Self-Associated 1,8-Naphthalimide as a Selective Fluorescent Chemosensor for Detection of High pH in Aqueous Solutions and Their Hg2+ Contamination

A novel diamino triazine based 1,8-naphthalimide (NI-DAT) has been designed and synthesized. Its photophysical properties have been investigated in different solvents and its sensory capability evaluated. The fluorescence emission of NI-DAT is significantly impacted by the solvent polarity due to its inherent intramolecular charge transfer character. Moreover, the fluorescence emission quenched at higher pH as a result of photo-induced electron transfer (PET) from triazine moiety to 1,8-naphthalimide after cleaving hydrogen bonds in the self-associated dimers. Furthermore, the new chemosensor exhibited a good selectivity and sensitivity towards Hg2+ among all the used various cations and anions in the aqueous solution of ethanol (5:1, v/v, pH = 7.2, Tampon buffer). NI-DAT emission at 540 nm was quenched remarkably only by Hg2+, even in the presence of other cations or anions as interfering analytes. Job’s plot revealed a 2:1 stoichiometric ratio for NI-DAT/Hg2+ complex, respectively.


Introduction
Mercury and its compounds are involved in several important industries such as pharmaceuticals, paints, agricultural chemicals, measuring instruments, chlor-alkali industry, etc. [1]. However, mercury is one of the most toxic heavy metals that threaten the human health and environment. All chemical forms of Hg 2+ are chronically toxic to humans, especially inorganic derivatives exhibit extremely severe toxicity that leads to a variety of diseases and organs damage [2]. Some organomercurials, particularly low molecularweight alkyl compounds, are reported to be highly toxic due to their strong affinity to bind with enzymes and proteins leading to mostly irreversible cell dysfunction [3]; nervous system damages [4,5], endocrine disorders [6], as well as failure of the immune [7] and digestive systems [8]. Hence, there is a growing interest in providing new reliable selective and sensitive methods to detect mercury in biological and environmental systems.
All used fine chemicals and solvents were of spectroscopic grade purity. 1 H spectrum was recorded on a Bruker Avance II + 600 spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany) operating at 600.13 MHz using DMSO-d 6 as a solvent at 25 • C. Absorption and emission spectra were recorded on a Varian Cary 5000 UV-Vis-NIR Spectrophotometer and a "Cary Eclipse" fluorometer (Agilent Technologies Deutschland GmbH, Darmstadt, Germany), respectively, using quartz cuvettes (Hellma, Munich, Germany). Origin pro 8 software was used for processing the absorption and emission data. The reaction was monitored on a TLC (Fluka F 60 254 20 × 20; 0.2 mm) with a 4:1 toluene:methanol solution as an eluent. TLC plates were investigated under UV light. Melting points were measured using a Hinotek-X4 micro melting point apparatus (Hinotek, Ningbo, China). All 2D structures were drawn using Chem3D ultra 9.0 (Chem Office 2008) program. A detailed description of the adopted computational protocol for the DFT calculations is given as Supplementary Information.

Synthesis of 4-Nitro-1,8-naphthalic anhydride (4)
4-Nitroacenaphthene (10 g, 50 mmol) were added to glacial acetic acid (100 mL) in a three-necked flask and heated to 95 • C. Then, sodium bichromate (58 g, 0.2 mol) were added in portions. The reaction temperature was kept 95 • C through the addition. After stirring for 4 h at 95 • C, the reaction mixture was poured into cold water (0.5 L) and the resulting precipitate was filtered off, washed with water, and dried. The crude product was treated with 150 mL aqueous NaOH (5%), filtration, and then the pH of the filtrate was adjusted to 4 using conc. HCl. The resulting precipitate was filtered, washed with water, and dried to give 3, which was dehydrated by heating at 130 • C for 10 h to give 4. Yield: 70%, mp 230-233 • C (lit. 233 • C) [70].

