Analytica Chimica Acta 958 (2017) 38e50

Contents lists available at ScienceDirect

Analytica Chimica Acta

journal homepage: www.elsevier.com/locate/aca

Electronically tuned sulfonamide-based probes with ultra-sensitivity

for Ga3þ or Al3þ detection in aqueous solution
Ashwani Kumar**, Pil Seok Chae*
Department of Bionanotechnology, Hanyang University, Ansan, 155-88, Republic of Korea

h i g h l i g h t s g r a p h i c a l a b s t r a c t

 Three electronically tuned

sulfonamide-based probes (probes 1,
2, and 3) were developed for metal
ion-sensing.
 Probes 2 and 3 exhibited AIE
behavior with increasing water-
content.
 Probes 2 and 3 displayed a selective
fluorescence “OFF-ON“ behavior for
Ga3þ detection with the LOD of
10 nM.
 PET suppression and AIE jointly
played a significant role for ultra-
sensitive detection of Ga3þ.

a r t i c l e i n f o a b s t r a c t

Article history: Three electronically tuned fluorescent probes (1e3) were synthesized by conjugating a fluorescent unit
Received 21 October 2016 to N,N-bis-(hydroxyethyl)ethylenediamine. Probe 1 bearing an electron-deficient naphthalenedimide
Received in revised form unit did not give a fluorescence response to the presence of various metal ions including monovalent
14 December 2016
metal ions (Naþ, Kþ, and Agþ), divalent metal ions (Ca2þ, Cd2þ, Co2þ, Ni2þ, Cu2þ, Hg2þ, Pb2þ, and Zn2þ)
Accepted 16 December 2016
Available online 27 December 2016
and trivalent metal ions (Al3þ, Ga3þ, Fe3þ, and Cr3þ) in an aqueous solution. By contrast, probes 2 and 3
possessing 1,8-naphthalimide and pyrene fluorophores, respectively, exhibited selective fluorescent
“OFF-ON” behaviors as a result of Ga3þ/Al3þ binding among the diverse metal ions, suggesting the
Keywords:
N,N-bis-(hydroxyethyl)ethylenediamine
importance of fluorophore electronic character with regard to metal ion sensing. The ethylenediamine
Naphthalenedimide analog of probe 3, corresponding to probe 4, was unable to yield a significant change in fluorescence
1,8-Naphthalimide intensity in the presence of any metal ions tested here, revealing the essential role of two hydroxyl
Pyrene sulfonamides groups for metal ion binding. A high association constant of Ka ¼ 2.99  105 M1 was obtained for probe
Aggregation-induced emission 3 with Ga3þ, with a limit of detection (LOD) of 10 nM. This LOD is the lowest value known for Ga3þ
Photo-induced electron transfer detection using chemical sensors. Along with an increase in aggregate sizes, PET suppression of probes
upon metal ion binding was the primary contributor to the enhancement in fluorescence emission
necessary for the sensitive detection of the target ions. The probe-metal ion complexes were fully
characterized via TEM, FE-SEM, 1H NMR, fluorescence spectroscopy techniques and DFT calculations.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction
* Corresponding author.
** Corresponding author.
Metal or metal ions are an absolute requirement for all forms of
E-mail addresses: [email protected] (A. Kumar), [email protected]
(P.S. Chae). life as they are involved in a number of cellular processes including

http://dx.doi.org/10.1016/j.aca.2016.12.032
0003-2670/© 2016 Elsevier B.V. All rights reserved.
 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50 39

metabolism, catalysis, osmotic regulation, bio-mineralization, and most abundant metal in the earth's crust. Pharmaceutical drugs
signaling. Specifically, transition metals such as iron, zinc, copper, containing this metal ion are also frequently utilized in human and
manganese are generally recognized as critical role players in veterinary medicine [24,25], exemplified by aluminum glycinate-
biology [1]. However, heavy metal ions such as cadmium, lead, containing aspirin [26]. The extensive use of aluminum for hu-
mercury, and arsenic ions are known to be hazardous to living or- man activities has increased the concentration of Al3þ in the
ganisms [2]. As the amounts of metal ions should be strictly regu- environment, a primary cause of Parkinson's disease, Alzheimer's
lated for cell's proper function, metal deficiency/richness is disease, and organ damage [27,28]. Thus, the development of new
responsible for many health issues and human diseases such as probes for the detection of Ga3þ/Al3þ are necessary to monitor the
cancer [3], neurodegenerative diseases [4], and diabetes [5]. For levels of these metal ions within environments or biosystems. In
instance, a high level of copper ions was shown to be associated literatures, a number of probes have been reported with regard to
with a variety of tumors and cancers [6], and an irregular zinc ion Al3þ detection, although most of the reported sensors possessed
concentration was also identified in cases involving diabetes [7] selectivity issues [29e33] and were often not applicable to aqueous
and cancers [8]. Consequently, the detection of metal ions in systems. Thus, the current study focused on the development of
aqueous solutions has gained great significance [9e13]. Of the large Ga3þ and Al3þ-sensing probes with high selectivity and sensitivity
number of metal ion detection methods currently available, few in aqueous media. In a continuation of our efforts on development
studies have been reported for Ga3þ sensing, particularly in of fluorescent sensors [34], here we present a chemo-sensor system
aqueous media [14e17]. Rare metal gallium has been used in the based on fluorescence that is a cost-effective, rapid, and facile
manufacture of electronics, such as LEDs and solar cells [18e20]. target detection method [35e37].
Gallium can also be used as a mercury replacement in high- Three probes (probes 1e3) were prepared by conjugating a
temperature thermometers and as a dopant in semiconductors fluorophore to N,N-bis-(hydroxyethyl)ethylenediamine (BHE)
and transistors [21]. Due to the remarkable ability of liquid gallium (Scheme 1). The fluorescence response of each probe to the pres-
to ‘wet’ surfaces, gallium alloys have been used as filling materials ence of metal ions was then investigated. In previous studies,
in human teeth [22]. Furthermore, gallium compounds and gallium several metal ion-binding units were used for the selective detec-
particles (e.g., GaAs) have been identified as potential health haz- tion of a target metal ion. Representatives include dipicolylamine
ards and are known to be carcinogenic. Acute exposure to GaCl3 can for Zn2þ [37e40], thiourea for Hg2þ [41,42], and N,N-dimethyle-
cause throat irritation, breathing difficulties, chest pains, and its thylenediamine for Cu2þ [43,44]. However, BHE was rarely used as a
fumes can lead to pulmonary edema and partial paralysis. The level metal-binding moiety [45]. We hypothesized that the presence of
of gallium within the environment is gradually rising due to its two hydroxyl groups as opposed to amine groups could impart the
increased consumption in various fields [23]. In contrast, Al is the BHE unit with selective metal ion-binding property. The probes also

Scheme 1. Chemical structures and synthetic schemes for probes (1e4). Probes 1 to 3 commonly bear N,N-bis-(hydroxyethyl)ethylenediamine (BHE) unit, but are different from
each other with regard to their fluorophores. Probes 1 to 3 contain as a fluorophore naphthalenedimide (NDI), 1,8-naphthalimide (NPI) and pyrene, respectively. Probe 4 with the
same fluorophore as probe 3 possesses an ethylenediamine group instead of BHE.
40 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50

