Naphthyridine derived colorimetric and fluorescent turn off sensors for Ni2+ in aqueous media

Highly selective and sensitive 2,7-naphthyridine based colorimetric and fluorescence “Turn Off” chemosensors (L1-L4) for detection of Ni2+ in aqueous media are reported. The receptors (L1-L4) showed a distinct color change from yellow to red by addition of Ni2+ with spectral changes in bands at 535–550 nm. The changes are reversible and pH independent. The detection limits for Ni2+ by (L1-L4) are in the range of 0.2–0.5 µM by UV–Visible data and 0.040–0.47 µM by fluorescence data, which is lower than the permissible value of Ni2+ (1.2 µM) in drinking water defined by EPA. The binding stoichiometries of L1-L4 for Ni2+ were found to be 2:1 through Job’s plot and ESI–MS analysis. Moreover the receptors can be used to quantify Ni2+ in real water samples. Formation of test strips by the dip-stick method increases the practical applicability of the Ni2+ test for “in-the-field” measurements. DFT calculations and AIM analyses supported the experimentally determined 2:1 stoichiometries of complexation. TD-DFT calculations were performed which showed slightly decreased FMO energy gaps due to ligand–metal charge transfer (LMCT).

www.nature.com/scientificreports/ Colorimetric sensors seem to be especially promising due to low cost, rapid detection and ease of use when compared to classical techniques as well as fluorescent sensors. Several organic molecules have been reported which showed colorimetric and fluorescent sensing of Ni 2+ but few were capable of parts per million level detection. The signals from absorption and emission changes of light by chromophores or fluorophores provide information about the mechanism of sensing of metal cation (Ni 2+ ) by electron transfer (ET), charge transfer (CT), photoinduced electron transfer (PET), excited state intramolecular proton transfer (ESIPT) mechanism 21,22 . In recent times, the Goswami group reported a quinoxaline 1 based ratiometric chemosensor for the naked eye detection of Ni 2+ in CH 3 CN media 23 . Sarkar et al. developed benzimidazole based 2 as a ratiometric and colorimetric chemosensor for Ni 2+ in DMSO 24 . Prabhu and co-workers developed a selective fluorescent turn on chemosensor for Ni 2+ based on pyrene-conjugated pyridine 13 2+ and Cu 2+ in aqueous methanol solution 26 (Fig. 1).
In continuation of our research work in the field of molecular recognition, we herein report benzo[c]pyrazolo [2,7]naphthyridine-5,6-diamine based efficient colorimetric and/or fluorescence off sensors that can detect Ni 2+ with sensitivity and selectivity in aqueous solutions. Many researchers have utilized amino group containing sensors like diaminonaphthalene 33 , 1,8-naphthyridine-amine 34 , 1,8-naphthyride-2-acetoamide 35 , diaminophenazine and 1,2-diaminoanthracene-dione 36 to sense different metal ions like Cu 2+ , Hg 2+ , Fe 3+ , Al 3+ . Here we report sensors in which vicinal amino groups are placed on the benzo[c]pyrazolo [2,7]naphthyridine skeleton. Diamines L1-L4 serve as sensors for Ni 2+ in aqueous solutions, utilizing the two amino groups as binding sites for the metal ion. The receptors detect the cation Ni 2+ by both Chelation Enhanced Fluorescence Quenching (CHEFQ and an instant change in color from yellow to red. The chemosensors L1-L4 could be used as practical sensors for quantitative determination of nickel at ppm level in real water samples and also used in test kits to observe the color changes for Ni 2+ by a dip stick method. Importantly the sensing potential of these derivatives can be tuned by the nature of the substituents R, i.e. electron donating (CH 3 ) and electron attracting (F, OCF 3 ), which affect the sensing properties. Quantum chemical computations (DFT) were performed to get detailed insight on the interaction between L4 and Ni 2+ and are in good agreement with experimental findings.

Results and discussion
The benzo[c]pyrazolo [2,7]naphthyridine based chemosensors were synthesized as shown in Scheme S1. The isatins 6 react with malononitrile 7 via Knoevenogel condensation to form arylidenes. The formed arylidenes are then reacted with 3-amino-5-methylpyrazole 8 to synthesize spiro-intermediates which then undergo basic hydrolysis, cyclization, decarboxylation and aromatization to form target naphthyridine receptors (L1-L4) 37 .    (Fig. 3a). Similar patterns of fluorescence spectral changes were observed for L1-L3 (Figs. S7-S9). The coordination behavior of L4 was examined via fluorescence titrations at room temperature. The sequential addition of Ni 2+ (0-12 µM) to receptor L4 caused a substantial decrease of emission intensity at 470 nm, indicating "turn-off" behavior of the receptor (Fig. 3b). The rigid and planar structure of L4 molecule makes it a highly fluorescent compound. However when chelation occurs between the NH 2 groups of receptor L4 and Ni 2+ , the amino groups lose their ability to donate electron density into the fluorophore due to ligand-metal charge transfer (LMCT) which causes the Chelation Enhancement Quenching Effect (CHEQ) and quenching of emission potential 39 (Fig. 3c). Similar coordination behavior was observed for L1-L3 (Figs. S10-S12).
