Propylimidazole Functionalized Coumarin Derivative as Dual Responsive Fluorescent Chemoprobe for Picric Acid and Fe3+ Recognition: DFT and Natural Spring Water Applications

A propylimidazole functionalized coumarin derivative (IPC) was fabricated for the first time and applied as a dual responsive fluorescent chemoprobe for sensitive and selective recognition of picric acid (PA) and Fe3+. Strong fluorescence quenching phenomena of the IPC were observed in H2O/ACN (5/95, v/v) medium (λem=408 nm) upon the additions of Fe3+or PA. The fabricated dual responsive IPC offered good selectivity and sensitivity with the low limit of detection values (0.92 µM for PA and 0.22 µM for Fe3+) lower than the acceptable amounts of Fe3+ and PA by the international official authorities. The validation study for the chemoprobe IPC for PA and Fe3+ was also performed. The interaction phenomena of IPC with PA and Fe3+ based on the findings of a range of experiments were considered and DFT computations were done to verify their recognition mechanisms. The sensing phenomena of IPC towards PA (1:1) and Fe3+ (3:1) were confirmed by the MALDI TOF–MS, FT–IR, 1H–NMR titration and Job's methods. Furthermore, the compound IPC was effectively applied as a fluorescent sensor for Fe3+ and PA detection in real natural spring water samples.


Introduction
Fluorescent sensor technology has attracted extensive interest to researchers for trace analyte detection in recent years due to its advantages; such as operational/instrumental simplicity, portability, cost efficiency, excellent sensitivity/selectivity, ease of visual detection and fast signal processing [1][2][3][4]. Fluorescent organic substances have been extensively utilized for the design of fluorescent sensors/probes to recognize a wide range of environmentally and biologically important analytes like heavy metal ions/ cations, anions, thiols, amino acids and nitroaromatic compounds [5]. Out of different nitroaromatic compounds, picric acid (PA; 2,4,6-trinitrophenol, TNP) is commonly utilized in blasting, manufacturing and chemical industries due to its better-quality explosiveness [6][7][8][9]. PA is known as a more influential secondary explosive substance than its structural related compounds like p-dinitrobenzene (p-DNB), 2,4-dinitrotoluene (DNT) and trinitrotoluene (TNT) and is also defined as a threat to public security and human health. Due to its extremely explosive nature, it could easily be employed by terrorists for illegal activities [10]. Moreover, it could induce fatal diseases like anemia, cancer, faintness, acute scratchiness and allergic reactions of the skin, eye irritation and damage to the functions of kidney and liver and does not degrade easily in nature [1,11,12]. On the other hand, highly sensitive and selective recognition of heavy metal ions in trace level with fluorescent organic compounds have been of great interest because of the fact that they have pivotal roles in various environmental and biological processes. For instance, Fe 3+ has vital roles in many living systems like electron transfer, oxygen uptake and transportation. Its superabundance (hyperferremia) in the body would injure bio-systems and induces various failures of limbs such as the heart, liver and kidney; as a result of producing reactive oxygen species. In the meantime, its absence (hypoferremia) could result in a number of critical diseases such as diabetes, insomnia, anemia diseases, and it induces iron homeostasis that is an important matter for the progression of Parkinson's, Huntington's and Alzheimer's diseases [13][14][15][16]. Thus; there is a great need to develop new analytical methodologies for the selective and sensitive recognition of PA and Fe 3+ .
To date, several analytical methods have been developed for the recognition of PA and Fe 3+ , for instance, inductively coupled plasma-optical emission spectrometry (IPC-OES), ion chromatography (IC), atomic absorption spectroscopy (AAS), high-pressure liquid chromatography (HPLC), etc. These methods have some drawbacks, such as they have need sophisticated instruments, specialized personnel, laborious sample pre-treatment procedures and the usage of costly chemical regents. To keep away from these drawbacks, fluorescent sensing systems have been recently developed due to their superior advantages mentioned above [8,[17][18][19]. However, the literatures based upon fabricating dual responsive fluorescent chemoprobes for the recognition of both PA and Fe 3+ with high selectivity and sensitivity, are still rare [11,[20][21][22]. Therefore, the need is great for more fluorescent chemoprobes able to detect the PA and Fe 3+ in real-time and reliably.
Herein, a new propylimidazole functionalized coumarin compound, N-(3-(1H-imidazol-1-yl)propyl) -2-oxo-2H-chromene-3-carboxamide (IPC) was prepared and utilized as a dual responsive chemoprobe for the detection of PA and Fe 3+ with high selectivity by fluorescence quenching behaviors. It is important to point out that fluorescence sensors capable of dual determination of PA and Fe 3+ are scarce. The optical properties and responses of IPC towards PA and Fe 3+ in H 2 O/ACN (5/95, v/v) media were determined. DFT computations were done to confirm the electronic and geometrical structural characteristics of IPC and its complexes. Moreover, IPC was used for the sensitive detection of PA and Fe 3+ in real natural spring water samples. The obtained results demonstrated that IPC is a good candidate for the determination of PA and Fe 3+ . Furthermore, the binding mechanisms were performed as well.