possess individual fluorophores with tunable electronic structures. by the addition of 1 M aqueous KOH solution (20 mL) and slow
Naphthalenedimide (NDI) was used as the sensing module [46e48] heating to obtain a clear solution. The pH of the reaction mixture
of probe 1 while 1,8-naphthalimide (NPI) and pyrene were used for was adjusted to approximately 6.0e6.5 using a 1 M H3PO4 solution,
probes 2 and 3, respectively. NPI possesses a higher electron den- verified with pH paper. Then n-butyl amine (0.37 ml, 3.73 mmol)
sity than NDI, but contains a lower electron density than pyrene, was added to the solution; the pH of the reaction mixture was
thus yielding a step-wise increase in fluorophore electron density readjusted to 6.4 with the 1 M H3PO4 solution. The reaction mixture
from probe 1 to 2 to 3. As a control agent, we also prepared another was refluxed for 15 h. The reaction mixture was then cooled to
probe (probe 4), possessing an ethylenediamine group in place of room temperature and filtered. Acetic acid (3 mL) was added to the
BHE, with the pyrene fluorophore to investigate the role of two BHE filtrate, forming a solid precipitate, the solids were filtered and
hydroxyl groups with regard to metal ion detection. The chemical dried under vacuum to yield compound 6 (0.90 g, 74% yield) as a
identities of the individual probes 1e4 were well confirmed via 1H, light brown solid. 1H NMR (400 MHz, DMSO-d6): d 0.92 (t,
13
C NMR spectra and HRMS spectrometry (see ESI for details). In the J ¼ 7.2 Hz, 3H, CH3), 1.33e1.38 (m, 2H, CH2), 1.60e1.64 (m, 2H, CH2),
presence of various metal ions, probes 2 and 3 exhibited ‘turn on’ 4.04 (t, J ¼ 7.2 Hz, 2H, NCH2), 8.14 (d, J ¼ 7.6 Hz, 2H, 2  ArH), 8.53
fluorescence behaviors upon metal ion-binding selective for Ga3þ (d, J ¼ 7.6 Hz, 2H, 2  ArH); 1H NMR (400 MHz, CD3OD): d 1.00 (t,
or Al3þ in an aqueous solution. In addition, these probes showed a J ¼ 7.2 Hz, 3H, CH3), 1.40e1.49 (m, 2H, CH2), 1.68e1.75 (m, 2H, CH2),
very sensitive detection for this metal ion; the limit of detection 4.16 (t, J ¼ 7.2 Hz, 2H, NCH2), 8.20 (d, J ¼ 7.6 Hz, 2H, 2  ArH), 8.63
(LOD) values of probe 3 were 10 and 50 nM for Ga3þ and Al3þ, (d, J ¼ 7.2 Hz, 2H, 2  ArH).
respectively. In contrast, probes 1 and 4 did not give any change in
fluorescence intensity in the presence of various metal ions. Thus, 2.3.2. 7-Butyl-1,3,6,8-naphthalenedimide-m-benzenesulfonic acid
the current study revealed that both an optimal electron density of (compound 7)
the sensing module and suitable functional groups in the metal ion- A suspended solution of 7-butyl-1,3,6,8-naphthalenedimide,
binding site are necessary for metal ion detection with high compound 6 (646.6 mg, 2 mmol), 3-aminobenzenesulfonic acid
selectivity and sensitivity. To the best of our knowledge, this is the (346 mg, 2 mmol), and triethylamine (1.01 g, 10 mmol) was refluxed
first report of sulfonamide-based BHE-containing probes used for in dimethylformamide (DMF, 50 mL), while stirring vigorously
metal ion detection in an aqueous solution. under an argon atmosphere for 15 h. Afterwards, the solvent was
evaporated under vacuum and the crude residue was dissolved in
2. Materials and methods 2:1 water/methanol (100 mL). Through a concentrated hydrochlo-
ric acid solution (6 M), the pH was adjusted to around 3. The re-
2.1. Chemicals and materials action mixture was carefully concentrated under vacuum and the
resulting precipitate was washed with methanol several times to
All reagents and solvents used in the experiment were of yield 800 mg (83% yield) of 7-butyl-1,3,6,8-naphthalenedimide-m-
analytical grade purchased from Aldrich and used without further benzenesulfonic acid (compound 7) as a light brown solid. 1H NMR
purification. Compounds 11 and 12 were synthesized according to (400 MHz, DMSO-d6): d 0.96 (t, J ¼ 7.2 Hz, 3H, CH3), 1.36e1.45 (m,
procedures reported in literature [34,49]. The 1H and 13C NMR 2H, CH2), 1.64e1.71 (m, 2H, CH2), 4.10 (t, J ¼ 7.2 Hz, 2H, NCH2), d 7.38
spectra were recorded using a Bruker AM-400 spectrometer with (d, J ¼ 8.0 Hz, 1H, ArH), 7.51 (t, J ¼ 7.6 Hz, 1H, ArH), 7.69e7.72 (m, 2H,
Me4Si as the internal standard. The FAB mass spectra were recorded 2  ArH), 8.71 (s, 4 H, 4  ArH); 13C NMR (100 MHz, DMSO-d6):
using a Jeol JMS 700 high-resolution mass spectrometer at the KBSI d 13.78 (CH3), 19.83 (CH2), 29.60 (CH2), one signal of NCH2 merged
(Korea Basic Science Institute) Daegu Center. The UVevisible ab- with DMSO signal, 124.20, 125.56, 128.57, 128.89, 130.15, 162.95,
sorption spectra were recorded using a Shimadzu UV-2600PC 168.72.
spectrophotometer. The fluorescence spectra were measured us-
ing a fluorescence spectrometer of Horiba Jobin Yvon FL3-11 (USA)
equipped with a Xe discharge lamp using 1-cm quartz cells with a 2.3.3. 7-Butyl-1,3,6,8-naphthalenedimide-m-benzenesulfonyl
3-nm slit width. All of the measurements were recorded at 298 K. chloride (compound 8)
To a solution of 7-butyl-1,3,6,8-naphthalenedimide-m-benzene-
2.2. Theoretical calculations sulfonic acid (compound 7) (478 mg, 1 mmol) in N,N’-dime-
thylformamide (5 mL), thionyl chloride (1.19 g, 10 mmol) was added
The geometries of all of the probe compounds were optimized dropwise at 0  C. This mixture was subsequently stirred for 3 h at
via gradient-correlated density functional theory (DFT) using 0  C under an argon atmosphere. After the reaction was complete,
Becke's three-parameter exchange functional [50] and the Lee- the mixture was poured onto ice. Brownish precipitates were
Yang-Parr (B3LYP) [51] exchange correlation functional with 6- filtered and washed with water. The solids were dried for 12 h at
37  C. The desired compound 8 (7-butyl-1,3,6,8-
31G* basis sets for C, H, N, and O. B3LYP/6-31G* calculations were
employed for the excited-state optimization of probes 1e4 and naphthalenedimide-m-benzenesulfonyl chloride) was obtained as
their metal complexes. For simplicity, all calculations were carried a light brownish solid at 90% yield (450 mg). 1H NMR (400 MHz,
out in vacuum (i.e., solvent effects were not included in the calcu- CDCl3): d 1.01 (t, J ¼ 7.2 Hz, 3H, CH3), 1.43e1.53 (m, 2H, CH2),
lations). All stationary points were verified as the minima via 1.72e1.79 (m, 2H, CH2), 4.23 (t, J ¼ 7.6 Hz, 2H, NCH2), d 7.74 (d,
complete calculations of Hessian and harmonic frequency analyses J ¼ 8.0 Hz, 1H, ArH), 7.85 (t, J ¼ 8.0 Hz, 1H, ArH), 8.06 (t, J ¼ 2.0 Hz,
[52,53]. 1H, ArH), 8.20 (t, J ¼ 8.0 Hz, 1H, ArH), 8.83 (s, 4 H, 4  ArH); 13C NMR
(100 MHz, CDCl3); d 14.02 (CH3), 20.56 (CH2), 30.36 (CH2), 41.12
2.3. Syntheses of probes 1e4 (NCH2), 126.23, 127.27, 127.65, 127.73, 127.99, 131.01, 131.31, 131.98,
135.93, 136.01, 145.46, 162.80, 162.85.
2.3.1. 7-Butyl-1,3,6,8-naphthalenedimide (compound 6)
Synthetic procedures were followed according to literature [54]. 2.3.4. N-(2-(bis(2-hydroxyethyl)amino)ethyl)-7-butyl-1,3,6,8-
A suspended water solution (200 ml) of 1,4,5,8- naphthalenedimide-m-benzene sulfonamide (probe 1)
naphthalenetetracarboxylic acid dianhydride, compound 5 (1.0 g, A solution of 7-butyl-1,3,6,8-naphthalenedimide-m-benzene
3.73 mmol) was vigorously stirred at room temperature followed sulfonyl chloride, compound 8 (248.5 mg, 0.5 mmol) in CH2Cl2
 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50 41