Binding stoichiometry, association constant and detection limit. The binding stoichiometry of the complexes were further explored by Job's continuous variation method 40 by plotting mole fraction versus changes in absorption intensity at 535 nm for L1, 538 nm for L2, 550 nm for L3 and 537 nm for L4, respectively. The Job's plots ( Fig. S13) indicate maximum values at 0.7 corresponding to the formation of complexes with 2:1 stoichiometry between L1-L4 and Ni 2+ .
The association constants K a of receptors L1-L4 with Ni 2+ were determined by analysis of the UV-visible and fluorescence data using the Benesi-Hildebrand equation 41 ( Figure S14, S15) and are listed in Table S1. The association constant values are in range of those 10 3 -10 6 reported for Ni 2+ sensing chemosensors 42 . The trend of K a values shows that L4-Ni 2+ complex is stronger than the other receptor complexes. Moreover the association constants obtained by UV-visible data were on the same order of trend for as those obtained from fluorescence data i.e. L4-Ni 2+ > L3-Ni 2+ > L2-Ni 2+ > L1-Ni 2+ .
The detection limits of L1-L4 for Ni 2+ as colorimetric sensors were determined by naked eye, absorption and fluorescence spectral changes Table 2.
For naked eye detection, the receptor L4 showed a distinct color change at a minimum concentration of 1 × 10 −6 M for Ni 2+ (Fig. 4/S16). Moreover, the detection limits determined by absorption and fluorescence spectral changes on the basis of 3S B /S 43 for L4 and Ni 2+ were found to be 2.43 × 10 −7 M and 4.03 × 10 −8 M respectively. These values are lower than EPA drinking water guidelines (1.2 × 10 −6 M for Ni 2+ 44 ) and revealed that L4 is highly efficient in sensing Ni 2+ even at minute levels.
Proposed sensing mechanism. ESI-MS, IR and 1 (Fig. S17). www.nature.com/scientificreports/ FT-IR titrations were performed by using a Bruker Alpha FT-IR. Figure 5 shows a comparison of the IR spectra of L4 before and after the addition of Ni 2+ . The sharp peaks present at 3420, 3294 and 3109 cm −1 due to NH stretching frequencies in free receptor L4 were broadened by adding Ni 2+ , suggesting the involvement of NH 2 group in coordination with Ni 2+ to form complex 45,46 . To further investigate the interaction behavior between L4 and Ni 2+ , we carried out 1 HNMR titrations but these were not successful due to paramagnetic property of L4-Ni 2+ complex 42 . The plausible sensing mechanism of L4 for Ni 2+ is shown in Scheme 1.
Metal ion selectivity. An important feature of receptor L4 is its selectivity towards analytes. This was examined by competitive titration experiments (Fig. 6a). The intensity of the absorption band at 537 nm due to complex formation of L4-Ni 2+ is not disturbed at all in the presence of other metal ions (Al 3+ , Ca 2+ , Cd 2+ , Co 2+ , Cr 3+ , Cu 2+ , Fe 3+ , Hg 2+ , K + , Mg 2+ , Mn 2+ , Na + , Pb 2+ , Sr 2+ and Sn 2+ ). Thus the receptor L4 shows excellent binding affinity for Ni 2+ , which should hold in physiological samples where Cu 2+ , Co 2+ , Fe 3+ , Hg 2+ and Pb 2+ usually coexist with analyte. This distinct selectivity for Ni 2+ may be due to matching of the geometry of the receptor with the ionic radius of Ni 2+ 13 .
The UV-visible absorption spectra of the L4-Ni 2+ complex with various anions was recorded to check the stability of complex. No change in the absorption band at 537 nm was observed (Fig. 6b-c). This clearly indicates that the stability of the complex is unaffected by the presence of various anions.   www.nature.com/scientificreports/ pH effect study. In order to investigate the effect of pH on the absorption response of receptor L4 to Ni 2+ , a series of solutions with pH values ranging from (2.0 -12.0) were prepared (Fig. S18). At pH 2.0-3.0, the receptor L4 shows no substantial response to Ni 2+ in absorption spectroscopy. The absorption of the L4-Ni complex at 537 nm is maximum and constant in pH range 7.0 -8.0. Above pH 8.0, absorbance decreased gradually. The results suggest that biological and environmental applications at physiological pH should be feasible. The color of L4-Ni 2+ complex remained red between pH 4-11, which indicated that Ni 2+ could be clearly detected over a wide range of pH 4-11.
