Array-based detection of isomeric and analogous analytes employing synthetically modified fluorophore attached β-cyclodextrin derivatives

Article abstract of DOI:10.1039/c7nj02968c

Reported herein is a sensitive and selective array-based sensing strategy based on differential interactions with three supramolecular cyclodextrin-fluorophore sensors. Each interaction results in a distinct fluorescence modulation response, and linear discriminant analyses of these responses results in 100% successful classification of three classes of isomeric analytes and two classes of analogous analytes. Calculated limits of detection for this system are at or near literature-reported levels of concern.

Full text of DOI:10.1039/c7nj02968c

NJC
View Article Online
View Journal | View Issue
PAPER
Array-based detection of isomeric and analogous
analytes employing synthetically modified
fluorophore attached b-cyclodextrin derivatives†
Cite this: New J. Chem., 2017,
41, 14431
Sauradip Chaudhuri, Dana J. DiScenza, Benjamin Smith, Reid Yocum and
Mindy Levine
*
Reported herein is
a sensitive and selective array-based sensing strategy based on differential
Received 10th August 2017,
Accepted 23rd October 2017
interactions with three supramolecular cyclodextrin–fluorophore sensors. Each interaction results in a
distinct fluorescence modulation response, and linear discriminant analyses of these responses results
in 100% successful classification of three classes of isomeric analytes and two classes of analogous
analytes. Calculated limits of detection for this system are at or near literature-reported levels of concern.
DOI: 10.1039/c7nj02968c
rsc.li/njc
non-aromatic compounds.13 This kind of situation requires the
development of a sensing system which is rapid, simple, and
Introduction
The selective detection and accurate quantification of structurally efficient in classifying a broad range of persistent organic
similar analytes is a major challenge for scientists, as structurally pollutants (POPs).14
similar analytes often have widely disparate toxicities.1 The most
Our group has previously reported the use of b-cyclodextrin
common strategy is to use mass spectrometry methods, such and g-cyclodextrin in array-based detection systems for the sensing
as liquid chromatography-mass spectrometry (LC-MS)2 or gas of a wide variety of environmental toxicants and POPs.15 The
chromatography-mass spectrometry (GC-MS).3 However, there are sensing strategy is based on cyclodextrin promoted analyte-to-
significant drawbacks associated with this approach, including the fluorophore energy transfer as well as cyclodextrin-promoted,
costs and time necessary to conduct such analyses,4 which limits analyte-induced fluorescence modulation. In the fluorescence
the ability to conduct high throughput assays.5
modulation systems, the fluorophore was added to the cyclo-
An alternate strategy is to use array-based sensing systems, dextrin solution prior to analyte addition, which can result in
which have recently gained in popularity.6 This approach relies on fluorophore–cyclodextrin binding that reduces the cyclodextrin’s
the development of a chemical signature for each analyte based on ability to bind the target analyte. As such, introduction of the
analyte-specific interactions with a sensor series. Array-based sensing analyte to the fluorophore–cyclodextrin solution requires the
systems can be combined with supramolecular sensors, which rely analyte–cyclodextrin association constants to be higher than
on differential non-covalent interactions of analytes with supra- those of the fluorophore–cyclodextrin (Fig. 1A), or it requires
molecular hosts, including cyclodextrins,7 fluorescent polymers,8 the formation of higher order association complexes between the
molecularly imprinted polymers,9 and metal–organic frameworks
(MOFs).10
Although supramolecular array-based systems overcome many
challenges associated with mass-spectrometry based detection
methods, the analyte scope explored in most of these reports
have been limited to aromatic small molecules.11 In a real-world
contaminated environment, the nature of the various pollutants
is highly complex,12 and includes mixtures of aromatic and
Department of Chemistry, University of Rhode Island, 140 Flagg Road, Kingston,
RI 02881, USA. E-mail: mlevine@chm.uri.edu, mindy.levine@gmail.com;
Fax: +1-401-874-5072; Tel: +1-401-874-4243
† Electronic supplementary information (ESI) available: Detailed synthetic procedures;
detailed procedures for fluorescence modulation, limit of detection, and array analysis
Fig. 1 Schematic illustration of this work compared to previously
published work.
