Imidazole-appended 9,10-anthracenedicarboxamide probe for sensing nitrophenols and selective determination of 2,4,6-trinitrophenol in an EtOH–water medium

Abstract

A novel imidazole-appended anthracenedicarboxamide “switch-off” probe, AIM-D, was designed and synthesized for nitrophenol sensing. AIM-D selectively detected 2,4,6-trinitrophenol (TNP) relative to other nitrophenol derivatives through a ratiometric fluorescence response in 1 : 1 EtOH–H₂O and exhibited a large Stern–Volmer quenching constant (K_sv = 2.69 × 10⁷ M⁻¹) and a binding ratio of 1 : 2. Mono- and dinitrophenols also quenched anthracene emission to a moderate (∼37%) or great (∼80%) extent depending on the position(s) of the nitro group(s) on the aromatic ring. The limit of detection of TNP with AIM-D was 1 ppb. The selective ratiometric discrimination of TNP over other nitrophenols was attributed mainly to protonation-induced electron transfer aided synergistic coulombic, multiple hydrogen bonding and π_an–π_TNP interactions in a symmetrical manner. Monomer quenching resulted from resonance energy transfer and a new enhancement at 520 nm occurred as a result of the effective inhibition of intramolecular charge transfer through intermolecular sequential opposite charge flow between AIM-D and TNP. However, mono- and dinitrophenol sensing attributed to the nitro group oriented resonance energy transfer from the latter to the former without inducing protonation at imidazole resulted in only monomer band quenching. An AIM-D coated paper strip was applied to the detection of nitrophenols in sea and river water samples and showed a limit of detection of 20 nM. The interaction of AIM-D with nitrophenols was examined by UV-visible, fluorescence, and ¹H NMR spectrometry and supported by DFT calculations.

---

1. Introduction

Nitrophenol derivatives are an important group of environmental pollutants and are well known for their detonation properties. As a result of their extensive use in the manufacture of insecticides, pesticides, dyes, plastics, and explosives, large quantities of nitrophenols and their derivatives have been released into the environment.¹ The nitro group not only increases the acidic nature of the hydroxyl group (–OH) in phenols, but also enhances detonation. For this reason, trinitrophenol has the highest detonation tendency, followed by mono- and dinitrophenol.² The exothermic reaction of nitrophenol ignition releases toxic gases and the polarity introduced by the nitro and hydroxyl substituents facilitates dissolution in water.³ Most nitrophenol derivatives are potential carcinogens, mutagens, or teratogens, even when present in only micromolar amounts in water, soil, or air. This may cause serious health problems in humans as well as animals.⁴ The monitoring of ultratrace amounts of these species in aqueous media is therefore essential for both pollution control and ecosystem sustainability.⁵

Instrumental techniques routinely used for the determination of nitrophenols include high-performance liquid chromatography, UV-visible and fluorescence spectrometry, capillary zone electrophoresis, optrodes, Raman spectrometry, energy-dispersive X-ray diffraction analysis, neutron activation analysis, ion mobility spectrometry, enzyme assay based biosensing, and electrochemistry.⁶ Among these techniques, fluorimetry is promising because of its simplicity, selectivity, sensitivity, and economy.⁷ Various fluorescent supramolecular assemblies, including polymers, metal–organic frameworks, sol gel aggregates, and nanocomposites, have been used to sense nitrophenols in organic or organic–aqueous mixtures.⁸ The rational design of nitroaromatic chemosensors has primarily used conjugated aromatic systems such as electron donors because the delocalized π* excited state significantly enhances donor ability. Pyrene, quinoline, anthracene, phenanthrene, and pentacene are the most promising candidates because their aromatic rings promote strong interactions with electron-deficient compounds through π–π interactions. Many reported supramolecular architectures have been designed with different fluorophores for sensing nitrophenols. Such architectures typically encounter serious problems in terms of the ease of synthesis, preparation time, solubility adjustment, fluorescence stability over a wide pH range, the sensing medium, and interferences. To avoid these problems, small anthracene-based luminophores attracted our attention due to their ease of synthesis, solubility tuning through functional group exchange, and variations in blue light emission on the binding of an analyte.⁹ Few results have been reported on the fluorescent sensing of nitrophenols and the selective determination of TNP in mixed aqueous solution based on quenching efficiency using small molecules. However, these molecules often suffer interference from protons, cations, and anions as a result of their electron-rich binding sites, which are typically connected to one or two methylene groups or without spacers for a rapid optical response to the interactions. To avoid such interference phenomena and to achieve a high photo-stability over a wide pH range, a new strategy was implemented with two principal binding sites. Specifically, the carboxamide functionality and imidazole unit were separated using a propyl spacer that could provide a rapid response towards specific analytes by creating the perfect space for favourable interactions using the most stable conformer. We recently designed imidazole-appended carboxamide-, sulphonamide-, and methylamine-based probes with different fluorophores to detect neutral molecules (electron-deficient aromatic species) by combined electroanalytical and optical methods.¹⁰ We report here the design and application of the new imidazole-appended 9,10-anthracenedicarboxamide fluorescent probe, AIM-D, to nitrophenol sensing in organic–aqueous media. The probe was investigated by UV-visible, fluorescence, and ¹H NMR spectrometry and further studied by DFT calculations. A real-world application of AIM-D is demonstrated using paper strip based sensing in marine and fresh water samples.

2. Experimental section

2.1 General

Melting points were determined using the OptiMelt (Stanford Research Systems) apparatus and were uncorrected. ¹H and ¹³C NMR spectra were recorded on a Bruker AM-400 spectrometer. Fast atom bombardment mass spectra were recorded on a JEOL JMS 700 high-resolution mass spectrometer at the Korea Basic Science Institute (KBSI) Daegu Branch, Korea. UV-visible absorption spectra were determined on a Shimadzu UV-1650PC spectrophotometer. Fluorescence spectra were measured with a Shimadzu RF-5301 spectrometer equipped with an Xe discharge lamp and 1 cm quartz cells using slit 3. All measurements were taken at 298 K. Analytical-grade ethanol was purchased from Merck. 1-(3-Aminopropyl)imidazole and other materials for synthesis were purchased from Aldrich Chemical and were used as received. Fluorescence quantum yields were determined by the integration of corrected fluorescence spectra. Quinine hemisulfate in 0.5 M H₂SO₄ was used as a standard (Φ = 0.54)¹¹ for the correction of the fluorescence spectra.