Design and Synthesis of NI-DAT
4-(4,6-diamino-1,3,5-triazin-2 -ylamino)-N-(2-dimethylaminoethyl)-1,8-naphthal-imide NI-DAT was prepared according to Scheme 1, presenting the nucleophilic substitution of 2-chloro-4,6-diaminotriazine by a 4-amino-N-(2-dimethylaminoethyl)-1,8-naphthalimide. The chemical structure of NI-DAT was confirmed by IR, NMR absorption, and emission spectra. 1 H-NMR spectrum, (see the Supplementary Information), confirms the formation of a stable self-associated dimer of the compound. The design of NI-DAT is based on appending 4,6-diaminotriazine, a well-known chelating moiety, to 4-amino-1,8-naphthalimide, whose photophysical properties are influenced by various stimuli. Moreover, the donoracceptor-donor feature of DAT that leads to self-associations in solutions (presumably by forming hydrogen-bonded dimers I-III, Figure 1), blocks PET triggering in the system that can be retrieved by stimuli able to separate the probe molecules of their dimers. (presumably by forming hydrogen-bonded dimers I-III, Figure 1), blocks PET triggering in the system that can be retrieved by stimuli able to separate the probe molecules of their dimers. (presumably by forming hydrogen-bonded dimers I-III, Figure 1), blocks PET triggering in the system that can be retrieved by stimuli able to separate the probe molecules of their dimers.   Table 1 presents the photophysical characteristics of NI-DAT in solvents of different polarity: absorption (λ A ) and emission maxima (λ F ), Stokes shifts (ν A − ν F ), and quantum yield of florescence (Φ F ) using fluorescein as a standard (Φ st = 0.79 in ethanol). The fluorescence quantum yield was calculated on the basis of the absorption and fluorescence spectra of NI-DAT by Equation (1).

Photophysical Characteristics
where Φ F is the emission quantum yield of the sample; Φ st is the emission quantum yield of fluorescein as standard (Φ st = 0.79 in ethanol), A st and A u represent the absorbance of the standard and sample at the excitation wavelength (λ ex. = 450 nm), respectively; while S st and S u are the integrated emission band areas of the standard and sample, respectively, and n st and n u are the solvent refractive index of the standard and sample; subscripts u and s refer to the unknown (sample) and standard, respectively. Stokes shifts were calculated using Equation (2).  As shown in Figure 2, the new NI-DAT has an emission band centered at 500-540 nm which is a mirror image of the corresponding absorption band centered at 417-440 nm. The observed bathochromic shift of the absorption and emission maxima upon increasing the solvent polarity (Figures 2 and 3) was ascribed to the charge transfer as a result of dipole-dipole interactions of the solvents with 1,8-naphthalimide chromophore system. That fact is in good accordance with the characteristics of the reported 1,8-naphthalimide containing compounds [73][74][75][76]. The red-shifts of the emission were higher than their absorption counterparts due to the large difference in the dipole moments in the excited and ground states.

Impact of the Medium pH
To investigate the effect of pH of the medium on the photophysical properties, the absorption and emission spectra of NI-DAT have been measured at different pH of the water/ethanol (5:1, v/v) solution. As shown in Figure 4, the absorption spectrum didn't change by varying the pH. On the other hand, the fluorescence emission quenched at higher pH due to the deprotonation of NH 2 at pKa = 12.1 which was calculated using the Henderson-Hasselbach Equation (3). where I max and I min are maximum and minimum fluorescence intensity, respectively; I is the fluorescence intensity at the given pH value.         At lower pH, the inclusion of nitrogen atoms (imine nitrogen of triazine moiety) in the hydrogen bonding of the dimer assembly, shown in Figure 1, blocks PET to 1,8naphthalimide and hence opens fluorescence emission like the related reported 1,8naphthalimides analogs [77][78][79][80][81]. In contrast, the probe responds to higher pH by an on-off response due to the deprotonation of NH 2 and concomitant switching of PET, Scheme 2.

