Synthesis and Sensing Applications of Fluorescent 3-Cinnamoyl Coumarins

We have synthesized two novel fluorescent 3-(4-diethylaminocinnamoyl) coumarins that exhibit fluorescence quenching upon exposure to a nerve agent simulant, diethylchlorophosphate (DCP), providing a basis for rapid and sensitive DCP chemosensing. Furthermore, these coumarin derivatives display two-photon fluorescence upon illumination with near-infrared laser pulses and their two-photon (TP) absorption cross-section was evaluated. The potential for TP bio-imaging of these compounds was investigated by their cellular uptake in HeLa cells by TP confocal microscopy.


Introduction
The extreme toxicity of organophosphorus (OP)-containing nerve agents such as Sarin, Soman, and Tabun poses a serious threat of chemical attack. Most of these nerve agents have chemical structures similar to those of insecticides and can irreversibly react with the enzyme acetylcholinesterase (a neurotransmitter), inhibiting its control over the central nervous system [1,2]. The use of such chemicals by terrorist groups in the past underscores the need to detect these odorless and colorless compounds. A variety of detection methods for nerve agents have been developed, including mass spectrometry-based techniques, enzyme-based sensors, colorimetric probes, and fluorometric methods [3][4][5][6][7][8][9][10]. However many of these systems suffer from their particular limitations, such as slow response times, operational complexity, and high cost, etc. Fluorescence-based detection methods are versatile due to their high sensitivity and much lower cost. Fluorescence-based sensors take advantage of the high electrophilicity of these nerve agents by reacting with a nucleophile embedded in a fluorophore leading to the change of fluorescence properties. The nerve agent simulants diethylcyanophosphate (DCNP), diethylchlorophosphate (DCP), and dimethylmethylphosphonate (DMMP) are normally used for research purposes as they offer similar reactivity as real nerve agents but lack the severe toxicity [11,12].
Coumarin-based derivatives have been widely used as fluorophores for constructing sensory systems for pH, metal cations, anions, and gases due to their desirable photophysical properties with large Stokes shifts and emissions in the visible spectral range [13][14][15][16][17][18][19]. Furthermore, these coumarin derivatives are also known to exhibit two-photon absorption (TPA) phenomena.

Quantum Efficiency Measurement
For quantum efficiency measurements, the compounds were dissolved in DMSO (1 mg/mL) and diluted further to get the required concentration for optical studies. Commercially available rhodamine (RhB) dye was used as reference. Solutions with similar optical density were prepared. The relative quantum efficiency was measured using the following equation: where Φ c = quantum efficiency of the 3-cinnamoyl coumarin, Φ r = quantum efficiency of the reference, F c = integrated fluorescence intensity of the 3-cinnamoyl coumarin, F r = integrated fluorescence intensity of the reference, n c = refractive index of the 3-cinnamoyl coumarin, n r = refractive index of the reference.

TPF Measurement
A mode-locked Ti:Sapphire laser (Quantronix, East Setauket, NY, USA) operated at 800 nm with a pulse width of 100 femto-second and a repetition rate of 1 kHz was used as the excitation source for the TPF study. The near-infrared (NIR) laser pulses were passed through the sample cuvette and the fluorescence signal was collected at 90˝with respect to the laser beam using a CCD and a monochromator. A short-pass optical filter at 750 nm was placed in front of the monochromator to eliminate the excitation radiation. A short focal length (4.5 cm) lens was employed to collect the TPF signal.

TPA Cross Section Measurement
The TPA cross-sections were determined using RhB in DMSO as a reference. The TPA cross-sections were calculated using the following formula: where Fs are the integrated TPF intensities; Cs are the concentrations; subscript c and r stand for the 3-cinnamoyl coumarin and the reference molecule, respectively [28].

Stern-Volmer Constant (Ksv), Binding Constant (K) and Number of Binding Sites (n) Measurements
The Stern-Volmer constant was calculated from the fluorescence titration data of compounds 7 or 8 (0.02 mM, 3 mL), with increasing concentration of DCP in chloroform and by plotting the relative fluorescence intensity (I o /I) against the [DCP] (Figure S13) using the following equation: The plot between log[(I o´I )/I] and log [DCP], gives rise to the binding constant (K) and number of binding sites (n) in compound 7 or 8 for DCP ( Figure S14)     Compound 13 was obtained as a yellow solid in 90% yield from the reaction of (E)-6-[3-(4-diethyl-aminophenyl)acryloyl]-7-hydroxy-4-methyl-2H-chromen-2-one (12)

