluminescence detector.











115

























N2 gas
N N2 gas

SV 1


SV 2
(3-way SV)


Legend
SV = solenoid valve


Figure 5-2. Photofragmentation set-up used for NO2 photolysis.


""* laser













3.0x105-


2.5x10 5-


2.0x10 5-


1.5x10s5-


1.0x105-


5.0x10-


0.01-
0.0


(a) 4-nitrotoluence
(b) 2,4-nitrotoluence
(c) 2,4,6-trinitrotoluence


2.5


5.0


193 nm laser pulse energy (mJ/pulse)

Figure 5-3. Chemiluminescence signals obtained for nitro-aromatic compounds at varying laser
energy.


7.5


10.0


12.5









Table 5-2. Determination of PF efficiency of nitro-aromatic compounds.
Explosive sample a MM (amu) VP at 25C (torr) Slope of the CL
signal vs. Energy
plot (C mJ-1)
NT 137.14 1.5 x 10 3.0x10-6 5.7x10-
(199.1 ppm)

2-4 NT 182.14 2.1 x 10-4 1.1x10-6+ 5.2x10-8
(0.28 ppm)

TNT 227.13 5.8 x 10 -6 1.7x10-6 8.7x10-8
(7.7 ppb)
a Abbreviations: NT, nitrotoluene; 2,4-NT, 2,4 dinitrotoluene; TNT, 2,4,6 trinitrotoluene












2.4x106


LEGEND
A1,A2 = without laser


2.0x10 0 B1,B2 = with laser


1.6x106-

m -6
S1.2x10-


0 8.0x10-


4.Oxl o-.


0.0
Al B1 A2 B2

Figure 5-4. Chemiluminescence signals obtained before and after PF of NO2.




















with CrO3





without Cr 3
._j -^without CrO3





-t I I/


0.0 2.5 5.0 7.5 10.0 12
laser energy (mJ/pulse)


... with CrO3


U0


.-- without CrO3


5.0 7.5
laser energy (mJ/pulse)


Figure 5-5. Effect of varying laser energy on the CL signal of NOx(x 1,2) fragments obtained
after photofragmentation of A) solid and B) vapor phase of PETN.










120


-2.0 -

-2.4

-2.8 -

-3.2

-3.6

-4.0

-4.4

-4.8-


-5.2


-3.4 -

-3.6 -

-3.8 -

-4.0

-4.2 -

-4.4 -

-4.6 -

-4.8 -

-5.0 -

-5.2


}



















-2.0-


-2.4-


-2.8-


-3.2-


-3.6-
0.
0) -4.0-
0

-4.4-


-4.8-


-5.2-


solid


U


...... vapor


-- - - -- - -


2000


4000


6000


8000


total number of laser pulses


Figure 5-6. Effect of varying total number of laser pulses on the CL signal of NOx(x=1,2)
fragments obtained after PF of solid and vapor phase of PETN.













8000


6000 ..,
S .. .....
^ 4000 ~ ~ ~ ~ ~ -- --.---^ ^^ ^ ^--------^
5000 --
o (ooo

3000 ..... .


2000





"a/ 100 120 0 <

9)


Figure 5-7. Effect of varying total number of laser pulses on the mass of PETN ablated and on
the CL signal of NOx(x 1,2) fragments obtained after PF of solid PETN.















10.0




7.5 ......


...

S 5.0







0.0 5 10





2)

Figure 5-8. Effect of varying laser energy on the mass of PETN ablated and on the CL signal of
NOx(x -1,2) fragments obtained after PF of solid PETN.














RDX


TNT


-I I I I2 3
-1 0 1 2 3


log (concentration, ppbv)



Figure 5-9. Calibration curves of PETN, RDX, and TNT.


-2.0 -


-2.4-


-2.8-


-3.2-


-3.6-


-4.0-


-4.4-


-4.8-









Table 5-3 Analytical figures of merit of PF-luminol CL detector for some nitro-based explosives.
Explosive sample a LOD at (S/N) = 3 Calibration sensitivity
(CL signal/ppb)
PETN 3 1 x 10 -
RDX 2 2 x 10 -5
TNT 34 1 x 10 -6
a Abbreviations:PETN, 1,3-dinitro-2,2-bis(nitramethylpropane); RDX, 1,3,5-trinitro-1,3,5-
triazacyclohexane; TNT, 2,4,6 trinitrotoluene









CHAPTER 6
DIRECT DETECTION OF EXPLOSIVE IN SOIL

Introduction

Rapid and on-site detection of nitro-based explosives in a complex matrix such as soil is

essential in providing an appropriate feedback during the characterization or remediation of

contaminated sites. This chapter begins with the discussion of explosives as an environmental

pollutant followed by a review of current monitoring explosive techniques, and finally an

evaluation of the analytical capability of PF-luminol CL detector for PETN analysis in soil.

Environmental Hazards of Explosive Contaminated Soil

Soil contamination of nitro compounds is a problem because of the scale on which

explosives have been manufactured, used, and tested. Contamination occurred during the

manufacturing of explosives which required large amounts of water for purification. The waste

water from this process is usually placed in lagoons or sedimentation basins.100 Many sites also

became contaminated through open detonation and burning of explosives at army depots,

evaluation facilities, artillery ranges, and ordinance disposal sites.

The United States Department of Defense has identified more than 1 000 sites with

explosive contamination of which 87% exceeded permissible contaminant levels. 101 Most

contamination exists in near surface soils and in the vicinity of firing range targets. A primary

concern is that soil contaminants may eventually migrate to groundwater and contaminate

drinking water supplies of nearby communities. TNT, RDX, HMX, and PETN are weakly to

moderately soluble in water with solubility of 130, 40, 5, and 43 mg/L, respectively at 25C.102

Nitro-based explosives are toxic and present harmful effects to all of life forms. TNT is on

the list of US EPA priority pollutants: it's a known mutagen and can cause pancytopenia as a

result of bone marrow failure. 103 TNT also causes the formation of methemoglobin on acute









exposures and anemia on chronic exposures. It also causes toxic hepatitis, leukocytosis,

peripheral neuritis, muscular pains, cardiac irregularities, renal irritation, and bladder tumors.

On the other hand, RDX is considered as a possible human carcinogen (Class C) of EPA. RDX

caused liver tumors in mice that were exposed to it in the diet. RDX also produced smaller

offspring in rats. Clinical findings in low level long term exposures may include tachycardia,

hematuria, proteinuria, mild anemia, neutrophilic leukocytosis, and electroencephalogram (EEG)

abnormalities.2 In general, nitro compounds are generally recalcitrant to the biosphere, where

they constitute a source of pollution due to both toxic and mutagenic effects on human, fish,

algae, and microorganisms.

Review of Explosive Detection in Soil

Various analytical methods for detection of nitro-based explosive present in soil have been

reported. These are mostly spectrometric such as Electron Capture Detector (ECD) 11; 104

Thermal Energy Analyzer (TEA) 12;105, Ion Mobility Spectrometry 15;16 and UV Absorption

coupled with several different separation methods such as extractions, gas and liquid

chromatography, and electrophoresis. Voltammetric and amperometry 106; 107 methods were also

commonly used in determination of trace levels of explosive substances in soil. The basic

principles of these methods were discussed in Chapter I.

The current techniques mentioned above are often cost prohibitive, time consuming, and

labor intensive. To characterize a site for explosive detection using these methods, it is necessary

to follow a standard protocol starting from sample collection. Then, the analyte from the soil is

extracted from the soil sample using organic solvent such as acetonitrile under at least 18 hour of

sonication in cooled ultrasonic water bath before the actual analysis. Indeed, these traditional

methods are laborious, time consuming, impractical and ineffective for field analysis.









Explosive field analysis is valuable due to extremely heterogeneous distribution of

explosives in contaminated soils requiring enormous sampling sites. With such enormous

sampling sites, a cost effective tool is necessary. Since cost per sample is lower, more samples

can be analyzed producing more reliable results. Also, the availability of near-real time results

permits re-design of the sampling scheme while in the field which facilitates more effective use

of off site laboratory facilities with more robust analytical methods.

Currently, field methods for detection of explosives in soil include colorimetric108 and

immunosensors methods.109 Both methods have several drawbacks. A colorimetric technique is

used only for rugged qualitative analysis and often gives unreliable results. Immunosensors

suffer because antibodies of all explosive materials likely to be encountered are needed, i.e.,

specific antibody is required for each compound of interest. This adds to the cost of an

immunosensors. Also, when multi-analyte sensors are used, there is poor signal discrimination

and as a result sensitivity is lost.

PF-Luminol CL detector is a promising technique for both on-site and off-site explosive

detection in soil. The newly developed method has the following attributes: (1) it is simple and

straightforward technique (2) it requires no sample preparation; (3) it is fast and provides a real

time response; (4) all components including CL sensors can be made rugged and field portable.

An actual analysis of PETN contaminated soil was demonstrated and the analytical

capability of the method was discussed.

Experimental Section

Chemical and Reagents

All the reagents and chemicals used for this experiment were exactly the same as Chapter

V. The soil matrix used was a standard reference material (SRM-2704) prepared by the National

Institute of Standards and Technology (NIST). SRM-2704 is freeze dried river sediment that was









sieved and blended to achieve a high degree of homogeneity. It is intended primarily for use in

the analysis of sediments, soils, or material of similar matrix.

Apparatus and Methodology

The experimental set-up is the same as Chapter V except for the CL cell reaction. During

the course of the dissertation, the geometry of the CL cell reaction underwent several revisions.

The final reaction cell was designed with several goals in mind. First, the area of the PMT

window was maximized. It was hoped that this would increase the detector sensitivity. A second

expectation was this configuration prevented condensing of the luminol solution on the surface

of the PMT window. Finally, by providing a compartment waste for the luminol solution,

necessary maintenance such as periodically cleaning the cell was no longer necessary. The cross

section of the CL cell reaction is shown in Figure 6-1.

The method of analysis is the same as described in Chapter V except for the laser

parameters. In each experimental run, 8 000 laser pulses with average laser energy of 8 mJ/pulse

was used. The laser repetition rate was set to 50 pulses per second.

Sample preparation. A known mass of soil was accurately weighed on a small piece of

weighing paper. The soil sample was quantitatively transferred and tightly packed into a small

aluminum cylindrical cup container (O.D. = 1.9 mm; Depth = 4.5 mm). Exactly 40 [L of the

standard PETN solution in acetone was deposited on the top of the soil sample. Then the sample

was dried in the oven at 60C for an hour to evaporate the solvent. To achieve varying

concentration of PETN in soil, different standard PETN solutions ranging from 0.4 to 0.16

mg/mL were prepared. The explosive concentration in soil (mg/g) was calculated by using the

following equation:

Explosive contamination (mg/g) = [PETN] (mg/mL) (40 uL) (10 -3) (5-1)
Weight of soil (g)










Results and Discussion

The analysis was done by imaging an ArF laser beam on the explosive contaminated soil.

The photofragmented products were then measured using Luminol-CL detector. Each

experimental run was at least repeated eight times on the same sample spot. A calibration curve

was generated by plotting peak area against concentration ranging from 0.4 to 9.0 parts per

thousand PETN. Figure 6-2 best fitted a polynomial equation ofy = 2.4x2 + 3.9x + 2.8 where y

equal to the CL peak area and x is the concentration of PETN in soil in parts per thousand. The

curve has a correlation coefficient value (R2) equal to 0.9996. Based on this curve, a detection

limit was established. The limit of detection (LOD) is defined as (3aBlk/m) where ablk is the

standard deviation of the blank and m is the calibration sensitivity or the slope of the line. This

definition is restricted for the cases of a calibration line where the slope is constant, however, the

calibration curve obtained for this analysis is curvilinear (slope is variable) and the definition

provided is not useful without modification. The above equation can be redefined as follows:

LOD = 3Cblk/{dy/dx[f(x)]} (5-2)

where the slope is replaced as a differential coefficient of function x, which in this case is the

PETN concentration in soil. The equation of the slope is then given by:

m = 2(2.4x) + 3.9 (5-3)

LOD range of 0.5 to 4.3 ppm of PETN was obtained with x values equal to 0.4 and 9.0 parts per

thousand of PETN. The precision of the method has been established at the lowest concentration

of PETN which gave a relative standard deviation (RSD) of 30 % where n = 10. Larger

variations obtained in the signal especially in the lower concentration range, is due to the

decreasing surface concentration of the contaminant as the laser ablates deeper into the soil

matrix.









The LOD was also estimated by considering only the linear plot obtained from the first

four data points. A linear equation of y = 15.662x 4.2364 was obtained where y is the peak area

of the signal and x is the PETN concentration in ppth. An LOD of 1.6 ppm was calculated from

this plot.

The LOD of the method is comparable to other current techniques as can be seen in Table

1-1 and 5-1.The analytical capability of this method can be applied in analysis of any typical

explosive contaminated soil given in Table 6-1. Based on the data tabulated in Table 6-1, the

TNT concentration in the soil ranges from 4 000 to 1 200 ppm whereas the concentration of

other explosive contaminants is less than 300 ppm. The nitrogen content of the soil was 7.5 ppm

as ammonium ion and nitrate concentrations vary from 6 to 12 ppm.102; 110

Pure NaNO2 and NaNO3 were photolyzed and analyzed with the same method in order to

investigate the potential interferent during the analysis but both compounds did not give a

significant CL signals as can be seen in Figure 6-3.

Conclusion

PF-Luminol-CL detector is a promising method for analysis of trace explosive in soil. The

LOD range of 0.5 to 4.3 ppm for PETN was established. The PF-luminol CL detector is

comparable to current techniques available. The most advantageous characteristic of the PF-

luminol CL detector compared to other traditional methods is its simplicity and reliability. The

technique also requires no sample preparation and provides fast real-time responses.

































luminol-N02 mist


Figure 6-1. Cross section of the CL cell.











300 -


250-


S200-


150



50-

50



0 2 4 6 8 10

S6PETN a soil



Figure 6-2. Calibration curve for PETN contaminated soil.









Table 6-1. Explosives concentration in the contaminated soil 102; 110
Explosive a Concentration range
(mg explosive/kg of soil)
TNT 4 000-12 000
TNB 175-300
2,4-DNT 50-200
RDX 50-125
HMX 50-100
a Abbreviations: TNT, 2,4,6-trinitrotoluene; TNB, trinitrobenzene; 2,4 DNT, 2,4-dinitrotoluene;
RDX, hexahydro-1,3,5-trinitro-1,3.5-triazine; HMX, octahydro-1,3,5,7-tetranitro-1,3,5,7-
tetraazocine



































0 10 20 30 40
tim e (s)


0 10 20 30
tim e (s]


50 60 70


40 50 60 70


0 10 20 30 40
tim e (s)


50 60 70


Figure 6-3. CL Spectra after PF of A) empty sample holder B) pure NaNO2 C) pure NaNO3.













135


A -5
-6

-7
a -8

R_ -9
.J
o -10

-11

-12

-13


B -5
-6
-7

S -8

P -9
S -1 0

-1 1

-1 2
-1 3


-4-

-6 -

-8


P -10
J -
o -12

-1 4 -

-1 6 -









CHAPTER 7
CONCLUSIONS AND FUTURE WORKS

Summary and Conclusions

The purpose of this research was to develop a simple, fast, reliable, sensitive and

potentially portable explosive detection device employing laser photofragmentation (PF)

followed by heterogeneous chemiluminescence (CL) detection. The work was divided into four

stages: (1) discerning the PF pathways of nitro-based explosives (2) luminol-CL method

development and optimization for improving NO2 sensitivity (3) construction of CrO3 oxidizer

for NO to NO2 conversion (4) evaluation of the device in detecting nitro-based explosives in

various media such as air and soil.

In the first stage of the research, the following PF pathways were verified using classical

colorimetric analysis and fast time resolved absorbance method.

R-N02 + hv -- R + N02; hv = 193 nm (7-1)

NO2 + hv NO + 0*; hv =193 nm (7-2)

Confirmation of NOx(x-1,2) generation after photolysis of nitro-based explosives was the basis for

developing an explosive detector based on PF-Fragment detection scheme.

The second stage involved the development and optimization for NO2 detection using

luminol CL. In this system, a stream of NO2 gas passed through a concentric nebulizer and used

to aspirate a spray of luminol solution. The aspiration process maximized N02/luminol contact

area, thereby, enhancing the CL signal. The current system was able to improve the detection

limit attained so far through optimization of all the feasible physical and chemical parameters

involved in the chemiluminescence reaction between luminol and NO2. Detection limit of 19 ppt

NO2 at (S/N) = 3 was reported. The optimal reagent solution from the viewpoint of sensitivity of

the response to NO2 (maximum signal/signal noise ratio) is 5 x 10-3 M luminol + 0.01 M p-









iodophenol + 0.2 M KOH. Luminol/H202 CL set-up was also explored where a bundle of porous

polypropylene fibers was used to bring the NO2 into contact with luminol solution. LOD of 178

ppt NO2 at (S/N) = 3 was obtained.

In the third stage, a CrO3 oxidizer was constructed for NO to NO2 conversion. Photolysis of

nitro compounds produced NO (reaction 7-2) which is not detectable by luminol-CL system. To

further increase the sensitivity of the device for NO2 fragments, NO must be converted back to

NO2.

Finally, the fourth stage was integration of PF and CL units by a CrO3 oxidizer into an

explosive detection device. The system was able to detect energetic materials in real time at

ambient conditions. Detection limits of 3.4 ppbv for PETN, 1.7 ppbv for RDX, and 34.5 ppbv for

TNT were obtained. It was also demonstrated that the presence of PETN residue within the range

of 61 ng/cm2 to 186 ng/cm2 can be detected at a given signal to background ratio of 10 using a

few micro joules of laser energy. The technique also demonstrated its potential for direct analysis

of trace explosive in soil. LOD range of 0.5 to 4.3 ppm for PETN was established, an analytical

capability comparable to current techniques available. The technique most advantageous

characteristic compared to other traditional methods is its simplicity and reliability. The

technique also requires no sample preparation and provides fast real-time response.

Future Research Directions

Based on the results presented in this dissertation, several paths can be followed for future

research studies. The future work can be grouped into (1) extending the systems applicability

towards peroxide-based explosives (2) instrument's field adaptability and finally (3) instrument's

performance in actual scenario.

Liquid peroxide explosives such as triacetoneperoxide (TATP) and hexamethylene

triperoxide diamine (HMTD) were home-made explosives recently used in Madrid and London









subway attacks of 2004 and 2005, respectively. TATP was also been used by suicide bombers in

Israel, and was chosen as a detonator in 2001 by the thwarted "shoe bomber" Richard Reid.2

Current analytical methods used for peroxide-based explosives include Infrared

spectroscopy (IR), Raman spectroscopy, Chemical Ionization-Mass Spectrometry (CI-MS), Ion

Mobility Spectroscopy (IMS), 1H and 13C NMR, and High Performance Liquid Chromatography

(HPLC) with post column UV irradiation and fluorescence detection. Most of these techniques

are applied in laboratory environment and usually not applicable to field screening scenarios. So

far, the only detection method used in field analysis involved UV irradiation (X = 254-415 nm) of

the liquid explosive samples followed by hydrogen peroxide (H202) fragment detection using

colorimetric technique.111 However, this particular method involves a lot of sample preparation,

used a lot of expensive reagents, time-consuming and relatively has poor sensitivity.

Furthermore, the technique is only limited to liquid samples. An alternative way to overcome

these issues is through direct photofragmentatation (PF) of peroxide-based explosive vapor and

subsequent H202 detection using luminol-CL detector (Figure 7-1). The high volatility of these

compounds is a very promising premise for developing the first known air analysis for peroxide-

based explosive using our method.

After extending the chemistry of the system with peroxide-based explosives, the second

direction to be considered is its portability for field adaptability. The over-all set-up including the

detection system can be readily miniaturized. The most challenging task is the replacement of

ArF laser but this can be feasibly done using commercially available excilamps which only

weigh about 2.2 kg.112 Excilamps provide narrow band light around a single wavelength by

radiative decomposition of excimer states created by a barrier discharge. Lamps operating at

wavelengths of 126, 146, 172, 193, 222, 282, and 308 are already available. Studies on









excilamps show a powerful radiation in UV and vacuum UV spectral range that allows

photochemical processes that were not previously possible with lamps.113 These lamps also

operate at low temperature, generate minimal IR radiation, and allow processing of temperature

sensitive substrates such as explosives. Furthermore, these equipment are highly efficient source

converting power into light effectively and therefore, reducing power requirements and heat

load. The system is also robust, inexpensive, and ecologically beneficial.113

After achieving the portability requirement, actual field testing of the instrument should be

evaluated. Foreseeable problem that can be encountered during the actual testing will be the

sensitivity issue of the device especially when dealing with direct explosive vapor detection with

an exemption for peroxide-based explosives. As mentioned in Chapter I, nitro-based explosives

significantly have low vapor pressures and the method that is suitable for its detection should be

able to detect concentrations down to less than 1 ng/L (ppb level). Our current method has a

LOD of 3.4, 1.7, and 34.5 ppbv for PETN, RDX, and TNT, respectively.

Improving the sensitivity of the device can be done by the following suggestions: (1)

replacing the PMT detector with another PMT system which has the least dark count rate and

highest photon gain and (2) eliminating the background noise due to interference present in air

such as native NOx(x-1,2), 03, and SO2.

Previous works 73-80 with atmospheric monitoring of NO2 were able to eliminate almost

100 % of 03 and SO2 from air by adding Na2SO3 in the basic luminol solution. 03 and SO2

preferred and readily react with a stronger reducing agent such as Na2SO3 rather than basic

luminol solution.

Interferences from native NOx(x-1,2) can be compensated in several ways. The first one

involved NOx(x-1,2) removal prior to sample analysis. This must be done without removing the









explosive vapor or the analyte. One feasible way to do this is through chemical adsorption of

NO2 on a surface of titanium dioxide (TiO2) photo catalyst and hygroxyapatite

(Caio(PO4)6(OH)2.114 In this method, NO was oxidized to HNO3 which is efficiently adsorbed on

the surface of TiO2. Removal efficiency of 98 % was achieved for this method. Another

alternative option is to purge the air sample with solutions of either triethanolamine (TEA) or

sodium arsenite prior to photofragmentation. Both solutions are known for their ability to pre-

concentrate NO2 in air by trapping and converting NO2 gas to nitrite ions in the bulk solution.

These absorber solutions have absorption efficiencies of 95 and 82 percent, respectively.59 The

stoichiometric factor of TEA for NO2 to NO2 is 0.85 while that of sodium arsenite is unity.115

The main concern of using these NOx(x 1,2) removal methods is that the feasibility of removing

some of the analyte as well. Explosive vapor is prone to get easily lost to wall adsorption during

transport.

The second way to compensate for the presence of native NOx(x-1,2) in air is to divide the

volume of the air sample pumped into the instrument each time. The first sample (sample A)

collected will be diverted to PF cell while the second sample (sample B) will be diverted to

another cell. Both samples will undergo the same reactions of oxidation and luminol-CL except

for photolysis part. The presence of explosive vapor can be confirmed by a significantly higher

CL signal of the air sample that undergoes photolysis than the air sample that was not treated

with UV irradiation. A positive explosive alarm will be based on the statistically established

differential value between sample A and B.

In conclusion, the main drawback of PF-CL detector is its relatively high background noise

due to interferents present in the air sample specifically the native NOx(x-1,2), thus, solid residue

analysis of nitro-based explosive is preferred and more compatible for this technique.










I(CH2--O-O-H2C\
N~CH2-O-_O-H2C-N
~CHg-0-0-H2C'*"


HMTD


.= 254 nm


H3CC .,CH3



HC -0 CH3
TATP


I Hydrogen Peroxide I


I luminol-CL detector I


Figure 7-1. Alternative scheme for detecting peroxide-based explosives.









LIST OF REFERENCES


1. Singh, S. J. Hazard. Mater. 2007, in press.

2. Mahoney, C.M.; Gillen, G.; Fahey, A.J. Forensic Science International 2006, 158, 39-51.

3. Vila, M.; Lorber-Pascal, S.; Laurent, F. Environmental Pollution 2007, 148, 148-154.

4. Ahmad, F.; Schnitker, S.P.; Journal of Contaminant Hydrology, 2007, 90, 1-20.

5. Agrawal,; J.P., Surve,; R.N.; Mehilal,; S. H.; Sonawane,; A.L. Journal of Hazardous
Materials 2000, 77, 11-31.

6. Hallowell, S. F. Talanta 2001, 54, 447-458.

7. Yinon, J. Trends Anal. Chem. 2002, 21, 292-301.

8. Hilmi, A.; Luong, J.T.; Environ. Sci. Technol. 2000, 34, 3046-3050.

9. Lu, Q.; Collins, G.E.; Smith, M.; Wang, J. Analytica ChimicaActa 2002, 469, 253-260.

10. Mullen, C.; Irwin, A.; Pond, B.; Huestis, D.L.; Coggiola, M.J.; Oser, H. Anal. Chem.
2006, 78, 3807-3814.

11. Walsh, M.E. Talanta 2001, 54, 427-438.

12. Lafleur, A.L.; Mills, K.M.; Anal. Chem. 1981, 53, 1202-1205.

13. Asbury, G.R.; Klasmier, J.; Hill, H.H. Talanta 2000, 50, 1291-1298.

14. Ewing, R.G.; Atkinson, D.A.; Eiceman, G.A.; Ewing, G.J. Talanta 2001, 54, 515-529.

15. Buryakov, I.A. J of Chromatogr. 2004, 800, 75-82.

16. Khayamian, T.; Tabrizchi, M.; Jafari, M.T. Talanta 2003, 59, 327-333.

17. Matz, L.M.; Tornatore, P.S.; Hill, H.H. Talanta 2001, 54, 171-179.

18. Kanu, A.B.; Haigh, P.E.; Hill, H.H. Analytica ChimicaActa 2005, 553, 148-159.

19. Otto, J.; Brown, M.F.; Long, W. AppliedAnimal Behaviour Science 2002, 77, 217-232.

20. Harper, R.J.; Almirall, J.R.; Furton, K.G. Talanta 2005, 67, 313-327.

21. Harvey, S.D.; Clauss, T.W. J. Chromatogr. A 1996, 753, 81-86.









22. Pnnaduwage, L.A.; Boiadjiev, V.; Hawk, J.E.; Thundat, T. Appl. Phys. Letter 2003, 83,
1471-1475.

23. Datskos, P.G.; Lavrik, N.V.; Sepaniak, M.J. Sens. Letters 2003, 25-37.

24. Ly, S.Y.; Kim, D.H. Talanta 2002, 58, 919-926.

25. Hilmi, A.; Luong, J.T.; Environ. Sci. Technol. 2000, 34, 3046-3050.

26. Wang, J.; Thongngadee, S. Analytica ChemicaActa 2003, 485, 139-143.

27. Albert, K.J.; Myrick, M.L.; Brown, S.B.; James, D.L.; Milanovich, F.P.; Watt, D.R.
Environ. Sci. Technol. 2001, 35, 3193-3200.

