Electronic Excited State Redox Properties for BODIPY Dyes Predicted

Article pubs.acs.org/JPCA

Electronic Excited State Redox Properties for BODIPY Dyes Predicted from Hammett Constants: Estimating the Driving Force of Photoinduced Electron Transfer Richard Lincoln, Lana E. Greene, Katerina Krumova, Zhutian Ding, and Gonzalo Cosa* Department of Chemistry and Center for Self-Assembled Chemical Structures (CSACS-CRMAA), McGill University, 801 Sherbrooke Street West, Montreal, Quebec H3A 0B8, Canada S Supporting Information *

ABSTRACT: Here we formulate equations based solely on empirical Hammett substituent constants to predict the redox potentials for the electronic excited state of boron-dipyrromethene (BODIPY) dyes. We utilized computational, spectroscopic, and electrochemical techniques toward characterizing the effect of substitution at the positions C2, C6, and C8 of the 1,3,5,7-tetramethyl BODIPY core. Working with a library of 100 BODIPY dyes, we found that highest occupied molecular orbital (HOMO) energies calculated at the B3LYP 6-31g(d) level correlated linearly with the Hammett σm value for substituents at position C8 and with Hammett σp values for substituents at positions C2 and C6. In turn, we observed that LUMO energies correlated linearly with Hammett σp at position C8 and with Hammett σm at positions C2 and C6. Focusing on a subset of 26 dyes for which reduction potentials were either previously available or measured herein and ranged from −1.84 to −0.52 V (a full 1.3 V), we found a linear relationship between redox potentials in acetonitrile and HOMO and lowest unoccupied molecule orbital (LUMO) energies determined via density functional theory (DFT). A linear correlation was thus ultimately established between redox potentials in acetonitrile and Hammett substituent constants. Combining this with equations derived for the linear relationship existing between the zero vibrational energy of the excited BODIPY and Hammett substituent constants enabled us to provide the parameters toward predicting the oxidizing/reducing power of photoexcited 1,3,5,7,-tetramethyl BODIPY dyes in their singlet excited state.

■

INTRODUCTION Over the past decades, the exploitation of photoinduced electron transfer (PeT)1 has played a pivotal role in the development of photocatalytic systems,2,3 dye sensitized solar cells,4 luminescence based sensors and switches,1,5 and logic gates.6 In this context significant research efforts have been dedicated toward developing new chromophores with the appropriate redox and optical properties to ensure the feasibility of PeT in the presence of a redox partner of interest. Careful tuning of redox properties in chromophores is particularly critical toward developing sensitive, luminescence based sensors relying on PeT. These sensors typically consist of a reporter (fluorescent chromophore) segment and a receptor or trap segment.1 Subtle changes in the latter segment following reaction with the analyte may reposition its highest occupied (HOMO) and lowest unoccupied molecule orbital (LUMO) energy levels with respect to those of the reporter, deactivating an otherwise effective, PeT mediated, intramolecular emission quenching. Initially nonemissive, the probe is rendered luminescent after reaction/trapping of the analyte takes place (fluorogenic probe, see also Scheme 1). Importantly, the choice of the receptor or trap segment utilized in preparing PeT based fluorogenic probes must ensure maximum specificity toward the analyte of interest. There is little to no room to tune its redox properties, as changes in the © XXXX American Chemical Society

chemical structure of the receptor may jeopardize its selectivity. Given that the receptor/trap may not be chemically modified, one is left with choosing a reporter with appropriate redox properties in order to ensure that sensitivity constraints are fulfilled. In other words, ensuring that the receptor−reporter probe transitions from the off to the on state upon reaction with the analyte of interest at the receptor segment. The above discussion is particularly true when developing probes where the receptor/trap segment emulates the chemical reactivity of an existing natural compound, as is the case for the fluorogenic α-tocopherol analogues we have pioneered.7−9 We and others have exploited boron-dipyrromethene (BODIPY) dyes as effective reporters in two-segment receptor−reporter fluorogenic probes relying on PeT.10−16 The ease of preparation of BODIPY dyes enables readily tuning their spectral and redox properties via convenient substitution along their backbone. A major obstacle in designing new luminescent sensors, however, lies in our current limitations to estimate the redox potential of BODIPY dyes, and their Special Issue: Current Topics in Photochemistry Received: June 13, 2014 Revised: July 26, 2014

A

dx.doi.org/10.1021/jp5059148 | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article

Scheme 1. Schematic Representation of Photoinduced Electron Transfer (PeT) Mechanism Operating in the Molecular Switching of Fluorogenic BODIPY Dyes1,10a

a

Emission quenching can occur via electron transfer from the HOMO energy level of an electron donor (quencher) segment to the semi-occupied HOMO of the photoexcited dye (see left panel). Alternatively, emission can occur by electron transfer from the semi-occupied LUMO energy level of the photoexcited dye to the LUMO of an electron acceptor segment (quencher, see right panel). Fluorescence is induced/restored upon deactivation of PeT pathways via chemical modification of the quencher (typically referred to as a receptor or trap segment, see middle panel) yielding an emissive product.

been reported for other systems, for example, the studies on 1,3,5-triarylpyrazoline backbone reported by Fahrni et al.23 Herein, we investigate the effect substituents at C2, C6, and C8 have on both HOMO and LUMO energy levels as derived from quantum mechanical calculations for a library of 100 dyes (Figure 1). Furthermore, we correlate the electrochemically determined redox potentials in acetonitrile with DFT results for a smaller subset of dyes. Combined, these results provide simple equations where from to estimate the redox potentials of BODIPY dyes in their first excited singlet state and thus the exergonic character of PeT processes for this family of BODIPY dyes.

