Fluorescence of aromatic amines and their fluorescamine derivatives for detection of explosive vapors

Article abstract of DOI:10.1366/000370206778397353

Nitroaromatics (such as dinitrotoluene, trinitrotoluene, and nitrobenzene) found in explosive vapors from buried landmines can be reduced to aminoaromatics by a novel process involving Pd metal nanocatalysts prepared in supercritical fluid carbon dioxide and supported on multi-walled carbon nanotubes. These aminoaromatics are fluorescent and, if desired, the fluorescence yield can be increased and the fluorescence maxima shifted further toward the red by reaction with appropriate derivatizing agents such as fluorescamine. Corrected spectra for these chemicals and their derivatives are included. Subpicomolar detection limits have already been achieved using a laboratory spectrofluorometer with a 150 W Xe arc lamp. Using lasers as excitation sources, this approach has the potential for developing a field sensor competitive with other methods currently used for detecting explosive vapors from land mines.

Full text of DOI:10.1366/000370206778397353

Fluorescence of Aromatic Amines and Their Fluorescamine
Derivatives for Detection of Explosive Vapors
DELYLE EASTWOOD,* CARLOS FERNANDEZ, B. YUNGHOON YOON,
CHRYSTAL N. SHEAFF, and CHIEN M. WAI
Department of Chemistry, University of Idaho, P.O. Box 442343 Moscow, Idaho 83844-2343
Nitroaromatics (such as dinitrotoluene, trinitrotoluene, and nitrobenzene)
found in explosive vapors from buried landmines can be reduced to
aminoaromatics by a novel process involving Pd metal nanocatalysts
prepared in supercritical fluid carbon dioxide and supported on multi-
walled carbon nanotubes. These aminoaromatics are fluorescent and, if
desired, the fluorescence yield can be increased and the fluorescence
maxima shifted further toward the red by reaction with appropriate
derivatizing agents such as fluorescamine. Corrected spectra for these
chemicals and their derivatives are included. Subpicomolar detection
limits have already been achieved using a laboratory spectrofluorometer
with a 150 W Xe arc lamp. Using lasers as excitation sources, this approach
has the potential for developing a field sensor competitive with other
methods currently used for detecting explosive vapors from land mines.
providing greater specificity, which might have advantages for
certain applications. An environmental high-performance
liquid chromatography (HPLC) method with fluorescence
7
detection for aminoaromatics in soil has been reported. This
is a laboratory method and requires the use of more solvent. To
the extent that we are talking about explosive vapors rather
than soil or ground water samples, most matrix and quenching
8
effects are eliminated. As stated in Guilbault, aromatic nitro
compounds are not good candidates for analysis by fluores-
cence spectroscopy unless they are converted chemically into
fluorescent compounds with amino groups by chemical
reduction. Fluorescence spectra of aminoaromatics have been
8
–10
11,12
Index Headings: Aminoaromatics; Explosive vapors; Fluorescence; Nitro-
aromatics; Palladium nanoparticles.
reported in the literature.
Kasha
has discussed the
theory of fluorescence for chemicals like the aromatic amines.
