A Comparison of the Photophysical, Electrochemical and Cytotoxic Properties of meso-(2-, 3- and 4-Pyridyl)-BODIPYs and Their Derivatives

Boron dipyrromethene (BODIPY) dyes bearing a pyridyl moiety have been used as metal ion sensors, pH sensors, fluorescence probes, and as sensitizers for phototherapy. A comparative study of the properties of the three structural isomers of meso-pyridyl-BODIPYs, their 2,6-dichloro derivatives, and their corresponding methylated cationic pyridinium-BODIPYs was conducted using spectroscopic and electrochemical methods, X-ray analyses, and TD-DFT calculations. Among the neutral derivatives, the 3Py and 4Py isomers showed the highest relative fluorescence quantum yields in organic solvents, which were further enhanced 2-4-fold via the introduction of two chlorines at the 2,6-positions. Among the cationic derivatives, the 2catPy showed the highest relative fluorescence quantum yield in organic solvents, which was further enhanced by the use of a bulky counter anion (PF6−). In water, the quantum yields were greatly reduced for all three isomers but were shown to be enhanced upon introduction of 2,6-dichloro groups. Our results indicate that 2,6-dichloro-meso-(2- and 3-pyridinium)-BODIPYs are the most promising for sensing applications. Furthermore, all pyridinium BODIPYs are highly water-soluble and display low cytotoxicity towards human HEp2 cells.

While there has been a large number of BODIPY dyes designed, synthesized and characterized over the last two decades, the development and synthesis of low molecular weight and water-soluble BODIPY dyes is yet to be fully explored [18]. Additional studies in this area are needed to provide water-soluble, membrane-permeable BODIPY derivatives and widen their biological applications [19]. The water solubility of BODIPY dyes has been enhanced via installation of solubilizing groups such as phosphates [20,21], sulfonates [22][23][24][25], carboxylates [26][27][28], carbohydrates [29,30], and oligoethylene glycol chains [31]. However, these solubilizing groups increase the dye's molecular size and tend to decrease their cell membrane permeability [32]. On the other hand, pyridinium groups are easy to install, and lead to cationic BODIPY derivatives with enhanced water-solubility, stability, and membrane permeability. In addition, meso-pyridyl-BODIPYs have been used as ligands in coordination chemistry, and as fluorescent pH sensors; therefore, this type of functionalized BODIPYs has found applications in bioimaging and biosensing. However, previous studies on a meso-(4-pyridinium)-BODIPY showed that upon excitation, the dye undergoes donor-photoinduced electron transfer (d-PeT) to a charge-transfer state that is quenched through intersystem crossing (ISC) [32]. This process occurs by the excitation (S 0 →S 2 ) and subsequent non-radiative decay from (S 2 →S 1 ) moving the electron from HOMO→LUMO+1→LUMO as shown in Figure 1. In this charge-transfer process, the 4-pyridinium group acts as an acceptor and the singlet-excited state of the BODIPY core acts as the donor. As a result, the fluorescence quantum yield of the dye is greatly reduced (see Figure 2a). While restoration of the meso-(4-pyridyl)-BODIPY fluorescence can be achieved by either inhibiting d-PeT or the ISC, the charge recombination emission can be enhanced to varying degrees by increasing the electrostatic charge of the 'hole' created by the electronic excitation and subsequent d-PeT process. We recently showed [33] that this can be accomplished through the installation of electron-withdrawing groups at the 2,6-positions. As an example, chlorination at the 2,6-positions induced an increase in the fluorescence quantum yield of a meso-4-pyridinium BODIPY from 0.019 to 0.048 in acetonitrile and from 0.004 to 0.038 in water [33]. Furthermore, Ueno and coworkers [34] reported that the installation of acetyl groups at the 2,6-positions of a meso-(nitroanisole)-BODIPY increased the fluorescence quantum yield of the dye in acetonitrile by over 20-fold (see Figure 2b). carbohydrates [29,30], and oligoethylene glycol chains [31]. However, these solubilizing groups increase the dye's molecular size and tend to decrease their cell membrane permeability [32]. On the other hand, pyridinium groups are easy to install, and lead to cationic BODIPY derivatives with enhanced water-solubility, stability, and membrane permeability. In addition, meso-pyridyl-BODIPYs have been used as ligands in coordination chemistry, and as fluorescent pH sensors; therefore, this type of functionalized BODIPYs has found applications in bioimaging and biosensing. However, previous studies on a meso-(4-pyridinium)-BODIPY showed that upon excitation, the dye undergoes donor-photoinduced electron transfer (d-PeT) to a charge-transfer state that is quenched through intersystem crossing (ISC) [32]. This process occurs by the excitation (S0→S2) and subsequent non-radiative decay from (S2→S1) moving the electron from HOMO→LUMO+1→LUMO as shown in Figure 1. In this charge-transfer process, the 4-pyridinium group acts as an acceptor and the singlet-excited state of the BOD-IPY core acts as the donor. As a result, the fluorescence quantum yield of the dye is greatly reduced (see Figure 2a). While restoration of the meso-(4-pyridyl)-BODIPY fluorescence can be achieved by either inhibiting d-PeT or the ISC, the charge recombination emission can be enhanced to varying degrees by increasing the electrostatic charge of the 'hole' created by the electronic excitation and subsequent d-PeT process. We recently showed [33] that this can be accomplished through the installation of electron-withdrawing groups at the 2,6-positions. As an example, chlorination at the 2,6-positions induced an increase in the fluorescence quantum yield of a meso-4-pyridinium BODIPY from 0.019 to 0.048 in acetonitrile and from 0.004 to 0.038 in water [33]. Furthermore, Ueno and coworkers [34] reported that the installation of acetyl groups at the 2,6positions of a meso-(nitroanisole)-BODIPY increased the fluorescence quantum yield of the dye in acetonitrile by over 20-fold (see Figure 2b).  Herein we report the synthesis and investigation of meso-(2-pyridyl) and meso-(3pyridyl)-BODIPYs, and of their 2,6-dichloro and pyridinium derivatives (see Figure 3) and compare their photophysical, electrochemical and cytotoxic properties relative to the known meso-(4-pyridyl) and meso-(4-pyridinium) derivatives and their 2,6-dichloro analogs. The moderate electron-withdrawing chlorine atoms are easily introduced at the 2,6-positions and expected to induce red-shifts in the absorption and emission wavelengths, increase reduction potentials, increase stability, and enhance the relative fluorescence quantum yields in both organic solvents and water. This type of BODIPY and halogenated derivatives could find multiple applications as optical sensors, fluorescent probes, or as sensitizers for phototherapy.  Herein we report the synthesis and investigation of meso-(2-pyridyl) and m (3-pyridyl)-BODIPYs, and of their 2,6-dichloro and pyridinium derivatives (see F ure 3) and compare their photophysical, electrochemical and cytotoxic proper relative to the known meso-(4-pyridyl) and meso-(4-pyridinium) derivatives a their 2,6-dichloro analogs. The moderate electron-withdrawing chlorine atoms easily introduced at the 2,6-positions and expected to induce red-shifts in the sorption and emission wavelengths, increase reduction potentials, increase sta ity, and enhance the relative fluorescence quantum yields in both organic solve and water. This type of BODIPY and halogenated derivatives could find multi applications as optical sensors, fluorescent probes, or as sensitizers for phototh apy.