HOMO−LUMO energy gap, which when combined, enables us to estimate the reducing/oxidizing character of photoexcited BODIPY dyes. Although density functional theory (DFT) calculations have proven to be invaluable tools during the design of PeT systems,17 some shortcomings can arise when relying on computation. For example, larger systems require longer computation times, and accurate orbital energies for dyes containing charged functional groups require an understanding and evaluation of solvation effects.18 We sought to find an empirical method to predict the driving force for PeT reactions with a given partner based on the BODIPY substitution pattern. We thus resorted to Hammett substituent constants to estimate the standard redox potential of the BODIPY radical cation (E°B+•/B) and the standard redox potential of BODIPY (E°B/B−•). These potentials, when combined with the vibrational zero electronic energy of the excited BODIPY (ΔE0,0), the redox potential of the partner of choice, and the electrostatic work term that accounts for the Coulombic attraction of the contact ion pair (ϖ) would enable us to estimate the reductive/oxidative power of photoexcited BODIPYs as a function of their substitution pattern upon calculating the Gibbs energy of photoinduced electron transfer via eq 1.19 ΔG°eT = [(E°D+• /D − E° A/A−•) − ΔE0,0 + ϖ]

■

EXPERIMENTAL SECTION Materials. 8-Acetoxymethyl-2,6-diethyl-1,3,5,7-tetramethylpyrromethene fluoroborate (PM605) was purchased from Exciton, Inc. (Dayton, OH). HPLC grade solvents for spectroscopy and column chromatography were purchased through Fisher Scientific. All other chemicals were supplied by Sigma-Aldrich, Co. and used without further purification. Synthesis. 2,6-Diethyl-1,3,5,7,8-pentamethyl-pyrromethene fluoroborate (2c);24,25 1,3,5,7,8-pentamethyl-pyrromethene fluoroborate (2d);25,26 2-chloro-1,3,5,7,8-pentamethyl-pyrromethene fluoroborate (2f) and 2,6-dichloro-1,3,5,7,8-pentamethyl-pyrromethene fluoroborate (2g);27 2,6-diethyl-1,3,5,7tetramethyl-pyrromethene fluoroborate (4c) and 1,3,5,7tetramethyl-pyrromethene fluoroborate (4d);28 and 2,6-diiodo-1,3,5,7-tetramethyl-pyrromethene fluoroborate were prepared as previously described in the literature.29 2-Chloro-1,3,5,7-tetramethyl-pyrromethene Fluoroborate, (4f). 1,3,5,7-Tetramethyl-pyrromethene fluoroborate 4d (0.10 g, 1 equiv) was dissolved in dry THF (5 mL) under argon at −78 °C. N-Chlorosuccinimide (0.11 g, 2.2 equiv) was added in portions. The reaction mixture was stirred for 15 min after which it was removed from the ice bath. The reaction mixture was stirred for an additional 3 h at room temperature, then quenched upon dilution with ethyl acetate and washed with water. The organic fraction was extracted and dried with anhydrous magnesium sulfate. The solvent was removed under reduced pressure. The solid residue was purified by silica flash chromatography and eluted with hexanes/dichloromethane.

(1)

We reasoned that if a relationship could be established between Hammett substituent constants and DFT calculated orbital energies, and in turn a relationship was established between orbital energies and both redox potentials, as well as zero vibrational electronic energies for the excited state, we could prepare formulas for the prediction of excited state redox potentials as a function of Hammett susbstituent constants. These formulas would ultimately allow us to estimate the reductive/oxidative power of photoexcited BODIPY dyes toward photoinduced electron transfer reactions based solely on the BODIPY substitution pattern. Previous studies have observed trends in the photophysical behavior of BODIPY dyes in relation to the Hammett substituent constants;20−22 yet to date, no one has conducted an in-depth investigation of the substituent effects on the BODIPY core. We note, however, that thorough studies have B



Article

flash chromatography and eluted with dichloromethane. Compound 4g was obtained as a red powder (0.16 g, 72%). 1 H NMR (500 MHz; CDCl3): δ 7.04 (s, 1H), 2.54 (s, 6H), 2.23 (s, 6H). 13C NMR (126 MHz; CDCl3): δ 153.6, 136.4, 131.0, 121.1, 120.8, 12.4, 9.6. HRMS (APCI) for C13H14BCl2F2N2 (M+) calcd 317.0592, found 317.0578. Instrumentation. Absorption spectra were recorded using a Hitachi U-2800 UV−vis−NIR spectrophotometer. Luminescence spectra were recorded using a PTI QuantaMaster spectrofluorimeter using 1 cm × 1 cm quartz cuvettes and corrected for detector sensitivity. 1H NMR and 13C NMR spectra were recorder on a Varian VNMRS 500 instrument at 500 and 125 MHz, respectively. Atmospheric pressure chemical ionization (APCI) mass spectra were measured on a Bruker maXis impact. Voltammetric experiments were conducted with a computer-controlled CHI760C potentiostat. Fluorescence Quantum Yield. Quantum yields of fluorescence were measured using PM605 in acetonitrile as a reference. Absorption and emission spectra of PM605 and the dye of interest were measured in acetonitrile at five different concentrations. The integrated intensity versus absorbance were then plotted and fitted linearly. Relative quantum yields of fluorescence for the unknown with respect to the standard were obtained from eq 2, where Φ, Δ, and n refer, respectively, to the quantum yield, the slope obtained from the above-mentioned plot, and the solvent refractive index for the unknown (x) or standard (st). ϕx = ϕst ×

Δx n2 × x2 Δst nst

(2)