The current paper includes an updated, corrected set of
fluorescence spectra using the purest aromatic amines available
including aniline, o- and p-toluidines, 2,4- and 2,6-diaminotol-
uene, and 2,4,6-triaminotoluene. For the sort of mixture of
explosive vapors in which we are most interested, 2,4,-DNT,
which can be reduced to 2,4-DAT, has been reported as the
INTRODUCTION
Thousands of forgotten buried landmines exist all over the
world. It is also common knowledge that plastic landmines and
improvised explosive devices (IEDs) are causing many
casualties in Iraq. Most of these mines contain TNT (2,4,6-
trinitrotoluene). TNT is also commonly the analyte of most
interest near munitions factories or storage areas. Some of the
techniques for detection assume metallic mines or require very
expensive instrumentation or are otherwise unreliable. The best
detection, especially for plastic mines, involves detection of
a mixture of vapors leaking from the mines. These are
1
major specie. Fluorescent fluorescamine derivatives of
,13
8
aromatic amines are known
examples for these aromatic amine derivatives. Recently,
a Siberian group
and we also include several
1
4,15
proposed a scheme for reducing nitro-
aromatics to aminoaromatics as a similar way of measuring
explosive vapors, but they used different reducing agents,
included no spectra of the aminoaromatics, and gave only one
example of a spectrum of a fluorescamine derivative. Our
proposed technique uses catalysis by metallic (Pd) nano-
impurities or degradation products such as 2,4-dinitrotoluene
1
(
DNT), which are more volatile than TNT by at least a couple
1
6–18
particles
synthesized in supercritical fluid carbon dioxide
of orders of magnitude. Other dinitrotoluenes or other aromatic
nitrated species such as TNT, nitrotoluene, or nitrobenzene
may be present, but can be handled similarly to the procedure
described in this paper. For this application very low levels of
vapors, lower than parts-per-trillion levels, must be measured.
References on the chemistry of explosives and related field
or water-in-hexane microemulsion and supported on carbon
nanotubes to efficiently reduce the nitroaromatics to amino-
aromatics, which are fluorescent themselves. UV-VIS fluores-
cence is the best and simplest ultratrace technique, being
capable with laser excitation as an option of even single
molecule detection. This fluorescence technique can be made
extremely sensitive (fluorescence derivatization with reagents
such as fluorescamine can make the derivatized amines even
more highly fluorescent) and the fluorescence can be shifted
further to the red where there is less background fluorescence.
In this paper we discuss the fluorescence spectra and chemistry
required as a framework for proving such a field sensor
2
–6
detection are included.
different techniques have also been proposed and perhaps the
Several electronic ‘‘noses’’ using
most promising is ‘‘FIDO’’ which uses fluorescent polymers
3
developed by T. Swager and incorporated in an electronic
nose by Nomadics Company (Stillwater, OK). In this case the
nitroaromatics such as DNT quench specially synthesized
fluorescent polymers. Fluorescence quenching may not be
specific enough for some forensic applications, and also the
fluorescent polymer may not be reliable and reproducible
enough, as well as being expensive to produce in bulk. In any
case, after a reduction step, direct fluorescence of the aromatic
amines or their derivatives provides an alternative approach,
feasible. Initial tests were made with a spectrofluorometer
TM
(Spex FluoroMax-3 ) having a 150 watt Xe arc lamp for
tunability. A small laser may be substituted for extra
sensitivity, once the desired wavelength (for the fluorescamine
derivative of DAT at about 391 nm) is determined. These
measurements are designed to show that direct detection of
aromatic amines or their derivatives can provide an alternative
approach for measuring explosives to fluorescence quenching
of nitroaromatic components of explosives.
Received 21 April 2006; accepted 8 June 2006.
Author to whom correspondence should be sent. E-mail: delyle@
uidaho.edu.
*
0
003-7028/06/6009-0958$2.00/0
9
58
Volume 60, Number 9, 2006
APPLIED SPECTROSCOPY
Ó 2006 Society for Applied Spectroscopy

EXPERIMENTAL
(silica gel with gypsum binder and fluorescent indicator) was
used.
Instrumentation. Standard fluorescence measurements of
the aminoaromatics and their derivatives with fluorescamine
were used to generate excitation and emission spectra using
a Spex (HORIBA Jobin Yvon) analytical grade spectrofluo-
Synthesis of Catalyst. Catalytic nanoparticles of Pd were
prepared in house for the reduction of nitro compounds such as
DNT and other aromatic compounds as discussed in more
16–18
detail in previous publications.