General
Commercially available reagents and solvents were used as received from VWR Sigma Aldrich unless noted otherwise. All reactions were monitored by thin-layer ch matography (TLC) using 0.2 mm silica gel plates (with UV indicator, polyester backed Å, pre-coated). Liquid chromatography was performed on preparative TLC plates or silica gel column chromatography (60 Å, 230-400 mesh). NMR spectra were measu on 400 or 500 MHz for 1 H and 500 or 750 MH for 13 13 C NM coupling constants (J) are given in hertz. High-resolution mass spectra (HRMS) were tained using an Agilent 6230-B ESI-TOF mass spectrometer.
Herein we report the synthesis and investigation of meso- (2-pyridyl) and meso-(3-pyridyl)-BODIPYs, and of their 2,6-dichloro and pyridinium derivatives (see Figure 3) and compare their photophysical, electrochemical and cytotoxic properties relative to the known meso-(4-pyridyl) and meso-(4-pyridinium) derivatives and their 2,6-dichloro analogs. The moderate electron-withdrawing chlorine atoms are easily introduced at the 2,6-positions and expected to induce red-shifts in the absorption and emission wavelengths, increase reduction potentials, increase stability, and enhance the relative fluorescence quantum yields in both organic solvent s and water. This type of BODIPY and halogenated derivatives could find multiple applications as optical sensors, fluorescent probes, or as sensitizers for phototherapy.

General
Commercially available reagents and solvents were used as received from VWR or Sigma Aldrich unless noted otherwise. All reactions were monitored by thin-layer chromatography (TLC) using 0.2 mm silica gel plates (with UV indicator, polyester backed, 60 Å, pre-coated). Liquid chromatography was performed on preparative TLC plates or via silica gel column chromatography (60 Å, 230-400 mesh). NMR spectra were measured on 400 or 500 MHz for 1 H and 500 or 750 MH for 13 C spectrometer. Chemical shifts (δ) are given in parts per million (ppm) In CDCl3 (7.27 ppm for 1 H NMR, 77.0 ppm for 13 C NMR) or (CD3)2CO (2.05 ppm for 1 H NMR, 206.68 and 29.92 ppm for 13 C NMR) or CD3CN (1.94(5) ppm for 1 H NMR, 118.69 and 1.39 ppm for 13 C NMR); coupling constants (J) are given in hertz. High-resolution mass spectra (HRMS) were obtained using an Agilent 6230-B ESI-TOF mass spectrometer.

