Design and Development of Fluorescent Sensors with Mixed Aromatic Bicyclic Fused Rings and Pyridyl Groups: Solid Mediated Selective Detection of 2,4,6-Trinitrophenol in Water.

For a strategic incorporation of both π-electron-rich moieties and Lewis basic moieties acting as hydrogen bonding recognition sites in the same molecule, two new fluorescent sensors, N,N'-bis(anthracen-9-ylmethyl)-N,N'-bis(pyridin-2-ylmethyl)butane-1,4-diamine (banthbpbn, 1) and N,N'-bis(naphthalen-1-ylmethyl)-N,N'-bis(pyridin-2-ylmethyl)butane-1,4-diamine (bnaphbpbn, 2), have been developed for the selective detection of highly explosive 2,4,6-trinitrophenol (TNP) in water. Each of the two identical ends of these sensors that are linked with a flexible tetra-methylene spacer contains a mixed aromatic bicyclic fused ring (anthracene or naphthalene) and a pyridyl group. These are synthesized via the simple reduced Schiff base chemistry, followed by the nucleophilic substitution reaction under basic conditions in high yields. Both 1 and 2 were characterized by Fourier transform infrared, UV-vis, and NMR (1H and 13C) spectroscopy, and high-resolution mass spectrometry. The bulk phase purity of 1 and 2 and their stability in water were confirmed by powder X-ray diffraction (PXRD). Utilizing the effect of solvents on their emission spectra as determined by fluorescence spectroscopy, spectral responses for 1 and 2 toward various nitro explosives were recorded to determine a detection limit of 0.6 and 1.6 ppm, respectively, for TNP in water via the "turn-off" quenching response. Also, the detailed mechanistic investigation for their mode of action through spectral overlap, lifetime measurements, Stern-Volmer plots, and density functional theory calculations reveals that resonance energy transfer and photoinduced electron transfer processes, and electrostatic interactions are the key aspects for the turn-off response toward TNP by 1 and 2. In addition, the selectivity for TNP has been found to be more in 1 compared to 2. Both exhibit good recyclability and stability after sensing experiments, which is confirmed by PXRD and field-emission scanning electron microscopy.

Introduction

Design and development of fluorescent sensors for the highly selective and sensitive detection of hazardous nitro explosives (NEs) are one of the essential concerns from the viewpoint of national/international security and safety because of unpredicted terrorist attacks and industrial disasters.[1,2] These NEs are also used in fireworks, dyes, and landmines, posing threat to the humankind and environment owing to their detonation properties.[3] Among the various NEs, such as 2,4,6-trinitrophenol (TNP), 2,4,6-trinitrotoluene (TNT), 2,6-dinitrotoluene (2,6-DNT), 1,3,5-trinitro-1,3,5-triazacyclohexane, and nitrobenzene (NB), TNP being highly acidic and soluble in water gets easily accumulated in the soil and agricultural land, causing contamination and health hazards.[4] Thus, highly selective and prompt detection of TNP in aqueous medium is significantly important. Diverse analytical techniques such as electrochemical, mass spectrometry, Raman spectroscopy, and fluorescence spectroscopy have been utilized for the detection of NEs, but the detection based on fluorescence emission is the most advantageous one because of its high sensitivity, quick response time, and real-time monitoring.[5]