Chemicals and Instrumentations
All chemical reagents and solvents were bought from commercial suppliers (Sigma Aldrich, Thermo Fisher Scientific and VWR International Chemicals) and used directly. Merck Milli-Q ® 7003/05/10/15 water purification machine in our laboratory was used to obtain ultra high-quality water, 18

Synthesis Protocol of IPC
Briefly, the compound A was synthesized according to the previous literature [23]. Then, the compound A (107 mg, 0.517 mmol) was added to a solution of 1-(3-aminopropyl) imidazole (64.61 mg, 0.517 mmol) in acetonitrile (MeCN, 10.0 mL). The reaction mixture was stirred at rt for 2h, and then the precipitated occurred. The obtained product was filtered by washing several times with MeCN (Scheme S1

Fluorescence Studies of IPC
For the fluorescence sensing studies, 1×10 -2 M stock solution of the chemoprobe IPC was prepared and then it was diluted to 50 µM in H 2 O/ACN (5/95, v/v) media. The stock solutions of nitroaromatic explosives (NBC, DNBA, BNBA, DNT, CDNB, DBNB, NB and PA) and metal ions (K + , Ag + , Cu 2+ , Cd 2+ , Zn 2+ , Fe 2+ , Hg 2+ , Al 3+ and Fe 3+ ) were prepared as 1×10 -2 M. The spectra of PA and Fe 3+ were gathered from the emission region of 340-800 nm (λ ex =330 nm, λ em =408 nm, slit widths 10.0 and 20.0 nm). Titration plots were constructed by plotting the emission intensities at 408 nm. For the selectivity studies, the same equivalents of 7 kinds of nitroaromatic explosives (50.0 eqv.) and 8 kinds of metal ions (50.0 eqv.) were employed by the chemoprobe IPC solution. The validation study for the chemoprobe IPC for PA and Fe 3+ was performed and all measurements were done at least three times.

Computational Studies
The molecular structures and HOMO/LUMO levels of the IPC and its complexes (IPC-PA and IPC-Fe 3+ ) were gained with the gas phase, by using DFT computations through the employ of Gaussian-09 [B3LYP/LANL2DZ (for PA and Fe 3+ ) and 6-31G (d,p) (for C, H, N, O)] and GaussView-5.0.8 software packages (Gaussian, Inc., Wallingford CT, UK).

