an asian journal

1 downloads 0 Views 3MB Size Report
Formation of the Excited Triplet State and Singlet Oxygen for. BODIPY ..... trapping of singlet oxygen by DPBF, which causes decomposition of DPBF.
CHEMISTRY AN ASIAN JOURNAL www.chemasianj.org

A Journal of

REPRINT

DOI: 10.1002/asia.201700794

Full Paper

BODIPY Photosensitizers

Photoinduced Electron Transfer-based Halogen-free Photosensitizers: Covalent meso-Aryl (Phenyl, Naphthyl, Anthryl, and Pyrenyl) as Electron Donors to Effectively Induce the Formation of the Excited Triplet State and Singlet Oxygen for BODIPY Compounds Xian-Fu Zhang*[a, b] and Nan Feng[a] Abstract: Pristine BODIPY compounds have negligible efficiency to generate the excited triplet state and singlet oxygen. In this report, we show that attaching a good electron donor to the BODIPY core can lead to singlet oxygen formation with up to 58 % quantum efficiency. For this purpose, BODIPYs with meso-aryl groups (phenyl, naphthyl, anthryl, and pyrenyl) were synthesized and characterized. The fluorescence, excited triplet state, and singlet oxygen formation properties for these compounds were measured in various solvents by UV/Vis absorption, steady-state and time-resolved fluorescence methods, as well as laser flash photolysis technique. In particular, the presence of anthryl and pyrenyl showed substantial enhancement on the singlet oxygen formation ability of BODIPY with up to 58 % and 34 % quantum

Introduction Photosensitizers (PS) that efficiently produce singlet oxygen (1Dg) play the key role in photodynamic therapy (PDT) of cancers.[1–3] Porphyrins, phthalocyanines, chlorophylls, and halogenated xanthene dyes are typical well-known photosensitizers.[4–6] Activatible PSs that produce singlet oxygen under specific conditions are highly desirable as additional selectivity can be obtained.[1] Such PSs, however, are still rarely available. The triplet excited state T1 of a PS is the key intermediate for generating singlet oxygen through an energy transfer process: T1 + O2 !S0 + 1O2.[7–9] T1 of a PS is generally produced through intersystem crossing (ISC) in which one unpaired electron of the S1 state is spin converted: S1!T1.[7, 8] For molecules

efficiency, respectively, owing to their stronger electron-donating ability. Upon the increase in singlet oxygen formation, the fluorescence quantum yield and lifetime values of the aryl-BODIPY showed a concomitant decrease. The increase in solvent polarity enhances the singlet oxygen generation but decreases the fluorescence quantum yield. The results are explained by the presence of intramolecular photoinduced electron transfer from the aryl moiety to BODIPY core. This method of promoting T1 formation is very different from the traditional heavy atom effect by I, Br, or transition metal atoms. This type of novel photosensitizers may find important applications in organic oxygenation reactions and photodynamic therapy of tumors.

that are not efficient in forming the T1 state, the external or internal heavy atom effect by I, Br, or transition metal atoms are often used to efficiently enhance ISC.[10, 11] This method has been used to make BODIPY (boron-dipyrromethene) compounds become singlet oxygen photosensitizers.[12] BODIPYs are environment insensitive, have high molar absorption coefficients, good resistance to photobleaching, and higher light– dark toxicity ratios than other PDT agents. These features make them ideal to act as good photosensitizers for PDT. We show now that BODIPYs (Scheme 1) are able to generate the T1 state based on a different mechanism. BODIPY and its

[a] Prof. Dr. X.-F. Zhang, N. Feng Institute of Applied Photochemistry & Center of Instrumental Analysis Hebei Normal University of Science and Technology 360 Hebeidajiexiduan, Qinhuangdao, Hebei Province, 066004 (China) E-mail: [email protected] [b] Prof. Dr. X.-F. Zhang MPC Technologies 124 Royal Ave., Hamilton, Ontario, L8S 3H4 (Canada) Supporting information for this article can be found under: https://doi.org/10.1002/asia.201700794. Chem. Asian J. 2017, 12, 2447 – 2456

Scheme 1. Molecular structures and abbreviations of the aryl-BODIPYs. Red dots show the linking positions.

