A fluorometric paper-based sensor array for the discrimination of volatile organic compounds (VOCs) with novel salicylidene derivatives

Article abstract of DOI:10.1016/j.dyepig.2018.06.044

In this paper, novel salicylidene fluorescent sensors (1, 2, and 3) were successfully synthesized by the one pot condensation reaction and coated onto filter paper to produce fluorometric paper-based sensor array. The compound 1, 2, and 3 paper-based sens

Full text of DOI:10.1016/j.dyepig.2018.06.044

DyesandPigments159(2018)378–383
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
A fluorometric paper-based sensor array for the discrimination of volatile
organic compounds (VOCs) with novel salicylidene derivatives
Thiti Jarangdeta,b, Kornkanya Pratumyota, Kittiwat Srikittiwannaa, Wijitar Dungchaia,
Withawat Mingvanisha, Ittipon Techakriengkraic, Mongkol Sukwattanasinittd,
Nakorn Niamnonta,b,∗
a
Organic Synthesis, Electrochemistry & Natural Product Research Unit, Department of Chemistry, Faculty of Science, King Mongkut's University of Technology Thonburi,
Bangkok, 10140, Thailand
b
Luminescence & Scintillation Materials Research Unit, Faculty of Science, King Mongkut's University of Technology Thonburi, Bangkok, 10140, Thailand
Department of Food Technology, Faculty of Science, Ramkhamhaeng University, Bangkok 10240, Thailand
Department of Chemistry, Faculty of Science and Nanotec-CU Center of Excellence on Food and Agriculture, Chulalongkorn University, Bangkok, 10330, Thailand
c
d
A B S T R A C T
In this paper, novel salicylidene fluorescent sensors (1, 2, and 3) were successfully synthesized by the one pot
condensation reaction and coated onto filter paper to produce fluorometric paper-based sensor array. The
compound 1, 2, and 3 paper-based sensor arrays exhibited rapid fluorescent signal in response to exposure to 15
different VOC vapors. A smartphone was used as an alternative device to monitor the appearance of fluorescent
signals. The images photographed using the smartphone camera were analyzed for RGB values by a conventional
image-processing program. The ΔRGB data set (3 fluorophores × 15 VOC vapors × 4 repeats) from fluorescent
images of 15 VOC vapors analysis have been statistically categorized by using principal component analysis
(PCA) into 15 clusters with 100% discriminating accuracy. The proposed sensor array showed high potential as a
portable electronic nose (e-nose) system for VOC vapors discrimination.
1. Introduction
development of highly sensitive and selective devices [5a–d]. Although
the detection of very small amounts of VOCs can be achieved by using
standard analytical techniques [6] such as GC or GC-MS, these techni-
ques require specialists to operate and take long time to analyze are
expensive and cannot be portable [7a,7b]. Therefore, simple, rapid and
economic devices which can on-site analyze VOCs are highly de-
manded. As a result, a portable electronic nose (e-nose) system based on
mobile sensors has been prevalently designed and developed [8a,8b].
The electrochemical [(5a)], color [9a–e], and fluorescent modes
[10a–e] applied for the VOCs detection and identification of sensor
array have been continually reported [11]. However, their applications
are still limited only to identify amine compounds, and are applied to
more number of sensors to discriminate VOCs [12a–f]. A real need for
the development of high-performance portable VOC sensor arrays
therefore still remains. Fluorescent technique has more advantages due
to its high sensitivity and selectivity, more safety, high-throughput
analysis, low cost, enabling portability, and simplicity to end-users
[13a,13b].
Volatile organic compounds (VOCs) are either subtracted from in-
spired air (by degradation and/or excretion in the body) or added to
alveolar breath as products of metabolism [1]. They do not include
carbon monoxide, carbon dioxide, carbonic acid, metallic carbides or
carbonates and ammonium carbonate [2]. They are found everywhere
both in indoor and outdoor environments because they are essential
ingredients in many products and materials. Consequently, VOCs may
have short- and long-term adverse health effects on humans, e.g. ner-
vous system impairment, asthma and cancer [3]. The measurement of
VOCs level both in indoor and outdoor environments has been im-
portant and may be applicable to simply diagnose some acute and
chronic diseases such as lung cancer from analysis of a patient's breath
[4a,4b]. Nowadays, many efforts have been made to actively develop
techniques for detecting and identifying volatile organic compounds
(VOCs) in indoor and outdoor environments and in acute and chronic
disease diagnostics [5a–d]. In order to determine VOCs at very low
levels of concentration, there is high necessity/need for the
Salicylidene fluorophores containing benzene, nitrophenol,
∗
Corresponding author. Organic Synthesis, Electrochemistry & Natural Product Research Unit, Department of Chemistry, Faculty of Science, King Mongkut's University of Technology
Thonburi, Bangkok, 10140, Thailand.