28. Charles, P.T.; Kustebeck, A.W. Biosens. Bioelectron. 1999, 14, 391-400.

29. Dorozhkin, L.M.; Nefedov, V.A.; Sobelnikov, A.G.; Sevastjanov, V.G. Sensors and
Actuators B 2004, 99, 568-570.

30. Sohn, H.; Calhoun, R.M.; Sailor, M.J.; Tragler, W.C. Angew. Chem. 2001, 13, 2162-
2165.

31. Wallenburg, S.R.; Bailey, C.G. Anal. Chem. 2000, 72, 1872-1875.

32. Wang, X.; Zeng, H.; Zhao, L.; Lin, J.M. Talanta 2006, 70, 160-164.

33. De Lucia, F.C.; Harmoun, R.S.; McNesby, K.L.; Winkel R.J.; Misiolek A.W. Applied
Optics 2003, 42, 6148-6150.

34. Lin, H.; Chen, Y. Optics Express 2006, 14, 415-424.

35. Shen, Y.C.; Lo, T.; Taday, P.F.; Cole, B.E.; Tribe, W.R.; Kemp, M.C. Applied Physics
Letters 2005, 86, 241116-1 to 241116-3.

36. Sylvia, J.M.; Janni. J.A.; Klein, K.D.; Spencer, K.M. Anal. Chem. 2000, 72, 5834-5840.

37. Hayward, I.P.; Kirkbride, T.E.; Batchelder, D.N.; Lacey, R.J. J. of Forensic Sciences
1995, 40, 883-884.

38. Janni, J,; Gilbert, B.D.; Field, R.W.; Steinfeld, J.I. Spectromica Acta Part A 1997, 73,
255-258.

39. Xu, S.; Sha, G.; Xie, J. Rev. Sci. Instrum. 2002, 73, 255-258.

40. Balint-kurti,;G.G., Shapiro, M. ChemicalPhysics 1981, 61,137-155.









41. Turro, N.J. Molecular Photochemistry. Amsterdam, 1967.

42. Mc Donnell L.; Heck, A.J. J. of Mass Spectrometry 1998, 33, 415-428.

43. Cabalo, J.; Sausa, R.C. Applied Optics 2005, 44, 1084-1091.

44. Simeonsson, J.B.; Sausa, R.C. Trends in Analytical Chemistry 1998, 17, 542-550.

45. Monts, D.L.; Singh, J.P.; Boudreaux, G.M. Encyclopedia of Analytical Chemistry, John
Wiley and Sons Ltd., Chichester, 2000, 2148-2171.

46. Ledingham, K.W.D.; Singhal, R.P. International Journal of Mass Spectrometry and Ion
Processes 1997, 163, 149-168.

47. Ostmark, H.; Carlson, M.; Ekvall, K. Combustion and Flame, 1996, 105, 381-390.

48. Butler, L.J.; Krajrovich, P.; Lee, Y.T. J. Chem. Phys. 1983, 79, 1708-1722.

49. Blais N. Journal of Chem. Phys. 1983, 79, 1708-1722.

50. Renlund, A.M.; Trott, W.M. Chemical Physics Letters 1984, 107, 555-560.

51. Capellos,C.; Papagrannakopoulos, P.; Liang, Y. Chemical Physics Letters 1989, 164,
533-538.

52. Galloway, D.B.; Bartz, J.A.; Huey, G.L.; Crim, F.F. J. Chem. Phys. 1993, 98, 2107-2114.

53. Lemire,G.W.; Simeonsson, J.B.; Sausa, R.C. Anal. Chem. 1993, 65, 529-533.

54. Simeonsson, J.B.; Lemire,G.W.; Sausa, R.C. Applied Spectroscopy 1993, 47, 1907-1912.

55. Marshall A.; Clark, A.; Ledingham, K.W.; Sanders, J.; Singhal, R.P.; Kosmidis, C.; Deas
R.M. Rapid Communications in Mass Spectrometry 1994, 8, 521-526.

56. Ledingham, K.W. Physica Scripta 1995, 158, 100-103.

57. Simeonsson, J.B.; Lemire,G.W.; Sausa, R.C. Army Research Laboratory Aberdeen
Ground, MD, USA Available NTIS Report 1994. 30 Rep. Announce. Index 95(19)
Abstract no. 5429519.

58. Yang, Mo; Ramsey, J.M.; Kim, B. Journal of Rapid Communications in Mass
Spectrometry 1996, 10, 311-315.

59. Gayathri, N.; Balasubramanian, N. Analusis 1999, 27, 174-181.









60. Finlayson, B.J.; Pitts J.N. Atmospheric Chemistry: Fundamentals and Experimental
Techniques. John Wiley and Sons Inc. Canada, 1986, 151-153.

61. Sun, F.; Glass, G.P.; Curl, R.F. Chemical Physics Letters 2001, 337, 72-78

62. Campana, A.G.; Baeyens, W.R. Chemiluminescence in Analytical Chemistry, Markel
Dekker Inc., Basel, Swizerland, 2001; Chapter 1.

63. Ingle, J.D.; Crouch, S.R. SpectrochemicalAnalysis, Prentice-Hall Inc., New Jersey, 1988;
Chapter 15.

64. Li, Y.; Qi, H.; Fang, F.; Zhang,; C. Talanta, 2007, 72, 1704-1709.

65. Isacsson, U.; Wettermark, G. Analytica ChimicaActa 1974, 68, 339-362.

66. Wendel, G.J. Development and Application of a Luminol-based NO2 Detector, University
of Michigan, 1985.

67. Tahirovic, A.; Copra, A.; Mikli6anin, E.O.; Kalcher, K. Talanta, 2007, 72, 1378-1385.

68. Su, Y.; Li, X.; Chen, X.; Lu, Y.; Hou, X. Microchemical Journal, 2007, In Press,
Corrected Proof

69. White, E.H.; Zafriou, H.H.; Hill J.H.M. J. Am. Chem. Soc. 1964, 86, 940-941.

70. Shevlin,P.B.; Neufeld, H.A. J. Org. Chem. 1970, 35, 2178-2182.

71. Hersh, B.B.; Wesely, M.L. J. Photochem. 1974, 10, 409-423.

72. Roswell, D.F.; White, E.H. Meth. Enzynol. 1978, 57, 409-423.

73. Maeda, Y.; Aoki, K.; Munemori, M. Anal.Chem.1980, 52, 307-11.

74. Wendel, G.J.; Stedman, H.; Cantrell, C.A. Anal.Chem. 1983, 55, 937-40.

75. Schiff, H.I.; Mackay, G.I.; Castledine, C.; Harns, G.W. Trans Q. Water, Air and Soil
Pollution. 1986.30, 105-114.

76. Kelly T.J., Spicer C.W., Ward G.F. Atmospheric Environment A 1990, 24, 2397-2403.

77. Mikuska, P.; Vecera, Z. Anal. Chem.1992, 64, 2187-91.

78. Collins, G.; Rose-Phersson, S.L. Anal. Chem. 1995, 67, 2224-30.

79. Robinson, J.K.; Bollinger, M.J.; Birks, J.W. Anal. Chem. 1999, 71, 5131-5136.










80. Kok, G.L.; Hoiler, T.P.; Lopez, M.B.; Nachitrieb, H.A.; Yuan, M. Environmental Science
and Technology 1978, 12, 1072-1076.

81. Thorpe, G.H.; Kricka, L.J.; Moseley, S.B.; Whitehead T.P. Clinical Chemistry 1985, 31,
1335-1341.

82. Robinson, J.K. Luminol-Hydrogen peroxide Detector for the Analysis of NO in Exhaled
Breath, University of Colorado, 1994.

83. Hutchinson, G.L.; Yang, W.X.; Andre, C.E. Atmospheric Environment 1999, 33, 141-
145.

84. Levaggi, P.; Kothny, E.L.; Belsky, T.; de Vera E.; Mueller, P.K. Environmental Science
and Technology 1974, 8, 348-350.

85. Nguyen Dao Hinh. US Patent no. 20040169495, 1994.

86. Nguyen Dao Hinh. US Patent no. 20040053421, 1994.

87. Wu, D.; Singh, J.; Yueh, F.; Monts, D. Applied Optics 1996, 35, 3998-4003.

88. Boudreax, G.M.; Miller, T.S.; Kunefke, A.J.; Singh, J.; Fang-Yu, Y.; Months D. Applied
Optics. 1999. 38, 1411-1417.

89. Arusi-Parpar, T.; Heflinger, D.; Lavi, R. Applied Optics 2001, 46, 6677-81.

90. Swayambunathan, V.; Singh, G.; Sausa, R.C. Applied Spectroscopy 2000, 54, 651-657.

91. Swayambunathan, V.; Singh, G.; Sausa, R.C. Applied Optics 1999, 40, 6447-6454.

92. Swayambunathan, V.; Singh, G.; Sausa, R.C. Proceedings of International Society for
Optical Engineering 1999, 176-184.

93. Cabalo, J.; Sausa, R.C. Applied Spectroscopy 2003, 57, 1196-1199.

94. Cabalo, J.; Sausa, R.C. Applied Optics 2005, 44, 1084-1091.

95. Lucero, D.P.; Boncyk, E.M. Journal of Energetic Materials 1986, 4, 473-510.

96. Chakraborty, D.; Muller, R.P.; Dasdupta, S.; Goddard, W.A. J. Phys. Chem. 2001, 105,
1302-1314.

97. Gonzalez, A.C.; Larson, C.W.; McMillen, D.F.; Golden, D.M. J. Phys. Chem. 1985, 89,
4809-4814.









98. He, Y.Z.; Cui, J.D.; Mallard, W.G.; Tsang, W. J. Am. Chem. Soc. 1988, 110, 3754-3759.

99. Kallman, H.P.; Spruch, G.M. Luminescence of Organic and Inorganic Materials. John
Wiley and Sons, Inc., NY, 1969.

100. Pennington, J.C.; Brannon, J.M. Themochimica Acta 2002, 384, 163-172.

101. Rodgers, J.D.; Nigel, J.B. Water Research 2001, 35, 2101-2111.

102. Halasz, J.D.; Groom, C.; Zhou E.; Paquet, L.; Ampleman, G.; Dubas, C.; Hawari, J.
Journal of Chromatog. A 2002, 963, 411-418.

103. Morley, M.C,; Yamamoto, H.; Speitel, G.; Clausen, J.; Journal of Contaminant
Hydrology 2006, 85, 141-158.

104. Jenkins, T.F.; Leggett, D.C.; Miyares, P.H.; Walsh, M.E.; Ranney, T.A.; Cragin, J.H.;
George V. Talanta 2001, 54, 501-513.

105. Bowerbank, C.R.; Smith, P.A.; Futterrolf, D.D.; Lee, M.L. Journal of Chromatog. A
2000, 902, 413-419.

106. Buttner, W.J.; Findlay, M.; Vickers, W.; Davis, W.M.; Cespedes, E.R.; Cooper, S.I.;
Adams, J.W. Analytica ChimicaActa 1997, 341, 63-71.

107. Hilmi, A.; Laong, J.H.; Nguyen, A. Journal of Chromatog. A 1999, 844, 97-110.

108. Uzer, A.; Ercag, E.; Apak, R. Analytical Chimica Acta 2005, 534, 307-317.

109. Gauger P.R.; Hoh D.B.; Patterson, C.H.; Charles, P.T.; Shriver-Lake L.; Kusterbeck,
A.W. Journal of Hazardous Materials 2001, 83, 51-63.

110. Boopathy R. International Biodeterioration and Biodegradation 2000, 46, 29-36.

111. Schutte-Ladbeck, R.; Karst, U. Analytica ChimicaActa 2003, 482, 183-188.

112. Li, Q.; Gu, C.; Di, Y.; Yin, H.; Zhang, J. Journal of Hazardous Materials 2006, 133, 68-
74.

113. Elsner, C.; Lenk, M.; Prager, L.; Mehner, R. Applied Surface Science 2006, 252, 3616-
3624.

114. Komazaki, Y.; Shimizu, H.; Tanaka, S. Atmos. Environ. 1999, 33, 4363-4371.

115. Yuen, W.K.; Horlick, G. Analytical Chemistry 1977, 49, 1448-1450.









BIOGRAPHICAL SKETCH

Maria Pamela Pineda Monterola was born on September 18, 1977, in San Pablo City,

Philippines. She is the third child of Nora Dizon Pineda and Conrado Cachero Monterola. She

graduated cum laude with bachelor's degree in chemistry in 1998 from the University of the

Philippines. She was immediately hired as a junior faculty in the same university while

completing her master's degree in computational chemistry and minor degree in mathematics. In

2000, she took and topped the Philippine Licensure Examination in chemistry. In 2001, she was

accepted as a research scholar in University of Tsukuba, Japan for a year. In 2007, she earned her

Ph.D. degree in analytical chemistry under the research supervision of Prof. James D.

Winefordner at the University of Florida.





PAGE 1

1 LASER PHOTOFRAGMENTATION AND H ETEROGENEOUS CHEMILUMINESCENCE FOR NITRO-BASED EX PLOSIVE DETECTION By MARIA PAMELA PINEDA MONTEROLA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Maria Pamela Pineda Monterola

PAGE 3

3 Gratefully dedicated to my family and to the loving memories of my father

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank many people for all their help in completing this dissertation and in my journey through graduate school. My main advi ser, Prof. James D. Winefordner, has been a wonderful mentor and inspirati on. I greatly admire his dedicati on to education and research. I would like to acknowledge Dr. Nicol Omenetto fo r his valuable research guidance. He never ceases to amaze me with his creative ideas, and hi s passion for science is really contagious. I am very grateful to Dr. Benjamin W. Smith, a.k.a. as the Mac Gyver of our la b, for all his advice in instrument design and his cool ways in trouble shooting. I would like to extend my appreciation to Dr. Igor Gornushkin for ma ny intellectual conversations about spectroscopy and life. I would like to thank my committee member s, Dr. William Harrison, Dr. David Hahn and Dr. Michael Scott for their valuab le comments and suggestions in the completion of this research dissertation. I would like to recognize the US Army as a main source of research funding. A special thanks to Dr. John Birks of Univer sity of Colorado for providing us the gasliquid exchange module and Dr. Mike Shepar d of US Naval Surface Warfare Center for providing us the with explosive samples. I w ould also like to acknowledge Duran and Tans research groups for allowing me to use their micro balance and spectrofluorometer. All the UVVis measurements were conducted in the laboratory of Dr. Kathryn Williams. I would like to express my deep gratitude to all support sta ff in chemistry, including the machine and electronic shops for a ll of their contribution in instru ment development. It has been a blessing to have Ms. Lori Clark from the Gr aduate Student Coordina tors Office and Ms. Jeanne Karably from the Analytical Chemistr y Division with their unwavering patience and assistance.

PAGE 5

5 My five years at the University of Florida w ould not have been so en joyable if it were not for the wonderful people of Omenetto-Smith -Winefordner group. I would like thank Ron Whiddon, Nick Taylor, Benoit Lauly and Jonathan Me rten for being ready to help with anything in the lab; with lasers, lifting, gas refilling and so on. I would like to recognize Jonathan Merten again for his editorial assistance. A special than ks to Dr. Joy Guingab, Mariela Rodriguez, Dr. Xi-hong Wu, Akua Oppong-Anane and Lydia Edward s for friendship. Those fun times and great memories will always be treasured and cherished. Heartfelt thanks goes to my fr iends who made my stay here at Gainesville pleasant these five years. I am very grateful to Ms. Anne Moore for being so kind to me and for being my second mom here in Gainesville. I would like to extend my gratitude to Merilla Stefan for her camaraderie, Alexandr Oblezov for being so reli able at all times, especially during my second year here at UF, Chen Liu and Ozlem Demir for their companionship and for setting examples of hard work and perseverance. I am so blessed to have a very good support group from my fellow Filipino graduate students. My heartfelt gratitude goes to Jhoanna Mendoza, Dr. Suzette Pabit, Dr. Jemy Gutierrez, Machel Ma lay, Joey Orajay, Attorney Tesi Lou Guanzon, and Ai-ai Cojungco. They helped make Gain esville a second home for me. Most of all, I would like to thank my family for their incr edible support and love. I would like to acknowledge my brothers, Conrad, Chris and Carlo. My brothers and I shared wonderful memories of playing basketball, chess, and play st ations which instilled in me the art of being a warrior in any game of life. I am forever grateful to my mother for her incredible strength and unconditional love. I am also fore ver in debt with my fathers legacy, of being so humble, kind, and handsome.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES................................................................................................................ .......10 LIST OF ABBREVIATIONS........................................................................................................13 ABSTRACT....................................................................................................................... ............15 CHAPTER 1 INTRODUCTION..................................................................................................................17 Background and Significance ................................................................................................17 Chemistry and History of Explosives..............................................................................17 Significance of Explosive Detection...............................................................................20 Review of Methods for Explosive Detection..................................................................22 Challenges on Explosive Detection.................................................................................27 Scope of Research Dissertation..............................................................................................27 2 PHOTOFRAGMENTATION PATHWAYS OF NITRO COMPOUNDS............................38 Introduction................................................................................................................... ..........38 Basic Theory of Photofragmentation...............................................................................38 Basic Theory of REMPI-TOFMS and LIF......................................................................40 PF Mechanism of Explosives based on PF-REMPI-TOF and PF-LIF............................42 Experimental Section........................................................................................................... ...46 Reagents and Chemicals..................................................................................................46 Apparatus and Methodology...........................................................................................47 Results and Discussion......................................................................................................... ..49 Detection of NO2 Fragments...........................................................................................49 Photofragmentation of NO2.............................................................................................51 Conclusion..................................................................................................................... .........53 3 NO2 DETECTION BY CHEMILUMINESCENCE..............................................................65 Introduction................................................................................................................... ..........65 Basic Theory of Chemiluminescence..............................................................................65 Luminol-NO2 Chemiluminescence.................................................................................67 Review of CL Methods for NO2 detection......................................................................68 Experimental Section........................................................................................................... ...70 Reagents and Chemicals..................................................................................................70 Apparatus and Methodology...........................................................................................70

PAGE 7

7 Results and Discussion......................................................................................................... ..73 Chemical Parameters.......................................................................................................73 Physical Parameters.........................................................................................................75 Performance of Luminol-CL Detector............................................................................76 Performance of Luminol/H2O2 CL Detector...................................................................77 Luminol CL Detector Compar ed to Other CL Set-ups...................................................77 Conclusion..................................................................................................................... .........78 4 CONVERSION OF NO TO NO2...........................................................................................98 Introduction................................................................................................................... ..........98 Experimental Section........................................................................................................... ...99 Reagents and Chemicals..................................................................................................99 Apparatus and Methodology...........................................................................................99 Results and Discussion.........................................................................................................100 Conclusion..................................................................................................................... .......100 5 LASER PHOTOFRAGMENTATION AND CHEMILUMINESCENCE FOR NITROBASED EXPLOSIVE DETECTION...................................................................................103 Introduction................................................................................................................... ........103 Experimental Section........................................................................................................... .105 Reagents and Chemicals................................................................................................105 Apparatus and Methodology.........................................................................................106 Results and Discussion.........................................................................................................107 Relative PF Efficiency of Nitro-aromatic Compounds.................................................107 NO2 Photolysis with Luminol-CL Detector...................................................................108 Analytical Capability of PF-CL Detector......................................................................108 Conclusion ............................................................................................................................ 111 6 DIRECT DETECTION OF EXPLOSIVE IN SOIL............................................................126 Introduction................................................................................................................... ........126 Environmental Hazards of E xplosive Contaminated Soil.............................................126 Review of Explosive Detection in Soil.........................................................................127 Experimental Section........................................................................................................... .128 Chemical and Reagents.................................................................................................128 Apparatus and Methodology.........................................................................................129 Results and Discussion.........................................................................................................130 Conclusion..................................................................................................................... .......131 7 CONCLUSIONS AND FU TURE WORKS.........................................................................136 Summary and Conclusions...................................................................................................136 Future Research Directions...................................................................................................137

PAGE 8

8 LIST OF REFERENCES.............................................................................................................142 BIOGRAPHICAL SKETCH.......................................................................................................148

PAGE 9

9 LIST OF TABLES Table page 1-1 Physical and chemical properties of nitrocompounds used in the study...........................31 1-2 Physical and chemical properties of some nitro-based explosives....................................32 1-3 Summary of the performance of various e xplosive detectors reviewed in this study.......34 2-1 Colorimetric analysis of NO2 fragments after photofrag mentation of some nitrobased explosives............................................................................................................... ..63 3-1 Summary of the LOD obtaine d from different luminol-NO2 CL set-ups..........................97 5-1 Comparative table showing summary of the performance of various explosive detectors based on PF-FD technique................................................................................113 5-2 Determination of PF efficien cy of nitro-aromatic compounds........................................118 5-3 Analytical figures of merit of PF-lu minol CL detector for some nitro-based explosives..................................................................................................................... ....125 6-1 Explosives concentration in the contaminated soil..........................................................134

PAGE 10

10 LIST OF FIGURES Figure page 1-1 Classification of explosives ba sed on composition and performance...............................29 1-2 Structural formulas and names of nitrocompounds used in the study...............................30 1-3 Chart showing some of th e technologies presently available and under development for the detection of trace explosives..................................................................................33 2-1 The fate of electronicall y excited species (AB*) formed after absorption of light, ( h ).............................................................................................................................. .......54 2-2 Schematic representation of the photodisso ciation of a triatomi c molecule (ABC)..........55 2-3 Various molecule ionization pr ocesses through photon absorption..................................56 2-4 Schematic of REMPI apparatus.........................................................................................57 2-5 Schematic of PF-LIF apparatus.........................................................................................58 2-6 Photofragmentation set-up. A) schematic of PF-NO2 fragment collection set-up B) actual picture of the PF cell...............................................................................................59 2-7 Nitrogen dioxide detection by colorimetric method..........................................................60 2-8 Schematic of time-resolved absorbance experimental set-up............................................61 2-10 Results obtained with fast time resolved experiments.......................................................64 3-1 Schematic diagram for static Chemiluminescence (CL) device........................................79 3-2 Schematic diagram for continuous Chemiluminescence (CL) device:..............................80 3-3 Luminol oxidation mechanism..........................................................................................81 3-4 Luminol-CL detector set-up...............................................................................................82 3-5 Schematic of spectrofluorometer used for acquiring luminol-NO2 CL emission spectra........................................................................................................................ ........83 3-6 Chemiluminescence Set-up................................................................................................84 3-7 Luminol/H2O2-CL detector................................................................................................85 3-8 Dilution set-up............................................................................................................ .......86 3-9 Effect of KOH concentration ............................................................................................87

PAGE 11

11 3-10 Effect of luminol c oncentration on CL signal....................................................................88 3-11 Effect of adding various t ype of alcohols on CL signal.....................................................89 3-12 Effect of p-iodophenol concentration on CL signal...........................................................90 3-13 Luminol-NO2 Emission Spectra........................................................................................91 3-14 Effect of gas sample flow rate on CL signal......................................................................92 3-15 Effect of the orie ntation of luminol-NO2 mist with respect to the PMT window on signal response................................................................................................................ ...93 3-16 Effect of the storage time of luminol solution on CL signal..............................................94 3-17 Calibration curve of NO2 using Luminol-CL set-up..........................................................95 3-18 Calibration curve of NO2 using Luminol/H2O2-CL set-up................................................96 4-1 Experimental set-up used for evaluating the efficiency of CrO3 oxidation unit..............101 4-2 CL spectra................................................................................................................. .......102 5-1 Photofragmentation-Chem iluminescence detector..........................................................115 5-2 Photofragmentation set-up used for NO2 photolysis.......................................................116 5-3 Chemiluminescence signals obtained for n itro-aromatic compounds at varying laser energy......................................................................................................................... ......117 5-4 Chemiluminescence signals obtai ned before and after PF of NO2..................................119 5-5 Effect of varying laser energy on the CL signal of NOx(x =1,2) fragments obtained after photofragmentation of A) solid a nd B) vapor phase of PETN........................................120 5-6 Effect of varying total number of laser pulses on the CL signal of NOx(x =1,2) fragments obtained after PF of so lid and vapor phase of PETN.....................................121 5-7 Effect of varying total number of la ser pulses on the mass of PETN ablated and on the CL signal of NOx(x =1,2) fragments obtained after PF of solid PETN......................122 5-8 Effect of varying laser energy on the ma ss of PETN ablated and on the CL signal of NOx(x =1,2) fragments obtained after PF of solid PETN....................................................123 5-9 Calibration curves of PETN, RDX, and TNT..................................................................124 6-1 Cross section of the CL cell.............................................................................................132 6-2 Calibration curve for PETN contaminated soil................................................................133

