Concise design and synthesis of water-soluble fluorescence sensor for sequential detection of Zn(II) and picric acid via cascade mechanism

Article abstract of DOI:10.1016/j.dyepig.2019.04.023

Under microwave assistance, 2,6-bis(benzimidazolyl)-pyridine has been concisely synthesized and further productively converted to the functional molecule BMBP via accessible N-alkylation reaction. For the introduction of alkoxyalkyl chain, this original substance BMBP with better solubility in water exhibits specific fluorescence response toward Zn²⁺ from colourless to blue in aqueous solution. A sequential detection for picric acid (PA) can be conducted in this following system, showing the high selectivity and sensitivity of quenching over other analogues. On the basis of the comparison with the control BMBB, a cascade sensing mechanism has been disclosed to accelerate the recognition of structure-property relationship, which is fully supported by LC-MS, ¹H NMR, lifetime measurement and theoretical calculation. Noteworthily, BMBP is also readily available for practical application not only in quantitative determination of Zn²⁺ and PA in real water samples, but also in visible detection of two analytes in multiple forms on paper test strips, offering convenient process for low-cost, portable and versatile sensing device.

Full text of DOI:10.1016/j.dyepig.2019.04.023

Accepted Manuscript
Concise design and synthesis of water-soluble fluorescence sensor for sequential
detection of Zn(II) and picric acid via cascade mechanism
Kai Jiang, Si-Hong Chen, Shi-He Luo, Chu-Ming Pang, Xin-Yan Wu, Zhao-Yang
Wang
PII:
S0143-7208(19)30316-X
https://doi.org/10.1016/j.dyepig.2019.04.023
DYPI 7478
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 10 February 2019
Revised Date: 17 March 2019
Accepted Date: 10 April 2019
Please cite this article as: Jiang K, Chen S-H, Luo S-H, Pang C-M, Wu X-Y, Wang Z-Y, Concise design
and synthesis of water-soluble fluorescence sensor for sequential detection of Zn(II) and picric acid via
cascade mechanism, Dyes and Pigments (2019), doi: https://doi.org/10.1016/j.dyepig.2019.04.023.
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
Graphic Abstract
Concise design and synthesis of water-soluble fluorescence sensor for sequential
detection of Zn(II) and picric acid via cascade mechanism
Author: Kai Jianga,b,#, Si-Hong Chena,#, Shi-He Luoa,b,*, Chu-Ming Pang, Xin-Yan Wu,
Zhao-Yang Wanga,b,*

ACCEPTED MANUSCRIPT
Concise design and synthesis of water-soluble fluorescence sensor for
sequential detection of Zn(II) and picric acid via cascade mechanism
Kai Jianga,b,#, Si-Hong Chena,#, Shi-He Luoa,b,*, Chu-Ming Panga, Xin-Yan Wua, Zhao-Yang
Wanga,b,*
a School of Chemistry and Environment, South China Normal University; Key Laboratory of Theoretical Chemistry of Environment,
Ministry of Education, Guangzhou 510006, China
b Key Laboratory of Functional Molecular Engineering of Guangdong Province, South China University of Technology, 381 Wushan
Road, Guangzhou 510640, China
# Kai Jiang and Si-Hong Chen contributed equally to this work.
* Corresponding Author
E-mail: wangzy@scnu.edu.cn. Tel: 8620-39310258.
E-mail: pinky_r@163.com.
1

ACCEPTED MANUSCRIPT
Abstract
Under microwave assistance, 2,6-bis(benzimidazolyl)-pyridine has been concisely synthesized
and further productively converted to the functional molecule BMBP via accessible N-alkylation
reaction. For the introduction of alkoxyalkyl chain, this original substance BMBP with better
solubility in water exhibits specific fluorescence response toward Zn2+ from colourless to blue in
aqueous solution. A sequential detection for picric acid (PA) can be conducted in this following
system, showing the high selectivity and sensitivity of quenching over other analogues. On the basis
of the comparison with the control BMBB, a cascade sensing mechanism has been disclosed to
accelerate the recognition of structure-property relationship, which is fully supported by LC-MS, 1H
NMR, lifetime measurement and theoretical calculation. Noteworthily, BMBP is also readily
available for practical application not only in quantitative determination of Zn2+ and PA in real
water samples, but also in visible detection of two analytes in multiple forms on paper test strips,
offering convenient process for low-cost, portable and versatile sensing device.
Keywords
Water-soluble sensor; Sequential detection; Cascade mechanism; Test strips
2

ACCEPTED MANUSCRIPT
1. Introduction
As the second abundant transition metal in the human body, Zn2+ assumes an irreplaceable role
in numbers of physiological processes such as enzymatic catalysis, structural organization, and
functional regulation [1]. However, the anomalous concentration of Zn2+ may not only cause
neurogenic disease, but also bring about contamination to soil as well as water [2, 3]. Due to many
merits of fluorescence sensing technique, such as high sensitivity, instant response and operational
simplicity, a plethora of fluorescence sensors with good performance on monitoring Zn2+ have been
reported in recent years [2, 4-7]. Even so, some drawbacks like water insolubility, complicated
synthesis or inferior selectivity, have still limited the application of the existing Zn2+ sensors [8-11].
Therefore, it will be greatly valuable for practicality to exploit fluorescence sensor that can
efficiently detect Zn2+ in aqueous solution.
Knowing as the highly explosive property, picric acid (PA) is also a prevalent sensing object as
well as TNT. Notably, PA is of certain water solubility and toxicity in lower concentration, which
may induce intense stimulation and health problem related to eye, skin and respiratory after contact
with its vapour, powder or dilute solution [12-14]. In light of implementing the rapid and handy
sensing of PA, fluorescence sensors become highly desirable in this field [15-22]. In fact, most of
the reported fluorescence PA sensors possess good selectivity and sensitivity [18-20, 23-25], but
some of them remain improvement owing to the high cost, slow response and dependence of
organic solvents [26-29]. Hence, the development of environmentally comparable fluorescence
sensor for on-site detection of PA in water samples is under pressing concern.
2,6-bis(benzimidazolyl)-pyridine (BBP) is always served as ligand coordinating with Cu2+,
Rh3+, Rh4+, etc., because of its structural characteristics on multiple binding sites and coordination
3

