Substituent Effects on the Photophysical Properties of a Series of 8(meso)-Pyridyl-BODIPYs: A Computational Analysis of the Experimental Data ^†

Abstract

Recently, a series of 8(meso)-pyridyl-BODIPYs (2-pyridyl, 3-pyridyl, and 4-pyridyl) and their 2,6-substituted derivatives were synthesized and their structure and photophysical properties were studied both experimentally and computationally. One of the main observed trends was that the 2-pyridyl-BODIPYs were consistently less fluorescent than their 3-pyridyl and 4-pyridyl analogs, regardless of the 2,6-substituents. Herein, we extend our previous computational studies and model not only the ground but also the excited states of the entire series of previously synthesized meso-pyridyl-BODIPYs with the aim of explaining the observed differences in the emission quantum yields. To better understand the trends and the effect of 2- and 2,6-substitution on the photophysical and electron-density-related properties, we also model the ground and excited states of BODIPYs that were not synthesized experimentally, however represent a logical part of the series. We calculate a variety of molecular properties and propose that the experimentally observed low quantum yields for all 2-pyridyl-BODIPYs could be due to the very flat potential energy surfaces with respect to the rotation of the 2-pyridyl ring in the excited states, and the stability of a non-planar and significantly less fluorescent meso-2-pyridyl-BODIPY structure.

1. Introduction

Boron dipyrromethene (BODIPY)-based fluorophores have attracted significant research attention in the last several decades [1,2,3,4,5,6] and have found a wide range of applications, including as medical imaging agents [7,8,9,10], biological labels [11,12], photosensitizers in cancer therapy [13,14,15], fluorescence switches [16,17], and laser dyes [18,19,20]. They have been shown to exhibit excellent physicochemical properties, among which are sharp absorption and emission bands in the visible region, high-fluorescence quantum yields, and high photo- and chemical stability [1,2,3,4,5,6]. One of the major advantages of BODIPYs is that they can be functionalized at all the carbon atoms and at the boron center (Figure 1), enabling the fine tuning of their chemical and photophysical properties toward a particular application [4,5,20,21]. This is particularly interesting from a computational chemistry point of view because the design and synthesis of BODIPY fluorophores with pre-determined physicochemical properties could be greatly simplified if it is possible to find a molecular descriptor that is related to the property of interest [22].

Of the various BODIPY derivatives, 8(meso)-pyridyl-substituted BODIPYs have attracted research interest because of the straightforward protonation or alkylation of the pyridyl groups, which allows the synthesis of low-molecular-weight water-soluble fluorescent dyes [23,24]. Recently, a series of meso-(2-, 3-, and 4-pyridyl)-BODIPYs (Figure 2) and their 2- and 2,6-substituted analogs was studied both experimentally and computationally [25,26]. The electron-withdrawing chloro-, nitro-, and methoxycarbonyl groups were used and their effect on the photophysical properties of the respective dyes were studied systematically. One of the main observed trends was that the 2-pyridyl-BODIPYs were consistently less fluorescent that their 3- and 4-pyridyl analogs. The reason for this interesting behavior was not found. Furthermore, the change in photophysical properties with the substitution of each electron-withdrawing group varied substantially. The introduction of a chloro substituent increased the fluorescence quantum yield, while the introduction of a nitro substituent quenched the fluorescence.

Herein, we extend our previous computational studies and model not only the ground but also the excited states of the entire series of previously synthesized meso-(2-, 3-, and 4-pyridyl)-BODIPYs with the aim to explain the observed drastic differences in the emission quantum yields of the dyes. To better understand the trends and the effect of 2- and 2,6-substitution on their photophysical and electron-density related properties, we also model the ground and excited states of compounds not yet synthesized experimentally, however represents a logical part of the series. These are the mono- or di-substituted analogs of the experimentally synthesized compounds, and BODIPYs with R_1,2 = CF₃, as a strongly σ-bond electron-withdrawing group. Thus, the entire series studied includes nPy, nPyCl, nPyCl₂, nPyNO₂, nPy(NO₂)₂, nPyCO₂Me, nPy(CO₂Me)₂, nPyCF₃, and nPy(CF₃)₂, where n = 2, 3, 4. Among these, the nPy, the mono-substituted nPyNO₂, and the di-substituted nPyCl₂ and nPy(CO₂Me)₂ have been previously synthesized experimentally and reported in Refs. [25,26]. To analyze the effect of meso-pyridyl substitution, we also included the meso-phenyl analogs Phe, PheCl, PheCl₂, PheNO₂, Phe(NO₂)₂, PheCO₂Me, Phe(CO₂Me)₂, PheCF₃, and Phe(CF₃)₂.