E-mail address: nakorn.nia@kmutt.ac.th (N. Niamnont).
https://doi.org/10.1016/j.dyepig.2018.06.044
Received 28 April 2018; Received in revised form 24 June 2018; Accepted 25 June 2018
Availableonline30June2018
0143-7208/©2018ElsevierLtd.Allrightsreserved.

T. Jarangdet et al.
DyesandPigments159(2018)378–383
down, the precipitated solid was filtered off and washed with cool
ethanol (3 × 15 mL) which was purified by recrystallization with
ethanol. Compound 1 was obtained as an orange crystal (205 mg, 62%
yield), mp: 188–190 °C; δH (400 MHz, CDCl3) 12.24 (1H, s), 9.22 (1H,
s), 7.99 (2H, d, J = 8.0 Hz), 7.86 (4H, d, J = 8 Hz), 7.64 (1H, d,
J = 2.4 Hz), 7.56–7.48 (1H, m), 7.41 (1H, t, J = 8.0 Hz), 6.98 (1H, d,
J = 12.0 Hz); δC (100 MHz, CDCl3) 168.4, 166.0, 160.9, 151.5, 137.9,
135.7, 134.8, 126.9, 125.5, 123.3, 121.8, 119.7, 111.3; HRMS-ESI m/z
calcd for C14H9BrN2OS: 334.9677 [M+H]+, Found: 334.9742 [M
Fig. 1. Structures of benzothiazole-salicylidene derivatives 1, 2, and 3.
+H]+
.
naphthalene, pyrene and bisphenol A have been reported to give high
sensitivity and selectivity via an excited state intramolecular proton
transfer (ESIPT) process and a Schiff base process towards fluoride
detection [14a–f]. However, most salicylidene derivatives have not
been studied to be the fluorescent paper-based sensor arrays for VOC
vapors detection. In this work, we reported the effectiveness of the
fluorescent paper-based sensor arrays embedded with novel ben-
zothiazole-salicylidene derivatives for detecting VOCs. The sensor ar-
rays generate different fluorescent signal profiles as exposed to VOC
analysts, which lead to the detection and qualitative recognition of the
15 different VOC vapors. Novel benzothiazole-salicylidene fluorophores
used for fluorescent paper-based sensor arrays for VOCs were shown in
Fig. 1. Numerical data of fluorescent signals to VOC vapors were con-
veniently collected by a smartphone in the form of images and further
analyzed by principal component analysis (PCA) program to identify 15
VOC vapors.
2.4. Synthesis of compound 2
2-Hydroxy-5-methoxybenzaldehyde (200 mg, 1.31 mmol), 2-ami-
nobenzothiazole (210 mg, 1.40 mmol) and ethanol (5 mL) were in-
troduced into a sealed tube (100 mL) under N2 atmosphere and the
reaction mixture was stirred for 3 h at 78 OC. After the reaction mixture
was cooled down and the precipitated solid was filtered off and washed
with cool ethanol (3 × 15 mL) which was purified by recrystallization
with ethanol. Compound 2 was obtained as a red crystal (268 mg, 72%
yield), mp; 144–146.5 OC; δH (400 MHz, CDCl3) 11.86 (1H, s), 9.25 (1H,
s), 7.97 (1H, d, J = 8.0 Hz), 7.85 (1H, d, J = 8.0 Hz), 7.50 (1H, td,
J = 1.2, 8.4 Hz), 7.39 (1H, td, J = 1.2, 8.4 Hz), 7.11 (1H, dd, J = 4.0,
8.0 Hz), 7.05–6.95 (2H, m), 3.02 (3H, s). δC (100 MHz, CDCl3) 169.1,
167.1, 156.6, 152.6, 151.5, 134.7, 126.7, 125.3, 123.7, 123.0, 121.8,
118.7, 117.8, 115.7, 115.2, 55.9; HRMS-ESI m/z calcd for
C
15H12N2O2S: 285.0692 [M+H]+, Found: 285.0778 [M+H]+
.
2. Experimental
2.5. Synthesis of compound 3
2.1. Reagents and apparatus
5-nitrosalicyladehyde (200 mg, 1.20 mmol), 2-aminobenzothaizole
(210 mg, 1.4 mmol) and ethanol (5 mL) were introduced into a sealed
tube (100 mL) under N2 atmosphere and the reaction mixture was
stirred for 3 h at 78 OC. After the reaction mixture was cooled down, the
precipitated solid was filtered off and washed with cool ethanol
(3 × 15 mL) which was purified by recrystallization with ethanol.