(40 mL) was added dropwise to a stirred solution of N,N0 -Bis-(2- dropwise to a stirred solution of N,N0 -Bis-(2-hydroxyethyl)ethyl-
hydroxyethyl)ethylenediamine (149 mg, 1 mmol) in CH2Cl2 enediamine (297 mg, 2 mmol) in CH2Cl2 (20 mL) at 0  C under an
(20 mL) at 0  C under an argon atmosphere. The reaction mixture argon atmosphere. The reaction mixture was subsequently stirred
was subsequently stirred at room temperature for another 2 h. at room temperature for another 2 h. The CH2Cl2 layer was washed
After completion of the reaction, solvent was removed under with H2O (2  50 mL) and with 100 mL of an aqueous solution of
reduced pressure. The reaction mixture was crystallized from a NaCl (5%). The organics were then dried over anhydrous sodium
mixture of CH3OHeCH2Cl2 (3:1) to obtain a solid precipitate that sulfate and filtered. The filtrate was evaporated under reduced
was filtered and washed with the same solvent followed by diethyl pressure and the residue was purified via column chromatography,
ether to yield the pure probe 1 as a light brown solid (250 mg, 82%). using CH2Cl2:CH3OH (8:2) as the eluent (Rf ¼ 0.15), to afford probe 2
1
H NMR (400 MHz, DMSO-d6): d 0.96 (t, J ¼ 7.2 Hz, 3H, CH3), as a white solid (390 mg, 80% yield). 1H NMR (500 MHz, DMSO-d6):
1.36e1.45 (m, 2H, CH2), 1.64e1.72 (m, 2H, CH2), 2.48 (t, J ¼ 6.0 Hz, d 2.48 (t, J ¼ 6.0 Hz, 4H, 2  CH2), 2.56 (t, J ¼ 6.5 Hz, 2H, CH2), 2.91 (bq,
4H, 2  CH2), 2.55 (t, J ¼ 6.4 Hz, 2H, CH2), 2.91 (t, J ¼ 6.4 Hz, 2H, 2H, CH2 of SO2NHCH2), 3.36 (t, J ¼ 6.0 Hz, 4H, 2  CH2 of CH2OH),
CH2), one signal of two CH2 protons of OCH2 group's merged with 4.33 (t, J ¼ 5.5 Hz, 2H, 2  OH), 7.57 (bs, SO2NH), 7.72 (d, J ¼ 8.0 Hz,
water, 4.09 (t, J ¼ 7.2 Hz, 2H, CH2), 4.34 (t, J ¼ 5.6 Hz, 2H, 2  OH), 1H, ArH), 7.78 (t, J ¼ 8.0 Hz, 1H, ArH), 7.90e7.94 (m, 4H, 4  ArH),
7.59 (br s, 1H, SO2NH), 7.75 (d, J ¼ 8.0 Hz, 1H, ArH), 7.81 (t, J ¼ 8.0 Hz, 8.52e8.55 (m, 4H, 4  ArH); 13C NMR (125 MHz, DMSO-d6): d 40.99,
1H, ArH), 7.94 (d, J ¼ 8.0 Hz, 1H, ArH), 7.97 (s, 1H, ArH), 8.71 (bs, 4H, 53.92, 56.53, 58.99, 122.59, 126.19, 127.22, 127.46, 127.91, 129.87,
4  ArH); 13C NMR (100 MHz, DMSO-d6): d 13.73 (CH3), 19.82 (CH2), 130.74, 131.46, 133.30, 134.54, 136.67, 141.34, 163.65; HRMS (EI):
29.56 (CH2), one signal of CH2 merged with DMSO signal, 40.99 calcd. for C24H26N3O6S (MþHþ): 483.1542, found: m/z 483.1544.
(CH2), 53.94 (CH2), 56.54 (2  CH2), 58.98 (2  CH2), 126.26, 126.50,
126.63, 126.85, 127.43, 130.07, 130.42, 133.17, 136.20, 141.46, 162.65,
2.3.8. N-(2-(bis(2-hydroxyethyl)amino)ethyl)pyrene-1-
162.90. HRMS (EI): calcd. for C30H33N4O8S (MþHþ): 609.2019,
sulfonamide (probe 3)
found: m/z 609.2022.
A solution of pyrenesulfonyl chloride, compound 12 (300 mg,
1.0 mmol) in CH2Cl2 (40 mL) was added dropwise to a stirred so-
2.3.5. 1,8-Naphthalimide-m-benzenesulfonic acid (compound 10) lution of N,N0 -Bis-(2-hydroxyethyl)ethylenediamine (297 mg,
A suspended solution of 1,8-naphthalic anhydride, compound 9 2 mmol) in CH2Cl2 (20 mL) at 0  C under argon. The reaction
(1.98 g, 10 mmol), 3-aminobenzenesulfonic acid (1.73 g, 10 mmol), mixture was subsequently stirred at room temperature for another
and triethylamine (2.02 g, 20 mmol) was refluxed in ethanol 2 h. The CH2Cl2 layer was washed with H2O (2  50 mL) and with
(50 mL) with vigorous stirring under an argon atmosphere for 12 h. 100 mL of an aqueous solution of NaCl (5%), dried over anhydrous
After completion of the reaction (as observed via TLC), the mixture sodium sulfate, and filtered. The filtrate was evaporated under
was cooled to room temperature and held for one day. The crys- reduced pressure and the residue was purified via column chro-
talline solids were then separated, filtered, and washed with matography using CH2Cl2:CH3OH (8:2) as the eluent (Rf ¼ 0.20), to
diethyl ether, dichloromethane, and methanol. The solids were yield probe 3 as a light yellow solid (335 mg, 81% yield). 1H NMR
dried to obtain a pure brownish white crystalline compound 10 (500 MHz, DMSO-d6): d 3.20e3.24 (m, 4H, 2  CH2), 3.28e3.30 (m,
(3.25 g, 92% yield). In the 1H and 13C NMR of compound 10, it was 4H, 2  CH2), 3.69 (br t, 4H, 2  OCH2), 5.25 (br s, 2H, 2  OH), 8.24
determined that triethylamine was entrapped within the crystal (t, J ¼ 7.5 Hz, 1H, ArH), 8.32 (d, J ¼ 9.0 Hz, 1H, ArH), 8.42 (d,
structure of the compound. 1H NMR (400 MHz, DMSO-d6): d 7.40 (d, J ¼ 9.0 Hz, 1H, ArH), 8.46e8.52 (m, 4H, 4  ArH), 8.62 (d, J ¼ 8.0 Hz,
J ¼ 7.6 Hz, 1H, ArHg), 7.55 (t, J ¼ 7.6 Hz, 1H, ArHf), 7.66 (s, 1H, ArHd), 1H, ArH), 8.95 (d, J ¼ 9.5 Hz, 1H, ArH), 9.42 (br s, 1H, SO2NH); 1H
7.75 (d, J ¼ 7.6 Hz, 1H, ArHe), 7.97 (t, J ¼ 8.0 Hz, 2H, 2  ArHb), 8.58 NMR (400 MHz, CD3OD): d 3.27 (t, J ¼ 6.4 Hz, CH2), 3.41 (t,
(d, J ¼ 8.0 Hz, 4H, 2  ArHa þ 2  ArHc), 8.91 (br s, OH); 13C NMR J ¼ 5.2 Hz, CH2), 3.43e3.49 (m, 4H, 2  CH2), 3.88 (t, J ¼ 5.2 Hz, 4H,
(100 MHz, DMSO-d6): d 122.63, 125.47, 126.30, 127.21, 127.86, 2  CH2), 8.19 (t, J ¼ 8.0 Hz, 1H, ArH), 8.23 (d, J ¼ 9.2 Hz, 1H, ArH),
128.34, 129.36, 130.68, 131.44, 134.42, 135.47, 148.97, 163.70. 8.34 (t, J ¼ 9.2 Hz, 1H, ArH), 8.37 (d, J ¼ 8.0 Hz, 1H, ArH), 8.42e8.44
(m, 3H, 3  ArH), 8.70 (d, J ¼ 8.0 Hz, 1H, ArH), 8.99 (d, J ¼ 9.6 Hz, 1H,
2.3.6. 1,8-Naphthalimide-m-benzenesulfonyl chloride (compound ArH); 13C NMR (125 MHz, DMSO-d6): d 37.09 (CH2), 52.48 (CH2),
11) 55.03 (2  CH2), 55.18 (2  CH2), 123.13, 123.21, 124.25, 124.33,
Thionyl chloride (1.19 g, 10 mmol) was added to a solution of 1,8- 126.81, 126.94, 127.09, 127.13, 127.20, 129.67, 129.75, 130.12, 130.51,
naphthalimide-m-benzenesulfonic acid, compound 10 (1.77 g, 131.48, 134.17; HRMS (EI): calcd. for C22H25N2O4S (MþHþ):
5 mmol) in N,N’-dimethylformamide (10 mL). This mixture was 413.1535, found: m/z 413.1537.
subsequently stirred for 3 h at 0  C under an argon atmosphere.
After completion of the reaction, the mixture was poured onto ice. 2.3.9. N-(2-aminoethyl)pyrene-1-sulfonamide (probe 4)
The brownish white precipitate was filtered and washed with A solution of pyrenesulfonyl chloride, compound 12 (300 mg,
water. The solids were dried for 12 h at 50  C. The desired 1,8- 1.0 mmol) in CH2Cl2 (80 mL) was added dropwise to a stirred so-
naphthalimide-m-benzene sulfonyl chloride (compound 11) prod- lution of ethylenediamine (300 mg, 5 mmol) in CH2Cl2 (40 mL) at
uct was obtained as a light brownish solid at 95% yield (1.76 g). 1H 0  C under an argon atmosphere and stirred for another 2 h. After
NMR (400 MHz, CDCl3): d 7.73 (d, J ¼ 8.0 Hz, 1H, ArHg), 7.80e7.86 completion of the reaction (as observed from TLC analysis), the
(m, 3H, 2  ArHb þ 1 ArHf), 8.06 (s, 1H, ArHd), 8.16 (d, J ¼ 8.0 Hz, 1H, CH2Cl2 layer was washed with 1 N HCl, leading to the formation of
ArHe), 8.33 (d, J ¼ 8.0 Hz, 2H, 2  ArHa), 8.68 (d, J ¼ 8.0 Hz, 2H, solid precipitates in a liquid solution of CH2Cl2 and water. This
2  ArHc); 13C NMR (100 MHz, CDCl3 þ DMSO-d6 (4:3)): d 122.12, solution was filtered and washed with CH2Cl2 to obtain the pure
125.61, 126.33, 126.69, 127.88, 128.51, 129.87, 130.83, 131.31, 134.21, probe 4 as a light yellow shining solid (280 mg, 86% yield). 1H NMR
135.19, 146.75, 163.48. (400 MHz, DMSO-d6): d 2.81 (t, J ¼ 6.8 Hz, 2H, CH2), 3.02 (t,
J ¼ 6.8 Hz, 2H, CH2), 7.93 (bs, 2H, protonated NH2), 8.24 (t, J ¼ 7.6 Hz,
2.3.7. N-(2-(bis(2-hydroxyethyl)amino)ethyl)1,8-naphthalimide- 1H, ArH), 8.33 (d, J ¼ 8.8 Hz, 1H, ArH), 8.42e8.53 (m, 5H, 5  ArH),
m-benzenesulfonamide (probe 2) 8.62 (d, J ¼ 8.0 Hz, 1H, ArH), 8.98 (d, J ¼ 9.2 Hz, 1H, ArH); 13C NMR
A solution of 1,8-naphthalimide-m-benzenesulfonyl chloride, (100 MHz, DMSO-d6): d 38.63 (CH2), one signal of 2nd CH2 merged
(compound 11) (372 mg, 1.0 mmol) in CH2Cl2 (40 mL) was added with the signal of DMSO, 123.19, 123.27, 124.25, 124.33, 124.38,
42 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50