General
Commercially available reagents and solvents were used as received from VWR or Sigma Aldrich unless noted otherwise. All reactions were monitored by thin-layer chromatography (TLC) using 0.2 mm silica gel plates (with UV indicator, polyester backed, 60 Å, pre-coated). Liquid chromatography was performed on preparative TLC plates or via silica gel column chromatography (60 Å, 230-400 mesh). NMR spectra were measured on 400 or 500 MHz for 1 H and 500 or 750 MH for 13 C spectrometer. Chemical shifts (δ) are given in parts per million (ppm) In CDCl 3 (7.27 ppm for 1 H NMR, 77.0 ppm for 13 C NMR) or (CD 3 ) 2 CO (2.05 ppm for 1 H NMR, 206.68 and 29.92 ppm for 13 C NMR) or CD 3 CN (1.94(5) ppm for 1 H NMR, 118.69 and 1.39 ppm for 13 C NMR); coupling constants (J) are given in hertz. High-resolution mass spectra (HRMS) were obtained using an Agilent 6230-B ESI-TOF mass spectrometer.
BODIPYs 2Py [35] and 3Py [36] were prepared following a procedure previously reported, by reacting 2,4-dimethylpyrrole with the corresponding pyridine 2-carbaldehyde in the presence of trifluoroacetic acid (TFA), followed by oxidation with 2,3-dichloro-5,6dicyano-1,4-benzoquinone (DDQ) and complexation with BF 3 •(OEt) 2 in triethylamine. The spectroscopic data for 2Py and 3Py are in agreement with the literature reports. BODIPY 2Py (13.31 mg, 0.041 mmol) was dissolved in dry acetonitrile (3 mL) in an oven dried 10 mL round bottom flask. An excess of methyl iodide (5 mL, 80.32 mmol) was added, and the mixture was refluxed for 8 h under N 2 atmosphere. The solvent was removed under reduced pressure. The grayish residue was washed with petroleum ether (5 mL) and the resulting solid recovered via filtration to afford 13 mg, 98% yield of 2catPy. 1  This compound was prepared as described above for 2catPy, using 3Py (13.3 mg, 0.041 mmol), dry acetonitrile (3 mL) and methyl iodide (5 mL, 80.32 mmol). The title BODIPY was obtained in 98% yield as a grayish powder. 1 A solution of BODIPY 2Py (10.7 mg, 0.03 mmol) in dry dichloromethane (6 mL) was stirred at room temperature in an oven dried round bottom flask under N 2 atmosphere. Trichloroisocyanuric acid (TCCA) (6.10 mg, 0.026 mmol) was dissolved in dichloromethane (2 mL) and added slowly to the solution. The reaction was stirred at room temperature for 2 h, until complete disappearance of the starting material. The solvent was removed under reduced pressure and the resulting residue was purified via column chromatography using 5-20% ethyl acetate in hexanes, to give 9 mg (84%) of the title compound. 1   This compound was prepared as described above for 2PyCl 2 , using BODIPY 3Py (12.1 mg, 0.04 mmol) and TCCA (7.70 mg, 0.03 mmol). The reaction was purified via column chromatography using 5-20% ethyl acetate in hexanes yielding 10 This compound was prepared as described above for 2catPy, using BODIPY 2PyCl 2 (15 mg, 0.038 mmol), anhydrous acetonitrile (10 mL), and methyl iodide (5 mL). The title compound was obtained (13.4 mg, 89% yield). 1 This compound was prepared as described above for 3catPy, using BODIPY 3PyCl 2 (10 mg, 0.025 mmol), anhydrous acetonitrile (10 mL), and methyl iodide (4 mL). The title compound was obtained (8.6 mg, 86% yield). 1

Spectroscopy Methods
UV−vis absorption spectra were collected on a Varian Cary 50 Bio spectrophotometer. Emission spectra were obtained on a PerkinElmer LS55 spectrophotometer, at room temperature. Spectrophotometric grade solvents and quartz cuvettes (1 cm path length) were used. Relative fluorescence quantum yields (Φ f ) were calculated using rhodamine 6G (Φ f = 0.86 in methanol) as reference for the neutral BODIPYs and Ru(bpy)Cl 2 .6H 2 O (Φ f = 0.028 in water) as the standard for the cationic BODIPYs using the following equation: 2 , where the Φ X and Φ ST are the quantum yields of the sample and standard, Grad X and Grad ST are the gradients from the plot of integrated fluorescence intensity vs. absorbance, and η represents the refractive index of the solvent (x is for the sample and st for the standard).

Electrochemistry
Cyclic voltammograms were collected on a Biologic SP-300 potentiostat using a threeelectrode setup with a platinum working electrode (2 mm diameter, CH Instruments), platinum counter electrode (2 mm diameter, CH Instruments), and a silver wire pseudo reference-electrode stored in 0.25 M NBu 4 PF 6 in anhydrous acetonitrile and separated from the electrochemical cell by a glass frit. The electrochemical cell was composed of a borosilicate glass scintillation vial with a custom-machined cap. Working electrodes were polished and then pre-treated by cycling between upper and lower boundaries of the electrochemical window for the experiment in a solution containing only the electrolyte. Measurements were taken in a solution of 0.25 M recrystallized NBu 4 PF 6 in dried and degassed anhydrous acetonitrile. BODIPYs were added to 5 mL of electrolyte solution to give a concentration of 1 mM. Each solution also contained 1 mM ferrocene as an internal reference. Solutions were sparged with N 2 before each measurement.