Impact of the Medium pH
To investigate the effect of pH of the medium on the photophysical properties, the absorption and emission spectra of NI-DAT have been measured at different pH of the water/ethanol (5:1, v/v) solution. As shown in Figure 4, the absorption spectrum didn't change by varying the pH. On the other hand, the fluorescence emission quenched at higher pH due to the deprotonation of NH2 at pKa = 12.1 which was calculated using the Henderson-Hasselbach Equation (3).
where Imax and Imin are maximum and minimum fluorescence intensity, respectively; I is the fluorescence intensity at the given pH value. At lower pH, the inclusion of nitrogen atoms (imine nitrogen of triazine moiety) in the hydrogen bonding of the dimer assembly, shown in Figure 1, blocks PET to 1,8-naphthalimide and hence opens fluorescence emission like the related reported 1,8-naphthalimides analogs [77][78][79][80][81]. In contrast, the probe responds to higher pH by an on-off response due to the deprotonation of NH2 and concomitant switching of PET, Scheme 2.    As Figure 4B shows, PET blocking of the probe at neutral pH can be exploited for further sensory applications of cations and/or anions detection. Coordination with a specific cation may lead to breaking the dimer assemblies giving a selective on-off PET response induced by the recovery of PET to naphthalimide moiety.

Sensory Applications towards Cations and Anions
The sensor ability of NI-DAT towards cations and anions has been investigated by recording the absorption and emission spectra in the presence of cations or anions. Several cations, including Cu 2+ , Co 2+ , Hg 2+ , Zn 2+ , Ni 2+ , Pb 2+ , Sn 2+ , Sr 2+ , Ba 2+ , Mg 2+ , Fe 3+ , and Al 3+ (as nitrate salts), and different anions, including CN − , S 2− , HPO 4 Figure 5, among all the used cations, only Hg 2+ affected the emission of the probe where it quenched its emission (EQ = 77%). This emission quenching (EQ) has been ascribed to the coordination of Hg 2+ with NH 2 groups switching the PET to 1,8-naphthalimide from triazine nitrogens. On the other hand, none of the cations under study significantly affected the absorption spectrum, which confirms that NH-group, attached directly to 1,8-naphthalimide, does not participate in the coordination with Hg 2+ . nitrate salts), and different anions, including CN − , S 2− , HPO4 2− , H2PO 4− , F − , S2O5 2− , SO NO2 − , CO3 2− , and CH3COO − (as sodium salts), have been used in the study. Moreover, experiments have been performed using aqueous solution of NI-DAT (5:1 water:etha in the presence of a Tampon buffer to keep a constant pH (7.4) and the variations in spectra due only to the coordination with the added analyte. As shown in Figure 5, am all the used cations, only Hg 2+ affected the emission of the probe where it quenched emission (EQ = 77%). This emission quenching (EQ) has been ascribed to the coordina of Hg 2+ with NH2 groups switching the PET to 1,8-naphthalimide from triazine nitrog On the other hand, none of the cations under study significantly affected the absorp spectrum, which confirms that NH-group, attached directly to 1,8-naphthalimide, d not participate in the coordination with Hg 2+ .   Figure 6, the stoichiometric ratio was found to be 2:1 for the complexion of NI-DAT and Hg 2+ , respectively.