Caution! TNP, 2,4-DNP, 3,4-DNP, and other nitroaromatic compounds should be used with extreme care employing the greatest safety precautions as a result of their explosive and carcinogenic properties. They should only be handled in small quantities.

2.2 Calculation of association constants and limit of detection

The Stern–Volmer quenching constants, K_sv, of AIM-D with nitrophenols were calculated from fluorescence titration data using the equation, I₀/I = 1 + K_sv[Q], where I₀ and I are the fluorescence intensities of the probe in the absence and presence, respectively, of nitrophenol, and [Q] is the nitrophenol concentration. Estimated error bounds and R values were calculated using the Origin 8.0 software package. Binding constants were calculated according to non-linear curve fitting methods up to the saturation point in fluorescence titration studies using the equation, I = I₀ + I_aK_n[Guest]ⁿ/1 + K_n[Guest]ⁿ in the GNU plot ver. 5 software package, where I is the intensity (calculated as a function of Y where f(Y) = I), I₀ is the intensity of the probe only, I_a is intensity at the saturation point, and n is the stoichiometric ratio between the probe and guest. A 0.2 × 10⁻⁶ M solution of AIM-D was prepared in a 50 mL flask (±0.025 mL) using distilled water (degassed using ultrasonic and vacuum methods), river water, and sea water samples. The river and sea water samples were collected from the Sincheon River, Daegu and the seashore of Busan, respectively. The water samples were passed through filter paper before use in UV-visible and fluorescence experiments (see ESI†). The limit of detection (LOD) of nitrophenols was calculated from fluorescence titration data in the water samples after calibration.

2.3 Limit of detection using paper strips

Paper strips (5 × 2 cm) were coated with AIM-D (1 × 10⁶ M) followed by the removal of solvent under vacuum at room temperature. The coated filter papers were then cut into pieces (1 × 1 cm) to produce the test strips used for the detection of explosives.

2.4 Fluorescence lifetime measurements

Fluorescence lifetime decays were measured using an inverted scanning confocal microscope (MicroTime-200, Picoquant, Germany) with a 20× objective. These measurements were performed at the KBSI, Daegu Branch. A single-mode pulsed diode laser (375 nm with a ∼240 ps pulse width and ∼5 μW laser power) was used as the excitation source. A dichroic mirror (Z375RDC, AHF), a long pass filter (HQ405lp, AHF), and an avalanche photodiode detector (PDM series, MPD) were used to collect the emission from the samples. Time-resolved fluorescence photon counting and exponential fitting of the fluorescence decays were performed with SymPhoTime software (version 5.3).

2.5 Theoretical calculations

Interactions between AIM-D and nitrophenols were calculated using the SPARTAN 10 program. Geometry optimizations were performed using the Becke three-parameter exchange functional and the Lee–Yang–Parr B3LYP exchange correlation functional with the 6-31+G* basis set for C, H, N, and O. Calculations were performed under vacuum and in EtOH, DMSO, and H₂O (SM8 solvent model) to verify the effect of the solvent. All stationary points were verified as minima through full calculations of the Hessian analysis and harmonic frequency analysis.¹²

2.6 Synthesis

Synthesis of AIM-D. SOCl₂ (2 mL) was added to a solution of 9,10-anthracenedicarboxylic acid (30 mg, 0.114 mmol) in CH₂Cl₂ (1 mL) under a nitrogen atmosphere at 0 °C. The reaction mixture was stirred for 40 h at room temperature and concentrated to dryness. The crude acid chloride was subjected to a coupling reaction with 1-(3-aminopropyl)imidazole without further purification. Triethylamine (25 mg, 0.25 mmol) and 1-(3-aminopropyl)imidazole (32 mg, 0.25 mmol) were added to a solution of the crude acid chloride (35 mg, 0.115 mmol) in dry CH₂Cl₂ (1.5 mL) at 0 °C under dinitrogen. After addition, the reaction mixture was stirred at room temperature for 48 h. The suspended solid was filtered and washed with NaHCO₃ solution followed by water. The crude product was purified by column chromatography using 10–15% MeOH in CH₂Cl₂ as an eluent (77% yield). Mp 283–285 °C; ¹H NMR (400 MHz, DMSO-d₆) δ 2.10 (m, 4H, NHCH₂CH₂CH₂, H_e), 3.43 (t, J = 7.7 Hz, 4H, NHCH₂CH₂, H_f), 4.12 (t, J = 6.8 Hz, 4H, Im CH₂CH₂, H_d), 6.92 (s, 2H, Im H_c), 7.25 (s, 2H, Im H_b), 7.59–7.66 (m, 4H, H_j), 7.69 (s, 2H, Im H_a), 7.96 (dd, J = 8.3, 0.8 Hz, 4H, H_i), 8.94 (t, J = 7.6 Hz, 2H, NH, H_g); ¹³C NMR (100 MHz, DMSO-d₆) δ 30.7, 36.5, 43.7, 119.3, 119.4, 125.6, 126.5, 126.7, 128.2, 128.4, 137.2, 137.3, 168.0; HR-FAB MS calcd for C₂₀H₂₀N₃O [M + H]⁺: 481.2356, found: m/z: 481.2354.

3. Results and discussion

The symmetrical extension of conjugation at the 9- and 10-positions of anthracene in AIM-D was designed to provide better fluorescence sensing. Extended electronic delocalization in the fluorophore facilitates symmetrical charge transfer between the carboxamide nitrogen and the fluorophore resulting in a high quantum yield.¹³ The optimum size of the propylene spacer provides flexibility and helps to form a more stable complex with nitrophenols. The amidic carboxamide proton and the imidazole group play a part in binding that promotes changes in fluorescence emission in the presence of the analyte. The aromatic plane of the anthracene fluorophore provided room for π–π interactions with electron-deficient nitrophenols for additional stability. The anthracene-based carboxamide probe AIM-M has recently been reported to show a greater selectivity towards TNP in ethanol. To attain a higher surface area for the guest interactions in a small molecular architecture, the anthracene conjugations were extended at the 9^th and 10^th positions in a symmetrical way, which pulled the dipoles away from the anthracene fluorophore. Such intramolecular charge transfer and photoexcitation phenomena can affect the FRET-based acceptor molecules (typically nitrophenols in the current studies), which could result in better sensitivity and selectivity than the single-armed probe with a different fluorogenic response (Fig. 1).