PAGE 12

12 6-3 CL Spectra after PF of A) empty sample holder B) pure NaNO2 C) pure NaNO3..........135 7-1 Alternative scheme for detec ting peroxide-based explosives..........................................141

PAGE 13

13 LIST OF ABBREVIATIONS Chemical Compounds 3-APA* 3-aminophthalate BuOH butanol DMNB 2,3 dimethyldinitrobutane 2,4 DNT 2,4 dinitrotoluene DNB dinitrobenzene EGDN ethylene glycol dinitrate EtOH ethanol HMX High Melting eXplosive MeOH methanol Luminol 5-amino-2,3-dihydr o-1,4-phthalazine dione NO2 nitrogen dioxide NO nitrogen oxide 4 NT 4-nitrotoluene NOx(x=1,2) NO + NO2 PETN Pentaerythridoltetranitrate RDX Research Department eXplosive TEA triethanolamine TNT trinitrobenzene TNT trinitrotoluene Techniques CE Capillary Electrophoresis CI-MS Chemical Ionization-Mass Spectrometry CL Chemiluminescence CRS Cavity Ringdown Spectroscopy DIAL Differential Absorption LIDAR DIRL Differential Reflection LIDAR ECD Electron Capture Detection EI Electron Ionization FL Fluorescence FNA Fast Neutron Analysis GC Gas Chromatography GC-TEA Gas ChromatographyThermal Energy Analyzer HPLC High Performance Liquid Chromatography IMS Ion Mobilty Spectrometry IR Infrared Spectroscopy LIBS Laser Induced Breakdown Spectroscopy LIDAR Laser Light De tection and Ranging PD photodissociation PF-FD Photofragmentati on-Fragment Detection PF photofragmentation PF-LIF Photofragmentation-La ser Induced Fluorescence

PAGE 14

14 Techniques (continued) PF-REMPI PF-Resonance Enhanced Multi-photon Ionization PL Photoluminescence TNA Thermal Neutron Analysis TOFMS Time-of-Flight Mass Spectrometry Units of Measure mC milli Coulomb ppm parts in 10 6 ppb parts in 10 9 pptr parts in 10 12 ppmv parts in 10 6 vapor ppbv parts in 10 9 vapor pptv parts in 10 12 vapor SLPM standard liter per minute Miscellaneous ArF Argon Fluoride laser BDE Bond Dissociation Energy EEG Electroencephalogram HE high explosives Hs heat of sublimation LE low explosives LOD Limit of Detection MEMS micro electrochemical systems NIST National Institute of Standards and Technology NDF neutral density filter PD photodiode PMT photo multiplier tube SAW surface acoustic wave (S/N) signal-to-noise ratio

PAGE 15

15 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy LASER PHOTOFRAGMENTATION AND H ETEROGENEOUS CHEMILUMINESCENCE FOR NITRO-BASED EX PLOSIVE DETECTION By Maria Pamela Pineda Monterola August 2007 Chair: James D. Winefordner Major: Chemistry The purpose of this research was to devel op a simple, fast, reliable, sensitive and potentially portable explosive detection device employing laser phot ofragmentation (PF) followed by heterogeneous chemiluminescence (C L) detection. The PF process involves the release of NOx(x=1,2) moieties from explosive compounds su ch as TNT, RDX, and PETN through a stepwise excitation-dissociation pro cess using a 193 nm ArF laser. The NOx(x=1,2) produced upon PF is subsequently detected by its CL reaction with basic luminol solution .The intensity of the CL signal was detected by ther moelectrically cooled photomultip lier tube with high quantum efficiency and negligible dark current counts. The research work was divided into four stag es: (1) discerning the PF pathways of nitrobased explosives (2) luminol-CL method de velopment and optimization for improving NO2 sensitivity (3) construction of CrO3 oxidizer for NO to NO2 conversion (4) evaluation of the device in detecting nitro-based explosives in various media including air and soil matrix. In the first stage of the research, the followi ng PF pathways were verified using classical colorimetric analysis and a fast time resolved absorbance method. R NO2 + hv R + NO2; hv = 193 nm (1) NO2 + hv NO + O*; hv = 193 nm (2)

PAGE 16

16 The second stage involved the development and optimization of NO2 detection using luminol CL. In this system, a stream of NO2 gas passed through a con centric nebulizer and was used to aspirate a spray of luminol so lution. The aspiration process maximized NO2/luminol contact area enhancing the CL signal. The curre nt system was able to improve the detection limit through optimization of all the feasible physical and chemical parameters involved in the CL reaction between luminol and NO2. Detection limit (LOD) of 19 ppt NO2 at (S/N) = 3 was reported. The optimal reagent solu tion from the viewpoint of sensitivity of the response to NO2 (maximum signal/signal noise ratio) is 5 x 10-3 M luminol + 0.01 M p-iodophenol + 0.2 M KOH. A Luminol/H2O2 CL set-up was also explored where a bundle of porous polypropylene fibers was used to bring the NO2 into contact with luminol solution. A LOD of 178 ppt NO2 at (S/N) = 3 was obtained. In the third stage, a CrO3 oxidizer was constructed for NO to NO2 conversion. Photolysis of nitro compounds produced NO (reaction 2) whic h is not detectable by luminol-CL system. To further increase the sensitivity of the device for NO2 fragments, NO must be converted back to NO2. Finally, the fourth stage was in tegration of PF and CL units into an explosive detection device via a CrO3 oxidizer. The system was able to detect energetic materials in real time at ambient conditions. Detection limits of 3.4 ppbv for PETN, 1.7 ppbv for RDX, and 34.5 ppbv for TNT were obtained. It was also demonstrated that the presence of PETN re sidue within the range of 61 to 186 ng/cm2 can be detected at a given signal to background ratio of 10 using a few micro joules of laser energy. The technique also demonstr ated its potential for di rect analysis of trace explosive in soil. LOD range of 0.5 to 4.3 pp m for PETN was estab lished, an analytical capability comparable to currently available techniques.

PAGE 17

17 CHAPTER 1 INTRODUCTION Background and Significance Chemistry and History of Explosives A chemical explosive is a co mpound or a mixture of com pounds which, when ignited by heat, impact, friction, or shock, undergoes very rapid and self propagating decomposition that results in formation of more stable material, the liberation of heat, development of a sudden pressure effect, and production of loud noi se and a flash which is called explosion.1 Explosives have been classified based on their composition and ra tes of decomposition (Figure 1-1). Low explosives (LE) were prepared by mixing solid oxidizers and fuels, a process that has remained virtually unchanged for centuri es. Propellants and pyrot echnics are included in this category. Propellants are us ed in guns and rockets. Most rocket propellants are composites based on a rubber binder, ammonium perchlorate oxidizer, and a powdere d aluminum fuel; or composites based on a nitrate ester (e.g., nitrocellulo se or nitramines) or nitroglycerine. On the other hand, pyrotechnics include il luminating and signaling flares. Py rotechnics are composed of organic oxidizer and metal powde r in a binder. Illuminating fl ares contain sodium nitrate (NaNO3), magnesium (Mg), and a binder while signali ng flares contain barium (Ba), strontium (Sr), or other metal nitrates. High explosives (H E) were prepared using mixtures of mononuclear energetic material such as trinitrotoluene (TNT ), in which each mol ecule contain both the oxidizing and the fuel components. Since, HE act as their own oxidant as well as fuel, their total energy is much lower than that of LE. Howeve r, the rate at which the combustion energy is released in HE is relatively faster when compar ed to the release rate of LE. The speed of HE reactions gives them a greater power than LE. HE detonate at velocity of km s-1 while LE rapidly burn or deflagrate at relativ ely low rate range of cm s-1.

PAGE 18

18 HE are further classified as primary and secondary based on their susceptibility to initiation. Primary explosives such as lead azid e and lead stryphrate ar e highly susceptible to initiation. Ther are also referred to as initiating explosives because they can be used to ignite secondary explosives. Secondary explosives which include nitroaromatics (e.g., TNT), nitramines (e.g., RDX) and nitrate ester (e.g., P ETN), are formulated to detonate only under specific circumstances. Secondary explosives ar e often used as main charge or bolstering explosives.2 PETN, RDX, TNT, NT, and 2,4 NT were the model compounds used in this study. Their structural formulas as well as the summary of their physical and chemi cal properties are shown in Figure 1-2 and Table 1-1, respec tively. Other common explosives ar e also listed on Table 1-2. Trinitrotoluene (TNT) was first synthesized in 1893 by a German chemist Joseph Wilbrand.3 The synthesis is performed in two steps. Firs t, toluene is nitrated with a mixture of sulfuric and nitric acid to produce mono and dinitrotoluence. Next, the mixture of mono and dinitrotoluence is further nitrated with a mixture of nitric acid and oleum, a more potent nitration reagent. TNT is a pale crystalline aromatic hydrocarbon that was originally used as a yellow dye. Its potential as an explosive wa s not appreciated for several year s mainly because of it does not detonate easily. In order to detona te, TNT must be confined in a casting or shell and subjected to severe pressures and/or temperatures (936oF), such as from a blasting cap or detonator. In fact, US army tests on pure TNT show that when struck by a rifle bullet, TNT failed to detonate 96% of the time and when dropped from an altitude of 4000 feet onto concre te, a TNT filled bomb failed to explode 92% of the time.3

PAGE 19

19 TNT was first used on a wide scale during World War I. The importance of TNT as a military explosive is based upon its safety in ma nufacturing, loading, transportation, and storage. As of today, TNT is mainly used as military mu nitions, quarrying activities, and civilian mining. TNT is a constituent of many explosives such as amatol, pentolite, tetrytol, torpex, tritonal, piciatol, and edratol. 2 Research Department eXplosive (RDX), or al so commonly known as cyclonite or hexogen was first prepared in 1899 by German chemist Hans Henning.4 Its explosive properties were not discovered until 1920. In 1940s, RDX was produced by the Bachmann process which reacts hexamine with nitric acid, ammonium nitrat e, glacial acetic acid, a nd acetic anhydride. The crude product is filtered and recrystallized to form RDX. RDX is a white crystalline nitramine compound. It deflagrates rather than explodes and only detonates with a detonator. There are many interpretations of its acronyms including Royal Demolition eXplosive, Research Department (Composition) X, and Research Department eXplosive. The latter is mostly likely correct. In the UK, new military explosives were given an identification number preceded by the letters R D indicating Research Department Number. For some reason, no number was given to this expl osive. Instead the lette r X was appended to indicate unknown with the intention of adding the a numbe r letter. Although a number was issued, the term RDX stuck. 4 RDX was widely used during World War II as a main component of the first plastic explosive. RDX has been used in several terroris t related explosions in India over the years such as the serial bomb blast in Mumbai of March 8, 1993 in which more than 300 people were killed and about 1500 injured. Again, on July 11 of 2006, where a series of powerful explosions took place in seven suburban Mumbais Western Railwa y train line, killing 209 people and injuring

PAGE 20

20 over 700. RDX is a main constituen t of many explosives such as Torpex, Tertrytol, Cyclotol, Composition A, A5, B, C, H6 and C4.2 Pentaerythritol tetranitrate (PETN) was firs t synthesized in 1891 by Tollens and Wiegand.5 PETNs preparation involves the nitration of pent aerythritol with a mixture of concentrated nitric and sulfuric acid. PETN is a white crystalline compound bel onging to the same chemical family as nitroglycerin-i.e., the nitrate esters. RDX is le ss sensitive than nitroglycerin but is easily detonated. It is the least stab le of secondary explosive. In 1912, after being patented by German gove rnment, the production of PETN started. PETN was used by the German Army in World War I.5 It is used in detonating fuses (Primacord) and in a mixture, called Pentolite, with an e qual amount of TNT in grenades and projectiles. During World War II, the Rocket Launcher (co mmonly known as a Bazooka) charged, with 8 ounce of pentolite, could penetrate up to five inches of armor. PETN is a main constituent of many explosives such as Datasheet and Semtre x, a Czech-made explosive that has been used commonly in many terrorist bombings.2 PETN is also used as vas odilator or a heart stimulant. The medicine for heart disease, Lentonitrat, is pure PETN. 5 Significance of Explosive Detection The field detection of explosives is an extremely relevant analytical issue in law enforcement and environmental applications. With the increasing use of explosives by terr orist groups and individu als, law enforcement and security agents are faced with the problem of detecting hidden explosiv es in luggage, mail, vehicles, aircraft, on traveler s and so on. Counter surveillan ce activities also include the detection of trace explosives duri ng investigation of bombing scenes Post explosion analysis is a

PAGE 21

21 great importance for such investigations, particul arly for drawing connections between cases and suspects. The September 11th disaster on the twin towers was one of the most terrible attacks worldwide. It was a temporary high point of numerous horrible attacks the recent years. Very often, nitro-based explosives are involved in these attacks including the detonation of 250 g plastic explosive to bring down Pan American Flight 103 in Scotland in 1998 and the bombings at the World Trade Center in 1993, Oklahom a City in 1995, Moscow in 1999 and 2002 and many more.6 To protect society against such attacks, early detection of explosive is needed. Explosive analysis is also necessary in order to address environmental and/or health related concerns. Explosive materials exhibit some toxi city, and ingestion or inhalation of explosive materials may cause a significant health problems. Sites where analysis may be required include military test ranges and industrial facilities where explosive materi als are manufactured in order to ensure the safety of the workers and nearby residents. In additions, sites may have become contaminated by improper storage and disposal of explosive materials, and analysis may be required as part of environmental clean-up. The detection of landmines is an urgent wo rldwide problem that requires specific and effective field detection methods. There are at least 125 million unexploded landmines buried in 70 countries around the world.7 Detection of landmines is essential to prevent the leaking of undetonated explosives into water and soil. Th e inherent difficulty in mine clearance is complicated by the great variety of materials in use. Most modern landmines contain only a small amount of metal, so that traditional dete ction techniques based on metal detectors are less useful.

PAGE 22

22 Review of Methods for Explosive Detection Extensive efforts have been devoted to the development of effective and innovative explosive detection systems. There are various techniques currently available and under development. Explosive detection techniques can be broadly classi fied into two categories: bulk detection and trace detection. In bulk detection, a macroscopic mass of explosive material is detected directly, usually by viewing images by X -ray scanners or similar equipment. In trace detection, the explosive is detected by chemi cal identification of microscopic residues of explosive compound. In trace explosive detection, the explosive residue can be in either or both of two forms: vapor and partic ulate. Vapor detection examines the vapor emanating from the concealed explosive samples and particulate de tection examines the microscopic residues of explosives that would be present on i ndividuals or contaminated materials. X-ray detection is a well esta blished conventional system wh ere X-ray energy is measured after passing through an object. This bulk expl osive detection approach is capable of differentiating substances according to their de nsities and atomic numbers. Commercial X-ray detection systems used for airport baggage insp ection have a throughput ra te of up to 800 bags per hour.7 No quantitative data on sensitivity and false alarm rate have been published for this method. Other types of bulk explosive detector s are nuclear inducing techniques which used neutrons instead of photons. Among of these me thods are Thermal Neutron Analysis (TNA) and Fast Neutron Analysis (FNA). Figure 1-3 illustrates the clas sification of explosive trace detectors based on the operating principle of the method. Commonly used explosive detectors can be broadly classified as: (1) separation-based techniques, (2) mass sensors, (3 ) electrochemical sensors, (4) biosensors and (5) optical sensors (6) photofragmenta tion-fragment detection approaches.

PAGE 23

23 Separation-based techniques include High Performance li quid Chromatography (HPLC), Capillary Electrophoresis (CE) 8; 9 and Gas Chromatography (GC) combined with Time-of-Flight Mass Spectrometry (GC-TOFMS)10, Electron Capture Detector (GC-ECD)11 or Thermal Energy Analyzer (GC-TEA)12. GC can also be coupled with Ion Mobility Mass spectrometry (GC-IMS) 13-18. GC-ECD has an advantage in quantitative explos ive analysis in soil due to its compatibility with acetonitrile, a preferred so lvent for extracting explosive resi dues from soil. ECD is based on the high electron affinity of nitr o moieties present in the molecu les of explosive which readily capture electrons produced by a radioactive 63Ni source. The decrease in a signal due to the consumption of free electrons is propor tional to the analyte concentration. In IMS, explosive vapors are converted to ions at atmospheric pressu re and these ions are characterized by their gas phase mobility in a we ak electric field. Currently, the most widely deployed explosives screening t echnology at airports is IMS, including the Ionscan 400 and Sabre 2000 which detect trace expl osive collected from the contaminated surface of baggage or paper tickets by either vacuum or wiping.14 Recently, a new IMS injection port inlet has been developed which collects the explosive sa mple using solid phase micro extraction.1 Solid phase micro extraction is a simple adsorption and de sorption technique for c oncentrating volatile compounds from air. GC can also be coupled with chemilumines cence detector which is also known as TEA analyzer. Pyrolysis of explosives produces NO fragments that are cons equently reacted with ozone (O3) to produce electronically excited NO2 (NO2*). The NO2* decays back to its ground state with emission of chemiluminescence light in the near-infrared region. The emitted light is

PAGE 24

24 detected by a photomultiplier (PMT). The light in tensity is proportional to the NO concentration and hence, to the explosive concentration. A mass sensor was used as the GC detect or in the EST-4200 Ultimate Vapor Tracer, a hand held instrument used in selected airports here in the U.S. It can also be an explosive detector by itself when combined in an array, which is often refe rred to as electronic noses. Mass sensors include surface acoustic wave (SAW) detectors, normally coated with different semi-selective polymeric layers and micro elect romechanical systems (MEMS), including micro cantilever detectors. Mass sensors typically adsorb the chemicals of interest into the surface and the device detects the change in mass. The dete ction can be through changes in acoustic waves propagating along the surface, as in the case of SAW, or by actual bending or a change in shape of the device as mass is accumulated as in the case of MEMS. Electrochemical sensors (ECD) de tect signal changes in an elec trical current being passed through electrodes that interact w ith the target chemicals. The f act that nitro-based explosives can be reduced into amines makes them an ideal candidate for electrochemical detection. Electrochemical sensors can be cl assified as potentiometric (meas ure of voltage), amperometric (measure of current) and conductometric (measure of conductivity). Biological explosive detectors, including specially trained dogs are considered the most successful and widespread system for explosive detecti on at airports and pub lic areas. In addition to canines, other animal species have been proposed as alternative methods of biological explosive detection. A research pr oject in Tanzania trains African Giant Pouched rats to detect landmines. Reports indicate that rats maybe capable of detecting similarly low level of explosive odor compared to dogs.19 Bees are also being studied as a biological explosive detection system. It has been demonstrated that bees have a sense of smell comparable to dogs and are capable of

PAGE 25

25 detecting explosives odors at concentr ations lower than most instruments.20 The bees can be imaged or traced to the source, or more comm only, used to survey areas by examining the chemical residues brought back to hives. Advantag es include that they ca n be trained quickly and will not set off any mines. Limitations include the bees do not fly at night, in heavy rain, or in cold weather (below 40oF). 20 The immunosensor is a class of explosive bios ensors and involves th e use of antibodies as the biosensing element. Reacti on takes place between a target analyte and a specific antibody. The basis of assay is the change in fluorescen ce emission intensity of a fluorescently labeled TNT analogue pre-bound to an anti-TNT antibody. The change in intensity occurs due to competitive displacement of the labeled TNT by the TNT from the samples. Optical techniques under inves tigation include absorption ba sed detection. This is the simplest method among different spectrometric met hods. Several kinds of color change chemical sensors have developed for rapid onsite de tection of explosives. On the other hand, photoluminescence (PL) detection is based on mon itoring the luminescence of an electron rich semi-conducting polymeric film on exposure to the analyte in a flowing air stream. The PL is quenched on exposure to nitroaromatic compounds by electron transfer. In Laser Induced Breakdown Spectroscopy (LIB S), a short laser pulse is focused on the sample. Laser energy heats, vaporizes, atomizes, and ionizes the sample material, generating a small plasma plume. As the plasma decays or cools down (approximately 1 s after the laser pulse), excited atoms and ions in the plasma emit a secondary light, which is collected and spectrally resolved by the spectrophotometer, then analyzed by a light detector. As a result, the multi-elemental composition of the sample can be determined. For explosive detection, different

PAGE 26

26 H, O, and N emission intensity ra tios were studied to identify th e analyte as organic explosive, organic non-explosive or non-organic sample. The Terahertz (THz) explosive sensor is ba sed on differential absorption. The given item or region is illuminated by THz radiation cont aining at least two dis tinct frequencies. The frequencies of radiation depend on the THz spectr a of the targeted explos ives and are chosen so as to maximize the contrast betw een the presence and absence of explosives. THz spectroscopy has not only the capability to de tect the presence of explosive but can also determined the specific type of explosive as different explosives have unique THz specific finger prints. Raman spectroscopy can also be used for de tecting explosive vapor by monitoring the NO2 stretching modes between 1300 and 1370 cm-1. This method is based on the inelastic light scattering effect due to interaction of light wi th a sample. This technique is hampered by the weak Raman signal. Stand-off technologies under development in clude Laser Light Detection and Ranging (LIDAR), differential absorption LIDAR (DIAL), and differential reflection LIDAR (DIRL) for imaging and LIBS as well. Cavity ringdown spectroscopy (CRS) is a suitable direct techni que for vapor detection that passes a high intensity pulse tunable laser beam through highly reflective mirrors, which form the cavity. The intensity of the time dependent de cay of the light leaking through the mirror is measured. The Photofragmentation-Fragment Detection (PF-FD) approach is based on detecting NO2 moieties after photofragmentation of explosive compounds. Photofragmentation is usually done by pyrolysis or UV la ser irradiation. NOx(x = 1,2) fragments are then detected by TOFMS, fluorescence, or TEA analyzer.

PAGE 27

27 Table 1-3 summarizes the results obtained w ith different types of explosive detectors except for PF-FD approach which will be discussed comprehensively in succeeding chapters. Challenges on Explosive Detection There are several obstacles to fast and easy vapor explosive detection. The vapor pressures of explosives are low (Table 1-1 and 1-2). The e quilibrium vapor pressure of most explosive at ambient temperature ranges between 10-4 to 10-9 torr and corresponds to equilibrium vapor concentrations in the region of 10 g/L down to 10 ng/L.20 This vapor concentration is measured for pure samples at equilibrium in a closed contai ner. However, in the field, the environment will not be a closed system and may contain possi ble interferents. In addition, the deliberate concealment (e.g. hard top case, suicide bombers, et c.) of the explosive influences the detection. As a rough rule of thumb, a method that is suit able for direct explosive vapor detection should be able to detect concentrations down to less than 1 ng/L (ppb level).20 The limited sample size available for analysis is not only a critical problem with explosive vapor but also with contaminated soil samples as well. Current methods used for detection of explosive in soil such as GC-ECD and HPLC are inadequate due to their slow detection time and tedious sample preparation. Scope of Research Dissertation There is no single explosive detector that pr omises speed, selectivity, and sensitivity all together. Although each of the methods mentione d above exhibits one or more inherent advantages and disadvantages, their common feature is a high degree of complexity which hinders their potential to be miniature mass-deployable devices. IMS are the only commercialized handheld trace explosive det ectors usually employed in airports. IMS instruments are quite expensive and use a radioactive source (63 Ni), which is not environmental

PAGE 28

28 friendly. Because of these limitations, there is a pressing need for improved instrumental technique for trace explosive detection. The focus of this dissertation is on the developm ent of a simple, fast, reliable, sensitive and potentially portable explosive detection de vice employing laser photofragmentation followed by chemiluminescence. In Chapter 2, generation of NO2 fragments and its further dissociation into NO and O* was confirmed after photolysis of nitro-contai ning explosives by a 193 nm ArF laser using classical colorimetric analysis and a fast time -resolved absorption experiment, respectively. In Chapter 3, NO2 sensitivity was improved using luminol-chemiluminescence technique. In Chapter 4, an oxidation unit was cons tructed and evaluated for NO to NO2 conversion. Chapter 5 involved the integration and evaluation of th e photofragmentation and chemiluminescence units as a single device for detecting nitro-based explos ives. In Chapter 6, the capability of the system to directly detect an explosive in soil was evaluated. General conclu sions and recommendations for future directions are described in Chapter 7.

PAGE 29

29 Figure 1-1. Classification of explosives based on composition and performance. Explosives LOW Explosives HIGH Explosives Primary explosives Secondary explosives

PAGE 30

30 NO 2 O 2 NNO 2 NN N NO 2 NO 2 O 2 N O O O O NO 2 NO 2 O 2 N NO 2TNT RDX PETN NO 2 NO 22,4 DNT NO 24 NT Figure 1-2. Structural formulas and na mes of nitrocompounds used in the study.

PAGE 31

31 Table 1-1. Physical and chemical propert ies of nitrocompounds used in the study. Properties TNT RDX PETN Chemical Formula C6H2(NO2)3CH3 C3H6N6O6 C5H8N4O12 IUPAC name 2-methyl-1,3,5trinitrotoluence 1,3,5-trinitro-1,3,5triazacyclohexane 1,3-dinitro-2,2bis(nitramethyl) propane Molar mass (g/mol) 227.131 222.117 316.140 Appearance yellow needles white crystalline solid white crystalline solid Density (g/cm3 at 20oC) 1.65 1.82 1.77 Melting point (oC) 80.4 205.5 141.3 Boiling point (oC) Decomposes at 295 Ignites at 234 Decomposes at 190 Solubility in water at 20oC (mg/L) 130 40 43 Solubility in ether,benzene,acetone, pyridine Acetone, acetronitrile benzene, sparingly solouble in alcohol,ether and acetone Vapor pressure at 25oC (torr) 5.8 x 10 -6 (7.7 ppb) 4.6 x 10 -9 (6 ppt) 1.4 x 10 -8 (18 ppt) Shock sensitivity insensitive low medium Friction sensitivity insensitive low medium Explosive velocity (m/s) 6900 8750 8400 RF Factor* 1.00 1.60 1.66 RF factor is a measurement of an explosives power for military demolition purposes. TNT equivalent is a measure of the energy released from the detonation of a nuclear weapon or a given quantity of fissionable material, in terms of the amount of TNT which could release the same amount of energy when exploded.

