Metal ion chemosensors based on diaza-18-crown-6 coupling with azobenzene dye

Article abstract of DOI:10.1016/j.saa.2019.02.052

Novel naked-eye cation-selective chemosensors were developed from derivatives of diaza-18-crown-6 bearing one or two azobenzenes functionalized with o-acetic acid or p-nitro groups. The chromogenic acid and nitro derivatives exhibited solvatochromic effects with a majority of quinone-hydrazone tautomer in DMSO. The study of sensing properties of all molecules towards several metal ions showed that the one and two nitro-bearing N1 and N2 derivatives responded selectively to Cr³⁺ in MeCN with a color change from pink to colorless. They also showed selectivity for Hg²⁺ in DMSO with a color change from purple to light yellow. The detection limit of N1 and N2 that render visual difference to the naked eye are 0.21 ppm and 0.31 ppm for Cr³⁺ in MeCN, and 1.20 ppm and 1.00 ppm for Hg²⁺ in DMSO, while that obtained from 3σ method are 0.08 ppm and 0.03 ppm for Hg²⁺ in DMSO. These results demonstrate a potential use of p-nitro azobenzene diaza-18-crown-6 as naked-eye sensors, notably for Cr³⁺ that brings acute toxic to aquatic life.

Full text of DOI:10.1016/j.saa.2019.02.052

Accepted Manuscript
Metal ion chemosensors based on diaza-18-crown-6 coupling
with azobenzene dye
Narongsak Koonrugsa, Saowarux Fuangswasdi
PII:
S1386-1425(19)30170-2
https://doi.org/10.1016/j.saa.2019.02.052
SAA 16828
DOI:
Reference:
Spectrochimica Acta Part A: Molecular and Biomolecular
Spectroscopy
To appear in:
Received date: 7 September 2018
Revised date:
Accepted
date:
28 December 2018
1
6 February 2019
Please cite this article as: N. Koonrugsa and S. Fuangswasdi, Metal ion chemosensors
based on diaza-18-crown-6 coupling with azobenzene dye, Spectrochimica Acta Part A:
Molecular and Biomolecular Spectroscopy, https://doi.org/10.1016/j.saa.2019.02.052
This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.

ACCEPTED MANUSCRIPT
1
Metal ion chemosensors based on diaza-18-crown-6 coupling with
azobenzene dye
Narongsak Koonrugsa , Saowarux Fuangswasdia,
a
a
Supramolecular Chemistry Research Unit, Department of Chemistry, Faculty of Science, Chulalongkorn University,
Bangkok, 10330, Thailand
Novel naked-eye cation-selective chemosensors were developed from derivatives of diaza-18-
crown-6 bearing one or two azobenzenes functionalized with o-acetic acid or p-nitro groups.
The chromogenic acid and nitro derivatives exhibited solvatochromic effects with a majority
of quinone-hydrazone tautomer in DMSO. The study of sensing properties of all molecules
towards several metal ions showed that the one and two nitro-bearing N1 and N2 derivatives
3
+
responded selectively to Cr in MeCN with a color change from pink to colorless. They also
2
+
showed selectivity for Hg in DMSO with a color change from purple to light yellow. The
detection limit of N1 and N2 that render visual difference to the naked eye are 0.21 ppm and
3
+
2+
0
.31 ppm for Cr in MeCN, and 1.20 ppm and 1.00 ppm for Hg in DMSO, while that
2
+
obtained from 3σ method are 0.08 ppm and 0.03 ppm for Hg in DMSO. These results
demonstrate a potential use of p-nitro azobenzene diaza-18-crown-6 as naked-eye sensors,
3
+
notably for Cr that brings acute toxic to aquatic life.
Keywords: Naked-eye sensor; Diaza-18-crown-6; Azobenzene; Tautomerization;
Chromium(III) ion

ACCEPTED MANUSCRIPT
2
1
. Introduction
The design and synthesis of new chemosensors for metal ions, notably toxic ones,
have received much attention especially in analytical and environmental science because of
their advantages such as high sensitivity, low operational cost, and practical use. Cation or
anion sensors are generally composed of binding unit which recognizes ion or guest molecule
and signaling unit or chromophore which gives signal changes, such as electrochemical or
optical signal, upon binding [1-5]. One of the desired properties for chemosensors is a distinct
optical change that is visible to the naked eye.
Azacrown ethers have been extensively used as receptors for metal ions especially
transition and heavy metal ions due to the presence of two nitrogen atoms in the macrocyclic
ring which strongly interact with these ions [1, 6]. This type of ligands however does not offer
human eye detection upon complexation. To be used as chemosensor, azacrown ether must be
modified with signaling unit.
Being an intensely colored moiety, azobenzene groups have been universally
employed as signaling unit [2, 7-10]. Many azobenzene derivatives show sensing behavior as
a result of structural change induced by interactions, for example, calixarenes containing
azobenzene show hypsochromic effect (blue shift) while their quinone-hydrazone analogues
give bathochromic effect (red shift). Azobenzene can be functionalized with various groups,
for example, addition of ester and carboxyl groups at ortho position to azo group which can
stabilize the quinone-hydrazone form [11-16].
This work is interested in developing ion-selective chemosensors to monitor heavy
metal ions. Novel ligands consisting of diaza-18-crown-6 as binding unit and functionalized
azobenzene, i.e., o-carboxyl and p-nitro group as signaling unit were synthesized and
characterized. Their complexation in various solvents (MeOH, DMSO, MeCN, and water)