The stoichiometric ratio of Hg 2+ and the probe in their complex was determined using Job's plot analysis by plotting (F0 − F)·(1 − X), where F and F0 are the emissions of the probe in the absence and presence of Hg 2+ , respectively, against the molar ratio of Hg 2+ (X = [Hg 2+ ]/([NI-DAT] + [Hg 2+ ]). As shown in Figure 6, the stoichiometric ratio was found to be 2:1 for the complexion of NI-DAT and Hg 2+ , respectively. Furthermore, the selectivity of the new chemosensor to Hg 2+ response has been examined by measuring the emission spectrum of the probe in the presence of 5 equivalents of both Hg 2+ and the interfering cation or anion. As shown in Figure 7, the emission response of the probe toward the presence of Hg 2+ was not affected by the coexistence of any of the interfering cations or anions under the study.   Furthermore, the selectivity of the new chemosensor to Hg 2+ response has been examined by measuring the emission spectrum of the probe in the presence of 5 equivalents of both Hg 2+ and the interfering cation or anion. As shown in Figure 7, the emission response of the probe toward the presence of Hg 2+ was not affected by the coexistence of any of the interfering cations or anions under the study.
The stoichiometric ratio of Hg 2+ and the probe in their complex was determined using Job's plot analysis by plotting (F0 − F)·(1 − X), where F and F0 are the emissions of the probe in the absence and presence of Hg 2+ , respectively, against the molar ratio of Hg 2+ (X = [Hg 2+ ]/([NI-DAT] + [Hg 2+ ]). As shown in Figure 6, the stoichiometric ratio was found to be 2:1 for the complexion of NI-DAT and Hg 2+ , respectively. Furthermore, the selectivity of the new chemosensor to Hg 2+ response has been examined by measuring the emission spectrum of the probe in the presence of 5 equivalents of both Hg 2+ and the interfering cation or anion. As shown in Figure 7, the emission response of the probe toward the presence of Hg 2+ was not affected by the coexistence of any of the interfering cations or anions under the study.   The sensitivity of NI-DAT to detect Hg 2+ has been estimated from the titration plot of the emission at 540 nm against the concentration of Hg 2+ , shown in Figure 8, using the formula of the limit of detection (LOD) = 3σ/b, where σ is the slandered deviation of the emission of the probe recorded at 540 after repeating the measurements 10 times, and b is the slop of the titration plot. The found LOD was found to be 2 × 10 −7 M.    The sensitivity of NI-DAT to detect Hg 2+ has been estimated from the titration plot of the emission at 540 nm against the concentration of Hg 2+ , shown in Figure 8, using the formula of the limit of detection (LOD) = 3σ/b, where σ is the slandered deviation of the emission of the probe recorded at 540 after repeating the measurements 10 times, and b is the slop of the titration plot. The found LOD was found to be 2 × 10 −7 M. The sensitivity of NI-DAT to detect Hg 2+ has been estimated from the titration plot of the emission at 540 nm against the concentration of Hg 2+ , shown in Figure 8, using the formula of the limit of detection (LOD) = 3σ/b, where σ is the slandered deviation of the emission of the probe recorded at 540 after repeating the measurements 10 times, and b is the slop of the titration plot. The found LOD was found to be 2 × 10 −7 M.