Electrochemical Studies. Electrochemical measurements were performed using a three-electrode system. The working electrode was a Pt wire, a Pt mesh wire was used as the counter electrode, and a Ag/AgCl electrode was used as the reference. A 0.1 M solution of tetrabutylammonium hexafluorophosphate in dry acetonitrile was used as the electrolyte solvent in which the compounds were dissolved to 1 mM or to saturation when the latter occurred before reaching a 1 mM concentration. Concentration was determined from their absorption and calculated extinction coefficient. Solutions also contained ferrocene with a concentration of 1 mM as an internal standard. The solutions were equilibrated with argon, and all measurements were conducted under inert atmosphere, with a minimum scan rate of 0.2 V s−1. When the acquired cyclic voltammograms were not reversible we increased the scan rate up to 1 V s−1. Formal redox potentials were calculated from the midpoint of the cathodic and anodic peak potentials observed in the cyclic voltammograms. All values were reported vs ferrocene, with the oxidation of ferrocene measured and corrected to zero for all experiments. Fluorescence Lifetime Studies. Fluorescence lifetime measurements were carried out using a Picoquant Fluotime 200 Time Correlated Single Photon Counting (TCSPC) setup employing an LDH 470 diode laser from Picoquant as the excitation source. The laser output was at 466 nm. The excitation rate was 10 MHz, and the laser power was adjusted to ensure that the detection frequency was less than 100 kHz. The laser was controlled by a PDL 88 B picosecond laser driver from Picoquant. Photons were collected at the magic angle of 54.7°. The IRF was 200 ps for our setup. Vibrational Zero Electronic Energy of the Excited BODIPY. The ΔE0,0 values were determined in acetonitrile

Figure 1. (A,B) Functionalized 1,3,5,7-tetramethyl BODIPY scaffold used for DFT calculations of HOMO and LUMO energies. Positions 2 and 6 (equatorial) and 8 (meso) are shown in yellow/green. (C) Table listing the various substitution patterns for the BODIPY core explored in our DFT calculations. Gray-filled squares indicate dyes either prepared as part of this work or recently reported and characterized by spectroscopic and electrochemical techniques.

Compound 4f was obtained as an orange powder (0.040 g, 35%). 1H NMR (500 MHz; CDCl3): δ 7.02 (s, 1H), 6.08 (s, 1H), 2.54 (s, 3H), 2.52 (s, 3H), 2.24 (s, 3H), 2.20 (s, 3H). 13C NMR (126 MHz; CDCl3): δ 159.2, 150.9, 143.0, 134.8, 134.2, 130.2, 120.4, 119.83, 119.81, 119.79, 14.8, 12.1, 11.3, 9.5. HRMS (APCI) for C13H15BClF2N2 (M+) calcd 283.0982, found 283.0978. 2,6-Dichloro-1,3,5,7-tetramethyl-pyrromethene Fluoroborate (4g). 1,3,5,7-Tetramethyl-pyrromethene fluoroborate 4d (0.18 g, 1 equiv) was dissolved in dry THF (10 mL) under argon at −78 °C. N-Chlorosuccinimide (0.36 g, 4 equiv) was added in portions. The reaction mixture was stirred for 15 min after which it was removed from the ice bath. The reaction mixture was stirred for an additional 24 h at room temperature, then quenched upon dilution with ethyl acetate and washed with saturated NaCl. The organic fraction was extracted and dried with anhydrous sodium sulfate. The solvent was removed under reduced pressure. The solid residue was purified by silica C



Article

Figure 2. Left column: Correlation surfaces of DFT derived energies (B3LYP 6-31g(d)) for the HOMO (εHOMO, blue circles) and LUMO (εLUMO, red triangles) of the 1,3,5,7-tetramethyl BODIPY core as a function of the Hammett substituent constants for functional groups at RI (meso), and RII and RIII (equatorial). Four possible combinations of Hammett substituent constants exist. Panel A: σp for RI, σp for RII and RIII. Panel B: σm for RI, σp for RII and RIII. Panel C: σm for RI, σm for RII and RIII. Panel D: σp for RI, σm for RII and RIII. Right column: One-dimensional residuals along σRI and (σRII + σRIII) following fitting according to eqs 3 and 4 for the HOMO and LUMO energies, respectively.

dyes that may be constructed upon combining the substitutions listed in Figure 1 (see the Supporting Information section for the values, reported relative to vacuum). Calculations were conducted at the B3LYP 6-31g(d)18 level to provide accurate results in relatively low computation time. Our work focused on substitution at C2 and C6 (equatorial) and C8 (meso) positions of the 1,3,5,7-tetramethyl BODIPY backbone, for which a plethora of synthetic literature is available for modifying the scaffold. This is our preferred BODIPY backbone when developing chemoselective fluorgenic probes.7−9,31 The added sterics of the methyl groups at positions C1 and C7 help

from the intercept of the normalized absorption and emission spectra. Computational Methods. All quantum mechanical calculations were performed using the Gaussian 09 package.30 HOMO and LUMO orbital energies were determined from molecular geometries optimized at the B3LYP 6-31G(d) level.18 Orbitals were visualized using the Gaussview 5 package.

■

RESULTS AND DISCUSSION Calculated Orbital Energies. We first used DFT to calculate the HOMO and LUMO energies for the 100 BODIPY D



Article

Table 1. Best Fitting Parameters Obtained for the Correlation of Hammett Substituent Constants with HOMO and LUMO Energy Levels Calculated via DFT (B3LYP 6-31g(d)) HOMO LUMO