2447

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper derivatives are well-known highly fluorescent materials for various applications,[13–17] but these dyes lack the ability to produce the T1 state upon photo excitation.[12, 18, 19] Unsubstituted BODIPY, for example, shows fluorescence efficiency near unity (0.97 : 0.03) and a long fluorescence lifetime of 6.89 ns in methanol,[20] which leaves it little space to form the T1 state. The heavy atom effect with iodine or bromine substitution and dimerization are often used to make BODIPY efficient in T1 formation.[21, 22] By covalently linking an aryl as an electron-donor unit (benzene, naphthalene, anthracene, or pyrene) onto the meso position of the BDP (Scheme 1), we could make BODIPY compounds generate the T1 state and hence singlet oxygen efficiently. This aryl substitution method extends our previous dimerization and bromination techniques to make BODIPY generate T1 and singlet oxygen effectively. Moreover, the capability is sensitive to external conditions that make them good activatible PSs.

Results and Discussion The synthesized compounds were characterized by NMR, HRMS, and UV/Vis spectra, which are all consistent with their chemical structure. The near-IR photoluminescence spectra were used to identify the formation of singlet oxygen with excitation at 505 nm (near the absorption maximum of ArTMBDPs) in air-saturated solutions, as shown in Figure 1. The 1270 nm band is the characteristic emission of singlet oxygen, which was observed for all the aryl-BODIPYs, but the band intensity of AN-TMBDP and PY-TMBDP are much higher. At 505 nm, only the BODIPY unit absorbs the light and is photoexcited, whereas the pyrene, anthracene, naphthalene, or benzene moiety absorbs only at wavelengths of less than 400 nm (Figure 1, right), therefore it is not due to the energy transfer from the aryl to the BODIPY core. In nitrogen-saturated solutions, however, the 1270 nm band for the aryl-BODIPYs was absent. These results indicate that the 1270 nm band is due to singlet oxygen (1Dg) generated by BODIPY’s light absorption. The variation of emission intensity in Figure 1 suggests that these aryl-substituted BODIPYs have different quantum yields for 1O2 formation (FD). Table 1 lists the FD values in different

Table 1. Quantum yield of singlet oxygen formation (FD) in five solvents

FD

Hexane

Toluene

THF

Ethanol

Acetonitrile

AN-TMBDP PY-TMBDP NP-TMBDP Ph-TMBDP TMBDP

0.024 0.13 0.050 0.050 0.034

0.045 0.086 0.043 0.023 0.061

0.21 0.20 0.13 0.13 0.091

0.58 0.34 0.041 0.030 0.058

0.22 0.34 0.057 0.017 0.069

solvents for these compounds. AN-TMBDP and PY-TMBDP show good FD values of 0.58 and 0.34 in ethanol, respectively, whereas the pristine TMBDP has a low FD value 0.063(: 0.029) in all solvents. Both compounds work well in polar solvents. Apparently, attaching an anthracene or a pyrene unit exhibits a tremendous effect on the photosensitizing capability of BODIPY compounds. Ph-TMBDP and NP-TMBDP, on the other hand, do not show significant promotional effects in the solvents studied. AN-TMBDP is remarkably more efficient than PYTMBDP in ethanol. In non-polar solvents, such as toluene and n-hexane, AN-TMBDP loses the enhanced capability toward photosensitization. This means that AN-TMBDP is a polarity activatible photosensitizer. It is complementary to our previously reported BODIPY dimer, which can generate singlet oxygen only in non-polar solvents.[12] To understand the phenomena, the detailed photophysical and spectral properties are measured to elucidate the mechanism. To confirm that the aryl-BODIPYs are indeed singlet oxygen photosensitizers, their excited triplet state T1 was then identified, as T1 is the precursor of singlet oxygen 1O2(1Dg), which generates 1O2(1Dg) through the energy transfer process: T1 + O2 !S0 + 1O2(1Dg). Laser flash photolysis was used to detect the T1 state of the aryl-BODIPYs, the case for AN-TMBDP is shown in Figure 2. No significant signal was detected for TMBDP, Ph-TMBDP, and NP-TMBDP in these solvents within the detection limit, indicating that the efficiency of direct intersystem crossing (ISC) for TMBDP is negligible: R-TMBDP (S1)!R-TMBDP (T1), for R = H, Ph, NP.