2. Computational Methods

The geometries of the ground and excited states of all compounds studied were optimized without symmetry constraints at the cam-b3lyp [27] /6-31+G(d,p) level in vacuum. Not including the solvent effects in the calculations could be considered a limitation. However, a previous detailed study has shown that using the cam-b3lyp [27] /6-31+G(d,p) level in vacuum correctly reproduced the experimental trends and the ordering of the molecular orbitals in meso-(4-pyridyl)-BODIPYs [23].

The UV-vis absorption and emission data were calculated using the TD-DFT [28] /cam-b3lyp/6-31+G(d,p) method in vacuum. The range-separated cam-b3lyp functional was used to capture the long-range interactions in the transitions between the BODIPY core and the pyridyl ring. The first three singlet excitations were considered, and the lowest-energy excited singlet state was optimized to calculate the energies, atomic charges, dipole moments, maximum emission wavelengths, and the other properties reported in this manuscript. The potential energy minima of the ground and excited states were confirmed with frequency calculations.

All calculations were performed using the Gaussian 09 program package [29].

3. Results and Discussion

3.1. Photophysical Properties

The calculated photophysical properties of the entire series of meso-(2-, 3-, and 4-pyridyl)-BODIPYs are presented in Table 1. The experimentally measured data are given when available. The dominant transition for all studied BODIPYs is S₀→S₁. The next singlet excited state is more than 0.9 eV higher in energy and, therefore, has minimal contribution. A previously study from our laboratory [25,26] reported that the experimentally measured maximum absorption (${\lambda }_{abs}$) and emission (${\lambda }_{em}$) wavelengths for nPy, nPyNO₂, nPyCl₂, and nPy(CO₂Me)₂ are almost identical within a given n = 2, 3, 4 series bearing the same 2,6-substituents. It was determined that this was related to the almost identical HOMO-LUMO gaps of the ground states for a given series. As seen in Table 1, the newly calculated hypothetical compounds nPyCl, nPy(NO₂)₂, nPyCO₂Me, nPyCF₃, and nPy(CF₃)₂, where n = 2, 3, 4, show similar trends—${\lambda }_{abs}$ and ${\lambda }_{em}$ are almost identical within a given n = 2, 3, 4 series. It is hypothesized that such trends will be observed regardless of the 2,6-substituents on the meso-(2-, 3-, and 4-pyridyl)-BODIPYs.

As could be expected, the computationally determined ${\lambda }_{abs}$ and ${\lambda }_{em}$ are significantly blue-shifted, as a known tendency of TD-DFT methods to overestimate singlet-singlet excitation energies [30,31]. However, the similarities in the experimentally observed ${\lambda }_{abs}$ and ${\lambda }_{em}$ within a series correlate very well with the similarities in the TD-DFT computational values. In addition, the experimentally observed red-shifts upon chlorination and blue-shifts upon nitration are in agreement with the calculated ${\lambda }_{abs}$ and ${\lambda }_{em}$ values. All these suggest that the computed ${\lambda }_{abs}$ and ${\lambda }_{em}$ can be used to determine the observed trends in the experimental data.

As previously observed [25,26], the experimental fluorescence properties of meso-(2-, 3-, and 4-pyridyl)-BODIPYs largely depend on the relative position of the nitrogen atom on the pyridine ring. The meso-(2-pyridyl)-BODIPY derivatives consistently show the lowest fluorescence quantum yields within an n = 1, 2, 3 series. This trend is valid in all experimentally observed values and is regardless of the substituents at the 2,6-positions. We will attempt to explain this phenomenon in the next sections.