σRI

σRII + σRIII

ρRI

ρRII,III

c

R2

σm σp

σp σm

−0.56 ± 0.02 −0.96 ± 0.03

−0.680 ± 0.008 −0.79 ± 0.03

−5.473 ± 0.006 −2.48 ± 0.01

0.98823 0.95215

prevent thermal relaxation of C8 functional groups.32−34 They also prevent conjugation from amino groups at position C8 that would otherwise yield hemicyanines (Figure 1C, first column).35 Additionally, methyl groups at positions C3 and C5 block undesired pyrrole polymerization during synthesis,36 and serve as modifiable positions via Knoevenagel condensation.11,37 Substitution at the available positions ranged from strongly electron donating (amine) to electron withdrawing (nitrile) groups as reflected by their characteristic Hammett substituent constants that ranged from σp of −0.66 to 0.66, respectively.38 The specific functional groups utilized in our DFT calculations and tabulated in Figure 1 were selected from the recent literature to ensure synthetic routes exist for their installation.31,35,39−42 Of note, not all of the substituent combinations that we calculated have been reported. Finally, only functional groups with reported Hammett substituent constants for both meta (σm) and para (σp) substituted benzoic acid were chosen,38 to allow distinction between resonance and inductive effects on orbital energies (see below). Correlation surfaces were next obtained by plotting the orbital energies (both εHOMO and εLUMO) versus the Hammett substituent constants for RI (σRI) and the sum of contributions from RII and RIII (σRII + σRIII) (Figure 2). Four possible combinations were considered, as each substituent has both a σp and σm value. To determine the most reliable combination of Hammett substituent constants for orbital energy prediction, each data set was fit with a plane according to eqs 3 and 4, where ρRI and ρRII,III represent the orbital energy sensitivity to modification at the respective positions and where c is a constant. εHOMO = (ρ RI )(σ RI) + (ρ RII,III )(σ RII + σ RIII) + c

(3)

εLUMO = (ρ′RI )(σ RI) + (ρ′RII,III )(σ RII + σ RIII) + c′

(4)

It is important to mention that regardless of which Hammett substituent constants were used for the fitting, the LUMO was always found to be significantly more sensitive than the HOMO to changes at RI. With the best correlation, ρRI values of −0.96 and −0.56 were obtained for the LUMO and HOMO, respectively. The negative sign of the ρ values we obtained is in line with the expected orbital stabilization (lower energy) resulting from more electron withdrawing groups (larger σ value) at the meso position.43 The sensitivity of the equatorial positions to substitution is similar for both energy levels, albeit larger for the LUMO (−0.79) than the HOMO (−0.68). These results are consistent with the shape of the molecular orbitals and the position of the nodal planes and allow for tuning of the HOMO−LUMO energy gap. Electrochemistry of BODIPY Dyes. We next sought to connect the DFT derived orbital energies (εHOMO and εLUMO) with observed ground state standard redox potentials (E°B+•/B and E°B/B−•) following the method derived by Moore and coworkers,18 according to the linear relationship below (eq 5). E°B+• /B(E°B/B−•) = a + bεHOMO(εLUMO)

(5)

Values of E°B /B and E°B/B were recorded in electrochemical studies that focused on a subset of 26 dyes tabulated in Figure 1 (gray shaded). The redox properties for 18 of these dyes were recently reported by us.31 To ensure a sufficiently large sample set, we expanded here our library of electrochemical redox potentials by preparing an additional eight dyes bearing either a hydrogen atom or an electron-donating methyl group at RI (Figure 3). Six of the dyes were prepared according +•

Table 1 lists the fitting parameters (ρ and c values) we obtained for the best possible correlations (see Supporting Information for the fitting results for all possible correlation surfaces). Although plotting orbital energies as a function of σp at all positions gave fairly good fitting results for both the HOMO and LUMO (r2 of 0.98190 and 0.94881 for the HOMO and LUMO, respectively), the best fitting results for the HOMO were achieved when plotted against σm for RI and the sum of σp for RII and RIII (r2 of 0.98823). Conversely, the best fit for the LUMO was achieved when plotted against σp for RI and the sum of σm at RII and RIII (r2 of 0.95215). This observation goes in line with the shape of the molecular orbitals for an unsubstituted BODIPY, see Figure 1B. The HOMO of a BODIPY contains a nodal plane at the C8 (RI) position and would therefore be much less sensitive to resonance effects, which are captured in σp but not in σm for substituents at RI. Relative to the HOMO, the LUMO contains significantly more electron density at C8 (RI) than at the equatorial C2 or C6 (RII and RIII). Therefore, the LUMO is more sensitive to resonance from substituents at RI than RII or RIII.

−•

Figure 3. Structure of the BODIPY dyes prepared and electrochemically characterized in this work. All dyes where prepared according to literature procedures, except 4f and 4g whose synthesis we report for the first time.

to recently reported procedures.24−28 The remaining two dyes, 4f and 4g, were prepared by chlorination of the unsubstituted BODIPY core (4d) with N-chlorosuccinimide (see the Experimental Section). Reversible redox potentials were obtained for all four of the meso-methyl BODIPY dyes. For the meso-hydrogen series, reversible oxidation events were recorded for all dyes studied; however, the reduction of the chlorinated dyes (4f and 4g) was not reversible even at the increased scan rate of 1 V s−1. The E



Article

Table 2. Measured Electrochemical Data, HOMO and LUMO Orbital Energies, and Photophysical Properties Determined in Acetonitrile unless Otherwise Indicated electrochemical dataa (V vs Fc+•/Fc) E°B/B−• 2c 2d 2f 2g 3cb 3db 3fb 3gb 3ib 3jb 4c 4d 4f 4g 5cb 5db 5fb 5gb 5ib 5jb 8cb 8db 8fb 8gb 8ib 8jb

Epc

−1.84c −1.73 −1.64c −1.38 −1.58 −1.48 −1.36 −1.23 −1.24

E°B+•/B 0.60c 0.77 0.85c 0.90c 0.63 0.77 0.82 0.92c

1.15 1.39

−0.90 −1.81 −1.55 −1.46c −1.32c −1.48 −1.44 −1.30 −1.18c

0.60 0.74 0.83c 0.91c 0.70 0.75c 0.93c 0.99c

−1.24 −0.77 −1.12 −1.05 −0.97 −0.89 −0.82 −0.52

Epa

1.31 1.39 0.73 0.86 1.07 1.05 1.23 1.42

calculated orbital energies

photophysical properties

εHOMO (eV)

εLUMO (eV)

abs λmax (nm)

em λmax (nm)

Φf

τdec (ns)

ε × 10−3 (M−1 cm−1)