Compound 3 was obtained as a red crystal (301 mg, 84% yield), mp:
231–233 OC; δH (400 MHz, CDCl3) 13.10 (1H, s), 9.40 (1H, s), 8.53 (1H,
d, J = 4 Hz), 8.36 (1H, dd, J = 4, 8 Hz), 8.02 (1H, d, J = 8 Hz), 7.89
(1H, d, J = 8 Hz), 7.55 (1H, t, J = 8 Hz), 7.44 (1H, t, J = 8 Hz), 7.17
(1H, d, J = 8 Hz). δC (100 MHz, CDCl3) 195.4, 166.6, 165.5, 131.7,
129.9, 129.7, 127.1, 125.9, 123.6, 121.9, 119.0, 118.7, 117.6, 77.0;
HRMS-ESI m/z calcd for C14H9N3O3S: 300.0437 [M+H]+, Found:
2-Hydroxy-5-methoxybenzaldehyde, 5-nitrosalicyladehyde, 5-bro-
mosalicyladehyde, 2-aminobenzothiazole were purchased from Tokyo
Chemical Industry (TCI). All other reagents were non-selectively pur-
chased from Sigma-Aldrich, Fluka, Merck or RCI Labscan and used
without further purification. Generally, solvents such as acetonitrile,
dichloromethane and ethanol were reagent grade and were stored over
4 Å molecular sieves before use. All reactions were carried out under N2
atmosphere. Additionally, silica gel 60 F254 thin layer chromatography
(TLC) was used for following the progress of a reaction.
2.2. Analytical instrument
The 1H and 13C NMR spectra were performed for benzothiazole-
salicylidene derivatives 1, 2, and 3 on a Bruker Avance Spectrometer
(400 MHz, 100 MHz for 1H and 13C) using tetramethylsilane as the in-
ternal standard. The number of absorption signals in the 1H NMR
spectra was designated as follows: s/singlet; d/doublet, t/triplet, dd/
double doublet, q/quartet, m/multiplet. Infrared spectra was performed
with a Thermo fisher scientific instrument (model Nicolet 8700) using
KBr pellets. High resolution electrospray ionization mass spectra
(HRMS) were acquired on a Bruker maXis using acetonitrile as a solvent
in the positive ionization mode. Melting points were measured using a
Thomas Hoover capillary instrument and their values were not subse-
quently corrected. Fluorescent spectra were measured on a Hitachi F-
2500 fluorescence spectrophotometer. UV-Vis absorption spectra were
measured on a Perkin Elmer ltd, lambda 35/fias 300 UV-vis spectro-
photometer. Fluorescent intensity in multisampling was recorded on
EnSpire multimode plate reader, Perkin-Elmer.
300.0522 [M+H]+
.
2.6. Photo-physical property of compounds 1, 2, and 3
The stock solutions of compounds 1, 2, and 3 (500 μM) were pre-
pared in methanol. 100 μM quinine sulfate in 0.05 M H2SO4 was used as
a standard fluorophore. The UV-Vis absorption spectra between 250 nm
to 600 nm were measured using a 1 cm quartz cell at room temperature.
The fluorescent spectra of compounds 1, 2, and 3 and quinine sulfate
were obtained under light irradiation of 380 nm, 410 nm, 310 nm and
345 nm, respectively. The fluorescence quantum yields of compounds
1, 2, and 3 were calculated by comparing to standard quinine sulfate
(Φ = 0.54) in 0.1 M H2SO4 [15].
2.7. Fabrication of compound 1, 2, and 3 coated papers
2.3. Synthesis of compound 1
The black circle ring for loading sample (5.0 mm of inner diameter)
were printed on Whatman filter paper No.1 by HP LaserJet Pro P1102w
Printer, and it was coated with the 3M adhesive tape on the backside.
Afterwards, 2 μL of the sensory solutions of compounds 1, 2, and 3 in
acetone (1 mM) were individually dropped to the filter papers directly
[16]. The fabricated papers were then dried using hair drier [12(c)].
5-bromosalicyladehyde (200 mg, 0.995 mmol), 2-aminobenzothia-
zole (165 mg, 1.1 mmol) and ethanol (5 mL) were introduced into a
sealed tube (100 mL) under N2 atmosphere and the reaction mixture
was stirred for 3 h at 78 OC. After the reaction mixture was cooled
379

T. Jarangdet et al.
DyesandPigments159(2018)378–383
2.8. Detection of VOCs and fluorometric measurements
Table 1
Photo-physical properties of compounds 1, 2, and 3 in methanol Solution.