Table 1. Experimental and calculated photophysical properties of meso-(2-, 3-, and 4-pyridyl)-BODIPYs: maximum absorption wavelength (${\lambda }_{abs}$), oscillator strength (f), maximum emission wavelength (${\lambda }_{em}$), and quantum yield ($\Phi$).

meso-BODIPY	${\mathit{\lambda }}_{\mathit{a}\mathit{b}\mathit{s}}$ (nm)		f	${\mathit{\lambda }}_{\mathit{e}\mathit{m}}$ (nm)		Φ
	exp	calc		exp	calc
Phe	501 (498) a	414	0.539	505 (508) a	426	0.56 (0.65) a
2Py	502 b	416	0.542	514 b	441	0.04 b
3Py	502 b	416	0.542	514 b	429	0.43 b
4Py	501 b	415	0.543	515 b	428	0.31 b
PheCl		434	0.508		437
2PyCl		424	0.552		446
3PyCl		424	0.551		441
4PyCl		423	0.551		443
PheCl2		430	0.561		447
2PyCl2	530 b	432	0.565	545 b	455	0.17 b
3PyCl2	527 b	432	0.565	542 b	452	0.58 b
4PyCl2	528 b	432	0.566	546 b	454	0.58 b
PheNO2		405	0.659		425
2PyNO2	491 c	408	0.655	509 c	436	0.05 c
3PyNO2	491 c	407	0.656	507 c	429	0.25 c
4PyNO2	490 c	406	0.660	508 c	431	0.26 c
Phe(NO2)2		409	0.677		434
2Py(NO2)2		413	0.673		443	0.13
3Py(NO2)2		412	0.675		438	0.28
4Py(NO2)2		412	0.676		440	0.36
PheCO2Me		410	0.658		427
2PyCO2Me		413	0.658		441	0.21
3PyCO2Me		413	0.659		431	0.61
4PyCO2Me		412	0.661		435	0.39
Phe(CO2Me)2		410	0.791		426
2Py(CO2Me)2	501 c	413	0.797	515 c	436	0.09
3Py(CO2Me)2	501 b	413	0.793	512 c	430	0.61
4Py(CO2Me)2	500 b	412	0.794	510 c	434	0.43
PheCF3		407	0.594		425
2PyCF3		409	0.596		437
3PyCF3		409	0.597		429
4PyCF3		408	0.599		432
Phe(CF3)2		404	0.6605		424
2Py(CF3)2		406	0.6639		434
3Py(CF3)2		406	0.6645		428
4Py(CF3)2		406	0.667		418

^a From Ref. [32] in THF and Ref. [1] in methanol. ^b From Ref. [25] in acetonitrile. ^c From Ref. [26] in acetonitrile.

Table 1. Experimental and calculated photophysical properties of meso-(2-, 3-, and 4-pyridyl)-BODIPYs: maximum absorption wavelength (${\lambda }_{abs}$), oscillator strength (f), maximum emission wavelength (${\lambda }_{em}$), and quantum yield ($\Phi$).

meso-BODIPY	${\mathit{\lambda }}_{\mathit{a}\mathit{b}\mathit{s}}$ (nm)		f	${\mathit{\lambda }}_{\mathit{e}\mathit{m}}$ (nm)		Φ
	exp	calc		exp	calc
Phe	501 (498) a	414	0.539	505 (508) a	426	0.56 (0.65) a
2Py	502 b	416	0.542	514 b	441	0.04 b
3Py	502 b	416	0.542	514 b	429	0.43 b
4Py	501 b	415	0.543	515 b	428	0.31 b
PheCl		434	0.508		437
2PyCl		424	0.552		446
3PyCl		424	0.551		441
4PyCl		423	0.551		443
PheCl2		430	0.561		447
2PyCl2	530 b	432	0.565	545 b	455	0.17 b
3PyCl2	527 b	432	0.565	542 b	452	0.58 b
4PyCl2	528 b	432	0.566	546 b	454	0.58 b
PheNO2		405	0.659		425
2PyNO2	491 c	408	0.655	509 c	436	0.05 c
3PyNO2	491 c	407	0.656	507 c	429	0.25 c
4PyNO2	490 c	406	0.660	508 c	431	0.26 c
Phe(NO2)2		409	0.677		434
2Py(NO2)2		413	0.673		443	0.13
3Py(NO2)2		412	0.675		438	0.28
4Py(NO2)2		412	0.676		440	0.36
PheCO2Me		410	0.658		427
2PyCO2Me		413	0.658		441	0.21
3PyCO2Me		413	0.659		431	0.61
4PyCO2Me		412	0.661		435	0.39
Phe(CO2Me)2		410	0.791		426
2Py(CO2Me)2	501 c	413	0.797	515 c	436	0.09
3Py(CO2Me)2	501 b	413	0.793	512 c	430	0.61
4Py(CO2Me)2	500 b	412	0.794	510 c	434	0.43
PheCF3		407	0.594		425
2PyCF3		409	0.596		437
3PyCF3		409	0.597		429
4PyCF3		408	0.599		432
Phe(CF3)2		404	0.6605		424
2Py(CF3)2		406	0.6639		434
3Py(CF3)2		406	0.6645		428
4Py(CF3)2		406	0.667		418