ΔE0,0 (eV)

−5.19 −5.36 −5.54 −5.70 −5.32 −5.50 −5.66 −5.81 −6.01 −6.45 −5.24 −5.38 −5.57 −5.75 −5.31 −5.49 −5.65 −5.80 −5.99 −6.44 −5.44 −5.63 −5.79 −5.94 −6.14 −6.58

−2.22 −2.31 −2.53 −2.74 −2.50 −2.58 −2.80 −3.00 −3.08 −3.56 −2.30 −2.42 −2.63 −2.83 −2.51 −2.61 −2.81 −3.01 −3.10 −3.57 −2.83 −2.98 −3.17 −3.34 −3.44 −3.89

514 491 501 514 536 510 521 537 498 509 525 501 511 526 542 515 526 543 503 515 535 509 520 538 499 510

528 498 514 530 552 523 535 554 517 525 530 503 519 534 561 530 546 560 526 531 N.A. N.A. N.A. N.A. N.A. N.A.

0.84 0.81 0.85 0.69 0.84 0.98 0.82 0.79 0.87 1.00 0.94 0.91 0.87 0.78 0.72 0.87 0.88 0.68 0.86 0.76 N.A. N.A. N.A. N.A. N.A. N.A.

6.24 5.61 6.02 6.14 6.90 6.65 6.31 6.46 4.65 5.05 6.36 5.50 5.99 6.14 6.76 6.66 6.31 6.46 4.68 5.24 N.A. N.A. N.A. N.A. N.A. N.A.

83 97 73 66 70 100 70 53 51 94 96 115 66 42 70 81 66 58 46 96 16 18 43 16 57 42

2.41 2.53 2.45 2.38 2.28 2.41 2.34 2.28 2.45 2.41 2.35 2.48 2.42 2.38 2.26 2.38 2.32 2.25 2.42 2.38 N.A. N.A. N.A. N.A. N.A. N.A.

a E°B/B−• = reversible reduction potential of the BODIPY dye; Epc = cathodic peak potential; E°B+•/B = reversible reduction potential of the BODIPY radical cation; Epa = anodic peak potential. bTaken from ref 31. cAcquired at 1000 mV s−1.

redox potentials and photophysical properties for the 26 dyes are listed in Table 2. When reversible redox potentials were not available, we listed the anodic (Epa) or cathodic (Epc) peak potential instead of E°B+•/B and E°B/B−•, respectively. For the meso-formyl BODIPY dyes (compounds 8c, 8d, 8f, 8g, 8i, and 8j), we reassigned the two reversible reduction waves previously reported by us31 based on the ordering of the unoccupied orbitals as calculated via DFT. The calculations showed that the LUMO is centered on the BODIPY and that the LUMO + 1 is centered on the carbonyl. The higher reduction potentials listed in our previous work31 were thus used when correlating BODIPY reduction with computationally derived orbital energies. An excellent linear correlation was observed for the experimentally measured redox potentials when plotted against the corresponding DFT-calculated orbitals (Figure 4). Fitting according to eq 4 resulted in slope and intercept values of −0.774 ± 0.007 and −3.50 ± 0.03, respectively (r2 = 0.995). The presence of anodic or cathodic peak potentials did not alter the overall fitting (see Supporting Information). Inspection of Figure 4 reveals that the redox properties of BODIPY dyes are extremely sensitive to substitution pattern. Values of E°B/B−• ranged from −1.84 to −0.52 V, a full 1.3 V. For E°B+•/B, values ranged from 0.60 to 1.42 V, a ca. 0.8 V spread in values. Vibrational Zero Electronic Energy of the Excited State. The combined results from the correlation of DFT calculations versus Hammett substituent constants and the

Figure 4. Correlation of redox potentials determined by cyclic voltammetry (E°B+•/B and E°B/B−•) with orbital energies calculated via DFT (εHOMO and εLUMO). Filled red triangles and blue circles correspond to E°B/B−• and E°B+•/B, respectively. Unfilled red triangles and blue circles indicate cathodic and anodic peak potentials, respectively. Shown in watermark is a cyclic voltammogram of a BODIPY dye measured in acetonitrile to illustrate the relationship between experimental and theoretical (DFT) parameters; the reversible peak at 0 V corresponds to the reversible oxidation of the ferrocene standard (1 mM).

correlation of redox potentials versus DFT calculations enabled us to establish the reducing/oxidizing power for ground state BODIPYs based on their substitution at C2, C6, and C8. In order to learn about the redox properties of BODIPYs in their excited state, it was necessary to determine the vibrational zero electronic energy for their excited states (ΔE0,0, see eq 1) and F



Article

To demonstrate the predictive power of our determined relationships, we prepared and characterized an iodine substituted dye (Figure 5). The iodine atom is not defined in

next to establish its relationship with the HOMO−LUMO gap and thus with the Hammett substituent constants. We found that the ΔE0,0 values ranged between 2.25 and 2.53 eV for the subset of 20 dyes explored (due to their nonfluorescent nature, ΔE0,0 values could not be determined for the meso-formyl dyes). The highest ΔE0,0 values were measured with those compounds bearing electron releasing and electron withdrawing groups at positions C8 and C2/C6, respectively. The smallest ΔE0,0 values were recorded with BODIPY dyes bearing electron withdrawing and electron releasing groups, respectively, at the same positions. As reported by Arbeloa et al. the trend is due to the different electronic density in the HOMO versus the LUMO at both positions C8 versus C2 and C6 (see also Figure 1). Electron withdrawing groups at C8 stabilize the LUMO and destabilize the HOMO,43 an opposite effect is found at C2 and C6. The energy for the first excited singlet state corresponds to the electronic transition from the HOMO to the LUMO. As expected, we obtained a linear correlation between experimentally determined ΔE0,0 and calculated HOMO−LUMO energy gaps for the 20 dyes explored (eq 6). The value of the slope (m) was 0.819 ± 0.003 (r2 = 0.9998). ΔE0,0 = m(εLUMO − εHOMO)

(6)

Equations for the Prediction of BODIPY Properties from Hammett Substituent Constants. Finally, we combined the results of the electrochemical studies (which provide parameters for the correlation of BODIPY redox potentials as a function of DFT orbital energies (eq 5) with the linear correlation between DFT and Hammett substituent constants (see eqs 3 and 4). We obtained in this manner equations for the reduction potentials of the radical cation (eq 7) and neutral forms (eq 8) of the 1,3,5,7-tetramethyl BODIPY core as a function of Hammett substituent constants at the meso (RI) and equatorial positions (RII + RIII) in acetonitrile.