ACCEPTED MANUSCRIPT
cavity [30-32]. Meanwhile, the outstanding emission property of benzimidazole moiety in BBP is
extensively applied in fluorescence sensors [33-35]. Herein, we selected BBP as the skeleton of the
fluorescence sensor, and searched for a reformative synthetic way to acquire good yield at the same
time. Utilizing the active N-H on imidazole for chemical modification, two hydrophilic alkyloxy
groups have been introduced to this skeleton to overcome the poor water solubility of BBP, giving
the target molecule BMBP as expected fluorescent probe (Scheme 1). It is revealed that BMBP
shows highly selective fluorescence response toward Zn2+ in entirely aqueous solution, and the
mixture of BMBP and Zn2+ can further detect PA by fluorescence quenching. These enable BMBP
to be developed as a promising fluorescence sensor for sequential detection of Zn2+ and PA not
merely in real water samples, but also on the paper test strips. In addition, the control compound
BMBB with a similar structure to BMBP have also been obtained through the same route (Scheme
1) to investigate the sensing mechanism.
Scheme 1. Synthetic route for target compound BMBP and control compound BMBB.
4

ACCEPTED MANUSCRIPT
2. Experimental section
2.1. Chemicals and apparatus
Melting point was performed on an X-5 digital melting point apparatus without correcting.
Low resolution electrospray ionization mass spectra (ESI-MS) were recorded on Thermo LCQ
DECA XP MAX mass spectrometer or Agilent Technologies 6120, while the high resolution mass
1
13
spectra (HRMS) were recorded on Bruker maXis impact. H and C NMR spectra were collected
on BRUKER DRX-600 spectrometer using TMS as an internal standard. Elemental analysis was
obtained with Perkin Elmer Series II 2400. UV-vis spectra were measured by using a Shimazu
UV-2700 ultraviolet absorption detector at room temperature. The fluorescence spectra were
recorded with a Hitachi F-4600 spectrophotometer at room temperature; and the slit width was 5 nm
for both excitation and emission. The pH values were measured by a PHS-25C meter. Fluorescence
lifetime was measured by FLS 920 fluorescence spectrometer. All chemicals used within this work
were of analytical grade purity. For the real water samples, tap water was collected from laboratory;
pond water and river water were collected from South China Normal University and the Pearl River
after centrifugation.
2.2. Synthesis
2.2.1. Synthesis of compound intermediate 2
Based on our previous report in terms of the microwave-assisted method for the synthesis of
benzimidazole compound [36], we conducted the synthesis of the intermediate 2 in a household
microwave oven. Acid 1 (1.5 mmol), 1,2-diaminobenzene (3.3 mmol) and 25 mL polyphosphoric
acid (PPA) were added into a 50 mL round bottom flask. The mixture was intermittently irradiated
5

ACCEPTED MANUSCRIPT
in a microwave cavity with an output at 40 % (320 W) for specified times (30 s per irradiation).
After that, the reaction was stopped by adding water, and the mixture was regulated to the pH range
of 9-10 with sodium hydroxide, and a dark green solid was obtained by an immediate filtration. At
last, the crude product was recrystallized by ethanol to give the intermediate 2.
2,6-Bisbenzimidazolylpyridine (2a, BBP), white solid, m.p. > 310 °C, Yield 48.5 % (46.0 % in
literature [37]); 1H NMR (DMSO-d6-TMS, 600 MHz): δ = 7.27-7.33 (2H, m, ArH), 7.34-7.41 (2H,
m, ArH), 7.75 (2H, d, J = 6.0 Hz, ArH), 7.79 (2H, d, J = 6.0 Hz, ArH), 8.19 (1H, t, J = 6.0 Hz,
PyH), 8.35 (2H, d, J = 6.0 Hz, PyH), 13.02 (2H, s, NH); 13C NMR (DMSO-d6, 150 MHz): δ = 112.3,
120.2, 121.8, 122.7, 124.2, 134.8, 139.7, 144.6, 148.2, 150.9; ESI-MS, m/z (%): Calcd for
-
C19H12N5 ([M-H]-): 310.12 (100), Found: 310.46 (100); Anal. Calcd for C19H13N5: C 73.30, H 4.21,
N 22.49, Found: C 73.41, H 4.23, N 22.36.
1,3-Bisbenzimidazolylbenzene (2b, BBB), white solid, m.p. 305.1-305.7 °C, yield 46.8 %
(87.0 % in literature [38]); 1H NMR (DMSO-d6, 600 MHz): δ = 7.19-7.30 (4H, m, ArH), 7.57 (2H,
d, J = 6.0 Hz, ArH), 7.71 (2H, d, J = 6.0 Hz, ArH), 7.75 (1H, t, J = 6.0 Hz, ArH), 8.27 (2H, d, J =
6.0 Hz, ArH), 9.06 (2H, s, ArH), 13.12 (2H, s, NH).
2.2.2. Synthesis of compound 3
According to the literature [39], we followed the N-alkylation reaction to carry out the
synthesis of compound 3. A 100 mL round bottom flask was charged with intermediate 2 (1 mmol),
1-bromo-2-(2-methoxyethoxy)ethane (2.2 mmol) and sodium hydroxide (4 mmol) in acetonitrile
(15 mL). The reaction mixture was stirred at 80 °C overnight. After removing the solvent, the
residue was dissolved in dichloromethane. The organic solution was washed with water three times
6