Figure 1. Left: NIR luminescence band of singlet oxygen (1Dg) by using the Ar-BODIPYs as photosensitizers in air-saturated CCl4 solutions with excitation at 505 nm (absorbance is 0.50 at 505 nm). Right: UV/Vis absorption spectra of the Ar-BODIPYs in air-saturated CCl4. Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

2448

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

Figure 2. Transient absorption spectra (left) and the decay of T1 absorption (right) of AN-TMBDP in argon purged ethanol solution (20 mm photosensitizer), excitation wavelength is 355 nm (4 ns Nd:YAG pulsed laser).

For AN-TMBDP, a positive absorption band was found at 420 nm in polar solvents ethanol or acetonitrile (Figure 2), but the band could not be detected in non-polar hexane and low polarity toluene, indicating that the observed transient absorption spectrum (TAS) is due to the presence of both anthracene and polar media. The positive absorption of Figure 2 is assigned to triplet– triplet (T1!Tn) TAS, because it exhibits the typical features of T1–Tn absorptions for BODIPYs and agreed with that reported previously.[19] For example, well-defined isosbestic points exist between the positive absorption and the well-separated ground state bleaching, whereas the minimum of the negative bands at 505 nm matches the maximum of the ground state absorption. Furthermore, the rise of the negative absorption and the decay of positive absorption are synchronous, indicating that the decay of the T1 state repopulated the ground state (T1!S0). The decay of T1!Tn absorption at 420 nm is also given in Figure 2, together with the rise of the absorption at 525 nm (ground state absorption minimum) for AN-TMBDP. Both curves are mono-exponential, suggesting that there was only one transient species and it was transformed to the S0 state within a ms timescale. The triplet lifetime (tT) thus obtained is 11 ms for AN-TMBDP in argon-saturated ethanol, which is sufficiently long for photosensitizing the production of singlet oxygen by an energy transfer process (T1 + O2 !S0 + 1O2). In airsaturated solution, the tT value is shortened dramatically. For example, tT of AN-TMBDP is decreased from 11 to 0.21 ms, the rate constant of oxygen quenching (kq) is calculated to be 2.3 V 109 m@1 s@1. This highly effective oxygen quenching also indicates that the positive bands are indeed due to T1 absorption. In summary, T1 was identified by laser flash pyrolysis (LFP), and it can effectively transfer energy to molecular oxygen. To understand the aryl substitution effect on the photosensitizing ability of BODIPY, we also measured the absorption and fluorescence properties of these compounds in different solvents. The absorption maximum of an aryl-BODIPY is slightly shifted depending on the aryl group owing to the weak electronic interaction between the aryl and the attached BODIPY (Figure 1 A). Figure 3 displays the normalized absorption and fluorescence emission spectra. Upon changing solvent polarity Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