experiments; summary figures and tables of all data; copies of NMR spectra of all new
compounds. See DOI: 10.1039/c7nj02968c
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2017
New J. Chem., 2017, 41, 14431--14437 | 14431

View Article Online
Paper
NJC
analyte, cyclodextrin and fluorophore (Fig. 1B). Such higher order of S1, S2, or S3 solutions (5 mM in DMSO) and 2 mL of deionized
association complexation is probable only for g-cyclodextrin.16
water were combined. The solution was excited at 320 nm, and
Herein, we report the development of an array-based detection the fluorescence emission spectra were recorded starting at
system using fluorophore-functionalized perbenzylated b-cyclo- 330 nm. Six repeat measurements were taken.
dextrin sensors, which enables binary complex formation between
Next, 2 mL of analyte (1 mg mLÀ1 in THF) was added, and
the functionalized cyclodextrin and the target analyte (Fig. 1C). again the solution was excited at the fluorophore’s excitation
Each sensor is selective, meaning the array is able to distinguish wavelength, and the fluorescence emission spectra were
three classes of isomeric analytes and two classes of structurally recorded. Six repeat measurements were taken. This step was
similar analytes, with 100% classification accuracy. High sensi- repeated for 4 mL of analyte, 6 mL of analyte, 8 mL of analyte,
tivity is demonstrated as well, with limits of detection appro- 10 mL of analyte, 12 mL of analyte, 14 mL of analyte, 16 mL of
aching or surpassing literature-reported levels of concern. Finally, analyte, 18 mL of analyte, and 20 mL of analyte.
preliminary efforts at using this system for the accurate identifi-
cation of binary analyte mixtures are also reported.
All of the fluorescence emission spectra were integrated
vs. wavenumber on the X-axis, and calibration curves were
generated. The curves plotted the analyte concentration in
mM on the X-axis, and the fluorescence modulation ratio on
the Y-axis. The curve was fitted to a straight line and the
equation of the line was determined.
Experimental section
Materials and methods
The limit of detection is defined according to eqn (2):
All the reagents were obtained from Sigma Aldrich or Fisher
Scientific and used without further purification, unless otherwise
noted. b-Cyclodextrin was dried in the oven prior to use. Reagent
grade solvents (99.9% purity) were used for the synthetic
reactions.
LOD = 3(SDblank)/m
(2)
where SDblank is the standard deviation of the blank sample and
m is the slope of the calibration curve.
Fluorescence modulation experiments
Results and discussion
Fluorescence emission spectra were obtained using a Shimadzu
RF-5301PC spectrophotofluorimeter with 3 nm excitation and
3 nm emission slit widths. 0.5 mL of S1, S2, or S3 solutions
(5 mM in DMSO) and 2 mL of deionized water were combined in
a quartz cuvette. The solution was excited at 320 nm, and the
fluorescence emission spectra were recorded.
The fluorescence emission spectra were integrated vs. wave-
number on the X-axis, and the fluorescence modulation was
measured as the ratio of the integrated emission of the fluoro-
phore in the presence of the analyte to integrated emission of
the fluorophore in the absence of the analyte (eqn (1)):
We employed a series of three cyclodextrin-based supramolecular
sensors (Fig. 2) for the detection of a broad variety of small
molecule analytes (Fig. 3). In these sensors, the perbenzylated
b-cyclodextrin cavity acts as the receptor domain, and the
attached fluorophore units act as the transducers, which are
responsible for fluorescence-based responses to changes in
their environment in the presence of the target analyte. The
covalent attachment strategy used in sensors S2 and S3, with
one and two degrees of functionalization on the primary rim,
respectively, ensures the close proximity of the fluorophore
Fluorescence modulation = Flanalyte/Flblank
(1)
where Flanalyte is the integrated fluorescence emission of the
fluorophore in the presence of 10 mL of analyte (1 mg mLÀ1 in
THF), and Flblank is the integrated fluorescence emission of the
fluorophore in the absence of the analyte.