ACCEPTED MANUSCRIPT
and then dried over anhydrous magnesium sulfate. Purified by column chromatography on silica gel
with gradient eluents of petroleum ether and ethyl acetate, the pure target compound 3 was afforded.
2,6-Bis(1-(2-(2-methoxyethoxy)ethyl)-2,6-bisbenzimidazolyl)pyridine (3a, BMBP), colourless
oily liquid, yield 43.2 %; 1H NMR (CDCl3, 600 MHz): δ = 3.16 (6H, s, OCH3), 3.20 (4H, t, J = 6.0
Hz, OCH2), 3.30 (4H, t, J = 6.0 Hz, OCH2), 3.75 (4H, t, J = 6.0 Hz, OCH2), 4.94 (4H, t, J = 6.0 Hz,
NCH2), 7.31-7.40 (4H, m, ArH), 7.55 (2H, d, J = 6.0 Hz, ArH), 7.87 (2H, d, J = 6.0 Hz, ArH), 8.07
13
(1H, t, J = 6.0 Hz, PyH), 8.34 (2H, d, J = 6.0 Hz, PyH); C NMR (CDCl3, 150 MHz): δ = 44.6,
58.8, 69.7, 70.4, 71.7, 107.8, 110.8, 120.1, 123.0, 123.7, 125.7, 136.4, 138.2, 142.4, 149.7, 150.3;
ESI-MS, m/z (%): Calcd for C29H34N5O4 ([M+H]+): 517.25 (100), Found: 516.85 (100); Anal.
+
Calcd for C29H33N5O4: C 67.55, H 6.45, N 13.58, Found: C 67.63, H 6.41, N 13.47.
1,3-Bis(1-(2-(2-methoxyethoxy)ethyl)-1H-benzimidazol-2-yl)benzene (3b, BMBB), colourless
1
oily liquid, yield 47.5 %; H NMR (CDCl3, 600 MHz): δ = 3.22 (6H, s, CH3), 3.39 (4H, t, J = 6.0
Hz, OCH2), 3.51 (4H, t, J = 6.0 Hz, OCH2), 3.90 (4H, t, J = 6.0 Hz, OCH2), 4.51 (4H, t, J = 6.0 Hz,
NCH2), 7.32-7.37 (4H, m, ArH), 7.51-7.55 (2H, m, ArH), 7.70 (1H, t, J = 6.0 Hz, ArH), 7.82-7.88
(2H, m, ArH), 8.02 (2H, d, J = 6.0 Hz, ArH), 8.28 (1H, s, ArH); 13C NMR (CDCl3, 150 MHz): δ =
45.0, 58.9, 69.3, 70.7, 71.8 110.5, 120.0, 122.7, 123.0, 125.1, 129.1, 131.2, 135.2, 135.8, 143.0,
+
153.4; ESI-MS, m/z (%): Calcd for C30H34N4O4 ([M+H]+): 515.26 (100), Found: 514.90 (100);
Anal. Calcd for C30H33N4O4: C 70.02, H 6.66, N 10.89, Found: C 70.03, H 6.63, N 10.94.
2.3. Methods
2.3.1. General procedure for optical spectral measurements
The compound 3 were dissolved in THF to acquire 10 mM stock solution. Then, the 10 µM
7

ACCEPTED MANUSCRIPT
aqueous solution of compound 3 was prepared for the measurements of both UV-vis absorption and
fluorescence spectra under room temperature. Both the excitation and emission slit widths were set
at 5 nm during the fluorescence experiments.
2.3.2. Limit of detection
According to the literature [40, 41], the limit of detection (LOD) was determined by using
equation: LOD = 3δ/K. Herein, δ is the standard deviation of the blank measurements (n = 10), and
the K is the slope of the calibration curve.
2.3.3. Calculation of Stern-Volmer constant
According to the literature [24, 41], the Stern-Volmer quenching constant (Ksv) was calculated
according to the Stern-Volmer equation: I0/I = Ksv[Q] + 1, where I0 and I are the fluorescence
intensities observed in the absence and presence of PA, respectively; [Q] is the quencher
concentration.
Caution! PA, 2,4-DNP, and NP used in the present study are highly explosive and should be
handled only in small quantities.
3. Results and discussion
3.1. Design and synthesis of target molecule BMBP
The existence of the multiple nitrogen atoms and semi-enclosed structure [30, 42, 43] in BBP
offer it significant binding affinity to metal ions, while the benzimidazole groups [33-36, 44] in
8