Till date, many fluorescent materials such as metal–organic frameworks,[6] covalent organic frameworks,[7] conjugated polymers,[8] carbon dots,[9] and cages[10] have been developed for the detection of TNP. However, their preparation requires high temperature conditions and multistep synthesis. Thus, there is still demand for the design of sensors with easy synthetic methods. Although many chemosensors have been developed for sensing of TNP in dimethylformamide (DMF), dichloromethane (CH2Cl2), and other organic solvents,[6a−6f,7,10−12] for practical purposes, selective sensing of TNP in water is desired.[6g−6m,8b,13] However, in many cases, it is not suitable because of aggregation-caused quenching (ACQ) in the presence of water and also because of the tendency of water to form hydrogen bonding with TNP.[13a] As a consequence, for developing sensors devoid of such limitations, it is still a challenging task. The essential criteria for designing suitable sensors for the detection of electron-deficient NEs are as follows: (i) π-electron-rich cloud in the molecule which can provide effective π–π interactions with the electron-poor NEs, (ii) the presence of Lewis basic sites that can further form noncovalent interactions like H-bonding with the nitrophenol derivatives, and (iii) the presence of some bulky substituents at the periphery that can prevent any self-aggregation which can lead to the formation of intermolecular excimers (causing ACQ).[14] Herein, considering these criteria, we have designed and synthesized two organic fluorescent sensors, N,N′-bis(anthracen-9-ylmethyl)-N,N′-bis(pyridin-2-ylmethyl) butane-1,4-diamine (banthbpbn, 1) and N,N′-bis(naphthalen-1-ylmethyl)-N,N′-bis(pyridin-2-ylmethyl)butane-1,4-diamine (bnaphbpbn, 2), with two different hydrophobic pendant moieties, namely, aromatic bicyclic fused rings (anthracene or naphthalene, respectively) and Lewis basic pyridyl groups. In our design, the fluorescent tags (anthracene or naphthalene) provide effective π–π stacking with the NEs while the pyridyl groups play dual roles: (i) act as the Lewis basic recognition sites for the acidic NEs via the hydrogen bonding interactions (H-bonding) and (ii) act as the bulky group to restrict the generation of any self-aggregation (see Figure ). Thus, we have envisaged that an interplay of the abovementioned criteria can be made possible with the strategic incorporation of both the fluorescent tags and the Lewis basic sites in these sensors. In fact, the fluorescent sensors 1 and 2 have been found to show solid-mediated selective detection of TNP in water through a turn-off response with a detection limit of 0.6 and 1.6 ppm, respectively.

Figure 1

Structures of 1 and 2 with an illustration of their different interaction sites.

Results and Discussion

Synthesis and Characterization

For the synthesis of 1 and 2, a general two-step method has been developed with excellent yields of 85 and 92%, respectively, as illustrated in Scheme . In the first step, a Schiff base reaction between anthracene or naphthalene carboxaldehyde and 1,4-diaminobutane (in a 2:1 ratio) followed by the reduction of the respective imine produces the intermediate Hbanthbn or Hbnaphbn, respectively. The second step involves the introduction of a methylpyridyl group at each alkyl nitrogen atom via the nucleophilic substitution under strongly basic conditions. Both 1 and 2 were extensively characterized by 1H and 13C NMR, Fourier transform infrared (FTIR), and UV–vis spectroscopy and high-resolution mass spectrometry (HRMS) (see Figures S1–S13). In the 1H NMR spectra, the presence of two singlets at 3.68 and 4.48 ppm in 1 and 3.75 and 4.04 ppm in 2 are attributed to the CH2 groups attached to 2-picolyl and anthracene or naphthalene, respectively. As expected, in their FTIR spectra, there were no peak at around 1700 and 3200 cm–1 for CHO and NH groups, confirming the formation of pure products. The HRMS data showing the presence of the parent ion peaks of 1 and 2 further confirmed their purity. Their phase purity and crystallinity of the bulk were confirmed by powder X-ray diffraction (PXRD).