One-Photon and TP Activity of 3-Cinnamoylcoumarins
UV-Vis absorption and fluorescence spectra of 3-(4-diethylaminocinnamoyl) coumarins 7 and 8 were recorded by dissolving the compounds in DMSO. These compounds exhibit strong absorption maxima at 521 and 523 nm, respectively. Upon exciting at their absorption maxima range, they emit strong fluorescence with λmax at 608 and 610 nm, respectively ( Figure S11). Stokes shift of 87 nm was observed for these compounds. The solvatochromism effect of these compounds was also studied and shown in Figure S12, Table S1. The quantum yields (Ф) of the compounds were measured using RhB as a reference in DMSO having quantum efficiency of 96% [31]. The quantum yields of these compounds were determined and summarized in Table 1. The TPF spectra of these compounds were recorded by exciting them with the NIR femtosecond laser pulses in dilute DMSO solution. Figure 1 shows the intensity dependent TPF spectra for compounds 7 and 8. In both of the cases, an increase in the TPF intensity with the increase of the exciting laser power was observed. The quadratic behavior of the TPF was confirmed by plotting log[peak TPF intensity] vs log[laser power] as shown in Figure 2. A slope close to two in the plot was determined, indicating that the fluorescence from the 3-(4-diethylaminocinnamoyl) coumarins 7 or 8 is indeed a TP phenomenon. The TPA cross-sections of 7 and 8 were then calculated, using the RhB as a reference having TPA cross-section of 28 GM at 800 nm [32]. The TPA cross-sections of the compounds studied (Table 1) were found to be comparable to the dyes used as biomarkers [33].

One-Photon and TP Activity of 3-Cinnamoylcoumarins
UV-Vis absorption and fluorescence spectra of 3-(4-diethylaminocinnamoyl) coumarins 7 and 8 were recorded by dissolving the compounds in DMSO. These compounds exhibit strong absorption maxima at 521 and 523 nm, respectively. Upon exciting at their absorption maxima range, they emit strong fluorescence with λ max at 608 and 610 nm, respectively ( Figure S11). Stokes shift of 87 nm was observed for these compounds. The solvatochromism effect of these compounds was also studied and shown in Figure S12, Table S1. The quantum yields (Φ) of the compounds were measured using RhB as a reference in DMSO having quantum efficiency of 96% [31]. The quantum yields of these compounds were determined and summarized in Table 1. The TPF spectra of these compounds were recorded by exciting them with the NIR femtosecond laser pulses in dilute DMSO solution. Figure 1 shows the intensity dependent TPF spectra for compounds 7 and 8. In both of the cases, an increase in the TPF intensity with the increase of the exciting laser power was observed. The quadratic behavior of the TPF was confirmed by plotting log[peak TPF intensity] vs log[laser power] as shown in Figure 2. A slope close to two in the plot was determined, indicating that the fluorescence from the 3-(4-diethylaminocinnamoyl) coumarins 7 or 8 is indeed a TP phenomenon. The TPA cross-sections of 7 and 8 were then calculated, using the RhB as a reference having TPA cross-section of 28 GM at 800 nm [32]. The TPA cross-sections of the compounds studied (Table 1) were found to be comparable to the dyes used as biomarkers [33].    A correlation of quantum efficiency and TPA cross-sections with the chemical structures of the coumarins was observed. An enhancement in quantum efficiency and TPA cross-section was observed on substituting the C-8 position of 3-(4-diethylaminocinnamoyl) coumarins with a methoxy group. A similar effect of electron donating group on δ value of same conjugated system was observed by Li et al. [24]. Higher TPA cross-section and quantum yield for coumarin derivative 8 were measured as compared to those for coumarin derivative 7. This observation could be helpful for the future design and development of the coumarin based TP fluorophores.

TPF Imaging
TPF microscopy confirmed the cellular uptake of these 3-cinnamoylcoumarins in HeLa cells, revealing the potentials of these coumarins as TP probes. For compounds 7 and 8, TP confocal imaging was carried out using a 710 Laser Scanning Microscope (Zeiss, Jena, Germany). HeLa cells were grown in DMEM containing 10% fetal bovine serum at 37 °C. Cinnamoyl -coumarins (2 g/mL) were incubated for 30 minutes in dish (P35G-1.5-10-C, MatTek Corporation, Ashland, MA, USA), and the uptake was directly analyzed under the microscope. Figure 3 shows the TPF confocal images of the A correlation of quantum efficiency and TPA cross-sections with the chemical structures of the coumarins was observed. An enhancement in quantum efficiency and TPA cross-section was observed on substituting the C-8 position of 3-(4-diethylaminocinnamoyl) coumarins with a methoxy group. A similar effect of electron donating group on δ value of same conjugated system was observed byLi et al. [24]. Higher TPA cross-section and quantum yield for coumarin derivative 8 were measured as compared to those for coumarin derivative 7. This observation could be helpful for the future design and development of the coumarin based TP fluorophores.