Figure 5. Top panel: Normalized absorption (solid) and emission (dashed) spectra in acetonitrile for the BODIPY dye featuring iodine at the RII and RIII positions. The green line corresponds to the ΔE0,0 value predicted according to eq 9. Bottom panel: Cyclic voltammogram measured in acetonitrile of a saturated (0.68 mM) solution of the above-mentioned BODIPY dye. The red and blue lines correspond to the predicted E°B/B−• and E°B+•/B, respectively.

the 6-31g(d) basis set, and thus this BODIPY cannot be simulated using the DFT methods we employed to derive eqs 3 and 4. Using our equations and the Hammett parameters for iodine, we successfully predicted the ΔE0,0 and electrochemical properties of the iodo-BODIPY within experimental error. In summary, the effect of substitution at the 2, 6, and 8 positions of the 1,3,5,7-tetramethyl BODIPY core was characterized by spectroscopic, electrochemical, and computational techniques. Experimental results on spectroscopic and electrochemical data were compiled for 26 different 1,3,5,7tetramethyl BODIPY dyes with varying substituents at the C2, C6, and C8 positions. We found that ΔE0,0s were not significantly affected by substitution patterns and ranged between 2.25 and 2.53 eV. This is in stark contrast with the large spread in redox potentials measured for the same 26 compounds. Values of E°B/B−• ranged from −1.84 to −0.52 V, a full 1.3 V. For E°B+•/B, values ranged from 0.60 to 1.42 V, a 0.8 V spread in values. Whereas the HOMO−LUMO energy gap is not significantly affected by the substitution pattern, the redox potentials can be readily tuned by appropriate choice of substituents. The HOMO and LUMO energy levels for 100 functionalized variations of the BODIPY scaffold were calculated at the B3LYP 6-31g(d) level, and correlated with the electron donating and withdrawing nature of the added substituents. The best fitting parameters for the HOMO were obtained when the Hammett σm value was used for substituents at position C8 and the sum of Hammett σp values was used for substituents at positions C2 and C6. Alternatively, the best fitting parameters for the LUMO were obtained when the Hammett σp value was used for substituents at position C8 and

E°B+• /B(V vs Fc+•/Fc) = (0.43 ± 0.02)(σmRI) + (0.53 ± 0.01)(σpRII + σpRIII) + (0.74 ± 0.7)

(7)

E°B/B−•(V vs Fc+•/Fc) = (0.74 ± 0.03)(σpRI) + (0.61 ± 0.03)(σmRII + σmRIII) + ( −1.58 ± 0.06)

(8)

We also combined eqs 3 and 4 with the results from the correlation between ΔE0,0 values and HOMO−LUMO gap energies (eq 6) to obtain an equation to predict the vibrational zero electronic energy of excited BODIPY dyes in acetonitrile as a function of the Hammett substituent constants (eq 9). ΔE0,0(eV) = ( −0.79 ± 0.03)(σpRI) + ( −0.65 ± 0.03) (σmRII + σmRIII) + (0.46 ± 0.02)(σmRI) + (0.557 ± 0.009)(σpRII + σpRIII) + (2.45 ± 0.04)

(9)

With these equations in hand, we are now equipped to rapidly estimate the redox potentials for any 1,3,5,7-tetramethyl BODIPY featuring functional groups with reported Hammett substituent constants. Drawing solely from the library of 530 Hammett substituent constants compiled by Hansch38 and computing all possible permutations of the three sites of functionalization, we can predict the redox and spectral behavior for over 75 million dyes and greatly enhance our ability to build BODIPY based PeT systems. G



Article

the sum of Hammett σm values was used for substituents at positions C2 and C6. The linear relationship between calculated orbital energies and redox potentials was quantified for a subset of the dyes, allowing for the preparation of equations directly relating the BODIPY reduction and oxidation potentials in acetonitrile with the Hammett substituent constants for functional groups at the equatorial and meso positions. A parallel set of equations was also derived to estimate the vibrational zero electronic energy of excited BODIPYs as a function of Hammett substituent constants. The equations reported here allow for the rapid, empirical prediction of BODIPY redox potentials, which when combined with ΔE0,0 values enable us to estimate the reducing and oxidizing power of photoexcited BODIPYs as a function of their substitution pattern. We anticipate that this work will greatly facilitate the selection of BODIPY chromophores for the preparation PeT based luminescent sensors and logic gates and to explore other systems where photoinduced electron transfer plays a critical role.