Array generation experiments
Array analysis was performed using SYSTAT 13 statistical
computing software with the following settings:
(a) Classical discriminant analysis
(b) Grouping variable: analytes
(c) Predictors: S1, S2, and S3
(d) Long-range statistics: mahal
Limit of detection experiments
The limit of detection (LOD) is defined as the lowest concen-
tration of analyte at which a signal can be detected. To determine
this value, the following steps were performed for each
cyclodextrin-analyte combination. In a quartz cuvette, 0.5 mL
Fig. 2 Structures of sensors S1–S3.
14432 | New J. Chem., 2017, 41, 14431--14437
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2017

View Article Online
NJC
Paper
Fig. 4 Fluorescence emission spectra of supramolecular hosts S1–S3
(1 mM) (inset shows the fluorescence of S2 and S3 in more detail) in
80 : 20 water–DMSO solution. (lex = 320 nm; 3 nm excitation slit width;
3 nm emission slit width).
Fig. 3 Structures of small molecule analytes 5–26.
units to the cyclodextrin receptor cavity, thereby facilitating
productive fluorophore–analyte interactions. In contrast, sensor S1
is a non-covalent combination of the perbenzylated b-cyclodextrin
and fluorophore 4 (1 : 1 molar ratio), and is included to enable a
direct determination of the benefits of covalent attachment in
sensor design.
The synthesis of supramolecular hosts S2 and S3 is shown in
Scheme 1. Perbenzylated b-cyclodextrin was obtained from the
reaction of b-cyclodextrin with excess benzyl chloride.17 Regio-
selective debenzylation of the primary rim was effected by
treating the perbenzylated b-cyclodextrin with DIBAL-H.18 This
was followed by esterification19 with the acid derivative of
fluorophore 4, yielding mono- and di-functionalized sensors
S2 and S3. Compounds S2 and S3 were fully characterized
by 1H NMR, 13C NMR, MALDI-TOF mass spectrometry, and
UV-visible and fluorescence spectroscopy.
The sensitivity of the fluorescence emission responses of
sensors S1–S3 to solvent composition were investigated, with
the goal of ensuring full dissolution of the sensor while
enabling strong binding of analytes in the cyclodextrin (optimal
in aqueous environments). These competing considerations led
us to choose an 80 : 20 water–DMSO mixture as the optimal
sensing solvent. Of note, covalent attachment of the fluoro-
phores in S2 and S3 led to a reduction of the fluorescence
emission compared to the free fluorophore in S1 (Fig. 4). This
decrease is in agreement with literature precedent in analogous
systems, and occurs as a result of increased non-radiative decay
pathways that are available through covalent attachment to
a highly flexible macromolecule. That decrease is offset by the
markedly improved fluorescence modulation results in the
presence of various analytes.20
The choice of perbenzylated b-cyclodextrin as a receptor is
due to the strong binding of organic guest molecules in the
extended hydrophobic cavity. A comparison of association
constants of analyte 5 revealed a 1000-fold increase in the
binding constant with perbenzylated b-cyclodextrin compared
to b-cyclodextrin, with further increases in the fluorophore-
functionalized cyclodextrins S2 and S3 (Table 1). These binding
constants are orders of magnitude higher than the highest
literature-reported binding constants for analyte 5 in b-cyclodextrin
(Ka = 50–215 MÀ1).21 Higher association constants for analyte-
sensor binding are known to lead to improved sensor
performance,22 a phenomenon that is also borne out in this
system (vide infra).
Similarly, in this case, strong binding of analytes 5–8 in
hosts S1–S3 induced marked changes in the resulting fluores-
cence emission due to proximity-induced interactions between
the analyte and the fluorophore. These changes were quantified
according to eqn (1).