Scheme 1

Synthetic Procedures of 1 and 2

Solvent Effect on Emission Spectra

Both 1 and 2 were soluble in DMF and chloroform (CHCl3), partially soluble in dimethylsulfoxide, and insoluble in water (H2O), methanol (CH3OH), and NB. Having high solubility in DMF and CHCl3, the emission intensities of 1 and 2 were fully quenched because of ACQ. In the case of NB, the presence of an electron-deficient nitro group led to the turn-off response via electron transfer from electron-rich sensors 1 and 2. On the other hand, the maximum emission intensity of 1 and 2 was observed in H2O followed by CH3OH and CH3CN, attributing to aggregation-induced enhanced emission (see Figure S14). To get better insight into the effect of solvents on their emission spectra, solid-state emission spectra of 1 and 2 were recorded upon excitation at the same wavelength of the aqueous suspension. A broad peak at 467 nm for 1 and a very broad peak at 427 nm for 2 are observed, with significantly lower intensity for 2 (see Figure S15). Comparing their emission peaks in aqueous solution, it is clear that an excimer formation takes place in 1 because of self-aggregation, which is much more pronounced in 2 because of less bulky naphthalene moiety. These observations confirm that H2O, CH3OH, and CH3CN solvents play an important role to prevent quenching due to self-aggregation and enhance the emission intensities of 1 and 2 in solution.

Emission Spectral Response toward NEs

The high luminescent property of 1 and 2 due to the presence of both π-conjugated cores and Lewis basic pyridyl sites has been utilized for sensing different electron-deficient NEs such as TNP, 2,4-dinitrophenol (2,4-DNP), 4-nitrophenol (4-NP), 2,6-DNT, TNT, 1,3-dinitrobenzene (1,3-DNB), and NB. Inspired from the aforementioned results of the solvent effect on emission spectra, H2O has been selected as the solvent medium. To investigate the sensing ability of 1 and 2 toward various NEs, their dispersed solutions in water (1 mg in 2 mL H2O) were prepared to facilitate close contact between the sensors and the NEs. The fluorescence intensities were monitored at 420 and 450 nm for 1 and 2, respectively, upon the incremental addition of aqueous solutions (0–200 μL of a 1 mM stock solution) of different NEs (see Figures and S16–S27). As anticipated, fluorescence spectra of both sensors upon the addition of TNP displayed turn-off responses. Also, the fluorescence quenching of 1 is higher than that of 2. This can be correlated to the decrease in the π-conjugation from anthracene to naphthalene, which may lead to less π–π interactions between the aromatic rings of the sensor and those of the NEs in 2 compared to 1.

Figure 2

Change in emission spectra of (a) 1 and (b) 2 dispersed in water upon the incremental addition of TNP solution (1 mM) in water.

The quenching efficiency percentage was calculated using the formula (I0 – I)/I0 × 100%, where I0 and I are the luminescent intensities of the sensors before and after the addition of the nitroanalytes, respectively (see Figure ). The quenching efficiency for 1 and 2 with 200 μL TNP is 98 and 90%, respectively. Also, 1 shows more selectivity toward TNP among all NEs with respect to quenching % compared to 2. Moreover, the detection limits of TNP calculated using the formula 3σ/m were found to be 0.6 and 1.6 ppm for 1 and 2, respectively (see Figures S28,S29).

Figure 3

Comparison of quenching % of different NEs in water for 1 and 2.