(from non-polar to highly polar), the band shape and the peak position of the absorption spectra show little variation for all compounds, indicating that the FD change upon changing solvent for a compound (Table 1) is not caused by its molecular ground state. The emission band at 510 nm exhibits the typical features of BODIPY fluorescence. For each aryl-substituted compound, the shape of the fluorescence spectra is not changed by the excitation wavelength tested within 300–500 nm, even though the AN, PY, and NP units each absorb in the 300–400 nm region. The results mean that each AN (S1), NP (S1), PY (S1) moiety in a conjugate R-TMBDP (R = AN, NP, PY) is non emissive, but only the attached TMBDP is fluorescent, owing to the energy transfer R(S1)-TMBDP!R-TMBDP(S1) and/or electron transfer R(S1)TMBDP!R + -TMBDP@ (R = AN, NP, PY). For clarity, we only used the wavelength > 480 nm to selectively photoexcite the TMBDP unit for the following study. For AN-TMBDP, however, the fluorescence spectral shape exhibits a dramatic change upon solvent polarity variation. In non-polar solvent, only a normal emission band (short wavelength emission owing to LE band) at 517(: 3) nm can be observed, the LE band is well known to originate from the lowest local excited state S1 on the TMBDP moiety, that is, process (2) after light absorption process (1). (1) AN-TMBDP(S0) + hn!AN-TMBDP(S1) (2) AN-TMBDP(S1)!AN-TMBDP(S0) + hn’ (LE emission: & 517 nm) In polar solvent, however, a new band appears at 650 nm; this band is broad and structureless, which has the typical features of charge transfer emission, owing to the charge separated state AN + –TMBDP@ produced by photoinduced electron transfer (PET) process (3) as shown in Scheme 2, PET can be slowed down by the competing heat releasing process (4). PYTMBDP also shows a weak emission band in the red region, but NP-TMBDP and Ph-TMBDP have only the LE emission band. The new emission (charge transfer (CT) band) is due to the charge transfer emission process (5), which competes with inverse electron transfer process (6), in which D stands for AN or another electron donor. (3) D-TMBDP(S1)!D + –TMBDP@ , PET (4) D-TMBDP(S1)!D-TMBDP(S0) + heat, internal conversion (5) D + –TMBDP@ !D-TMBDP + hnCT, CT emission: & 650 nm

2449

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

Figure 3. Normalized absorption (left) and fluorescence spectra (right) in different solvents. Emission spectra were measured with excitation at 485 nm. Substance concentration is 20 mm.

Scheme 2. Photophysical mechanism for the triplet state and singlet oxygen formation. (1) Light absorption of TMBDP unit. (2) Local excited state emission from the TMBDP unit with rate constant kf. (3) Photoinduced electron transfer (PET) that generates charge separated states (CSS) with rate constant kPET. (4) Internal conversion (IC) with rate constant kIC. (5) Charge recombination (CR) that yields triplet state (T1) with rate constant kCR,ISC. (6) Charge transfer emission from CSS with rate constant kf,CT. (7) Charge recombination with rate constant kCR. (8) Singlet oxygen formation through energy transfer from T1 to molecular oxygen with rate constant kET. (9) Decay of T1 state to ground state S0 with rate constant kD. (10) Emission of singlet oxygen with 1270 nm band. (11) Chemical trapping of singlet oxygen by DPBF, which causes decomposition of DPBF. Techniques that demonstrate the occurrence of the process: (1) UV/Vis absorption; (2), (3), and (5): fluorescence spectra and lifetimes; (6), (8), and (9): laser flash photolysis; (10) NIR fluorescence; (11) UV/Vis monitoring of DPBF decomposition.

(6) D + –TMBDP@ !D-TMBDP(T1), T1 formation: 420 nm (7) D + –TMBDP@ !D-TMBDP + heat, charge recombination (8) D-TMBDP(T1) + O2 !D-TMBDP + 1O2, energy transfer Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

The relative emission intensity of the new band (intramolecular charge transfer, ICT) for AN-TMBDP is higher in polar solvent, so is the FD value of AN-TMBDP. This consistency be2450