ACCEPTED MANUSCRIPT
BBP provide it excellent fluorescence behaviour. For these reasons, the basic skeleton of BBP was
chosen as the donator and fluorophore of the fluorescence sensor for metal ion. Thus, BBP as an
intermediate was concisely synthesized via an improved approach within an hour by the aid of
microwave technology.
To improve the water solubility of the sensor and make it more practical, two hydrophilic
alkoxy groups were considered to chemically combine with BBP on the reactive imidazole N-H
where the substitution reaction can be easily carried out. Indeed, as designed, the successfully
obtained compound BMBP has largely improved water solubility and expectable sensing
performance, which will be confirmed in the following discussions.
Meanwhile, we also synthesized the control compound BMBB via the same synthetic route
(Scheme 1) to support the building strategy and sensing mechanism of BMBP. Both of the novel
1
13
compounds BMBP and BMBB were systematically characterized by H, C NMR, ESI-MS (the
corresponding spectra, Figs. S1-S10, see Supplementary Material) and elemental analysis, and the
obtained data were consistent with the anticipation.
3.2. Fluorescence sensing of Zn2+
At the beginning, we investigated the water solubility of BMBP by comparing the UV-vis
absorption spectra with BBP. After 0.1 mmol of these two compounds were respectively deposited
in 1 L water, the 10-4 M solutions were acquired with adequate dissolution. As shown in Fig. S11, a
clear absorption band at 310 nm belongs to BMBP, while obviously no absorption band can be
observed for BBP. The absolutely larger absorbance of BMBP indicates the significant
enhancement of water solubility of BMBP versus BBP, which can be ascribed to the contribution of
9

ACCEPTED MANUSCRIPT
the hydrophilic alkoxy chains.
Fig. 1. Fluorescence spectra of compound BMBP (10-4 M in water, λex = 310 nm) with different
metal ions (5 equiv.) dissolved in water. Inset: visible fluorescent change of compound BMBP (10-4
M in water) with various metal ions (5 equiv.) under 365 nm UV lamp.
Subsequently, we investigated the selectivity of the fluorescence sensing studies of BMBP
solution (10-4 M in water) towards 24 different metal ions, i.e. Na+, K+, Ag+, Zn2+, Fe2+, Pd2+, Ca2+,
Ba2+, Mg2+, Pt2+, Cd2+, Pb2+, Mn2+, Cu2+, Hg2+, Ni2+, Co2+, Al3+, Cr3+, Fe3+, Ru3+, Tb3+, Au3+, Rh4+.
As depicted in Fig. 1, the BMBP exhibits the sole emission peak at 383 nm upon 310 nm excitation,
and the fluorescence is colourless under 365 nm UV light as the emission wavelength locates at
nearly ultraviolet region. After adding 5 equiv. of these 24 metal ions into BMBP solution
10

ACCEPTED MANUSCRIPT
respectively, a distinct red shift of the emission peak to visible region can be observed only for Zn2+,
while there is no obvious shift occurrence for the other metal ions. By viewing under 365 nm UV
light, the addition of Zn2+ causes the clear fluorescence change of BMBP solution from colourless
to blue, while the others give slight changes as the emission wavelength is still beyond the visible
region (Fig. 1, inset). The unique “off-on” fluorescent response towards Zn2+ implies the potential
of BMBP being a highly selective fluorescence sensor for Zn2+ over other metal ions in water via
either fluorescence spectroscopic analysis or naked-eye inspection. Besides, the control compound
BMBB was also proceeded with the selectivity studies, whereas none of the metal ions in BMBB
(10-4 M in water) solution can cause conspicuous shift of the emission peak including Zn2+ (Fig. 2).
The weak interaction between BMBB and Zn2+ is supposed to be the reason for this phenomenon.
Fig. 2. Fluorescence spectra of control compound BMBB (10-4 M in water, λex = 310 nm) with
different metal ions (5 equiv.) dissolved in water.
11

ACCEPTED MANUSCRIPT
The fluorometric titration experiment of BMBP (10 µM) was then carried out with Zn2+ in
water. As shown in Fig. 3, the initial emission peak at 383 nm gradually shifts to 415 nm upon the
incremental addition of Zn2+, and this bathochromic shift is almost terminated at 5 equiv. of Zn2+.
To inspect the relationship between the amount of Zn2+ and the wavelength of maximum emission,
we integrated these two groups of data and built up a bi-exponentially fitted curve in good
correlation (R2 = 0.992) as depicted in the inset of Fig. 3. Moreover, the LOD for the Zn2+ is
calculated to be 1.83 Õ 10-7 M (11.97 ppb) by virtue of the partial linear fitted data (Fig. S12). This
value of LOD is superior to many other Zn2+ sensors [8, 9, 45-48] (For more examples, please see
Table S1, the comparison of BMBP with other Zn2+ sensors in Supplementary Material), which
indicates that BMBP can sensitively monitor Zn2+ in water.
Fig. 3. Fluorescence spectra of compound BMBP (10-4 M in water) upon the addition of Zn2+
dissolved in water (λex = 310 nm); Inset: the functional relationship between the maximum λem and
equivalents of Zn2+.
12