Mechanistic Studies

To gain better insight into the quenching mechanism for TNP, Stern–Volmer (SV) quenching constants were calculated using the SV equation I0/I = 1 + KSV[A], where, I0 and I are fluorescence intensities before and after the addition of NEs, respectively, [A] is the molar concentration of the nitroanalyte, and KSV is the quenching constant (M–1), as depicted in Figure . The KSV values were found to be 5.78 × 104 and 3.26 × 104 for 1 and 2, respectively (see Figures S30,S31). As evident from these values, the interaction with TNP is much stronger in 1 compared to 2. These KSV values are comparable or better than those reported in the literature (see Table S1). The Ksv values of 1 and 2 are greater than those of sensors containing fluorescein (3.94 × 104 and 2.10 × 104 M–1)[13a] and naphthalene (2.69 × 104 and 3.43 × 103 M–1),[10a,13b] whereas they are comparable to those of anthracene-tagged sensors (8.08 × 104 and 7 × 104 M–1).[10a,13c] Also, the detection limit is greater than those with pyrene (2.98 ppm)[12]- and naphthalene (3.48 ppm)[13b]-based sensors, respectively. The SV plots for TNP in both cases exhibit linearity at lower concentrations and an upward bent curve at higher concentrations. The nonlinearity observed can be attributed either to self-absorption or resonance energy transfer processes.[6e,6g,15] To further validate the mechanism relevant for this turn-off quenching, the spectral overlap between the absorption spectra of various NEs and emission spectra of the sensors was examined. It was observed that the extent of overlap is more between the emission spectrum of 1 and 2 with the absorption band of TNP compared to other NEs, as shown in Figure . It is noteworthy to observe a larger extent of overlap for 1 than 2. This was further proved by calculating the overlap integral J(λ) values by using the equation given belowwhere FD(λ) = corrected fluorescence intensity of the donor in the range of λ to λ + Δλ with the total intensity normalized to unity and εA = extinction coefficient of the acceptor at λ in M–1 cm–1.[6e,16] For the TNP analyte, the overlap integral values were calculated (based on the area from an overlap between the start of an emission curve of the sensor and the end of the absorption curve of TNP) to be 2.30 × 1014 (1) and 1.24 × 1014 (2). This confirms that more resonance energy transfer takes place in 1 compared to 2. This is also reflected in the SV plot where a stronger interaction is observed for 1.

Figure 4

3D representation of SV plots for different NEs in (a) 1 and (b) 2. Insets: Cuvette images of 1 and 2 before and after the addition of 200 μL TNP upon UV light illumination (λ = 365 nm).

Figure 5

Spectral overlap between absorbance spectra of NEs and emission spectra of 1 and 2.

Furthermore, the effective quenching response was more prompted in case of nitrophenol derivatives, which is in accordance with the acidity of the phenolic protons: TNP > 2,4-DNP > 4-NP. A decrease in the quenching efficiency with the reduction in the acidity can also be ascribed to the presence of Lewis basic pyridyl moieties where the pyridyl nitrogen can form H-bonding interactions with the phenolic protons of the nitrophenol derivatives, which are absent in the other NE analytes.[13b] A probable mechanism for the selective sensing of TNP from these observations has been illustrated in Scheme , where an energy transfer via π–π interactions between TNP and anthracene or naphthalene moieties and the abovementioned H-bonding with the more acidic phenolic proton of TNP play important roles.

Scheme 2

Probable Mechanism for TNP Sensing by 1 and 2

In addition to these, highest occupied molecular orbital–lowest unoccupied molecular orbital (HOMO–LUMO) energies of the sensors and NEs were calculated using density functional theory with the basis set B3LYP/6-31G+(dp) and the Gaussian 09 package program.[17] The energy-optimized structures obtained for 1 and 2 are shown in Figure . The length of the methylene chain is 6.028 and 6.361 Å in 1 and 2, respectively, which shows that because of more rigid and π-conjugated anthracene moieties compared to naphthalene moieties, the chain gets squeezed to stabilize the structure. On the other hand, the angles between the anthracene moiety and the pyridyl group in 1 and the naphthalene moiety and the pyridyl group in 2 (see Figure S32) are determined. On the basis of this analysis, it is clear that the angle on one side is very similar (51.83° and 52.72°, respectively) while the angle on the other side is much different (71.99° and 88.54°, respectively) in 1 and 2. With this difference in angle between the groups, there appears to be an impact of electronic and steric effects on the interaction between 1 or 2 and TNP (and other analytes). With a wider angle in 2, it can be assumed that the access to both the naphthalene moiety and the pyridyl group by the nitroanalytes is more than the corresponding situation in 1. This affected the quenching % of 1 and 2 by the analytes. In general, the LUMO of the nitroanalytes lie in between the conduction and valence band of the sensor, resulting in photoinduced electron transfer (PET) to the electron-deficient analytes. From these HOMO–LUMO energy gaps (see Figure S33 and Table S2), it is inferred that PET takes place from the HOMO of the sensor to the LUMO of electron-deficient TNP compared to other NEs. This corroborates well with the observed maximum quenching efficiency for TNP. However, when the quenching efficiency of other nitroanalytes was compared with the corresponding spectral overlap integral values (wherever possible) and LUMO energies, a similar trend was not observed. This suggests that both PET and resonance energy transfer processes interplay together for the observed quenching response.