PAGE 32

32 Table 1-2. Physical and chemical propert ies of some nitro-based explosives. Explosive class Explosive/ Symbol Formula Molar mass (amu) Vapor pressure at 25oC (torr) Acid salt Aliphatic nitro Ammonium nitrate nitromethane NH4NO3 CH3NO2 80.04 61.04 5.0 x 10 -6 2.8 x 10 1 2,3-Dimethyldinitrobutane/ DMNB C6H12N2O4 176.17 2.1 x 10 -3 Aromatic nitro 2-Nitrotoluene/ C7H7NO2 137.14 1.5 x 10 -1 o -MNT 4-Nitrotoluene/ p -MNT C7H7NO2 137.14 4.1 x 10 -2 2,4-Dinitrotoluene/ DNT C7H6N2O4 182.14 2.1 x 10 -4 2,4,6Trinitrophenol/ Picric acid C6H3N3O7 229.11 5.8 x 10 -9 Nitrate ester Ethylene glycol dinitrate/EGDN C2H4N2O4 152.06 2.8 x 10 -2 Trinitroglycerin/ NG C4H5N3O9 227.09 2.4 x 10 -5 Nitramine TetranitroN methylamine/Tetryl C7H5N5O8 287.15 5.7 x 10 -9 Tetranitrotetrazacyclooctane/ HMX Hexanitrohexaaisowurzitane/ CL20 C4H8N8O8 C6H6N12O12 296.16 438.19 1.6 x 10 -13 N/A N/A not available.

PAGE 33

33 SAWMEMS Potentiometric Amperometric Conductometric Immunosensors DOGS RATS BEES Absorption PL FL LIBS THz Raman IR LIDAR CRS LIQUID HPLCCE GC-MS GC-ECD IMS GC-TEA TRACE EXPLOSIVE DETECTION GAS SEPERATION MASS SENSORS OPTICAL SENSORS PF-REMPI-TOFMS PF-REMPI-ION PROBE PF-LIF PyrolysisChemiluminescence BIOSENSORS PF-FD APPROACH ELECTROCHEMICAL SENSORS Figure 1-3. Chart showing some of the technolo gies presently available and under development for the detection of trace explosives.

PAGE 34

34Table 1-3.Comparative table showing summary of the performance of various explos ive detectors reviewed in this study. Transducer Field of Application Explosive Detected Limit of detection Comments References HPLC-UV Absorption, 254 nm CE GC-Laser Ionization Time-of-Flight-MS GC ECD Aquifer samples collected in the vicinity of military installation Laboratory liquid samples laboratory vapor sample soil sample TNT RDX TNT, TNB tetryl NB 2,4-DNT 4-NT 2,4-NT,TNT RDX HMX PETN,NG 0.10 ng/mL 0.10 ng/mL 60, 160 g/L 200 g/L 17-24 ppb; (S/N=2) ~40 ppb; (S/N=2) 10 g/kg 1 g/kg 3 g/kg 25 g/kg 10-40 g/kg time consuming and tedious sample preparation time consuming and tedious sample preparation, costly bulky time consuming and tedious sample preparation costly; not environmental friendly 21 9 10 11

PAGE 35

35Table 1-3. (Continued). Transducer Field of Application Explosive Detected Limit of detection Comments References GC-TEA laboratory vapor sample 4-NT 2,4-NT TNT 137 g/L 18.2 g/L 22.7 g/L Poor performance for explosive vapor detection, bulky, 12 Electrospray-IMS laboratory vapor sample TNT RDX HMX EGDN 15 g/L 20 g/L 86 g/L 190 g/L IMS in general has a higher % of false positive and negative responses; low mass spectroscopic resolution; uses radio active source which is not environment friendly 13 GC-IMS (multicapillary chromatographic column) laboratory liquid sample PETN TNT 2,4-DNT 8 ng/L 14 ng/L 16 ng/L 15 Corona dischargeIMS laboratory solid sample PETN TNT RDX 8 x 10 -11 g 7 x 10 -11 g 3 x 10 -10 g 16

PAGE 36

36Table 1-3. (Continued). Transducer Field of Application Explosive Detected Limit of detection Comments References MEMS laboratory vapor sample PETN,RDX,TNT low femto gram level 520 ppt Limited applications due to noisy chemical background 22,23 Electrochemical Electrochemical Electrochemical soil sample marine water soil extract and ground water RDX TNT RDX TNT 2,4-DNT ,4-NT 0.12 ppm 25 ppb 60 ppb 0.11 ppm 0.15 ppm In general, electrochemical sensors suffer from limited sensitivity and mobile electrolyte. Also, electrodes can be easily fouled 24,25,26 Canines airports, public places almost all approximately low pptr level subject to fatigue and behavioral variations 20,27 Immunosensors seawater TNT 250 ppt Limited application,anti-bodies are not re-usable, unreliability 28 Absorption laboratory samples of contaminated finger prints DNT DNB N/A Applicable only for qualitative analysis 29 PL based air and sea water TNT picric acid 4 ppb TNT in air,1.5 ppt TNT in sea water, 6 ppb picric acid in sea water bulky, relatively low sensitivity 30

PAGE 37

37Table 1-3.(Continued). Transducer Field of Application Explosive Detected Limit of detection Comments References FL based laboratory samples TNB,TNT,DNB, tetryl,2,4-DNT 1 ppm for all of these explosives 31 FL based Water samples DNP DNT, NB 1 x 10 -6 mol/L 40 ppb 17-24 ppb 32 LIBS Laboratory samples HMX,PETN,HMX, TNT,C4,M43 not available qualitative analysis 33 Tetrahertz Laboratory solid samples RDX not available qualitative analysis 34,35 Raman Laboratory vapor sample 2,4-DNT 5 ppb qualitative analysis 36 Optic probe Raman Laboratory samples of contaminated finger prints RDX, PETN g range qualitative analysis 37 IR Laboratory vapor sample TNT,RDX,PETN not available qualitative analysis 38 Mid-IR CRS Laboratory vapor sample TNT 75 ng/L bulky 39

PAGE 38

38 CHAPTER 2 PHOTOFRAGMENTATION PATHW AYS OF NITRO COMPOUNDS Introduction This chapter begins with the discussion of the basic theory of photofragmentation (PF) followed by a review of previous works elucid ating the PF pathways of nitro compounds using Resonance Enhanced Multi-photon Ionization Ti me-of-Flight Mass spectrometry (REMPITOFMS) and Laser Induced Fluores cence (LIF). Generation of NO2 fragments and their further dissociation into NO and O* was confirmed afte r photolysis of nitro-containing explosives by 193 nm ArF laser using classical colorimetric method and fast time-resolved absorption experiment, respectively. Basic Theory of Photofragmentation Photochemical processes can be classified ba sed on the fate of electronically excited species formed after absorption of light. As shown in Figure 2-1, when a molecule (AB) has absorbed a quantum of radiation, ( hv ), it becomes energy rich or ex cited in the process. There are several pathways by which an electronically species (AB*) may lose its energy. Some of these pathways are through chemical reaction of the reactive fragments formed, isomerization or ionization processes.40 The process whereby the excited st ate species is split into simpler fragments is called photofragmentati on (PF) or photodissociation (PD). There are two important PF rules that may be stated as follows: (1) Only the light absorbed by a system is e ffective in producing a PF in a molecule. This is also known as the Grottus-Draper Law. 41 (2) Each photon absorbed activates one molecule in the primary excitation step of a photochemical sequence. 41 Over the past four decades, various theoreti cal models and experimental techniques have been developed to advance the understanding of molecular dissoci ation induced by photon excitation. For instance, theoretical ly constructed potential energy curves or surfaces for PF of

PAGE 39

39 diatomic or polyatomic molecules have been widely used to predict or ex plain the experimental phenomena.42 As an example, consider the follow ing PF for a hypothetical molecule ABC. ABC + hv AB + C (2-1) In Figure 2-2, the ground and excited state pot ential energy surface ar e termed G and E, respectively while R is the internuclear distance between atom C and the AB moiety and hv is the energy of the laser photon. The ground state pot ential energy surface is bound and evolves at long AB C internuclear distances to an asympto tic energy which corresponds to the energy difference of the separated AB and C fragments. The energy difference between the ground state of the ABC molecule and the gr ound state energy of the two AB and C fragments is the bond dissociation energy (DE). Following the absorption of a photon (of energy h ), the ground state molecule is, in this case, excited to a repulsi ve excited state, (E). Depending on the repulsive character of the excited state, the molecule will dissociate promp tly or more slowly, into the fragments AB and C. The exce ss energy, which is the differen ce between the energy of the photon and the bond dissociation energy, may be di stributed among all available degrees of freedom of the nascent fragments (translational and internal). The rele ase of kinetic energy causes the AB and C fragments to move away from the point in space where they were generated. The proportion of the excess energy rel eased as kinetic energy and the masses of the AB and C fragments determine their recoil speeds. The residual excess energy is divided over the other degrees of freedom of the fragments and may, for instance, lead to rotational, vibrational, and/or electronic excitati on of the fragments. By consideri ng the law of conservation of energy, an expression for the partitioning of energy can be written as follows: E ex = h + Ep int Do(AB C) = EE + ET + EV + ER (2-2)

PAGE 40

40 where available or residual energy ( E ex) consists of the photon energy (h ) plus the internal energy of the parent molecule (Ep int) minus the energy Do(AB C) required to break the B C bond. This residual energy is partitioned into the electronic (EE), translational (ET), vibrational (EV) and rotational (ER) degrees of freedom of the fragments.42 Basic Theory of REMPI-TOFMS and LIF Propellants and explosives cont ain molecules with characteri stic functional groups, namely the NO2 moiety bonded to carbon, ( C NO2 for nitrocompounds), oxygen, ( C O NO2 for nitric ester) or nitrogen (( C N NO2 for nitramines). This functional group is weakly bound to the main skeletal portion of the parent molecule by approximately 40 to 60 kcal/mole depending upon the specific molecule.43 It is this moiety that is respon sible for the weak and structureless absorption observed in the UV region near 230 nm.44 Resonance Enhanced Multi-photon Ionization (REMPI) technique ty pically involves a resonant single or multiple phot on absorption to an electronica lly excited intermediate state followed by another photon which ionizes the atom or molecule. An ion and a free electron will result if the photons have imparted enough energy to exceed the ionization threshold energy of the system. Typically, small molecules composed of light atoms will have ionization energies around 10-15 eV, corresponding to the absorption of three UV photons from the ground state of the neutral molecule.45 The cross section of the ionization process is greatly enhanced if there is a real excited state resonant at the energy of one or two absorbed photons (Figure 2-3).45 Multi-photon fragmentation and ionization can occur by two di stinct mechanisms. First, molecules absorb photons, exciting them to a disso ciative state below the ionization limit. If the laser pulse duration is much shorter than the life time of dissociative stat e, then the up pumping rate may be so high that ionization occurs before dissociation. The ions may absorb additional photons before fragmenting. This process is ionization followed by dissociation (ID) or ladder

PAGE 41

41 climbing.45 However, if the laser pulse duration is longe r than the dissociative state lifetime, then the laser intensity is sufficiently high, these fr agments may absorb additional photons to further dissociate or ionize. This pro cess is dissociation followed by i onization (DI) or ladder switching. Thermally labile compounds, which include ma ny nitro compounds, fragment primarily by the DI route while many aromatic comp ounds fragment by ID route alone.45 A schematic of a typical REMPI apparatus is presented in Figure 2-4. A gas phase sample is introduced into the apparatus, typically in the form of a mo lecular beam. A high power, pulsed laser is used to induce multi-photon absorpti on, resulting in fragmentation and ionization.46 The resulting ions are then detected by Time -of Flight-Mass Spectrometry (TOF-MS). Laser Induced Fluorescence (LIF) is simply the optical emission from molecules that have been excited to higher levels by absorption of laser radiation. LI F has been used to study the electronic structures of the molecules and to make quantitative measurements of their concentrations.46 LIF for explosive detection is usually comb ined with photofragmentation in order to dissociate a target molecule and subsequently de tect the fluorescence of the generated fragments. This is done because the fluorescence of most explosives is weak while the fluorescence intensity of NO(X=1, 2) fragments produced afte r photolysis is strong.47 Figure 2-5 shows a schematic of a PF-LIF expe rimental system. A UV laser is used to dissociate and induce fluorescence in the sample A photomultiplier (PMT) detects the LIF signal which can be averaged and recorded on a digital oscilloscope. The laser wavelength can also be scanned to collect LIF spectra. These spectra can be used to identify the wavelength of the highest intensity LIF signal.

PAGE 42

42 PF Mechanism of Explosives based on PF-REMPI-TOF and PF-LIF In 1983, Butler, et al. 48 studied the dissociation of nitromethane (CH3NO2) at 193 nm using LIF and molecular beam photofragment tran slational energy spectroscopy which is similar to REMPI-TOFMS except for the addition of an el ectron ionization (EI) sour ce at the entrance of TOF analyzer. The primary photod issociation process is shown to be cleavage of the C N bond to yield CH3 and NO2 radicals with quantum yield almost e qual to unity. The translational energy distribution for this chemical pr ocess indicates that there are tw o distinct mechanisms by which CH3 and NO2 radicals are produced. The dominan t (major) mechanism produces CH3 and NO2 in its electronically excited ( 2B2) state. The electronically excited NO2 has sufficient internal energy to further dissociate to NO and O* as follows: NO2 ( 2B2) NO (X2 ) + O (1D) (2-3) In the minor mechanism, the NO2 fragment is produced in the excited ( 2B2) state, which has a large cross section for absorbing another photon. The subsequent absorption of an additional 193 nm photon results in furt her photodissociation of NO2 into NO and O* as shown below. NO2 ( 2B2) + h NO (A2+); h = 193 nm (2-4) Almost at the same time, Blais 49 obtained the same results as Butler et al. and determined that about 30% of the NO2 produced from primary photodissociat ion participates in the second absorption. Fragments from photodissociation were observed at m/z values of 15, 16, and 30 corresponding to CH3, O, and NO. A very small signal at m/z = 46 (NO2) was measured. The NO2 + signal was significantly lower than that of NO+ signal, presumably due to extensive fragmentation of vibr ationally excited NO2 molecules in the ionizer. In 1984, Renlund 50 reported the decomposition of nitromethane (CH3NO2), nitrobenzene (C6H5NO2), and n-propylnitrate (CH3(CH2)2ONO2). The emission of C*, CH*, and NO2* was monitored from 380 to 650 nm. Of particul ar note is the behavior of the NO2 yield as a function

PAGE 43

43 of laser fluence. In the case of n itromethane and n-propylnitrate, the NO2 yield varies linearly with fluence until it saturates; however, satura tion was not observed with nitrobenzene, but instead the NO2 yield increases more sharply with hi gher fluence conditions. It was concluded then that by increasing the rate of energy deposition, the yield of NO2 can be accelerated suggesting that the primary disso ciation of nitrobenzene into C6H5 and NO2 may change as a function of laser intensity. In 1989, Capellos et al. 51 used PF-LIF technique to study the photodecomposition of RDX vapor at 428K using KrF laser with 248 nm a nd 15 ns pulse radiation. They observed prompt emission at 305-320 nm and at 380-420 nm followi ng photolysis of RDX. Based on the spectra recorded, they attributed the UV emission to electronically excited OH (2+) and the structureless visible emission centered at 608 nm to electronically excited NO2 (2B2). Monitoring the signal intensity as a function of laser fluence revealed that NO2 (2B2) was formed by a onephoton process while OH (2+) was formed by a two-photon pr ocess. The production of NO2 (2B2) was claimed to result primarily from the N N bond scission of electronically excited RDX and OH (2+) possibly via dissociation of vibrationa lly excited nitrous acid (HONO) which is generated by a H atom rearrangement in RDX, followed by five member ring formation and HONO elimination. In 1993, Galloway et al. 52 used PF-REMPI-TOFMS in study ing the pathways and kinetic energy disposal of the photodisso ciation of nitrobenzene. At several photolysis wavelengths between 220 and 320 nm, the following photodiss ociation pathways for nitrobenzene were observed: C6H5NO2 C6H5 + NO2 (2-5) C6H5NO2 C6H5NO + O (2-6)

PAGE 44

44 C6H5NO2 C6H5O + NO (2-7) C6H5NO2 C5H5 + CO + NO (2-8) The relative yield of the pathways producing NO2 and NO varies strongly with the photolysis wavelength. The production of NO2 exceeds that of NO by about 50% for the 280 nm photolysis, but increases to almost six fold excess in 222 nm dissociation. Lemire et al.53; 54(1993), Marshall et al. 55(1994), and Ledingham et al. 56(1995) used the same technique and all confirmed the char acteristic fingerprint at m/z = 30 (NO+) for nitrobenzene; 2,4 DNB; TNT; EGDN; P ETN; RDX; and Semtrex sample at = 226 nm. Simeonsson et al. 57(1994) repeat the same experiment using a 193 nm laser instead. The same mass spectra were obtained ex cept for a higher signal intensity at m/z =30 (NO), which can be attributed to the higher pulse en ergy of an argon fluoride (ArF) laser ( = 193 nm) compared to a dye laser ( = 266 nm). In 1996, Moyang et al. 58 studied the photodissociation of nitrocellulose by means of electron impact ionization ion trap mass spectrometer (EI-MS). In this article, nitrocellulose was irradiated at 532, 355, and 266 nm and with a tuna ble dye laser operating in the range of 310-330 nm. The first mass spectrum obtained with 532 nm laser light only showed mass peaks corresponding to background gases: water, nitr ogen and oxygen. However, irradiation with 355 nm light produced three major peaks at m/z = 19, 30, and 48, corresponding to protonated water (H3O+), nitric oxide (NO+), and hydrated nitric oxide complex (NOH2O+). When 266 nm laser light was used, the intensities of the mass peaks occurring at m/z 30 and 48 increased approximately four times in comparison with the same peaks observed using 355 nm light. The wavelength dependence of removing the NO2 moieties from its parent molecule (or denitration) in the 310-330 nm range has been measured, and the NO yield normalized by laser intensity

PAGE 45

45 showed a linear decrease of NO yield with incr easing laser wavelength. All of these results suggest the simple trend that shorter wavelengt h UV light induces more denitration. The weight loss rate of sample due to denitration induced by 266 nm laser light was measured to be 4.0.4 g J-1. This value represents a 5.8% photon-to-n itro group conversion efficiency. A laser intensity threshold about 6 mJ cm-2 was observed; above which thermal decomposition occurs (detonation of sample occurs). Research on photofragmentation of nitro-c ontaining compounds within the last 20 years did not experimentally c onfirm the detection of NO2 fragments which is fundamentally essential in supporting the primary photodissociation of nitro compounds as shown below: R NO2 R + NO2 (2-9) PF-REMPI-EI-TOFMS during 1980s claimed that NO2 + was not detected due to extensive fragmentation of NO2 molecules in the ionizer while PF-REMPI-TOFMS during the 1990s attributed the absence of NO2 + signal due to the absence of a real excited state resonant with the energy of two absorbed photons to ionize NO2 to NO2 +. However, NO2 fragments were detected by its prompt fluorescence emission. Detecting NO2 using this method is not that reliable since the NO2* emission is irradiated over a large spectral region (from visible to near IR) where other excited fragments are also expected to emit. The work presented in this dissertat ion was able to detect directly NO2 fragments from explosives after their photolysis with an Ar F laser by a classical colorimetric method. Furthermore, the photodissociation of NO2 fragments into NO and O* was confirmed using a time resolved absorption experiment.

PAGE 46

46 Experimental Section Reagents and Chemicals Explosive samples. Five percent (v/v) solutions of PETN, RDX, and TNT in acetone were provided by Naval Surface Warfare Center (M D, USA). Acetone was allowed to evaporate at room temperature to obtain the crystallin e samples. Solid reagent grade chemicals (99% purity) of 4-NT and 2,4 NT were used and purch ased from Sigma Aldrich (St. Louis, MO). Colorimetric experiment. All chemicals used were analytical reagents from Fischer Scientific Inc. (Hampton, NJ) unless otherwis e stated. Deionized water from Millipore Milli-Q deionizing system was used for pr eparing all the reagent solutions. Triethanolamine (TEA) solution was prepared by dissolving 15 g of TEA in 1 L of water. Standard Nitrite solution was prepared by dissolving 0.2 g dried NaNO2 (105oC for 1 hour) in 100 mL of water. Exactly one mL of this solution was diluted using TEA solution to get 4 g/mL of nitrite which se rves as a stock solution. Neutral red solution was prepared by dissolving 0.1 g of neutral red (Basic red 5 C.I. 50040) in 1 L of water cont aining 2.5 mL of 4.25 M H2SO4. The dye solution was stable for thirty days. 1 % (wt./v) Sodium hypophosphite(NaH2PO2) was prepared by dissolving 1 g of NaH2PO2 in 100 mL of water. 1 % (wt./v) Potassium bromide (KBr) was prepared by dissolving 1 g of KBr in 100 mL of water. Time-resolved absorption experiment. Copper powder (325 mesh) was purchased from Alfa Products (Danvers, MA) and r eacted with 98 %(v/v) nitric acid (HNO3).

PAGE 47

47 Apparatus and Methodology Photofragmentation-NO2 fragment collection. Figure 2-6A shows the photofragmentation (PF) set-up used for NO2 fragment collection. The set-up consists of the PF cell containing the explosive sample, an argon fluo ride (ArF) laser to photolyze the sample, and a delivery gas system to purge all the photofragm ented products into the glass bubbler containing TEA solution. The PF cell (Figure 2-6B) is made of cylindrical stainless st eel with an O.D. of 1.93 cm and a length of 4.85 cm. Both ends of th e PF cell were enclosed with quartz window (D = 1.80 cm) to allow passage of UV radiation into the cell. The delivery gas system was coupled into the PF cell through solenoid valv es (Asco Inc., NJ) that were attached to both sides of the cell. A tank containing nine ty nine percent pure N2 gas was attached to the entrance of solenoid valve I while the exhaust end of solenoid valv e II connects the PF cell to the gas bubbler. The gas flow rate was controlled by a mass flow cont roller manufactured by A licat Instrument Inc. (Tucson, AZ). In these experiments, solid explosive sample was deposited and tightly packed into a small aluminum cylindrical cup (O.D. = 1.9 mm; Dept h = 4.5 mm). The sample was then securely placed into the circular quartz disk having its center drilled and hollowed to exactly fit the sample holder. This circular disk serves as the backside window panel of the PF cell. An argon fluoride (ArF) excimer laser (GAM Laser Inc., Orlando, FL) was used to generate the photolysis wavelength of 193 nm. Each experimental run used 80 000 laser pulses with an average energy of 8 mJ/pulse. The pulse length and the repetition rate are 15 ns and 50 Hz, respectively. The beam diameter is 3.1 mm. It should be noted that PF is distinct from LaserInduced Breakdown Spectroscopy (LIBS). The laser flux employed in PF is always sufficiently low to avoid laser induced breakdown of the sample medium or the creation of plasma.