Computational Studies on the Structure of NI-DAT and Its Hg 2+ and Mg 2+ Complexes
The titled compound NI-DAT was investigated theoretically by means of DFT calculations. A full geometry optimization of the sensor molecule was performed at B3LYP/6-31+G(d,p) level of theory in the gas phase (ε = 1.0). The gas phase optimized structure of the low-energy isomer of NI-DAT is visualized in Figure S3A. The coordination mode between the sensor molecule and Hg 2+ ions was probed computationally by modeling a ligand/metal architecture from simplified ligand model-2,4,6-triamino-1,3,5-triazine (TAT). TAT was also used in modeling H-bonded dimeric structures, and the intermolecular NH . . . N distances were measured to be 2.02 Å ( Figure S3B).
The stoichiometry proposed for the binding between NI-DAT and Hg 2+ is 2:1 ( Figure 5, Job's plot experiment). Our complex was modeled by placing Hg 2+ spatially close to one of the endocyclic N atoms of TAT for a complex with 2:1 (TAT:metal) stoichiometry. It is known that Hg 2+ is hydrated by a first solvation shell of six water molecules [82]. Thus, the TAT:metal complex was modeled with a six coordinate octahedral geometry with four water molecules at the metal cation. Figure 9A depicts the simplified model TAT complexing Hg 2+ as a monodentate ligand at 2:1 stoichiometry. The titled compound NI-DAT was investigated theoretically by means of DFT calculations. A full geometry optimization of the sensor molecule was performed at B3LYP/6-31+G(d,p) level of theory in the gas phase (ε = 1.0). The gas phase optimized structure of the low-energy isomer of NI-DAT is visualized in Figure S3A. The coordination mode between the sensor molecule and Hg 2+ ions was probed computationally by modeling a ligand/metal architecture from simplified ligand model-2,4,6-triamino-1,3,5-triazine (TAT). TAT was also used in modeling H-bonded dimeric structures, and the intermolecular NH…N distances were measured to be 2.02 Å ( Figure S3B).
The stoichiometry proposed for the binding between NI-DAT and Hg 2+ is 2:1 ( Figure  5, Job's plot experiment). Our complex was modeled by placing Hg 2+ spatially close to one of the endocyclic N atoms of TAT for a complex with 2:1 (TAT:metal) stoichiometry. It is known that Hg 2+ is hydrated by a first solvation shell of six water molecules [82]. Thus, the TAT:metal complex was modeled with a six coordinate octahedral geometry with four water molecules at the metal cation. Figure 9A depicts the simplified model TAT complexing Hg 2+ as a monodentate ligand at 2:1 stoichiometry. In the 2:1 complexes, each model ligand is coordinated to the metal cation via a single nitrogen bridging atom. The metal cation has a six-coordinate octahedral geometry. Each water molecule is bound to two amino groups (one from each of the two ligands).
The Gibbs energies of the complex formation reaction 2TAT + [Hg(H2O)6] 2+ → TAT@[Hg(H2O)4] 2+ @TAT + 2H2O are given in Table 2. The results for the Gibbs energy of the complex formation reaction with Hg 2+ indicate spontaneous and energy-favorable complex formation processes in the gas phase and in water environment. Table 2. Gibbs energies for the complex formation reactions with hydrated Hg 2+ and Mg 2+ cations in the gas phase (ΔG 1 ) and in water environment (ΔG 78 ), in kcal mol −1 . In addition, possible complexes formed as a result of the complexation of hexahydrated Mg 2+ cations, [Mg(H2O)6] 2+ , with TAT model system were modeled. The positive value (11.9 kcal mol −1 ) for the Gibbs energy of the complex formation reaction with hydrated magnesium ions in water indicates non-spontaneous and energy-unfavorable complex formation processes. Magnesium ion coordinates six water molecules more strongly than mixed ligand (TAT)/water molecules. The results obtained for the complexation of hydrated Hg 2+ and Mg 2+ cations with a simplified NI-DAT model are consistent with the experimental data and outline a difference in the complexation behavior of metal cations from the ligand.  Table 2. The results for the Gibbs energy of the complex formation reaction with Hg 2+ indicate spontaneous and energy-favorable complex formation processes in the gas phase and in water environment. In addition, possible complexes formed as a result of the complexation of hexahydrated Mg 2+ cations, [Mg(H 2 O) 6 ] 2+ , with TAT model system were modeled. The positive value (11.9 kcal mol −1 ) for the Gibbs energy of the complex formation reaction with hydrated magnesium ions in water indicates non-spontaneous and energy-unfavorable complex formation processes. Magnesium ion coordinates six water molecules more strongly than mixed ligand (TAT)/water molecules. The results obtained for the complexation of hydrated Hg 2+ and Mg 2+ cations with a simplified NI-DAT model are consistent with the experimental data and outline a difference in the complexation behavior of metal cations from the ligand.

