Fluorescent discrimination between traces of chemical warfare agents and their mimics

Article abstract of DOI:10.1021/ja500710m

An array of fluorogenic probes is able to discriminate between nerve agents, sarin, soman, tabun, VX and their mimics, in water or organic solvent, by qualitative fluorescence patterns and quantitative multivariate analysis, thus making the system suitabl

Full text of DOI:10.1021/ja500710m

Communication
pubs.acs.org/JACS
Fluorescent Discrimination between Traces of Chemical Warfare
Agents and Their Mimics
Borja Díaz de Grenu,† Daniel Moreno,† Tomas Torroba,*,† Alexander Berg,‡ Johan Gunnars,‡
́
̃
Tobias Nilsson,‡ Rasmus Nyman,‡ Milton Persson,‡ Johannes Pettersson,‡ Ida Eklind,§
and Par Wasterby*,§
̈
̈
†Department of Chemistry, Faculty of Sciences, University of Burgos, 09001 Burgos, Spain
‡Department of Physics, Umea University, SE-90187 Umea, Sweden
̊
̊
§Department for CBRN Defense and Security, Swedish Defense Research Agency (FOI), SE-90182 Umea, Sweden
̊
S
* Supporting Information
Scheme 1. Synthesis of Fluorescent Probes and Their Action
ABSTRACT: An array of fluorogenic probes is able to
discriminate between nerve agents, sarin, soman, tabun,
VX and their mimics, in water or organic solvent, by
qualitative fluorescence patterns and quantitative multi-
variate analysis, thus making the system suitable for the in-
the-field detection of traces of chemical warfare agents as
well as to differentiate between the real nerve agents and
other related compounds.
erve agents are highly toxic volatile liquids that
Nirreversibly block the enzyme acetylcolinesterase in the
neuronal synapsis, thus disrupting nerve impulse transmission
and causing death through the paralysis of respiratory muscles.1
They are used as chemical warfare agents (CWA) for dirty war
in undeveloped countries, causing hundreds of victims,
although their use as chemical weapons is prohibited.2 Their
quick detection can be achieved by hand-held instruments that
are costly and prone to false positives3 so the availability of safe
and easy to use portable devices is most sought-after. More
importantly, the investigation of chemical weapons allegations
is a very slow process that implies unequivocal detection of
CWA residuals in water and organic samples,4 with the risk of
long delays in the environment of worrying war scenarios.5
Colorimetric6 or fluorimetric7 reactive dyes in solution or as
arrays,8 as well as supported in nanomaterials,9 have been used
for fast detection of CWA as good alternatives to classic
methods, but most of these methods are implemented for nerve
agents mimics, and so there is no clear proof that they will work
for real CWA.10 To complement the existing methodologies,
we have developed a series of new highly solvatochromic
fluorescent indicators for phosphorylating reagents capable of
developing large differences in fluorescence. In this paper, we
report our findings upon the selective fluorescent discrim-
ination of real nerve agents from their mimics.
We have previously prepared some charge-transfer fluoro-
genic probes, bearing conjugated donor and acceptor groups in
their structure, that were useful for the detection of significant
analytes.11 For our current purpose we have designed new
fluorescent probes (Scheme 1).