The sensor S1 shows a fluorescence modulation value close
to 1.00 for all the tested analytes, indicating minimal to no effect
on the fluorescence emission of the fluorophore with the intro-
duction of the analyte. In contrast to this, fluorescence modulation
values measured for sensors S2 and S3 are significantly different
from that of S1, and display widespread variability between
Table 1 Association constants of analyte 5 in perbenzylated b-cyclodextrin,
S2, and S3a
Host
Association constant (MÀ1
)
Perbenzylated b-cyclodextrin
S2
S3
3.6 (0.1) Â 104
4.8 (0.5) Â 104
24.9 (0.5) Â 104
a
Association constants calculated using 1H NMR titrations in 80 : 20
water–DMSO mixture. Values in parentheses indicate the error in the
association constant values.
Scheme 1 Synthesis of supramolecular hosts S2 and S3.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2017 New J. Chem., 2017, 41, 14431--14437 | 14433

View Article Online
Paper
NJC
Table 2 Fluorescence modulation of supramolecular sensors in the
presence of aromatic alcohol analytes 5–8a
Analyte
S1
S2
S3
5
6
7
8
1.00 Æ 0.00
1.01 Æ 0.00
0.99 Æ 0.00
1.01 Æ 0.01
1.04 Æ 0.01
0.82 Æ 0.01
0.90 Æ 0.00
0.87 Æ 0.01
0.98 Æ 0.01
0.88 Æ 0.01
1.05 Æ 0.02
0.75 Æ 0.01
a
Results were calculated using eqn (1). All results represent an average
of at least 3 trials.
different classes of analytes as well as within each analyte class
(Table 2). These results clearly demonstrate the effect of the sensor
architecture, and in particular the effects of covalent fluorophore
attachment and the number of fluorophore units. The covalent
attachment ensures close proximity between the cyclodextrin-
bound analyte and the fluorophore moiety(ies), causing various
degrees of fluorescence modulation to occur. An example of
analyte-induced fluorescence modulation for analyte 8 is shown
in Fig. 5.
Fig. 6 Linear discriminant analysis showing 100% differentiation between
analytes 5–8 based on their interactions with supramolecular hosts S1–S3.
The fluorescence signals of sensors S1–S3 in the presence
of analytes 5–8 were subjected to linear discriminant analysis,
and enabled 100% selectivity between the different aromatic
alcohol isomers (Fig. 6). This selectivity is particularly note-
worthy as such isomers are challenging to separate using other
analytical techniques.23 The binding of other structural isomers
and analogues in supramolecular hosts S1–S3 also led to analyte-
specific changes in the fluorescence emission (Table 3), with
selected results highlighted in Fig. 7–10.
Analytes 9–12 represent a class of aliphatic alcohols consisting
of cyclohexylmethanol (11) and its isomers. These compounds are
widely used as alkene precursors,24 and a structurally similar
analogue was part of a recent chemical spill.25 While all the
analytes are structural isomers, analytes 10 and 12 are also
stereoisomers. Distinct fluorescence modulation values are
noted for sensor S3 in combination with stereoisomers 10
and 12, highlighting the power of the cyclodextrin-based sensor
in differentiating even small structural changes. Overall, the use
of sensors S1–S3 in combination with these analytes enabled
100% differentiation using linear discriminant analysis (Fig. 7).