X-ray Crystallography
The crystal structures of 2Py, 2PyCl 2 and 3Py were determined at T = 90 K from data collected on a Bruker Kappa Apex-II diffractometer equipped with Mo Kα (for 2Py and 2PyCl 2 ) or CuKα (for 3Py) source. Absorption corrections were by the multi-scan method. Non-hydrogen atoms were treated with anisotropic displacement parameters. The H atoms were visible in difference maps and were placed in idealized positions in the refinements, with torsional parameters refined for methyl groups.

Computational Methods
The geometries of the ground and excited states of all compounds were optimized without symmetry constraints at the cam-b3lyp/6-31+G(d,p) level in vacuum. This method has been shown to correctly reproduce the experimental trends [18]. The UV-vis absorption data were calculated using the TD-DFT method at the same level. The rotational barriers were calculated in CH 3 CN taking the solvent effects into account using the Polarized Continuum Model (PCM). All calculations were performed using the Gaussian 09 program package [37].

Cell Toxicity
Human carcinoma HEp2 cells were obtained from ATCC and maintained in a 75 cm 2 flask (Chemglass) with 1× MEM (Eagle's minimal essential medium) containing 10% FBS (fetal bovine serum) and 1% antibiotic (penicillin-streptomycin) in a humidified, 5% CO 2 incubator at 37 • C. Stock solutions with a concentration of 32 mM in DMSO were prepared for each BODIPY. The stock solution was diluted to the final working concentrations with culture medium. The HEp2 cells were plated at 7500-10,000 cells per well in a Costar 96-well plate and allowed to grow for 24 h. The BODIPY stock solution was diluted to a 200 µM working solution and two-fold serial dilutions were prepared with final concentrations of 200, 100, 50, 25, 12.5, 6.25, and 0 µM. Each BODIPY was incubated for 24 h at 37 • C under 5% CO 2 . For the dark cytotoxicity experiments, after incubation, excess BODIPY was removed by washing the cells three times with 1× PBS (phosphate-buffered saline) and replacing with media containing 20% Cell Titer Blue (Promega, Madison, WI, USA). The cells were then incubated for additional 24 h at 37 • C under 5% CO 2 and the cell toxicity was determined by fluorescence intensity at ex 570/em 615 nm using a FluoStar Optima microplate reader (BMG Labtech, Cary, NC, USA).
For the phototoxicity experiments, after incubation the cells were exposed to a light dose of~1.5 J/cm 2 generated using a 600 W Quartz Tungsten Halogen lamp (Newport Corporation, Irvine, CA, USA). The cells were exposed for 20 min while the 96-well plate rested on an EchoTherm chilling plate (Torrey Pines Scientific, Carlsbad, CA, USA) set to 5 • C. Following light exposure, the loading media was removed and media containing 20% Cell Titer Blue was added. After incubation for 24 h at 37 • C under 5% CO 2 , the cell viability was determined as described above.

Synthesis
The 2Py [35] and 3Py [36] BODIPYs were prepared in 48% and 55% yields, respectively, by condensation of 2,4-dimethylpyrrole with the corresponding pyridine carboxaldehyde in presence of TFA, followed by oxidation with DDQ and boron complexation using boron trifluoride etherate, as previously reported. Treating BODIPYs 2Py and 3Py with excess of methyl iodide in acetonitrile gave the corresponding pyridinium BODIPYs, 2catPy and 3catPy, in nearly quantitative yields. Anion exchange was performed on these BODIPYs using ammonium hexafluorophosphate (NH 4 PF 6 ). The resulting salts were purified by washing with organic solvents and characterized by mass spectrometry.

X-ray Analysis
Crystals of 2Py, 3Py and 2PyCl 2 were obtained by slow evaporation of dichloromethane and pentane, and the results are shown in Figure 4. BODIPYs 2Py, 3Py, and 2PyCl 2 all have their 12-atom BODIPY cores nearly planar, with mean deviations from coplanarity of 0.029, 0.023 and 0.038 Å, respectively. The dihedral angles between the BODIPY core planes and the pyridyl planes are also similar, 79.3 • for 2Py, 82.3 • for 3Py (average of two independent molecules), and 85.9 • for 2PyCl 2 . These values agree well with those for the published structure of 4Py [36,38,39], for which the mean deviation from the BODIPY core is 0.018 Å and the BODIPY/Py dihedral angle is 83.9 • (average of two). Bond distances and angles also show excellent agreement across the three structures and in agreement with the published structure of 3Py [36]. The relatively lower dihedral angle observed for 2Py relative to the 3Py and 4Py suggests a slight increased flexibility for 2Py, although the difference is very small. Therefore, for all three isomers the pyridine is interlocked between the 1,7-methyl substituents and does not affect the molecular structure of the BODIPY core.
iodide afforded their pyridinium derivatives 2catPyCl2 and 3catPyCl2. Anion exchange was performed on these BODIPYs using ammonium hexafluorophosphate. The structures of all the BODIPYs were confirmed by 1 H, 13 C and 11 B NMR (see Supplementary Materials, Figures S3-S26) and by high-resolution mass spectrometry (ESI-TOF).