In this case, they have a secondary donor group that was not
involved in the charge-transfer process. Thus, the Suzuki
Received: January 22, 2014
Published: March 5, 2014
© 2014 American Chemical Society
4125
dx.doi.org/10.1021/ja500710m | J. Am. Chem. Soc. 2014, 136, 4125−4128

Journal of the American Chemical Society
Communication
reaction of aryl boronates 1a−c and 5-bromoindanone 2
catalyzed by Pd(PPh3)4 in tetrahydrofuran/water in the
presence of Na2CO3 gave arylindanones 3a−c in 85−95%
yields. Knoevenagel reaction of 3a−c and malononitrile in the
presence of DABCO in toluene at reflux for 24 h gave
arylindanes 4a−c in 55−68% yields. N-Boc deprotection with
trifluoroacetic acid for 15 min from 3a−c and 4a−c gave the
unprotected amine derivatives BD03, BD13, BD68, BD69,
BD77 and BD78, in which the initial fluorescence of 3a−c and
4a−c is quenched in some extension by a photoinduced
electron transfer from the free amine group. Subsequent
acylation or phosphorylation of the amine group should
therefore increase fluorescence of these compounds, thus
making these compounds suitable for phosphorylating agents
detection (Scheme 1). Fluorescence of these compounds can
be also affected by protic acids, therefore we added to the series
a fluorogenic dye, BD00,12 which is not fluorescent but
develops a blue fluorescence in the presence of common protic
acids. In this way, false positives are prevented. We next tested
10−4 M solutions of the seven fluorescent probes in
dimethylsulfoxide (DMSO) or acetonitrile (MeCN) with 1
equiv of 5 × 10−3 M solutions of nerve agent simulants (DCP,
DFP, CNP) (Scheme 1) and phosgene13 (Cl2CO) in MeCN or
water and recorded all changes that the fluorescent probes
underwent with every analyte under a common TLC-UV light,
λ = 366 nm, by qualitative (photographs) and quantitative
in water mixed with probe DM13 solved in DMSO was called
GB_W13D. For the mixtures with only probes the name begins
with NaN. To photograph the samples they were placed under
a 366 nm UV-lamp in a dark room. A color reference sheet
illuminated with white light was placed nearby (Figure S89).
Copies of the RAW-files were edited, all in the same way (batch
process), before being converted to JPG for extraction of the
colors as RGB-values. Both the colors from the edited images
and from the original images were analyzed. As an example, a
photograph of Soman samples is seen in Figure 1.
Figure 1. Samples contained Soman in MeCN mixed with each of the
seven probes in DMSO. From left to right the samples contained
probes BD00 to BD78.
Since there were three images of every set of seven samples,
the mean values in R, G, and B had to be computed. A table of
these colors in the form of colored squares was then created as
seen in Figure 2. Looking at the tables of observed colors it was
abs
fluo
measurements such as initial and final λmax and λmax
,
variations in the relative intensity of fluorescence and kinetics of
processes. The qualitative measurements gave clear and distinct
fingerprints of every nerve agent mimic used for testing the
probes, undoubtedly discriminating between them. The
quantitative measurements were subjected to hierarchical
cluster analysis (HCA).8 HCA dendrogram obtained from
fluorescent measures showed a clear clustering for all the nerve
agent simulants, blank and phosgene, giving a good separation
of every analyte (Figure S65b). Absorbance or mixed data from
absorbance and fluorescence afforded a poor separation
between some analytes (Figures S64 and S65a), therefore
establishing that discrimination between analytes is best
obtained by fluorescence measurements. Likewise, principal
components analysis (PCA)14 of the same data afforded good
discrimination between each one of the CWA mimics as well as
phosgene (Figure S66), therefore probing that the array of
fluorescent dyes is able to discriminate between closely related
phosphorylating or acylating reagents by both their fingerprints,
HCA or PCA. The next step was testing the system with real
nerve agents, but because of the extreme toxicity we performed
the tests at the laboratories of the FOI CBRN Defense and
Figure 2. Observable colors in 366 nm excitation light. CWA: Sarin
(GB), Soman (GD), Tabun (GA), VX (VX). CWA-simulants: CNP
and DCP. CWA/simulants solvent: water (W) and acetonitrile (A).
Probes: BD00, BD03, BD13, BD68, BD69, BD77, BD78. Probe
solvents: dimethylsulfoxide (D) and acetonitrile (A).