Analytes 13–16 represents aromatic pesticide p,p-DDT
(compound 15), its known metabolites DDE (compound 13) and
DDD (compound 14),26 and its co-occurring structural isomer
o,p-DDT (compound 16).27 These compounds are suspected
carcinogens28 and toxicants,29 and are important targets for
detection. Despite the structural similarity between the analytes,
Table 3 Fluorescence modulation of sensors S1–S3 in the presence of
analytes 9–26a
Analyte
S1
S2
S3
9
1.01 Æ 0.00
1.01 Æ 0.00
1.01 Æ 0.00
0.99 Æ 0.00
1.00 Æ 0.00
1.01 Æ 0.00
0.98 Æ 0.01
0.99 Æ 0.01
1.00 Æ 0.00
1.05 Æ 0.00
0.98 Æ 0.00
1.00 Æ 0.00
1.03 Æ 0.01
1.03 Æ 0.00
1.01 Æ 0.01
1.01 Æ 0.00
1.05 Æ 0.00
1.00 Æ 0.01
0.89 Æ 0.00
0.90 Æ 0.00
0.99 Æ 0.03
0.89 Æ 0.00
0.93 Æ 0.01
0.95 Æ 0.01
1.17 Æ 0.01
1.08 Æ 0.01
1.01 Æ 0.01
1.06 Æ 0.00
1.09 Æ 0.01
0.99 Æ 0.01
1.03 Æ 0.02
1.06 Æ 0.06
1.02 Æ 0.04
1.07 Æ 0.04
0.56 Æ 0.01
0.92 Æ 0.03
1.07 Æ 0.05
0.97 Æ 0.01
0.77 Æ 0.06
1.14 Æ 0.01
1.33 Æ 0.03
1.07 Æ 0.04
1.35 Æ 0.05
1.04 Æ 0.05
0.94 Æ 0.02
0.93 Æ 0.02
0.95 Æ 0.02
1.01 Æ 0.01
0.89 Æ 0.01
0.85 Æ 0.01
0.98 Æ 0.03
0.89 Æ 0.02
0.98 Æ 0.01
1.14 Æ 0.02
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
a
Fluorescence modulation results were calculated using eqn (1). All
results represent an average of at least 3 trials.
100% accurate classification was achieved (Fig. 8). Interestingly,
although sensor S3 demonstrated nearly identical fluorescence
modulation values in response to analytes 13 and 15, sensor S2
was able to clearly differentiate between those two analytes. These
results illustrate that altering the degree of functionalization of
the sensor can alter its response.
Analytes 17–21 represent aliphatic n-hexane (compound 17),
its commonly occurring structural isomers (compounds 18–20,
generated in 10–30% yield from industrial production of
hexane)30 and its cyclopentane analogue (compound 21). The
fact that hexanes co-occur as isomeric mixtures complicates a
variety of applications that require accurate characterization.31
Using this supramolecular sensing strategy, 100% accurate
classification between these analytes is achieved (Fig. 9).
Analytes 22–26 represent polychlorinated biphenyls (PCBs), a
class of POPs that cause neurotoxicity32 and endocrine disruption.33
Fig. 5 Fluorescence emission of (A) sensor S1; (B) sensor S2; and
(C) sensor S3 in the presence of analyte 8. (lex = 320 nm; 3 nm excitation
slit width; 3 nm emission slit width).
14434 | New J. Chem., 2017, 41, 14431--14437
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2017

View Article Online
NJC
Paper
Fig. 7 (A) Fluorescence response of host S1 in the presence of analytes
9–12; (B) linear discriminant analysis of the fluorescence responses,
leading to 100% differentiation of the analyte signals. (lex = 320 nm;
3 nm excitation slit width; 3 nm emission slit width).
Fig. 10 (A) Fluorescence response of host S2 in the presence of analytes
22–26; (B) linear discriminant analysis of the fluorescence responses,
leading to 100% differentiation of the analyte signals (lex = 320 nm;
3 nm excitation slit width; 3 nm emission slit width).
Table 4 Calculated limits of detection and comparisons to known levels
of concern
Analytes
Sensors
LOD calculated (mM)
Limit of concern (mM)
a
5
6
6
S2
S1
S3
S1
S2
S1
S2
S3
S1
S2
S3
S3
S1
S2
7.1 Æ 0.9
5.5 Æ 0.2
21.2735
7.3 Æ 0.5
21.2735
a
11
11
15
15
15
18
19
21
22
25
26
1.2 Æ 0.01
1.4 Æ 0.1
a
0.43 Æ 0.04
0.48 Æ 0.05
2.1 Æ 0.03
2.1 Æ 0.2
2.8236
2.8236
2.8236
Fig. 8 (A) Fluorescence response of host S2 in the presence of analytes
13–16; (B) linear discriminant analysis of the fluorescence responses,
leading to 100% differentiation of the analyte signals (lex = 320 nm;
3 nm excitation slit width; 3 nm emission slit width).