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper tween ICT and FD values (Table 1) hints that the charge separated state is related to the singlet oxygen production, because the charge recombination (inverse electron transfer of PET) process (6) leads to T1 formation on the TMBDP unit, and then the energy transfer from T1 to oxygen yields singlet oxygen (process 8). This mechanism is also supported by the laser flash photolysis study, because T1 absorption (420 nm band, Figure 2) could only be observed in polar solvents ethanol and acetonitrile, but not in non-polar hexane, this solvent dependency of T1 absorption is the same as that of CT emission (650 nm, Figure 3), confirming that D + –TMBDP@ is the common precursor for both T1 and CT emission, processes (5) and (6). The fluorescence quantum yields and lifetime values in different solvents for the compounds were also measured and the results are listed in Table 2, which also provides evidence for PET. The fluorescence decay curves are shown in Figure 4,

which shows that the fluorescence lifetime is affected by both the aryl structure and solvent polarity. In polar solvent acetonitrile, the rate of emission decay (Figure 4, left) is ranked as: AN-TMBDP > PY-TMBDP > TMBDP > NP-TMBDP. In non-polar solvent hexane, however, the decay rate (Figure 4, right) is ranked as: TMBDP (5.85 ns) > PY-TMBDP (6.00 ns) > NP-TMBDP (6.41 ns) > AN-TMBDP (6.83 ns). AN-TMBDP showed the slowest mono-exponential decay in hexane, but became the fastest and bi-exponential in acetonitrile (because intramolecular PET occurs in polar solvents and shortens the lifetime, PET forms a charge separated state (CSS), which has an ICT emission band, the dual emission causes the bi-exponential decay). For AN-TMBDP, the rate of fluorescence decay is increased by the increase in solvent polarity and ranked as: ethanol > acetonitrile > acetone > CH2Cl2 > hexane (Figure 4, right). For TMBDP, the Ff and tf values were measured in eight solvents with different polarities, it is clear that the Ff value (0.95 :

Table 2. Photophysical properties of R-TMBDP in different solvents.[a]

R

Solvent

labs [nm]

lem [nm]

Ff

tf [ns]

kf [109 s@1]

kPET [109 s@1]

PY

n-hexane toluene CCl4 CH2Cl2 THF ethanol CH3CN DMF c-hexane benzene CCl4 THF CH2Cl2 acetone ethanol CH3CN n-hexane toluene CCl4 CH2Cl2 THF ethanol CH3CN n-hexane c-hexane toluene CCl4 CH2Cl2 EtOAc THF acetone MeOH CH3CN c-hexane benzene toluene CCl4 CH2Cl2 acetone CH3CN ethanol

503 506 507 504 503 502 500 503 504 508 507 502 506 503 503 502 503 506 507 504 505 502 500 501 503 503 504 501 498 500 498 498 497 508 510 509 510 506 502 501 503

513 518 517 515 514 512 509 513 516 520 518 515 518 514, 625 514, 640 517, 648 513 518 517 514 514 511 510 511 513 516 515 513 510 512 509 510 508 514 517 518 516 513 509 508 510

0.83 0.83 0.79 073 0.94 0.70 0.59 0.31 0.87 0.64 0.84 0.36 0.10 0.026 0.036 0.009 0.87 0.77 0.81 0.91 0.85 0.90 0.83 0.56 0.60 0.62 0.54 0.63 0.58 0.56 0.54 0.58 0.52 0.96 0.90 0.92 1.00 0.92 1.00 1.00 0.99

6.00 5.24 5.60 5.72 5.80 5.58 3.58, 5.34 4.94 6.83 5.47 5.77 5.83 5.33 1.01, 3.54 0.32, 3.53 0.26, 3.00 6.41 5.42 5.76 6.06 6.01 6.35 6.35 3.37 4.04 4.28 3.98 3.97 3.98 3.73 3.65 3.90 3.81 5.85 5.17 5.46 5.66 5.73 5.77 5.70 5.95

0.14 0.16 0.14 0.16 0.16 0.12 0.16 n.c.[a] 0.13 0.12 0.15 n.c. n.c. n.c. n.c. n.c. 0.14 0.14 0.14 0.14 0.14 0.14 0.13 0.17 0.15 0.14 0.14 0.16 0.15 0.15 0.15 0.15 0.14 0.16 0.17 0.17 0.18 0.16 0.17 0.18 0.17