126.94, 127.04, 127.16, 127.24, 127.32, 129.74, 129.86, 130.24, 130.58, imide groups at both sides of the naphthalene ring (Scheme 1).
131.49, 134.25; HRMS (EI): calcd. for C18H17N2O4S (MþHþ): Thus, this probe possesses the most electron-deficient fluorophore
325.1011, found: m/z 325.1011. among the probes in this study. The UVeVisible spectrum of probe
1 possessed three absorbance peaks at 340, 361, and 381 nm,
2.4. Preparation for fluorescent study spectral characteristics of the NDI fluorophore when solubilized in
ethanol. The absorption peaks were slightly shifted when
The probes 1e4 stock solutions (1 mM) are generally prepared increasing the water content in ethanolic solutions (Fig. S1a). The
in DMSO/EtOH. The probes test solutions (1 mM) for AIE charac- intensity of the absorbance tail tended to increase as the water
terization were prepared in water/EtOH mixtures with different content increased, particularly for a 100% water solution, suggestive
water fractions. The solutions of metal ions (104 M) were prepared of occurrence of probe aggregation. When probes 2 and 3 were
from their perchlorate salts. For all fluorescent measurements, the tested, probe 2 with NPI unit exhibited broad absorption spectra
excitation wavelength was 340 nm. Each measurement was per- with a maximum at ~340 nm and an increase in tail intensity
formed in triplicate. whereas probe 3 with pyrene ring yielded absorption spectra with
multiple maxima about 340, 351, and 355 nm with a slight increase
3. Results and discussion in tail intensity (Fig. S1b,c). This result suggested that aggregates
formed by probe 2 were rather large at high water concentrations
3.1. Aggregation-induced emission (AIE) of the probes while probe 3 formed no aggregates or small aggregates under the
same conditions.
Probe 1 contains an NDI unit with two electron-withdrawing The fluorescence behaviors of the probes were further

Fig. 1. Spectral changes in the fluorescence spectra of probes 1 (a), 2 (b), and 3 (d), in addition to the corresponding bar diagrams for the emission maxima of probes 2 (c) and 3 (e)
with a variation in water content from 0% to 100% for binary solutions of ethanol and water. The probes were used at 1 mM and excited at 340 nm. The inset in (a) shows a magnified
image of the spectral region indicated by the red rectangle. E and W denote ethanol and water, respectively. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)
 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50 43

investigated at a probe concentration of 1 mM. A change in fluo- displayed an increase in fluorescence intensity at 398 nm and
rescence intensity was measured with an increase in water content 383 nm, respectively, with an increase in water content, a typical
for ethanolic solutions (Figs. 1 & S2). An excitation wavelength of feature of AIE behavior. The fluorescence spectra of these probes
340 nm was employed as this wavelength yielded the highest were also measured with increasing probe concentration in pure
fluorescence intensity (Figs. S3 & S4). In the case of probe 1, we ethanol (Fig. S5). In the case of probe 1, the emission maximum at
could not find aggregation-induced emission (AIE) enhancement 381 nm gradually decreased with an increase in the probe con-
[55] or aggregation-caused quenching (ACQ) [56,57], known to be centration, along with a concomitant increase in peak intensities at
popular for many NDI-containing sensor systems [58,59]. Instead, 406 and 530 nm, indicating that the aggregate sizes were slightly
the emission maximum was red-shifted from 381 nm to 387 nm, dependent on the probe concentration. By contrast, probes 2 and 3
probably due to a change in the polarity of the binary solvent with the different fluorophores exhibited an increase in fluores-
systems. Combined with the UVeVisible spectrum, the results cence intensity under the same conditions, consistent with their
indicated that probe 1 appeared to form aggregates, but the sizes of potential AIE behaviors. However, probe 2 yielded a rather gradual
the aggregates did not vary to a great extent with increasing water increase in fluorescence intensity, as opposed to probe 3, which
content within the ethanolic solutions. By contrast, probes 2 and 3 exhibited a sudden jump in fluorescence intensity at a high probe

Fig. 2. FE-SEM images of probes 1 (aef), probe 2 (gei), and probe 3 (jel) in three different water concentrations (0, 50, and 90%). D, E, and W denote DMSO, ethanol, and water,
respectively. Due to solubility issues, DMSO was added to the sample solutions of probes 1 and 2 while the original binary mixture of ethanol and water was used for probe 3.
Morphological changes of the aggregates were observed for probe 1 with no substantial size variation whereas the sizes of aggregates formed by probes 2 and 3 increased
significantly with an increased amount of water.
44 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50

concentration (e.g., 20 mM). Based on these results, it was expected observed for probe 2, but with a lesser degree of response towards
that probe 1 may undergo changes in morphology rather than size Ga3þ or Al3þ (Fig. S10). By contrast, probe 1 did not yield a signif-
of the aggregates with an increase in water concentration whereas icant change in fluorescence intensity in the presence of individual
probes 2 and 3 were likely to form larger aggregates with an in- metal ions (Fig. S11). The difference in sensitivity for metal ion-
crease in water concentration. sensing between the individual probes was likely due to varia-
In order to support this expectation, FE-SEM images of indi- tions in electron density for the sensing module (i.e., fluorophore).
vidual probes in three different water concentrations (0, 50, and Probe 3, the most effective for sensing Ga3þ and Al3þ, contains an
90%) were obtained. Due to the solubility issue, an equal amount of electron-rich pyrene fluorophore while probe 1 with an electron-
DMSO was added to the sample solutions of probes 1 and 2 (Fig. 2). poor NDI unit was unable to detect those metal ions. Probe 2
The results showed that, with an increase in water content, probe 1 containing an NPI fluorophore with an intermediate electron den-
underwent morphological changes from a nest-like structure with sity demonstrated the good sensing behavior toward these metal
multiple layers (10e15 mm) to a beautiful flower with sharp edges ions, but was less effective than probe 3. Fluorescence intensity was
(15 mm) to long leaves (6e23 mm) (Fig. 2aef), as expected. By measured at 396 and 383 nm for probes 2 and 3, respectively. The
contrast, the FE-SEM images of probes 2 and 3 revealed that the fluorescence emission of probe 2 or 3 was not increased in the
aggregates formed by these probes were enlarged when the water presence of other metal ions (Fig. 3a and b, Figs. S9 and S10). This
content increased from 0% to 90%. Specifically, the sizes of aggre- result indicates the high selectivity of these probes for Ga3þ or Al3þ
gates increased from ~1.7 mm to ~5 mm to ~13 mm for probe 2. Ag- detection in aqueous environments. From metal ion titration ex-
gregates formed by probe 3 underwent much more dramatic size periments, we also observed that the fluorescence emission of each
changes than probe 2; aggregate sizes varied from no aggregates to probe (probe 2 or 3) was enhanced along with concentrations of
~500 nm to 1.5 mm (Fig. 2gei). The size increase observed for the metal ions (Ga3þ and Al3þ) (Fig. 3c and S9e10). Association con-
aggregates of probe 3 was further confirmed via TEM analysis un- stants (Ka) between each probe and Ga3þ or Al3þ were calculated by
der the same conditions. Specifically, the aggregate sizes of probe 3 fitting the titration data (Fig. 3d). The complex of probe 3 with Ga3þ
increased from 100-200 nm to 1e2 mm when the water content was and Al3þ yielded a higher association constant (Ka ¼ 2.99  105 and
increased from 50% to 90% in ethanolic solutions (Fig. S6), further 1.23  105 M1, respectively) than the counterparts of probe 2
supporting AIE behavior of this probe with increasing water con- (Ka ¼ 1.42  105 and 1.01  105 M1, respectively).
centration. Note that we did not take TEM image of the probe in an Probe 3 was bound to Ga3þ two times stronger than probe 2
ethanol-only solvent as no aggregates were detected in the FE-SEM (Ka ¼ 2.99  105 M1 vs 1.42  105 M1) although both probes
image at this condition (Fig. 2j). AIE behavior is known to occur due possess the same metal-binding motif (i.e., BHE). This fitting data
to restricted molecular motion of the probe, resulting in inhibition suggested 1:1 complex formation between each probe and the
of non-radiative relaxation and increase in fluorescence intensity metal ion, which was confirmed by job's plot (Fig. S12). The limit of
[60]. Because of their suitability in practical applications, several detection (LOD) of probe 3 for Ga3þ was observed to be 10 nM when
probes with AIE behavior have been successfully applied as metal a linear relationship between I383 nm and [Ga3þ] was used (inset in
ion [61] and biomolecule sensors [62] as well as gas and pH sensors Fig. 3d). The LOD of the probe for Al3þ was found to be 50 nM using
[63,64]. Based on the significant contribution of AIE-based probes a similar method (Fig S9 and S24). It is notable that the LOD value
to chemical sensor field, probes 2 and 3 are likely to be used to obtained for probe 3 was the lowest of those previously reported
detect metal ions dissolved in aqueous media. for Ga3þ (Table S1) [65e70] or Al3þ (Table S2) [29e33]. We have
also performed the UVeVis. studies of probes 2 and 3 with various
3.2. Fluorescence detection for diverse metal ions metal ions (Fig. S13). There was no significant change in the spectra,
except an increase in the tail-off absorbance intensity selective for
In order to investigate the potential for these probes to be Ga3þ or Al3þ. This result suggests the presence of large aggregates
applied toward metal ion-sensing, we first employed electron-rich formed by the probes in the presence of either metal ion.
probe 3 for the detection of various metal ions dissolved in ethanol. As probes 2 or 3 tended to form rather large aggregates under
The probe and individual metal ion concentrations were 1 mM and the tested solvent composition (1:1 mixture of pH 7.4 HEPES:
100 mM, respectively. Metal ions included monovalent (Naþ, Kþ, EtOH), we investigated the effects of Ga3þ/Al3þ binding to the
and Agþ), divalent (Ca2þ, Cd2þ, Co2þ, Ni2þ, Cu2þ, Hg2þ, Pb2þ, and probes on their aggregate sizes. FE-SEM or TEM images for the
Zn2þ) and trivalent metal ions (Al3þ, Ga3þ, Fe3þ, and Cr3þ). Probe 3 individual probe-metal complexes revealed substantial increases in
triggered a “switch on” enhancement state with regard to the the sizes of aggregates for both probes (probes 2 and 3) in the
fluorescence emission intensity in the presence of Ga3þ or Al3þ presence of either Ga3þ or Al3þ (Fig. S14 for probe 2 & Fig. 4 for
(Fig. S7), but did not exhibit a noticeable response to the presence probe 3). The potential interference of other metal ions with regard
of other metal ions (Naþ, Kþ, Ca2þ, Cd2þ, Co2þ, Ni2þ, Cu2þ, Hg2þ, to the detection of Ga3þ and Al3þ was monitored by observing any
Agþ, Pb2þ, Zn2þ, and Cr3þ) with the exception of Fe3þ; the probe changes in fluorescence intensity following the addition of various
responded to Fe3þ to a small degree. In water-ethanol composition metal ions at 100 mM into a solution of [probe 3 þ Ga3þ or Al3þ
for metal ion-sensing, we monitored the fluorescence intensity of (20 mM)] (Fig. S15). In the presence of five-fold excess of another
probe 3 at 383 nm in the presence of Ga3þ with increasing water metal ion, probe 3 has an ability to selectively detect target metal
content (Fig. S8). A large enhancement in the fluorescence emission ion (Ga3þ or Al3þ).
was observed with Ga3þ for ethanol-water composition (pH 7.4
HEPES buffer) changed from 10:0 to 3:7. However, the fluorescence 3.3. Selective and sensitive binding of probes 2 and 3 with Ga3þ or
emission tended to decrease with further increases in water con- Al3þ
tent. Among a wide range of water concentrations, we selected a
50% water-ethanol solution for metal ion detection (1:1 mixture of In order to understand the role of two hydroxyl groups within
pH 7.4 HEPES buffer and EtOH) as this solvent composition contains the BHE moiety as pertaining to metal ion binding and sensing, an
water amount enough to apply for an aqueous system and the ethylenediamine analog of probe 3 (i.e., probe 4) was synthesized.
probe alone gave minimal fluorescence intensity. Under these Fluorescence experiments showed that this probe tended to yield
conditions, probe 3 exhibited “switch on” fluorescence emission in AIE behavior with increasing water content (Fig. S3g,h) or probe
the presence of Ga3þ or Al3þ (Figs. 3 and S9). Similar results were concentration (Fig. S4d). When applied at 14 mM, the probe reached
 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50 45