Natural Spring Water Analysis
Natural spring water samples were collected from local water resources in Konya City. The samples were analyzed without sample pre-treatment; just they were centrifugated at 10.000 rpm for 3 min. 3 mL of the IPC sensing solution (50 μmol.L -1 ) was transferred into quartz-cuvette, and then 15.0 μL sample was added into this solution. Afterward, the standard addition method was applied with the addition 15 µL of PA or Fe 3+ (0.10 and 0.20 μmol.L -1 ). After the adding of PA or Fe 3+ into the solution of chemoprobe IPC, their fluorescence intensities were recorded. The recovery and the standard deviation (RSD) values were calculated for the analytical evaluation. All the fluorescence measurements were performed at least three times and the statistical calculations were done.

Fabrication and Characterization of IPC
IPC was obtained as a white solid and well-characterized using FT-IR, 1 H-NMR and MALDI TOF-MS techniques ( Fig. S1-S3). 1 H NMR spectrum shows α-hydrogen of coumarin and aliphatic peaks of propylimidazole groups clearly. In addition, 298.526 (m/z) was observed which corresponded to IPC (chemical formula: C 16

Fluorescence Sensing Studies of IPC
The effect of different solvents on the emission intensity of IPC was studied. For this purpose, the emission intensities were obtained upon excitation at 330 nm for the chemoprobe IPC prepared in acetonitrile (ACN), methanol (MeOH), ethanol (EtOH), dimethylformamide (DMF), water (H 2 O), and dimethylsulfoxide (DMSO) (Fig. S4). The maximum fluorescence of IPC was achieved at 408 nm as it was prepared in ACN. The influence of water percentage used in ACN on the fluorescence response was also studied for the IPC, IPC-PA and IPC-Fe 3+ . The differentiation of emission intensities between the chemoprobe IPC and its complexes were the most appropriate when the both volume ratio of ACN/H 2 O was 95/5. Thus, the solvent media of H 2 O/ ACN (5/95, v/v) was employed for the further fluorescence measurements under this study.
To reveal the sensing ability of chemoprobe IPC, its response towards a pool of analytes including different nitroaromatic explosives (NBC, DNBA, BNBA, DNT, CDNB, DBNB, NB and PA) and metal ions (K + , Ag + , Cu 2+ , Cd 2+ , Zn 2+ , Fe 2+ , Hg 2+ , Al 3+ (Fig. 1b) and Fe 3+ (0-20.0 equiv.) (Fig. 2b). Upon the adding of PA into a solution of chemoprobe IPC (50 μM), the emission intensity progressively quenched due to the π-π stacking and deprotonation phenomena [24] and it arrived at a minimum intensity level after the adding of 20.0 equiv. of PA. On the contrary, after the addition of different amounts of Fe 3+ , the emission intensity of IPC was quenched because of the paramagnetic quenching effect with the transferring of energy and/or electron, which is known as ''ligand-metal charge transfer mechanism (LMCT)''. The fluorescence intensity reached a stable value after the concentration of Fe 3+ reached 20.0 equiv. (Fig. 2b) [16,25,26].
From the findings of fluorescence titration studies, the detection limits (LOD) of chemoprobe IPC for PA and Fe 3+ were computed by fluorescence alterations on the basis of 3σ/k equation; where ''σ'' is the deviation of the blank Scheme 1 Schematic illustration of chemoprobe IPC emission intensity and ''k'' is the slope of the linear calibration graph. They were found to be 0.92 µM for PA and 0.22 µM for Fe 3+ , which is less than the daily uptake level of iron ion (2 mg.L -1 ) suggested by the WHO [16]. These titration data were also employed to compute the association constants (log K) of IPC-PA and IPC-Fe 3+ complexes, and they were found to be 6.72×10 2 M -1 for PA and 3.62 M -1/3 for Fe 3+ on the basis of Benesi-Hildebrand equation (Fig. 3) [27][28][29]. Thus, these findings recommended that the chemoprobe IPC has a great potential of recognition quantitatively unidentified concentrations of PA or Fe 3+ and could be utilized for the sensitive recognition of PA and Fe 3+ in H 2 O/ACN (5/95, v/v) media. The sensing properties of the chemoprobe IPC are also comparable to those of some dual-responsive fluorescent probes for PA and Fe 3+ , which reveals that our chemoprobe system has significant improvements ( Table 1).