X-ray Analysis
Crystals of 2Py, 3Py and 2PyCl2 were obtained by slow evaporation of dichloromethane and pentane, and the results are shown in Figure 4. BODIPYs 2Py, 3Py, and 2PyCl2 all have their 12-atom BODIPY cores nearly planar, with mean deviations from coplanarity of 0.029, 0.023 and 0.038 Å, respectively. The dihedral angles between the BODIPY core planes and the pyridyl planes are also similar, 79.3° for 2Py, 82.3° for 3Py (average of two independent molecules), and 85.9° for 2PyCl2. These values agree well with those for the published structure of 4Py [36,38,39], for which the mean deviation from the BODIPY core is 0.018 Å and the BODIPY/Py dihedral angle is 83.9° (average of two). Bond distances and angles also show excellent agreement across the three structures and in agreement with the published structure of 3Py [36]. The relatively lower dihedral angle observed for 2Py relative to the 3Py and 4Py suggests a slight increased flexibility for 2Py, although the difference is very small. Therefore, for all three isomers the pyridine is interlocked between the 1,7-methyl substituents and does not affect the molecular structure of the BODIPY core.

Spectroscopic Properties
The absorption and emission spectra of BODIPYs 2Py, 3Py, 2PyCl 2 , 3PyCl 2 and their cationic derivatives 2catPy, 3catPy, 2catPyCl 2 , 3catPyCl 2 were obtained at room temperature in acetonitrile, methanol, and water, and the results are as summarized in Table 1, and in Figure S1 and Table S1 of the Supplementary Materials. As expected, the absorption spectra for 2Py, 3Py and 4Py featured a sharp peak at ca. 501 nm which is attributed to the BODIPYs strong S 0 -S 1 (π-π*) transition, as confirmed by molecular modeling. The broad high energy shoulder bands are attributed to the BODIPYs vibrational transitions. The emission peak for these BODIPYs is seen at ca. 514 nm, and the relatively small Stokes shift indicates that the molecular structure of the excited state is not very different than that of the ground state. The maximum absorption and emission wavelengths for the neutral 2Py, 3Py and 4Py are almost identical in both acetonitrile and methanol (see Tables 1 and S1). In agreement with this observation, computational modeling of the absorption spectra in the gas phase confirms the almost identical maximum absorption wavelengths, which correspond well to the almost identical HOMO-LUMO gaps, as can be seen in Table 2. For the entire series of 2Py, 3Py and 4Py, the HOMO is almost entirely localized on the BODIPY core, whereas the LUMO partially involves the pyridyl group (see Figure 5a, and Figures A1 and A2 of Appendix A). LUMO+1 is predominantly on the pyridyl group.   The presence of a 2-, 3-or 4-pyridyl group at the meso-position has a strong effect on the relative fluorescence quantum yields using rhodamine 6G as the reference (Фf = 0.86 in methanol) [40]. BODIPY 3Py was found to be the most fluorescent (f = 0.43) and the 2Py is the least (f = 0.04). The high fluorescence of 3Py and the 4Py in organic solvents has previously been observed [36,41,42]. On the other hand, the 2Py displays much lower fluorescence compared with its regioisomers, as previously reported [35,43]. This is due to the closer proximity of the nitrogen atom to the BODIPY core in 2Py relative to the 3Py and 4Py compounds. The 2pyridyl group has a lower rotational barrier compared with the 3-and 4-pyridyl groups (see Table 2), suggesting that in the case of 2Py more energy is released via non-radiative decay pathways. The meso-pyridyl rotation causes distortion of the BODIPY core that disrupts electron conjugation and could be the reason for the observed lower quantum yields for 2Py. Another possible explanation is the difference in the S1 LUMO. In the case of 3Py and the 4Py, the S1 LUMO is almost entirely localized on BODIPY, whereas for 2Py, the orbital is delocalized over BODIPY and pyridyl. The presence of a 2-, 3-or 4-pyridyl group at the meso-position has a strong effect on the relative fluorescence quantum yields using rhodamine 6G as the reference (Φ f = 0.86 in methanol) [40]. BODIPY 3Py was found to be the most fluorescent (Φ f = 0.43) and the 2Py is the least (Φ f = 0.04). The high fluorescence of 3Py and the 4Py in organic solvents has previously been observed [36,41,42]. On the other hand, the 2Py displays much lower fluorescence compared with its regioisomers, as previously reported [35,43]. This is due to the closer proximity of the nitrogen atom to the BODIPY core in 2Py relative to the 3Py and 4Py compounds. The 2-pyridyl group has a lower rotational barrier compared with the 3-and 4-pyridyl groups (see Table 2), suggesting that in the case of 2Py more energy is released via non-radiative decay pathways. The meso-pyridyl rotation causes distortion of the BODIPY core that disrupts electron conjugation and could be the reason for the observed lower quantum yields for 2Py. Another possible explanation is the difference in the S1 LUMO. In the case of 3Py and the 4Py, the S1 LUMO is almost entirely localized on BODIPY, whereas for 2Py, the orbital is delocalized over BODIPY and pyridyl.
Methylation of the pyridyl group leads to a decrease in the molar extinction coefficients, as well as bathochromic shifts in the maximum absorption wavelengths of the cationic derivatives in acetonitrile, that was more pronounced for the 2catPy (18 nm) relative to the 3catPy (8 nm) and 4catPy (8 nm). This tendency is also seen in the calculated maximum absorption wavelengths; all cationic BODIPYs show bathochromic shifts with that of 2catPy (19 nm) being more pronounced than for 3catPy (13 nm) and 4catPy (12 nm), demonstrating similar shifts. For this series of methylated compounds, the HOMO is again almost entirely located on the BODIPY core but the LUMO and LUMO+1 switch compared with the non-methylated analogs. In this case, LUMO is mostly on the pyridinium group, whereas LUMO+1 is mostly on BODIPY core (see Figure 5 and Figures A1 and A2 of Appendix A). This suggests that PeT occurs from the BODIPY core to the pyridinium group upon excitation, which quenches a high percentage of the fluorescence. The gaps between HOMO and LUMO+1 (LUMO BODIPY ) are given in Table 2, as this is the transition corresponding to the fluorescence activity. The electron-withdrawing effect lowers both HOMO and LUMO BODIPY with greater effect on LUMO BODIPY , thus decreasing the gap between these two orbitals, which is in agreement with the experimentally observed bathochromic shift.