^a From Ref. [32] in THF and Ref. [1] in methanol. ^b From Ref. [25] in acetonitrile. ^c From Ref. [26] in acetonitrile.

3.2. Electron-Density-Related Properties of the Ground and Excited States

The apparent difference between meso-(2-pyridyl)- and meso-(3-, and 4-pyridyl)-BODIPYs is the closer proximity of the nitrogen atom to the BODIPY core in the former compounds. The different positions of the nitrogen atom in the pyridyl group result in different strengths of the electron-withdrawing effect of the pyridyl nitrogen and different electron densities in these compounds. To evaluate these differences, we analyzed the shapes and the energies of the frontier molecular orbitals. We also analyzed the dipole moments and their changes upon excitation.

The shapes of the frontier molecular orbitals (Figure 3, Figure S1, Supplementary Materials) clearly demonstrate that the electron-withdrawing effect of the substituent featuring electron density spread to Cl, NO₂, CO₂Me, and CF₃ moieties, respectively. This is observed in both S₀ and S₁ states. However, regardless of the substituent, both HOMO and LUMO are almost entirely localized on the BODIPY core, indicating no significant charge-transfer effect and no drastic difference between meso-(2-pyridyl)-BODIPYs and meso-(3-, and 4-pyridyl)-BODIPYs. Thus, while the frontier molecular orbitals clearly indicate the effect of different electron-withdrawing substituents at the 2,6 positions, they cannot explain the drastic difference in the experimentally observed fluorescence quantum yield between meso-(2-pyridyl)-BODIPYs and their meso-(3-, and 4-pyridyl)-analogs.

The energies of the HOMO and LUMO for the ground states of all studied compounds are given in

Table 2. Calculated ground-state HOMO and LUMO energies, band gap ($E_{g,0}$), dipole moment (${\mu }_0$), and the scalar ($\Delta \mu ={\mu }_1-{\mu }_0$) and absolute value of the vector change ($\overrightarrow{\Delta \mu }=\left|\overrightarrow{{\mu }_1}-\overrightarrow{{\mu }_0}\right|$) of the dipole moment upon excitation for the series of meso-pyridyl-BODIPYs studied. The subscript 0 refers to the ground state and the subscript 1 refers to the first singlet excited state.