The water-in-hexane
TM
rometer (FluoroMax-3 ) with a 150 watt Xe arc lamp. The
microemulsions were prepared at 25 8C by mixing 0.1 M of
Na PdCl solution (0.864 mL) with 200 mL of n-hexane and
sodium dioctyl sulfosuccinate (AOT, 1.7782 g, 4 mmol), which
instrument was calibrated by measuring the spectrum of the
xenon lamp for the excitation monochromator (wavelength
calibration and intensity), measuring the water Raman (in-
tensity), and a Starna standard of ovalene in polymethylmeth-
acrylate (PMMA) for the emission monochromator
2
4
18
gave a W value of 12; W value is defined as the water to the
surfactant molar ratio ([water]/[surfactant]). Multi-walled
carbon nanotubes (MWCNTs) (1.0 g, 60–100 nm in diameter,
Nanostructure & Amorphous Materials Inc, Los Alamos, NM)
(
wavelength calibration and intensity). Wavelength accuracy
was better than 1 nm and the water Raman was initially above
manufacturer’s specifications in intensity, indicating that the
xenon lamp was more intense than expected. Excitation and
emission spectra were measured using several excitation and
detection wavelengths and several bandwidths (typically 5 or
were pretreated by sonication in 14 M HNO
refluxed for 12 h in a mixture of HNO (50 mL, 14 M) and
SO (98 % 50 mL). The treated MWCNTs (0.030 g) were
3
for 1 h and then
3
H
2
4
added to the microemulsion solution with continuous stirring.
Hydrogen gas at 1 atm was then bubbled through the solution
1
0 nm), all of which were narrow compared to the broad
2þ
for 30 min to reduce the Pd ions dissolved in the water core
of the microemulsion. It is known that hydrogen gas can cause
emission and excitation spectra for the aromatic amines or their
derivatives. Fluorescence references applicable to these types
2þ
reduction of Pd ions in aqueous solutions to their elemental
1
9,20
1
7
of samples are included for reference.
For the aromatic
state. After hydrogen reduction, the CNT-supported metal
nanoparticles would precipitate to the bottom of the flask
without stirring. The CNT-supported Pd nanoparticle catalyst
was collected from the flask, washed with methanol, and dried
in an oven for hydrogenation experiments.
amines, monochromator bandwidths were 5.0 nm, 1 nm
increments, and 0.1 second integration time. For the derivatized
aromatic amines, monochromator bandwidths of 10 nm, 1 nm
increments, and integration time of 0.1 second were used to
lower the detection limit. Starna fused silica 1 cm path length
cells were used with right-angle excitation for the solutions.
For a few solutions, absorption spectra were recorded for
comparison with the excitation spectra using a Spectral
Instruments, Inc (Tucson, AZ) charge-coupled device (CCD)
Formation and Derivatization of DAT starting with 2,4-
DNT. 2,4-Dinitrotoluene (0.0182 g, 0.1 mmol) in 10 mL of
ꢀ2
ethanol (10 M) with Pd/CNT (0.010 g) was stirred and then
hydrogen was bubbled through the solution for 2 min at room
temperature to make sure that DNT conversion to DAT was
complete; TLC (eluent; 25% ethyl acetate and 75% n-hexane)
was checked under UV light (254 nm) and then visualized by
1
array UV-VIS spectrophotometer with fiber optics. H nuclear
magnetic resonance (NMR) spectra were measured on a Bruker
(
AMX300, CDCl
3
) spectrometer at room temperature. Data
ninhydrin agent (red color). No starting materials were detected
1
were recorded as chemical shift values in ppm on the d scale,
multiplicity, integration, and coupling constants.
by TLC (DNT, R
NMR (300 MHz, CDCl
f
value ¼ 0.45) and the product was DAT: H
3
): d¼6.83 (s, 1H), d¼6.11 (d, 2H, J¼
Reagents. The reagents were all used as received and were
the highest grade obtainable. Only ethanol (200 proof USP or
from Aldrich for microbiology) was used as solvent for the
measurements reported here. Nitric acid (singly distilled from
Aldrich) and HPLC grade water, also from Aldrich, were used
for cell cleaning. Dinitrotoluene (2,4-DNT) was used from
Sigma-Aldrich for the reduction reaction described below to
produce diaminotoluene (2,4-DAT). 2,4-DAT was first ob-
tained from Sigma-Aldrich (98%) and later from AccuStandard
Inc (99.4%), 2,6-diaminotoluene (2,6-DAT) (99.7%) was also
obtained from AccuStandard. Triaminotoluene (2,4,6-TAT)
and aniline (aminobenzene) as likely typical minor components
of an explosive vapor mixture after reduction were also
obtained from AccuStandard Inc. O-Toluidine (97% or better)