Figure 6

Energy-optimized structures of (a) 1 and (b) 2; (color code: carbon, gray, but the carbon atoms of anthracene and naphthalene are shown in orange and green colors, respectively for clarity; nitrogen, blue; hydrogen, purple).

Also, the selectivity of TNP in the presence of other nitroanalyte congeners was studied to analyze the competing effect of NEs. In such an experiment, the fluorescence intensity of sensors (1 mg dispersed in 2 mL of water) was recorded, followed by the 100 μL addition of a particular nitroanalyte and then the 100 μL addition of TNP (1 mM, water). This result vividly depicts insignificant change in the intensity for other NEs while tremendous decrease in intensity for TNP in the case of 1 (see Figure a), whereas for 2, in the presence of 2,4-DNP and 4-NP, significant decrease was observed followed by moderate change after the TNP addition. Thus, selectivity for TNP is more in 1 than in 2 (see Figure a,b).

Figure 7

Bar diagrams of the fluorescence intensity (%) of (a) 1 and (b) 2 before and after the addition of nitroanalytes (100 μL, 1 mM) separately followed by the addition of the same amount of TNP. Purple bars represent the suspension of compounds in water, red bars represent the addition of various nitroanalytes to the suspension of the compound (compound + nitro-analytes), and green bars represent the subsequent addition of TNP (100 μL, 1 mM) to the abovementioned suspensions (compound + nitro-analytes + TNP); from left to right (i) 2,4-DNP, (ii) 4-NP, (iii) TNT, (iv) 2,6-DNT, (v) 1,3-DNB, and (vi) NB.

Both 1 and 2 were recovered for the recyclability test by centrifugation and washed several times with water after quenching experiments. These were then utilized for quenching experiments of TNP for up to five recycles to exhibit their quenching ability (see Figure S34). To study the hydrolytic and chemical stability of 1 and 2, 12 mg of each sample was soaked in 2 mL of water or aqueous solution of TNP (1 mM) for 3 days, filtered, and washed thoroughly with water for PXRD studies. Comparing the PXRD patterns of the pristine and treated samples of 1 and 2, it is confirmed that their crystallinity is intact because of such exposure (see Figure a,b). Furthermore, an analysis of surface morphologies of both 1 and 2 by field-emission scanning electron microscopy (FESEM) indicates that there is no change between before and after immersing in a 1 mM aqueous solution of TNP (see Figure c). Thus, 1 and 2 can be used for the long term in-field detection of highly explosive TNP.

Figure 8

PXRD profiles of 1 (a) and 2 (b) before and after immersing in water and aqueous TNP solution. (c) FESEM images of 1 (i,ii) and 2 (iii,iv) before and after soaking in 1 mM aqueous TNP solution, respectively.

Lifetime Measurements

On the basis of the abovementioned observation of deviation from linearity in the SV plot, which suggests that there is an amalgamation of both static and dynamic quenching, time-resolved lifetime measurements were recorded for 1 (Figure a) and 2 (Figure b) for further justification. In general, if the lifetime of the sensor remains unchanged with the change in the concentration of the analyte, the process is termed as static quenching.[16] From the lifetime decay curves, the average lifetime for both the sensors was calculated before and after the addition of TNP (Table S3). For 1, average lifetimes were 10.36 ns (before the TNP addition), 10.38 ns (after the 20 μL TNP addition), 10.23 ns (after the 60 μL TNP addition), and 10.11 ns (after the 200 μL TNP addition). On the other hand, for 2, average lifetimes were found to be 6.97 ns (before the TNP addition), 6.75 ns (after the 20 μL TNP addition), 6.50 ns (after the 60 μL TNP addition), and 6.42 ns (after the 200 μL TNP addition). This corroborates well with the aforementioned results, signifying that resonance energy transfer is more dominant at higher concentrations of the analyte. The quenching rate constant (Kq) values were also calculated to be 5.58 × 1012 and 4.68 × 1012 M–1 s–1, respectively, for 1 and 2 using the following equation: Kq = KSV/τavg, where KSV is the quenching constant and τavg is the average lifetime of the sensor.