PAGE 48

48 The ArF laser ablated and photolyzed the explosive sample while N2 gas (0.5 SLPM flowrate) which serves as the carrier ga s is continuously introduced to purge the photofragmented products into the bubbler cont aining 10 mL of TEA solution. The TEA solution trapped the NO2 fragments into the bulk solution by c onverting them to nitrite ions (NO2 -) as shown by the following equation: N(HOCH2CH2)3 + 2NO2 + 2OH2NO2 + -O+N(OHCH2CH2)3 + H2O (2-10) Colorimetric analysis. The nitrite ions in TEA solu tion were then analyzed using colorimetric method by mixing exactly 2 mL of aliquot solution of TEA into 25 mL standard flask containing 2 mL of 0.01 % neutral red, 1 mL of 4.25 M H2SO4, and 1 mL KBr. This was followed by addition of 1 mL of 1 % NaH2PO2 and mixed thoroughly. The mixed solution was then diluted to mark and the absorbance was measured at 530 nm against distilled water. Preparation of the Calibration Graph Aliquots of the standa rd stock solution of NO2 containing 0 to 85 g of nitrite were added to a series of 25 mL standa rd flask containing the same reagents used for preparing the samples. Absorbance was measured at 530 nm against distilled water. The calibrati on graph obtained was linear but with a negative slope. Nature of Species Responsible for Color. A known excess amount of neutral red on treatment with nitrite solution undergoes diazotiza tion in acidic medium (Figure 2-7A). KBr acts as a catalyst for this reaction On treatment with NaH2PO2, the diazonium salt undergoes deamination and causes a decrease in absorbance proportional to the concentration of added nitrite ions. The absorption spectrum of unr eacted neutral red is shown on Figure 2-7B. PF-time resolved absorption experiment. Figure 2-8 shows the PF set-up used for confirming the further dissociation of NO2 fragments into NO and O* upon absorption of a photon with 193 nm wavelength. The set-up co nsists of the PF cell containing the NO2 gas

PAGE 49

49 sample, an ArF laser to photolyze the sample, and a detection system to monitor the change in the profile of the laser pulse be fore and after photolysis. Neutral density filters were added to attenuate the laser pulse energy and to prevent the saturation of the ET-2030 photodiode detector (Electro-Optics Technology, Traverse City, MI ). The set-up is capable of monitoring the intensity versus time profile of the la ser pulse after photofragmentation of NO2 gas. Nitrogen dioxide gas was produ ced by reacting pulverized Cu with concentrated nitric acid in a sealed 250 mL Erlenmeyer flask (reaction 11). The NO2 gas generated from this reaction was collected using a gas syringe and injected into a quartz cell (Diameter = 2.0 cm; Length = 10.0 cm) through its rubber septum cover. Cu(s) + 4HNO3 (aq.) Cu(NO3)2(s) + 2 H2O (l) + 2NO2(g) (2-11) The ArF laser beam was focused in to the cell and used to photolyze the NO2 gas sample. The photodiode detector is triggered by the PMT syst em to collect the signal from the laser. The intensity versus time profile of the laser pulse was monitored and plotted by fast oscilloscope (HP54542C Oscilloscope, Hewlett Packard, Pab Ator California). The ArF laser was calibrated using an MPE 2500 power meter from Spir icon Power Products Inc., Logan, Utah. Results and Discussion Detection of NO2 Fragments The calibration graph (Figure 2-9) for NO2 fragment detection wa s obtained by plotting absorbance values at 530 nm against the mass of nitrite in g. Linear dynamic response was found within the range of 0 to 85 g NO2 -. The calibration plot is a stra ight line; with correlation coefficient(R) value of -0.98506 and obeying the following equation: A (dye) = -0.00753(m) + 0.65734 (2-12) where A is the absorbance and m is the amount of nitrite in g. The limit of detection (LOD) was 0.78 g (0.02 mmol) nitrite. LOD is de fined by the following equation:

PAGE 50

50 LOD = (ksblk) / m (2-13) where k is the confidence fact or which is equal to 3, sblk is the standard deviation of the blank and m is the slope of the line or the calibration sensitivity. The precision of the method has been establis hed at 19 g nitrite which gave a relative standard deviation (RSD) of 5.66% where n = 10. The amount of NO2 fragments detected for each solid explosive sample is shown in Table 2-1, which experimentally shows the primary photodi ssociation of explosives after photolysis at 193 nm. The amount of the sample ablated was determined from another experimental run where there is no flow of carrier gas. The aluminum cup sample holder was weighed before and after photofragmentation using a microbalance. The fraction of NO2 detected was calculated by the following equation: Fraction of NO2 detected = moles of NO 2 detected in TEA solution NO2 moles ablated (2-14) The fraction of NO2 detected for all the explosive samples was approximately the same with the exception of RDX. RDX signi ficantly higher fracti on of detected NO2 might be due to its PF mechanism as elucidated by Capellos et al.51 Nitramines have a characteristic dissociation pathway which involve HONO elimination from a parent molecule. HONO produced during photolysis can be readily trapped and absorbed in a basic TEA solution, subsequently increasing the fraction of NO2 detected. The validity of the colorimetric analysis has been established using NO2 gas standards having concentrations of 5 and 10 ppm. The NO2 gas was purged into the same bubbler containing 10 mL TEA solution at flow rate of 0.3 SLPM. Exactly two mL of this solution was

PAGE 51

51 analyzed for nitrite content following the pr ocedure described under th e calibration graph. The following equation was used to convert the g NO2 determined into ppm concentration. NO2 (ppm) = (mD)/(40.9VMAS) (2-15) where m is the mass of nitrite ions in g, D is th e dilution factor, V is the volume of the gas in L collected at standard temperature a nd pressure, M is the molar mass of NO2 in g/mol, A is the fraction of NO2(g) absorbed by TEA solution and S is the fraction of NO2 gas converted to NO2 ions. The numerical values of A and S are 0.95 and 0.85, respectively.59 Based on this equation, NO2 concentrations of 6.1.0 and 11.9.4 ppm were obtained after analyzing 5 and 10 ppm NO2 standards, respectively. From these sets of data, th e direct confirmation of NO2 fragments was established after photolysis of explosives at 193 nm. Relatively low fraction of NO2 detected for each sample indicates the further photodissociation of NO2 into NO and O* upon absorption of an additional photon. Photofragmentation of NO2 Nitrogen dioxide exposed to troposph eric UV radiation photodissociate at < 420 nm to give NO and O*.60 NO2 + h NO + O*; h < 420 nm (2-16) A fast time resolved absorption experiment ha s been used to verify the photodissociation of NO2 at 193 nm. Light from ArF excime r laser was used to decompose NO2 at ambient conditions. The laser pulse has a duration of 10 ns and delivered output energy of 10 mJ/pulse at a repetition rate of 20 Hz. The oscilloscope was set to obtain the average of 32 pulses for each run of the experiment under closed batch system. The NO2 concentration diluted in air was approximately around 120 ppm or about 3% in terms of the part ial pressure of NO2 gas with

PAGE 52

52 respect to the atmospheric pres sure. The concentration of NO2 was calculated using the following equation: % NO2 (Po NO2 /Patm) = A R T C x 100% (2-17) 0.434 LPN where A is the initial absorbance (0.7), R is th e universal gas constant (0.08206 L-atm/mol-K), T is the room temperature (298K), C is the conversion factor (1000 cm3/L), is the absorption cross section of NO2 at 193 nm (2.0 x 10-19 cm2)61, L is the length of th e PF cell (10 cm), N is Avogadros number (6.02 x 10 23 molecules/mole). Figure 2-10A shows the profile of the ArF la ser pulse before and af ter the photolysis of NO2 gas sample. Plot A represents the actual prof ile of the laser pulse. Plot B represents the profile of the laser pulse after passing it through an empty PF cell and an added neutral density filter. Plot C represents the profile of the laser pulse after passing it through the PF cell containing NO2 gas sample. From these sets of data, the time-resolved absorbance plots were obtained (Figure 2-10B). If the absorption of light is uniform during the entire duration of the pulse and the detector responds lin early to light, then the plot of time-resolved absorbance versus time will show a constant value for the entire dura tion of the pulse. This is exactly the case after adding a neutral density filter which is shown by the A1(t) plot. On the other hand, if the first part of the pulse induces a dissociation process and the fragments produced do not absorb, the shape of the pulse will be distorted since now the second part of the pulse will show no absorption. The overall result would be a non-cons tant behavior of the time-reso lved absorbance plot during the entire duration of the pulse as shown by the A2(t) plot.

PAGE 53

53 Conclusion The following PF pathways were confirmed using classical colorimetric analysis and fast time resolved absorbance. R NO2 + h R + NO2; h = 193 nm (2-18) NO2 + h NO + O*; h = 193 nm (2-19) Confirmation of NOx(x = 1,2) generation after photolysis of nitr o-based explosives was the basis for developing an explosive detector ba sed on PF-Fragment detection scheme

PAGE 54

54 Figure 2-1. The fate of electr onically excited species (AB*) formed after absorption of light, ( h ). AB + h AB* BA Isomerization A + B Dissociation AE + B AB+ + eIonization Direct reaction

PAGE 55

55 Figure 2-2. Schematic representation of the phot odissociation of a triato mic molecule (ABC). R is the distance between C and the AB diat omic fragment, DE is the bond dissociation energy and h is the energy of the photon. Th rough absorption of a photon, the molecule is excited to a repulsive state via which it dissociates. The excess energy ( Eex) is the energy difference between the photon energy and bond dissociation energy. 0 h DE Eex V(R) R ABC ( G ) ABC ( E ) AB + C

PAGE 56

56 Figure 2-3. Various molecule ionization processes through phot on absorption. A) Ionization via direct one-photon aborption B) non-res onant multi-photon ionization and (c) resonance enhanced multi-phot on ionization (REMPI). Proce ss C) is an example (2 + 1) REMPI scheme; two photons are required to excite the molecule AX to a resonant excited intermediate state AX* and one more to ionize AX*. AX AX* AX+ A B C

PAGE 57

57 Figure 2-4. Schematic of REMPI apparatus.

PAGE 58

58 tunable UV laser sample cell PMT Digital oscilloscope data analysis computer filter Figure 2-5. Schematic of PF-LIF apparatus.

PAGE 59

59 ArF laser =193 nm N2gas SV2Mass flow controller SV1 PF cellTEA soln A BLegend : SV = solenoid valve Figure 2-6. Photofragmentation set-up. A) schematic of PF-NO2 fragment collection set-up B) actual picture of the PF cell

PAGE 60

60 + 2H2O N N NH2N N NN2 +NBrN N N+ N2 +Br-NO2 -Br -2H+ colorless[H+]+ neutral red dye max= 530 nmA B 0.00 0.25 0.50 0.75 150300450600750900wavelen g th ( nm ) Absorbance (a.u.) Figure 2-7. Nitrogen dioxide detection by colorime tric method. A) species responsible for color B) absorption spectrum of neutral red dye.

PAGE 61

61 PMT ArF laser Beam splitter Focusing lens cell 2-3 NDF PD oscilloscope Channel 1 triggering output Channel 2 data output NDF = neutral density filter Figure 2-8. Schematic of time-resolv ed absorbance experimental set-up.

PAGE 62

62 0.020.040.060.080.0100.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Adye= -0.00753(g NO2 -) + 0.65734 R = -0.98506 BLANKAbsorbance at max= 530 nmg NO2 Figure 2-9. NO2 calibration curve.

PAGE 63

63 Table 2-1. Colorimetric analysis of NO2 fragments after photofragment ation of some nitro-based explosives. Explosive sample Actual NO2 ablated(mol)a NO2 detected (mol)b % NO2 detected 4 nitrotoluene 1.04.22 0.140.008 13.5.4 2,4 nitrotoluene 1.24.05 0.190.011 15.3.8 trinitrotoluene 0.78.06 0.150.008 19.2.8 PETN 7.96.10 1.000.056 12.6.7 RDX 2.46.22 1.69.096 68.7.3 a based on the ablated mass of the explosive samp le obtained by weighing the sample before and after photofragmenta tion using microbalance b amount of NO2 detected using colorimetric method

PAGE 64

64 -30-20-100102030 0.00 0.02 0.04 0.06 0.08 0.10 Intensity (mV)time (ns)A B-6-4-20246810 0.4 0.5 0.6 0.7 Absorbance (a.u.)time (ns)Legend: IA= initial intensity IB= intensity with NDF IC= intensity with NO2A2(t) = -log (IC/IA) A1(t) = -log (IB/IA)IAIBIC Figure 2-10. Results obtained with fast time reso lved experiments. A) profile of ArF laser pulse before and after NO2 photofragmentation B) time resolved absorbance plots for NO2 photofragmentation.

PAGE 65

65 CHAPTER 3 NO2 DETECTION BY CHEMILUMINESCENCE Introduction Highly sensitive detection of nitrogen dioxide gas (NO2) is essential not only for air monitoring purposes, but also for developing ex plosive detection. Common explosives can be detected based on their characterist ic functional group, specifically the NO2 moieties that can be photofragmented from the parent molecule usi ng a UV laser. This chapter begins with a discussion of the theory of chemilumines cence, followed by a review of previous NO2 detection systems, and finally, an evaluation and assessm ent of a newly developed instrument based on luminol chemiluminescenece (CL) for the analysis of NO2 fragments from explosives. Basic Theory of Chemiluminescence Chemiluminescence (CL) is defined as the pr oduction of electromagne tic radiation (UV, Vis, or IR) as a result of a chemical reaction.62 A significant fraction of reaction intermediates or products are produced in excited el ectronic states and emit light on re turning to the ground state. The following reactions are three common sequen ces used to produce CL: A + B I + I*; I* I + hv P (3-1) A + B P + P*; P* P + hv (3-2) I* (or P*) + F I (or P) + F F + hv (3-3) where A, B are reactants, I is an intermediate, P is the product, and F is a fluorescence acceptor. Reactions 1 and 2 illustrate that the luminescing species can be an intermediate or a product in a reaction. Reaction 3 represents sensitized CL in which the excited species produced transfers its energy to an efficient fluorophor e acceptor which then fluoresces.63 The CL quantum efficiency (CL) is defined as: (CL) = (number of photons emitted/number of photons reacted) (3-4)

PAGE 66

66 It can be represented as the pr oduct of two efficiencies or (CL) = ex L (3-5) where ex is the efficiency of the production of excited species (the fraction of reacting molecules that produces an excited molecule) and L is the luminescence efficiency (or the fluorescence quantum efficiency) of the luminescing species.63 For CL to occur, three general conditions mu st be met. First, there must be sufficient energy to produce an excited state. Thus, the reaction must be exothermic such that G > (hc/ ex) > (2.86 x 104) / ex (3-6) where G is the free energy change (kcal mol-1) of the reaction and ex is the upper wavelength limit in nm for excitation of the luminescing species. For CL emission in the visible region, G must be 40 to 70 kcal mol-1.62 Second, there must be a favorable pathway to produce the excited state. Third, photon emission must be a favorable deactivation process for the excited product in relation to other competit ive non-radiative process.62 CL has the potential for being the basis of chemical analysis. Chemiluminescence detection is based on measuring th e emission of photons from the excited molecules. Often, it is possible to arrange reaction conditi ons such that the luminescence si gnal is related to the analyte concentration. The analytical performance of CL technique s is governed primarily by the chemistry of the reaction utilized and the luminescence charact eristics of the emitting species. Careful control of the reaction conditions is mandatory. By ch emical parameter optimization, the analytical advantages of CL including high sensitivity, low de tection limits, and wide linear ranges can be maximized and achieved with very simple equipment. These attributes also result from the fact that no light source is require d, significantly eliminating the b ackground noise from the source.

PAGE 67

67 CL is up to 100 000 times more sensitive than ab sorption spectroscopy and at often 1000 times more sensitive than fluorometry.64 One attractive feature of CL technique is the simplicity of the instru mentation. All that is required is a reaction cell, a detector, and a signal processor-readout system. The emitted light intensity is measured by means of a photom ultiplier (PMT). Because only one emitter is normally present in the reaction medium, no monoc hromator is required to discriminate among different wavelengths. There are two ways of measuring CL, namely : (1) static mode and (2) continuous flowing stream mode.63 In the static mode, (also known as batch or discrete sampling), the CL reagents and the analyte are mixed rapidly and the CL inte nsity versus time profile is measured (Figure 31). The initial increase in intens ity is due to the finite time for mixing of reagents or an induction period in the reaction. The decay of the signal is caused by the consumption of the reagents and changes in the CL quantum efficiency with time. The analytical signal can be taken as the peak signal or the signal after some fi xed delay time from the point of mixing, (the integral of the signal over some period) or the signal (area) of the entire peak. In the continuous flow mode the CL reagents and the an alyte are continually pumped and combined at tees and sent to a flow cell or mixed and observed in an integral reactor flow cell as shown in Figure 3-2. When the cell is tota lly filled with the reaction mixture, a maximum CL signal is obtained. Then, the system will equi librate reaching a steady state signal. The signal represents the output integrated over the residence time of the reaction mixture in the cell. Luminol-NO2 Chemiluminescence While there are many chemical reactions that are chemiluminescent, the oxidation of luminol (C8H7N3O2, which is also known as 5-amino-2,3-di hydro-1,4-phthalazined ione) is one of the most commonly known. Luminol was first s ynthesized in 1893, and its ability to yield a

PAGE 68

68 chemiluminescent reaction in ba sic solution in the presence of an oxidant such as hydrogen peroxide was first reported in 1928 by H.O. Albrecht.65 The luminol CL system consists of water, ba se (sodium or potassium hydroxide), oxidant, catalyst, and luminol. A variet y of oxidants including perma nganate, hypochlorite, hydrogen peroxide, and nitrogen dioxide have been used in the luminol reaction. T ypical catalysts include peroxidase, hemin, and transition metal ions (Co 2+, Cu 2+, Fe 3+ etc.).66 The CL reaction used for NO2 detection is based on the same principle. NO2 is detected by an increase in CL intensity that results when NO2 reacts with an alkaline solution of luminol. The CL reaction is as follows: 2C8H7N3O2 + 4OH+ 2NO2 2C8H5O4 + 3N2(g) + 4H2O + hv (3-7) The CL spectrum has maximum emission intensity at 425 nm with an approximately 100 nm bandwidth.67 Despite intensive studies, the mechanis m of luminol oxidation in water has not been fully elucidated. However, the general outli ne of the reaction has been agreed upon and is shown in Figure 3-3. Base oxidizes the luminol leaving negative charges on nitrogen atoms which move onto the carbonyl oxygen to form wh at is known as enolate anion. The oxidizing agent next performs a cyclic addition to the two carbonyl carbons which leads to formation of 3aminophthalate (3-APA*).68 White et al. have shown that th e CL spectrum of luminol matches the fluorescence spectrum of the 3-aminophthala te anion and, hence, identifying it as the emitting species.69 Since then, several other investigator s have confirmed the identity of the emitter. 70-72 Review of CL Methods for NO2 detection The detection of nitr ogen dioxide gas (NO2) by means of chemilumi nescent reaction with luminol in aqueous solution has been the subject of considerable interest and development .The first prototype detector employing this appr oach was developed in 1980s by Maeda et al.73 by

PAGE 69

69 directly reacting the NO2(g) at the surface of continuously flowing luminol solution. The intensity of luminescent radiation emitted by the reac tion was monitored by photomultiplier (PMT) detector. This system was very sensitive to movements of the compartment and had relatively slow time response. An alternate desi gn for chemiluminescent detection of NO2(g) was subsequently presented by Wendel et al.66;74, who allowed a stream of NO2(g) to pass over filter paper drenched with basic luminol solution. The CL reaction occurs on the surface of the filter paper which is adjacent to the PMT window. A fabric wick has also been used instead of filter paper to make a compact commercial NO2 detector (Luminox LMA-3, by Scintrex/Unisearch Associates Inc.); a detailed description of this instrument was given by Schiff et al.75 and Kelly et al.76 A further variation on the luminol chemiluminescent detector for NO2(g) was introduced by Mikuska and Vecera;77 this configuration employed a cont inuous spray of luminol solution, immediately below the PMT. CL was generated from the stream of analyzed gas. Another detection scheme modification wa s made by Collins and Pehrsson78 wherein NO2(g) reacted on the surface of a glass containing immobilized lumi nol molecules within a hydrogel or polymeric thin film positioned in front of a PMT tube Yet another variation was developed by Robinson et al.79 where a bundle of porous polypropylene hollow fiber membranes was used to bring the NO2(g) into contact with luminol solution. Chemilumi nescence occurring w ithin the translucent hollow fibers was detected using a miniature PMT tube. The present work was adapted from the ch emiluminescent aerosol detector of Mikuska and Vecera77 for its simplicity, efficiency and high sensitivity. The aim of this work is to evaluate, assess and further improve the analytical capability of detection of NO2(g) with the luminol reaction.

PAGE 70

70 Experimental Section Reagents and Chemicals Luminol powder of 98% purity (5-amino-2,3dihydro-1,4-phthalazinedione) was obtained from Acros Organics (New Jersey, USA) and used without further purification. Potassium hydroxide, methanol, ethanol, tert -butanol, hydrogen peroxide, s odium bicarbonate, and sodium carbonate were all obtained from Fischer Scie nce Company (Pittsburg, PA). Para-iodophenol was obtained from Sigma Aldrich Chemical Comp any (St. Louis, MO). Deionized water was obtained by passing distilled water through a Millipore Milli-Q deionizing system All gases purchased were of the highest pur ity. A tank of 99% nitrogen gas was purchased from BOC Gases Inc. (Murray Hill, NJ). Nitrog en dioxide gas buffered with nitrogen gas was purchased from Safety Products Inc. (Lakeland, FL). Apparatus and Methodology Luminol-CL detector Figure 3-4 is the instrumental set-up used for measuring and optimizing the chemiluminescence signal arising fr om the basic luminol solution and nitrogen dioxide gas. The three main sec tions are a gas delivery system reaction chamber and measuring electronics. The gas delivery system consisted of the gas samples and mass flow controllers (Alicat Instrument Inc.,Tuczon.AZ) used fo r dilution. The chemiluminescence reaction set-up was adapted from the NO2 atmospheric monitoring system developed by Mikuska and Vecera.77 A stream of NO2 gas passed through a concentric nebuliz er (Type C-Q nebulizer manufactured by Precision Glass Blowing, Cent ennial, Colorado) allowing the basic luminol solution to be aspirated into the reaction chamber and dispersed to form a fine mist with a large interface area for NO2 and luminol reaction. The aero solized luminol solution is directed toward the bottom wall of the reaction cell with PMT orthogonal to the directio n of the mists propagation. The reaction was done in a batch mode which last for twenty seconds. The luminescent radiation

PAGE 71

71 resulting from the reaction between the luminol alkaline solution aerosol and NO2 gas is detected by a cooled R3896 photomultiplier tube (PMT).Th e PMT has a spectral response range from 185 to 900 nm with maximum sensitivity at 450 nm. The PMT is operated at -850 V and a temperature of 10 to 15oC. The photocurrent signal from the PMT is amplified by Kiethley Model 6514 electrometer which is connected to a computer by a GPIB-USB-HS digital-analog converter from the National Instrument Laborator y. EXCELINK software is used to control all the data acquisition. All data are then pro cessed using Origin 6.1 and Excel software. The reaction cell is made of a Teflon tube w ith an inner diameter of 2.10 cm and a length of 4.57 cm. It is encased in an aluminum housing to completely seal the reaction from external light sources. One end of the reaction cell is open to the PMT tube, separated only by a fused silica window with the diameter of a 1.50 cm. The concentric nebulizer is connected to the side of the reaction ce ll and oriented parallel with respect to the PMT detector. CL emission spectrum set-up. A Fluorolog-Tau-3 spectrof luorometer (Jovin Yuon Inc., Edison, NJ) in chemiluminescence mode (excita tion source turned off) is used to obtain the emission spectra. A block diagram of the instrument is depicted in Figure 3-5. In the first set-up, a 25 ppm NO2 standard gas was purged into the quart z cell containing 0.01 M luminol in 1 M KOH. In the second set-up, a peristaltic pump (Rai nin, Model Rabbit) was set at desired flow rate to deliver H2O2 solution into the quartz cel l containing luminol solution. Gas-liquid exchange module set-up. A diagram of the gas liquid exchange module is shown in Figure 3-6A. This de vice was provided by Dr. John W. Birks from University of Colorado. The device is made up of microporous membrane allowing the NO2 gas to react with alkaline luminol-H2O2 solution. The membrane consists of a bundle of approximately fifty

PAGE 72

72 polypropylene capillary fibers having an outer diameter of 380 m, a wall thickness of 50 m, and a pore size of 0.2 m (Akzo Nobel, Scottsbor o, AL). The fibers are encased in a polymer housing and the entire unit will be referred as th e gas-liquid exchange modul e. The actual picture of the set-up is shown in Figure 3-6B. The luminol solution flows through the interior of the fibers, while the NO2 gas flows around the exterior of the fibers. The aqueous soluti on remains in the interior of the fibers due to the hydrophobic nature of the pol ypropylene. Gases diffuse thr ough the pores of the membrane and into solution, resulting in the chemiluminescent reaction. NO2 was detected above the background signal due to the reaction of H2O2 and luminol. The fibers and the polymer are translucent, allowing the transmitted light to be detected by a PMT. Luminol/H2O2-CL detector Figure 3-7 is a schematic diagram for a luminol/H2O2-CL detector. The gas-liquid exchange module is used as the CL reac tion cell, and a miniature headon R647 photomultiplier tube (PMT) tube is used to detect the light emitted. The PMT has a spectral response range from 300 to 650 nm with a maximum sensitivity at 420 nm. The PMT is operated at -850 V at ambient te mperature. The photocurrent si gnal from the PMT is amplified by a Kiethley Model 6514 electrometer which is connected to a co mputer by GPIB-USB-HS digital-analog converter from National Instrument s. EXCELINK software is used to control all the data acquisition. All data are then pro cessed using Origin 6.1 and Excel software. A series of peristaltic pumps is used to pump the luminol and H2O2 solutions into the gasliquid exchange module. The two so lutions are mixed in a tee just prior to the exchange module. The flow rate was set at 1.0 L/min for each solu tion. The gas outlet is connected to the exhaust fume hood.