Fig. 3. (a) Fluorescence spectra of probe 3 in the presence of various metal ions in an aqueous solution (pH 7.4 HEPES: EtOH ¼ 1:1). The probe and metal ions were used at 1 mM and
100 mM, respectively. (b) Bar diagram for the relative fluorescence emission enhancement of probe 3 in the presence of individual metal ions and a corresponding solution image
including Al3þ or Ga3þ (inset). (c) Changes in the fluorescence spectrum in the course of probe 3 titration with Ga3þ at an excitation wavelength of 340 nm. (d) Fluorescence
emission intensity of probe 3 at 383 nm as a function of [Ga3þ] (mM). (Inset) an enlarged view in the low [Ga3þ] region indicated by a green rectangle, which was used to determine
the LOD of the probe for Ga3þ detection. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4. FE-SEM and TEM images of aggregates formed by probe 3 (center; b & e) alone or complexed with Al3þ (left; a & d) or Ga3þ (right; c & f) in a 50% water-ethanol solution. The
upper panel (aec) presents FE-SEM images while the lower panel (def) presents TEM images. Aggregate sizes increased due to complex formation of individual probes with each
metal ion. Note the different scale bars for each image.

out of the detection limit of fluorescence emission (Fig. S4d). essential for the strong binding of probe 3 toward Ga3þ or Al3þ.
However, in contrast to probe 3, probe 4 with an ethylenediamine Combined together, it was evident that both an electron-rich flu-
unit did not exhibit any significant changes in fluorescence emis- orophore (i.e., pyrene) and a suitable metal ion-binding unit (e.g.,
sion intensity upon the addition of various metal ions (Fig. S16) BHE) were essential for the selective and sensitive sensing of Ga3þ
under the same conditions used for probe 3. The comparison of or Al3þ.
probe 3 vs 4 indicates that the presence of two hydroxyl groups is In order to determine which part of the probe is involved in
46 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50

Fig. 5. 1H NMR spectral changes of probes 2 (A) and 3 (B) alone (i) or upon the addition of 1 equiv. (ii) and 2 equiv. (iii) of Ga3þ in D2O:CD3OD ¼ 3:5. (iv) Ga3þ-only solution was used
as a control. Metal ions were added in the form of Ga(ClO4)3.

metal ion binding, 1H NMR titrations of probes 2 and 3 were per- observed. Upon the addition of one equivalent of Ga3þ, the signals
formed with Ga3þ or Al3þ in D2O-CD3OD (3:5) (Fig. 5 and Fig. S18). for the BHE protons, labelled b, d, and e, shifted downfield from
The associated changes in chemical shifts for the individually d 3.450 to 3.498, from d 3.400 to 3.450, and from d 3.885 to 3.920,
labelled protons are summarized in Tables 1 and 2. The NMR peaks respectively, whereas the signals corresponding to the aromatic
in the aromatic region were partially assigned to pyrene protons protons were not significantly affected by metal ion binding.
based on the coupling contants obtained from the 1H NMR spec- Associated chemical shift changes for the labelled protons are
trum and proton-proton correlations found in 1H-1H COSY spec- summarized in Table 2. The addition of extra equivalents of Ga3þ
trum (Fig. S17). Upon the addition of one equivalent of Ga3þ, the did not further affect the chemical shifts of the BHE and aromatic
BHE proton signals of probe 2, labelled b, c, d, and e, shifted protons, again supportive of a 1:1 complex formation between
downfield from d 3.076 to 3.517, from d 2.656 to 3.428, from d 2.625 probe 3 and Ga3þ (Fig. 5). When Al3þ was used instead of Ga3þ, a
to 3.480, and from d 3.554 to 3.943, respectively, whereas the ar- similar trend was observed, although the degree of chemical shift
omatic proton signals, labelled g, h, i, and j, shifted slightly down- change was much smaller for this metal ion (Fig. S18; Table 2). This
field from d 7.873 to 7.934, d 8.033 to 8.072, d 7.851 to 7.885, and result is likely attributed to stronger binding of probe 3 to Ga3þ than
d 7.672 to 7.720, respectively. During the addition of extra equiva- Al3þ, which was reflected by the higher association constant. With
lents of Ga3þ to the probe, no further changes were observed, regard to the system of probe 2 and Ga3þ/Al3þ, differences in
supporting the formation of a 1:1 complex between probe 2 and chemical shift change (Dd) and the association constants (Ka) be-
Ga3þ (Fig. 5). Similar chemical shift changes were detected during tween Ga3þ and Al3þ binding of the probe were relatively small
the addition of Al3þ to the probe (Table 1; Fig. S18). When probe 3 compared to the corresponding probe 3 system. This result in-
was used for 1H NMR titration experiment, similar trends were dicates the greater differentiation of probe 3 for Ga3þ over Al3þ

Table 1
Changes in proton chemical shifts (Dd)a measured for probe 2 induced by the complexation of Ga3þ or Al3þ in D2O:CD3OD ¼ 3:5.

Protons a b c d e f g h i j

Dd [2 þ Ga3þ] …. 0.441 0.772 0.855 0.399 … 0.061 0.039 0.034 0.048

[2 þ Al3þ] …. 0.441 0.764 0.855 0.392 … 0.059 0.037 0.036 0.048
a
Positive and negative values indicate a downfield- or upfield-shift of the proton peaks in the 1H NMR spectra upon complexation.

Table 2
Changes in proton chemical shift (Dd)a measured for probe 3 induced by the complexation with Ga3þ or Al3þ in D2O:CD3OD ¼ 3:5.