Figure 9

Lifetime decay curves of (a) 1 and (b) 2 before and after the TNP addition.

In-Field Detection of TNP

For fast and effective detection of TNP, test strips were prepared using Whatman filter papers immersed in aqueous suspensions of 1 and 2. These coated paper strips showed blue emission under UV light at 365 nm. Also, after adding a small quantity of an aqueous solution of each NE as a spot, these paper strips displayed a distinguishable dark color change for TNP only, as shown in Figure a,b. In the case of 1, more darkened spots were depicted for TNP compared to those for 2. In 2, moderate darkening was also observed for 2,4-DNP and 4-NP, further confirming more selectivity in 1. Furthermore, a negligible change was observed for the rest of the NEs in both cases. In addition to these, different concentrations of TNP were also added (10 μL each, 10–8 to 10–3 M) to 1 and 2 (see Figure c,d), leading to a darkened color change with an increase in the concentration of TNP after illumination with UV light at 365 nm for visual and facile detection of TNP with naked eye.

Figure 10

Photographs of test paper strips coated with (a) 1 and (b) 2 and their respective responses after adding various NEs under UV light (λ = 365 nm). Photographs of paper strips coated with (c) 1 and (d) 2 for visual detection with an increasing concentration of TNP (10–8 to 10–3 M).

Conclusions

In conclusion, two new sensors 1 and 2 based on mixed bicyclic fused rings (anthracene or naphthalene) with pendant Lewis basic pyridyl groups have been designed and synthesized via simple reduced Schiff base condensation reaction. Their bulk phase purity and crystallinity were confirmed by PXRD. These sensors are capable of detecting highly explosive TNP in water with detection limits of 0.6 and 1.6 ppm, respectively. On the basis of the spectral overlap between the absorption spectrum of TNP and the emission spectrum of 1 or 2, resonance energy transfer plays a dominant role in both cases with more overlap integral value in 1, resulting in more selectivity toward TNP compared to 2. Also, the competitive test demonstrated that 1 has more selectivity than 2 in the presence of different commonly interfering NEs. The resonance energy transfer, PET, and electrostatic interactions play key roles for the turn-off quenching response for TNP in 1 and 2. Both were recyclable for up to five recycles without much loss in sensitivity. Their stability after sensing experiments was confirmed by PXRD and FESEM. For an in-field detection, handy test paper strips coated with 1 or 2 also confirmed that these can be used for visual detection of TNP for environmental monitoring. The present results could provide a pathway for the development of different fluorescent sensors with the incorporation of various recognition sites to end up with increased sensing capabilities.

Experimental Section

Materials and Methods

All chemicals for synthesis were purchased commercially and used as received without further purification. All reactions were carried out under aerobic conditions.

Caution! TNP and TNT are highly explosive in nature and should be handled carefully and in small amounts. The explosives should be handled as dilute solutions and with safety measures to avoid an explosion.