clear that several probes could be used to guarantee the absence
of Sarin, Soman, Tabun, and VX. For some choices of solvents
there was a possibility to make the distinction between Soman
and the other CWA. If a sample with probe BD69 in
acetonitrile fluoresced very weakly (as in probe BD69 in
acetonitrile, known as NaN A in the table of observed colors),
then the risk of there being Sarin, Soman, Tabun, or VX in the
samples is low since the corresponding CWA samples, with
nerve agent in water and probe in acetonitrile, all fluoresced
green. Probe BD78 acts in a similar way, but here the CWA-
samples fluoresced in orange, while for probe BD03 it is the
other way around. The NaN A sample with probe BD03
fluoresced in bright green, while there is barely any fluorescence
from the corresponding CWA samples. Probe BD77 also gave
valuable information but in a different way. For this probe both
the probe and the CWA samples fluoresced clearly, but the
Security (Umea, Sweden), where handling of nerve agents was
̊
performed under appropriate conditions. Again, the seven
different fluorescent probes were mixed with a series of nerve
agents, Soman, Sarin, Tabun, and VX and chemically similar
substances diethylchlorophosphate (DCP) and diethylcyano-
phosphonate (CNP), in the same conditions used for CWA
mimics. The acquired samples of mixtures were then subjected
to light (300−500 nm) in which they fluoresced with different
colors. Light intensities were registered with a spectrofluor-
ometer and photographs were taken for a chart of visible colors
of all the test samples. The probes and CWA were solved in
two different solvents, DMSO and MeCN for the probes, and
MeCN and water for the CWA/CWA-simulants. The probes
were also tested without CWA or simulant. The acquired
mixtures were named as in the following example: Sarin solved
4126
dx.doi.org/10.1021/ja500710m | J. Am. Chem. Soc. 2014, 136, 4125−4128

Journal of the American Chemical Society
Communication
probe samples did so in blue, while the CWA samples all
fluoresced in green. It was probe BD78 that indicated that there
was a possibility to distinguish Soman from the rest of the
CWA. It is visible in the Figure 2 of observed colors that the
mixture of Soman in MeCN and probe in DMSO fluoresced in
a clear orange color, while the rest of the CWA samples with
the same solvents fluoresced with weak obscure blue color. A
table with the colors from the unedited images can also be
found in Figure S90. The colors from preliminary experiments
with only simulants have been included in a similar table in
Figure S92. For quantitative measurements we used a calibrated
spectrofluorometer. In the analysis of the spectral data a
multivariate data analysis with Simca15 software was used. To
analyze the data we used a couple of approaches. Some of the
basic analysis was made just by looking at the plots of the
spectroscopy data. We were able to see that some of the
mixtures just gave fluctuations in the data, while other gave
clear tops. We found that the probes BD03, BD68, and BD77
were the probes that gave the highest number of clear tops,
while the other only gave a few clear tops. In Figure 3 six
Figure 4. (Upper) A plot of the first two principal axes when running a
multivariate analysis over the agent-probe combinations. In the plot we
see a clear separation of the CWA (G and V) and the simulants (C).
In this plot only the probes BD03, BD68, and BD77 were used, i.e.,
the probes that did not have a large tendency to fluctuate. (Down) A
plot of the first two principal axes when running a multivariate analysis
over the agent-solvent combinations. In the plot we can see a clear
separation of the agents solved in water and those solved in
acetonitrile.
Figure 3. Plots of the spectrofluorometer data for CWA Soman (GD)
solved in MeCN with the probes BD03 to BD78 in DMSO. In the
plots we see how probe BD13 (top middle) causes fluctuations over
the whole spectra while the probes BD03 and BD77 (top left and
bottom middle) have very clear spectra. One of the few tops of probe
BD78 (bottom right) can be seen and it is also this clear orange color
that makes Soman stand out from the other CWA in the photographs.
each of the four solvent combinations. Before the PCA analysis
was run all data were transformed logarithmically and grouped
in blocks of λem/λex, fl_max, and area50 before unit variance
was run. From the load vectors for the analysis of the agent-
probe combinations (Figure S72) we could see how the areas
and florescence were the main parameters for the second
component. In addition, the simulants generally had a bit
higher fluorescence and for some a bit smaller areas, therefore
they tended toward the lower values on the second component.
When studying the combinations of agent and solvents against
probes we were able to see a clear separation between the
agents that were solved in water and those that were solved in
acetonitrile. This can clearly be seen in Figure 4, the main
reason behind this separation seems to be that the intensity of
probe BD68 becomes higher for those agents solved in water
(Figure S88).