Fig. 1 Anthracenecarboxamide probes AIM-D and AIM-M.

In designing candidates to sense electron-deficient nitrophenol derivatives, AIM-D attracted our attention as a result of its high quantum yield (Φ_AIM-D = 0.146) and uniform fluorescence in many solvents compared with single-armed AIM-M (Fig. S1†).¹⁴ Single-armed AIM-M showed a lower quantum yield (Φ_AIM-M = 0.052) in EtOH (0.2 μM) at λ_ex = 362 nm and variable fluorescence emission in solvents such as DMSO, isopropanol, glycerol, 1,4-dioxane, acetonitrile, dichloromethane, and chloroform (Fig. S2†). The incremental addition of water to 0.2 μM solutions of AIM-D in DMSO, isopropanol, glycerol, 1,4-dioxane, and acetonitrile eventually decreased its quantum yield. However, a high quantum yield was observed at a low AIM-D concentration (0.2 μM) in all solvents, which provides a basis for nitrophenol analysis in organic–aqueous mixtures.

To find the best solvent for UV-visible and fluorescence studies of TNP, the quenching efficiency with AIM-D was examined in various solvents. EtOH provided the greatest quenching among all solvents tested (Fig. S3†). Various aqueous–organic solvent systems were investigated in different proportions. The fluorescence emission decreased gradually with the incremental addition of water to a 0.2 μM AIM-D solution. An 1 : 1 EtOH–H₂O mixture was chosen because the AIM-D fluorescence suddenly dropped to a very low level (Φ = 0.02) when the water content exceeded 50% (Fig. S4†).

AIM-D was synthesized by the reaction of 9,10-anthracenedicarboxylic acid chloride with 1-(3-aminopropyl)imidazole in good yield (Scheme 1). The ¹H NMR spectrum of AIM-D in DMSO-d₆ displayed three imidazole protons (H_a, H_b, and H_c) at δ 7.69, 7.25, and 6.92, respectively, whereas the amidic –NH appeared as a triplet at δ 8.94. The proton signals of anthracene appeared in the aromatic region. The imidazole C-2 appeared at δ 137.7 in the ¹³C NMR spectrum of AIM-D and the fast atom bombardment mass spectrum showed the molecular ion [M + H]⁺ at m/z = 481.2354 (see ESI†).

Scheme 1 Synthesis of AIM-D.

The UV-visible spectrum of AIM-D (0.2 μM, EtOH–H₂O, 1 : 1) exhibited anthracene absorption bands at λ_max = 317, 349, 366, and 388 nm (Fig. S5†). On excitation at 366 nm, AIM-D (0.2 μM) exhibited fluorescence emission maxima characteristic of anthracene at λ_max = 393, 417, and 442 nm with a good quantum yield (Φ = 0.130). Although AIM-D exhibited several absorption maxima, 366 nm was used as the excitation wavelength for fluorescence studies as a result of the higher fluorescence emission (Φ_{349 nm} = 0.0897 and Φ_{388 nm} = 0.110) in EtOH–H₂O (1 : 1) compared with other values. On the addition of 5.0 equiv. of phenol derivatives, including phenol (P), 2-aminophenol (2-AP), 2-fluorophenol (2-FP), 2-nitrophenol (2-NP), 3-nitrophenol (3-NP), 4-iodophenol (4-IP), 4-bromophenol (4-BP), 4-chlorophenol (4-CP), 4-cyanophenol (4-CNP), 4-nitrophenol (4-NP), 3,4-dinitrophenol (3,4-DNP), 2,4-dinitrophenol (2,4-DNP), 2,4,6-trichlorophenol (TCP), 2,3,5,6-tetrafluorophenol (TFP), 2,3,4,5,6-pentafluorophenol (PFP), and 2,4,6-trinitrophenol (TNP), only TNP exhibited 96% quenching of fluorescence emission (Φ = 0.012) (Fig. 2). 2,4-DNP, 3,4-DNP, 4-NP, and 2-NP showed 85, 76, 49, and 40% quenching, respectively.

Fig. 2 Fluorescence spectra of AIM-D (0.2 μM, EtOH–H₂O, 1 : 1) in the presence of various phenols (5.0 equiv. each) at pH 7.0, λ_ex = 366 nm.

Fluorescence titrations were performed to explore the stoichiometry and strength of AIM-D binding with TNP in EtOH–H₂O (1 : 1) (Fig. 3). The fluorescence intensity gradually decreased at 393, 417, and 442 nm and increased at 520 nm with the addition of TNP aliquots to AIM-D (0.2 μM, EtOH–H₂O, 1 : 1). Fluorescence quenching continued until the addition of 2.0 equiv. of TNP and was saturated at 393, 417, and 442 nm. However, the fluorescence intensity at 520 nm increased regularly until the addition of 2.0 equiv.; quenching was saturated at this point, but resumed at greater concentrations (Fig. 3d).

Fig. 3 (a) Fluorescence titration of AIM-D (0.2 μM) with TNP in EtOH–H₂O (1 : 1) pH 7.0 at λ_ex = 366 nm; (b) molar ratio plot of fluorescence quenching at λ_em = 393 nm; (c) enlarged portion of the fluorescence spectrum at 500–525 nm after the addition of 0–2.0 equiv. of TNP; and (d) variation of fluorescence intensity versus equiv. of TNP at λ_em = 520 nm.

To investigate the selectivity among electron-deficient aromatic systems, aromatic carboxylic acids and hazardous nitroaromatics such as 1,4-dinitrobenzene (1,4-DNB), 1,3-dinitrobenzene (1,3-DNB), 2-nitrobenzoic acid (2-NBA), and 3,5-dinitrobenzoic acid (3,5-DNBA) were investigated. However, none of these systems showed a response in fluorescence studies. The increase in fluorescence intensity at 520 nm clearly indicated that AIM-D selectively detected TNP among various phenol derivatives based on the ratio of responses shown in Fig. 4.