The solution's color changed from red to light yellow (original color of L4). Upon addition of Ni 2+ again the absorbance at 537 nm was recorded. The absorption changes in spectral bands were reversible even after several cycles with alternative sequential addition of Ni 2+ and EDTA. These results indicate that L4 could be recyclable as an off-on-off receptor by the interaction of EDTA with L4-Ni 2+ (Scheme 1). Such regeneration and reversibility could be valuable for the fabrication of Ni 2+ sensors. The ESI-MS of the synthesized L4-Ni 2+ complex (Fig. 8a) showed the molecular ion peak at m/z 685.41 which matched very well with the calculated molar mass of [2L4 + Ni + Cl 2 + H]. Furthermore, SEM analysis was carried out to obtain a better understanding of morphological difference before and after the addition of Ni 2+ to L4 receptor (Fig. 8b-c). SEM images of receptor L4 show dense sprinkled elliptical shapes which are transformed into Scheme 1. The plausible sensing mechanism of Ni 2+ and with L4. Practical application. In order to investigate the potential use of receptor L4 in real water samples, a calibration curve was drawn, which showed a good linear relationship (R2 = 0.9996, n = 3) between the absorbance of the L4-Ni 2+ complex and the Ni 2+ concentration (0-5 µM) at 537 nm (Fig. S19). The receptor L4 was used for the estimation of Ni 2+ in drinking water, tap water and industrial waste water samples (Table 3). All water samples were analyzed in triplicate with good recoveries and RSD values. The results indicate that receptor L4 is highly specific and sensitive for Ni 2+ estimations in environmental samples.
To explore another application of receptor L4, test kits were prepared by immersing filter paper in receptor L4 (1 × 10 −3 M, HEPES buffer, pH = 7.4) and then air drying to investigate a "dip-stick" method for detection of Ni 2+ . When the prepared test strips were immersed into aqueous solutions of Ni 2+ with different concentrations, clear color changes from yellow to red were observed (Fig. 9). The results showed that discernible concentrations of Ni 2+ can be as low as 1 × 10 −5 M. The "dip-stick" method did not require any additional equipment for detection of Ni 2+ and should be highy attractive for "in-the-field" measurements.
Theoretical details on structural aspects of L4-Ni 2+ . Computational details. In order to evaluate the interaction between L4 and Ni 2+ theoretical calculations on the L4-NiCl 2 complex were performed. Geometry optimization calculations were performed for the free species L4 and NiCl 2 , and complexes L4-NiCl 2 and 2L4-Ni 2 Cl 2 (with two L4 molecules). The calculations employed the DFT method by applying the M06L 48 functional and def2-SVP 49 Fig. 10. The relative energies are shown in Table S2 and demonstrate that the complexation between L4 and NiCl 2 is spontaneous. The energies for complexation with one or two L4 molecules showed that the complex with two L4 molecules is more stable than with  FMO and GRD analysis. The energies for frontier molecular orbitals of free L4 and the 2L4-NiCl 2 complex are shown in Table S3. It was verified through the FMO analysis species that the complexation between L4 and NiCl 2 causes a slight decrease of the frontier molecular orbitals and the 2L4-NiCl 2 complex showed lower HOMO/ LUMO energy gaps, particularly for HOMO-1/LUMO + 1 and HOMO-2/LUMO + 2 which are related to the interaction with nickel. The surfaces for frontier molecular orbitals of L4 and 2L4-NiCl 2 are shown in Fig. 11.    Table S4 and their respective molecular graphs are shown in Fig. 12a-b respectively. The characterization of strengthening of the interaction between L4 and NiCl 2 was performed through the topological analysis of the electronic density. Therefore, the presence of an interaction between atoms was featured by the presence of a Bond Path (BP) between two attractor (atoms in a molecule) and their characteristics such covalence and strength were determined though the AIM properties in their Bond Critical Points (BCPs). The main AIM properties that were evaluated in this work were the electronic density, ρ(r), Laplacian of density, ∇ 2 ρ(r), and density of potential energy, V(r). Thus, the presence of the interaction between NH 2 and Ni was revealed by the presence of a BP between N and Ni atoms. The AIM molecular graphs with their respective labels are shown in Fig. 12a-b. The BCPs a and b are related to the Ni-N bonds in the complexes L4-NiCl 2 and 2L4-NiCl 2 . The complexes' formation showed BCPs a and b with positive value of ∇ 2 ρ(r) with can indicate a non-covalent interaction or ionic bond. When the values of electronic density, ρ(r), were analyzed for these BCPs was verified that the BCPs a