BODIPY	${\mathit{H}\mathit{O}\mathit{M}\mathit{O}}_0$ (eV)	${\mathit{L}\mathit{U}\mathit{M}\mathit{O}}_0$ (eV)	${\mathit{E}}_{\mathit{g},0}$ (eV)	${\mathit{\mu }}_0$ (D)	$\mathit{\Delta }\mathit{\mu }$ (D)	$\overrightarrow{\mathit{\Delta }\mathit{\mu }}$ (D)
Phe	−6.79	−1.62	5.17	5.03	0.07	0.07
2Py	−6.77	−1.62	5.14	6.05	0.03	0.05
3Py	−6.92	−1.78	5.14	3.94	0.02	0.08
4Py	−6.96	−1.82	5.15	2.13	0.02	0.07
PheCl	−6.94	−1.82	5.12	5.61	0.13	0.29
2PyCl	−6.91	−1.82	5.10	6.53	−0.08	0.10
3PyCl	−7.06	−1.97	5.09	4.53	−0.04	0.22
4PyCl	−7.10	−2.01	5.10	3.14	0.08	0.16
PheCl2	−7.07	−2.01	5.07	5.23	0.20	0.27
2PyCl2	−7.05	−2.00	5.05	6.22	0.16	0.58
3PyCl2	−7.19	−2.15	5.04	4.02	0.28	0.53
4PyCl2	−7.23	−2.19	5.04	2.30	0.14	0.14
PheNO2	−7.46	−2.21	5.25	9.04	−0.33	0.45
2PyNO2	−7.44	−2.22	5.21	9.51	−0.58	0.87
3PyNO2	−7.58	−2.37	5.21	8.02	−0.60	0.71
4PyNO2	−7.62	−2.40	5.22	7.54	−0.41	0.46
Phe(NO2)2	−8.01	−2.80	5.21	6.09	0.17	0.17
2Py(NO2)2	−7.98	−2.81	5.17	7.08	0.14	0.41
3Py(NO2)2	−8.13	−2.96	5.17	4.91	0.15	0.28
4Py(NO2)2	−8.17	−3.00	5.18	3.11	0.11	0.11
PheCO2Me	−7.03	−1.83	5.21	3.98	0.02	0.39
2PyCO2Me	−7.01	−1.83	5.18	4.86	−0.30	0.71
3PyCO2Me	−7.16	−1.99	5.17	3.02	−0.38	0.61
4PyCO2Me	−7.20	−2.02	5.18	1.92	−0.07	0.35
Phe(CO2Me)2	−7.24	−2.02	5.22	1.92	0.11	0.11
2Py(CO2Me)2	−7.21	−2.02	5.19	3.00	0.16	0.42
3Py(CO2Me)2	−7.36	−2.18	5.18	1.89	0.04	0.31
4Py(CO2Me)2	−7.40	−2.21	5.19	1.05	−0.03	0.03
PheCF3	−7.20	−1.97	5.24	6.78	−0.04	0.18
2PyCF3	−7.18	−1.97	5.21	7.56	−0.36	0.71
2PyCF3	−7.33	−2.13	5.21	5.85	−0.32	0.53
2PyCF3	−7.37	−2.16	5.21	4.87	−0.11	0.18
Phe(CF3)2	−7.57	−2.31	5.26	5.43	0.14	0.14
2Py(CF3)2	−7.54	−2.31	5.24	6.46	0.03	0.61
3Py(CF3)2	−7.69	−2.46	5.23	4.30	0.13	0.47
4Py(CF3)2	−7.73	−2.49	5.24	2.50	0.11	0.11

. As expected, both HOMO and LUMO are stabilized upon 2- and 2,6-substituion with electron-withdrawing groups, with the di-substituted compounds showing a greater effect. The stabilization is most pronounced in the case of nitration and least pronounced in the case of chlorination. All n = 2, 3, 4 series follow the same trends when compared to the meso-phenyl analogs. Meso-2-pyridyl substitution has a slight effect on HOMO (a destabilization by only 0.02–0.03 eV), while meso-(3- and 4-pyridyl)-substitutions stabilize HOMO by 0.1–0.2 eV. The effect of meso-(3- and 4-pyridyl)-substitution is almost identical but consistently features 0.04 eV lower HOMO energies in the case of meso-4-pyridyl-BODIPYs. The effect on LUMO is consistent stabilization, which is almost negligible in the case of meso-2-pyridyl substitution (0–0.1 eV) and significantly more pronounced in the case of meso-(3-, and 4-pyridyl)-substitution (0.14–0.16 eV). Again, the effect is slightly greater for meso-4-pyridyl than for meso-3-pyridyl. The analysis of the HOMO and LUMO energies demonstrates that the effect of meso-n-pyridyl substitution is similar for n = 3,4 but different in the case of n = 2. This might be related to the observed quantum yield differences.