Protons a b c d e f g
3þ
Dd [3 þ Ga ] …. 0.048 …. 0.050 0.035 … 0.005
[3 þ Al3þ] …. 0.014 …. 0.017 0.007 … 0.005
a
Positive and negative values indicate downfield- and upfield-shift in the proton peaks in the 1H NMR spectra upon complexation.
 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50 47

detection than probe 2. As the chemical shifts of the BHE protons (b, electron-rich pyrene produced the highest LUMO energy level.
c, d, and e) were changed more significantly than those of the ar- DFT calculations were also used to predict the chemical struc-
omatic protons of the fluorophore upon metal ion addition, this tures of the probe-metal ion complexes. From a literature [71],
moiety is primarily responsible for strong metal ion binding. probe 2 and 3 were predicted to form an octahedral complexes with
Additionally, the 1H NMR signals for the protons (c and d) were Ga3þ or Al3þ. Based on this structural information, density func-
shifted more downfield than for the protons (b and e) upon metal tional theory (DFT) calculations were performed at the B3LYP/6-
ion binding, suggesting that the tertiary amine rather than the 31G* level to find optimized structures, orbital energy levels, and
sulfonamide nitrogen is strongly involved in Ga3þ or Al3þ binding HOMO and LUMO localizations for the probe-metal complexes
(Tables 1 and 2). In order to investigate aggregation effect on the (Fig. 6 and Fig. S22e23). As expected, one tertiary amine and two
chemical shifts of the probe, we have carried out the 1H NMR ti- hydroxy oxygens were in coordination with the metal ion (Ga3þ or
trations of probe 3 with increasing content of water in CD3OD. In Al3þ). Surprisingly, a sulfonamide oxygen rather than sulfonamide
the absence of Ga3þ, the two most downfield signals were gradually nitrogen occupied the fourth coordination site, along with two
upfield-shifted whereas the other aromatic signals were ethanol molecules positioned at the other sites. Upon complexation
downfield-shifted with increasing water content (Fig. S19). In with Ga3þ or Al3þ, the HOMOs of both probes were localized on the
contrast, the aliphatic signals underwent rather minor changes in fluorophores whereas the LUMO locations were dependent on the
the chemical shifts, but instead were significantly broadened under probe-metal ion combination (Fig. S24). The LUMOs were primarily
the condition. These results indicate that probe aggregates were distributed over the metal binding site when forming a metal
enlarged with increasing water content in the absence of Ga3þ. A complex with Ga3þ. In the case of Al3þ complexation, the LUMO
similar trend in chemical shifts was observed for the aromatic was localized over the benzene ring and pyrene fluorophore for
protons of probe 3 in the presence of Ga3þ (Fig. S20). However, the probes 2 and 3, respectively. The HOMO e LUMO energy gaps (DE)
aliphatic peak broadenings were less obvious in the presence of the of each probe were significantly decreased upon complexation with
metal ion, presumably ascribed to the rigidification of this alkyl the metal ion (Ga3þ or Al3þ). The probe-Ga3þ complexes were also
moiety as a result of complex formation with the metal ion. Thus, smaller than their corresponding Al3þ counterparts in this regard.
probe aggregates became enlarged with increasing water concen- Specifically, the energy gap of the [probe 3 þ (Ga3þ/Al3þ)] complex
tration irrespective of the absence or presence of Ga3þ, which is in was smaller than that of probe 3 alone (i.e., DE[3 þ
3þ 3þ
well agreement with the FE-SEM and TEM data. (Ga )] ¼ 2.51 eV < DE[3 þ (Al)] ¼ 2.90 eV < DE3 ¼ 3.70 eV).
Similarly, the energy gap of the [probe 2 þ (Ga3þ/Al3þ)] complex
3.4. Density functional theory (DFT) calculations was smaller than that of probe 2 by itself (i.e., DE[2 þ
3þ 3þ
(Ga )] ¼ 1.82 eV < DE[2 þ (Al)] ¼ 2.40 eV < DE2 ¼ 2.99 eV)
Further insight regarding fluorescence behavior was obtained (Fig. S24). The relatively small orbital energy gaps for the Ga3þ-
by calculating the energy levels and orbital distributions for the probe complex compared to those of the Al3þ-probe complex could
HOMO and LUMO of individual probes via density functional theory partly be attributed to the relatively strong binding of Ga3þ, as
(DFT)-based calculations at the B3LYP/6-31G* level using Spartan observed via fluorescence and 1H NMR titration experiments.
10 software. The fluorescence emission of probes 1 to 3 could be
observed through the electronic transitions of an excited electron 3.5. Fluorescence turn-on mechanism for the probe-Ga3þ/Al3þ
from the LUMO to HOMO (Fig. S21). DFT calculations showed that systems
the HOMO and LUMO were primarily distributed on the pyrene ring
for probe 3 while the HOMO and LUMO were localized within the The observed large fluorescence emission enhancement of
BHE and NDI/NPI regions, respectively, for probes 1 and 2. The probe 2 or 3 as a result of metal ion (Ga3þ or Al3þ) binding in this
HOMO e LUMO energy gap (DEH/L) of probes 1 to 3 increased with study is likely attributed to the suppression of photo-induced
an increased electron density of the fluorophore in the following electron transfer (PET) (Scheme 2). In the absence of metal ion
order: DE3 ¼ 3.70 eV > DE2 ¼ 2.99 eV > DE1 ¼ 2.08 eV). The binding, one of lone pair electrons on the tertiary nitrogen within
HOMO energy levels for the probes were in the range of 5.55 eV the BHE unit could be transferred to the HOMO of the photo-excited
to 5.71 eV, being relatively similar to each other. However, the probes prior to the relaxation of an excited electron to the ground
LUMO energy levels varied from 3.63 eV to 1.98 eV depending state, resulting in fluorescence quenching. In the course of the
on the electron density of the fluorophore. Probe 3 with the most complex formation with Ga3þ/Al3þ, however, the lone pair of the

Fig. 6. Energy-minimized structures of probe 3 in the absence or presence of Ga3þ or Al3þ. The individual structures were obtained from DFT calculations at the B3LYP/6-31G* level.
For the probe 3-Al3þcomplex, bond distances were as follows; a' ¼ 1.943 Å, b' ¼ 1.972 Å, c' ¼ 2.115 Å, d' ¼ 1.860 Å, e' ¼ 1.939 Å, f' ¼ 1.950 Å. For the probe 3-Ga3þ complex, bond
distances were as follows; a' ¼ 2.039 Å, b' ¼ 2.220 Å, c' ¼ 2.032 Å, d' ¼ 2.053 Å, e' ¼ 1.953 Å, f' ¼ 2.061 Å.
48 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50

Scheme 2. A plausible fluorescence turn-on mechanism for probe 3 upon binding to Ga3þ or Al3þ. In the absence of the metal ion, the lone pair of tertiary amine quenches the
fluorescence emission of pyrene via photo-induced energy transfer (PET). The PET of probe 3 is suppressed by Ga3þ or Al3þ binding to the tertiary amine, resulting in restoring
fluorescence emission of the fluorophore.

probes should be used to coordinate with the metal ion, thus being demonstrated via Ga3þ/Al3þ detection in a real sample. A fluores-
incapable of its electronic transition to the HOMO with an unpaired cence titration experiment of probe 3 was performed using Korean
electron (i.e., PET suppression). Thus, probe-metal complexation beer with Ga3þ/Al3þ. The LOD values of the probe for Ga3þ and Al3þ
led to fluorescence emission enhancement. Another significant sensing in this system were 10 and 50 nM, respectively (Fig. S25).
contribution to Ga3þ/Al3þ detection by probes 2 and 3 could be These were identical to the same values obtained from the exper-
attributed to a substantial increase in the size of their aggregates iments using a 1:1 mixture of ethanol and pH 7.4 buffer.
upon metal ion binding as these probes exhibited AIE behaviors
(Scheme 2).
4. Conclusions

3.6. Solution pH effect and real sample application for Ga3þ or Al3þ Three new sulfonamide-based probes with BHE (probes 1, 2, and
detection 3) were synthesized via the simple synthetic protocols. Among
these, probes 2 and 3 exhibited AIE behaviors with an increased
Additionally, pH effects on the abilities of the probes to detect water content in an EtOH-water mixed solvent system, indicated by
Ga3þ/Al3þ were investigated by measuring the fluorescence in- UVevisible absorption and fluorescence spectroscopy in addition to
tensity of individual probes and their complexes with solution pH FE-SEM/TEM analysis. Probes 2 and 3 bearing NPI and pyrene flu-
variations. The fluorescence emission intensity of probe 2 remained orophores, respectively, displayed “OFF-ON” fluorescence emission
low in the range of pH > 6.0, but increased abruptly when the so- behaviors in the presence of Ga3þ or Al3þ, but not in the presence of
lution pH decreased to less than 6.0 (Fig. 7a). This was likely due to other metal ions. Sensitivity to the metal ion was slightly higher for
the protonation of the tertiary amine in acidic media. Upon the probe 3 than probe 2. Probe 1 possessing the electron-poor NDI unit
addition of Ga3þ/Al3þ, fluorescence emission of the probe remained and probe 4 possessing only the ethylenediamine group gave little
high in a solution pH < 9.0, but decreased significantly when the response to the presence of Ga3þ or Al3þ with regard to their
solution pH was greater than >9.0. At high solution pH values, the fluorescence emission intensities, indicating that both high elec-
sulfonamide NH was likely to be deprotonated, inducing a complete tron density of a fluorophore and two hydroxyl groups of the
change in the probe-metal binding mode. This titration result in the binding unit are necessary for selective binding and sensing of the
absence or presence of Ga3þ/Al3þ indicates that the probe is capable metal ions. Probes 2 and 3 formed a 1:1 complex with Ga3þ or Al3þ
of detecting Ga3þ/Al3þ between a range of solution pH values from while the highest association constant (Ka ¼ 2.99  105 M1) was
6.0 to 9.0, typical of most aqueous solutions. A similar result was obtained for the probe 3-Ga3þ complex. The limit of detections
obtained for probe 3 (Fig. 7b). The utility of probe 3 was further (LODs) of probe 3 for Ga3þ and Al3þ were 10 and 50 nM,