Physical Measurements

The 1H and 13C NMR spectra of the 1 and 2 ligands were obtained in CDCl3 solution at 25 °C on a Bruker ARX-400 spectrometer. Chemical shifts are reported relative to the residual solvent signals. FTIR spectra were measured in the 4000–400 cm–1 range on a PerkinElmer spectrum I spectrometer with samples prepared as KBr pellets. The UV–vis spectra were recorded using 1 mM solution on an Agilent Cary 6000 spectrometer. Electrospray ionization (ESI) mass spectrometry was performed using a Thermo Scientific LTQ XL LC–MS instrument for the 50–2000 amu range. PXRD data were recorded on a Rigaku Ultima IV diffractometer equipped with 3 kW sealed-tube Cu Kα X-ray radiation (generator power settings: 40 kV and 40 mA) and a D/tex Ultra detector using the BB geometry over the angle range 5–50° with a scanning speed of 2°/min with 0.02° step. (2.5° primary and secondary solar slits and 0.5° divergence slit with 10 mm height limit slit). The surface morphology of the as-prepared samples was examined using FESEM (JEOL, 8 or 15 kV). Fluorescence spectra were recorded using a HORIBA Jobin-Yvon fluorescence spectrophotometer with stirring attachment. Lifetime measurements of 1 and 2 were carried out using the time-resolved HORIBA scientific single photon counting controller.

Synthesis of N,N′-Bis(anthracen-9-ylmethyl)butane-1,4-diamine (Hbanthbn)

This was prepared with some modification of the literature procedure.[18] To a solution of 9-anthracene carboxaldehyde (206 mg, 2 mmol) in a mixture of CH2Cl2/CH3OH (1:2 v/v, 6 mL) placed in a 10 mL RB flask, a methanolic solution of 1,4-diaminobutane (0.1 mL, 1 mmol in 1 mL methanol) was added dropwise with continuous stirring. The reaction mixture was refluxed at 50 °C for 6 h, leading to the formation of a yellow precipitate. An excess of NaBH4 (3 equiv, 114 mg) was added portion-wise at 0 °C and stirred for 24 h at 30 °C to obtain a clear yellow solution. The solvent was removed under reduced pressure, washed with water, and extracted into a chloroform layer. To the organic layer, anhydrous Na2SO4 was added and filtered, and solvent was evaporated and washed with diethyl ether to get a yellow solid. Yield: 384.3 mg (82%). mp 114–116 °C. 1H NMR (400 MHz, CDCl3): δ 1.68 (t, 4H), 2.91 (t, 2H), 4.72 (s, 4H), 7.45–7.54 (m, 8H), 8.01 (d, 4H), 8.33 (d, 4H), 8.42 (s, 2H) ppm. HRMS (ESI-TOF) m/z: calcd for [M + H]+, 469.2599; found, 469.2578.

Synthesis of N,N′-Bis(anthracen-9-ylmethyl)-N,N′-bis(pyridin-2-ylmethyl)butane-1,4-diamine (banthbpbn, 1)

In a 25 mL round-bottom (RB) flask, 384.3 mg (0.82 mmol) of Hbanthbn was added to 3 mL of water. To this solution, 269 mg (1.64 mmol) of 2-picolyl chloride dissolved in 2 mL of water was added dropwise, followed by the addition of 4 equiv NaOH (132 mg in 2 mL water) to maintain the pH = 12. The resulting mixture was stirred for 24 h at 30 °C. The desired product was extracted in the chloroform layer and obtained as light yellow oily liquid by evaporating the solvent under reduced pressure. The liquid was washed with diethyl ether to form a yellow solid. Yield: 553 mg (85%). mp 177–180 °C. 1H NMR (400 MHz, CDCl3): δ 1.55 (t, 4H), 2.51 (t, 4H), 3.68 (s, 4H), 4.48 (s, 4H), 7.03 (t, 2H), 7.19 (d, 2H), 7.43–7.50 (m, 12H), 7.97 (d, 4H), 8.39 (d, 2H), 8.47 (d, 4H) ppm. 13C NMR (100 MHz, CDCl3): δ 24.9, 51.0, 54.8, 60.1, 121.6, 123.1, 124.8, 125.1 (2C), 125.5 (2C), 125.6 (2C), 127.5 (2C), 128.9 (2C), 130.3 (2C), 131.4, 136.0, 148.4, 160.6 ppm. HRMS (ESI-TOF) m/z: calcd for [M + H]+, 651.3443; found, 651.3421.