for both complexes showed high negative values of density of potential energy, V(r), which indicates that the bonds related to BCPs a have the characteristic of ionic bonds. The BCP b L4-NiCl 2 also showed an ionic characteristic. However, the 2L4-NiCl 2 complex showed the weakening of one of their Ni-NH 2 interaction to formation of another Ni-NH 2 interaction with the second L4 molecule. Thus, the 2L4-NiCl 2 showed two ionic bonds between N-Ni (BCPs a and b*) with ρ(r) = 0.0922 a.u. and ρ(r) = 0.0901 a.u., respectively, and two non-covalent interactions between N-Ni (BCPs b and a*), that were revealed by the presence of ∇ 2 ρ(r) > 0, though, with low values of ρ(r) (ρ(r) = 0.0202 a.u. and ρ(r) = 0.0181 a.u., respectively) and low negative values of V(r). While the L4-NiCl 2 complex showed two ionic bonds with the same L4 with BCPs a and b, the 2L4-NiCl 2 complex showed two ionic bonds and several noncovalent interactions (BCPs c, d, c* and d*) between L4 and NiCl 2 , Thus, such non-covalent interactions should be related with the stabilization of 2L4-NiCl 2 complex in relation to L4-NiCl 2 . Beyond the interaction between N atoms from L4 and Ni 2+ was verified the presence of NH···Cl interaction (BCPs d and d* with ρ(r) = 0.0165 a.u. and ρ(r) = 0.0133, respectively) in the 2L4-NiCl 2 complex which should increase the NH stretching frequencies shift in relation to free L4 which agree with experimental results.

Conclusion
In summary, we have successfully characterized the photophysical properties of benzo[c]pyrazolo [2,7]naphthyridines (L1-L4) which were prepared by a green synthetic route. The receptors (L1-L4) allow selective and sensitive sensing of Ni 2+ in aqueous media over a wide range of pH (4-11) even in the presence of competitive ions i.e. Fe 3+ , Cu 2+ , Co 2+ . A unique colorimetric response to Ni 2+ is observed (yellow to red) through coordination of receptor and Ni 2+ complex could be recyclable through treatment with EDTA. The detection limit of Ni 2+ was found to be in range of 0.2-0.5 µM for (L1-L4) by UV-Visible data, and 0.040-0.47 µM by fluorescence data, which is lower than the permissible value of Ni 2+ (1.2 µM) in drinking water and opens a potential application of the receptors in recognizing Ni 2+ in environment. The binding stoichiometries of (L1-L4) with Ni 2+ were found to be 2:1 through Job's plot and ESI-MS analysis. The fluorescence properties of the receptors were evaluated with fluorescence quenching by coordination with Ni 2+ . As a practical application, the most efficient receptor L4 could be used to quantify and detect Ni 2+ in real water samples and also applied for fabrication of test kits using the "dip-stick" method. To the best of our knowledge, receptors (L1-L4) are the first reported multifunctional, naked eye chemosensors for sensing of Ni 2+ in aqueous solutions. Theoretical calculations demonstrate that the stoichiometry of complexation between L4 and NiCl 2 should be 2:1 as experimentally verified. The non-covalent interactions between L4 and NiCl 2 in the formed complex, particularly involving the -NH 2 groups can broaden the NH stretching frequencies, which is consistent with the experimental results. The FMO demonstrate the complexation with Ni decreases the frontier molecular orbitals and their respective energy gaps.   Water sample collection and Ni 2+ determination. The drinking water samples, tap water and industrial waste water samples were collected, preserved and stored in plastic containers for Ni 2+ analysis. Industrial waste water samples were filtered prior to analysis. Each sample was analyzed in triplicate using receptor L4 and ICP-OES as standard method (Table 3). Spiking and recovery method was used in order to validate chemosensing performance of our newly developed sensor L4. UV-visible spectral measurement of water samples containing Ni 2+ was carried out by adding 0.5 mL of receptor L4 to 2.5 mL of sample solutions and pH of solution was maintained at 7.4 using HEPES buffer. The solutions were allowed to stand for 10 min at room temperature and absorption measurements were taken at 537 nm. Filtered water samples were directly used for ICP-OES analysis.
Colorimetric test strips. The test kits were prepared by immersing filter paper strips in to receptor L4 solution 1 × 10 −3 M (DMSO-H 2 O (v/v 1:2) using HEPES buffer of pH = 7.4) and then dried in air. Then the pure water solution with different Ni 2+ concentrations were prepared and the prepared test strips were immersed in water samples and color change from yellow to red was observed.
Magnetic moments. Magnetic moments of the prepared solid complex were estimated by magnetic susceptibility balance Sherwood (Auto, 2005) at room temperature.