Fig. 7. pH dependency of fluorescence intensity for probes 2 (a) and 3 (b) (1 mM, H2O: EtOH (1:1 v/v)) and their Ga3þ/Al3þ-complexes in a pH range from 2.0 to 13.5. The fluo-
rescence intensities for probes 2 and 3 were taken from maxima at 396 nm and 383 nm, respectively, with lex ¼ 340 nm.
 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50 49

respectively. We also observed that the size of aggregates for these High-brightness light emitting diodes using dislocation-free indium gallium
nitride/gallium nitride multiquantum-well nanorod arrays, Nano Lett. 4
probes increased significantly upon the binding of either Ga3þ or
(2004) 1059e1062.
Al3þ, revealed by the FE-SEM and TEM analysis. The binding be- [20] R. Trotta, P. Atkinson, J.D. Plumhof, E. Zallo, R.O. Rezaev, S. Kumar, S. Baunack,
haviors of these probes with Ga3þ/Al3þ were further supported by J.R. Schroter, A. Rastelli, O.G. Schmidt, Nanomembrane quantum-light-
1 emitting diodes integrated onto piezoelectric actuators, Adv. Mater. 24
H NMR titration. The DFT calculations enabled us to understand
(2012) 2668e2672.
binding mode and mechanism of interaction on the basis of en- [21] J. Goldberger, R. He, Y. Zhang, S. Lee, H. Yan, H.J. Choi, P. Yang, Single-crystal
ergies and shapes of the frontier orbitals for the probes or probe- gallium nitride nanotubes, Nature 422 (2003) 599e602.
metal ion systems. It is likely that the observed sensitivity and [22] R.M. Waterstrat, R.W. Longton, Gallium-palladium alloys as dental filling
material, Public Health Rep. 79 (1964) 638e642.
selectivity of probes 2 and 3 for Ga3þ or Al3þ sensing were due to [23] H. Tavallal, P. Vahdati, E. Shaabanpur, Developing a new method of 4-(2-
suppression of PET and an increase in aggregate sizes upon pyridylazo)-resorcinol immobilization on triacetylcellulose membrane for
complexation with the Ga3þ or Al3þ. The binding and sensing for selective determination of Ga3þ in water samples, Sens. Actuator B Chem. 159
(2011) 154e158.
Ga3þ or Al3þ were successfully achieved through the judicious use [24] C. Exley, L.M. Charles, L. Barr, C. Martin, A. Polwart, P.D. Darbre, Aluminium in
of the unique binding unit (BHE) and the electron-rich fluorophore human breast tissue, J. Inorg. Biochem. 101 (2007) 1344e1346.
(pyrene). [25] E.B. Lindblad, Aluminium compounds for use in vaccines, Immunol. Cell Biol.
82 (2004) 497e505.
[26] A. Lione, Aluminum intake from non-prescription drugs and sucralfate, Gen.
Acknowledgement Pharmacol. Vasc. Syst. 16 (1985) 223e228.
[27] S. Verstraeten, L. Aimo, P. Oteiza, Aluminium and lead: molecular mechanisms
of brain toxicity, Arch. Toxicol. 82 (2008) 789e802.
This work was supported by the National Research Foundation [28] V. Hornung, F. Bauernfeind, A. Halle, E.O. Samstad, H. Kono, K.L. Rock,
of Korea (NRF) funded by the Korean government (MSIP) (grant K.A. Fitzgerald, E. Latz, Silica crystals and aluminum salts activate the NALP3
inflammasome through phagosomal destabilization, Nat. Immunol. 9 (2008)
number 2008-0061891 to P.S.C. and A.K.). We are very thankful to
847e856.
Prof. Youngbok Lee (Hanyang university, Ansan) for providing the [29] N.R. Chereddy, P. Nagaraju, M.V.N. Raju, V.R. Krishnaswamy, P.S. Korrapati,
1
He1H COSY (2D-NMR) data of probe 3. P.R. Bangal, V.J. Rao, A novel FRET ‘offeon’ fluorescent probe for the selective
detection of Fe3þ, Al3þ and Cr3þ ions: its ultrafast energy transfer kinetics and
application in live cell imaging, Biosens. Bioelectron. 68 (2015) 749e756.
Appendix A. Supplementary data [30] S. Chemate, N. Sekar, Highly sensitive and selective chemosensors for Cu2þ
and Al3þ based on photoinduced electron transfer (PET) mechanism, RSC Adv.
5 (2015) 27282e27289.
Supplementary data related to this article can be found at http:// [31] W. Cao, X.-J. Zheng, J.-P. Sun, W.-T. Wong, D.-C. Fang, J.-X. Zhang, L.-P. Jin,
dx.doi.org/10.1016/j.aca.2016.12.032. A highly selective chemosensor for Al(III) and Zn(II) and its coordination with
metal ions, Inorg. Chem. 53 (2014) 3012e3021.
[32] A. Ghorai, J. Mondal, S. Chowdhury, G.K. Patra, Solvent-dependent
References fluorescent-colorimetric probe for dual monitoring of Al3þ and Cu2þ in
aqueous solution: an application to bio-imaging, Dalton Trans. 45 (2016)
[1] I. Bertini, H.B. Gray, E.I. Stiefel, J.S. Valentine (Eds.), Biological Inorganic 11540e11553.
Chemistry, first ed., University Science Books, Sausalito, CA, 2007. [33] S. Chemate, N. Sekar, A new rhodamine based OFFeON fluorescent chemo-
[2] H.N. Kim, W.X. Ren, J.S. Kim, J. Yoon, Fluorescent and colorimetric sensors for sensors for selective detection of Hg2þ and Al3þ in aqueous media, Sens.
detection of lead, cadmium, and mercury ions, Chem. Soc. Rev. 41 (2012) Actuator B Chem. 220 (2015) 1196e1204.
3210e3244. [34] A. Kumar, P.S. Chae, New 1,8-naphthalimide-conjugated sulfonamide probes
[3] D.R. Williams, Metals, ligands, and cancer, Chem. Rev. 72 (1972) 203e213. for TNP sensing in water, Sens. Actuator B Chem. 240 (2017) 1e9.
[4] S. Bolognin, L. Messori, P. Zatta, Metal ion physiopathology in neurodegen- [35] X.-D. Wang, O.S. Wolfbeis, Optical methods for sensing and imaging oxygen:
erative disorders, Neuro Mol. Med. 11 (2009) 223e238. materials, spectroscopies and applications, Chem. Soc. Rev. 43 (2014)
[5] Y.W. Chen, C.Y. Yang, C.F. Huang, D.Z. Hung, Y.M. Leung, S.H. Liu, Heavy 3666e3761.
metals, islet function and diabetes development, Islets 1 (2009) 169e176. [36] J.F. Zhang, Y. Zhou, J. Yoon, J.S. Kim, Recent progress in fluorescent and
[6] A. Gupte, R.J. Mumper, Elevated copper and oxidative stress in cancer cells as a colorimetric chemosensors for detection of precious metal ions (silver, gold
target for cancer treatment, Cancer Treat. Rev. 35 (2009) 32e46. and platinum ions), Chem. Soc. Rev. 40 (2011) 3416e3429.
[7] B. Jones, Medical oncology. Cardiovascular events linked to levels of cortisol [37] Y. Chen, Y. Bai, Z. Han, W. He, Z. Guo, Photoluminescence imaging of Zn2þ in
production, Nat. Rev. Endocrinol. 10 (2014) 251. living systems, Chem. Soc. Rev. 44 (2015) 4517e4546.
[8] C. Hogstrand, P. Kille, R.I. Nicholson, K.M. Taylor, Zinc transporters and cancer: [38] P. Rivera-Fuentes, S.J. Lippard, Metal-based optical probes for live cell imaging
a potential role for ZIP7 as a hub for tyrosine kinase activation, Trends Mol. of nitroxyl (HNO), Acc. Chem. Res. 48 (2015) 2927e2934.
Med. 15 (2009) 101e111. [39] R. Guliyev, S. Ozturk, Z. Kostereli, E.U. Akkaya, From virtual to physical:
[9] X. Qian, Z. Xu, Fluorescence imaging of metal ions implicated in diseases, integration of chemical logic gates, Angew. Chem. Int. Ed. 50 (2011)
Chem. Soc. Rev. 44 (2015) 4487e4493. 9826e9831.
[10] D.J. Hare, E.J. New, M.D. de Jonge, G. McColl, Imaging metals in biology: [40] Z. Kowser, H. Tomiyasu, X. Jiang, U. Rayhan, C. Redshaw, T. Yamato, Solvent
balancing sensitivity, selectivity and spatial resolution, Chem. Soc. Rev. 44 effect and fluorescence response of the 7-tert-butylpyrene-dipicolylamine
(2015) 5941e5958. linkage for the selective and sensitive response toward Zn(II) and Cd(II) ions,
[11] S. Singha, D. Kim, H. Seo, S.W. Cho, K.H. Ahn, Fluorescence sensing systems for New J. Chem. 39 (2015) 4055e4062.
gold and silver species, Chem. Soc. Rev. 44 (2015) 4367e4399. [41] S. Kim, N.H. Lee, S.H. Seo, M.S. Eom, S. Ahn, M.S. Han, Selective colorimetric
[12] K.P. Carter, A.M. Young, A.E. Palmer, Fluorescent sensors for measuring metal sensor for Hg2þ ions using a mixture of thiourea derivatives and gold nano-
ions in living systems, Chem. Rev. 114 (2014) 4564e4601. particles stabilized with adenosine triphosphate, Chem. Asian J. 5 (2010)
[13] J. Yin, Y. Hua, J. Yoon, Fluorescent probes and bioimaging: alkali metals, 2463e2466.
alkaline earth metals and pH, Chem. Soc. Rev. 44 (2015) 4619e4644. [42] Y.-K. Yang, K.-J. Yook, J. Tae, A rhodamine-based fluorescent and colorimetric
[14] G.R.C. Hamilton, S.K. Sahoo, S. Kamil, N. Singh, N. Kaur, B.W. Hyland, J.F. Callan, chemodosimeter for the rapid detection of Hg2þ ions in aqueous media, J. Am.
Optical probes for the detection of protons, and alkali and alkaline earth metal Chem. Soc. 127 (2005) 16760e16761.
cations, Chem. Soc. Rev. 44 (2015) 4415e4432. [43] A. Kumar, V. Vanita, A. Walia, S. Kumar, N,N-dimethylaminoethylaminoan-
[15] H. Zhu, J. Fan, B. Wang, X. Peng, Fluorescent, MRI, and colorimetric chemical throne e a chromofluorogenic chemosensor for estimation of Cu2þ in aqueous
sensors for the first-row d-block metal ions, Chem. Soc. Rev. 44 (2015) medium and HeLa cells imaging, Sens. Actuator B Chem. 177 (2013) 904e912.
4337e4366. [44] P. Singh, K. Kumar, G. Bhargava, S. Kumar, Self-assembled nanorods of bay
[16] K. Kaur, R. Saini, A. Kumar, V. Luxami, N. Kaur, P. Singh, S. Kumar, Chemo- functionalized perylenediimide: Cu2þ based ‘turn-on’ response for INH,
dosimeters: an approach for detection and estimation of biologically and complementary NOR/OR and TRANSFER logic functions and fluo-
medically relevant metal ions, anions and thiols, Coord. Chem. Rev. 256 rosolvatochromism, J. Mater. Chem. C 4 (2016) 2488e2497.
(2012) 1992e2028. [45] H. Lan, Y. Wen, Y. Shi, K. Liu, Y. Mao, T. Yi, Fluorescence turn-on detection of
[17] S.T. Manjare, Y. Kim, D.G. Churchill, Selenium- and tellurium-containing Sn2þ in live eukaryotic and prokaryotic cells, Analyst 139 (2014) 5223e5229.
fluorescent molecular probes for the detection of biologically important [46] S.V. Bhosale, S.V. Bhosale, M.B. Kalyankar, S.J. Langford, A core-substituted
analytes, Acc. Chem. Res. 47 (2014) 2985e2998. naphthalene diimide fluoride sensor, Org. Lett. 11 (2009) 5418e5421.
[18] J. Yoon, S. Jo, I.S. Chun, I. Jung, H.S. Kim, M. Meitl, E. Menard, X. Li, J.J. Coleman, [47] S.R. Bobe, R.S. Bhosale, S.P. Goskulwad, A.L. Puyad, S.V. Bhosale, S.V. Bhosale,
U. Paik, J.A. Rogers, GaAs photovoltaics and optoelectronics using releasable A highly selective colorimetric Cys sensor based on core-substituted naph-
multilayer epitaxial assemblies, Nature 465 (2010) 329e333. thalene diimides, RSC Adv. 5 (2015) 100697e100701.
[19] H.M. Kim, Y.H. Cho, H. Lee, S.I. Kim, S.R. Ryu, D.Y. Kim, T.W. Kang, K.S. Chung, [48] N.V. Ghule, R.S. Bhosale, A.L. Puyad, S.V. Bhosale, S.V. Bhosale,
50 A. Kumar, P.S. Chae / Analytica Chimica Acta 958 (2017) 38e50