ACCEPTED MANUSCRIPT
Since the response time is a vital parameter for metal ions sensor, we measured the
time-dependent Zn2+ detection by recording the instantaneous fluorescence intensity of BMBP at
415 nm. As shown in Fig. S13, the fluorescence intensity decreases rapidly within few seconds after
the addition of 5 equiv. of Zn2+ into the BMBP solution (10-4 M in water), and finally keeps
constant in the next 4 min, whereupon the fluorescence response of BMBP sensing Zn2+ can be
concluded to approximately 4 min.
The tolerability of pH is also essential to survey for the fluorescence sensor applied in solution
[49]. Herein, we marked down the maximum fluorescence intensity of the BMBP (10-4 M) in
aqueous system with different pH values from 1 to14 before and after introduction of 5 equiv. Zn2+.
It can be learnt from Fig. S14 that when the pH is below 5, the intensity is strongly declined
whether there is or not Zn2+ in BMBP solution. When the pH reaches to 5-12, the intensity as well
as the case when sensing Zn2+ may be backed to the normal. When the pH is over 12, the original
intensity is still normal as before but remains unchanged after the addition of Zn2+. These results
denote that the suitable pH values of the aqueous system for BMBP sensing Zn2+ range from 5 to
12, which is broader than many other Zn2+ sensors [9, 48, 50-54].
3.3. Sequential sensing of PA
For a good sensor, to implement the sequential detection means to not only reduce the cost of
making sensor, but also improve the sensing efficiency [55, 56]. Thus, many sensor with sequential
sensing performance on bis- or multi-analytes received particular attention in recent years [57-59].
Generally, these sensors may respond to multiple ions [60-63] or small molecules [64], even some
of them can sequentially detect ion to molecule [65-67]. In view of knowing the “off-on”
13

ACCEPTED MANUSCRIPT
fluorescent response towards Zn2+, we conjectured that BMBP might be further employed to
visualize the sequential detection of PA after Zn2+ within the same aqueous system. Therefore, the
following PA sensing experiments have been observed.
To validate if the BMBP can sequentially monitor PA after Zn2+, the BMBP-Zn2+ solution
stemming from the foregoing experiments of Zn2+ detection was then coped with the selectivity
studies of PA among a bunch of analogues, including nitroaromatic compounds (DNP, TNT, NP, etc.,
their corresponding structures are presented in Fig. S15) and other interferents (HBAc, Phenol).
With the same condition, 2 equiv. PA and the other analogues were respectively added into
BMBP-Zn2+ to perform the fluorescence assay. As described in Fig. 4, the introduction of PA
induces significant fluorescence quenching of BMBP-Zn2+ from blue to colourless, and the
quenching efficiency reaches to 84.3 %. On the other hand, the same amount of DNP and NP only
brings to 48.6 % and 36.2 % quenching efficiency, which cannot lead to the distinguishable
fluorescence change as PA. This can be attributed to the more favorable electrostatic interaction
between BMBP-Zn2+ and PA permitted by the more electron-deficient property of PA. This
interaction may further facilitate the photoinduced electron transfer (PET) process to deliver the
remarkable quenching of BMBP-Zn2+ selectively toward PA [16, 22, 68]. In addition, the other
nitro compounds show negligible impact to the fluorescence of BMBP-Zn2+. We have also
investigated the selectivity towards 10 different anions commonly in water. As anticipated, these
anions cannot produce any obvious changes to the fluorescence spectra of BMBP-Zn2+ (Fig. S16).
Thus, the above observations prove that BMBP-Zn2+ can serve to selectively detect PA in water
through a visible “on-off” fluorescence signal.
14

ACCEPTED MANUSCRIPT
Fig. 4. Comparison of fluorescence quenching efficiency and visible fluorescent colour changes
under 365 nm UV light of BMBP-Zn2+ after interaction with 2 equiv. PA or other analogues
respectively.
We further investigated the fluorometric titration of PA to analyze the sensing behaviour. As
elucidated in Fig. 5, with the addition of PA from 0 to 4 equiv., the fluorescence intensity of
BMBP-Zn2+ at 415 nm is gradually descended. The fluorescence quenching efficiency reaches to
99.5 % after dropping with 4 equiv. PA into BMBP-Zn2+. To place the fluorescence intensity and
the equivalent of PA together, an exponential curve with 0.997 of R2 can be fitted, which offers the
possibility of BMBP-Zn2+ for quantitative PA determination. The LOD is calculated to be 2.16 Õ
10-7 M (49.49 ppb) from the fitted linear equation of converted titration data (Fig. S17). This value
of the LOD possesses certain advantage among the reported PA sensors [26-29, 69-73], which
reflects the good sensitivity of BMBP-Zn2+ towards PA in water (For more examples, please see
15

ACCEPTED MANUSCRIPT
Table S2, the comparison of BMBP-Zn2+ with other PA sensors in Supplementary Material).
Fig. 5. Fluorescence spectra of BMBP-Zn2+ upon the addition of different equivalents of PA
dissolved in water (λex = 310 nm); Inset: the functional correlation between fluorescence intensity at
415 nm and equivalent of PA.
To further evaluate the quenching efficiency, the fluorometric titration data were transformed
to the Stern-Volmer plot showing in Fig. S18. The plots show upward linearity at a lower
concentration of PA (0-0.7 × 10-4 M), and the quenching constant (Ksv) is therefore calculated to be
2.72 × 104 M-1. This value is higher than some reported PA sensor [70-73], which further
demonstrates the sensitive detection of BMBP-Zn2+ toward PA. In addition, a steep curve can be
viewed at a higher concentration of PA, which is presumably due to the existence of superamplified
quenching effect [18].
The response time of sensing PA was also investigated by scanning the time-dependent
16