The observed bathochromic shifts are similar in the case of 3catPy and 4catPy but different in the case of 2catPy. This is likely due to the stronger electron-withdrawing effect of the 2-pyridinium group due to the proximity of the nitrogen atom to the BODIPY core. While in the case of the 3catPy and 4catPy the pyridinium nitrogen pulls electron density nearly equally from both sides of the BODIPY core, in the case of 2catPy there is a direct pull of electron density from the nearby carbon-2 of the pyridyl group, which induces the observed enhanced bathochromic shift. This is in agreement with the observed similar orbital energies of HOMO and LUMO BODIPY in the case of 3catPy and 4catPy, and lower orbital energies in the case 2catPy (see Figure 5 and Figures A1 and A2 of Appendix A). Furthermore, moderate redshifted emissions were observed for 2catPy and 3catPy, centered at 544 and 532 nm, respectively. In the case of 4catPy a more pronounced redshift to 596 nm was observed, as we previously reported [33], likely due to the highly dipolar species formed upon excitation, with a pronounced charge-transfer character [44].
The fluorescence in acetonitrile of 3catPy was drastically reduced to 0.006, as previously observed for 4catPy [33], due to intramolecular charge transfer, since pyridinium is more easily reduced than pyridine. This has led to the exploration of photo-or chemical induced dealkylation of a 4-pyridinium quencher as a strategy to restore the fluorescence of the BODIPY chromophore [45][46][47]. On the other hand, the fluorescence of 2catPy actually increased to 0.04 to 0.13 or 0.23 with I − or PF 6 − as the counter ion, respectively. The observed lower quantum yields for 3catPy and 4catPy compared with 3Py and 4Py are likely due to intramolecular charge transfer, since pyridinium is more easily reduced than pyridine. The energy difference between LUMO and LUMO+1 is only 0.30 eV in the case of 2catPy and 0.79/0.89 eV in 3catPy/4catPy, respectively (see Figure 5a, and Figures A1 and A2 of Appendix A), suggesting easier charge-recombination and higher quantum yield in 2catPy. Furthermore, 3catPy and 4catPy show higher probability for intersystem crossing compared with 2catPy, as suggested by the lower S1→T energy differences ( Table 2). The increase in fluorescence upon alkylation of 2Py has been previously observed [35], and is in part due to the bulkiness of the alkyl group on the pyridinium nitrogen atom in close proximity to the BODIPY core. The larger counter ion (PF 6 − ) induces the highest relative fluorescence in the 2catPy, but this effect is not observed for the 3catPy nor the 4catPy. We hypothesize that in the case of the cationic 3-and 4-pyridinium-BODIPYs with iodide as counter ion, there is likely a close interaction of the BODIPYs with the iodide anion, since the pyridinium nitrogen in 3catPy and 4catPy is easily accessible. Therefore, the iodide anion in these cases is able to exert a heavy-atom effect that leads to ISC from the lowest singlet excited state (S 1 ) to the triplet state of the cationic system [48]. On the other hand, in BODIPY 2catPy the sterically hindered cationic nitrogen center prevents close interaction with the iodide ion.
In water, the fluorescence quantum yields decreased further due to the increase in solvent polarity which stabilizes the charge-transfer state leading to deactivation through ISC, as previously observed for 2catPy [35] and 4catPy [33]. However, the maximum absorption and emission observed for the 2catPy and 3catPy remained unchanged in acetonitrile, methanol, and water (see Table 1 and Supplementary Table S1 in Supplementary Materials).
The installation of the 2,6-dichloro substituents led to significant redshifted absorptions (25-27 nm), decreased molar absorptivities, red-shifted emissions (28-31 nm), and slight increased Stokes shifts of the neutral BODIPYs, as previously observed [33,49], due to the electron-withdrawing effect of the chlorine atoms. The chlorines at the 2,6-positions stabilize both the HOMO and LUMO, and this effect is more pronounced on the LUMO (see Figure 5 and Figures A1 and A2 of Appendix A). As a result, the HOMO-LUMO gaps are smaller for the chlorinated derivatives, causing the observed redshifts of the absorption wavelengths. The performed calculations show an increase in the oscillator strengths upon introduction of the chlorine atoms. In addition, the experimental relative fluorescence quantum yields increased upon 2,6-dichlorination, likely due to reduction in nonradiative deactivation, as it was previously shown that the nonradiative rate constant significantly decreases with the introduction of chlorine groups [49]. Chlorination at the 2,6-positions also increased the fluorescence quantum yields of the BODIPYs due to the decrease in electronic charge density on the BODIPYs. Compound 2PyCl 2 again showed relatively reduced fluorescence (Φ f = 0.17) in comparison with 3PyCl 2 and 4PyCl 2 (both Φ f = 0.58) due to its closer proximity of the nitrogen atom to the BODIPY core.
The corresponding methylated 2,6-dichlorinated analogues showed increase in fluorescence quantum yields in acetonitrile, particularly in the case of 3catPyCl 2 which showed a 38-fold enhancement in fluorescence relative to 3catPy . In water, the absorption and emission wavelengths of the cationic derivatives did not change significantly, except in the case of 2catPyCl 2 . Furthermore, the fluorescence quantum yields were partially enhanced upon introduction of the 2,6-chlorines, as we previously reported in the case of 4catPyCl 2 [33]. A drastic increase was observed in the case of 2catPyCl 2 (Φ f = 0.24) which displayed over 200-fold increase in fluorescence relative to 2catPy, whereas 100-and 10-fold enhancements were observed for 3catPyCl 2 and 4catPyCl 2 , respectively. The observed fluorescence enhancements are likely due to an alteration of the excited state lifetime resulting from the increase in the BODIPY oxidation potential of the chlorinated BODIPY derivatives in addition to the naturally higher oxidation potential of the 2-pyridinium BODIPY (see below), which increases the rate of charge recombination by magnifying the electron-hole electrostatic interaction. In agreement with these findings, up to 100-fold fluorescence enhancement was reported for a 2catPy derivative bearing methylmercaptophenoxy groups in place of the fluorines on the boron atom, upon oxidation to the corresponding sulfoxides [50]. Furthermore, our results show that the fluorescence of 3catPyCl 2 in water (Φ f = 0.47) is the highest among the series of pyridinium-BODIPYs.