The validation parameters were calculated using the "Youden approach" and "Fisher and Dixon tests". All parameters were determined during the assessment of the fluorescence method are given in Table S1. The precision of the fluorescence method was estimated using "repeatability" and "intermediate precision". The repeatability study (Table S2) was performed for two concentrations (2 × 10 −5 and 1 × 10 −5 M) in multiple time periods (16 h, 20 h and 24 h) (N=10). Using the obtained data, the "RSD" and the "Hor. Rat-ratio" (RSD/RSD Horwitz) were calculated (Eqs. (1) and (2)). The ratio of HorRat is smaller than 1. In addition, the Dixon test was carried out to obtain Q values (Eq. (3)). When we compare Q calculated values with Q critical (Q critical = 0.96) value, it is obviously seen in Table 2 that the Q calculated values are lower than Q critical . The results indicate that the variation of measurement time and concentration were not statistically notable.  Table S3, the HorRat ratios were determined <1. Besides, the Fisher test (F-test) was performed using Eq. (4). It is clearly seen that the obtained F calculated values are lower than the F critical value (F calculated <F critical ). Thus, it has been determined that the proposed method is reproducible.
In addition, the robustness study of the fluorescence method was performed using the Youden approach. The diversifying conditions of this study were presented in Table S4. All conditions were determined with different factorial combinations (Ci) ( Table S5). Pareto's graphs for PA and Fe 3+ (Fig. S5) were achieved using the Eq. (5). All conditions including storage time, type of water, before the analysis, solvent sytem, temperature of analysis (°C), nitrogen atmosphere, pH were notable, verifying the robustness of the fluorescence method.
To use the chemoprobe IPC as a selective fluorescent sensor for PA and Fe 3+ , the impacts of competing for nitroaromatic explosives (NBC, DNBA, BNBA, DNT, CDNB, DBNB, NB and PA) and metal ions (K + , Ag + , Cu 2+ , Cd 2+ , Zn 2+ , Fe 2+ , Hg 2+ , Al 3+ and Fe 3+ ) have been also studied. As seen from Fig. 4a, the emission quenching was observed for the mixtures of PA with other nitroaromatic explosives was similar to that stimulated by PA alone; therefore, the presence of competing explosives could not make interference with the recognition of PA. Likewise, the Fe 3+ sensing system in H 2 O/ACN (5/95, v/v) media was not influenced by a pool of metal ions (Fig. 4b).The competition studies showed that the fluorescence response of the chemoprobe (5) For the parameter 3 = [(C 2 + C 4 + C 6 + C 8 )∕4  IPC toward PA or Fe 3+ was not interfered with the studied competing analytes; therefore, IPC could be used as a ''turn-off'' fluorescent chemoprobe for the recognitions of both PA and Fe 3+ .
The response time is another critical parameter for a newly designed chemoprobe in real applications. As seen from Fig. S6, the fluorescence quenching for PA and Fe 3+ occurred only within just 30 seconds, and their fluorescence intensities were reached equilibrium at the same response time. Thus, a response time was decided on 30 seconds for the following experiments and the chemoprobe IPC depicted an excellent response time towards PA and Fe 3+ respecting the formerly developed a lot of fluorogenic dual chemoprobes [11,[20][21][22].