Table 2 also lists the calculated dipole moments for the ground states and the scalar ($\Delta \mu ={\mu }_1-{\mu }_0$) and vector ($\overrightarrow{\Delta \mu }=\left|\overrightarrow{{\mu }_1}-\overrightarrow{{\mu }_0}\right|$) changes in the dipole moment upon excitation, where the subscript 0 refers to the ground state and subscript 1 refers to the first singlet excited state of the respective compound. Analysis of the magnitude of the dipole moment (${\mu }_0$) indicates that the meso-2-pyridyl-BODIPYs are consistently the most polar and the meso-4-pyridyl-BODIPYs are consistently the least polar BODIPYs within a given n = 2, 3, 4 series. The meso-2-pyridyl-BODIPYs are also consistently more polar than their meso-phenyl analogs. These two trends exist in both the S₀ and S₁ states, regardless of the 2,6-substituent. Interestingly, upon excitation, the dipole moment decreases in the mono-2-substituted BODIPYs but increases in the 2,6-di-substituted BODIPYs. The change in polarity upon excitation depends on the 2,6-substituent—it is most pronounced for nitration and least pronounced for chlorination. Whether this is related to the fact that chlorination increases the experimental quantum yield, while nitration decreases it, is currently under study in our laboratory.

Next, we look at the orientation of the dipole moment. Since nitrogen is more electronegative than carbon, the different position of the pyridyl nitrogen will create a different orientation of the dipole moment vector. Figure 4 shows a side view of the dipole moment vectors for the 2,6-dichlorinated and 2,6-dinitrated meso-(2-, 3-, and 4-pyridyl)-BODIPYs, as two examples of the effect. The meso-phenyl-BODIPY analogs and the unsubstituted BODIPYs are included for comparison. All compounds were optimized without symmetry constrains; however, for simplicity, we will refer to pseudo symmetry elements to explain the vector orientations. As with all di-substituted BODIPYs, the dipole moments lie in the pseudo mirror plane of symmetry (if the molecule was symmetric), i.e., the plane of the 2D representation in Figure 4. The meso-(4-pyridyl)-BODIPYs have their dipole moments oriented along a pseudo C₂ axis (if the molecule was symmetric). The meso-(2-, and 3-pyridyl)-BODIPYs have their dipole moments oriented at an angle of roughly 30° for meso-(2-pyridyl)-BODIPYs and roughly 60° for meso-(3-pyridyl)-BODIPYs with respect to the pseudo C₂ axis.

According to a recent study of a series of conjugated terpyridine derivatives [22], the fluorescence quantum yield could be quantified by the change in dipole moment between the ground and excited states. To check the applicability of this approach for our series, we calculated the absolute value of the vector change in the dipole moment ($\overrightarrow{\Delta \mu }=\left|\overrightarrow{{\mu }_1}-\overrightarrow{{\mu }_0}\right|$) upon excitation. These data are presented in the last column of Table 2. For almost all 2,6-substituents, the $\overrightarrow{\Delta \mu }$ value suggests the greatest quantum yield for meso-(2-pyridyl)-BODIPYs, which contradicts the experimental findings. The observed significant decrease in the fluorescence of meso-(2-pyridyl)-BODIPYs must be due to a different reason.

3.3. Rotational Barrier in the Excited States

As mentioned above, the major difference between the meso-(2-pyridyl) and meso-(3- and 4-pyridyl)-BODIPYs is the closer proximity of the nitrogen atom of the pyridine group to the BODIPY core in the former compounds. In addition to the different strengths of the electron-withdrawing effect of the pyridyl nitrogen and the different electron densities, the different positions of the nitrogen atom result in different steric effects. In meso-(2-pyridyl) compounds, position 2 is occupied by a nitrogen atom but in meso-(3- and 4-pyridyl)-BODIPYs it is occupied by a carbon atom bonded to a hydrogen atom. Because of the methyl groups in the 1,7-positions for the entire series studied, this could affect the rotation of the pyridyl ring. We hypothesized that the 2-pyridyl ring might be easier to rotate than the 3- and 4-pyridyl rings.

Analysis of the geometries of the ground states shows that the meso-pyridyl ring is oriented approximately perpendicular with respect to the BODIPY core ($\approx 90°$) for all but the mono- and di-nitro compounds (Table S1, Supplementary Materials). This orientation is consistent with the experimental findings for the crystal structure of the previously synthesized BODIPYs from the series [26]. The slight difference in the case of 2PyNO₂, 3PyNO₂, and 4PyNO₂ is also consistent with the experiment and could be related to the lower fluorescence quantum yields observed in the case of these compounds compared to the other BODIPYs from the series, which is likely due to partial delocalization of LUMO to the pyridyl ring (Figure 3).