PAGE 73

73 Dilution set-up. NO2 standards in ppb range were prepared by dynamic dilution of 5 to 25 ppm standard NO2/N2 mixtures. Pure N2 gas was used as the diluent. A block diagram of the setup is shown in Figure 3-8. Gas flow rates were c ontrolled by mass flow controllers. An in-line gas mixer filled with 4 mm glass beads housed in a polystyrene casing was used to ensure efficient gas mixing. Mass flow controllers were calibrated el ectronically using a Humonics Optiflow 650 Digital flow meter and ma nually using a GC bubble flow meter. Results and Discussion In order to have a highly sensitive method for NO2 determination based on its chemiluminescence reaction with luminol, both the physical and ch emical parameters of the system were optimized. Physical parameters are related to overall design of the reaction cell/detector configuration, whereas chemical parameters are aff ected by the preparation of the chemical components of the basic luminol solution. Chemical Parameters The optimal solution composition for NO2 detection was determined by systematically varying the concentration of potassium hydroxide (KOH) and chemiluminescent enhancer reagents such as p-iodophenol, methanol, ethanol and t-butanol. The effect of the concentration of these reagents on chemiluminescence intensit y was studied by varying the concentration of one component over a range while hold ing the other components constant. The effect of KOH concentration on the emi ssion intensity is show n in Figure 3-9A. As the concentration of KOH increased, the decay rate of the reaction becomes faster, which consequently decreased the integrated chemilumi nescence signal. By obtai ning the integral or area of the entire peak for each corresponding signal from Figure 3-9A and plotting it against the pOH, the optimum KOH concentration of 0.2 M can be inferred which corresponds to

PAGE 74

74 pH=13.1 as seen from Figure 3-9B. The resu lt is consistent w ith previous studies 75; 77 verifying that excess OHquenches the chemiluminescence radiation. Maeda et al.73 obtained a maximum CL signal for luminol concentrations of 3 x 10-4 M to 3 x 10-3 M with no variation within that range. Mikuska and Vecera77 as well as Wendel 74 agreed with this observation. Kok et al.80 obtained peak chemilumine scence intensity at 2.4 x 10-4 M luminol and a 50% decrease for an order of magnit ude change about this va lue. The result of our experiment is presented in Figure 3-10 where a peak occurs for 5 x 10-3 M luminol which is in agreement with previous results 73;77. Based on the experimental results acquired by Wendel et al., the addition of water soluble alcohols such as methanol, ethanol or t-butanol to the alkaline luminol so lution resulted in a two-fold increase in the chem iluminescence signal with NO2(g). Based on our results, addition of alcohols had negligible effect on the r eaction, as can be seen in Figure 3-11. Thorpe and et al. 81 used phenol derivatives such as p-iodophenol and p-phenyl phenol to enhance light emission for the r eaction between hydrogen peroxide (H2O2) and luminol. They were able to observe that the light emitted in the reaction decayed slowly and its intensity was about 1000 fold greater than the unenhanced re action. Addition of p-iodophenol to the basic luminol solution was investigated and it was found out that the presence of 0.01 M p-iodophenol significantly enhanced the integrated chemilumi nescence signal up to 10 times, as shown in Figure 3-12. The mechanism of the p-iodophenol enhancement of chemiluminescence signal is still unresolved. Here, it was shown that the emissi on spectra of phenol-enhanced and unenhanced reactions are remarkably similar (Figure 3-13), suggesting two plausible hypotheses that (1) piodophenol as a chemiluminescent enhancer reagen t is neither a more efficient emitter nor

PAGE 75

75 fluorescent acceptor, but exerts its action earlier in the complex reaction between luminol and the oxidant and (2) p-iodophenol enhanced the ch emiluminescence signal of aerosolized luminolNO2(g) system by significantly decreasing the dielect ric constant of the aerosolized luminol solution, which enhances the desolvation of lumi nol enolate anions on th e surface of the aerosol and thus, increasing the frequency of interaction of the reagents involved. This is supported by the fact that the signal en hancement using p-iodophenol can only be observed with the chemiluminescence aerosol detector and not by simply bubbling the NO2(g) into the bulk of the basic luminol solution as seen on Fi gure 3-13B and 3-13C,respectively. Physical Parameters Gas and liquid flow rates. The gas flow rate was varied with a mass flow controller and it was noted that above 0.7 SLPM (~24 psi), the chemiluminescence signal saturates (Figure 3-14). The luminol uptake at this gas flow is 1.10 mL/min. Since the re sponse was independent of gas flow rate, the instrument is a concentra tion detector rather than a mass detector. A flow rate of 0.7 SLPM was used for further studies. Cell design. The change in chemiluminescence intens ity as a function of the orientation of the concentric nebulizer with respect to the detection window wa s investigated. It was observed that if the NO2-luminol mist entered parallel to the PMT window, a signal enhancement of about 8 fold occurred comp ared to introduction of the NO2-luminol mist perpendicular to the detection window (Figure 3-15). Two different sets of reaction cell were used for this experiment. The set-up described under the methodology section was used for the experiment where the orientation NO2-luminol mist was parallel with respect to the PMT window wh ich will be referred as Set-up A. For the set-up where the NO2-luminol mist was oriented orthogonal to the PMT window, a completely different reaction cell of the same design and configuration as Set-up A was used except that the distance

PAGE 76

76 between the nebulizer and the de tection window was approximately four times farther than Setup A and a stream of N2 gas was allowed to blow con tinuously on the PMT window as the reaction was in progress. In this way, it could be argued that the signi ficant difference in the signal was due to impaction of aerosolized solution on the surface of the PMT window. Shelf life of the alkaline luminol solution. It was observed from previous studies 74; 80 that luminol-NO2 chemiluminescence intensity increased with the age of the lumi nol solution. Hence, a controlled study of the chemilumi nescence intensity as a function of time from the preparation of the solution was made. An optimized luminol solution was prepared (5 x 10-3 M luminol, 0.01 M p-iodophenol in 0.2 M KOH) and the chemilumi nescence reaction from a 5 ppm standard source of NO2 was measured. This measurement was done as soon as the solution was made and repeated periodically for at least 2 months. According to Wendel et al.74, the chemiluminescence intensity increased by as much as 20-fold over the course of days and after this initial aging period, it was found out that the signal did not change apprecia bly over longer storage period. However, our results (Figure 3-16) did not observe this trend. Instead, it was observed that the basic luminol solution gave a slight change in chemiluminescence readings from day 0 up to day 71. Performance of Luminol-CL Detector All the optimized parameters obtained were then applied to construct the calibration curve for NO2 (Figure 3-17). A calibration working curv e was generated by making three replicate measurements of the NO2 standards. The linear range of the detector was measured using known concentrations of NO2 from 33 ppbv to 25 ppmv. The calculate d detection limit with a signal to noise ratio of 3 was 19 ppt NO2. A calibration sensitivity of 7.08 x 10 -8 C/ppb NO2 was obtained. The non-zero intercept is a reflect ion of the background of the instrument.

PAGE 77

77 Performance of Luminol/H2O2 CL Detector The luminol/ H2O2 CL detector was tested for direct detection of NO2. NO2 reacted with H2O2 to form a potent oxidizer peroxynitrite (ONOO-) as shown by the following reaction: 2NO2 + H2O2 + OH2ONO2 + H2O (3-8) Peroxynitrite is capable of oxidi zing luminol to an excited st ate of 3-aminophthalate, which relaxes by emitting light as can be seen in reaction 3-9.82 2ONO2 + luminol 3-aminophthalate + N2 + hv (3-9) The NO2 signal resulting from peroxynitrite forma tion is detected on top of the background signal resulting from the sl ow oxidation of luminol by H2O2. Optimal conditions obtained previously were used to gene rate the calibration curve for NO2. The concentrations of H2O2 and luminol in a bicarbonate/carbonate buffer (with pH=10.25) were 2 mM and 4 mM, respectively. Responses of the detector to NO2 standards in the range of 33 ppb to 25 ppm are depicted in Figure 3-18. The working curve was generated by making three replicate measurements of the NO2 standards. The calculated detection limit and the calibration sensitivity were 178 pptr NO2 and 2.43 x 10-8 C/ppb, respectively. Luminol CL Detector Compared to Other CL Set-ups The LOD ratio of luminol CL detector to luminol/H2O2 CL detector is 9 which can be attributed to the following reasons: (1) in th e first set-up surface interaction between the NO2 gas and luminol solution was maximized through ne bulization (2) the first set-up also had a significantly lower background sign al due to the absence of H2O2 and the peristaltic pumps (3) the side-on PMT used for the first set-up was more sensitive compared to the head-on PMT used in the second set-up.

PAGE 78

78 In general, the luminol-CL detector gave a be tter LOD compared to other existing CL setups as shown on Table 3-1. Conclusion The luminol-CL detector for quantifying NO2 is based on the NO2 CL reaction with aerosolized basic luminol solution. The aerosol is formed by continuous nebulization of reagent solutions by the stream of analy zed gas in a sealed reaction chamber. The intensity of the CL signal was detected by a PMT. The current system improved the detection limit obtained previously by others optimizati on of all the feasible physical a nd chemical parameters involved in the chemiluminescence reaction between luminol and NO2. A detection limit of 19 ppt NO2 at (S/N) = 3 was obtained. The optimal reagent solu tion from the viewpoint of sensitivity of the response to NO2 (maximum signal/signal noise ratio) was 5 x 10 -3 M luminol + 0.01 M piodophenol + 0.2 M KOH. A Luminol/H2O2 CL set-up was also explored where a bundle of porous polypropylene fibers was used to bring the NO2 into contact with luminol solution. A LOD of 178 ppt NO2 at (S/N) = 3 was obtained. The newly developed CL instrument benefits from superior sens itivity, simplicity, and reliability, which are very attractive in chemical sensor. It is anticipated that the CL detector developed can be used as a diagnostic tool and aid in developing a bett er explosive detector.

PAGE 79

79 PMT I CL reaction time reaction starts A B Figure 3-1. Schematic diagram for static CL devi ce. A) the sample and the CL reagent(s) are introduced in the reaction cell and the final reagent is injected to initiate the CL emission, then light is monitored by the PMT detector B) curve showing CL intensity as a function of time after the reagents are mixed to ini tiate the reaction.

PAGE 80

80 A B Figure 3-2. Schematic diagram for continuous CL device: A) the sample and the CL reagent(s) are introduced simultaneously and continuously into the flow cell B) a steady state signal is achieved when the cell is tota lly filled with the reaction mixture.

PAGE 81

81 O2 .-Enolate anion Figure 3-3. Luminol oxidation mechanism.

PAGE 82

82 reaction chamber PMT power supply electrometer A/D converter computer PMT cooling system NO2 (g)buffer with N2 (g) fused silica window aluminum housing luminol solution Figure 3-4. Luminol-C L detector set-up.

PAGE 83

83 Excitation source (turn-off) monochromator cell reaction monochromator PMT Parameters Scanning wavelength : 200-650 nm Slit width : 10 nm Excitation source : turn-off HV (PMT) : 850 V Integration time : 0.3 s data processor Figure 3-5. Schematic of spectrofluo rometer used for acquiring luminol-NO2 CL emission spectra.

PAGE 84

84 gas inlet gas outlet solution in solution waste cross sectional view of gas exchanger hollow fibers O.D. : 380 m Wall thickness: 50 m Pore size : 0.2 m gas inletgas outlet solution in solution waste Bundle of porous polypropylene capillary fibersA B Figure 3-6. Chemiluminescence Set-up II.A) sche matic of gas-liquid exchange module set-up B) the actual picture of the set-up.

PAGE 85

85 solution waste PMT H2O2luminol Peristaltic pump gas inlet gas outlet Peristaltic pump power supply electrometer A/D computer gas-liquid exchange module Figure 3-7. Luminol/H2O2-CL detector.

PAGE 86

86 NO2 gas N2 gas MFC I MFC II gas mixer reaction cell Figure 3-8. Dilution set-up.

PAGE 87

87 050100150200250300 0 2 4 6 8 ---0.05 M KOH ---0.2 M KOH ---1.0 M KOH ---2.5 M KOHA-0.6-0.30.00.30.60.91.21.5 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 peak area (a.u.)-(log [OH-])pH = 13.1 0.2 M KOHB Figure 3-9. Effect of KOH concentrat ion on A) decay rate of luminol-NO2 reaction and B) on CL signal.

PAGE 88

88 1.01.52.02.53.0 0.0 0.1 0.2 0.3 0.4 peak area (a.u.)-log([Luminol])5.0 x 10-3M luminol Figure 3-10. Effect of luminol concentration on CL signal.

PAGE 89

89 w/o alcoholMe(OH)Et(OH)Bu(OH) 0.0 0.1 0.2 0.3 0.4 0.5 peak area (a.u.)type of alcohol Figure 3-11. Effect of adding vari ous type of alcohols on CL signal.

PAGE 90

90 0.000.020.040.060.080.10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 peak area (a.u.)p-iodophenol concentration (M) Figure 3-12. Effect of p-iodophe nol concentration on CL signal.

PAGE 91

91 200300400500600700 0 4000 8000 12000 16000 20000 24000 28000 32000 CL intensity(cps)emission wavelength(nm)a b c Figure 3-13. Luminol-NO2 Emission Spectra. A) chemilumine scence spectrum obtained from reaction of luminol with H2O2 B) chemiluminescence spectrum obtained from reaction of luminol with NO2 C) chemiluminescence spectrum obtained from enhanced reaction of luminol with NO2 using p-iodophenol.

PAGE 92

92 0.40.81.21.62.0 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 peak area (a.u.)NO2 flowrate (SLPM) Figure 3-14. Effect of gas sa mple flow rate on CL signal.

PAGE 93

93 0255075100125150 -2.0 -1.5 -1.0 -0.5 0.0 time (s)7.6E-5C9.0E-6 5.8E-4C8.0E-6 PMT PMTCL signal (A) Figure 3-15. Effect of th e orientation of luminol-NO2 mist with respect to the PMT window on signal response.

PAGE 94

94 01020304050607080 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 peak area (mC)duration of storage (days) Figure 3-16. Effect of the storage tim e of luminol solution on CL signal.

PAGE 95

95 1.52.02.53.03.54.04.5 -4.0 -3.6 -3.2 -2.8 log (peak area) log [NO 2 ] (pp b ) Figure 3-17. Calibration curve of NO2 using Luminol-CL set-up.

PAGE 96

96 1.52.02.53.03.54.04.5 -5.5 -5.0 -4.5 -4.0 -3.5 -3.0 log (peak area)log [NO2]ppb Figure 3-18. Calibration curve of NO2 using Luminol/H2O2-CL set-up.

PAGE 97

97 Table 3-1 Summary of the LOD obt ained from different luminol-NO2 CL set-ups. Researchers LMN-NO2 reaction set-up Detection limit Maeda et al. (1980)73 Passing NO2 gas on the surface of continuous flowing luminol solution 50 pptr Wendell et al. (1983)74 Blowing NO2 gas on the surface of filter paper on which luminol solution was allowed to flow 30 pptr Schiff et al. (1986)75 Blowing NO2 gas on the surface of fabric wick on which luminol solution was allowed to flow 5 pptr Mikuska et al. (1992)77 Reacting NO2 gas with aerosol of luminol solution 0.24 ppm (lowest concentration detected) Collins et al. (1995)78 Monterola et al. (2007) Blowing NO2 gas on the surface of polymeric thin film of luminol Reacting NO2 gas with aerosol of luminol solution 0.46 ppbv 19 pptr

PAGE 98

98 CHAPTER 4 CONVERSION OF NO TO NO2 Introduction Photofragmentation of nitro-based explosiv es using a UV laser produced both nitrogen dioxide (NO2) and nitrogen oxide (NO) fragments. De tection of these fragments indicates the presence of explosives. However, NO cannot be detected by luminol chemiluminescence and hence must be converted back to NO2 to enhance the sensitivity for explosive detection. Several methods for NO to NO2 conversion has been reported for atmospheric monitoring purposes. This chapter begins with a review of those methods and is followed by a description of the oxidation unit used to integrate the photof ragmentation and chemiluminescence set-ups into a single device for explosive detection. Review of Methods for NO oxidation to NO2 Several methods have been used for oxidizing NO to NO2. Among these are potassium permanganate and sulfuric acid (KMnO4/ H2SO4), iodine pentoxide (I2O5), acidified MnO2, persulfate, ozone (O3), heated chlorine dioxide (ClO2), oxygen (O2), and chromium oxide (CrO3).82 Each method has certain disadva ntages especially during real time analysis. Persulfate and KMnO4/ H2SO4 are not quantitative oxidi zers while acidified MnO2 is not stable.82 An excess of O3 will convert NO all the way to nitric acid (HNO3) in the presence of moisture.83 Also, O3 cannot be used as an oxidant because it pr oduces an interfering signal with luminol solution. A disadvantage in the use of I2O5 is that it is sensitive to many other reducing agents. ClO2 has been demonstrated to yield good conversions but must be generated in situ which is quite inconvenient. Gaseous O2 oxidizes NO slowly and can only be used for concentrations above 100 ppm. 83

PAGE 99

99 CrO3 coated on an inert substance has been s hown to give the best results for NO to NO2 conversion. Oxidation efficiency close to 100 % wa s achieved in the range of 30 to 80 % relative humidity.82-84 The work presented here described the fabric ation of an oxidation unit for conversion of NO to NO2 conversion that was used to integrate the photofragmentation and chemiluminescence units into an explosive detector device. Experimental Section Reagents and Chemicals Reagent grade CrO3 (~98%) was purchased for Sigma Aldrich (St. Louis, MO). Glass beads (D = 4 mm) were purchased from Fischer Scientific (Hampton, NJ). A tank of 10 ppm NO and 10 ppm NO2 both buffered with N2 were purchased from Airgas South Inc. (Gainesville, FL) and Safety Products Inc. (Lakeland, FL), respectively. Apparatus and Methodology CrO3 Oxidation unit. The CrO3 oxidation unit was prepared by coating CrO3 on glass beads. Glass beads were chosen since it is an inert substance and beads were used to provide high surface area. The first step for preparing the CrO3 converter was soaking the glass beads in a 20 %(w/v) aqueous solution of CrO3 for about 30 minutes. After soaking, the glass beads were filtered and dried in an oven at 105oC for 30 to 45 minutes. The glass beads were then packed in a 1 diameter by 8 length stainless steel tubing. CrO3 Oxidation Unit/Luminol-CL Detector Tandem. Figure 4-1 is the experimental set-up used for evaluating the efficiency of the CrO3 oxidation unit. A 10 ppm NO in N2 was used as reference sample. The gas sample flow rate was controlled by th e mass flow controller which was then connected to the CrO3 oxidation unit. Oxidized products were then introduced

PAGE 100

100 into the CL unit through the side arm of the con centric nebulizer. The operating conditions of the instrument were essentially the same as those described in Chapter 3. Results and Discussion The CrO3 oxidation unit was tested for effi ciency in converting NO to NO2 by passing a 10 ppm NO/N2 sample through the converter. The sample flow rate was set to 100 mL/minute. The NO2 produced upon oxidation was detected by the luminol-CL detector. Figure 4-2A shows the CL signal acquired with 10 ppm NO in N2 without CrO3 oxidation unit whereas Figure 4-2B shows the signal obtained after passing the same gas sample through the CrO3 converter. Figure 4-2C shows the signal obse rved after passing 10 ppm NO2 in N2 through the same converter. Based on the three replicate measurements, average peak areas for 10 ppm NO and 10 ppm NO2 after passing through the same converter were 857.1.5 and 1488.5.3, respectively. Conversion efficiency of 58% was obtained. The lower conversion effi ciency can be attributed to relative humidity of the NO sample. It is well esta blished that the efficiency of NO oxidation by CrO3 varies with sample water content, being grea test at mid-range relative humidity (30 to 80 %) but much poorer near both extremes. 82-84 Conclusion CrO3 oxidation unit has approximately 58% NO to NO2 conversion efficiency which was anticipated to significantly enhance the sensi tivity of the Photofragmentation-Chemiluminescence apparatus for explosive detection.

PAGE 101

101 power supply Signal processor and read-out PMTLuminol solution Optical window NO2 MFC oxidation unit Legend : MFC = mass flow controller Figure 4-1. Experimental set-up used for evaluating the efficiency of CrO3 oxidation unit.

PAGE 102

102 -20020406080100120140160180 -70 -60 -50 -40 -30 -20 -10 0 10 CL signal (nA)time (s)-20020406080100120140 -170 -160 -150 -140 -130 -120 -110 CL signal (nA)time (s)-20020406080100120140160180 -35 -30 -25 -20 -15 -10 -5 0 5 CL signal (nA)time (s)A B C Figure 4-2. CL spectra of A) 10 ppm NO/N2 B) 10 ppm NO/N2 after passing through CrO3 oxidation unit C) 10 ppm NO2/N2 after passing through CrO3 oxidation unit.

PAGE 103

103 CHAPTER 5 LASER PHOTOFRAGMENTATION AND CHEM ILUMINESCENCE FOR NITRO-BASED EXPLOSIVE DETECTION Introduction Common explosives contain NO2 functional group and sin ce there are relatively few naturally occurring sources of n itro compounds, its presence can be used as a signature for explosive detection. Explosive de tection using Photofragmentatio n-Fragment Detection (PF-FD) has been explored. This chapter begins with a review of those methods and is followed by a description of a new instrument based on laser photofragmentation and luminol chemiluminescence. Review of PF-FD Methods for Explosive Detection The PF-FD approach most often utilized when the analyte molecule does not lend itself to direct spectroscopic detection. In general, atoms and small molecules (composed of 2 to 3 atoms) can be detected directly by absorptio n, fluorescence, or ionization techniques.44 This is due to a favorable combination of usually strong optical transitions and sharp, we ll resolved spectral features that provide effective optical activity. However, it is often the case for larger molecules such as the explosive compounds th at the transitions are weaker and the spectral features are broad and poorly defined. In these cases, direct detection of the molecule by any of the above techniques is not analytically practical. While direct detection may not be feasible, the PF products of explosive such as NOx (x =1, 2) can be readily detected. Since NOx (x =1, 2) fragments are characteristic of the chemical composition of th e explosive compounds, they also contribute to the selectivity of the method. A patented instrument based on PF-FD tec hnique for explosive compounds was reported by Nguyen et al.85;86 The instrument is based on coupling pyrolyis with luminol chemiluminescence detector. Explosive samples were collected and pyrolyzed under

PAGE 104

104 electronically heated Pt-Rh wire at 700oC. The NO2 fragments that were generated after pyrolysis were swept into a r eaction cell containing basic luminol solution. The luminol solution was separated from the remainder of the re action cell by a semi-per meable hydrophobic PTFE membrane. The membrane was permeable to NO2, thereby allowing diffusion to bring the NO2 into contact with luminol solution. Light emitted from the reaction was detected by a photomultiplier tube (PMT) coupled to the reaction cell. The method claimed a detection limit of 100 ppt for DMBA. There are several disadvantage s regarding the pyro lysis method. First of all, pyrolysis produced many decomposition pathways, one of which is the detonation reaction of the explosive samples. TNT, for instance, is known to detonate at 502oC and upon detonation, TNT decomposes as follows: 3 2C7H5N3O6 + 3N2 + 5H2O + 7CO + 7C (5-1) Reaction 5-1 significantly diminished NO2 fragments detectable. Secondly, pyrolysis is a time consuming method which is definitely a dr awback for a fast real time analysis. Due to limitations of the pyrolysis met hod, research on PF-FD techniques was pursued using laser photofragmentation. Laser photofragmentation followed by fragment detection using Resonance Enhanced Multi-photon Ionization Time-of-Flight Mass Spectrometer (REMPITOFMS) or Laser Induced Fluorescence (LIF) was not only performed for elucidating the photodissociation pathways of explosive compounds but for real time explosive monitoring as well. Lemire et al.53 and Simeonsson et al.54 used PF-REMPI-TOFMS with a laser operating at 226 nm to photofragment the parent explos ive compound and photoioni ze the resulting NO fragments. The NO+ ions were subsequently detected using a TOFMS. The technique was

PAGE 105

105 demonstrated using nitrocompounds such as nitromethane, DMNA, RDX, NB, o -NT, m -NT and TNT. Limit of detection for each an alyte is tabulated in Table 5-1. Wu et al.87, Boudreax et al.88, Parpar et al.89, and Swayambunathan et al.90 used PF-LIF in detecting explosives in various media including soil. Analytical figures of merit for these works are also tabulated in Table 5-1. In 1999 and onwards, Sausa et al.91-94 replaced TOFMS with an ion probe made up of miniature square electrodes to collect the resulting electrons and ions from PF of explosive compounds. The modification makes the device s impler and portable without trading its sensitivity for NO(x = 1, 2). Pertinent results regarding this technique are noted at Table 5-1. Finally, a new prototype inst rument was developed in this study by coupling the accuracy and convenience of laser photofragmentation over pyrolysis and the sensitivity for NO(x =1,2) detection of luminol-chemiluminescence (CL) detector over TOFMS and LIF. Experimental Section Reagents and Chemicals All of the reagents and chemicals were essentia lly the same as those described in Chapters II, III, and IV. Basic luminol solution was pr epared by dissolving luminol and p-iodophenol in 0.2 M KOH to make 5 x 10 -3 M luminol and 0.01 M p-i odophenol, respectively. Paraiodophenol served as the chemiluminescent enhan cer reagent and was adde d after dissolving the luminol into the KOH solution. Standard solutions of aromatic compounds su ch as NT, TNT, and 2,4 NT were purchased from Accustandards (New Haven, CT).

PAGE 106

106 Apparatus and Methodology Photofragmenation-Chemilum inescence Detector (PF-CL). Figure 5-1 is a schematic of the experimental set-up for explosive detec tion using PF followed by CL. It consists of three main sections: (1) PF unit (2) Oxidation unit (3) CL unit. The PF unit is essentially the same as descri bed in Chapter II except for an additional heating system allowing the generation of the expl osive vapor. After PF, the products were swept into the oxidation unit where NO frag ments were converted back to NO2. The NO2 fragments were then subsequently detected by the CL un it. The complete description of the CL and oxidation units was discussed in Chapter III and IV, respectively. Varying explosive vapor concentration can be generated as a function of temperature based on Clausius-Clapeyron Equation. The following equati ons were used to ca lculate the explosive vapor concentration: 91; 95 log [PETN] ppt = -7243/T(K) + 25.56 (5-2) log [RDX] ppt = -6473/T(K) + 22.50 (5-3) log [TNT] ppt = -5481/T(K) + 19.37 (5-4) The temperature of the vapor was assumed to be the same as the temperature of the PF cell. The temperature was monitored using a thermocouple thermometer. The sample preparation for solid explosive wa s the same as described in Chapter II. The total mass ablated for each experimental run was determined using Sartorious micro balance. PF efficiency of Nitro-aromatic compounds. Uniform thin films of NT, TNT, and 2-4 NT were prepared by depositing 40 L of soluti on of the target compound dissolved in acetone in a small aluminum cylindrical cup contai ner (O.D. = 1.9 mm; Depth = 4.5 mm).The solvent was then allowed to evaporate at room temp erature overnight. For each measurement, five different containers containing uniform thin film of selected explosive were used.