Each VOC (10 mL) was individually added into a TLC developing
chamber with a lid (25 mL), and allowed it to reach equilibrium for
60 min. After that, the VOC saturated chamber was covered with the lid
having the paper coated with the sensor molecule inside for 5 min at
room temperature. The photographic images of the sensory paper sa-
turated with VOC vapor were then recorded under the UV light of
365 nm using i-phone SE with Shutter application of the highest ISO
and medium white balance modes. Each image was converted to TIF
format file and cropped into a circle of 0.065 cm2 areas. The RGB values
of the cropped images were analyzed by Image J program. The ΔR, ΔG
and ΔB values were obtained from a comparison of the RGB values
before and after exposure to VOC saturated vapors. The raw 540 RGB
numerical data (3 RGB values × 15 solvents × 3 sensors × 4 repeats)
were analyzed by Principal Component Analysis (PCA) was performed
by Unscrambler v. 9.7 (CAMO A/S, Oslo, Norway) using full cross va-
lidation method. All variables were setting at 1.0/standard deviations
for their weights.
Compounds
λabs (nm)
log ε
λem (nm)
Φ
1
2
3
270, 380
270, 410
270, 310
4.05, 3.29
3.93, 3.60
4.13, 3.90
550
560
540
0.0069
0.0855
0.0008
3. Results and discussion
Figure 2. Preparation of paper-based salicylidene sensors.
3.1. Synthesis of salicylidene derivatives (1, 2, and 3)
decrease in the extent of π conjugated system. The fluorescent quantum
yields of all compounds were very weak in methanol media, especially
that for compound 3. We expected that the electron-withdrawing nitro
group enhanced the excited state intramolecular proton transfer
(ESIPT) process in protic solvents [19a,19b], resulting in low fluor-
escent quantum yields of compound 3 in methanol.
A series of salicylidene derivatives (1, 2, and 3) were synthesized via
the one-step condensation reaction of salicylaldehyde derivatives and
2-aminobenzothiazole with high yield (62–84%) [17]. Indeed, the re-
action of 2-aminobenzothiazole with various salicylaldehyde can be
considered as one of the green chemistry process due to the utilization
of ethanol solvent possessing little or no toxicity to the human health
and environment (Scheme 1) [18]. The chemical structures of com-
pounds 1, 2, and 3 have been confirmed by 1H NMR, 13C NMR and FT-
IR techniques as shown in supporting information (Fig S1-S9). Ob-
viously, 1H NMR spectra showed the characteristic imine proton signals
of compounds 1, 2, and 3 at 9.30, 9.30 and 9.40 ppm, respectively. In
addition, the presence of the HRMS-ESI molecular ion peaks [M+H]+
of compounds 1, 2, and 3 at m/z 334.9742, 285.0778 and 300.0522,
respectively, corresponded to the proposed molecular structures as
shown in supporting information (Fig. S10-S12).
3.3. Preparation of compound 1, 2, and 3 coated papers
Compounds 1, 2, and 3 were successfully coated on filter papers
with the simple process stated previously (Fig. 2). On the exposure of
VOCs vapor, the fabricated paper coated with compounds 1, 2, and 3
exhibited fluorescent emission that can be observed by naked eye under
UV light at 365 nm. Compounds 1, 2, and 3 fabricated on the papers
were stable upon storage at least four weeks in the dark at room tem-
perature when exposed to the UV light of 365 nm.
3.2. Photo physical properties of compound 1, 2, and 3
3.4. Fluorescence responses of compounds 1, 2, and 3 to VOCs
The photo-physical properties of compounds 1, 2, and 3 are sum-
marized in Table 1. The UV−Vis absorption spectrum of each com-
pound in methanol displayed the maximum absorption wavelength
(λmax) ranging from 310 to 410 nm with two major absorption bands
corresponding to the π−π* and n−π* transitions. The maximum
emission wavelength of each compound was observed in the range
between 460 and 560 nm (Fig. S13–S15). The absorption and emission
spectra of compounds 1 and 2 exhibited remarkable red-shift phe-
nomenon in comparison with those of compound 3. This could be ex-
plained by the effect of electron withdrawing nitro (–NO2) group on the
Earlier research mainly focused on luminescent properties of sali-
cylidene derivatives towards volatile solvents in liquid phase [20].