Synthesis of N,N′-Bis(naphthalen-1-ylmethyl)butane-1,4-diamine (Hbnaphbn)

This was prepared with a little modification reported in the literature procedure.[19] To a solution of 1-naphthalene carboxaldehyde (0.303 mL, 2.2 mmol) dissolved in 6 mL CH2Cl2 placed in a 25 mL RB flask, 1,4-diaminobutane (0.11 mL, 1.1 mmol in 1 mL methanol) was added dropwise with continuous stirring at 30 °C for 2 h. The solvent was evaporated to obtain yellow oily liquid and suspended in 8 mL CH3OH. To this abovementioned solution, an excess of NaBH4 (3 equiv, 125.4 mg) (acting as the reducing agent) was added portionwise at 0 °C and stirred for 20 h at 30 °C. The solvent was evaporated under reduced pressure and washed with water, and the product was extracted into the chloroform layer. To the organic layer, anhydrous Na2SO4 was added and filtered, and the reduced Schiff base was obtained in the oily form by evaporating the solvent under reduced pressure. Yield: 375.7 mg (90%). 1H NMR (400 MHz, CDCl3): δ 1.63 (t, 4H), 2.77 (t, 4H), 4.24 (s, 4H), 7.42–7.55 (m, 8H), 7.80 (d, 2H), 7.88 (d, 2H), 8.13 (d, 2H) ppm. HRMS (ESI-TOF) m/z: calcd for [M + H]+, 369.2286; found, 369.2270.

Synthesis of N,N′-Bis(naphthalen-1-ylmethyl)-N,N′-bis(pyridin-2-ylmethyl)butane-1,4-diamine (bnaphbpbn, 2)

In a 25 mL RB flask, 375.7 mg (1.02 mmol) of Hbnaphbn was added to 3 mL of water. To this, 334.6 mg (2.04 mmol) of 2-picolyl chloride was added dropwise and dissolved in 2 mL water, followed by the addition of 4 equiv NaOH (164 mg in 2 mL water) to maintain the pH = 12. The resulting mixture was stirred for 24 h at 30 °C. The desired product was extracted in the chloroform layer and obtained as an off-white solid by evaporating the solvent under reduced pressure. Yield: 517 mg (92%). mp 143–145 °C. 1H NMR (400 MHz, CDCl3): δ 1.55 (t, 4H), 2.50 (t, 4H), 3.75 (s, 4H), 4.04 (s, 4H), 7.09 (t, 2H), 7.34–7.38 (m, 4H), 7.39–7.54 (m, 8H), 7.76 (d, 2H), 7.84 (t, 2H), 8.23 (t, 2H), 8.47 (d, 2H) ppm. 13C NMR (100 MHz, CDCl3): δ 24.8, 54.7, 57.5, 60.5, 121.7, 122.9, 124.6, 125.2, 125.5, 127.3, 127.7, 128.4 (2C), 132.3, 133.8, 135.1, 136.2, 148.6, 160.5 ppm. HRMS (ESI-TOF) m/z: calcd for [M + H]+, 551.3110; found, 551.3106.

Fluorescence Study for NEs

One milligram of finely ground samples of 1 and 2 were weighed and added to a fluorescence cuvette (path length of 1 cm) containing 2 mL of Milli-Q water under constant stirring. To the abovementioned dispersed solution of the samples in water, NEs were added incrementally from 1 mM stock solution (stock solutions of those without a phenolic group were prepared in 9.5 mL water and 0.5 mL methanol), and the corresponding photoluminescence spectral responses were monitored.

Preparation of the Test Paper Strip

A test paper strip was prepared by coating the fluorescent samples on a Whatman filter paper by immersing them into an aqueous suspension of the respective samples and drying in air. These strips were used to detect nitroanalytes by dropcasting a small volume of an aqueous solution of each of the nitroanalytes (10 μL, 1 mM in water) onto the dried filter paper strips coated with the samples.