ACCEPTED MANUSCRIPT
3
with several types of metal ions (alkali, alkaline earth, transition, and heavy metal) was
carried out using color observation and UV-Visible spectrophotometry.
2
. Experimental
2
.1. Material
+
+
+
2+
2+
3+
2+
All reagents for synthesis, metal nitrate salts (Li , Na , K , Mg , Ca , Cr , Co ,
2
+
2+
2+
+
2+
2+
3+
2+
2+
Ni , Cu , Zn , Ag , Cd , and Pb ), and chloride salts (Fe , Fe and Hg ) for sensing
study were purchased from Merck, Sigma-Aldrich, and TCI and used without further
purification. All analytical grade salts and ligands were vacuum dried before use.
Tetrabuthylammonium hexafluorophosphate from Fluka was used as supporting electrolyte.
All solutions were prepared in spectroscan-grade solvents (MeOH, MeCN and DMSO) from
LAB-SCAN and Merck.
2
.2. Instruments
1
13
H and C NMR spectra were recorded at room temperature with Varian 400 and
Bruker 500 instruments operating at a frequency of 400 and 500 MHz, respectively. UV-Vis
absorption measurement was acquired on a Varian Cary 50 Probe UV-Vis spectrophotometer
using a 1 cm path length cuvette at 25 ºC with a Julabo F33 temperature controller. Mass
spectra (MS) were obtained with a matrix-assisted laser desorption ionization mass
spectrometer (MALDI-MS) using 2-cyano-4-hydroxycinnamic acid (CCA) or dithranol as a
matrix. Infrared spectra were recorded on a Nicolet Impact 412 FT-IR spectrometer.
2
.3. Color change and UV-Vis spectroscopic properties
The concentrations of ligand stock solutions and metal stock solutions were
−3
–1
–1
approximately 1×10 mol·L and 0.02 mol·L , respectively. The exact concentration of

ACCEPTED MANUSCRIPT
4
metal salt stock solutions was determined by complexometric titration with EDTA. The ionic
−1
strength of all solutions was kept constant at 0.1 mol·L
using tetrabuthylammonium
hexafluorophosphate. The spectral change after complexation was recorded after addition of 5
−5
−4
eq of metal ion solution into ligand solution of 1×10 – 1×10 M.
2
.4. Competitive assay
equivalents of Cr or Hg were added to a mixture of 20 M of N1 or N2 and
3
+
2+
5
another cation in MeCN or DMSO. UV-Vis spectra before and after competition were
recorded and the absorbances at the wavelength of maximum absorption around 370 or 390
nm were then compared.
2
.5. The detection limit from visual color change
Photographs of mixtures of sensor solution with varied equivalents of target ion (1-10
−6
−6
eq) were taken. The sensor concentrations were either 1×10 or 2×10 M.
3
. Synthesis and characterization
Syntheses of o-acetic and p-nitro azobenzene diazacrown ether derivatives A2, N1,
and N2 are shown in Scheme 1.
<<Scheme 1>>
3
.1. Synthesis of receptors A2
3
.1.1 Synthesis of trietyleneditosylate 1
Triethyleneglycol (26 mmol), triethylamine (78 mmol), and DMAP was stirred in
dichloromethane for 30 minutes. Anhydrous solution of tosyl chloride (52 mmol) was added
into the mixture, and the stirring continued for 12 hours. After water-extraction, CH2Cl2 was

ACCEPTED MANUSCRIPT
5
removed by rotary evaporator and the crude product was recrystallized in methanol. The
resulting solid was filtered and dried (7.643 g, 64 %yield). Melting point: 82-83 °C. Rf
1
(
CH2Cl2): 0.78. H-NMR (CDCl3, TMS):  1.57 (s, H2O),  2.44 (s, 6H, CH3-Ar),  3.53 (s,
4
H, CH2-CH2-OTs),  3.65 (t, 4H, CH2-CH2-O),  4.13(t, 4H, CH2-CH2-O),  7.34 (d, 4H,
1
3
Ar),  7.79 (d, 4H, Ar) ppm. C NMR (CDCl3, TMS):  129.4, 118.7, 116.4, 71.3, 52.8, 48.2,
+
4
4
7.3 ppm. IR: 1509.28 [ν(C=C)aromatic]. MALDI-MS C20H26O8S2 found m/z [M+Na]
81.255, calcd 458.11.
3
.1.2 Synthesis of diazacrown ether 2
Trietyleneditosylate (9.65 mmol), ethylenedioxy amine (9.65 mmol), and Na2CO3 (96
mmol) were dissolved in anhydrous acetonitrile. After refluxing for 3 days, the reaction
mixture was filtered and acetonitrile was removed by rotary evaporator. The crude product
was purified by column chromatography on alumina using 1:9 ethanol:THF as eluent which
was later removed by rotary evaporator. The resulting was white solid (0.658 g, 26 %yield).
1
Melting point: 115-116 °C. H-NMR (CDCl3, TMS):  1.42 (s, H2O),  2.20 (s, 2H, NH), 
1
3
2
6
.79 (m, 8H, NH-CH2-CH2),  3.59 (m, 16H, -OCH2-) ppm. C NMR (CDCl3, TMS):  77.5,
+
9.8, 48.8 ppm. IR: 3364.19 [ν(N-H)]. MALDI-MS C12H26N2O4 found m/z [M] 262.979,
calcd 262.19.
3
.1.3 Synthesis of diazacrown phenol 3
Diazacrown phenol 3 was synthesised according to the procedure reported previously
17]. The mixture of diazacrown (2 mmol), phenol (4.80 mmol), and paraformaldehyde (4.90
[
mmol) in toluene was refluxed for 12 hours and the solvent was evaporated under vacuum. A
small amount of methanol was added to help the crystallisation. After sonication for 30
minutes, the resulting solid was filtered and dried (0.135 g, 14 %yield). Melting point: 93-94