Fig. 4 Relative fluorescence changes (I_{520 nm}/I_{417 nm}) of AIM-D (0.2 μM) with the addition of various phenol derivatives (5.0 equiv.) in EtOH–H₂O (1 : 1) at pH 7.0, λ_ex = 366 nm.

The relative fluorescence quenching ratio, [I₀ − I/I₀] × 100, and the Stern–Volmer plot, I₀/I = 1 + K_sv[Q], where I₀ and I are the fluorescence intensities in the absence and presence of TNP, K_sv is the Stern–Volmer quenching constant, and [Q] is the concentration of the quencher (Fig. S6†), indicated the formation of a 1 : 2 complex with TNP. The Stern–Volmer quenching constant (K_sv = 2.69 × 10⁷ M⁻¹) is the largest among the results reported in organic–aqueous mixtures.¹⁵ A plot of the AIM-D fluorescence intensity at λ_max = 393 nm versus [TNP] was linear with R² = 0.9978 and provided a detection limit of 1 nM (Fig. S7†).

The formation of a 1 : 2 complex between AIM-D and TNP was further supported by a Job's plot (Fig. S8†). The Stern–Volmer plot indicated a linear response range from 20 to 400 × 10⁻⁹ M with no further change beyond the addition of 2.0 equiv. of TNP due to the nearly complete quenching of anthracene emission. Excitation of a TNP solution (0.2 μM, EtOH : H₂O 1 : 1) at 366 nm produced no fluorescence emission at 380–440 nm (anthracene) and none at 520 nm. This further demonstrated that the generation of the unique band at 520 nm did not arise solely from TNP. Similar fluorescence titrations and Job's plot experiments of AIM-D with 2,4-DNP, 3,4-DNP, 2-NP, and 4-NP revealed the formation of 1 : 2 complexes with quenching constants, K_sv, of 1.82 × 10⁷, 1.00 × 10⁷, 7.21 × 10⁶, and 5.03 × 10⁶ M⁻¹, respectively (Fig. S9–S14†). These quenching constants are the largest among values reported in aqueous–organic media using small molecules.¹⁶

To differentiate TNP sensing among the nitrophenols, binding strengths ratios were calculated at two emission wavelengths. Binding constants were obtained through the non-linear curve fitting method up to the saturation point from the fluorescence titration studies and were denoted as K_{397 nm} and K_{520 nm}. The binding constant ratios obtained at 397–520 nm (Table 1) clearly demonstrated that TNP binds more strongly to AIM-D recognized through the ratiometric approach than to the other nitrophenols.

Table 1 Stern–Volmer quenching constants, binding constant ratios, and detection limits of various mono-, di-, and trinitrophenols in EtOH–H₂O (1 : 1) at pH 7.0^a

Nitrophenol	Binding stoichiometry	Ksv (M^−1)	K397 nm (M^−1)	K520 nm (M^−1)	K520 nm/K397 nm	Detection limit
a Stern–Volmer quenching constants and binding constants were obtained from fluorescence titration studies in EtOH–H2O (1 : 1). NR: no response and value taken as 0 for calculating the binding ratios.
TNP	1 : 2	2.69 × 10^7	8.82 × 10^7	4.32 × 10^7	0.489	1 ppb
2,4-DNP	1 : 2	1.82 × 10^7	2.20 × 10^7	NR	0	8 nM
3,4-DNP	1 : 2	1.00 × 10^7	7.34 × 10^6	NR	0	20 nM
2-NP	1 : 2	7.21 × 10^6	3.03 × 10^6	NR	0	40 nM
4-NP	1 : 2	5.03 × 10^6	1.12 × 10^6	NR	0	40 nM

---

The detection limit is shown in Fig. S15–S17.† The quenching constant and quenching ratio of each nitrophenol with AIM-D were clearly correlated with the position and number of nitro groups in the phenol derivatives. As nitro groups were introduced to the 2-, 4-, and 6-positions of the phenol (ortho and para), the quenching percentage increased, resulting in a large K_sv value. When the nitro groups were present at the 3- and 5-positions (meta), the quenching percentage decreased, creating a smaller quenching constant. The association constants of AIM-D with different nitrophenols decreased in the order TNP > 2,4-DNP > 3,4-DNP > 2-NP > 4-NP, which supports an oriented interaction contribution to the quenching phenomenon based on the position of the nitro group in the aromatic ring, which directly affects the acidity of the phenolic –OH and electron-deficient benzene ring of the respective phenols.

With DMSO, a negligible solvent effect (neither a shift nor a change) was observed in UV-visible and fluorescence studies (Fig. S18†). To investigate the nature of the interaction and binding sites of TNP complexation with AIM-D, an ¹H NMR study was carried out in DMSO-d₆–D₂O (2 : 1). In 1 : 1 DMSO-d₆–D₂O containing 5.0 × 10⁻³ M AIM-D, a proper spectrum was not obtained as a result of shimming and solvent locking problems.

The addition of 2.0 equiv. of TNP to a solution of AIM-D led to a pronounced downfield shift of the imidazole proton signals (a, b, and c in Fig. 5): H_a from δ 7.69 to 9.01 ppm (Δδ 1.32), H_b from δ 7.22 to 7.65 ppm (Δδ 0.43), and H_c from δ 6.92 to 7.45 ppm (Δδ 0.53). The amide proton (–CONH), H_g, showed a downfield shift from δ 8.96 to 9.47 ppm (Δδ 0.51) with increased intensity, confirming that rapid exchange of the amidic –NH proton is hampered by its involvement in the interaction with TNP. The propylene spacer protons (f) adjacent to the amide group showed an upfield shift from δ 3.42 to 3.13 ppm (Δδ −0.29), suggesting that the amide proton is involved in multiple hydrogen bonding interactions with TNP. In addition, the anthracene protons (i and j) showed an upfield shift; H_i from δ 7.89 to 7.76 ppm (Δδ −0.13) and H_j from δ 7.58 to 7.47 ppm (Δδ −0.11) (Table 2). The addition of one more equiv. of TNP resulted in no further shift in the proton signals, confirming the formation of a 1 : 2 complex between AIM-D and TNP (Fig. 6), as expected from the fluorescence titration studies. The large changes in chemical shift of the imidazole (a, b, and c), –CONH (g), and propylene spacer (f) protons, plus the upfield shift of the anthracene protons, clearly indicates binding of two TNP molecules to AIM-D. The downfield shifts of the imidazole protons reveal that, although the molecules possess two linkages with freely rotating propylene groups (which would give different signals after complexation), all the spacer protons (d, e, and f) shift together, demonstrating that two molecules of TNP are symmetrically bound to AIM-D.