and p-toluidine (97% or better), also likely minor components
of an explosive vapor mixture after reduction, were obtained
from Sigma. A fluorescence derivatization agent fluorescamine
8.4 Hz), d ¼ 3.50 (s, 4H), d ¼ 2.08 (s, 3H).
To increase the fluorescence yield, DAT was reacted with
ꢀ7
ꢀ
7
fluorescamine reagent: 10
M DAT and 4 3 10
M
fluorescamine (diluted down by a factor of 10 from the ethanol
ꢀ6
ꢀ6
solution of 10 M DAT and 4 3 10 M fluorescamine (4 3
.00278 mg), in 10 mL of ethanol and the solution was shaken
0
for 2 min (longer than needed) at room temperature (RT). The
solution was diluted by a factor of 10 after detection of
fluorescence peaks and then this dilution procedure was
ꢀ12
repeated to reach 10
M of the solution. These DAT
derivatives and other solutes being measured by fluorescence
were stored in amber glass vials and further protected from
ambient room light by aluminum foil to reduce risk of
photoreactivity.
RESULTS AND DISCUSSION
A summary of the data for excitation and emission spectral
peaks for the aromatic amines and their fluorescamine
derivatives are given in Tables I and II. Figures 1–8 show
the actual excitation and emission fluorescence data for the
corrected and solvent-subtracted spectra: (Fig. 1) aniline, (Fig.
2) p-toluidine and o-toluidine, (Fig. 3) 2,6-DAT and 2,4-DAT,
(Fig. 4) 2,4,6-TAT, (Fig. 5) aniline þ fluorescamine, (Fig. 6) p-
toluidine þ fluorescamine and o-toluidine þ fluorescamine,
(Fig. 7) 2,4-DAT þ fluorescamine at concentrations ranging
(
98%) was obtained from Molecular Probes, now a division of
Invitrogen Inc. Although fluorescamine is supposed to be
completely non-fluorescent in the absence of reactive amines,
so that a surplus would not matter, it was found to be weakly
fluorescent, presumably due to an unidentified impurity.
Finally, fluorescein from Fluka (now affiliated with Sigma-
Aldrich) was used to check the limit of detection of the
instrument in a basic (NaOH) solution. For thin-layer
chromatography (TLC) analysis, a Sigma-Aldrich TLC plate
ꢀ8
ꢀ12
M to 10
from 10
M, and (Fig. 8) 2,4,6-TAT þ
APPLIED SPECTROSCOPY 959

TABLE I. Fluorescence data for aromatic amines.
Excitation Emission
Chemical
(nm)
285
290
282
283
287
294
(nm)
341
346
339
343
336
347
Comments
ꢀ6
Aniline
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ6
p-Toluidine
o-Toluidine
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ6
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ6
2
2
2
,4-DAT (Diaminotoluene)
,6-DAT (Diaminotoluene)
,4,6- TAT (Triaminotoluene)
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ5
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ4
Ex & Em 10 M,
bandwidths 5.0 nm
fluorescamine. The aromatic amines shown in Figs. 1–4 are
likely fluorescent impurities in a mixture of explosive vapors
after reduction, with 2,4-DAT being the major component. The
aniline fluorescence spectrum was previously recorded in Ref.
9
. The fluorescent derivatives of these compounds with
1
3
fluorescamine were also measured (Figs. 5–8) and in the
case of 2,4-DAT as a function of several concentrations. DAT
þ fluorescamine led to a detection limit of 10ꢀ M when
exciting at 392 nm (Fig. 7). The measured spectra, as are
typical of aromatic phenols and amines, are broad (full-width at
half-maximum (FWHM) of 56 nm for the emission spectra of
ꢀ6
FIG. 1. Excitation and emission spectra of aniline 10 M. Bandwidth: 5 nm.