PAGE 107

107 PF of NO2 with luminol-CL detector. Figure 5-2 shows the PF set-up for verifying the further photodissociation of NO2 to NO and O* upon absorpti on of a photon in 193 nm wavelength. The set-up consists of the analyte-transport system and luminol-CL detector. The experiment was performed by generating the NO2 gas through reacting Cu and concentrated HNO3 in a reaction vessel A The NO2 generated from the reacti on was swept into the quartz reaction cell B by opening solenoid valves 1 and 2 simulta neously. This step allowed the reaction cell B to be filled up by NO2 gas. After quite sometime, solenoi d valves 1 and 2 were closed and the laser was fired into reaction cell B A total of 1000 pulses were used for each experimental run with pulse energy of 5 mJ/pulse and a repe tition rate of 50 Hz. After PF, solenoid valves number 2 and 3 were opened to sweep all the PF products into the CL unit. Then, the luminol-CL detector measured the remaining NO2 left in the cell. Then, a re ference signal was acquired by repeating the whole process without turning the laser on. Results and Discussion Relative PF Efficiency of Nitro-aromatic Compounds PF of solid explosive samples can be describe d in the following stepwise manner. The first step involved the absorption of sufficient laser radiation by th e solid explosive sample. Upon absorption, a plume of explosive will form and e xpand away from the sample surface as a result of pressure difference between the rapidly vaporized sample and the surrounding atmosphere. The subsequent photon absorption of the explos ive vapor will produce PF products, such as NO and NO2. Photolysis of solid structur ally similar compounds such as NT, 2,4 NT, and TNT were performed. The parent molecules of thes e compounds contained one, two, and three NO2 moieties, respectively. Figure 5-3 shows the luminol-CL signal obtained from the NOx(x = 1,2) fragments of these compounds at varying laser energy. The slopes acquired from these sets of

PAGE 108

108 data (Table 5-2) suggest nonlinearity in the PF efficiency of these compounds where PF efficiency is defined as follows: % PF efficiency = (moles of NO2 detected by CL)/(moles of NO2 available) x 100% (5-5) The non-linear behavior in the PF efficiency can be attributed to the difference in the latent heat of sublimation ( Hs) for each compound. As Hs increases, the more energy is required for a substance to vaporize. In the PF process for a solid explosive sample, this will consequently result in less explosive vapor for photolysis. Actual numerical values of Hs for these nitrocompounds were not available, however their vapor pressure at ambient condition have been well studied and are gi ven in Table 5-2. The vapor pre ssure is linearly related to Hs. From Table 5-2, NT has the lowest molecu lar mass implying that it has the weakest intermolecular force of attraction resulting in the highest vapor pr essure at ambient condition. As a consequence, the laser irradiat ed spot of NT produced a plume with the highest concentration of explosive vapor favoring the most efficien t photolysis. The higher slope value of TNT compared to 2,4 NT can be explained by a si ngle photon absorption capab le of simultaneous multi-NO2 scission. This process defies the higher va por pressure of 2,4 NT compared to TNT. NO2 Photolysis with Luminol-CL Detector Results obtained from the time resolved abso rption experiment in Chapter 2 on photolysis of NO2 into NO and O* using ArF laser was furthe r confirmed by using the luminol-CL detector. Figure 5-4 shows the CL signal of the NO2 gas before and after photodissociation. Significantly lower CL signal was obse rved after the PF of NO2 gas sample. Analytical Capability of PF-CL Detector The interaction of laser phot ons with both solid and vapor phase PETN was investigated by the quantity of NOx (x =1,2) fragments generated after photol ysis. Figure 5-5 presented the effect of varying laser pulse ener gy on CL signal obtai ned from the NOx (x =1,2) generated. In the

PAGE 109

109 absence of CrO3 converter, it can be assumed that the CL signal is due to NO2 fragments alone. The amount of NO2 released for both solid and vapor phase of PETN increased linearly within lower fluence range (e.g., a range of 2.5 to 5 mJ/pulse for solid PETN and 2.5 to 7.5 mJ/pulse for PETN vapor). However, at higher laser fluence, the quantity of NO2 fragments produced does not significantly changed. The pattern signifies that higher laser fluenc es yield faster photon deposition on the sample which favorably induced the sequential two-step PF process (Reactions 5-4 and 5-5) to take place. C(CH2)4O4(NO2)4 + h C(CH2)4O4(NO2)3 + NO2; h = 193 nm (5-4) NO2 + h NO + O*; h = 193 nm (5-5) The presence of CrO3 oxidizer for NO to NO2 conversion yielded an increasing plot of NO2 fragments within the entire range of laser fluence for both solid and vapor phase of PETN indicating that highe r photon intensity per pulse c onsequently produced more NOx (x =1,2) fragments. Figure 5-6 shows the effect of varying total number of lase r pulses at constant pulse energy for both solid and vapor phase of P ETN. The trend shows th at the detected NOx (x =1,2) fragments increase with increasi ng pulse number as long as there is an available sample to photolyze. The analytical capability of the method for tr ace explosive detection was obtained. In trace explosive detection, the e xplosive residue can be present in two forms: vapor and particulate. Vapor detection examines the vapor emanating fr om concealed explosive sample and particulate detection refers to the microscopic residues of ex plosives that would be present on individuals or material which have been through contamination.

PAGE 110

110 Figure 5-7 and 5-8 show the effect of laser pa rameters such as the total number of pulses and energy on the amount of solid PETN ablate d which directly affects the corresponding CL signal obtained. The dependence of the CL signa l on the amount of PETN ablated was projected on x,y axis for both graphs (solid points). Both sets of data best fitted into exponential equations of y = (4x10-6) m1.5021 and y = (2x10-5) m 1.6266 for the total number of pulses and laser energy dependent graphs, respectively. The y and m va riables correspond to th e CL signal and the amount of the ablated PETN, respectively. Both of these equations were used to estimate the amount of PETN residue that can be pos itively confirmed by the method. A positive confirmation is defined as the amount of PETN residue that provides a signal to background ratio of 10. A surface concentration range of 61 to 186 ng/cm2 of PETN was calculated. By using identical laser parameter conditions calibration curves for the vapor of RDX, PETN, and TNT were obtained (Figure 5-9). Ro ugh limits of detection (LOD) as well as the calibration sensitivity for each analyte were determined. The LOD is defined as (3 Blk )/m where Blk is the standard deviation of the backgr ound signal and m is the calibration sensitivity or the slope of the CL peak area versus the expl osive concentration (in pp b) plot. Table 5-3 gives the LOD values of 3.45 ppb for PETN, 1.73 ppb for RDX and 34.5 ppb for TNT. Ranking the compounds by LOD yields RDXRDX PETN. A priori, the anticipated result should be that the order of compounds LOD is parallel with their absorbance order. Surprisingly, this is not the case: The TNTs

PAGE 111

111 absorption coefficient is higher than that of RDX and PETN, yet its L OD is poorer. Clearly, the molecules absorption coefficient plays less of a role in its LOD. The LOD ratio of TNT to PETN is ~ 10 which is similar to the value reported previously92. A plausible argument for TNTs L OD being higher than PETN a nd RDX might be due to TNTs several alternative decomposition pathways that compete with R-NO2 bond scission. These include the oxidation of CH3 to form anthranil, 96; 97 nitro-nitrite isomerization98 and possibly rearrangement of the ring substituents.99 These pathways decrease the initial production of NOx(x= 1,2) and contribute to TNTs lower sensi tivity relative to RDX and PETN. PETN has a LOD that is a factor of nearly 2 times greater than that of RDX. This suggest that the process of NO2 release in RDX molecules is more complicated than the simple cleavage of a single nitro functional group and may involve the loss of more than one nitro group for each molecule. The energy for the rings N-N bond cleavag e is lowered after the removal of a nitro group, and further decomposition generating additional NO2 is feasible even without photon absorption.51 Conclusion PF efficiency of nitro compounds is not linearly related to the quantity of NO2 moieties of the parent molecule which was proven by photo dissociation of NT, TNT, and 2,4 NT. Photolysis of NO2 fragments into NO and O* was ve rified using luminol-CL detector. The interaction of laser photons with solid and vapor phase of PETN was investigated. At lower laser fluence, a linear increase on the quantity of NO2 fragments produced was observed for both solid and vapor phase of PETN. Howeve r, at higher laser flue nces, saturation of NO2 production occurred which implies that at highe r photon intensities, the proposed sequential twostep PF mechanism as shown below took place.

PAGE 112

112 R NO2 + h R + NO2; h = 193 nm (5-6) NO2 + h NO + O*; h = 193 nm (5-7) The presence of CrO3 oxidizer for NO to NO2 conversion significantly enhanced the sensitivity of the detector in analyzing both the solid and vapor phase of PETN. By using laser PF followed by NO2 detection with luminol-CL technique, it is feasible to detect energetic materials in real time at am bient conditions. Detecti on limits of 3.4 ppb for PETN, 1.7 ppb for RDX, and 34.5 ppb for TNT were obt ained. It was also demonstrated that the presence of PETN residue w ithin the range of 61 ng/cm2 to 186 ng/cm2 can be detected at a signal to background ratio of 10 using a few micro joules of laser energy.

PAGE 113

113 Table 5-1. Summary of the performance of various explosive detectors based on PF-FD technique. Research group/ Year Method/Comments LOD GW Lemire, JB Simeonsson, RC Sausa (1993)53 PF-REMPI-TOFMS CH3NO2 1000 ppb DMNA 450 ppb RDX 8 ppb TNT 24 ppb NB 2400 ppb JB Simeonsson, GW Lemire, RC Sausa (1993)54 PF-REMPI-TOFMS LOD (ppm) at =193nm CH3NO2 0.18 DMNA 0.51 NB 0.49 o-NT 0.12 m-NT 0.10 TNT 0.21 LOD (ppm) at =226nm 1.0 0.45 2.4 15 36 1.7 D Wu, J Singh, F. Yuch, D. Monts (1996)87 PF-LIF TNT in soil TNT = 40 ppb at 373 K G Boudreax, T. Miller, A. Kurefke, J Singh, F. Yuch, D Monts (1999)88 PF-LIF TNT H2O 500 ppm TNT soil 100 ppm T. Arusi-Parpar, D. Heflinger, R. Laui (2001)89 PF-LIF at 1atm, 24oC TNT 8 ppb at (S/N) = 10 V.Swayambunathan, G. Singh, RC Sausa (1999)90 PF-LIF both CH3NO2 and TNT are not detected at 454 nm and (355 + 450)nm LOD at 227 nm CH3NO2 4.3 ppm TNT 37 ppm LOD at (355+227 ) 3.3 ppm 2.6 ppm V.Swayambunathan, G Singh, RC Sausa (1999)91 PF-REMPI with ion probe detector PF-LIF LOD with REMPI TNT 70 ppb RDX 7 ppb PETN 2 ppb LOD with LIF 37 ppm NOT DETECTED NOT DETECTED

PAGE 114

114 Table 5-1. (Continued) Research group/ Year Method/Comments LOD Research group/ Year V.Swayambunathan, G Singh, RC Sausa (1999)92 PF-REMPI with ion probe detector PF-LIF LOD with PFREMPI at 227 and 454 nm (ppm), respectively PETN 0.5;20.4 RDX 0.4;not detected TNT 4.4 ppm;not detected LOD with PF-LIF at 227 and 454 nm (ppm), respectively 2.2;140 1.6 ppm;not detected 3.8 ppm;not detected J. Cabalo, R Sausa, (2003)93 Laser Surface PFFragment Detection Spectroscopy (REMPI-ion probe detector) RDX 1.4 ng/cm2 J. Cabalo, R Sausa, (2005)94 Laser Surface PFFragment Detection Spectroscopy (REMPI-ion probe detector) RDX 1.4 ng/cm2 HMX 2.0 ng/cm2 CL20 7.1 ng/cm2 TNT 15.4 ng/cm2 Monterola, M.P.P., Smith,B.W.,Omenetto, N.,Wineforner,J.W. (2007) PhotofragmentationLuminol-NO2 Chemiluminescence Method RDX 1.7 ppbv PETN 3.4 ppbv TNT 34.5 ppbv (S/N) = 3 PETN 61 to 186 ng/cm2

PAGE 115

115 Laser Power supply electrometer D/A converter N2gas PF cell hot plate computer Photofragmentation unit Oxidation unitluminol PF productsChemiluminescence unit Figure 5-1. Photofragmentati on-Chemiluminescence detector.

PAGE 116

116 N2gas N2gas Chemiluminescence set-up laser SV 1 A SV 3 SV 2 (3-way SV)Legend SV = solenoid valve B Figure 5-2. Photofragmentation set-up used for NO2 photolysis.

PAGE 117

117 Figure 5-3. Chemiluminescence signals obtained fo r nitro-aromatic compounds at varying laser energy. 0.02.55.07.510.012.5 0.0 5.0x10-61.0x10-51.5x10-52.0x10-52.5x10-53.0x10-5 c b a(a) 4-nitrotoluence (b) 2,4-nitrotoluence (c) 2,4,6-trinitrotoluence peak area of CL signal (C)193 nm laser pulse energy (mJ/pulse)

PAGE 118

118 Table 5-2. Determination of PF efficiency of nitro-aromatic compounds. Explosive sample a MM (amu) VP at 25oC (torr) Slope of the CL signal vs. Energy plot (C mJ-1) NT 137.14 1.5 x 10 -1 (199.1 ppm) 3.0x10-6 5.7x10-8 2-4 NT 182.14 2.1 x 10 -4 (0.28 ppm) 1.1x10-6 5.2x10-8 TNT 227.13 5.8 x 10 -6 (7.7 ppb) 1.7x10-6 8.7x10-8 a Abbreviations: NT, nitrotoluene; 2,4-NT, 2,4 dinitrotoluene; TNT, 2,4,6 trinitrotoluene

PAGE 119

119 A1B1--A2B20.0 4.0x10-78.0x10-71.2x10-61.6x10-62.0x10-62.4x10-6 peak area(C) LEGEND A1,A2 = without laser B1,B2 = with laser Figure 5-4. Chemiluminescence signals obtained before and after PF of NO2.

PAGE 120

120 0.02.55.07.510.012.5 -5.2 -5.0 -4.8 -4.6 -4.4 -4.2 -4.0 -3.8 -3.6 -3.4 without CrO3with CrO3log (peak area) laser energy (mJ/pulse)0.02.55.07.510.012.5 -5.2 -4.8 -4.4 -4.0 -3.6 -3.2 -2.8 -2.4 -2.0 without CrO3with CrO3log (peak area)laser energy (mJ/pulse)A B Figure 5-5. Effect of varying la ser energy on the CL signal of NOx(x =1,2) fragments obtained after photofragmentation of A) solid and B) vapor phase of PETN.

PAGE 121

121 02000400060008000 -5.2 -4.8 -4.4 -4.0 -3.6 -3.2 -2.8 -2.4 -2.0 solid vaporlog (peak area)total number of laser pulses Figure 5-6. Effect of varying total numb er of laser pulses on the CL signal of NOx(x =1,2) fragments obtained after PF of solid and vapor phase of PETN.

PAGE 122

122 0 20 40 60 80 100 120 0 1000 2000 3000 4000 5000 6000 7000 8000 0 2 4 6 8 10 tot a l num be r of l a s e r p ul s e s p e a k a r e a ( a u )m a s s o f P E T N a b l a t e d ( 1 06 g ) Figure 5-7. Effect of varying total number of laser pulses on the mass of PETN ablated and on the CL signal of NOx(x =1,2) fragments obtained after PF of solid PETN.

PAGE 123

123 0 5 10 15 20 25 30 35 0.0 2.5 5.0 7.5 10.0 0 2 4 6 8 10 l as e r en e r g y ( m J/ p u l s e)p e a k a r e a ( a u )m a s s o f P E T N a b l a t e d ( 1 0 6 g ) Figure 5-8. Effect of varying laser energy on th e mass of PETN ablated and on the CL signal of NOx(x =1,2) fragments obtained after PF of solid PETN.

PAGE 124

124 -10123 -4.8 -4.4 -4.0 -3.6 -3.2 -2.8 -2.4 -2.0 TNT PETN RDXlog (peak area)log (concentration,ppbv) Figure 5-9. Calibration curves of PETN, RDX, and TNT.

PAGE 125

125 Table 5-3 Analytical figures of merit of PF-luminol CL detector for some nitro-based explosives. Explosive sample a LOD at (S/N) = 3 Calibration sensitivity (CL signal/ppb) PETN 3 1 x 10 -5 RDX 2 2 x 10 -5 TNT 34 1 x 10 -6 a Abbreviations:PETN, 1,3-dini tro-2,2-bis(nitram ethylpropane); RDX, 1,3,5-trinitro-1,3,5triazacyclohexane; TNT, 2,4,6 trinitrotoluene

PAGE 126

126 CHAPTER 6 DIRECT DETECTION OF EXPLOSIVE IN SOIL Introduction Rapid and on-site detection of nitro-based explos ives in a complex matrix such as soil is essential in providing an appropriate feedback during the characterizati on or remediation of contaminated sites. This chapter begins with th e discussion of explosives as an environmental pollutant followed by a review of current m onitoring explosive techniques, and finally an evaluation of the analytical capab ility of PF-luminol CL detector for PETN analysis in soil. Environmental Hazards of Explosive Contaminated Soil Soil contamination of nitro compounds is a problem because of the scale on which explosives have been manufactured, used, a nd tested. Contamination occurred during the manufacturing of explosives which required la rge amounts of water for purification. The waste water from this process is usually pla ced in lagoons or sedimentation basins.100 Many sites also became contaminated through open detonation an d burning of explosives at army depots, evaluation facilities, artillery range s, and ordinance disposal sites. The United States Department of Defense has identified more than 1 000 sites with explosive contamination of which 87% ex ceeded permissible contaminant levels.101 Most contamination exists in near surf ace soils and in the vicinity of firing range targets. A primary concern is that soil contaminants may even tually migrate to groundwater and contaminate drinking water supplies of nearby communities. TNT, RDX, HMX, and PETN are weakly to moderately soluble in water with solubili ty of 130, 40, 5, and 43 mg/L, respectively at 25oC.102 Nitro-based explosives are toxi c and present harmful effects to all of life forms. TNT is on the list of US EPA priority pollutants: its a known mutagen and can cause pancytopenia as a result of bone marrow failure.103 TNT also causes the formation of methemoglobin on acute

PAGE 127

127 exposures and anemia on chronic exposures. It also causes toxic hepatitis, leukocytosis, peripheral neuritis, muscular pain s, cardiac irregularities, rena l irritation, and bladder tumors.1 On the other hand, RDX is considered as a po ssible human carcinogen (Class C) of EPA. RDX caused liver tumors in mice that were exposed to it in the diet. RDX also produced smaller offspring in rats. Clinical findings in low leve l long term exposures may include tachycardia, hematuria, proteinuria, mild anemia, neutroph ilic leukocytosis, and electroencephalogram (EEG) abnormalities.2 In general, nitro compounds are generally recalcitrant to the biosphere, where they constitute a source of po llution due to both toxic and mu tagenic effects on human, fish, algae, and microorganisms. Review of Explosive Detection in Soil Various analytical methods for detection of n itro-based explosive present in soil have been reported. These are mostly spectrometric su ch as Electron Capture Detector (ECD) 11; 104, Thermal Energy Analyzer (TEA) 12; 105, Ion Mobility Spectrometry 15; 16 and UV Absorption coupled with several different separation me thods such as extractions, gas and liquid chromatography, and electrophoresis. Voltammetric and amperometry 106; 107 methods were also commonly used in determination of trace levels of explosive substances in soil. The basic principles of these methods were discussed in Chapter I. The current techniques menti oned above are often cost prohibitive, time consuming, and labor intensive. To characterize a site for explosive detection usi ng these methods, it is necessary to follow a standard protocol starting from sample collection. Then, the analyte from the soil is extracted from the soil sample us ing organic solvent such as acet onitrile under at least 18 hour of sonication in cooled ultrasonic water bath before the actual analysis. Indeed, these traditional methods are laborious, time consuming, impract ical and ineffective for field analysis.

PAGE 128

128 Explosive field analysis is valuable due to extremely heterogeneous distribution of explosives in contaminated soils requiring e normous sampling sites. With such enormous sampling sites, a cost effective tool is necessar y. Since cost per sample is lower, more samples can be analyzed producing more re liable results. Also, the availabi lity of near-real time results permits re-design of the sampling scheme while in the field which facilitates more effective use of off site laboratory facilities w ith more robust analytical methods. Currently, field methods for detection of explosives in soil include colorimetric108 and immunosensors methods.109 Both methods have several drawb acks. A colorimetric technique is used only for rugged qualitative analysis and of ten gives unreliable results. Immunosensors suffer because antibodies of all explosive material s likely to be encountered are needed, i.e., specific antibody is required for each compound of interest. This adds to the cost of an immunosensors. Also, when multi-analyte sensors are used, there is poor signal discrimination and as a result sensitivity is lost. PF-Luminol CL detector is a promising techni que for both on-site and off-site explosive detection in soil. The newly developed method ha s the following attributes: (1) it is simple and straightforward technique (2) it requires no sample preparation; (3) it is fa st and provides a real time response; (4) all co mponents including CL sensors can be made rugged and field portable. An actual analysis of PETN contaminated soil was demonstrated and the analytical capability of the method was discussed. Experimental Section Chemical and Reagents All the reagents and chemicals used for this experiment were exactly the same as Chapter V. The soil matrix used was a standard refere nce material (SRM-2704) prepared by the National Institute of Standards and Technology (NIST). SRM-2704 is freeze dried river sediment that was

PAGE 129

129 sieved and blended to achieve a high degree of homogeneity. It is intended primarily for use in the analysis of sediments, soils, or material of similar matrix. Apparatus and Methodology The experimental set-up is the same as Ch apter V except for the CL cell reaction. During the course of the dissertation, the geometry of the CL cell reaction underwent several revisions. The final reaction cell was designed with severa l goals in mind. First, the area of the PMT window was maximized. It was hoped that this w ould increase the detector sensitivity. A second expectation was this configuration prevented co ndensing of the luminol solution on the surface of the PMT window. Finally, by providing a co mpartment waste for the luminol solution, necessary maintenance such as periodically cl eaning the cell was no longe r necessary. The cross section of the CL cell reaction is shown in Figure 6-1. The method of analysis is the same as described in Chapter V except for the laser parameters. In each experimental run, 8 000 laser pulses with average lase r energy of 8 mJ/pulse was used. The laser repetition rate was set to 50 pulses per second. Sample preparation A known mass of soil was accurate ly weighed on a small piece of weighing paper. The soil sample was quantitativel y transferred and tightly packed into a small aluminum cylindrical cup container (O.D. = 1.9 mm; Depth = 4.5 mm). Exactly 40 L of the standard PETN solution in acetone was deposited on the top of the soil sample. Then the sample was dried in the oven at 60oC for an hour to evaporate the solvent. To achieve varying concentration of PETN in soil, different st andard PETN solutions ranging from 0.4 to 0.16 mg/mL were prepared. The explosive concentrati on in soil (mg/g) was calculated by using the following equation: Explosive contamination (mg/g) = [PETN] (mg/mL) (40 L) (10 -3) (5-1) Weight of soil (g)

PAGE 130

130 Results and Discussion The analysis was done by imaging an ArF lase r beam on the explosive contaminated soil. The photofragmented products were then m easured using Luminol-CL detector. Each experimental run was at least re peated eight times on the same sample spot. A calibration curve was generated by plotting peak ar ea against concentration rang ing from 0.4 to 9.0 parts per thousand PETN. Figure 6-2 best fi tted a polynomial equation of y = 2.4x2 + 3.9x + 2.8 where y equal to the CL peak area and x is the concentr ation of PETN in soil in parts per thousand. The curve has a correlation coefficient value (R2) equal to 0.9996. Based on th is curve, a detection limit was established. The limit of detection (LOD) is defined as (3 Blk/m) where blk is the standard deviation of the blank a nd m is the calibration sensitivity or the slope of the line. This definition is restricted for the cases of a calibra tion line where the slope is constant, however, the calibration curve obtained for this analysis is curvilinear (slope is variable) and the definition provided is not useful without modification. The above equation can be redefined as follows: LOD = 3 blk/{dy/dx[f(x)]} (5-2) where the slope is replaced as a differential coef ficient of function x, whic h in this case is the PETN concentration in soil. The equa tion of the slope is then given by: m = 2(2.4x) + 3.9 (53) LOD range of 0.5 to 4.3 ppm of PETN was obtaine d with x values equal to 0.4 and 9.0 parts per thousand of PETN. The precision of the method has been established at the lowest concentration of PETN which gave a relati ve standard deviation (RSD) of 30 % where n = 10. Larger variations obtained in the signal especially in the lower concentration range, is due to the decreasing surface concentration of the contaminan t as the laser ablates deeper into the soil matrix.

PAGE 131

131 The LOD was also estimated by considering onl y the linear plot obtained from the first four data points. A linear equation of y = 15.662x 4.2364 was obtained where y is the peak area of the signal and x is the PETN concentration in ppth. An LO D of 1.6 ppm was calculated from this plot. The LOD of the method is comparable to othe r current techniques as can be seen in Table 1-1 and 5-1.The analytical capabil ity of this method can be applie d in analysis of any typical explosive contaminated soil given in Table 6-1. Based on the data tabulated in Table 6-1, the TNT concentration in the soil ranges from 4 000 to 1 200 ppm whereas the concentration of other explosive contaminants is less than 300 pp m. The nitrogen content of the soil was 7.5 ppm as ammonium ion and nitrate concen trations vary from 6 to 12 ppm.102; 110 Pure NaNO2 and NaNO3 were photolyzed and analyzed wi th the same method in order to investigate the potential interf erent during the analysis but bo th compounds did not give a significant CL signals as can be seen in Figure 6-3. Conclusion PF-Luminol-CL detector is a pr omising method for analysis of trace explosive in soil. The LOD range of 0.5 to 4.3 ppm for PETN was es tablished. The PF-luminol CL detector is comparable to current techniques available. Th e most advantageous characteristic of the PFluminol CL detector compared to other traditional methods is its simplicity and reliability. The technique also requires no sample preparat ion and provides fast real-time responses.