5801.8137
21.1 Æ 1.4
8.4 Æ 0.6
5801.8138
a
5.2 Æ 0.2
1.0038
1.7139
1.0038
0.30 Æ 0.01
0.17 Æ 0.01
a
Limits of concern have not been established for these compounds.
The limits of detection for each sensor S1, S2 and S3 for
each class of analytes were calculated, to determine their ability
to sense analytes at environmental levels of concern and
at levels that induce toxicity. In every case, the calculated limits
of detection were at or below the literature reported limits of
concern (Table 4), highlighting the sensitivity of this method.
Practical applications of this system require the capability to
identify analyte mixtures, because environmental contamina-
tion scenarios almost always involve such mixtures. To that
end, preliminary work focused on identification of 1 : 1 binary
mixtures of aromatic alcohol analytes 5–8. Using the supra-
molecular sensors combined with linear discriminant analytical
Fig. 9 (A) Fluorescence response of host S3 in the presence of analytes
17–21; (B) linear discriminant analysis of the fluorescence responses,
leading to 100% differentiation of the analyte signals (lex = 320 nm;
3 nm excitation slit width; 3 nm emission slit width).
As a result of these effects, the use of PCBs has been banned in techniques, 83% accurate identification of the 1 : 1 binary
many countries; however, their environmental persistence mixtures was obtained (Fig. 11). Interestingly, the mixture of
means that significant amounts of PCBs are still found in the analytes 5 + 7 is grouped near the mixtures of analytes 6 + 8 and
environment.34 100% accurate classification has been achieved 5 + 8, which reduces the overall classification accuracy slightly.
for these analytes (Fig. 10), which is particularly crucial because This kind of co-clustering of analyte groups has been observed
these analytes have widely disparate toxicities.
previously, and can be attributed to similar sensor responses
The ability of this detection method to generate well- originating from competing interactions between each component
separated signals was further investigated by generating an of the mixture. Other than those combinations, the mixtures
array with all analytes from all classes. In this case, the array demonstrated excellent signal separation and accurate identifi-
exhibited well-separated clusters based on compound class, cation. Current work in our group is focused on improving
as well as excellent separation within each class. Overall, classification accuracy of analyte mixtures, expanding such
100% accurate identification was obtained (see ESI† for more techniques to multiple analyte classes, and moving from binary
details).
mixtures to ternary and even quaternary mixtures of analytes.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2017 New J. Chem., 2017, 41, 14431--14437 | 14435

View Article Online
Paper
NJC
References
1 K. L. Willett, E. M. Ulrich and R. A. Hites, Environ. Sci.
Technol., 1998, 32, 2197.
2 I. Rodriguez, M. Fraga, A. Alfonso, D. Guillebault, L. Medlin,
J. Baudart, P. Jacob, K. Helmi, T. Meyer, U. Breitenbach,
N. M. Holden, B. Boots, R. Spurio, L. Cimarelli, L. Mancini,
S. Marcheggiani, M. Albay, R. Akcaalan, L. Koeker and
L. M. Botana, Environ. Toxicol. Chem., 2017, 36, 645.
3 M. Bergknut, A. Kucera, K. Frech, E. Andersson, M. Engwall,
U. Rannug, V. Koci, P. L. Andersson, P. Haglund and
M. Tysklind, Environ. Toxicol. Chem., 2007, 26, 208.
4 U. Shafique, S. Schulze, C. Slawik, S. Kunz, A. Paschke and
G. Schueuermann, Anal. Chim. Acta, 2017, 949, 8.
5 J. E. Rager, M. J. Strynar, S. Liang, R. L. McMahen, A. M.
Richard, C. M. Grulke, J. F. Wambaugh, K. K. Isaacs,
R. Judson, A. J. Williams and J. R. Sobus, Environ. Int.,
2016, 88, 269.