ACCEPTED MANUSCRIPT
6
1
°C. Rf (20% MeOH/CH
2
Cl
2
): 0.83. H-NMR (CDCl3, TMS): 2.86 (m, 8H, NCH2CH2),  3.61
(m, 8H, -CH2CH2O-),  3.67 (m, 8H, -OCH2CH2-),  3.81 (m, 4H, NCH2Ar),  6.60-7.16 (m,
1
3
8
H, Ar) ppm. C NMR (CDCl3, TMS):  144.9, 133.0, 129.8, 127.9, 70.7, 69.2, 68.7, 21.6
ppm. IR: 3266.21 [ν(O-H)], 1567.66 [ν(C=C)aromatic]. MALDI-MS C26H38N2O6 found m/z
+
[
M+H] 475.139, calcd 474.27.
3
.1.4 Synthesis of azo-diaza-18-crown-6 ester derivatives 4
The aqeuous solution of ethyl-2-aminobenzoate (2 mmol), NaNO2 (2.1mmol), and
HCl was stirred at 0 °C for 15 minutes before addition of diazocrown phenol (1.05 mmol) in
5 ml of THF:pyridine (5:2). After stirring at 0 °C for 24 hours, the mixture was extracted by
2
CH2Cl2 which was evaporated later. The crude products were purified and separated by
chromatography on silica gel using 10% of methanol in dichloromethane as eluent. After
evaporation, the separated product 4 appeared as orange oil (0.154 g, 24 %yield). Rf (20%
1
MeOH/CH
2
Cl
2
): 0.93. H-NMR (CDCl3, TMS):  1.30 (t, 6H, OCH2CH3),  2.86 (m, 8H,
NCH2CH2),  3.62 (m, 8H, -CH2CH2O-),  3.69 (m, 8H, -OCH2CH2-),  3.90 (m, 4H,
NCH2Ar),  4.36 (m, 4H, OCH2-),  6.75-7.30 (m, 6H, Ar-OH),  7.50-7.85 (m, 6H, Ar-CO-)
1
3
ppm. C NMR (CDCl3, TMS):  164.8, 159.4, 152.8, 140.3, 131.4, 126.6, 126.4, 125.8,
1
23.7, 123.5, 122.6, 115.4, 113.0, 69.4, 68.7, 62.8, 59.6, 56.2, 16.2 ppm. IR: 3267.25 [ν(O-
H)], 1559.77 [ν(C=O)], 1400.82 [ν(C=C)aromatic]. MALDI-MS C44H54N6O10 found m/z
+
[
M+H] 827.579, calcd 826.93.
3
.1.5 Synthesis of azo-diaza-18-crown-6 acid derivatives A2
Azo-diaza-18-crown-6 ester derivatives 4 (0.2 mmol) and NaOH (0.4 mmol) were
dissolved in 1:1 EtOH:water and refluxed for 3 hours. After that, the solution’s pH was
adjusted to 7 by 3 M HCl and the mixture was extracted by 20 ml of CH2Cl2 3 times. After

ACCEPTED MANUSCRIPT
7
CH2Cl2 was evaporated, the crude product was purified by chromatography on silica gel using
0% of methanol in dichloromethane as eluent. The resulting product (A2) appeared as
2
orange solid after evaporation (0.092 g, 60 %yield). Melting point: 205-207 °C. Rf (20%
1
MeOH/CH2Cl2): 0.70. H-NMR (CDCl3, TMS):  2.86 (m, 8H, NCH2CH2),  3.62 (m, 8H, -
CH2CH2O-),  3.69 (m, 8H, -OCH2CH2-),  3.81 (m, 4H, NCH2Ar),  7.71-7.91 (m, 6H, Ar-
1
3
OH),  8.24-8.38 (m, 8H, Ar-CO-) ppm. C NMR (CDCl3, TMS):  163.7, 160.4, 133.8,
1
39.3, 134.5, 127.6, 127.4, 124.3, 123.8, 123.5, 122.6, 115.4, 114.6, 69.7, 69.1, 59.2, 57.2
ppm. IR: 3243.50 [ν(O-H)], 1542.11 [ν(C=O)], 1400.82 [ν(C=C)aromatic]. MALDI-MS
+
C40H46N6O10 found m/z [M+H] 771.586, calcd 770.33.
3
.2. Synthesis of receptors N1 and N2
The p-nitroaniline (2 mmol), NaNO2 (2.1 mmol) and HCl were dissolved in water and
stirred at 0 °C for 15 minutes before addition of diazocrown phenol (1.05 mmol) in 20 ml of
Na2CO3 solution. After stirring at 0 °C for 24 hours, the mixture was extracted by CH2Cl2
which was then removed by rotary evaporator. The crude product was purified and separated
by chromatography on silica gel using 20% of methanol in dichloromethane as eluent. The
first product (N2) appeared after evaporation as dark brown solid (0.132 g, 17 %yield).
1
Melting point: 122-123 °C. Rf (20% MeOH/CH2Cl2): 0.89. H-NMR (CDCl3, TMS):  3.51
(t, 8H, -CH2CH2O-),  3.62 (m, 8H, -CH2CH2O-),  3.78 (m, 8H, -CH2CH2O-),  3.91 (m,
1
3
4
H, NCH2Ar),  7.79 (m, 4H, Ar-NO2),  7.82 (m, Ar-OH),  8.25 (m, 6H, Ar-OH) ppm. C
NMR (CDCl3, TMS): 165.3, 157.1, 153.2, 144.5, 126.7, 124.0, 122.9, 122.7, 119.3, 113.8,
7
0.3, 69.1, 57.9, 55.5 ppm. IR: 1515.62 [ν(C=C)aromatic]. MALDI-MS C38H44N8O10 found m/z
+
[
M] 772.432, calcd 772.32.
The dark brown solid was also obtained after evaporation of the second product (N1)
1
(
0.182 g, 29 %yield). Melting point: 134-135 °C. Rf (20% MeOH/CH2Cl2): 0.77. H-NMR