In summary, we have synthesized a new series of fluorogenic
probes that are able to discriminate between traces of CWA and
their mimics, in water or organic solvent. Discrimination is
achieved by means of the different fluorogenic response
triggered by CWA or their mimics on the fluorogenic probes
in different solvent combinations of CWA and probes.
The different response given by the series of fluorogenic
probes is charted as a fingerprint of the fluorescent response of
every CWA/probe/solvent combination under a common 366
nm UV light, thus permitting a fast visual differentiation
between CWA and their mimics. More accurate discrimination
spectra for Soman are plotted, showing all types of spectra that
occurred, along with the distinct spectra for probe BD78 that
distinguished Soman from the other probes in the spectra as
well as in the photographs.
By use of multivariate data analysis we found that we were
able to detect in which solvent the CWA were solved. We were
also able to see a clear difference between the probes that gave
clear tops in the spectrofluorometer data, and those that did
not. In the analysis of the spectrofluorometer data the measured
values were emission (λem) and excitation (λex) wavelengths,
and intensity of the maximum (fl_max) in each of the produced
two-dimensional spectra. We also calculated the area in the
spectra with intensities of 50% and 75% or more of the
maximum (area50 and area75, respectively). We were able to
see a clear difference between the simulants and the CWA
when performing a multivariate analysis on agent-probe
combinations (Figure 4). After our analysis we can conclude
that there is a large probability that the probes are able to detect
the most important CWA from their mimics. In the analysis
only probes BD03, BD68, and BD77 were used to avoid the
fluctuations, as variables we used area50, λex, λem, and fl_max for
4127
dx.doi.org/10.1021/ja500710m | J. Am. Chem. Soc. 2014, 136, 4125−4128

Journal of the American Chemical Society
Communication
Minami, T.; Mosca, L.; Lynch, V. M.; Zyryanov, G. V.; Anzenbacher,
P. Chem. Sci. 2013, 4, 3617.
(5) Organisation for the Prohibition of Chemical Weapons, United
Nations Joint Mission for the Elimination of the Chemical Weapons
Programme of Syrian Arab Republic, http://opcw.unmissions.org/,
accessed November 8, 2013.
is achieved by multivariate analysis by using quantitative
measurements in fluorescence spectroscopy. In this way we
have obtained a complete differentiation between CWA and
their mimics, so the system is suitable for the accurate in-the-
field detection of traces of CWA. We have seen that the
response given by CWA mimics is very different to the
response given by the real CWA, because of the slightly
different chemical functionality of CWA and their mimics.
Since most of the chromogenic and fluorogenic probes hitherto
studied for the detection of CWA are based in the study of the
response given by their mimics, there is no guarantee that
previously known probes for CWA mimics will work with real
CWA samples. Our work clearly shows that the response can be
very different. In addition, the synthesis of the reported
fluorogenic probes is simple and straightforward, therefore
these fluorescent probes are suitable for the development of
upcoming practical methodology.
(6) Review: (a) Royo, S.; Martínez-Man
́
ez, R.; Sancenon
́
, F.;
Costero, A. M.; Parra, M.; Gil, S. Chem. Commun. 2007, 4839. E.g.:
(b) Gotor, R.; Costero, A. M.; Gil, S.; Parra, M.; Martínez-Manez, R.;
̃
́
̃
Sancenon
M.; Parra, M.; Gil, S.; Martínez-Man
́
, F. Chem.Eur. J. 2011, 17, 11994. (c) Royo, S.; Costero, A.
́
ez, R.; Sancenon
́
, F. Chem.Eur.
̃
J. 2011, 17, 6931. (d) Nourmohammadian, F.; Wu, T.; Branda, N. R.
Chem. Commun. 2011, 47, 10954. (e) Gotor, R.; Royo, S.; Costero, A.