Similar analysis in the case of the excited states demonstrates that the meso-pyridyl ring forms different dihedral angles with the BODIPY core that do not necessarily follow a trend (Table S1, Supplementary Materials). For this reason, we decided to concentrate on the analysis of the excited states and the differences in the rotation of the 2-, 3-, and 4-pyridyl rings upon excitation.

We modeled the rotation of the meso-pyridyl group and estimated the rotational barrier in the excited states in the cases of the 2,6-unsubstituted meso-(2-, 3-, and 4-pyridyl)-BODIPYs. The results of the scan are shown in Figure 5. In the case of meso-(3-pyridyl and 4-pyridyl)-BODIPYs, the potential energy surfaces of the meso-rotation in the excited states are very flat. All energies required for the rotation up to 40° are lower compared to the roughly 20 kcal/mol barrier that could be overcome at room temperature. Therefore, we believe that, at room temperature, the meso-pyridyl ring rotates freely (Figure 5a). In the case of 3Py and 4Py, this rotation causes gradual structural changes. The BODIPY core becomes slightly distorted with the gradually increasing contribution of the pyridyl ring to LUMO, as the meso-dihedral angle, $\theta$, decreases. This is reflected in a gradual red-shift in the maximum emission wavelength, ${\lambda }_{em}$, and a slight decrease in the oscillator strength (Figure 5b,c). In the case of 2Py, however, when $\theta$ approaches 40°, the structure of the molecule becomes significantly distorted (Figure 6). The BODIPY core is no longer planar and the 2Py ring becomes tilted. The LUMO drops by 0.7 eV, ${\lambda }_{em}$ shifts by almost 80 nm, and the oscillator strength drops significantly (Figure 5b,c).

An explanation for the significant drop in the oscillator strength can be seen in Figure 7. When $\theta =90°$, almost the entire electron density is localized on the BODIPY core and the molecule will fluoresce. However, as the meso-dihedral angle, $\theta$, decreases, the electron density becomes delocalized onto the pyridyl ring, indicating a gradual increase in the charge transfer effect that lowers the fluorescence. This is observed until $\theta$ reaches $40°$, when the distorted structure bearing a non-planar, non-fluorescent BODIPY core and tilted pyridyl ring becomes the lowest-energy conformation. We propose that this effect is the reason for the experimentally observed drastic decrease in the fluorescence of 2Py. Indeed, similar distorted structures exist in the case of 3Py and 4Py; however, they are very high in energy, 31 kcal/mol and 61 kcal/mol, respectively, compared to the roughly 20 kcal/mol barrier that could be overcome at room temperature. Therefore, these structures do not significantly affect the fluorescence of 3Py and 4Py to the same extent as for 2Py, where the distorted structure is located only 15 kcal/mol above the $\theta =90°$ minimum.

4. Conclusions

The performed computational modeling of a series of meso-(2-, 3-, and 4-pyridyl)-BODIPYs and their meso-phenyl analogs allows for several conclusions to be drawn.

Electron-withdrawing substituents at the 2,6-positions lower both HOMO and LUMO and affect both the energies and the shapes of the molecular orbitals. The calculated HOMO-LUMO gaps correlate with the experimentally observed red- or blue-shifts in the maximum absorption and emission wavelengths.

The analysis of the HOMO and LUMO energies demonstrates that the effect of meso-pyridyl substitution is significantly more pronounced for the 3- and 4-pyridyl-BODIPYs compared to their 2-pyridyl analogs.

The excited states are consistently less polar than the ground states in all 2-mono-substituted BODIPYs. The change in polarity upon excitation depends on the 2,6-substituent(s)—it is most pronounced for nitration and least pronounced for chlorination.

The experimentally observed consistent low quantum yields for the meso-(2-pyridyl) BODIPYs could be due to the very flat potential energy surfaces with respect to rotation of the 2-pyridyl ring in the excited states, and the stability of a crooked, non-planar, and significantly less fluorescent meso-(2-pyridyl) excited state structure.

The results from this study might be used to investigate other fluorescence quenching effects in fluorophores that might be due to structural changes in the excited states. Scanning the PES of the excited states of other BODIPYs might provide more insight into the mechanism of fluorescence quenching when no charge-transfer quenching occurs. It would also be interesting to account for the low-energy vibrational soft modes and examine their effects on the absorption and emission spectra of the compounds.