■

(6) Andreasson, J.; Pischel, U. Smart molecules at work: mimicking advanced logic operations. Chem. Soc. Rev. 2010, 39, 174−188. (7) Oleynik, P.; Ishihara, Y.; Cosa, G. Design and synthesis of a BODIPY-alpha-tocopherol adduct for use as an off/on fluorescent antioxidant indicator. J. Am. Chem. Soc. 2007, 129, 1842−1843. (8) Krumova, K.; Friedland, S.; Cosa, G. How lipid unsaturation, peroxyl radical partitioning, and chromanol lipophilic tail affect the antioxidant activity of alpha-tocopherol: direct visualization via highthroughput fluorescence studies conducted with fluorogenic alphatocopherol analogues. J. Am. Chem. Soc. 2012, 134, 10102−10113. (9) Krumova, K.; Greene, L. E.; Cosa, G. Fluorogenic alphatocopherol analogue for monitoring the antioxidant status within the inner mitochondrial membrane of live cells. J. Am. Chem. Soc. 2013, 135, 17135−17143. (10) Krumova, K.; Cosa, G. In Photochemistry; The Royal Society of Chemistry: London, U.K., 2013; Vol. 41, pp 279−301. (11) Loudet, A.; Burgess, K. BODIPY dyes and their derivatives: syntheses and spectroscopic properties. Chem. Rev. 2007, 107, 4891− 4932. (12) Yuan, M.; Li, Y.; Li, J.; Li, C.; Liu, X.; Lv, J.; Xu, J.; Liu, H.; Wang, S.; Zhu, D. A colorimetric and fluorometric dual-modal assay for mercury ion by a molecule. Org. Lett. 2007, 9, 2313−2316. (13) Sunahara, H.; Urano, Y.; Kojima, H.; Nagano, T. Design and synthesis of a library of BODIPY-based environmental polarity sensors utilizing photoinduced electron-transfer-controlled fluorescence ON/ OFF switching. J. Am. Chem. Soc. 2007, 129, 5597−5604. (14) Ulrich, G.; Ziessel, R.; Harriman, A. The chemistry of fluorescent bodipy dyes: versatility unsurpassed. Angew. Chem., Int. Ed. 2008, 47, 1184−1201. (15) Domaille, D. W.; Que, E. L.; Chang, C. J. Synthetic fluorescent sensors for studying the cell biology of metals. Nat. Chem. Biol. 2008, 4, 168−175. (16) Li, Y.; Liu, T.; Liu, H.; Tian, M.-Z.; Li, Y. Self-assembly of intramolecular charge-transfer compounds into functional molecular systems. Acc. Chem. Res. 2014, 47, 1186−1198. (17) Krumova, K.; Oleynik, P.; Karam, P.; Cosa, G. Phenol-based lipophilic fluorescent antioxidant indicators: a rational approach. J. Org. Chem. 2009, 74, 3641−3651. (18) Mendez-Hernandez, D. D.; Tarakeshwar, P.; Gust, D.; Moore, T. A.; Moore, A. L.; Mujica, V. Simple and accurate correlation of experimental redox potentials and DFT-calculated HOMO/LUMO energies of polycyclic aromatic hydrocarbons. J. Mol. Model. 2013, 19, 2845−2848. (19) Rehm, D.; Weller, A. Kinetics of fluorescence quenching by electron and H-atom transfer. Isr. J. Chem. 1970, 8, 259−&. (20) Komatsu, T.; Urano, Y.; Fujikawa, Y.; Kobayashi, T.; Kojima, H.; Terai, T.; Hanaoka, K.; Nagano, T. Development of 2,6-carboxysubstituted boron dipyrromethene (BODIPY) as a novel scaffold of ratiometric fluorescent probes for live cell imaging. Chem. Commun. 2009, 7015−7017. (21) Komatsu, T.; Oushiki, D.; Takeda, A.; Miyamura, M.; Ueno, T.; Terai, T.; Hanaoka, K.; Urano, Y.; Mineno, T.; Nagano, T. Rational design of boron dipyrromethene (BODIPY)-based photobleachingresistant fluorophores applicable to a protein dynamics study. Chem. Commun. 2011, 47, 10055−10057. (22) Banuelos, J.; Martin, V.; Gomez-Duran, C. F.; Arroyo Cordoba, I. J.; Pena-Cabrera, E.; Garcia-Moreno, I.; Costela, A.; Perez-Ojeda, M. E.; Arbeloa, T.; Lopez Arbeloa, I. New 8-amino-BODIPY derivatives: surpassing laser dyes at blue-edge wavelengths. Chemistry 2011, 17, 7261−7270. (23) Verma, M.; Chaudhry, A. F.; Fahrni, C. J. Predicting the photoinduced electron transfer thermodynamics in polyfluorinated 1,3,5-triarylpyrazolines based on multiple linear free energy relationships. Org. Biomol. Chem. 2009, 7, 1536−1546. (24) Shah, M.; Thangaraj, K.; Soong, M.-L.; Wolford, L. T.; Boyer, J. H.; Politzer, I. R.; Pavlopoulos, T. G. Pyrromethene-BF2 complexes as laser dyes. Heteroat. Chem. 1990, 1, 389−399.

ASSOCIATED CONTENT

S Supporting Information *

1

H NMR and 13C NMR spectra for compounds 4f and 4g. Complete ref 30. Gaussian 09 output data. Complete fitting parameters obtained for the correlation of Hammett substituent constants with HOMO and LUMO energy levels calculated via DFT. Fitting parameters obtained for the correlation of redox potentials determined by cyclic voltammetry with orbital energies calculated via DFT. Correlation of ΔE0,0 values with HOMO−LUMO gap energies calculated via DFT. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*(G.C.) E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS G.C. is grateful to the Natural Sciences and Engineering Research Council (NSERC) and Canadian Foundation for Innovation (CFI) for funding. R.L. is thankful to NSERC for a postgraduate scholarship; L.E.G. is thankful to Vanier Canada for a postgraduate scholarship; and K.K. is thankful to the Drug Discovery and Training Program (CIHR) for a postgraduate scholarship.