Fig. 11 Linear discriminant analysis results of binary mixtures of analytes 5–8.
6 O. R. Miranda, B. Creran and V. M. Rotello, Curr. Opin.
Chem. Biol., 2010, 14, 728; K. L. Diehl and E. V. Anslyn,
Chem. Soc. Rev., 2013, 42, 8596.
7 X. Xu, Z. Liu, X. Zhang, S. Duan, S. Xu and C. Zhou,
Electrochim. Acta, 2011, 58, 142; X. Chen, X. Cheng and
J. J. Gooding, Anal. Chem., 2012, 84, 8557.
8 S. Hussain, A. H. Malik and P. K. Iyer, J. Mater. Chem. B,
2016, 4, 4439.
9 B. B. Prasad and I. Pandey, Sens. Actuators, B, 2013, 181, 596.
10 J. Zhang, X. Zhang, J. Chen, C. Deng, N. Xu, W. Shi and
P. Cheng, Inorg. Chem. Commun., 2016, 69, 1.
11 S. Hatanaka, T. Ono and Y. Hisaeda, Chem. – Eur. J., 2016,
22, 10346; S. Mukherjee, A. V. Desai, A. I. Inamdar, B. Manna
and S. K. Ghosh, Cryst. Growth Des., 2015, 15, 3493; B. Gole,
W. Song, M. Lackinger and P. S. Mukherjee, Chem. – Eur. J.,
2014, 20, 13662.
12 E. A. Cohen Hubal, A. Richard, L. Aylward, S. Edwards,
J. Gallagher, M.-R. Goldsmith, S. Isukapalli, R. Tornero-
Velez, E. Weber and R. J. Kavlock, J. Toxicol. Environ. Health,
2010, 13, 299.
13 L. J. Gensburg, C. Pantea, C. Kielb, E. Fitzgerald, A. Stark
and N. Kim, Environ. Health Perspect., 2009, 117, 1265.
14 D. Megson, E. J. Reiner, K. J. Jobst, F. L. Dorman, M. Robson
and J.-F. Focant, Anal. Chim. Acta, 2016, 941, 10.
15 D. J. DiScenza and M. Levine, New J. Chem., 2016, 40, 789;
N. Serio, D. F. Moyano, V. M. Rotello and M. Levine, Chem.
Commun., 2015, 51, 11615; N. Serio, K. Miller and M. Levine,
Chem. Commun., 2013, 49, 4821.
Conclusions
In conclusion, we have developed an efficient array-based detection
strategy for isomeric and analogous analytes. The array employs
three architecturally unique perbenzylated b-cyclodextrin–fluoro-
phore sensors for identification of a particular isomer within a
class of isomeric or structurally similar analytes. The binding of
analytes to the cyclodextrin induces a distinct change in the
fluorescence emission of the attached fluorophores, which is then
statistically translated into array clusters of maximum separation
via linear discriminant analysis. We demonstrate 100% successful
classification of three isomeric (aromatic alcohols, aliphatic
alcohols, aliphatic hexanes) and two analogous (DDT pesticides,
PCB congeners) analyte classes. Sensitivity measurements highlight
limits of detection at or near literature-reported levels of concern.
Preliminary attempts on binary mixtures demonstrated fairly selec-
tive levels of classification with 83% accuracy. This method in
tandem with chromatographic analysis of complex isomeric mix-
tures would complement each other in determining the nature
of each isomer. Current work in our laboratory is focused on
expanding the classes of analytes detectable via this system,
improving analyte mixture identification, and developing a practical
cyclodextrin-based detection device. The results of these and other
investigations will be reported in due course.
16 K. Higashi, S. Ideura, H. Waraya, W. Limwikrant, K. Moribe
and K. Yamamoto, Chem. Pharm. Bull., 2010, 58, 769.
17 J. Bjerre, T. H. Fenger, L. G. Marinescu and M. Bols,
Eur. J. Org. Chem., 2007, 704.
Conflicts of interest
There are no conflicts of interest to declare.