M.; Parra, M.; Gil, S.; Martínez-Man
2012, 68, 8612. (f) Chulvi, K.; Gavina, P.; Costero, A. M.; Gil, S.;
́
ez, R.; Sancenon
́
, F. Tetrahedron
̃
̃
Parra, M.; Gotor, R.; Royo, S.; Martínez-Man
́
ez, R.; Sancenon
́
, F.;
Vivancos, J.-L. Chem. Commun. 2012, 48, 10105. (g) Barba-Bon, A.;
Costero, A. M.; Parra, M.; Gil, S.; Martínez-Manez, R.; Sancenon, F.;
̃
́
́
̃
Gale, P. A.; Hiscock, J. R. Chem.Eur. J. 2013, 19, 1586. (h) Walton,
I.; Davis, M.; Munro, L.; Catalano, V. J.; Cragg, P. J.; Huggins, M. T.;
Wallace, K. J. Org. Lett. 2012, 14, 2686.
ASSOCIATED CONTENT
* Supporting Information
Experimental details and characterization data. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
(7) Review: (a) Burnworth, M.; Rowan, S. J.; Weder, C. Chem.Eur.
S
J. 2007, 13, 7828. E.g.: (b) Costero, A. M.; Parra, M.; Gil, S.; Gotor,
́ ́
R.; Mancini, P. M. E.; Martínez-Manez, R.; Sancenon, F.; Royo, S.
̃
Chem. Asian J. 2010, 5, 1573. (c) Han, S.; Xue, Z.; Wang, Z.; Wen, T.
B. Chem. Commun. 2010, 46, 8413. (d) Wu, X.; Wu, Z.; Han, S. Chem.
Commun. 2011, 47, 11468. (e) Wild, A.; Winter, A.; Hager, M. D.;
Schubert, U. S. Chem. Commun. 2012, 48, 964. (f) Ajami, D.; Rebek, J.,
Jr. Org. Biomol. Chem. 2013, 11, 3936.
AUTHOR INFORMATION
Corresponding Authors
ttorroba@ubu.es
■
(8) Reviews: (a) Diehl, K. L.; Anslyn, E. V. Chem. Soc. Rev. 2013, 42,
8596. (b) Askim, J. R.; Mahmoudi, M.; Suslick, K. S. Chem. Soc. Rev.
2013, 42, 8649.
(9) (a) Rathfon, J. M.; Al-Badri, Z. M.; Shunmugam, R.; Berry, S. M.;
Pabba, S.; Keynton, R. S.; Cohn, R. W.; Tew, G. N. Adv. Funct. Mater.
par.wasterby@foi.se
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
2009, 19, 689. (b) Climent, E.; Martí, A.; Royo, S.; Martínez-Manez,
́
̃
■
R.; Marcos, M. D.; Sancenon
Parra, M. Angew. Chem., Int. Ed. 2010, 49, 5945. (c) Candel, I.;
Bernardos, A.; Climent, E.; Marcos, M. D.; Martínez-Manez, R.;
́
, F.; Soto, J.; Costero, A. M.; Gil, S.;
We gratefully acknowledge financial support from the Minister-
́
io de Economia y Competitividad, Spain (Project CTQ2012-
́
̃
́
́ ́
31611), Junta de Castilla y Leon, Consejeria de Educacion y
Sancenon
2011, 47, 8313. (d) Coronado-Puchau, M.; Saa, L.; Grzelczak, M.;
Pavlov, V.; Liz-Marzan, L. M. Nano Today 2013, 8, 461.
́
, F.; Soto, J.; Costero, A.; Gil, S.; Parra, M. Chem. Commun.
Cultura y Fondo Social Europeo (Project BU246A12-1), the
European Commission, Seventh Framework Programme
(Project SNIFFER FP7-SEC-2012-312411) and the Swedish
Ministry of Defence (no. A403913). This paper is dedicated to
the memory of Dr. Stefano Marcaccini who passed away
October 1, 2012.
́
(10) Real CWA sensing: (a) Sambrook, M. R.; Hiscock, J. R.; Cook,
A.; Green, A. C.; Holden, I.; Vincent, J. C.; Gale, P. A. Chem. Commun.
2012, 48, 5605. (b) Ordronneau, L.; Carella, A.; Pohanka, M.;
Simonato, J.-P. Chem. Commun. 2013, 49, 8946. (c) Pardasani, D.;
Tak, V.; Purohit, A. K.; Dubey, D. K. Analyst 2012, 137, 5648.