Fig. 5 Partial ¹H NMR spectra of (i) AIM-D (5.0 × 10⁻³ M); after the addition of (ii) 2.0 equiv. and (iii) 3.0 equiv. of TNP; (iv) TNP in DMSO-d₆–D₂O (2 : 1).

Table 2 Proton chemical shift differences (Δδ) of AIM-D induced by complexation with TNP and 2,4-DNP in DMSO-d₆–D₂O (2 : 1)

Probe	Δδ^a
	Ha	Hb	Hc	Hd	He	Hf	Hg	Hi	Hj
a Chemical shift values are in ppm. Positive values indicate a downfield shift; negative values indicate an upfield shift.
TNP	1.32	0.43	0.53	0.35	0.16	−0.29	0.51	−0.13	−0.11
2,4-DNP	0.22	0.15	0.19	0.02	−0.07	−0.12	0.15	0.12	0.03

---

Fig. 6 Stern–Volmer plots for TNP with AIM-D (0.2 μM, EtOH–H₂O, 1 : 1) at pH 7.0 prepared with distilled water (black), river water (red), and sea water (blue) samples; λ_ex = 366 nm and λ_em = 393 nm.

To see the effect of binding 2,4-DNP to AIM-D, partial ¹H-NMR studies were carried out in DMSO-d₆–D₂O (2 : 1) (Fig. S19†). The addition of 2.0 equiv. of 2,4-DNP to AIM-D led to downfield shifts of the imidazole protons (a, b, and c) and an upfield shift of the propylene spacer (f) protons: H_a from δ 7.69 to 7.91 ppm (Δδ 0.22), H_b from δ 7.22 to 7.37 ppm (Δδ 0.15), H_c from δ 6.92 to 7.11 ppm (Δδ 0.19), and H_f from δ 3.42 to 3.30 ppm (Δδ −0.12) (Table 2). The amide proton, H_g, showed a downfield shift from δ 8.96 to 9.11 ppm (Δδ 0.15). The downfield shifts with 2,4-DNP were comparatively small, which accords with the greater binding strength of TNP indicated by the UV-visible and fluorescence results. This finding demonstrates that the additional –NO₂ group in the ortho position leads to stronger interactions.

Most reported TNP sensors have been based on electron-rich aromatic systems with conjugated amines or on fluorophores connected to the nitrogen atom of a heterocyclic aromatic moiety (usually with one or two methylene spacer groups),¹⁷ a metal–organic framework, or an imidazolium salt.¹⁸ Such probes usually suffer interference from protons, cations, and photo-induced electron transfer (PET), which is based on quenching phenomena in salt-based systems. As a result of these limitations, nitrophenol analysis must be performed in neutral solutions. The presence of the carboxamide (–CONH–) group, which is a typical functional group that can act as a proton donor (–NH–) and proton acceptor (–C=O–), can effectively provide fluorescence stability from pH 1.0 to 11.75 (Fig. S20†) and contributes well to binding. Fluorescence titration studies were performed in strongly acidic, acidic, and basic media (Fig. S21–S23†) with TNP to determine the response and binding strength. The observed Stern–Volmer plot and quenching ratios were similar under acidic (pH 3.5) and basic (pH 9.5) conditions with a linear relationship; however, the quenching ratios decreased due to the lower enhancement at 520 nm in extremely acidic conditions (pH 1.5). The highly acidic nature of TNP (pK_a 0.23) could efficiently protonate the imidazole rings, forming imidazolium ions, which interact with picrate ions through coulombic forces. However, this process was effectively inhibited in extremely acidic conditions due to the already protonated imidazole group and the weakly hydrogen-bonded carbonyl oxygen from TNP. The observed K_sv values and quenching ratios under strongly acidic (pH 1.5), acidic, and basic pH conditions showed that simple protonation at the imidazole group is not the only factor in the selective sensing process (Table 3). It also justified its separation from the fluorophore to attain low interference from the protons/cations. Mono- and dinitrophenols were unable to protonate the imidazole rings due to their weakly acidic nature (pK_a > 4.2). This is one of the main reasons for differentiating nitrophenols by the ratiometric response. The binding ratios in extremely acidic, acidic, and basic media clearly show the importance of the imidazole unit in selective sensing. As a result of the effective protonation of the interactions of the imidazole unit with the AIM-D probe, TNP can exist in the form of a coulombic mode in mixed aqueous media. Fluorescence titration studies were performed in the presence of a 10 mM (1000 times more concentrated than [AIM-D][TNP]) NaNO₃ solution (Fig. S24†) to evaluate this process. A negative deviation was observed in the Stern–Volmer plot with a lower quenching efficiency than that of the acidic and basic pH solutions. These results clearly revealed that the coulombic interactions also played a significant part in the selective sensing process. In addition, the AIM-D fluorescence emission did not vary in the presence of 10.0 equiv. of metal ions (Fig. S25†). This result indicates the selective binding of nitrophenols over metal ions in mixed aqueous–organic media.

Table 3 Binding constants, binding constant ratios at different emission wavelengths, and detection with TNP in EtOH–H₂O (1 : 1)^a

Nitrophenol	K397 nm (M^−1)	K520 nm (M^−1)	K520 nm/K397 nm
a Binding constants at different wavelengths of emission were obtained through the non-linear curve fitting of fluorescence titration data at specified wavelengths with the least error (reduced χ^2 analysis) bound.
pH 1.5	1.82 × 10^7	8.28 × 10^6	0.249
PH 3.5	7.92 × 10^7	3.66 × 10^7	0.462
PH 7.0	8.82 × 10^7	4.32 × 10^7	0.489
pH 9.5	8.80 × 10^7	4.28 × 10^7	0.486

---

To demonstrate the utility of AIM-D for quantitative analysis of nitrophenols in real water samples, the fluorescence efficiency of nitrophenols spiked into sea water and river water was evaluated in comparison with distilled water. The Stern–Volmer plots obtained from fluorescence titrations of river and distilled water samples containing TNP were nearly identical; the sea water samples showed a small negative deviation, which supported the interruptions in coulombic interactions by dissolved salts (strong electrolytes) in the sea water sample (Fig. 6).