– – – – – – 0.10 – – – – – – 0.82 2.96 3.67 – – – – – – – – – – – – – – – – – – – – – – – – –

AN

NP

Ph

H

(71 %)

(97 %) (81 %) (70 %), 6.70 (17 %)

[a] n.c.: not calculated because of PET. Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

2451

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper

Figure 4. Structure and solvent effect on the fluorescence lifetime. Left: the fluorescence decay of AN-TMBDP in different solvents (n-hexane, CH2Cl2, acetone, acetonitrile, ethanol, from top to bottom) with excitation at 509 nm (70 ps diode laser) and emission at 520 nm, [AN-TMBDP] = 20 mm. Right: the fluorescence decay of different aryl-substituted R-TMBDP (R = naphthyl, H, pyrenyl, phenyl, anthryl, from top to bottom) in CH3CN with excitation at 509 nm (70 ps diode laser) and emission at 520 nm, [Ar-TMBDP] = 20 mm.

0.05) is close to 1.0 and only slightly affected by solvent polarity (Table 2), its tf is 5.66(: 0.30) ns, which is also not determined by solvent polarity (Table 2). As FD & Fisc & 1@Ff, the very high fluorescence efficiency of TMBDP explains why its FD is low (Table 1) and no triplet excited state signal was detected by LFP. The emission rate constant of TMBDP (kf = Ff/tf) is 0.17 V 109 s@1, which is independent of solvent (Table 2). Upon the aryl substitution on the meso position of TMBDP, the fluorescence properties became related to not only the aryl but also solvent polarity. For AN-TMBDP, both the Ff and tf values are smaller than that of TMBDP in each solvent (Table 2), indicating that the attached anthracene unit quenches the fluorescence of the TMBDP moiety. In particular, the increase in solvent polarity leads to the sharp decrease in both Ff and tf values of AN-TMBDP, which suggests that the intramolecular PET from the AN moiety to the BODIPY unit causes the quenching of fluorescence. Upon solvent variation, the decrease in Ff and tf values coincides with the increase in FD values for AN-TMBDP and PY-TMBDP, which tells that the FD value changes with solvent polarity and is also closely related to intramolecular PET, that is, the product of PET is the precursor of singlet oxygen: (9) AN-TMBDP(S1)!AN + –TMBDP@ !AN-TMBDP(T1), (10) AN-TMBDP(T1) + O2 !D-TMBDP + 1O2 The solvent effect on the fluorescence properties of PYTMBDP is similar but not so strong as that of AN-TMBDP, meaning that PY also quenches the fluorescence of TMBDP through intermolecular PET or ICT, therefore PY-TMBDP generates singlet oxygen with the same mechanism as that of ANTMBDP. However, Table 2 shows that PY is less efficient than AN in quenching TMBDP fluorescence, owing to the fact that AN is a better electron donor (AN has a lower oxidation potential) than PY. As shown in Table 2, the Ff and tf values of Ph-TMBDP in ten solvents are 0.57 : 0.05 and 3.87 : 0.50 ns, respectively, smaller than those of TMBDP. The phenyl does quench the fluorescence of TMBDP but the slight fluctuation of Ff and tf values show no clear trend with solvent polarity, indicating that no intermolecular PET occurs in Ph-TMBDP. The mesophenyl quenching is due to the phenyl rotation along the C@C Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