ACCEPTED MANUSCRIPT
fluorescence intensity of BMBP-Zn2+ at 415 nm. Once 4 equiv. PA is added into the BMBP-Zn2+
system, the fluorescence intensity sharply goes down in few seconds and turns to stable after 6 s
(Fig. S19). This quick response enables BMBP-Zn2+ to be an on-site fluorescence sensor for PA
detection in aqueous solution, which is faster than many reported PA sensors [27, 70, 71, 74].
To test the stability of sensing PA, we further recorded the fluorescence intensity of
BMBP-Zn2+ at 415 nm varying different pH values from 1 to 14 before and after adding with PA
(Fig. S20). Similar with the behaviour of Zn2+ detection, the emission intensity of BMBP-Zn2+ is
insusceptible to the solution with pH values in 5-14. And within this range, the quenching
efficiency can also be kept over 95 %. Thus, BMBP-Zn2+ is confirmed to be pH-tolerable to
monitor PA in water above plenty of reported PA sensors [13, 21, 41, 72, 74].
3.4. Sensing mechanism
To ascertain the sensing mechanism of BMBP towards Zn2+, the UV-vis absorption spectra of
BMBP with Zn2+ in water have been carried out at first (Fig. S21). It can be learnt that BMBP may
witness a decrease of absorbance at 310 nm along with an emergence of new absorbance at around
345 nm. This new absorbance may derive from the formation of BMBP-Zn2+ complex instead of
the absorbance of Zn2+ as known by the absorption spectrum for only 5 equiv. Zn2+ (blue line in Fig.
S21). What’s more, the existence of BMBP-Zn2+ complex was further confirmed by HRMS. The
sample for measurement was collected from the Zn2+ sensing experiment after dilution by water,
and the results were shown in Fig. 6. Clearly, three dominant peaks have been captured. The one at
m/z 516.2160 is assigned to [BMBP + H]+ along with the nearby [BMBP + Na]+. The other one at
m/z 624.1796 is assigned to [BMBP + Zn2+ + C2H5O-]+, which can be the symbol of 1:1 interaction
17

ACCEPTED MANUSCRIPT
between BMBP and Zn2+ in the sensing system. Referring to the reported method [54], the binding
constant (Ka) of BMBP-Zn2+ is calculated to be 4.88 × 104 M-1 according to a Benesi-Hildebrand
plot (Fig. S22). On the contrary, there isn’t any possible peak for BMBB-Zn2+ that can be found in
the HRMS of BMBB with Zn2+ (Fig. S23). As a coherent conclusion with selectivity assay, this
possibly points to the weak interaction between BMBB and Zn2+.
Fig. 6. HRMS spectrum of Zn2+ sensing system.
To unveil the interaction pattern of BMBP with Zn2+ and PA in solution, we have made a
1
parallel comparison of the partial H NMR spectra before and after their interaction in mixed
solvent of THF-d6 and D2O (Fig. 7). It is not difficult to find that there is a downshift for the
chemical shift of the whole aromatic hydrogen after the addition of Zn2+, especially the largest shift
for Hb from 8.02 to 8.14. This implies that Zn2+ may strongly bind to BMBP and the N atom on
pyridine is likely to be one of the binding sites. Moreover, the Hc has the larger downshift (7.68 to
7.75) than Hd (7.54 to 7.60), suggesting that the unalkylated N atom in imidazole ring may
participate in binding while the alkylated N atom doesn’t. However, the further addition of 2 equiv.
18

ACCEPTED MANUSCRIPT
PA seems not bring any shift of the aromatic H, which indicates the absence of specific interaction
1
between BMBP and PA. To our surprise, the comparative analysis on H NMR spectra using
control BMBB (Fig. S24) shows that, there is a downshift for most aromatic H except for Hi after
the addition of 5 equiv. Zn2+, and the downshift of Hc (7.68 to 7.80) is larger than that of Hb (7.60 to
7.65). These indicate the fact that BMBB may also interact with Zn2+ by combination on imidazole
N atom, but the binding mode is certainly different with BMBP-Zn2+. This difference in binding
mode between BMBB and Zn2+ may not bring to the effect of the extended conjugate as well as
BMBP, which is confirmed by the later theoretical calculation.
Fig. 7. Partial 1H NMR spectra (8.8-7.0 ppm region) of BMBP (10-4 M in THF-d8 and D2O)
before and after the sequential addition of Zn2+ and PA.
19

ACCEPTED MANUSCRIPT
Fig. 8. Spectral overlap between UV-vis absorption spectra of PA, excitation and emission spectra
of BMBP-Zn2+.
Aiming to gain in insight into the quenching mechanism of sensing PA in water, the absorption
spectrum of PA is merged with the excitation and emission spectra of BMBP-Zn2+ as description in
Fig. 8. Not only a large area of overlap appears between the absorption spectrum of PA and the
excitation spectrum of BMBP-Zn2+, but also a middle area of overlap between the absorption
spectrum of PA and the emission spectrum of BMBP-Zn2+ can be observed. Based on this
observation, it is assumed that both the inner filter effect (IFE) and Förster resonance energy
transfer (FRET) are probably responsible for the fluorescence quenching [75-77]. We also
investigated the spectral overlap for the other tested PA analogues (Fig. S25). Likewise, DNP and
NP exhibit partial overlap with the excitation or emission spectra of BMBP-Zn2+, but the other
tested nitro compounds do not. This indicates that the IFE or FRET effect may occur after
20