12
excite as far towards the blue as possible, where better and
cheaper sources exist.) The emission spectra of the fluoresc-
amine derivatives show a single peak in the range of 480 to 510
nm. The instrumental conditions are given in each figure
caption. For some compounds it was necessary to narrow the
bandwidths to avoid entering the nonlinear saturation region
for the photon-counting photomultiplier tube (PMT) detector,
which was above 2.5 million counts per second (cps). All
spectra were analyzed at least in duplicate; most were measured
in triplicate or higher. Spectral peak heights for 2,4-DAT þ
fluorescamine were linear over the concentration range of
ꢀ6
2
,4-DAT 10 M) and without vibrational structure. They all
have an excitation peak in the range of 280–300 nm and an
emission peak between 330 and 350 nm. Although these
spectra are similar, they can probably be distinguished by using
synchronous luminescence and pattern recognition techniques,
which we plan to discuss in follow-up publications. The
excitation and emission spectra of the fluorescamine deriva-
tives are red-shifted, broader (FWHM of 86 nm for the
ꢀ
9
emission spectrum of 2,4-DAT at 10 M), and more intense.
The emission intensity of a 10 M solution of DAT þ
ꢀ10
ꢀ12
1
0
to 10 M, but were not linear above this concentration
ꢀ6
fluorescamine derivative was found to be several orders of
ꢀ6
magnitude greater than a 10 M solution of DAT under the
same experimental conditions (same bandwidths). The excita-
tion peaks appear in a range of 380 to 410 nm. (For excitation
purposes in a field instrument or sensor, it is preferable to
TABLE II. Fluorescence data for derivatized aromatic amines.
Excitation Emission
Chemical
(nm)
398
410
383
391
(nm)
491
511
480
496
Comments
ꢀ8
Aniline þ Fluorescamine
p-Toluidine þ Fluorescamine
o-Toluidine þ Fluorescamine
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ6
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ6
Ex & Em 10 M,
bandwidths 5.0 nm
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
8
2
,4-DAT þ Fluorescamine
Ex & Em 10
Ex & Em 10
Ex & Em 10
Ex & Em 10
Ex & Em 10
M
M
9
497
483
478
463
10
11
12
M
M
M,
detected 492 nm,
bandwidths 10 nm
ꢀ6
2,4,6-TAT þ Fluorescamine
390
487
Ex & Em 10 M,
detected 491 nm,
bandwidths 10 nm
ꢀ6
FIG. 2. Excitation and emission spectra of (top) p-toluidine 10 M and
(bottom) o-Toluidine 10 M. Bandwidths: 5 nm.
ꢀ6
9
60
Volume 60, Number 9, 2006

ꢀ6
FIG. 3. Emission and excitation spectra of (top) 2,6-DAT 10 M and
FIG. 5. Emission and excitation spectra of Aniline þ Fluorescamine 10ꢀ M.
8
ꢀ6
(bottom) 2,4-DAT 10 M. Bandwidths: 5 nm.
Bandwidths: 10 nm.
even with solvent blank subtraction. In addition, a slight blue
shift in the emission peak of 2,4-DAT after ethanol subtraction
was observed as the concentration of DAT decreased probably
due to a slight under-correction of the solvent blank on a very
weak signal. 2,4-DAT was linear over the concentration range
excitation and emission peak positions, instrumental band-
widths, and concentrations (aniline, p-toluidine, o-toluidine,
2,4-DAT, 2,6-DAT, and 2,4,6-TAT). Table II gives fluores-
cence spectral data for derivatized aromatic amines including
excitation and emission peak positions, instrumental band-
widths, and concentrations (fluorescamine derivatives of
aniline, p-toluidine, o-toluidine, 2,4-DAT, 2,6-DAT, and
ꢀ
4
ꢀ6
to 10 M. Provided the data are sufficiently
reproducible, a calibration curve can be constructed even
of 10
outside the linear range. For six replications of 2,4-DAT þ
ꢀ6
2,4,6-TAT).
fluorescamine at 10 M, the relative standard deviation was
1
1%. Most of the spectra are normalized to the highest
fluorescence emission peak in the figure. Table I gives
fluorescence spectral data for aromatic amines including
Fluorescein standard (quantum yield approximately 0.9) in
ꢀ12
a basic solution was measured to give a detection limit of 10
M. This result was similar to what Spex had claimed for this
ꢀ4
FIG. 4. Emission and excitation spectra of 2,4,6-TAT 10 M. Bandwidths: 5
nm.