Naphthalenediimide amphiphile based colorimetric probe for recognition of [61] L.N. Neupane, E.T. Oh, H.J. Park, K.H. Lee, Selective and sensitive detection of
Cu2þ and Fe3þ ions, Sens. Actuator B Chem. 227 (2016) 17e23. heavy metal ions in 100% aqueous solution and cells with a fluorescence
[49] A. Kumar, A. Pandith, H.-S. Kim, Pyrene-appended imidazolium probe for chemosensor based on peptide using aggregation-induced emission, Anal.
2,4,6-trinitrophenol in water, Sens. Actuator B Chem. 231 (2016) 293e301. Chem. 88 (2016) 3333e3340.
[50] A.D. Becke, Density-functional thermochemistry.III. The role of exact ex- [62] X. Wang, J. Hu, G. Zhang, S. Liu, Highly selective fluorogenic multianalyte
change, J. Chem. Phys. 98 (1993) 5648e5652. biosensors constructed via enzyme-catalyzed coupling and aggregation
[51] C. Lee, W. Yang, R.G. Parr, Development of the Colle-Salvetti correlation-en- induced emission, J. Am. Chem. Soc. 136 (2014) 9890e9893.
ergy formula into a functional of the electron density, Phys. Rev. B Condens [63] S. Chen, Y. Hong, Y. Liu, J. Liu, C.W.T. Leung, M. Li, Full-range intracellular pH
Matter 37 (1988) 785e789. sensing by an aggregation-induced emission-active two-channel ratiometric
[52] M.W. Schmidt, K.K. Balbridge, J.A. Boatz, S.T. Elbert, M.S. Gordon, J.H. Jensen, fluorogen, J. Am. Chem. Soc. 135 (2013) 4926e4929.
S. Koseki, N. Matsunaga, K.A. Nguyen, S. Su, T.L. Windus, M. Dupuis, [64] W. Zhang, W. Liu, P. Li, F. Huang, H. Wang, B. Tang, Rapid-response fluorescent
J.A.J. Montgomery, General atomic and molecular electronic structure system, probe for hydrogen peroxide in living cells based on increased polarity of C-B
J. Comput. Chem. 14 (1993) 1347e1363. bonds, Anal. Chem. 87 (2015) 9825e9828.
[53] R.G. Huber, M.A. Margreiter, J.E. Fuchs, S.V. Grafenstein, C.S. Tautermann, [65] J.Y. Noh, S. Kim, I.H. Hwang, G.Y. Lee, J. Kang, S.H. Kim, J. Min, S. Park, C. Kim,
K.R. Liedl, Heteroaromatic p-stacking energy landscapes, J. Chem. Inf. Model. J. Kim, Solvent-dependent selective fluorescence assay of aluminum and
54 (2014) 1371e1379. gallium ions using julolidine-based probe, Dye Pig 99 (2013) 1016e1021.
[54] S.K.M. Nalluri, C. Berdugo, N. Javid, P.W.J.M. Frederix, R.V. Ulijn, Biocatalytic [66] H. Kim, K.B. Kim, E.J. Song, I.H. Hwang, J.Y. Noh, P.-G. Kim, K.-D. Jeong, C. Kim,
self-assembly of supramolecular charge-transfer nanostructures based on n- Turn-on selective fluorescent probe for trivalent cations, Inorg. Chem. Com-
type semiconductor-appended peptides, Angew. Chem. Int. Ed. 53 (2014) mun. 36 (2013) 72e76.
5882e5887. [67] H. Wang, Y. Wang, J. Guo, Y. Su, C. Sun, J. Zhao, H. Luo, X. Daia, G. Zou, A new
[55] S. Basak, N. Nandi, K. Bhattacharyya, A. Datta, A. Banerjee, Fluorescence from chemosensor for Ga3þ detection by fluorescent nitrogen-doped graphitic
an H-aggregated naphthalenediimide based peptide: photophysical and carbon dots, RSC Adv. 5 (2015) 13036e13041.
computational investigation of this rare phenomenon, Phys. Chem. Chem. [68] J. Kimura, H. Yamada, H. Ogura, T. Yajima, T. Fukushima, Development of a
Phys. 17 (2015) 30398e30403. fluorescent chelating ligand for gallium ion having a quinazoline structure
[56] J.B. Birks, Photophysics of Aromatic Molecules, Wiley, London, 1970. with two Schiff base moieties, Anal. Chim. Acta 635 (2009) 207e213.
[57] S.W. Thomas III, G.D. Joly, T.M. Swager, Chemical sensors based on amplifying [69] B.-Y. Kim, H.-S. Kim, A. Helal, A fluorescent chemosensor for sequential
fluorescent conjugated polymers, Chem. Rev. 107 (2007) 1339e1386. recognition of gallium and hydrogen sulfate ions based on a new phenyl-
[58] M.A. Kobaisi, S.V. Bhosale, K. Latham, A.M. Raynor, S.V. Bhosale, Functional thiazole derivative, Sens. Actuator B Chem. 206 (2015) 430e434.
naphthalene diimides: synthesis, properties, and applications, Chem. Rev. 116 [70] A. Kumar, J.Y. Lee, H.-S. Kim, Sensitive and selective fluorescence OFF-ON-OFF
(2016) 11685e11796. sensor for cascade detection of Ga3þ cation and I anion based on
[59] R.T.K. Kwok, C.W.T. Leung, J.W.Y. Lam, B.Z. Tang, Biosensing by luminogens pyrenesulfonamide-functionalized inorganic/organic hybrid nanoparticles,
with aggregation induced emission characteristics, Chem. Soc. Rev. 44 (2015) Sens. Actuator B Chem. 239 (2017) 85e93.
4228e4238. [71] M.L. Ramos, A.R.E. de Sousa, L.L.G. Justino, S.M. Fonseca, C.F.G.C. Geraldes,
[60] J. Mei, N.L.C. Leung, R.T.K. Kwok, J.W.Y. Lam, B.Z. Tang, Aggregation-induced H.D. Burrows, Structural and photophysical studies on gallium(III) 8-
emission: together we shine, united we soar, Chem. Rev. 115 (2015) hydroxyquinoline-5-sulfonates. Does excited state decay involve ligand
11718e11940. photolabilisation? Dalton Trans. 42 (2013) 3682e3694.