ACCEPTED MANUSCRIPT
BMBP-Zn2+ interacting with PA, DNP and NP, while the other nitro compounds cannot.
Noticeably, it is a fact that DNP and NP containing less nitro groups show smaller quenching
efficiency than PA as shown in Fig. 4. This potentially releases a signal that a synergistic
mechanism in addition to IFE or FRET may contribute to selective fluorescence quenching of
BMBP-Zn2+ with respect to PA over other tested PA analogues.
To further verify the major reason for quenching induced by PA, the lifetime of BMBP was
subsequently measured in the absence and presence of PA preceded by Zn2+. As shown in Fig. S26,
the time-correlated single photon counting (TCSPC) plots of BMBP, BMBP-Zn2+ and BMBP-Zn2+
with PA show almost complete overlap. And their lifetimes calculated from the single-exponentially
fitted TCSPC plots are almost the same also (Table 1). These represent that the possibility of FRET
effect should be ruled out and the IFE should be one of the major effects during the quenching
process [75-77].
Table 1. Fluorescence lifetime data for BMBP with various analytes.
Substance in solution
τ (ns)
2.88
2.90
2.86
2.87
χ2
1.006
1.019
1.105
1.043
BMBP
BMBP + 5 eq Zn2+
BMBP-Zn2+ + 1 eq PA
BMBP-Zn2+ + 2 eq PA
To better understand the mechansim of fluorescence response, according to the reported
method [5, 78], we performed the theoretical research by employing the density functional theory
21

ACCEPTED MANUSCRIPT
(DFT) in Gaussian 09. B3LYP/6-31g* basis set. Especially, water was chosen as the solvent for the
structural optimization of BMBP, BMBP-Zn2+, and PA. The results are shown as Fig. 9. The
calculated HOMO-LUMO energy levels and their energy gaps (∆E) of BMBP, BMBP-Zn2+, and
1
PA are summarized in Fig. 9a. Based on the results of HRMS spectrum and H NMR studies, a
reasonable structure for BMBP-Zn2+ has been optimized and presented in Fig. 9b.
Fig. 9. (a) HOMO-LUMOs of BMBP, BMBP-Zn2+ and PA calculated by the DFT methods; (b)
Optimized structure of BMBP-Zn2+ and length of partial bonds.
Initially, the simulation of BMBP displays a distort morphology in water owing to the rotation
of benzimidozle moieties. After binding with Zn2+, the planarity of BMBP has been drastically
promoted and the LUMO of BMBP massively transfers from benzimidazole moieties to pyridine
ring, which indicates the activation of intramolecular charge transfer (ICT) process leading to the
22

ACCEPTED MANUSCRIPT
fluorescent red shift [40, 79]. Moreover, the ∆E for BMBP-Zn2+ is smaller than BMBP, which can
also explain the red shift of fluorescence resulting in the “off-on” response for sensing Zn2+ [40, 80].
In addition, we also finished the calculation and simulation of BMBB with Zn2+ (Fig. S27). The
decrease value of ∆E for BMBB after binding with Zn2+ is 0.2 eV (Fig. S27a), which is much more
smaller than 0.65 eV of BMBP (Fig. 9a). Perhaps, for the steric hindrance of Ar-H, the distance of
Zn to N atoms in BMBB is shown to be longer than that in BMBP-Zn2+, and the middle benzene in
the optimized structure of BMBB-Zn2+ is nonplanar with the benzimidazole groups (Fig. S27b).
These differences between BMBB and BMBP may possibly indicate the less likely activation of
ICT process for BMBB-Zn2+ and further account to the irresponsive fluorescence of BMBB
towards Zn2+.
The aforementioned PET process is sugggested to be one of the quenching mechanisms. We
can directly get the proof from comparing the orbital energy of BMBP-Zn2+ and PA [16, 81]. It
should be noted from Fig. 9a that the LUMO energy of electron-deficient PA is higher than the
HOMO energy of BMBP-Zn2+ but lower than the LUMO energy of BMBP-Zn2+, allowing the
facile PET process from the LUMO of the excited BMBP-Zn2+ to the LUMO of PA as
non-radiative decay [81, 82]. For the other tested nitro compounds, their LUMO energy levels are
higher than PA (Fig. S28), which indicates their less driving force of PET process and the highest
electron-transfer efficiency from BMBP-Zn2+ to PA. This also demonstrates the selective
fluorescence quenching response of BMBP-Zn2+ towards PA.
Consequently, a convinced cascade sensing mechanism with a suggested interaction mode for
the sequential fluorescence detection of Zn2+ to PA was presented in Fig. 10 according to the above
investigations. Upon the addition of Zn2+ in BMBP aqueous solution, the ICT effect was switched
23

ACCEPTED MANUSCRIPT
on by the restricted rotation of benzimidazole moieties and the extension of molecular planarity
after 1:1 coordination, leading to the visible “off-on” fluorescent response. The sequential addition
of PA into the following system causes significant “on-off” fluorescence change via a synergetic
effect of IFE and PET. To make an overview of this sensing process, we can conclude that a
cascade mechanism of “ICT-IFE&PET” occurs with the visible “off-on-off” fluorescent response in
sequential detection of Zn2+ and PA in water.
Fig. 10. Proposed cascade sensing mechanism of compound BMBP towards Zn2+ and PA.
3.5. Reversible detection of Zn2+
The reversiblity is widely recognized as an important nature of sensors [39, 83]. According to
the reference [7], we introduced EDTA to perform the reversible test of BMBP sensing Zn2+. As
illustrated in Fig. 11, the maximun λem can back to the initial 383 nm from 415 nm after a certain
amount of EDTA adding into BMBP-Zn2+ in water, and the fluorescent colour correspondingly
backs to invisible from blue within the first cycle. Such behaviour can be maintained in 5 cycles,
which encourages BMBP to be a reusable fluoresence sensor for Zn2+ detection in water [7, 84].
24