ACCEPTED MANUSCRIPT
8
(CDCl3, TMS):  2.81 (t, 4H, -NCH2CH2-),  3.07 (dt, 4H, OCH2CH2),  3.41 (d, 4H, -
CH2CH2O-),  3.44 (m, 4H, -OCH2CH2-),  3.60 (m, 4H, -OCH2CH2-),  3.68 (m, 4H,
NCH2CH2-),  3.75 (m, 2H, NCH2Ar),  3.98 (m, 2H, NCH2Ar),  6.34-7.01 (m, 4H, Ar-OH),
1
3

1
1
7.78, 8.35 (m, 4H, Ar-NO2),  7.83 (m, 3H, Ar-OH) ppm. C NMR (CDCl3, TMS):  164.3,
56.2, 153.4, 152.3, 145.4, 130.2, 128.7, 125.8, 125.4, 124.9, 123.0, 122.9, 122.8, 119.3,
15.6, 113.8, 71.6, 71.5, 68.9, 68.3, 59.1, 58.3, 56.7, 56.3 ppm. IR: 1584.24 [ν(C=C)aromatic].
+
MALDI-MS C32H41N5O10 found m/z [M] 623.245, calcd 623.70.
4
. Sensing Properties
4
.1. Absorption of azo-diaza-18-crown-6 derivatives
*
Azobenzene compounds usually show two absorption bands, i.e., π→π transition
*
band of the azo group (-N=N-) at 360-400 nm and n→π transitions of the hydrazone group
(
-NH-N) at 500-576 nm. These bands correspond to the tautomerization between azobenzene
and quinone-hydrazone as shown in Figure 1 [11-16].
<Figure 1>>
<
The wavelengths at maximum absorption (λmax) of all receptors in different solvents
were presented in the first row of Table 1. Compounds A2, N1 and N2 exhibited
solvatochromic properties in all solvents. In water, A2 displayed only one band with
maximum absorption at 380 nm corresponding to π-π* transition in azobenzene form [11-13].
However, shoulder band at 510 nm was observed in addition to λmax at 360 nm in the cases of
A2 in MeOH and MeCN. The absorption at 510 nm corresponding to quinone-hydrazone
form became a major band when A2 was dissolved in DMSO. These results could be
explained by interaction between A2 and solvents: water, MeOH and MeCN could interact
with hydrogen atom in carboxylic group, resulting in azobenzene form. However, DMSO
could not interact with proton in acid group and an intramolecular H-bonding between acid

ACCEPTED MANUSCRIPT
9
and nitrogen could occur leading to quinone-hydrazone tautomer. The solvatochromic effects
were also observed in N1 and N2: major quinone-hydrazone band and shoulder azobenzene
band in aprotic solvents (DMSO, MeCN) were seen while only one strong band of
azobenzene was observed in MeOH. The difficulty of quinone-hydrazone formation in protic
solvent could be ascribed to H-bonding between MeOH and lone pair electrons in azo group
[15, 18]. In short, the tautomerization of ligands to quinone-hydrazone depends on the
distribution of electron density between hydroxyl group and substituent groups which
increase in order of p-nitro > o-acid.
4
.2. Complexation of azo-diaza-18-crown-6 derivatives with metal ions
The sensing properties of all receptors towards different cations were investigated
using UV-Vis absorption spectrophotometry and color observation. The wavelengths at
maximum absorption (λmax) after addition of 5 equivalents of metal ions were shown in Table
1
with selected values of large displacement of wavelength at maximum absorption (Δλmax)
more than 100 nm given in parentheses. Note that a negative sign in Δλmax indicates a shift to
lower wavelength or blue shift.
4
.2.1. Complexation of azo-diaza-18-crown-6 acid derivatives (A2) with metal ions
Among three synthesized receptors, only acid derivative A2 is soluble in water. The
color of A2 and various metal ions in aqueous solution, together with their absorption
3
+
2+
spectrum showed that addition of Cr and Hg to A2 could turn its yellow color to colorless
(Figure S1). However, this color change was associated with one absorption band with a
small blue shift of only 35 nm which was impractical for naked eye detection. In methanol,
however, the changes in absorption spectra in the presence of metal ions were more obvious
2
+
2+
2+
(
Figure 2). Worth mentioning are Ca , Cu , and Cd : a small red shift of azobenzene band
(Δλ 40-50 nm) accompanied by an increasing of absorption intensity of quinone-hydrazone

ACCEPTED MANUSCRIPT
1
0
band, sometimes with a small blue shift, was recorded. The quinone-hydrazone band
becoming major band has been reported before in complexation of o-ester
diazophynylcalix[4]arene [12]. To explain the interaction of these metal ions, a UV-Vis
titration of A2 with NaOH was performed. The absorption after addition of 5 eq of base to
deprotonate acidic group showed only one peak at 400 nm with a small red shift of 40 nm.
2
+
2+
2+
Using this observation, it is likely that complexation of A2 with Ca , Cu , and Cd not only
induces some deprotonation of azobenzene but also tautomerization to quinone-hydrazone.
2
+
2+
This suggests that large cations like Ca and Cd might interact with oxygen in acid group
2
+
while a small ion like Cu , known to prefer nitrogen to oxygen donor atom, possibly binds
with lower N in azo group close to diaza-18-crown-6.
<<Figure 2>>
Data in Table 1 also show that complexation of A2 in DMSO with transition and
heavy metal ions are stronger than that with alkali and alkaline earth ions. A large blue shift
of up to 165 nm with an obvious color change from orange to yellowish or greenish is usually
observed as presented in Figure 3. Regrettably even if the displacement of λmax of A2 and
metal ions in DMSO is much larger than that observed in other solvents, the selectivity could
not be realized as the colors are indistinguishable to human eye.
<<Figure 3>>
4
.2.2. Complexation of azo-diaza-18-crown-6 nitro derivatives (N1, N2) with metal ions
Similar to A2, alkali and alkaline earth ions do not induce any significant change
regardless of solvent type (Table 1). Nevertheless, complexation of nitro derivatives with
transition and heavy metal ions did depend on solvent type resulting in desirable selectivity.
When treating N1 and N2 with metal ions in MeOH, a small shift of less than 50 nm was