■

REFERENCES

(1) de Silva, A. P.; Gunaratne, H. Q.; Gunnlaugsson, T.; Huxley, A. J.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Signaling recognition events with fluorescent sensors and switches. Chem. Rev. 1997, 97, 1515−1566. (2) Beeson, T. D.; Mastracchio, A.; Hong, J. B.; Ashton, K.; Macmillan, D. W. Enantioselective organocatalysis using SOMO activation. Science 2007, 316, 582−585. (3) Schultz, D. M.; Yoon, T. P. Solar synthesis: prospects in visible light photocatalysis. Science 2014, 343, 1239176. (4) O’Regan, B.; Gratzel, M. A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO2 films. Nature 1991, 353, 737− 740. (5) de Silva, A. P.; Moody, T. S.; Wright, G. D. Fluorescent PET (photoinduced electron transfer) sensors as potent analytical tools. Analyst 2009, 134, 2385−2393. H



Article

(25) Nepomnyashchii, A. B.; Bard, A. J. Electrochemistry and electrogenerated chemiluminescence of BODIPY dyes. Acc. Chem. Res. 2012, 45, 1844−1853. (26) Nepomnyashchii, A. B.; Broring, M.; Ahrens, J.; Bard, A. J. Synthesis, photophysical, electrochemical, and electrogenerated chemiluminescence studies. Multiple sequential electron transfers in BODIPY monomers, dimers, trimers, and polymer. J. Am. Chem. Soc. 2011, 133, 8633−8645. (27) Duran-Sampedro, G.; Agarrabeitia, A. R.; Garcia-Moreno, I.; Costela, A.; Banuelos, J.; Arbeloa, T.; Arbeloa, I. L.; Chiara, J. L.; Ortiz, M. J. Chlorinated BODIPYs: surprisingly efficient and highly photostable laser dyes. Eur. J. Org. Chem. 2012, 2012, 6335−6350. (28) Wu, L.; Burgess, K. A new synthesis of symmetric boraindacene (BODIPY) dyes. Chem. Commun. 2008, 4933−4935. (29) Yogo, T.; Urano, Y.; Ishitsuka, Y.; Maniwa, F.; Nagano, T. Highly efficient and photostable photosensitizer based on BODIPY chromophore. J. Am. Chem. Soc. 2005, 127, 12162−12163. (30) Frisch, M. J.; et al. Gaussian 09, revision D.01; Gaussian, Inc.; Wallingford, CT, 2004. (31) Krumova, K.; Cosa, G. Bodipy dyes with tunable redox potentials and functional groups for further tethering: preparation, electrochemical, and spectroscopic characterization. J. Am. Chem. Soc. 2010, 132, 17560−17569. (32) Prieto, J. B.; Arbeloa, F. L.; Martinez, V. M.; Lopez, T. A.; AmatGuerri, F.; Liras, M.; Arbeloa, I. L. Photophysical properties of a new 8-phenyl analogue of the laser dye PM567 in different solvents: internal conversion mechanisms. Chem. Phys. Lett. 2004, 385, 29−35. (33) Kee, H. L.; Kirmaier, C.; Yu, L.; Thamyongkit, P.; Youngblood, W. J.; Calder, M. E.; Ramos, L.; Noll, B. C.; Bocian, D. F.; Scheidt, W. R.; et al. Structural control of the photodynamics of boron-dipyrrin complexes. J. Phys. Chem. B 2005, 109, 20433−20443. (34) Hedley, G. J.; Ruseckas, A.; Harriman, A.; Samuel, I. D. Conformational effects on the dynamics of internal conversion in boron dipyrromethene dyes in solution. Angew. Chem., Int. Ed. 2011, 50, 6634−6637. (35) Osorio-Martinez, C. A.; Urias-Benavides, A.; Gomez-Duran, C. F.; Banuelos, J.; Esnal, I.; Lopez Arbeloa, I.; Pena-Cabrera, E. 8AminoBODIPYs: cyanines or hemicyanines? The effect of the coplanarity of the amino group on their optical properties. J. Org. Chem. 2012, 77, 5434−5438. (36) Wood, T. E.; Thompson, A. Advances in the chemistry of dipyrrins and their complexes. Chem. Rev. 2007, 107, 1831−1861. (37) Zhu, S. L.; Zhang, J. T.; Vegesna, G.; Tiwari, A.; Luo, F. T.; Zeller, M.; Luck, R.; Li, H. H.; Green, S.; Liu, H. Y. Controlled Knoevenagel reactions of methyl groups of 1,3,5,7-tetramethyl BODIPY dyes for unique BODIPY dyes. RSC Adv. 2012, 2, 404−407. (38) Hansch, C.; Leo, A.; Taft, R. W. A survey of Hammett substituent constants and resonance and field parameters. Chem. Rev. 1991, 91, 165−195. (39) Allik, T. H.; Hermes, R. E.; Sathyamoorthi, G.; Boyer, J. H. Proc. SPIE-Int. Soc. Opt. Eng. 1994, 2115, 240−248. (40) Alamiry, M. A. H.; Benniston, A. C.; Hagon, J.; Winstanley, T. P. L.; Lemmetyinen, H.; Tkachenko, N. V. The fluorine effect: photophysical properties of borondipyrromethene (bodipy) dyes appended at the meso position with fluorinated aryl groups. RSC Adv. 2012, 2, 4944−4950. (41) Leen, V.; Yuan, P.; Wang, L.; Boens, N.; Dehaen, W. Synthesis of meso-halogenated BODIPYs and access to meso-substituted analogues. Org. Lett. 2012, 14, 6150−6153. (42) Esnal, I.; Banuelos, J.; Arbeloa, I. L.; Costela, A.; Garcia-Moreno, I.; Garzon, M.; Agarrabeitia, A. R.; Ortiz, M. J. Nitro and amino BODIPYS: crucial substituents to modulate their photonic behavior. RSC Adv. 2013, 3, 1547−1556. (43) López Arbeloa, F.; Bañuelos, J.; Martínez, V.; Arbeloa, T.; López Arbeloa, I. Structural, photophysical and lasing properties of pyrromethene dyes. Int. Rev. Phys. Chem. 2005, 24, 339−374.

I


Electronic Excited State Redox Properties for BODIPY Dyes Predicted

Recommend Documents