Electrochemical Properties
The electrochemical data for the pyridinium-BODIPYs 2catPy, 3catPy and 4catPy with PF 6 − as counterion are summarized in Table 3 and shown in Figure 6. Each pyridinium-BODIPY shows one quasi-reversible oxidation corresponding to the formation of the p-radical cation [51]. Due to the stronger electron-withdrawing effect of the 2-pyridinium group in 2catPy, the oxidation potential is shifted positive by 100 mV compared with 4catPy and 3catPy. Each pyridinium-BODIPY shows reversible reductions that occur either through an apparent two-electron reduction or two separate single-electron reductions. 4catPy undergoes a single two-electron reduction which corresponds to the reductions localized on the BODIPY core and the pyridinium group. It is likely that the initial reduction forms a radical which is more easily reduced than the initial BODIPY compound, resulting in the second electron transfer occurring almost immediately upon the formation of the neutral radical. These results are in agreement with previous published data [32]. Meanwhile, 2catPy and 3catPy show two sequential reversible one electron reductions, which suggest an increased stability of the radical formed upon the initial reduction ( Figure 6a). Overall, 2catPy and 3catPy are more resistant to reduction to either the singly or doubly reduced state compared with 4catPy. BODIPY 2catPy undergoes its first reduction 30 mV more negative and its second reduction 230 mV more negative than the two-electron reduction seen in 4catPy. BODIPY 3catPy is more significantly affected as its first reduction appears 200 mV more negative and its second reduction appears 480 mV more negative than the two-electron reduction in 4catPy. The first reduction for both 2catPy and 3catPy lies in the potential range usually seen for the reduction in the BODIPY core to its singly reduced state, whereas the second reduction may correspond to the reduction in the pyridinium group [51]. Thus, the influence of meso-pyridinium substituents on the BODIPY redox potentials are moderate (~170 mV) and consistent with the electron-withdrawing effects of the pyridinium groups. On the other hand, the introduction of the 2,6-dichloro groups induces larger effects, as discussed below. compared with 4catPy. Meanwhile 2catPyCl2 and 3catPyCl2 undergo two se tial one-electron reductions. The first reduction wave is reversible for 2catPyCl2 and 3catPyCl2 and is shifted positive by 220 mV and 240 mV, r tively, when compared with their unchlorinated BODIPYs. The second, mor ative reduction wave, is reversible in 2catPyCl2 but quasi-reversible in 3ca and is shifted positive by 130 mV and 120 mV when compared with 2catP 3catPy, respectively. As expected, the introduction of the 2,6-dichloro group larger influence on the electrochemical properties of BODIPYs than the meso dinium substituents.