Interferences from cations and anions in the aqueous–organic mixtures were negligible. The similar quenching constants, K_sv, obtained attest to the strength of interaction between AIM-D and TNP. The results in fresh and marine waters confirm that AIM-D can be used to sense nitrophenols in these media. The LOD for TNP in river and sea water was 20 nM; this is one of the lowest detection limits reported previously and primarily used a small, simple luminophore over a wide pH range.¹⁹

The real-world application of AIM-D was investigated using contact-mode recognition of nitrophenols on paper strips prepared by dipping into a 1 μM solution in EtOH–H₂O (1 : 1), followed by drying under vacuum at room temperature. Tests with various concentrations of aqueous TNP solutions (2.5 μL) spotted onto the probe-impregnated paper strips showed dark spots amidst strong luminescence on UV illumination at 365 nm (Fig. 7).

Fig. 7 Photographs of Whatman filter paper strips coated with AIM-D and spotted with different concentrations of TNP: (a) 0 M (water blank) and (b) 20, (c) 50, (d) 100, (e) 200, and (f) 400 nM TNP, respectively, on illumination at 365 nm with an UV lamp.

The different intensities of the dark spots provided immediate visualization for trace amounts of TNP. The LOD of TNP by naked eye observation on illumination at 365 nm was 20 nM. 2,4-DNP, 3,4-DNP, 4-NP, and 2-NP also were tested and provided LODs of 40, 75, 100, and 100 nM, respectively (Fig. S26†).

Stern–Volmer plots showed linear behaviour from 0 to 400 nM for TNP, 2,4-DNP, 3,4-DNP, 4-NP, and 2-NP. A saturation zone was observed from 400 nM to 1 μM (Fig. S27–S30†). A positive deviation was observed at concentrations beyond the saturation zone for the mono- and dinitrophenols, which could result from spectral overlap. For this reason, 2-NP and 4-NP showed positive deviations at concentrations >2000 nM. AIM-D can be used effectively to sense nitrophenols below 2.0 μM in mixed aqueous media. The limits of detection for 2,4-DNP, 3,4-DNP, 2-NP, and 4-NP are tabulated in Table 4.

Table 4 Detection limits of nitrophenols in various water samples with an AIM-D-coated paper strip^a

Nitrophenols	Limit of detection (nM)
	Distilled water	Paper strip	Sea water	River water
a Detection limits were calculated using fluorescence titration data.
4-NP	40	100	50	50
2-NP	40	100	50	50
3,4-DNP	20	75	30	30
2,4-DNP	8	40	20	20
TNP	1	20	20	20

---

The DFT calculations support the presence of hydrogen bonding in AIM-D²⁺·(TNP⁻)₂ and confirm the evidence from the UV-visible, fluorescence, and ¹H NMR experiments. Intermolecular distances of 1.96 and 1.92 Å, respectively, were calculated for the hydrogen bonds formed by the amidic proton of the anthracenecarboxamide in AIM-D²⁺ with the deprotonated phenolic oxygen (O⁻) moiety of picrate (depicted as a in Fig. 8) and with the nitro group in the 2-position of TNP (b). We previously reported the involvement of the acidic C2H proton in multiple hydrogen bonding interactions with the electronegative heteroatoms (most likely O, F, and N atoms) of salicylic acid derivatives and the electron-deficient phenols for highly selective recognition. By contrast, the C2-H proton became more acidic than the neutral form as a result of the protonation of the N_sp² of imidazole induced by the highly acidic TNP. Such protonation substantially enhanced the tendency for interactions with the C-2 proton (H_a in Fig. 5) and the nitro group at the 2-position (f, NO₂⋯HC-2 = 1.99 Å) as well as the deprotonated phenolic oxygen (O⁻) (c, 1.90 Å) of picrate due to its more highly acidic character. In addition, the proton of the protonated nitrogen co-operatively assisted by forming a hydrogen bond with one of the ortho-nitro groups (e, 1.98 Å) and O⁻ (d, 1.95 Å) of picrate, which stabilized the complex through coulombic interactions. The furthest distance between the carboxamide amidic proton and the ortho-nitro group (b) was 1.92 Å, which indicates that it is nearer to the probe AIM-D (clearly visualized in Fig. 8). The interplanar π–π stacking distance (g) was 3.90 Å. The energy-minimized structure showed symmetrical interactions above and below the anthracene plane with the picrate ions. The energy of the LUMO of the protonated AIM-D²⁺ probe (−2.88 eV) was higher than the LUMO of the picrate ion (−4.74 eV) that was responsible for the electron transfer (ET) process from the former to the latter. The band gap energy, ΔE_{(HOMO–LUMO)}, decreased significantly from −3.47 to −2.13 eV with the formation of AIM-D²⁺·(TNP⁻)₂ from AIM-D. This decrease (−1.34 eV) was much greater than the differences observed on the formation of the other 1 : 2 complexes, which were −0.85, −0.03, −0.08, and −0.04 eV for 2,4-DNP, 3,4-DNP, 2-NP, and 4-NP, respectively (see Table S1†). The weaker acidic strength of the di- and mononitrophenols, which was insufficient to protonate the imidazole, was one of the reasons for the decreased stabilization of the complex with AIM-D compared with TNP. The stability of the various nitrophenol complexes showed that the interaction sites and modes (coulombic/non-coulombic) were important, and their orientations were responsible for the strength of the complex.

Fig. 8 DFT based studies showing the HOMO, LUMO and band gap energies (eV) of (i) probe AIM-D²⁺ and TNP⁻ and (ii) the energy-minimized structure of AIM-D²⁺·(TNP⁻)₂ calculated by B3LYP/6-31G* in EtOH (SM8 model) and (iii) the vertical view. The labelled vital symmetrical interactions points are identified by colour and prime notation (see text).