However, the best of our knowledge, there is no published paper re-
porting their detection of VOCs in gaseous phase. Therefore, the com-
pound 1, 2, and 3 coated papers were prepared and tested against a
wide range of VOCs vapor, i.e. hexane, toluene, benzene, diethyl ether,
dichloromethane (CH2Cl2), chloroform (CHCl3), tetrahydrofuran (THF),
ethyl acetate (EtOAc) acetone, acetonitrile (MeCN), dimethylforma-
mide (DMF), dimethylsulfoxide (DMSO), 2-propanol, ethanol (EtOH)
and methanol (MeOH). Upon exposure to UV light at 365 nm for 5 min,
the spots of compounds 1, 2, and 3 coated on the paper exhibited a
difference in fluorescent emission responses, which was attributed to
the different interactions of the sensoring compounds with the VOCs
tested (Fig. 3). Specifically, the enol-imine group in compounds 1, 2,
and 3 form intramolecular (O−H) hydrogen bond, which is highly
sensitive to local polarity and local hydrogen bond to their oxygen
atoms. In polar protic solvents such as methanol, the enol (-OH) group
of compounds 1, 2, or 3 formed double intermolecular hydrogen bonds
with the hydroxyl (-OH) group of methanol and caused the enol-imine
group to tautomerize into the keto-enamine group (Scheme 1S in sup-
plementary data) [21]. The excited state intramolecular proton transfer
(ESIPT) process that was accounted for this tautomerization was
Scheme 1. Synthetic route to compounds 1, 2, and 3.
380

T. Jarangdet et al.
DyesandPigments159(2018)378–383
Fig. 3. Fluorescent images of Compound 1, 2, and 3 based paper sensory arrays exposed by various vapors of volatile organic solvents under black light of 365 nm.
responsible for the decrease in fluorescent intensity on exposure of the
compound 1, 2, and 3 coated on papers to methanol and ethanol (see
Fig. S17(a)-S19(c) in supplementary data) [22]. On the contrary, there
was lower ESIPT generated in polar aprotic solvents by hydrogen bond
with other solvents, thus the fluorescent intensity of the sensoring pa-
pers were maintained upon contact with these solvents. In addition, the
hydrophobic interaction (London) of compounds 1, 2, and 3 with non-
polar solvents helped to stabilize the monomeric form of compounds 1,
2, and 3 from aggregation that caused their fluorescent emission to be
quenched. Therefore, the fluorescent intensity of the sensor spots in
contact with all VOCs except polar protic ethanol and methanol were
still strong emission.
patterns, while the polar solvents display a large change in ΔRGB pat-
terns with a remarked change in ΔR values. In particular, the polar
protic solvents, i.e. methanol and ethanol displayed the most unique
ΔRGB patterns on the sensor arrays that allows them to be distinguished
from other VOCs. Therefore, the paper-based sensor arrays of com-
pounds 1, 2, and 3 could selectively detect and identify the presence of
methanol and ethanol as compared with other VOCs. The standard
deviation as depicted by the error bars in Fig. 4, was calculated from 12
measurements (3 sensors × 4 repeats), was small which demonstrates
the precision of this method of measurement.
3.6. Multivariate statistical analyses of the fluorescent responses of
compounds 1, 2 and 3 to VOCs
3.5. Evaluation of fluorescent response patterns of compounds 1, 2 and 3 to
VOCs
Variability in fluorescence response patterns can be used to dis-
criminate different compounds. To organize an enormous data set with
many variables (15 VOCs × 3 sensors × 4 repeats × 3 values of RGB),
the principal analysis (PCA) is an alternative tool to reduce di-
mensionality of multivariate data by grouping similar observations to-
gether. PCA is a statistical multivariate analysis to generate principal
component (PC) scores as a data set. Fig. 5 showed that PC1 and PC2
were taken into account for 52% and 25% of the data variance, re-
spectively. The 2D PCA plot clearly shows the high discriminating
ability of the compound 1, 2, and 3 paper-based sensor array towards
15 VOCs. As expected, methanol, which showed the unique ΔRGB
pattern among other VOCs, had the highest PC1 and PC2 scores. To
evaluate the classification accuracy, factorial discriminant analysis
(FDA) was applied on the PC scores to cross validate the discriminating
ability using a leave-one-out technique [24]. The FDA cross validation
gave 100% classification accuracy. The results emphasized the high
effectiveness of the novel salicylidene derivatives 1, 2, and 3 applied on
a fluorometric paper-based sensor array for detection and identification
of VOCs.
Due to the most distinguishable PC score of methanol, the com-
pound 1, 2, and 3 paper-based sensor array have been selected for
testing the real methanol samples. In order to gather more details of the
methanol detection, we have set out to test the effect of concentration
of methanol on color change of the compound 3 spot on paper-based
sensor array. The compound 3 spot was used in this experiment because
it showed the most obvious color change upon contact with methanol as
compared with compounds 1 and 2. To investigate the fluorescent be-
havior of compound 3 towards methanol concentration, a paper-based
sensor array of compound 3 spots was exposed to 6 different con-
centrations of methanol solution in water. Upon exposure to methanol
vapors in water for 5 min, the compound 3 paper-based sensor array
The 45 fluorescent responses of 3 sensors with 15 VOC vapors were
transformed into images of TIF type with a digital camera, and then the
color of the TIF images was determined and reported to the RGB values
(R = red, G = green and B = blue) using ImageJ software program.