bond that links the meso-phenyl to TMBDP. The rotation is evident from the viscosity effect on fluorescence. In viscous ethylene glycol, the Ff and tf values of Ph-TMBDP were measured as 0.96 and 5.02 ns, remarkably larger than 0.64 and 3.79 ns in ethanol. For NP-TMBDP, the Ff and tf values were measured in eight solvents, Ff is 0.85 : 0.08 and tf is 6.05 : 0.60 ns. Both Ff and tf values are not dependent on solvent polarity, also suggesting that no intermolecular PET occurs within NP-TMBDP. The larger size of naphthyl than phenyl reduces its rotation rate, therefore the Ff and tf values of NP-TMBDP are larger than that of Ph-TMBDP. The absence of PET within Ph-TMBDP and NP-TMBDP explains why the phenyl and naphthyl substitution on TMBDP does not lead to the promotion of singlet oxygen formation. Oxidation potential (Eox) is a measurement of the electrondonating ability of a compound. The Eox of benzene, naphthalene, anthracene, and pyrene in acetonitrile is 2.30, 1.54, 1.09, and 1.16 V respectively;[23, 24] the larger the Eox is, the weaker the electron-donating ability is. The presence or absence of PET is determined by the associated free energy change (DGPET). For PET to occur, its DGPET must be negative. DGPET is calculated by Equation (1): DGPET ¼ E ox @E Red @C@E S ¼ E ox þ 1:15@0:06@2:43 ¼ E ox @1:34 ð1Þ in which ERed is the reduction potential of TMBDP, and the value is @1.15 V,[25] ES is the excitation energy of S1 state (510 nm, or 2.43 eV), C is a constant (0.06 eV). Thus, we have DGPET = 0.96, 0.20, @0.25, and @0.18 eV for Ph-TMBDP, NPTMBDP, AN-TMBDP, and PY-TMBDP, respectively. As DGPET is positive for Ph-TMBDP and NP-TMBDP, no PET can occur. This explains why phenyl and naphthyl substitution shows no effect to enhance singlet oxygen formation. DGPET of ANTMBDP is more negative than that of PY-TMBDP, which explains why PY is less efficient in quenching the fluorescence of TMBDP. To further support the above explanation, we also made the compound Et2N-Ph-TMBDP (in which a para-(C2H5)2N- was at-

2452

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper tached to the benzene unit, which lowered the oxidation potential from 2.30 V of Ph-TMBDP to 0.76 V), so we have a negative DGPET = @0.58 eV in this case, instead of + 0.96 eV for PhTMBDP. Correspondingly, the FD value of Et2N-Ph-TMBDP is increased to 0.75, much larger than 0.091 of Ph-TMBDP. We can conclude that any electron donor attached to TMBDP can make BODIPY a good photosensitizer as long as DGPET is negative. The presence of PET within AN-TMBDP is also supported by the HOMO/LUMO distribution of the compounds in acetonitrile (Figure 5). The HOMO of AN-TMBDP is located on AN, whereas

Figure 5. Top: HOMO@1, HOMO, and LUMO of AN-TMBDP in acetonitrile. Middle: HOMO@1, HOMO, and LUMO of PY-TMBDP in acetonitrile. Bottom: HOMO@1, HOMO, and LUMO of NP-TMBDP in acetonitrile.

the LUMO is located on the TMBDP moiety, which means that upon de-excitation, one electron will be relocated from AN to TMBDP [AN-TMBDP(S1)!AN + –TMBDP@], that is, PET occurs. This is also the case for PY-TMBDP. For NP-TMBDP and PhTMBDP, however, both HOMO and LUMO are located on the BODIPY moiety; this means that electron transfer from naphthyl or phenyl to the BODIPY unit is not likely to occur. To further confirm the photosensitizing capability of the aryl-substituted TMBDPs, we examined the photosensitized oxidation of DPBF (1,3-diphenylisobenzofuran) in the presence of the compounds in different solvents. As DPBF can be specifically oxidized by 1O2, DPBF + 1O2 !DPBF-O2 decomposed product. The rate constant of this reaction (kr & 108–109 m@1 s@1) is known to be much faster than the natural decay rate of singlet oxygen (kD & 103–105 s@1), and it is generally considered that 1 O2 is essentially 100 % trapped by DPBF. DPBF shows an absorption maximum at 410 nm, but its oxidized product shows no absorption in the visible region. By monitoring the absorbance decrease of DPBF at 410 nm, we can calculate FD and evaluate the photosensitizing ability of R-TMBDP in different solvents. Figure 6 shows three examples, that is, the change of UV/Vis absorption spectra of DPBF in the presence of R-BODIPY (R = AN, PY, and NP) in air-saturated solutions. The 410 nm band of DPBF decreases with photo-irradiation of an R-BODIPY at 505 nm in the presence of oxygen, whereas the absorption peak of R-TMBDP had no change. However, the absence of any one of R-TMBDP, light irradiation, and oxygen led to the absence of DPBF decomposition. These results show that R-BODIPYs do indeed act as the photosensitizer of singlet oxygen.