ACCEPTED MANUSCRIPT
1
1
observed and the color change was not apparent. In contrast, large blue shifts were seen in
MeCN and DMSO. The colors and selected UV-Vis absorption spectra of N1 and N2 when
3
+
Δλmax > –100 nm are given in Figure 4. The evident blue shift of 155 nm in the case of Cr in
MeCN led to an obvious color change form pink to light yellow. Their absorption spectrum
showed one characteristic peak of azobenzene: λmax = 365 nm for N1 and 370 nm for N2.
Similar observation, i.e., large spectral and pronounced color change, was also observed for
N1 and N2 in DMSO. The nitro receptors yielded a large blue shift of around 165 nm and
2
+
turned purple solution to colorless upon addition of Hg .
<<Figure 4>>
3
+
The large blue shift observed in the cases of N1 and N2 with Cr in MeCN (520 to
2
+
3
65 nm and 525 to 370 nm) and with Hg in DMSO (560 to 390 nm and 555 to 390 nm)
allowed a ratiometric absorption response in the presence of 5 eq of individual metal ion
3
+
2+
(
Figure 5). The selectivity of these sensors towards Cr in MeCN and Hg in DMSO using
2
+
3+
absorbance ratio is obvious: only Cu could interfere with the detection Cr of N1 in MeCN.
2
+
N1 and N2 show an impressive selectivity towards Hg in DMSO with a ratiometric
absorbance ratio of 52 and 64, respectively, which is much larger than those of previous
ratiometric sensors like oxazine-thione (19) and quinoline-based sensor (4.1) [19, 20].
<<Figure 5>>
3
+
2+
Considering that N1 and N2 bound selectively with Cr in MeCN and Hg in DMSO
with both ions induced azobenzene formation, solvent type must play an important role.
3
+
However, since the size of these metal ions are largely different, Cr actually being the
smallest ion studied, these metal ions should interact with the ligands differently. It is likely
3
+
that Cr interacts not only with crown ether moiety but also with oxygen atom in hydroxyl

ACCEPTED MANUSCRIPT
1
2
group which prevents n-π* transition in quinone-hydrazone form and tautomerization to
azobenzene is preferred [15]. To proof this assumption, a UV-Vis titration of NaOH into a
3
+
mixture of Cr (5 eq) and N1 or N2 derivative was performed. The characteristic peaks of
3
+
quinone-hydrazone, fully observed after addition of base 20 eq, suggest that Cr could not
–
bind with O in the deprotonated ligand.
2
+
2+
For Hg , an increase in the absorption band at 390 nm after complexation with Hg
2
+
which is much higher in N2 than in N1 suggests that a larger Hg could interact with both
3
+
nitrogen atoms in azo groups of N2. This increment is not seen with Cr in MeCN, because
2
+
both N1 and N2 contain 2 hydroxyl groups. The selectivity for Hg in DMSO is likely due to
2
+
the interactions of Hg with the solvent.
3
+
2+
4
.2.3. Effect of pH on colors of N1 and N2 in the presence of Cr or Hg
Both N1 and N2 showed different colors depending on pH as can be seen in Figures 6
and S2. In acetonitrile, both sensors were pale yellow at pH 1-4, then they became more red
as the pH increased with N1 being purple at pH 14. The light yellow were also observed at pH
1
-4 in DMSO and both sensors continuously changed from violet (N1) or purple (N2) to
intense blue along with increasing pH. At low pH, the absorption spectra showed only the
+
characteristeic peak of azobenzene, suggesting a protonation of H on oxygen in hydroxyl
group which allows π-π* transition and preferable quinone-hydrazone to azobenzene
tautomerization [15]. At high pH, the more intense red in MeCN and blue in DMSO observed
were a result of a small blue shift of increasing quione-hydrazone intensity. This suggests that
−
OH could deprotonate hydrogen from hydroxyl group and induce tautomerization from
azobenzene to quinone-hydrozone.
<<Figure 6>>