The three-dimensional vectors (ΔR, ΔG, ΔB) of each salicylidene spot on
the image were calculated its RGB values before and after exposure to
the VOC vapors. Histogram plot (Fig. 4) shows the changes in ΔR, ΔG,
and ΔB values of the three sensors upon exposure to various VOC va-
pors. Notably, most of the samples exhibited considerable change in the
red (R) and green (G) values as compared to the blue (B) values. The
images of compound 1 and 3 spots have decreased red and green va-
lues, resulting in the negative values of ΔR and ΔG. Additionally, the
decrease in the R, G, and B values of compound 3 spots are more dra-
matic than that of compound 1 spots for most VOCs except methanol
and ethanol. The images of compound 2 spots, in contrast, have in-
creased R, G, and B values, leading to the positive values of ΔR, ΔG, and
ΔB. These responses were possibly due to the substituent effects that
stablilize or destabilize tautomeric form of compounds 1, 2, and 3 over
another. The electron donating –OCH3 group in compound 2 stabilizes
the enol-imine tautomeric form by strengthening the O−H bond of the
enol, thus reduces the ESIPT and consequently promotes the increased
fluorescent signal intensity [23]. On the other hand, the electron
withdrawing –Br and -NO2 groups in compounds 1 and 3 respectively
destabilizes the enol-imine form by weakening the O−H bond of the
enol [14(e)], driving the tautomerization toward keto-enamine form.
During the tautomerization, the ESIPT process is responsible for the
decrease in fluorescent signal intensities. On comparison of the ΔRGB
patterns among 15 VOCs, it showed that the non polar solvents, i.e.
hexane, toluene and benzene exhibited a subtle change in ΔRGB
381

T. Jarangdet et al.
DyesandPigments159(2018)378–383
Fig. 4. Histogram plot of ΔRGB profile of a fluorometric paper-based sensor array of compounds 1, 2, and 3 after exposure to 15 saturated vapors of VOCs.
showed various color changes, as shown in Fig. 6. As the concentration
of methanol increased, the emission intensity of compound 3 decreased.
This result was attributed to the ESIPT process that was caused by
hydrogen bonding of methanol.
4. Conclusion
A fluorometric paper based sensor array for detecting 15 different
VOCs in the vapor phase with 100% accuracy has been successfully
constructed by using salicylidene derivatives. Principle component
analysis (PCA) of ΔRGB values obtained from images suggested that this
novel fluormetric paper based sensor array has high reproducibility and
Fig. 5. PCA score plot of ΔRGB patterns of the images photographed from exposure of 15 VOC vapors onto fluorometric paper-based sensor arrays of compounds 1, 2,
and 3 (3 sensor × 15 VOCs × 4 measurements).
382

T. Jarangdet et al.
DyesandPigments159(2018)378–383
Fig. 6. Images illustrating the fluorescence responses of compound 3 paper based sensor array to various concentrations of methanol vapor in water.
discriminating ability towards 15 VOCs. Therefore, this portable
fluorometric paper-based sensor array of salicylidene derivatives
showed high potential in analyzing VOC vapors. In addition, compound
3 was also applicable to qualitatively determine the contamination of
methanol in water. Thus, the salicylidene derivatives themselves should
be valuable for the development of electronic tongue for VOC vapors
related food analysis.
2015;113:47–54;
(e) Kim YJ, Gwon SY, Kim SH. Dyes Pigments 2016;133:184–8.
[11] EC. Commission regulation 2870/2000 Laying down community reference methods
for the analysis of sprits drinks. Off J Eur Comm L 2000;333:20–46.
[12] (a) Yoon J, Chae SK, Kim JM. Colorimetric sensors for volatile organic compounds
(VOCs) based on conjugated polymer-embedded electrospun fibers. J Am Chem Soc
2007;129:3038–9;
(b) Suslick KS, Rakow NA, Sen A. Colorimetric sensor arrays for molecular re-
cognition. Tetrahedron Lett 2004;60:11133–8;
(c) Zhao W, Berg AVD. Lab on paper. Lab Chip 2008;8:1988–91;
(d) Soga T, Jimbo Y, Suzuki K, Citterio D. Inkjet-printed paper-based colorimetric
sensor array for the discrimination of volatile primary amines. Anal Chem
2013;85:8973–8;
Acknowledgements
(e) Fraiwan A, Lee H, Choi S. A paper-based cantilever array sensor: monitoring
volatile organic compounds with naked eye. Talanta 2016;158:57–62;
(f) Zhu X, Zhang H, Wu J. Chemiresistive ionogel sensor array for the detection and
discrimination of volatile organic vapor. Sensor Actuator B Chem
2014;202:105–13.