PAGE 132

132 CL cell luminol-NO2mist waste compartment Figure 6-1. Cross section of the CL cell.

PAGE 133

133 Figure 6-2. Calibration curve fo r PETN contaminated soil.

PAGE 134

134 Table 6-1. Explosives concentra tion in the contaminated soil 102; 110 Explosive a Concentration range (mg explosive/kg of soil) TNT 4 000-12 000 TNB 175-300 2,4-DNT 50-200 RDX 50-125 HMX 50-100 a Abbreviations: TNT, 2,4,6-trinitrotoluene; TNB, trinitrobenzene; 2,4 DNT, 2,4-dinitrotoluene; RDX, hexahydro-1,3,5-trinitro -1,3.5-triazine; HMX, oc tahydro-1,3,5,7-tetranitro-1,3,5,7tetraazocine

PAGE 135

135 -10010203040506070 -13 -12 -11 -10 -9 -8 -7 -6 -5 CL signal (pA)time (s) -10010203040506070 -13 -12 -11 -10 -9 -8 -7 -6 -5 CL signal (pA)time (s) -10010203040506070 -16 -14 -12 -10 -8 -6 -4 CL signal (pA)time (s) A B C Figure 6-3. CL Spectra after PF of A) empty sample holder B) pure NaNO2 C) pure NaNO3.

PAGE 136

136 CHAPTER 7 CONCLUSIONS AND FUTURE WORKS Summary and Conclusions The purpose of this research was to devel op a simple, fast, reliable, sensitive and potentially portable explosive detection device employing laser phot ofragmentation (PF) followed by heterogeneous chemiluminescence (CL) detection. The work was divided into four stages: (1) discerning the PF pathways of n itro-based explosives (2) luminol-CL method development and optimization for improving NO2 sensitivity (3) construction of CrO3 oxidizer for NO to NO2 conversion (4) evaluation of the device in detecting nitro-based explosives in various media such as air and soil. In the first stage of the research, the followi ng PF pathways were verified using classical colorimetric analysis and fast time resolved absorbance method. R NO2 + hv R + NO2; hv = 193 nm (7-1) NO2 + hv NO + O*; hv = 193 nm (7-2) Confirmation of NOx(x=1,2) generation after photolysis of nitrobased explosives was the basis for developing an explosive detector ba sed on PF-Fragment detection scheme. The second stage involved the development and optimization for NO2 detection using luminol CL. In this system, a stream of NO2 gas passed through a concen tric nebulizer and used to aspirate a spray of luminol solutio n. The aspiration process maximized NO2/luminol contact area, thereby, enhancing the CL signal. The curre nt system was able to improve the detection limit attained so far through optimization of all th e feasible physical and chemical parameters involved in the chemiluminescence reaction between luminol and NO2. Detection limit of 19 ppt NO2 at (S/N) = 3 was reported. The optimal reagent solution from the viewpoint of sensitivity of the response to NO2 (maximum signal/signal noise ratio) is 5 x 10-3 M luminol + 0.01 M p-

PAGE 137

137 iodophenol + 0.2 M KOH. Luminol/H2O2 CL set-up was also explored where a bundle of porous polypropylene fibers was used to bring the NO2 into contact with luminol solution. LOD of 178 ppt NO2 at (S/N) = 3 was obtained. In the third stage, a CrO3 oxidizer was constructed for NO to NO2 conversion. Photolysis of nitro compounds produced NO (reac tion 7-2) which is not detect able by luminol-CL system. To further increase the sensitivity of the device for NO2 fragments, NO must be converted back to NO2. Finally, the fourth stage was integr ation of PF and CL units by a CrO3 oxidizer into an explosive detection device. The sy stem was able to detect energe tic materials in real time at ambient conditions. Detection limits of 3.4 ppbv for PETN, 1.7 ppbv for RDX, and 34.5 ppbv for TNT were obtained. It was also demonstrated that the presence of PETN re sidue within the range of 61 ng/cm2 to 186 ng/cm2 can be detected at a given signal to background ratio of 10 using a few micro joules of laser energy. The technique also demonstrated its potential for direct analysis of trace explosive in soil. LOD range of 0.5 to 4.3 ppm for PETN was established, an analytical capability comparable to current techniques av ailable. The technique most advantageous characteristic compared to other traditional methods is its simplicity and reliability. The technique also requires no sample preparat ion and provides fast real-time response. Future Research Directions Based on the results presented in this disserta tion, several paths can be followed for future research studies. The future work can be gr ouped into (1) extending th e systems applicability towards peroxide-based explosives (2) instruments field adaptability and finally (3) instruments performance in actual scenario. Liquid peroxide explosives such as tria cetoneperoxide (TATP) and hexamethylene triperoxide diamine (HMTD) were home-made e xplosives recently used in Madrid and London

PAGE 138

138 subway attacks of 2004 and 2005, re spectively. TATP was also been used by suicide bombers in Israel, and was chosen as a detonator in 2001 by the thwarted shoe bomber Richard Reid.2 Current analytical methods used for per oxide-based explosives include Infrared spectroscopy (IR), Raman spectroscopy, Chemical Ionization-Mass Spectrometry (CI-MS), Ion Mobility Spectroscopy (IMS), 1H and 13C NMR, and High Performance Liquid Chromatography (HPLC) with post column UV irradiation and fl uorescence detection. Most of these techniques are applied in laboratory environm ent and usually not applicable to field screening scenarios. So far, the only detection method used in fi eld analysis involved UV irradiation ( = 254-415 nm) of the liquid explosive samples followed by hydrogen peroxide (H2O2) fragment detection using colorimetric technique.111 However, this particular method i nvolves a lot of sample preparation, used a lot of expensive r eagents, time-consuming and relatively has poor sensitivity. Furthermore, the technique is only limited to liquid samples. An alternative way to overcome these issues is through direct photofragmentatati on (PF) of peroxide-bas ed explosive vapor and subsequent H2O2 detection using luminol-CL detector (Figure 7-1). The high volatility of these compounds is a very promising premise for devel oping the first known air analysis for peroxidebased explosive using our method. After extending the chemistry of the system with peroxide-based explosives, the second direction to be considered is it s portability for field adaptabil ity. The over-all set-up including the detection system can be readily miniaturized. Th e most challenging task is the replacement of ArF laser but this can be feasibly done using commercially available excilamps which only weigh about 2.2 kg.112 Excilamps provide narrow band light around a single wavelength by radiative decomposition of excime r states created by a barrier discharge. Lamps operating at wavelengths of 126, 146, 172, 193, 222, 282, and 308 are already available. Studies on

PAGE 139

139 excilamps show a powerful radiation in UV a nd vacuum UV spectral range that allows photochemical processes that were no t previously possible with lamps.113 These lamps also operate at low temperature, generate minimal IR radiation, and allow pr ocessing of temperature sensitive substrates such as explosives. Furthermor e, these equipments are highly efficient source converting power into light effectively and ther efore, reducing power requirements and heat load. The system is also robust, inex pensive, and ecologically beneficial.113 After achieving the portability re quirement, actual field testing of the instrument should be evaluated. Foreseeable problem that can be enco untered during the actual testing will be the sensitivity issue of the device esp ecially when dealing with direct explosive vapor detection with an exemption for peroxide-based explosives. As mentioned in Chapter I, nitro-based explosives significantly have low vapor pressu res and the method that is suitable for its detection should be able to detect concentrations down to less th an 1 ng/L (ppb level). Our current method has a LOD of 3.4, 1.7, and 34.5 ppbv for PETN, RDX, and TNT, respectively. Improving the sensitivity of the device can be done by the follow ing suggestions: (1) replacing the PMT detector with another PMT sy stem which has the least dark count rate and highest photon gain and (2) elim inating the background noise due to interferences present in air such as native NOx(x=1,2), O3, and SO2. Previous works 73-80 with atmospheric monitoring of NO2 were able to eliminate almost 100 % of O3 and SO2 from air by adding Na2SO3 in the basic luminol solution. O3 and SO2 preferred and readily react with a stronger reducing agent such as Na2SO3 rather than basic luminol solution. Interferences from native NOx(x=1,2) can be compensated in se veral ways. The first one involved NOx(x=1,2) removal prior to sample analysis. This must be done without removing the

PAGE 140

140 explosive vapor or the analyte. One feasible way to do this is through chemical adsorption of NO2 on a surface of titanium dioxide (TiO2) photo catalyst and hygroxyapatite (Ca10(PO4)6(OH)2.114 In this method, NO was oxidized to HNO3 which is efficiently adsorbed on the surface of TiO2. Removal efficiency of 98 % was achieved for this method. Another alternative option is to purge th e air sample with solutions of either triethanolamine (TEA) or sodium arsenite prior to photofragmentation. Both solutions are known for their ability to preconcentrate NO2 in air by trapping and converting NO2 gas to nitrite ions in the bulk solution. These absorber solutions have absorption effi ciencies of 95 and 82 percent, respectively.59 The stoichiometric factor of TEA for NO2 to NO2 is 0.85 while that of sodium arsenite is unity.115 The main concern of using these NOx(x=1,2) removal methods is that the feasibility of removing some of the analyte as well. Explosive vapor is prone to get easily lost to wall adsorption during transport. The second way to compensate for the presence of native NOx(x=1,2) in air is to divide the volume of the air sample pumped into the inst rument each time. The first sample (sample A) collected will be diverted to PF cell while the second sample (sample B) will be diverted to another cell. Both samples will undergo the same reactions of oxidation and luminol-CL except for photolysis part. The presence of explosive va por can be confirmed by a significantly higher CL signal of the air sample that undergoes photolys is than the air sample that was not treated with UV irradiation. A positive explosive alarm w ill be based on the stat istically established differential value between sample A and B. In conclusion, the main drawback of PF-CL detector is its relativ ely high background noise due to interferents present in the air sample specifically the native NOx(x=1,2), thus, solid residue analysis of nitro-based explos ive is preferred and more comp atible for this technique.

PAGE 141

141 luminol-CL detector Figure 7-1. Alternative scheme for de tecting peroxide-b ased explosives.

PAGE 142

142 LIST OF REFERENCES 1. Singh, S. J. Hazard. Mater. 2007 in press. 2. Mahoney, C.M.; Gillen, G.; Fahey, A.J. Forensic Science International 2006, 158 39-51. 3. Vila, M.; Lorber-Pascal, S.; Laurent, F. Environmental Pollution 2007, 148 148-154. 4. Ahmad, F.; Schnitker, S.P.; Journal of Contaminant Hydrology 2007, 90, 1-20 5. Agrawal,; J.P., Surve,; R.N.; Mehilal,; S. H.; Sonawane,; A.L. Journal of Hazardous Materials 2000, 77, 11-31. 6. Hallowell, S. F. Talanta 2001 54 447-458. 7. Yinon, J. Trends Anal. Chem. 2002 21 292-301. 8. Hilmi, A.; Luong, J.T.; Environ. Sci. Technol 2000 34 3046-3050. 9. Lu, Q.; Collins, G.E.; Smith, M.; Wang, J. Analytica Chimica Acta 2002 469 253-260. 10. Mullen, C.; Irwin, A.; Pond, B.; Huestis, D.L.; Coggiola, M.J.; Oser, H. Anal. Che m. 2006 78 3807-3814. 11. Walsh, M.E. Talanta 2001 54 427-438. 12. Lafleur, A.L.; Mills, K.M.; Anal. Chem. 1981 53 1202-1205. 13. Asbury, G.R.; Klasmier, J.; Hill, H.H. Talanta 2000 50 1291-1298. 14. Ewing, R.G.; Atkinson, D.A.; Eiceman, G.A.; Ewing, G.J. Talanta 2001 54 515-529. 15. Buryakov, I.A. J of Chromatogr 2004 800 75-82. 16. Khayamian, T.; Tabrizchi, M.; Jafari, M.T. Talanta 2003 59 327-333. 17. Matz, L.M.; Tornatore, P.S.; Hill, H.H. Talanta 2001 54 171-179. 18. Kanu, A.B.; Haigh, P.E.; Hill, H.H. Analytica Chimica Acta 2005 553 148-159. 19. Otto, J.; Brown, M.F.; Long, W. Applied Animal Behaviour Science 2002, 77, 217-232. 20. Harper, R.J.; Almirall, J.R.; Furton, K.G. Talanta 2005 67 313-327. 21. Harvey, S.D.; Clauss, T.W. J. Chromatogr A 1996 753 81-86.

PAGE 143

143 22. Pnnaduwage, L.A.; Boiadjiev, V.; Hawk, J.E.; Thundat, T. Appl. Phys. Letter 2003 83 1471-1475. 23. Datskos, P.G.; Lavrik, N.V.; Sepaniak, M.J. Sens. Letters 2003 25-37. 24. Ly, S.Y.; Kim, D.H. Talanta 2002 58 919-926. 25. Hilmi, A.; Luong, J.T.; Environ. Sci. Technol 2000 34 3046-3050. 26. Wang, J.; Thongngadee, S. Analytica Chemica Acta 2003 485 139-143. 27. Albert, K.J.; Myrick, M.L.; Brown, S.B.; Ja mes, D.L.; Milanovich, F.P.; Watt, D.R. Environ. Sci. Technol 2001 35 3193-3200. 28. Charles, P.T.; Kustebeck, A.W. Biosens. Bioelectron 1999 14 391-400. 29. Dorozhkin, L.M.; Nefedov, V.A.; Sobe lnikov, A.G.; Sevastjanov, V.G. Sensors and Actuators B 2004 99 568-570. 30. Sohn, H.; Calhoun, R.M.; Sailor, M.J.; Tragler, W.C. Angew. Chem 2001 13 21622165. 31. Wallenburg, S.R.; Bailey, C.G. Anal. Chem 2000 72, 1872-1875. 32. Wang, X.; Zeng, H.; Zhao, L.; Lin, J.M. Talanta 2006 70 160-164. 33. De Lucia, F.C.; Harmoun, R.S.; McNes by, K.L.; Winkel R.J.; Misiolek A.W. Applied Optics 2003 42 6148-6150. 34. Lin, H.; Chen, Y. Optics Express 2006 14 415-424. 35. Shen, Y.C.; Lo, T.; Taday, P.F.; Cole, B.E.; Tribe, W.R.; Kemp, M.C. Applied Physics Letters 2005 86 241116-1 to 241116-3. 36. Sylvia, J.M.; Janni. J.A.; Klein, K.D.; Spencer, K.M. Anal. Chem. 2000 72 5834-5840. 37. Hayward, I.P.; Kirkbride, T.E.; Batchelder, D.N.; Lacey, R.J. J. of Forensic Sciences 1995 40 883-884. 38. Janni, J,; Gilbert, B.D.; Fi eld, R.W.; Steinfeld, J.I. Spectromica Acta Part A 1997 73 255-258. 39. Xu, S.; Sha, G.; Xie, J. Rev. Sci. Instrum 2002 73 255-258. 40. Balint-kurti,;G.G., Shapiro, M. Chemical Physics 1981, 61, 137-155

PAGE 144

144 41. Turro, N.J. Molecular Photochemistry Amsterdam, 1967 42. Mc Donnell L.; Heck, A.J. J. of Mass Spectrometry 1998 33 415-428. 43. Cabalo, J.; Sausa, R.C. Applied Optics 2005 44, 1084-1091. 44. Simeonsson, J.B.; Sausa, R.C. Trends in Analytical Chemistry 1998 17 542-550. 45. Monts, D.L.; Singh, J.P.; Boudreaux, G.M. Encyclopedia of Analytical Chemist ry, John Wiley and Sons Ltd., Chichester, 2000 2148-2171. 46. Ledingham, K.W.D.; Singhal, R.P. International Journal of Mass Spectrometry and Ion Processes 1997 163 149-168. 47. stmark, H.; Carlson, M.; Ekvall, K. Combustion and Flame 1996, 105 381-390. 48. Butler, L.J.; Krajrovich, P.; Lee, Y.T. J. Chem. Phys. 1983 79 1708-1722. 49. Blais N. Journal of Chem. Phys 1983 79 1708-1722. 50. Renlund, A.M.; Trott, W.M. Chemical Physics Letters 1984 107 555-560. 51. Capellos,C.; Papagrannakopoulos, P.; Liang, Y. Chemical Physics Letters 1989 164 533-538. 52. Galloway, D.B.; Bartz, J.A.; Huey, G.L.; Crim, F.F. J. Chem. Phys 1993 98 2107-2114. 53. Lemire,G.W.; Simeonsson, J.B.; Sausa, R.C. Anal. Chem 1993 65 529-533. 54. Simeonsson, J.B.; Lemire,G.W.; Sausa, R.C. Applied Spectroscopy 1993 47 1907-1912. 55. Marshall A.; Clark, A.; Ledingham, K.W.; Sanders, J.; Singhal, R.P.; Kosmidis, C.; Deas R.M. Rapid Communications in Mass Spectrometry 1994 8 521-526. 56. Ledingham, K.W. Physica Scripta 1995 158 100-103. 57. Simeonsson, J.B.; Lemire,G.W.; Sausa, R.C. Army Research Laboratory Aberdeen Ground, MD, USA Available NTIS Report 1994 30 Rep. Announce. Index 95(19) Abstract no. 5429519. 58. Yang, Mo; Ramsey, J.M.; Kim, B. Journal of Rapid Communications in Mass Spectrometry 1996 10 311-315. 59. Gayathri, N.; Balasubramanian, N. Analusis 1999 27, 174-181.

PAGE 145

145 60. Finlayson, B.J.; Pitts J.N. Atmospheric Chemistry: Fundamentals and Experimental Techniques John Wiley and Sons Inc. Canada, 1986 151-153. 61. Sun, F.; Glass, G.P.; Curl, R.F. Chemical Physics Letters 2001 337 72-78 62. Campana, A.G.; Baeyens, W.R. Chemiluminescence in Analytical Chemistry Markel Dekker Inc., Basel, Swizerland, 2001 ; Chapter 1. 63. Ingle, J.D.; Crouch, S.R. Spectrochemical Analysis Prentice-Hall Inc., New Jersey, 1988 ; Chapter 15. 64. Li, Y.; Qi, H.; Fang, F.; Zhang,; C. Talanta 2007 72 1704-1709. 65. Isacsson, U.; Wettermark, G. Analytica Chimica Acta 1974 68 339-362. 66. Wendel, G.J. Development and Application of a Luminol-based NO2 Detector University of Michigan, 1985 67. Tahirovi A.; opra, A.; Mikli anin, E.O.; Kalcher, K. Talanta 2007, 72 1378-1385. 68. Su, Y.; Li, X.; Chen, X.; Lu, Y.; Hou, X. Microchemical Journal 2007 In Press, Corrected Proof. 69. White, E.H.; Zafriou, H.H.; Hill J.H.M. J. Am. Chem. Soc. 1964 86 940-941. 70. Shevlin,P.B.; Neufeld, H.A. J. Org. Chem. 1970 35 2178-2182. 71. Hersh, B.B.; Wesely, M.L. J. Photochem. 1974 10 409-423. 72. Roswell, D.F.; White, E.H. Meth. Enzynol 1978 57 409-423. 73. Maeda, Y.; Aoki, K.; Munemori, M. Anal.Chem .1980, 52 307-11. 74. Wendel, G.J.; Stedman, H.; Cantrell, C.A. Anal.Chem 1983 55 937-40. 75. Schiff, H.I.; Mackay, G.I.; Castledine, C.; Harns, G.W. Trans Q. Water, Air and Soil Pollution. 1986 30 105-114. 76. Kelly T.J., Spicer C.W., Ward G.F. Atmospheric Environment A 1990 24 2397-2403. 77. Mikuska, P.; Vecera, Z. Anal. Chem 1992 64 2187-91. 78. Collins, G.; Rose-Phersson, S.L. Anal. Chem 1995 67 2224-30. 79. Robinson, J.K.; Bollinger, M.J.; Birks, J.W. Anal. Chem 1999 71 5131-5136.

PAGE 146

146 80. Kok, G.L.; Hoiler, T.P.; Lopez, M.B.; Nachitrieb, H.A.; Yuan, M. Environmental Science and Technology 1978 12 1072-1076. 81. Thorpe, G.H.; Kricka, L.J.; Moseley, S.B.; Whitehead T.P. Clinical Chemistry 1985 31 1335-1341. 82. Robinson, J.K. Luminol-Hydrogen peroxide Detector fo r the Analysis of NO in Exhaled Breath University of Colorado, 1994 83. Hutchinson, G.L.; Yang, W.X.; Andre, C.E. Atmospheric Environment 1999 33 141145. 84. Levaggi, P.; Kothny, E.L.; Belsky, T. ; de Vera E.; Mueller, P.K. Environmental Science and Technology 1974 8 348-350. 85. Nguyen Dao Hinh. US Patent no. 20040169495, 1994. 86. Nguyen Dao Hinh. US Patent no. 20040053421, 1994 87. Wu, D.; Singh, J.; Yueh, F.; Monts, D. Applied Optics 1996 35 3998-4003. 88. Boudreax, G.M.; Miller, T.S.; Kunefke, A.J.; Singh, J.; Fang-Yu, Y.; Months D. Applied Optics. 1999 38 1411-1417. 89. Arusi-Parpar, T.; Heflinger, D.; Lavi, R. Applied Optics 2001, 46 6677-81. 90. Swayambunathan, V.; Singh, G.; Sausa, R.C. Applied Spectroscopy 2000 54 651-657. 91. Swayambunathan, V.; Singh, G.; Sausa, R.C. Applied Optics 1999 40 6447-6454. 92. Swayambunathan, V.; Singh, G.; Sausa, R.C. Proceedings of International Society for Optical Engineering 1999 176-184. 93. Cabalo, J.; Sausa, R.C. Applied Spectroscopy 2003 57 1196-1199. 94. Cabalo, J.; Sausa, R.C. Applied Optics 2005 44 1084-1091. 95. Lucero, D.P.; Boncyk, E.M. Journal of Energetic Materials 1986 4 473-510. 96. Chakraborty, D.; Muller, R.P.; Dasdupta, S.; Goddard, W.A. J. Phys. Chem 2001 105 1302-1314. 97. Gonzalez, A.C.; Larson, C.W.; McMillen, D.F.; Golden, D.M. J Phys. Chem. 1985 89 4809-4814.

PAGE 147

147 98. He, Y.Z.; Cui, J.D.; Mallard, W.G.; Tsang, W. J. Am. Chem. Soc. 1988 110 3754-3759. 99. Kallman, H.P.; Spruch, G.M. Luminescence of Organic and Inorganic Materials John Wiley and Sons, Inc., NY, 1969 100. Pennington, J.C.; Brannon, J.M. Themochimica Acta 2002 384 163-172. 101. Rodgers, J.D.; Nigel, J.B. Water Research 2001 35 2101-2111. 102. Halasz, J.D.; Groom, C.; Zhou E.; Paquet, L. ; Ampleman, G.; Dubas, C.; Hawari, J. Journal of Chromatog. A 2002 963 411-418. 103. Morley, M.C,; Yamamoto, H.; Speitel, G.; Clausen, J.; Journal of Contaminant Hydrology 2006 85 141-158. 104. Jenkins, T.F.; Leggett, D.C.; Miyares, P.H. ; Walsh, M.E.; Ranney, T.A.; Cragin, J.H.; George V. Talanta 2001 54 501-513. 105. Bowerbank, C.R.; Smith, P.A.; Futterrolf, D.D.; Lee, M.L. Journal of Chromatog. A 2000 902 413-419. 106. Buttner, W.J.; Findlay, M.; Vickers, W.; Da vis, W.M.; Cespedes, E.R.; Cooper, S.I.; Adams, J.W. Analytica Chimica Acta 1997 341 63-71. 107. Hilmi, A.; Laong, J.H.; Nguyen, A. Journal of Chromatog. A 1999, 844 97-110. 108. Uzer, A.; Ercag, E.; Apak, R. Analytical Chimica Acta 2005 534 307-317. 109. Gauger P.R.; Hoh D.B.; Patterson, C.H.; Char les, P.T.; Shriver-Lake L.; Kusterbeck, A.W. Journal of Hazardous Materials 2001 83 51-63. 110. Boopathy R. International Biodeterioration and Biodegradation 2000 46 29-36. 111. Schutte-Ladbeck, R.; Karst, U Analytica Chimica Acta 2003 482 183-188. 112. Li, Q.; Gu, C.; Di, Y.; Yin, H.; Zhang, J. Journal of Hazardous Materials 2006 133, 6874. 113. Elsner, C.; Lenk, M.; Prager, L.; Mehner, R. Applied Surface Science 2006 2 52, 36163624. 114. Komazaki, Y.; Shimizu, H.; Tanaka, S. Atmos. Environ 1999 33 4363-4371. 115. Yuen, W.K.; Horlick, G. Analytical Chemistry 1977 49 1448-1450.

PAGE 148

148 BIOGRAPHICAL SKETCH Maria Pamela Pineda Monterola was bor n on September 18, 1977, in San Pablo City, Philippines. She is the third child of Nora Di zon Pineda and Conrado Cachero Monterola. She graduated cum laude with bachelors degree in chemistr y in 1998 from the University of the Philippines. She was immediately hired as a junior faculty in the same university while completing her masters degree in computational chemistry and minor degree in mathematics. In 2000, she took and topped the Philippine Licensure Examination in chemistry. In 2001, she was accepted as a research scholar in University of Tsukuba, Japan for a year. In 2007, she earned her Ph.D. degree in analytical chemistry under th e research supervision of Prof. James D. Winefordner at the University of Florida.