ACCEPTED MANUSCRIPT
1
3
3
+
Figure 6 also demonstrated that the presence of Cr in MeCN induced obvious color
change from red to pale yellow at pH 5-10 for both sensors, notably at pH 10. The similar
2
+
observation was also seen when Hg was present in DMSO where both receptors change
from violet or purple to yellow at pH 5-10 as shown in Figure S2. A very large blue shift was
3
+
observed in the cases of N1 and N2 with Cr in MeCN (520 to 365 nm and 525 to 370 nm)
2
+
and with Hg in DMSO (560 to 390 nm and 555 to 390 nm). These results show that N1 and
3
+
2+
N2 could detect Cr and Hg at pH 5-10.
4
.4. Selectivity: competitive assay
Since N1 and N2 offer distinct color change in the presence of Cr in MeCN and
3
+
2
+
Hg DMSO, the selectivity of these compounds towards specific ions was explored. A
competitive assay by adding ion of interest to a mixture of sensor and another metal ion was
carried out and the absorbances at either 365, 370 or 390 nm corresponding to λmax of target
ion complex before and after competition were compared. The results, shown in Figure 7,
3
+
2+
demonstrate a consistent selectivity of sensors N1 and N2 towards Cr in MeCN and Hg in
3
+
2+
DMSO. Addition of target ion Cr or Hg to a mixture of ligand and another metal ion
3
+
2+
yielded absorption spectrum similar to that of ligand with Cr or Hg alone, with an
2
+
exception of Cu in DMSO where absorbance around 390 nm increased in an almost linear
fashion as a result of copper absorption in this region. The color change to colorless induced
3
+
2+
by addition Cr and Hg even if the receptors already bound with another metal ion as seen
by orange tint color means that the complexation with these two ions are much stronger. In
3
+
2+
other words, N1 and N2 are selective towards Cr in MeCN and Hg DMSO.
<<Figure 7>>
4
.5. Stability constants of complexes between azo-diaza-18-crown-6 nitro derivatives (N1,
3
+
2+
N2) and target metal ions (Cr , Hg )

ACCEPTED MANUSCRIPT
1
4
3
+
2+
To get a deeper insight into the complexation of Cr and Hg with N1 and N2
receptors, spectrometric titrations were performed and the apparent stability constants were
refined from these data by a nonlinear least square algorithm using the program Sirko [21]. In
3
+
3+
the case of Cr , the spectral changes of N1 in MeCN with Cr were not repeatable and could
not be refined because of the ligand’s instability in the open air. However, a formation of two
3
+
3+
complexes is evident for both N1 and N2 with Cr . For instance, addition of Cr into N2
induced a progressive blue shift and decreasing of quinone-hydrozone band (525 nm) with an
increasing of azo band (400 nm) when the metal to ligand ratio was up to 0.5 eq (Figure 8).
After that, the quinone-hydrozone band disappeared and the azo band shifted slightly towards
a shorter wavelength, yiedling one isosbestic point at 390 nm. These results suggest a
formation of two complexes (ML2 and ML) whose refined apparent stability constants in
logarithmic units, average from at least 3 separate determinations with the standard deviation
2
+
σn−1, are equal to 11.370.01 and 5.700.01, respectively. On the contrary, addition of Hg
until 10 eq caused only a decreasing of quinone-hydrozone band and an increasing of azo
band without any λmax shift. This resulted in one isosbestic point around 450 nm and a
formation of only one complex, usually a simple ML. However, a refinement of ML model in
2
+
the system of N1 and Hg was inacceptable (Rf factor > 12%) but that of ML2 was much
improved (Rf factor < 0.9%) with a good curve-fitting throughout the entire data range and
2
+
the stability constant (log10 ML) of 10.530.04 was obtained. For N2 + Hg , the best
refinement corresponded to the ML model with the apparent stability constant in logarithmic
units of 6.020.01.
<<Figure 8>>
4
.6. Detection limit of azo-diaza-18-crown-6 nitro derivatives (N1, N2)

ACCEPTED MANUSCRIPT
1
5
3
+
2+
The detection limit (LOD) of N1 and N2 for Cr and Hg were calculated using three
times of the standard deviation of a blank signal over the slope between intensity versus metal
3
+
ion concentration (Figure S3). In the case of N2 and Cr , the linear range was found only in
−6
−5
2+
the range of 5.0×10 to 2.0×10 M, and LOD was not determined. For N1 and Hg , the
2
+
absorbance at 390 nm showed a good linear relationship with concentration of Hg ranging
−6
−7
from 0 to 6.0×10 M and the detection limit was 3.76×10 M (0.08 ppm). Similarly, N2 also
2
+
–5
displayed a good linear response for Hg in the range of 0 to 1.6×10 M with the detection
–
7
limit of 1.63×10 M (0.03 ppm).
The colors of N1 and N2 at varied concentration were shown in Figure S4. A solution
−6
−6
of 1×10 M was chosen as acceptable for naked eye detection for all systems, except 2×10
M for N1 in DMSO. Photographs of the sensor solutions containing various concentrations of
3
+
2+
target Cr and Hg ions (1-10 eq), compared in Figure 9, indicate that the lowest
3
+
2+
concentrations of Cr and Hg in organic solution that can eliminate the color of the sensors
3
+
−6
and are visible to the naked eye are: a) N1 with Cr = 4.0×10 M (0.21 ppm), b) N2 with
3
+
−6
2+
−6
2+
Cr = 6.0×10 M (0.31 ppm), c) N1 with Hg = 6.0×10 M (1.2 ppm), and d) N2 with Hg
−6
2+
=
5.0×10 M (1.0 ppm). Although the detection limits for Hg are rather high but those for
3
+
Cr are lower than 0.57 ppm which is a value recommended by United States Environmental
Protection Agency (EPA) for freshwater criterion maximum acute concentration for aquatic
3
+
life. This means that N2 could be used to monitor a critical concentration of Cr in
environment.
To explore the possibility of using the nitro-derivatives to detect metal ions in water,
the effect of water contents on the colors of N1 and N2 in mixed H2O/MeCN or H2O/DMSO
was also studied by varying the percentage of water up to 30% (Figure S5). Both compounds
gradually changed to yellow or orange shade with less intensity when the water percentage
increased. However, their color in 5% H2O/organic solvents were similar to those in pure