The authors gratefully thank the research fund for the Thailand
Research Fund (MRG6180070). Master of Science would like to ac-
knowledge the financial supports from Faculty of Science, the KMUTT
55th Anniversary Commemorative Fund. M.S. would also like to ac-
knowledge the financial supports from National Nanotechnology Center
(NANOTEC), NSTDA, Ministry of Science and Technology, Thailand,
through its program of Center of Excellence Network.
[13] (a) Dolai S, Bhunia SK, Beglaryan SS, Kolusheva S, Zeiri L, Jelinek R. Colorimetric
polydiacetylene-aerogel detector for volatile organic compounds (VOCs). Appl
Mater Interface 2017;9:2891–8;
(b) Yoon B, Park IS, Shin H, Park HJ, Lee CW, Kim JM. A litmus-type colorimetric
and fluorometric volatile organic compound sensor based on inkjet-printed poly-
diacetylenes on paper substrates. Macromol Rapid Commun 2013;34:731–5.
[14] (a) Saravanakumar D, Devaraj S, Iyyampillai S, Mohandossb K, Kandaswamy M.
Schiff’s base phenol-hydrazone derivatives as colorimetric chemosensors for
fluoride ions. Tetrahedron Lett 2008;49:127–32;
Appendix A. Supplementary data
(b) Li J, Lin H, Cai Z, Lin H. A high selective anion colorimetric sensor based on
salicylaldehyde for fluoride in aqueous media. Spectrochim Acta 2009;72:1062–5;
(c) Dalapati S, Alam MA, Jana S, Karmakar S, Guchhait N. “Test kit” for detection
of biologically important anions: a salicylidene-hydrazine based Schiff base.
Spectrochim Acta 2013;102:314–8;
(d) Mukherjee S, Paul AK, Evans HS. A family of highly selective fluorescent sen-
sors for fluoride based on excited state proton transfer mechanism. Sensor Actuator
B Chem 2014;202:1190–9;
(e) Zang L, Jiang S. Substituent effects on anion sensing of salicylidene schiff base
derivatives: tuning sensitivity and selectivity. Spectrochim Acta 2015;150:814–20;
(f) Erdemir S, Kocyigit O. Reversible “OFF-ON” fluorescent and colorimetric sensor
based benzothiazole-bisphenol A for fluoride in MeCN. Sensor Actuator B Chem
2015;221:900–5.
Supplementary data related to this article can be found at http://dx.
doi.org/10.1016/j.dyepig.2018.06.044.
References
[1] Phillips M, Herrera J, Krishnan S, Zain M, Greenberg J, Cataneo RN. J Chromatogr B
1999;729:75–88.
[2] Tisser R, Young R. 6-The respiratory system. Essential oil safety: a guide for health
care professionals. 2014. p. 99–110.
[3] Natale CD, Macagnano A, Martinelli E, Paolesse R, D’Arcangelo G, Roscioni C,
Finazzi- Agro A, D’Amico A. Biosens Bioelectron 2003;18:1209–18.
[4] (a) Peng G, Trock E, Haick H. Nano Lett 2008;8:3631–5;
(b) Chatterjee S, Castroa M, Feller JF. J Mater Chem B 2013;36:4563–75.
[5] (a) Lin H, Jang M, Suslick KS. J Am Chem Soc 2011;133:16786–9;
(b) Chang Y, Tang N, Qu H, Liu J, Zhang D, Zhang H, Pang W, Duan X. Sci Rep
2016;6:23970–82;
(c) Fair JD, Bailey WF, Felty RA, Gifford AE, Shultes B, Volles LH. J Environ Sci
2009;21:1005–8;
(d) Maeda T, Onodera S, Ogino H. J Chromatogr A 1995;710:51–9.
[6] Silva ML, Malcata FX. J Enol Vitic 1998;49:56–64.
[7] (a) Vesely P, Jusk L, Basarova G, Seabrooks J, Ryder D. J Agric Food Chem
2003;51:6941–4;
(b) Gerogiannaki-Christopoulou M. J Food Technol 2008;6:196–202.
[8] (a) Lorwongtragool P, Sowade E, Watthanawisuth N, Baumann RR, Kerdcharoen T.