The overlap of the emission spectrum of AIM-D and the absorption spectra of nitrophenols (Fig. S31†) is responsible for resonance energy transfer (RET).²⁰ However, the quenching intensity differed in individual cases. The observed differences in quenching intensity are consistent with the presence of RET, negligible optical inner filter effects, and synergistic hydrogen bonding and π–π interactions between the electron-rich fluorophore and electron-deficient nitrophenols. As observed by UV-visible, fluorescence, and ¹H NMR spectrometry, the ortho- and para-substituted nitrophenols showed more pronounced interactions with AIM-D. Binding was strengthened by the presence of these substituents, which provided a point of attachment and resonance-induced electron withdrawal from the benzene ring that directly affected the acidic nature of phenolic –OH. Stronger bonding shortens the intermolecular distance, resulting in more effective RET from AIM-D to nitrophenol. This point is supported by the energy-minimized structures of the AIM-D complexes with the nitrophenols (Fig. S32†). The optimized solvation energies for nitrophenols showed the effect of the solvent on the strength of interaction (Fig. S33†). Fluorescence decay measurements for AIM-D in the presence and absence of TNP showed the same behaviour (Fig. S34†). The linearity of the Stern–Volmer plots (Fig. 9) and the invariant fluorescence emission irrespective of the oxygen concentrations confirmed that the quenching intensity and strength of the interaction of AIM-D with nitrophenols are due to the formation of 1 : 2 ground state complexes and that RET from the excited state of AIM-D to the ground state of the nitrophenol depends mainly on the interaction sites and the acidity of the phenol groups.²¹ Fluorescence studies in the absence and presence of dissolved oxygen were found to be similar, suggesting a negligible quenching effect from dissolved oxygen.

Fig. 9 Stern–Volmer plots of AIM-D (0.2 μM) with nitrophenols in an EtOH–H₂O (1 : 1) mixture at pH 7.0.

Thus the binding mode of the probe AIM-D towards the nitrophenols over a wide pH range (1.0–11.5) can be attributed to the multiple hydrogen bonds and π–π interactions that occurred in a synergistic manner and which decreased the intermolecular distance for an efficient RET from anthracene to that of each respective nitrophenol. However, the high sensitivity and ratiometric behaviour observed for TNP could be due to the additional proton transfer from TNP to the imidazole unit of the AIM-D probe, followed by electron transfer from anthracene to picrate (see ESI,† the LUMO distribution of the AIM-D²⁺·(TNP⁻)₂ complex located on the benzene ring of picrate ions in both planes). The former caused strong intermolecular opposing charge flows and coulombic interactions between the picrate and imidazolium ions, together with synergistic multiple hydrogen bonds and π–π interactions. The substantial monomer quenching in the nitrophenols was due to RET (dependent on the probe–analyte interaction point); the intermolecular sequential proton and electron charge flow in the opposite direction can symmetrically inhibit the intramolecular charge transfer process from the anthracene unit to the carboxamide nitrogen via a conjugation mode, which resulted in a new band at 520 nm in the case of TNP.²² By contrast, the higher pK_a values of the di- and mononitrophenols could not efficiently protonate the imidazole unit. This is one of the crucial factors in obtaining lower monomeric quenching (only due to less efficient RET) without a ratiometric response. The vital protonation site of the imidazole unit and the wide pH fluorescence stability of the CO–NH group were isolated through the propyl spacer and clearly justified a rational design of the probe, more suitable for nitrophenol sensing and TNP discrimination in mixed aqueous media. Henceforth, successful new strategies were implemented for a more efficient recognition of the nitrophenols and to discriminate TNP from the di- and mononitrophenols with various factors (Fig. 10) that decided the fluorogenic behaviour of the AIM-D probe based on the protonating ability towards imidazole and the synergistic molecular interactions (such as multiple hydrogen bonding, π–π interactions, and coulombic interactions).

Fig. 10 Plausible sensing mechanism of the AIM-D probe with nitrophenols.

4. Conclusion

A novel, simple, and rapid method was developed with a new strategy for the recognition of nitrophenols without interference in organic–aqueous media over a wide pH range (1.0–11.5). AIM-D differentiated TNP from other nitrophenols through the ratiometric fluorescence response and produced an LOD of 1 ppb. Higher selectivity through the ratiometric response towards TNP occurred as a result of selective protonation and synergistic interactions, such as multiple hydrogen bonding, π–π interactions, and coulombic interactions. These interactions efficiently reduced the intermolecular distance to give a higher probability of RET. This resulted in monomeric band quenching and sequential intermolecular opposite charge flow aided the inhibition of intramolecular charge transfer from the anthracene fluorophore to the carboxamide nitrogen, which was responsible for the ratiometric response with TNP. Mono- and dinitrophenols were effectively differentiated based on the quenching efficiency at the sub-micromolar level in mixed aqueous media, a process which mainly relied on the positions of the nitro groups that can modulate the RET efficiency through the points of contacts between the respective nitrophenols and AIM-D. The practical application of AIM-D was demonstrated by coating a paper strip with AIM-D and observing the presence of nitrophenols with the naked eye under UV irradiation. The LODs for TNP, 2,4-DNP, 3,4-DNP, 2-NP, and 4-NP were 20, 40, 75, 100, and 100 nM, respectively. Nitrophenols also were detected in river and sea water samples with LODs of 20, 20, 30, 50, and 50 nM for TNP, 2,4-DNP, 3,4-DNP, 2-NP, and 4-NP, respectively.

Acknowledgements

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, Future Planning, Republic of Korea (2013R1A1A2006777).