TPF Imaging
TPF microscopy confirmed the cellular uptake of these 3-cinnamoylcoumarins in HeLa cells, revealing the potentials of these coumarins as TP probes. For compounds 7 and 8, TP confocal imaging was carried out using a 710 Laser Scanning Microscope (Zeiss, Jena, Germany). HeLa cells were grown in DMEM containing 10% fetal bovine serum at 37˝C. Cinnamoyl -coumarins (2 µg/mL) were incubated for 30 minutes in dish (P35G-1.5-10-C, MatTek Corporation, Ashland, MA, USA), and the uptake was directly analyzed under the microscope. Figure 3 shows the TPF confocal images of the uptake of the 3-cinnamoylcoumarin molecules in HeLa cells. The cellular uptake results suggest that the synthesized coumarin compounds could be used as potential fluorescent probes for TP imaging.
Sensors 2015, 15 9 uptake of the 3-cinnamoylcoumarin molecules in HeLa cells. The cellular uptake results suggest that the synthesized coumarin compounds could be used as potential fluorescent probes for TP imaging.

DCP Sensing Study
To study the DCP sensing capability of 3-cinnamoylcoumarins 7 and 8, the absorption and the fluorescence spectra of the compounds was recorded in chloroform with the addition of different

DCP Sensing Study
To study the DCP sensing capability of 3-cinnamoylcoumarins 7 and 8, the absorption and the fluorescence spectra of the compounds was recorded in chloroform with the addition of different concentrations of DCP. The solutions of compounds 7 and 8 in chloroform absorb strongly around 506 and 505 nm, and emit bright orange fluorescence with λ max at 573 and 572 nm upon excitation, respectively. The fluorescence quenching of both compounds 7 and 8 with DCP in the chloroform solutions were observed (Figure 4). Both solutions turned colorless upon addition of DCP. The absorption peaks at 506 and 505 nm for compounds 7 and 8 kept decreasing with the appearance of a new absorption band at 372 nm. Figure 5 presents the absorption spectra of compounds 7 and 8 upon titrating with DCP (0.85-7.68 mM). A blue shift in the absorption maxima of compounds 7 and 8 was determined to be about 133 nm on the incremental addition of DCP. This change in UV absorption that accompanies DCP binding i.e., one peak goes down and the other goes up may be useful for ratiometric measurements.  Upon addition of DCP to chloroform solution of cinnamoylcoumarins 7 and 8, the fluorescence intensities at 573 and 572 nm decreased as shown in Figure 6. The corresponding Stern-Volmer plots are shown in Figure S13 and the Stern-Volmer constant (Ksv), the binding constant (K) and number of binding sites (n) ( Figure S14) were calculated and are listed in Table 2. It has been observed that the substitution of the methoxy group at the C-8 position of the coumarin increases the value of the Stern-Volmer constant and the binding constant. This may be due to the electron donating effect of methoxy group which enriched the nucleophilic attack of compound on phosphorous atom of DCP. The Ksv values for both compounds 7 and 8 were found to be higher than the value reported for the DCP sensor  Upon addition of DCP to chloroform solution of cinnamoylcoumarins 7 and 8, the fluorescence intensities at 573 and 572 nm decreased as shown in Figure 6. The corresponding Stern-Volmer plots are shown in Figure S13 and the Stern-Volmer constant (Ksv), the binding constant (K) and number of binding sites (n) ( Figure S14) were calculated and are listed in Table 2. It has been observed that the substitution of the methoxy group at the C-8 position of the coumarin increases the value of the Stern-Volmer constant and the binding constant. This may be due to the electron donating effect of methoxy group which enriched the nucleophilic attack of compound on phosphorous atom of DCP. The Ksv values for both compounds 7 and 8 were found to be higher than the value reported for the DCP sensor based on poly(fluorene-quinoxaline) [34]. To confirm if the fluorescence quenching is a selective phenomenon for the cinnamoylcoumarins 7 and 8, we have synthesized three structural analogues (11−13) of compounds 7 and 8 (Scheme 2). The TPA activity of compound 11 was previously reported by our group, but the DCP sensing application has not been explored earlier [27]. For these analogues Upon addition of DCP to chloroform solution of cinnamoylcoumarins 7 and 8, the fluorescence intensities at 573 and 572 nm decreased as shown in