Comparison between NI-DAT and Some of the Reported Sensors for Detecting Hg 2+
A comparison between the NI-DAT chemosensor and some of the reported sensors for detecting Hg 2+ is shown in Table 3. NI-DAT can be considered as a good candidate for detecting Hg 2+ due to its good solubility in water and low limit of detection.

Comparison between NI-DAT and Some of the Reported Sensors for Detecting Hg 2+
A comparison between the NI-DAT chemosensor and some of the reported sensors for detecting Hg 2+ is shown in Table 3. NI-DAT can be considered as a good candidate for detecting Hg 2+ due to its good solubility in water and low limit of detection.

Comparison between NI-DAT and Some of the Reported Sensors for Detecting Hg 2+
A comparison between the NI-DAT chemosensor and some of the reported sensors for detecting Hg 2+ is shown in Table 3. NI-DAT can be considered as a good candidate for detecting Hg 2+ due to its good solubility in water and low limit of detection.

Comparison between NI-DAT and Some of the Reported Sensors for Detecting Hg 2+
A comparison between the NI-DAT chemosensor and some of the reported sensors for detecting Hg 2+ is shown in Table 3. NI-DAT can be considered as a good candidate for detecting Hg 2+ due to its good solubility in water and low limit of detection.

Comparison between NI-DAT and Some of the Reported Sensors for Detecting Hg 2+
A comparison between the NI-DAT chemosensor and some of the reported sensors for detecting Hg 2+ is shown in Table 3. NI-DAT can be considered as a good candidate for detecting Hg 2+ due to its good solubility in water and low limit of detection.  [58] A comparison between the NI-DAT chemosensor and some of the reported sensors for detecting Hg 2+ is shown in Table 3. NI-DAT can be considered as a good candidate for detecting Hg 2+ due to its good solubility in water and low limit of detection.

Conclusions
The present work reporting on a simple appending of 2,4-diaminotriazine to 4amino-1,8-naphthalimide has demonstrated that NI-DAT is able to form in aqueous solution stable self-associated dimers by hydrogen bonding. This hydrogen bonding enhances the emission of the probe by blocking PET from triazine moiety to 1,8-naphthalimide. At higher pH, the hydrogen bonding does not occur due to NH2 deprotonation. Hence, the PET is retrieved, and the emission quenched. Moreover, Hg 2+ selectively and sensitively quenches the fluorescence emission of the probe solution due to its binding to the probe through NH2 groups that detach the probe molecules from their dimers. The stoichiometric ratio of NI-DAT/Hg 2+ complex was found to be 2:1, respectively. The coordination behavior of the studied sensor molecule towards metal ions is supplemented with computational (DFT) data and possible structures of Hg 2+ and Mg 2+ complexes in water solutions are proposed.

Conclusions
The present work reporting on a simple appending of 2,4-diaminotriazine to 4amino-1,8-naphthalimide has demonstrated that NI-DAT is able to form in aqueous solution stable self-associated dimers by hydrogen bonding. This hydrogen bonding enhances the emission of the probe by blocking PET from triazine moiety to 1,8-naphthalimide. At higher pH, the hydrogen bonding does not occur due to NH2 deprotonation. Hence, the PET is retrieved, and the emission quenched. Moreover, Hg 2+ selectively and sensitively quenches the fluorescence emission of the probe solution due to its binding to the probe through NH2 groups that detach the probe molecules from their dimers. The stoichiometric ratio of NI-DAT/Hg 2+ complex was found to be 2:1, respectively. The coordination behavior of the studied sensor molecule towards metal ions is supplemented with computational (DFT) data and possible structures of Hg 2+ and Mg 2+ complexes in water solutions are proposed.

Conclusions
The present work reporting on a simple appending of 2,4-diaminotriazine to 4amino-1,8-naphthalimide has demonstrated that NI-DAT is able to form in aqueous solution stable self-associated dimers by hydrogen bonding. This hydrogen bonding enhances the emission of the probe by blocking PET from triazine moiety to 1,8-naphthalimide. At higher pH, the hydrogen bonding does not occur due to NH2 deprotonation. Hence, the PET is retrieved, and the emission quenched. Moreover, Hg 2+ selectively and sensitively quenches the fluorescence emission of the probe solution due to its binding to the probe through NH2 groups that detach the probe molecules from their dimers. The stoichiometric ratio of NI-DAT/Hg 2+ complex was found to be 2:1, respectively. The coordination behavior of the studied sensor molecule towards metal ions is supplemented with computational (DFT) data and possible structures of Hg 2+ and Mg 2+ complexes in water solutions are proposed.

Conclusions
The present work reporting on a simple appending of 2,4-diaminotriazine to 4-amino-1,8-naphthalimide has demonstrated that NI-DAT is able to form in aqueous solution stable self-associated dimers by hydrogen bonding. This hydrogen bonding enhances the emission of the probe by blocking PET from triazine moiety to 1,8-naphthalimide. At higher pH, the hydrogen bonding does not occur due to NH 2 deprotonation. Hence, the PET is retrieved, and the emission quenched. Moreover, Hg 2+ selectively and sensitively quenches the fluorescence emission of the probe solution due to its binding to the probe through NH 2 groups that detach the probe molecules from their dimers. The stoichiometric ratio of NI-DAT/Hg 2+ complex was found to be 2:1, respectively. The coordination behavior of the studied sensor molecule towards metal ions is supplemented with computational (DFT) data and possible structures of Hg 2+ and Mg 2+ complexes in water solutions are proposed.