References

1. R. G Smith, N. D'Souza and S. Nicklin, Analyst, 2008, 133, 571–584.
2. X. Sun, Y. Wang and Y. Lei, Chem. Soc. Rev., 2015, 44, 8019–8061.
3. Y. Gui, C. Xie, J. Xu and G. Wang, J. Hazard. Mater., 2009, 164, 1030–1035.
4. (a) N. Venkatesan, V. Singh, P. Rajakumar and A. K. Mishra, RSC Adv., 2014, 4, 53484–53489; (b) L. Wenfeng, M. Hengchang and L. Ziqiang, RSC Adv., 2014, 4, 39351–39358.
5. R. H. Yang, K. M Wang, D. Xiao, K. Luo and X. H Yang, Analyst, 2000, 125, 877–882.
6. A. Ding, L. Yang, Y. Zhang, G. Zhang, L. Kong, X. Zhang, Y. Tian, X. Tao and J. Yang, Chem.–Eur. J., 2014, 20, 12215–12222.
7. (a) Z. Yang, J. Cao, Y. He, J. H. Yang, T. Kim, X. Peng and J. S. Kim, Chem. Soc. Rev., 2014, 43, 4563–4601; (b) Y. X. Shi, F. L. Hu, W. H. Zhang and J. P. Lang, CrystEngComm, 2015, 17, 9404–9412; (c) S. Pramanik, Z. Hu, X. Zhang, C. Zheng, S. Kelly and J. Li, Chem.–Eur. J., 2013, 19, 15964–15971.
8. (a) M. Mohan and D. K. Chand, Anal. Methods, 2014, 6, 276–281; (b) R. Chopra, P. Kaur and K. Singh, Anal. Chim. Acta, 2015, 864, 55–63; (c) M. Mei, X. Huang, J. Yu and D. Yuan, Talanta, 2015, 134, 89–97; (d) B. Horstkotte, O. Elsholz and V. C. Martin, Talanta, 2008, 76, 72–79; (e) R. H. Yang, K. M. Wang, D. Xiao, K. Luo and X. H. Yang, Analyst, 2000, 125, 877–882.
9. J. Pan, F. Tang, A. Ding, L. Kong, L. Yang, X. Tao, Y. Tiana and J. Yang, RSC Adv., 2015, 5, 191–195.
10. (a) J. Yoon, J. R. Jadhav, J. M. Kim, M. Cheong, H.-S. Kim and J. Kim, Chem. Commun., 2014, 50, 7670–7672; (b) A. Kumar, A. Pandith and H.-S. Kim, Sens. Actuators, B, 2016, 231, 293–301; (c) A. Kumar, M. K. Ghosh, C. H Choi and H.-S. Kim, RSC Adv., 2015, 5, 23613–23621; (d) A. Kumar, A. Pandith and H.-S. Kim, J. Lumin., 2016, 172, 309–316.
11. A. Kumar, A. Pandith and H.-S. Kim, Dyes Pigm., 2015, 122, 351–358.
12. (a) Spartan, Wavefunction, Inc., 2008, http://www.wavefun.com/index.html; (b) F. De-Proft and P. Geerlings, Chem. Rev., 2001, 101, 1451–1464; (c) A. Pandith, A. Kumar and H.-S. Kim, RSC Adv., 2015, 5, 81808–81816.
13. M. Zhu and C. Yang, Chem. Soc. Rev., 2013, 42, 4963–4976.
14. A. Pandith, A. Kumar, J.-Y. Lee and H.-S. Kim, Tetrahedron Lett., 2015, 56, 7094–7099.
15. (a) L. Bingxin, C. Tong, L. Feng, C. Wang, Y. He and C. Lu, Chem.–Eur. J., 2014, 20, 2132–2137; (b) X. Sun, X. Ma, C. V. Kumar and Y. Lei, Anal. Methods, 2014, 6, 8464–8468; (c) A. Ding, L. Yang, Y. Zhang, G. Zhang, L. Kong, X. Zhang, Y. Tian, X. Tao and J. Yang, Chem.–Eur. J., 2014, 20, 12215–12222; (d) Z. F. An, C. Zheng, R. F. Chen, J. Yin, J. J. Xiao, H. F. Shi, Y. Tao, Y. Qian and W. Huang, Chem.–Eur. J., 2012, 18, 15655–15661; (e) K. D. Prasad and T. N. G. Row, RSC Adv., 2014, 4, 45306–45310.
16. (a) S. Paliwal, M. Wales, T. Goodc, J. Grimsley, J. Wild and A. Simonian, Anal. Chim. Acta, 2007, 596, 9–15; (b) H. T. Feng and Y. S. Zheng, Chem.–Eur. J., 2014, 20, 195–201; (c) A. Gupta, J. H. Lee, J. H. Seo, S. G. Lee and J. S. Park, RSC Adv., 2014, 5, 73989–73992; (d) J. H. Ko, J. H. Moon, N. Kang, J. H. Park, H. W. Shin, N. Park, S. Kang, S. M. Lee, H. J. Kim, T. K. Ahn, J. Y. Lee and S. U. Son, Chem. Commun., 2015, 51, 8781–8784; (e) A. Yadav and R. Boomishankar, RSC Adv., 2015, 5, 3903–3907; (f) S. Madhu, A. Bandela and M. Ravikanth, RSC Adv., 2014, 4, 7120–7123.
17. (a) W. Wei, X. Huang, K. Chen, Y. Tao and X. Tang, RSC Adv., 2012, 2, 3765–3771; (b) L. Wenfeng, M. Hengchang and L. Ziqiang, RSC Adv., 2014, 4, 39351–39358.
18. (a) B. Roy, A. K. Bar, B. Gole and P. S. Mukherjee, J. Org. Chem., 2014, 78, 1306–1310; (b) R. Kumar, S. Sandhu, P. Singh, G. Hundal, M. S. Hundal and S. Kumar, Asian J. Org. Chem., 2014, 3, 805–813.
19. O. Pinrat, K. Boonkitpatarakul, W. Paisuwan, M. Sukwattanasinitt and A. Ajavakom, Analyst, 2015, 140, 1886–1893.
20. K. Acharyya and P. S. Mukherjee, Chem.–Eur. J., 2015, 21, 6823–6831; A. K. Bandela, S. Bandaru and C. P. Rao, Chem.–Eur. J., 2015, 21, 1–12.
21. S. Shanmugaraju, H. Jadhav, R. Karthik and P. S. Mukherjee, RSC Adv., 2013, 3, 4940–4950.
22. (a) X. Y. Wan, F.-L. Jiang, C.-P. Liu, K. Zao, L. Chen, Y.-L. Gai, Y. Yang and M.-C. Hong, J. Mater. Chem. A, 2015, 3, 22369–22376; (b) V. Vij, V. Bhalla and M. Kumar, ACS Appl. Mater. Interfaces, 2013, 5, 5373–5380; (c) N. Venkatramaiah, A. D. G. Firmino, F. A. A. Paz and J. P. C. Tome, Chem. Commun., 2014, 50, 9683–9686; (d) A. I. Novaira, V. Avilaa, G. G. Montich and C. M. Previtalia, J. Photochem. Photobiol., B, 2001, 60, 25–31.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra17407h

---