Figure 6. The corresponding Stern-Volmer plots are shown in Figure S13 and the Stern-Volmer constant (Ksv), the binding constant (K) and number of binding sites (n) ( Figure S14) were calculated and are listed in Table 2. It has been observed that the substitution of the methoxy group at the C-8 position of the coumarin increases the value of the Stern-Volmer constant and the binding constant. This may be due to the electron donating effect of methoxy group which enriched the nucleophilic attack of compound on phosphorous atom of DCP. The Ksv values for both compounds 7 and 8 were found to be higher than the value reported for the DCP sensor based on poly(fluorene-quinoxaline) [34]. To confirm if the fluorescence quenching is a selective phenomenon for the cinnamoylcoumarins 7 and 8, we have synthesized three structural analogues (11´13) of compounds 7 and 8 (Scheme 2). The TPA activity of compound 11 was previously reported by our group, but the DCP sensing application has not been explored earlier [27]. For these analogues (11´13), although a similar tendency of fluorescence quenching has been observed on incremental addition of DCP (Figures S15-S17 in SI) as for the compounds 7 and 8. However, the measured Stern-Volmer constants revealed that these analogues of 7 and 8 exhibit weaker quenching responses with DCP as described in Table S2. The binding constants for the synthesized compounds are comparable with the value reported for rhodamine-based sensors for DCP [35].  In order to evaluate the selectivity of compounds 7 and 8 for different organophosphorous compounds, we studied the change in the fluorescence intensity of compound 7 towards two more organophosphorous compounds, i.e., orthophosphoric acid and dimethyl methylphosphonate. Quite interestingly, no significant change in the emission spectra of compound 7 was observed with the addition of any of the two compounds ( Figure S18) even with addition of 30 times higher amount of orthophosphoric acid/dimethyl methylphosphonate in comparison to DCP (9.23 mM). This result clearly showed that compounds 7 and 8 are highly selective towards sensing the DCP.   In order to evaluate the selectivity of compounds 7 and 8 for different organophosphorous compounds, we studied the change in the fluorescence intensity of compound 7 towards two more organophosphorous compounds, i.e., orthophosphoric acid and dimethyl methylphosphonate. Quite interestingly, no significant change in the emission spectra of compound 7 was observed with the addition of any of the two compounds ( Figure S18) even with addition of 30 times higher amount of orthophosphoric acid/dimethyl methylphosphonate in comparison to DCP (9.23 mM). This result clearly showed that compounds 7 and 8 are highly selective towards sensing the DCP.

NMR Study and the Sensing Mechanism
A 1 H-NMR titration study was performed to find the nucleophilic center (NEt 2 or OH) involved in the binding of DCP. Interestingly, it was noticed that the continuous addition of DCP (0-4 eq.) to compound 8 resulted in a deshielding of the protons ortho (H-3 11 & H-5 11 ) to the NEt 2 group (Figure 7). A downfield shift in the peak signal of the methylene protons [N(CH 2 CH 3 ) 2 ] as well as meta-protons (H-2 11 & H-6 11 ) was observed on addition of DCP ( Figure S19), while the benzenoid protons of the coumarin skeleton ring (H-5 & H-6) did not exhibit any significant change in δ value, this may be due to the safe distance from the reactive nucleophilic center involved with DCP interaction. These facts suggest that the participation of NEt 2 group in nucleophilic attack on phosphorous atom of DCP.
A correlation of the chemical structure and TPF was established. The TPF confocal microscopic investigation confirmed the internalization of these 3-cinnamoyl coumarins by HeLa cell lines. These compounds showed large fluorescence quenching upon addition of DCP. The Ksv values obtained for the compounds 7 and 8 by the fluorescence titration method were found to be 791 and 1070 M´1, respectively. These values are comparable or even higher than those reported for other nerve gas sensors. Thus, we have successfully developed a reagent that exhibit good selectivity to interact with DCP in comparison to its structural analogues DMMP and orthophosphoric acid. Our research findings could be useful for future design of coumarin based biological imaging probes and nerve gas sensors.