FIG. 6. Emission and excitation spectra of (top) p-Toluidine þ Fluorescamine
ꢀ6
ꢀ6
10 M and (bottom) o-ToluidineþFluorescamine 10 M. Bandwidths: 5 nm.
APPLIED SPECTROSCOPY
961

FIG. 8. Emission and excitation spectra of 2,4,6-TAT þ Fluorescamine 10ꢀ
6
ꢀ
8
FIG. 7. Emission and excitation spectra of 2,4-DAT þ Fluorescamine at 10
ꢀ11
M (multiplied
by 5), and 10
subtracted). Bandwidths: 10 nm.
ꢀ
9
ꢀ10
M. Bandwidths: 10 nm.
M, 10 M (multiplied by 3), 10
M (multiplied by 4), 10
ꢀ12
M (multiplied by 5) in consecutive order (solvent blank
ꢀ12
levels could be lowered further than the 10
M reported here,
or the aromatic amines could be measured directly without
need for derivatization, which would permit greater specificity,
if needed. Reaction kinetics could be altered by changing
experimental conditions to make the speed of the reactions
more competitive with other procedures. The present con-
ditions are optimized for 2,4-diaminotoluene reduced from 2,4-
dinitrotoluene by a novel method. 2,4-DNT is likely to be the
main component in explosive vapors leaking from a plastic
2
1
instrument for fluorescein in a recent application note.
TM
A
FluoroLog-3 instrument with a cooled PMT (which would
further reduce instrumental noise and scattered light) can detect
2
1
as low as 50 femtoMol of fluorescein.
Recent experiments in our laboratory, using small stainless
steel tubing with a diameter of 3 mm and a length of 13 cm
packed with carbon nanotube-supported palladium nanopar-
ticles for conversion of DNT to DAT in air with saturated DNT
vapors and hydrogen gas (bubbling through an ethanol solution
containing 10 M fluorescamine at 0.7 L per min at room
temperature), showed that 1 min of flow was enough to detect
the fluorescence emission of DAT–fluorescamine derivative in
the solution.
1
landmine containing TNT. An alternative approach might call
for deposition on a solid substrate (cold finger) or to measure
the DAT directly in the vapor phase. This method could be
modified to apply to other explosive mixtures or be applied to
explosives extracted from soil with supercritical fluid carbon
dioxide.
ꢀ6
ACKNOWLEDGMENTS
CONCLUSION
This work was partially supported by a grant from DOD-AFOSR (f49620-
3-1-0361). The authors (D.E. and C.M.W.) also want to acknowledge helpful
discussions with technical contacts from HORIBA Jobin Yvon (Spex
Fluorescence Division) (Ray Kaminski, James Mattheis, and Karen Rubelow-
sky). We also wish to acknowledge useful conversations with Ernesto Cespedes
of the Idaho National Laboratory.