Sensors 2014;14:19700–12;
[15] Fery-Forgues S, Lavabre D. Are fluorescence quantum yields so tricky to measure? a
demonstration using familiar stationery products. J Chem Educ 1999;76:1260–4.
[16] Eaidkong T, Mungkarndee R, Phollookin C, Tumcharern G, Sukwattanasinittc M,
Wacharasindhu S. Polydiacetylene paper based colorimetric sensor array for vapor
phase detection and identification of volatile organic compounds. J Mater Chem
2012;22:5970–7.
[17] Surpateanu GG, Landy D, Lungu CN, Fourmentin S, Surpateanu G, Rethoreb C,
Avarvari N. Synthesis and inclusion capability of a β-cyclodextrintetrathiafulvalene
derivative. Tetrahedron Lett 2006;62:9701–4.
[18] AMc Cluskey, Robinson PJ, Hill T, Scottc JL, Edwards JK. Green chemistry ap-
proaches to the Knoevenagel condensation: comparison of ethanol, water and sol-
vent free (dry grind) approaches. Tetrahedron Lett 2002;43:3117–20.
[19] (a) Kotaka H, Konishi GI, Mizuno K. Synthesis and photoluminescence properties
of π-extended fluorine derivatives: the first example of a fluorescent solvatochromic
nitro-group-containing dye with a high fluorescence quantum yield. Tetrahedron
Lett 2010;51:181–4;
(b) Doroshenko AO, Posokhov EA, Verezubova AA, Ptyagina LM, Skripkina VT,
Shershukov VM. Radiationless deactivation of the excited phototautomer form and
molecular structure of ESIPT-compounds. Photochem Photobiol Sci 2002;1:92–9.
[20] Sun W, Li S, Hu R, Qian Y, Wang S, Yang G. Understanding solvent effects on
luminescent properties of a triple fluorescent ESIPT compound and application for
white light emission. J Phys Chem 2009;113:5888–95.
[21] Houjou H, Shingai H, Yagi K, Yoshikawa I, Araki K. Mutual interference between
intramolecular proton transfer sites through the adjoining π-conjugated system in
schiff bases of double-headed, fused salicylaldehydes. J Org Chem
2013;78:9021–31.
[22] Ziolek M, Burdzinski G, Karolczak J. Influence of intermolecular hydrogen bonding
on the photochromic cycle of the aromatic schiff base N,N′-Bis(salicylidene)-p-
phenylenediamine in Solution. J Phys Chem 2009;113:2854–64.
[23] Dalapati S, Alam MA, Jana S, Guchhait N. Naked-eye detection of F- and AcO- ions
by Schiff base receptor. J Fluorine Chem 2011:132536–40.
(b) Arshak K, Moore E, Lyons GM, Harris J, Clifford S. Sens Rev 2004;24:181–98.
[9] (a) Ratnarathorn N, Chailapakul O, Henry CS, Dungchai W. Talanta
2012;99:552–7;
(b) You GR, Jang HJ, Jo TG, Kim C. RSC Adv 2016;6:74400–8;
(c) Lin Q, Chen P, Liu J, Fu YP, Zhang YM, Wei TB. Dyes Pigments 2013;98:100–5;
(d) Wang S, Men G, Zhao L, Hou Q, Jiang S. Binaphthyl-derived salicylidene Schiff
base for dual-channel sensing of Cu, Zn cations and integrated molecular logic gates.
Sensor Actuator B Chem 2010;145:826–31;
(e) Murale DP, Kim H, Choi WS, Churchill DG. Highly selective excited state in-
tramolecular proton transfer (ESIPT)-based superoxide probing. Org Lett
2013;15:3946–9.
[10] (a) Becuwe M, Landy D, Delattre F, Cazier F, Fourmentin S. Fluorescent indolizine-
β-cyclodextrin derivatives for the detection of volatile organic compounds. Sensors
2008;8:3689–705;
(b) Davis BW, Burris AJ, Niamnont N, Hare CD, Chen CY, Sukwattanasinitt M,
Cheng Q. Dual-mode optical sensing of organic vapors and proteins with poly-
diacetylene (PDA)-embedded electrospun nanofibers. Langmuir 2014;30:9616–22;
(c) Ji SM, Gwon SY, Kim SH. Dyes Pigments 2014;108:93–7;
(d) Martini G, Martinelli E, Ruggeri G, Galli G, Pucci A. Dyes Pigments
[24] Hammami M, Rouissi H, Salah N, Selmi H, Otaibi MA, Blecker C, Karoui R.
Fluorescence spectroscopy coupled with factorial discriminant analysis technique to
identify sheep milk from different feeding systems. Food Chem 2010;122:1344–50.
383