ACCEPTED MANUSCRIPT
Fig. 11. Reversible changes in the maximun λem of BMBP (10-4 M in water) upon the alternate
addition of Zn2+ and EDTA.
3.6. Application in real water samples
Attempt to develop the application of BMBP in real water samples, tap water, pond water and
river water were prepared for the tested BMBP solution. For the procedure [52, 53], different
amounts of Zn2+ were added into the as-prepared BMBP solution. The readout of their maximum
λem was processed to the theoretical value by puting into the bi-exponentially fitted curve in Fig. 3,
and then compared with the actual value to output the recoveries. Similarly, the readout of
fluorescence intensity after sequential addition of PA was put into the exponentially fitted curve in
Fig. 5. Each group of water samples was carried out for three independent tests to afford the
Relative Standard Deviation (RSD). The results were summarized in Table 2. It can be seen that,
the recovery for Zn2+ is 100 ± 2.3 %, while for PA is 100 ± 5.9 %. This indicates that the theoretical
25

ACCEPTED MANUSCRIPT
value of either Zn2+ or PA is found to be largely aligned with the actual value in the three tested
water samples. Moreover, the repeatability (RSD ≤ 6.2 %) is also satisfactory [85]. Thus, it is
confirmed that BMBP has the potency to sequentially and quantatitively detect Zn2+ and PA in real
water samples.
Table 2. Sequential detection of BMBP towards Zn2+ and PA in real water samples.
Zn2+ added
(10-4 M)
Zn2+ found
(10-4 M)
Recovery
(%)
RSD
(%)
PA added
(10-4 M)
PA found
(10-4 M)
Recovery
(%)
RSD
(%)
Samples
Tap water
Pond water
River water
2.34
0.87
1.91
2.39
0.85
1.94
102.1
97.7
4.6
2.6
5.1
1.52
3.26
0.34
1.49
3.25
0.36
98.0
99.7
2.7
1.8
6.2
101.6
105.9
3.7. Paper strips for sequential detection
In consideration of developing convenient, economical and efficient sensing materials, we
fabricated the BMBP-coated fluorescent paper strips for the investigation as the reported procedure
[15, 67]. To investigate the sensing performance of the BMBP-coated paper strip, we dropped the
Zn2+ solution on it and tested with fluorescence spectra. Shown as the line a and b in Fig. 12(1), the
BMBP-coated paper strip dealt with Zn2+ shows red-shifted emission which is kept in consistency
with the foregoing experiments in water. Furthermore, its fluorescent colour changes from
colourless to blue towards Zn2+, which is shown as Fig. 12(2).
26

ACCEPTED MANUSCRIPT
Fig. 12. (1): Fluorescence spectra of BMBP-coated test strips before (a) and after dropping with
Zn2+ solution (b) and sequential fuming by PA vapour (d); (2)&(3): Photographs of visible
detection of BMBP-coated test strips towards Zn2+ and sequential detection of PA in different
forms (c: dissolved in water, d: vapour, e: powder) under 365 nm UV light.
In addition, we used the BMBP-Zn2+ coated paper strips to sequentially detect PA dissolved in
water, PA vapour and PA powder. The results are shown in Fig. 12(3). Obviously, dark area can be
found in the place where the paper strips have been contacted to the PA solution (c), PA vapour (d)
and powder (e), respectively. These note that detecting PA in multiple forms with the BMBP-Zn2+
coated paper strips can be visualized. Of course, the fluorescence intensity of BMBP-Zn2+ coated
paper strip largely decreases, e.g. upon exposing in PA vapour, which is shown as the line d in Fig.
12(1). Thus, the above studies support that this readily accessible BMBP-coated paper strip can be
served to sequentially detect Zn2+ and PA as a promising sensing device.
4. Conclusions
27

ACCEPTED MANUSCRIPT
In summary, we have successfully synthesized a water-soluble fluorescence sensor BMBP in
concise two steps. This novel sensor displays an intriguing sequential fluorescence sensing from
Zn2+ to PA in the same aqueous system, concomitantly with “off-on-off” visible fluorescence
response. Besides, BMBP possesses good sensing performance toward Zn2+ and PA on sensitivity,
selectivity (distinguished response from multitude of analogues and common analytes) and pH
tolerability. Delightedly, BMBP is also achieved to the wide application of sequential detection for
Zn2+ and PA not only in real water samples, but also on the versatile paper strip. Importantly, the
cascade fluorescence sensing mechanism from activated ICT to IFE&PET has been unveiled with
the aid of comprehensive studies, which provides a good guidance to the building strategy of
fluorescence sensor for sequential detection of ions and small molecules.
Acknowledgements
We are grateful to the Guangdong Provincial Science and Technology Project (No.
2017A010103016), Guangzhou Science and Technology Project Scientific Special (General Items,
No. 201607010251), Open Fund of the Key Laboratory of Functional Molecular Engineering of
Guangdong Province in SCUT (2017kf01) and NSF of Guangdong Province (No.
2014A030313429) for financial support.
References
[1]
Maret W, Li Y. Coordination dynamics of zinc in proteins. Chem. Rev. 2009; 109: 4682-
707.
[2]
Goldberg JM, Wang F, Sessler CD, Vogler NW, Zhang DY, Loucks WH, et al. Photo-
activatable sensors for detecting mobile zinc. J. Am. Chem. Soc. 2018; 140: 2020-3.
28