ACCEPTED MANUSCRIPT
1
6
MeCN or DMSO, suggesting a possibility of studying these receptors when the water
contents did not exceed at 5%. This was proved right as addition of 5 eq of target metal ions
into the mixed H2O/organic solution of N1 and N2 showed an almost identical color change
in pure organic solvents and in 5% H2O/organic solutions. At higher water percentage,
however, less pronounced color change was observed.
3
+
2+
A potential sensing of Cr and Hg in environment was demonstrated by performing
−4
visual detection similar to Figure 9 but using 0.1-1×10 M aqueous metal solutions instead
3
+
2+
(
Figure S6). The lowest concentrations of aqueous Cr and Hg that render colorless
solution were exactly identical to what was found using metal ion in organic solvents, i.e.
3
+
2+
0
.21 (N1) and 0.31 (N2) ppm for Cr , and 1.2 (N1) and 1.0 (N2) ppm for Hg .
3
+
2+
The sensing properties of N1 and N2 for Cr and Hg are compared with some
3
+
previously reported chemosensors in Table 2. N1 and N2 show a detection limit for Cr in
the same concentration level compared with quinoline coupling hydroxynaphthalene [22] and
antipyrine Schiff base chemosensors [23]. However, N1 and N2 offer a more noticeable
visual change from purple to yellow compared to the two mentioned chemosensors which
2
+
change from yellow to colorless. N1 and N2 also show a similar detection limit for Hg
compared to azo dyes featuring pyrene [24], bodipy-derived probe [25] and nitophynyl-
aminothiourea [26]. Even if the LOD from this work’s chemosensors is not as low as what
obtained from the azo thioaxetoals [27], the color change from pink to pale yellow is more
easily to observe than yellowish to colorless by other sensors.
<<Figure 9>>
5
. Conclusions
A series of chemosensor was developed by coupling diaza-18-crown-6 with
functionalized azobenzene i.e., ortho-carboxyl and para-nitro group. The chromogenic acid

ACCEPTED MANUSCRIPT
1
7
(A2) and nitro derivatives (N1 and N2) exhibited solvatochromic properties. The
3
+
chemosensors bearing nitro substituents are capable of selective sensing towards Cr in
2
+
MeCN and Hg in DMSO even in the presence of another metal ion. The visual detection
3
+
limits of N1 and N2 are 0.21 ppm and 0.31 ppm for Cr in MeCN and 1.20 ppm and 1.00
2
+
3+
ppm for Hg in DMSO, respectively. The low detection limit for Cr could be used to
3
+
monitor Cr that bring acute toxic to aquatic life.
Acknowledgments
The authors would like to express their gratitude towards The Thailand Research Fund
and Commission on Higher Education (RTA5380003) and the Ratchadaphisaksomphot
Endowment Fund, Chulalongkorn University, for partial financial support of this research.
References
[1] P.-E. Danjou, J. Lyskawa, F. Delattre, M. Becuwe, P. Woisel, S. Ruellan, S. Fourmentin,
F. Cazier-Dennin, New fluorescent and electropolymerizable N-azacrown carbazole as a
selective probe for iron (III) in aqueous media, Sens. Actuators, B, 171–172 (2012) 1022-
1
028.
2] K. Kaur, A. Singh, N. Singh, Azo dye coupled imine linked dipodal chemosensor: Anion
recognition with counter anion displacement assay, Sens. Actuators, B, 191 (2014) 734-740.
3] J.Y. Kim, G. Kim, C.R. Kim, S.H. Lee, J.H. Lee, J.S. Kim, UV Band Splitting of
Chromogenic Azo-Coupled Calix[4]crown upon Cation Complexation, J. Org. Chem., 68
2003) 1933-1937.
4] K. Kurzątkowska, S. Sayin, M. Yilmaz, H. Radecka, J. Radecki, Calix[4]arene derivatives
as dopamine hosts in electrochemical sensors, Sens. Actuators, B: Chemical, 218 (2015) 111-
21.
[
[
(
[
1

ACCEPTED MANUSCRIPT
1
8
[5] K. Rezaeian, H. Khanmohammadi, V. Arab, Rational design of a novel azoimine
appended maleonitrile-based Salen chemosensor for rapid naked-eye detection of copper(II)
ion in aqueous media, Spectrochim. Acta, Part A, 151 (2015) 848-853.
[6] H. Zhou, J. Wang, Y. Chen, W. Xi, Z. Zheng, D. Xu, Y. Cao, G. Liu, W. Zhu, J. Wu, Y.
Tian, New diaminomaleonitrile derivatives containing aza-crown ether: Selective, sensitive
and colorimetric chemosensors for Cu(II), Dyes Pigm., 98 (2013) 1-10.
[
7] T. Aksungur, Ö. Arslan, N. Seferoğlu, Z. Seferoğlu, Photophysical and theoretical studies
on newly synthesized N,N-diphenylamine based azo dye, J. Mol. Struct., 1099 (2015) 543-
50.
8] E. Hrishikesan, P. Kannan, Azobenzene chemosensor based on nitrogen chelator for the
detection of Cu (II) ion in aqueous medium, Inorg. Chem. Commun., 37 (2013) 21-25.
9] T.-L. Kao, C.-C. Wang, Y.-T. Pan, Y.-J. Shiao, J.-Y. Yen, C.-M. Shu, G.-H. Lee, S.-M.
Peng, W.-S. Chung, Upper Rim Allyl- and Arylazo-Coupled Calix[4]arenes as Highly
5
[
[
2
+
Sensitive Chromogenic Sensors for Hg Ion, J. Org. Chem., 70 (2005) 2912-2920.
10] F. Karcı, İ. Şener, H. Deligöz, Azocalixarenes. 1: synthesis, characterization and
investigation of the absorption spectra of substituted azocalix[4]arenes, Dyes Pigm., 59
2003) 53-61.
11] A.M. Khedr, M. Gaber, R.M. Issa, H. Erten, Synthesis and spectral studies of 5-[3-
1,2,4-triazolyl-azo]-2,4-dihydroxybenzaldehyde (TA) and its Schiff bases with 1,3-
[
(
[
(
diaminopropane (TAAP) and 1,6-diaminohexane (TAAH). Their analytical application for
spectrophotometric microdetermination of cobalt(II). Application in some radiochemical
studies, Dyes Pigm., 67 (2005) 117-126.
[12] T.H. Kim, S.H. Kim, L.V. Tan, Y.J. Seo, S.Y. Park, H. Kim, J.S. Kim, Transition metal
ion selective ortho-ester diazophenylcalix[4]arene, Talanta, 71 (2007) 1294-1297.