18 T. Lecourt, A. Herault, A. J. Pearce, M. Sollogoub and
P. Sinay, Chem. – Eur. J., 2004, 10, 2960.
19 B. S. Choi, J. Choi, S. Bak and S. Koo, Eur. J. Org. Chem.,
2015, 514.
Acknowledgements
This work was supported by the National Science Foundation
(Grant Number 1453483) and the National Cancer Institute 20 C. Huo, J.-C. Chambron and M. Meyer, New J. Chem., 2008,
(Grant Number CA185435).
32, 1536.
14436 | New J. Chem., 2017, 41, 14431--14437
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2017

View Article Online
NJC
Paper
21 G. W. Meindersma, T. van Schoonhoven, B. Kuzmanovic 32 G. Winneke, J. Neurol. Sci., 2011, 308, 9.
and A. B. de Haan, Chem. Eng. Process., 2006, 45, 175.
22 R. N. Dsouza, U. Pischel and W. M. Nau, Chem. Rev., 2011,
111, 7941.
33 T. Grzeskowiak, B. Czarczynska-Goslinska and A. Zgola-
Grzeskowiak, TrAC, Trends Anal. Chem., 2016, 75, 209.
34 N. S. Shifrin and A. P. Toole, Environ. Eng. Sci., 1998, 15, 247.
23 J. Peng, Y. Zhang, X. Yang and M. Qi, J. Chromatogr. A, 2016, 35 Center for Disease Control and Prevention. The National
1466, 148.
Institute for Occupational Safety and Health (NIOSH):
Cresols (o, m, p isomers) https://www.cdc.gov/niosh/idlh/
cresol.html, accessed Dec 20, 2016.
24 J. J. Cawley and P. E. Lindner, J. Chem. Educ., 1997, 74, 102.
25 A. J. Whelton, L. McMillan, M. Connell, K. M. Kelley,
J. P. Gill, K. D. White, R. Gupta, R. Dey and C. Novy, Environ. 36 Center for Disease Control and Prevention. The National
Sci. Technol., 2015, 49, 813.
26 M. Ricking and J. Schwarzbauer, Environ. Chem. Lett., 2012,
10, 317.
Institute for Occupational Safety and Health (NIOSH): DDT
https://www.cdc.gov/niosh/idlh/50293.html, accessed Dec
20, 2016.
27 J. P. Arrebola, M. Cuellar, J. P. Bonde, B. Gonzalez-Alzaga 37 Center for Disease Control and Prevention. The National
and L. A. Mercado, Environ. Res., 2016, 151, 469.
28 A. M. Soto, K. L. Chung and C. Sonnenschein, Environ.
Health Perspect., 1994, 102, 380.
Institute for Occupational Safety and Health (NIOSH):
2-methylpentane. https://www.cdc.gov/niosh/ipcsneng/neng
1262.html, accessed Dec 20, 2016.
29 A. Mansouri, M. Cregut, C. Abbes, M.-J. Durand, A. Landoulsi 38 Center for Disease Control and Prevention. The National
and G. Thouand, Appl. Biochem. Biotechnol., 2017, 181, 309.
30 R. A. Myers, Handbook of Petroleum Refining Process, McGraw-
Hill, New York, 2004; R. Clavier, Wiley Critical Content:
Institute for Occupational Safety and Health (NIOSH):
3-methylpentane. https://www.cdc.gov/niosh/ipcsneng/neng
1263.html, accessed Dec 20, 2016.
Petroleum Technology, Wiley-Interscience, Hoboken, New Jersey, 39 Center for Disease Control and Prevention. The National
2007.
Institute for Occupational Safety and Health (NIOSH): Poly-
chlorinated Biphenyls (PCB’s). https://www.cdc.gov/niosh/
docs/86-111/default.html, accessed Dec 20, 2016.
31 R. Hayati, S. Z. Abghari, S. Sadighi and M. Bayat, Korean
J. Chem. Eng., 2015, 32, 629.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2017 New J. Chem., 2017, 41, 14431--14437 | 14437