Binding Mechanisms of IPC Towards PA and Fe 3+
To understand the binding stoichiometry of the complexes between IPC and PA/Fe 3+ , the MALDI TOF-MS, FT-IR and Job's plot methods were applied. The stoichiometric ratios of chemoprobe IPC toward Fe 3+ were found to be 3:1.
To determine the binding stoichiometry of IPC-Fe 3+ complex, Job's plot study was performed (Fig. S7). The fluorescence intensities at 408 nm are graphed against the molar fractions of the chemoprobe IPC. The maximum spot was monitored at a mole fraction of 0.75 for Fe 3+ and this result has shown that it was 3:1 stoichiometry of the binding mode of IPC-Fe 3+ . In addition, it is clearly observed that the N-H peak of the chemosensor IPC, which was present at 3259  S9) (Scheme 2). The stoichiometric ratios of chemoprobe IPC towards PA were found to be 1:1 (Fig. S7). To understand the binding stoichiometry of IPC-Fe 3+ complex, Job's plot study was performed. The maximum spot was monitored at a mole fraction of 0.5 for PA and this result has shown that it was 1:1 stoichiometry of the binding mode of IPC-PA. Also, 1 H-NMR measurements were performed to obtain an insight into the interaction mechanism between IPC and PA. As depicted in Fig. S10, the slight peak was shifted up-field in the presence of PA. Therefore, the formation of π-π stacking and the intermolecular H-bonds between IPC and PA caused the quenching of fluorescence intensity [6,24]. (Scheme 2)

Theoretical Computations
Theoretical computations of the chemoprobe IPC and its complexes have been done to obtain their HOMO-LUMO energy levels. The orbital energies were computed using Gaussian-09 [B3LYP/6-31G (d,p) (for IPC and IPC-PA) and LANL2DZ (for IPC-Fe 3+ )] and GaussView-5.0.8 software packages. To confirm the suggested interaction pathway of the chemoprobe IPC toward PA, a DFT calculation was performed based on the reported study [6]. The optimal structure of IPC-PA displayed intermolecular hydrogen bonding between IPC and PA due to π-π interactions (Fig. 5). As depicted in Fig. 5, the computed energy gaps between HOMO and LUMO orbitals of IPC, IPC-PA and IPC-Fe 3+ were found as 3.32, 2.19 and 0.7 eV, respectively, showing good interactions. Therefore, these findings revealed that the interaction of IPC towards PA and Fe 3+ stabilizes the systems as evident from the lower HOMO-LUMO energy gaps of the complexes compared to IPC (Fig. 5).

Natural Spring Water Application
In the last part of the study, the applicability of the developed fluorescence chemoprobe system for the recognition of PA or Fe 3+ in natural spring water samples was tested. For real water sample measurements, the samples were spiked with known concentrations (0.10 and 0.20 mol.L -1 ) of PA or Fe 3+ , according to the standard addition method. After the adding of PA or Fe 3+ into the solution of chemoprobe IPC, their fluorescence intensities were recorded. The findings were given in Table 2, and the recovery values of PA and Fe 3+ were between 90.89 and 104.71% with the lower relative standard deviation (RSD) values. Thus, these findings revealed that the chemoprobe IPC system could specifically and accurately labor to recognize PA or Fe 3+ in natural spring waters.

Conclusion
In summarize, a new of dual-responsive fluorescence chemoprobe (IPC) based on propylimidazole functionalized coumarin structure for the recognition of PA and Fe 3+ has been successfully developed. The IPC revealed ''on-off'' fluorescence responses towards PA and Fe 3+ at 408 nm within only 30 seconds in H 2 O/ACN (5/95, v/v) media.
The LOD values for PA and Fe 3+ were found to be 0.92 µM and 0.22 µM, respectively, which are satisfactorily low to permit the recognition of these analytes in realistic applications. The stoichiometry of the complexes [IPC-PA (1:1) and IPC-Fe 3+ (3:1)] was identified by MALDI TOF-MS, FT-IR, 1 H-NMR titration and Job's plot experiments. The binding mechanism of IPC toward PA or Fe 3+ was also supported by the DFT computation study. Furthermore, the chemoprobe IPC was employed for detecting PA or Fe 3+ in natural spring waters with good recovery values. Therefore, these promising findings will make a great contribution to researchers studying nitroaromatic explosives and metal ions in different systems.