ACCEPTED MANUSCRIPT
1
9
[
13] T.H. Kim, S.H. Kim, L. Van Tan, Y. Dong, H. Kim, J.S. Kim, Diazo-coupled
calix[4]arenes for qualitative analytical screening of metal ions, Talanta, 74 (2008) 1654-
658.
14] M. Kumar, J.N. Babu, V. Bhalla, Azophenol appended (thia)calix[4]arenes for
colorimetric sensing of anions: A complexation induced extended conjugation, Talanta, 81
2010) 9-14.
15] M.R. Mahmoud, S.A. Ibrahim, M.A. Hamed, Low excitation energy band of 4-
hydroxyazobenzene derivatives, Spectrochim. Acta, Part A, 39 (1983) 729-733.
16] K. Rezaeian, H. Khanmohammadi, Naked-eye detection of biologically important anions
by a new chromogenic azo-azomethine sensor, Spectrochim. Acta, Part A, 133 (2014) 31-37.
17] N. Su, J.S. Bradshaw, P.B. Savage, K.E. Krakowiak, R.M. Izatt, S.L. De Wall, G.W.
1
[
(
[
[
[
Gokel, Syntheses and aggregate study of bisphenol-containing diaza-18-crown-6 ligands,
Tetrahedron, 55 (1999) 9737-9742.
[18] B. Shaabani, Z. Shaghaghi, A.A. Khandar, Optical spectroscopy studies of the
2
+
2+
complexation of bis(azophenol)calix[4]arene possessing chromogenic donors with Ni , Co ,
2
+
2+
2+
Cu , Pb and Hg , Spectrochim. Acta, Part A, 98 (2012) 81-85.
19] Q.-Q. Zhang, J.-F. Ge, Q.-F. Xu, X.-B. Yang, X.-Q. Cao, N.-J. Li, J.-M. Lu, A selective,
sensitive probe for mercury(II) ions based on oxazine-thione, Tetrahedron Lett., 52 (2011)
95-597.
20] Y. Zhang, Y. Yan, S. Chen, Z. Gao, H. Xu, ‘Naked-eye’ quinoline-based ‘reactive’
[
5
[
2
+
sensor for recognition of Hg ion in aqueous solution, Bioorg. Med. Chem. Lett., 24 (2014)
373-5376.
21] V.I. Vetrogon, N.G. Lukyanenko, M.J. Schwing-Weill, F. Arnaud-Neu, A PC compatible
5
[
computer program for the calculation of equilibrium constants by the simultaneous processing
of different sets of experimental results, Talanta, 41 (1994) 2105-2112.

ACCEPTED MANUSCRIPT
2
0
[22] Y.J. Jang, Y.H. Yeon, H.Y. Yang, J.Y. Noh, I.H. Hwang, C. Kim, A colorimetric and
3
+
3+
fluorescent chemosensor for selective detection of Cr and Al , Inorg. Chem. Commun., 33
2013) 48-51.
23] L.K. Kumawat, N. Mergu, M. Asif, V.K. Gupta, Novel synthesized antipyrine derivative
(
[
3
+
3+
based “Naked eye” colorimetric chemosensors for Al and Cr , Sens. Actuators, B, 231
2016) 847-859.
24] C.-C. Cheng, Z.-S. Chen, C.-Y. Wu, C.-C. Lin, C.-R. Yang, Y.-P. Yen, Azo dyes
(
[
featuring a pyrene unit: New selective chromogenic and fluorogenic chemodosimeters for
Hg(II), Sens. Actuators, B, 142 (2009) 280-287.
[25] K. Liu, X. Zhao, Q. Liu, J. Huo, Z. Li, X. Wang, A novel multifunctional BODIPY-
2
+
−
derived probe for the sequential recognition of Hg and I , and the fluorometric detection of
3
+
Cr , Sens. Actuators, B, 239 (2017) 883-889.
26] J. Liu, M. Yu, X.-C. Wang, Z. Zhang, A highly selective colorimetric sensor for Hg
based on nitrophenyl-aminothiourea, Spectrochim. Acta, Part A, 93 (2012) 245-249.
27] Z. Yu, Z. Tian, Z. Li, Z. Luo, Y. Li, Y. Li, J. Ren, A new chromogenic and fluorogenic
2
+
[
[
2
+
chemosensor for Hg(II) with high selectivity based on the Hg -promoted deprotection of
thioacetals, Sens. Actuators, B, 223 (2016) 172-177.

ACCEPTED MANUSCRIPT
2
1
Figure 1

ACCEPTED MANUSCRIPT
2
2
Figure 2

ACCEPTED MANUSCRIPT
2
3
Figure 3

ACCEPTED MANUSCRIPT
2
4
Figure 4

ACCEPTED MANUSCRIPT
2
5
Figure 5

ACCEPTED MANUSCRIPT
2
6
Figure 6

ACCEPTED MANUSCRIPT
2
7
Figure 7

ACCEPTED MANUSCRIPT
2
8
Figure 8

ACCEPTED MANUSCRIPT
2
9
Figure 9