(11) (a) Gom
́
ez, T.; Moreno, D.; Díaz de Grenu, B.; Fernandez, A.
́
̃
REFERENCES
■
C.; Rodríguez, T.; Rojo, J.; Cuevas, J. V.; Torroba, T. Chem. Asian J.
2013, 8, 1271. (b) Moreno, D.; Díaz de Grenu, B.; García, B.; Ibeas, S.;
Torroba, T. Chem. Commun. 2012, 48, 2994. (c) Moreno, D.; Cuevas,
J. V.; García-Herbosa, G.; Torroba, T. Chem. Commun. 2011, 47, 3183.
́
(12) Ballesteros, E.; Moreno, D.; Gomez, T.; Rodríguez, T.; Rojo, J.;
García-Valverde, M.; Torroba, T. Org. Lett. 2009, 11, 1269.
(13) Phosgene sensors: (a) Kundu, P.; Hwang, K. C. Anal. Chem.
2012, 84, 4594. (b) Wu, X.; Wu, Z.; Yang, Y.; Han, S. Chem. Commun.
2012, 48, 1895. (c) Feng, D.; Zhang, Y.; Shi, W.; Li, X.; Ma, H. Chem.
Commun. 2010, 46, 9203. (d) Ziessel, R.; Ulrich, G.; Harriman, A.;
Alamiry, M. A. H.; Stewart, B.; Retailleau, P. Chem.Eur. J. 2009, 15,
1359.
̌
(14) Reviews: (a) Anzenbacher, P., Jr.; Lubal, P.; Bucek, P.; Palacios,
M. A.; Kozelkova, M. E. Chem. Soc. Rev. 2010, 39, 3954. (b) Stewart,
S.; Ivy, M. A.; Anslyn, E. V. Chem. Soc. Rev. 2013, 43, 70.
(1) Reviews: (a) Russell, A. J.; Berberich, J. A.; Drevon, G. F.;
Koepsel, R. R. Annu. Rev. Biomed. Eng. 2003, 5, 1. (b) Eubanks, L. M.;
Dickerson, T. J.; Janda, K. D. Chem. Soc. Rev. 2007, 36, 458.
(c) Ganesan, K.; Raza, S. K.; Vijayaraghavan, R. J. Pharm. BioAllied Sci.
2010, 2, 166. (d) Sambrook, M. R.; Notman, S. Chem. Soc. Rev. 2013,
42, 9251. See also: (e) Chemical Warfare Agents: Toxicology and
Treatment, 2nd ed.; Marrs, T. C., Maynard, R. L., Sidell, F., Eds.; John
Wiley & Sons: Chichester, 2007.
(2) Convention on the prohibition of the development, production,
stockpiling and use of chemical weapons and on their destruction,
http://www.opcw.org/chemical-weapons-convention/download-the-
cwc/, accessed November 8, 2013.
(3) Sferopoulos, R. A Review of Chemical Warfare Agent Detector
Technologies & Commercial-Off-The-Shelf Items; DSTO Defence
Science and Technology Organisation: Victoria, Australia, 2009.
(4) Review: (a) Aleksenko, S. S. J. Anal. Chem. 2012, 67, 82. E.g.:
(b) Cordell, R. L.; Willis, K. A.; Wyche, K. P.; Blake, R. S.; Ellis, A. M.;
Monks, P. S. Anal. Chem. 2007, 79, 8359. (c) Kolakowski, B. M.;
D’Agostino, P. A.; Chenier, C.; Mester, Z. Anal. Chem. 2007, 79, 8257.
(d) Nilles, J. M.; Connell, T. R.; Durst, H. D. Anal. Chem. 2009, 81,
6744. (e) Knaack, J. S.; Zhou, Y.; Magnuson, M.; Silvestri, E.; Johnson,
R. C. Anal. Chem. 2013, 85, 2611. (f) Esipenko, N. A.; Koutnik, P.;
̃
(15) SIMCA, version 13, Multivariate Analysis Software; Umetrics
AB: Umea, Sweden, 2012.
̊
4128
dx.doi.org/10.1021/ja500710m | J. Am. Chem. Soc. 2014, 136, 4125−4128