The purpose of this preliminary paper was to show, using
a commercial instrument, the feasibility of developing
a fluorescence method or miniaturized field sensor for detection
of explosive vapors as an alternative to fluorescence quenching
methods currently being field tested. We have included
a corrected updated UV-VIS fluorescence spectral library
0
(
Figs. 1–4) for the aromatic amines in Table I, and for their
1. A. B. Crockett, T. F. Jenkins, H. D. Craig, and W. F. Sisk, ‘‘Overview of
On-Site Analytical Methods for Explosives in Soil’’, Special Report 98-4
fluorescamine derivatives (Figs. 5–8), which have lower
detection limits but broader spectra, in Table II. For 2,4-
diaminotoluene þ fluorescamine, we have measured the
fluorescence intensity as a function of concentration, with
a standard deviation of 11% at 10 M. The detection limit was
found to be 10
We also demonstrated the reduction of 2,4-DNT to the
corresponding DAT and its reaction with fluorescamine. This
was made by bubbling the DNT vapor through the Pd
nanocatalysts, suspended on multiwalled carbon nanotubes,
in the presence of hydrogen gas, and an ethanol solution of
fluorescamine. Evidently, by using more powerful laser sources
and more sensitive detectors with lower noise, the detection
(
NTIS, Springfield, VA, February, 1998).
2
. J. Akhavan, The Chemistry of Explosives (The Royal Society of Chemistry,
London, 1998).
3. T. Swager, in Electronic Noses and Sensors for the Detection of Explosives,
J. W. Gardner and J. Yinon, Eds. (Kluwer Academic Publishers, Boston,
2004).
ꢀ
6
ꢀ12
M under the conditions of the experiment.
4
. M. Fisher and J. Sikes, in Electronic Noses and Sensors for the Detection
of Explosives, J. W. Gardner and J. Yinon, Eds. (Kluwer Academic
Publishers, Boston, 2004).
5. J. Yinon, Forensic and Environmental Detection of Explosives (John
Wiley and Sons, New York, 1999).
6
. J. Yinon and S. Zitrin, Modern Methods and Application in Analysis of
Explosives (John Wiley and Sons, New York, 1993).
7
. C. M. Pace, J. R. Donnelly, J. L. Jeter, W. C. Brumley, and G. W.
Sovocool, J.—Assoc. Off. Anal. Chem., Interntl. 79, 777 (1996).
9
62
Volume 60, Number 9, 2006

8
. G. G. Guilbault, Ed., Practical Fluorescence (Marcel Dekker, New York,
990), 2nd ed., Revised and Expanded, pp. 99–100.
. I. B. Berlman, Handbook of Fluorescence Spectra of Aromatic Molecules
Academic Press, New York, 1965), pp. 59–60.
Razvitiya 11, 353 (Publisher Siberian Branch of Russian Academy of
Science) (2003).
5. V. A. Nadolinnyi, V. N. Ivanova, and S. N. Lunegov, Khimiya v
Interesakh Ustoichivogo Razvitiya, 11, 391 (2003).
1
1
9
(
1
1
1
1
2
6. M. Ohde, H. Ohde, and C. M. Wai, Langmuir 21, 1738 (2005).
7. B. Yoon, H. Kim, and C. M. Wai, Chem. Commun. 9, 1040 (2003).
8. B. Yoon and C. M. Wai, J. Am. Chem. Soc. 127, 17174 (2005).
9. C. A. Parker, Photoluminescence of Solutions (Elsevier, New York, 1968).
0. J. R. Lakowicz, Principles of Fluorescence Spectroscopy (Kluwer
Academic/Plenum Publishers, New York, 1999), 2nd ed., pp. 79–80.
1
0. J. T. Brownrigg, D. A. Busch, and L. P Giering, A Luminescence Survey of
Hazardous Materials, Report No. CG-D-53-79 (NTIS, May, 1979).
1. M. Kasha and R. Rawls, Photochem. Photobiol. 7, 561 (1968).
2. M. Kasha, Discuss. Faraday Soc. 9, 14 (1950).
1
1
1
3. R. P. Haugland, The Handbook, A Guide to Fluorescent Probes and
Labeling Technologies (Invitrogen Corp., Eugene, OR, 2005), 10th ed.
4. V. N. Ivanova and V. A. Nadolinnyi, Khimiya v Interesakh Ustoichivogo
21. Spex Fluorescence Group, Application Notes F-1, ‘‘Sub-Picomolar
Fluorescein’’ (HORIBA Jobin Yvon Inc, Edison, NJ, 2000).
1
APPLIED SPECTROSCOPY
963