Cytotoxicity Properties
The dark and photocytoxicity of this series of BODIPYs was investiga human carcinoma HEp2 cells and the results are summarized in Table 4 and s in Figure S2 of the Supplementary Materials. Very low cytotoxicity was obs for all the BODIPYs without 2,6-chlorine substitution (IC50 > 200 mM in th and >100 mM at 1.5 J/cm 2 ). The neutral 2,6-dichloro derivatives showed sligh hanced cytotoxicity, particularly the 3PyCl2 due to the introduction of the ch atoms. The most cytotoxic were found to be the cationic 2,6-dichloro deriv Similar trends were seen with chlorinated versions 2catPyCl 2 , 3catPyCl 2 , and 4catPyCl 2 ( Figure 6b and Table 3). The 2,6-dichlorinated BODIPYs have the same number of redox events as their unchlorinated counterparts but the reduction and oxidation waves are shifted positive. The oxidation wave is irreversible for each at 100 mV/s but becomes more reversible at 1 V/s for 4catPyCl 2 , and at 10 V/s for 2catPyCl 2 . Compared with the corresponding unchlorinated BODIPYs at 100 mV/s, the anodic peak for each was shifted positive by 70 mV, 160 mV, and 180 mV for 2catPyCl 2 , 3catPyCl 2 , and 4catPyCl 2 , respectively. As in 4catPy, 4catPyCl 2 undergoes a reversible two-electron reduction, which is shifted 180 mV positive compared with 4catPy. Meanwhile 2catPyCl 2 and 3catPyCl 2 undergo two sequential one-electron reductions. The first reduction wave is reversible for both 2catPyCl 2 and 3catPyCl 2 and is shifted positive by 220 mV and 240 mV, respectively, when compared with their unchlorinated BODIPYs. The second, more negative reduction wave, is reversible in 2catPyCl 2 but quasi-reversible in 3catPyCl 2 and is shifted positive by 130 mV and 120 mV when compared with 2catPy and 3catPy, respectively. As expected, the introduction of the 2,6-dichloro groups has a larger influence on the electrochemical properties of BODIPYs than the meso-pyridinium substituents.

Cytotoxicity Properties
The dark and photocytoxicity of this series of BODIPYs was investigated in human carcinoma HEp2 cells and the results are summarized in Table 4 and shown in Figure S2 of the Supplementary Materials. Very low cytotoxicity was observed for all the BODIPYs without 2,6-chlorine substitution (IC 50 > 200 mM in the dark and >100 mM at 1.5 J/cm 2 ). The neutral 2,6-dichloro derivatives showed slightly enhanced cytotoxicity, particularly the 3PyCl 2 due to the introduction of the chlorine atoms. The most cytotoxic were found to be the cationic 2,6-dichloro derivatives 2catPyCl 2 and 3catPyCl 2 . These results are in agreement with previous observations showing that the introduction of heavier halogens at the 2,6-positions, such as iodides, drastically reduce the fluorescence and increase the phototoxicity of meso-(4-pyridinium)-BODIPYs toward human cancer cells [38] and pathogenic microorganisms [52][53][54].

Conclusions
The photophysical, electrochemical, and cytotoxic properties of meso-(2-pyridyl) and meso-(3-pyridyl)-BODIPYs, and their 2,6-dichloro and pyridinium derivatives were compared with those of their corresponding meso-(4-pyridyl) and meso-(4-pyridinium) analogs. Due to the stronger electron-withdrawing effect of the 2-pyridinium group in 2catPy, its oxidation potential was shifted positive by 100 mV compared with 4catPy and 3catPy. BODIPYs 2catPy and 3catPy were found to be more resistant to reduction compared with 4catPy. Unlike the presence of heavy halogens (Br, I) at the BODIPY 2,6-positions, which is known to enhance ISC, the presence of 2,6-chlorines was shown to enhance the relative fluorescence quantum yields of the BODIPYs in both organic solvents and water. Furthermore, the 2catPy isomer showed the highest relative fluorescence quantum yield in organic solvents, which was further enhanced by the use of a bulky counter anion (PF 6 − ). In water, over a 200-fold increase in fluorescence was observed for 2catPyCl 2 , 100-fold increase for 3catPyCl 2 and 10-fold increase for 4catPyCl 2 . The fluorescence of 3catPyCl 2 in water (F f = 0.47) was found to be the highest among the series of BODIPYs. Our results show that the introduction of 2,6-chlorines in BODIPYs has a larger influence on their electrochemical properties than the meso-pyridyl substituents.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.