Figure 6. Top and bottom left: the change of absorption spectrum upon irradiation time in air-saturated CH2Cl2 solution containing 35 mm DPBF and 6 mm photosensitizer with irradiation at 505 nm. Bottom right: the plot of ln([DPBF]0/[DPBF]t) at 412 nm against irradiation time. Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

2453

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper From the time evolution of DPBF absorption spectra, it is clear that the photosensitizing ability is ordered AN-TMBDP > PYTMBDP @ NP-TMBDP. Quantitatively, the DPBF concentration change follows the first-order kinetics very well with rate constant k (Figure 6, bottom right), that is, ln([DPBF]0/[DPBF]t) = kt, in which k = IaFD, and Ia = I0(1@10@A), I0 is the intensity of incident light at 505 nm, and Ia is the light intensity absorbed by a photosensitizer. The rate constant k values are obtained as the slope of the linear plot of ln([DPBF]0/[DPBF]t) against time (Figure 6, bottom right). From the slope of the linear plot, it is easy to see that the photosensitizing ability is: Ph-TMBDP < NP-TMBDP ! PY-TMBDP < AN-TMBDP. By using a reference compound in the same solvent, FD values were calculated and listed in Table 1.

Conclusions We synthesized and characterized meso-aryl-substituted BODIPYs (aryl = one of phenyl, naphthyl, anthryl, and pyrenyl). We showed that anthryl- and pyrenyl-substituted BODIPYs exhibited remarkable enhancement of the singlet oxygen generation for BODIPY with FD up to 0.58 and 0.34, respectively, in contrast to the FD of 0.063 for TMBDP. This is due to the electrondonating ability of the aryl, which makes the difference, as the triplet state T1 is formed by the charge recombination after PET or ICT within donor-TMBDP (donor = AN, PY, etc.). Any attached electron donor can make BODIPY a good photosensitizer as long as DGPET is negative. This method of promoting T1 formation opens a new avenue for designing new singlet oxygen photosensitizers. It can be polarity selective and different from the traditional heavy atom effect method using I, Br, or transition metal atoms. This type of PET-based or ICT-assisted halogen-free photosensitizers may find important applications in organic oxygenation reactions and photodynamic therapy of tumors.

Experimental Section General methods All reagents for synthesis were of analytical grade and used as received. All solvents for spectroscopic studies were dried and redistilled before use. 1H and 13C NMR spectra were recorded at room temperature with a Bruker AVANCE III HD 600 MHz NMR spectrometer. MS spectra were recorded with a Thermal Fisher LCQ FleetQ mass spectrometer. IR spectra were recorded at room temperature with a Shimadzu FTIR-8900 spectrometer. UV/Visible spectra were recorded with an Agilent 8454 spectrophotometer using 1 cm matched quartz cuvettes.

General procedure for the synthesis of 8-aryl-substituted BODIPY Anhydrous dichloromethane (CH2Cl2, 50 mL) containing a selected aldehyde (2.0 mmol) and 2,4-dimethyl-pyrrole (4.0 mmol) was stirred under argon protection. After 15 min, one drop of trifluoroacetic acid (CF3COOH) was added, and the solution was stirred for 12 h at room temperature. Then, DDQ (2,3-dichloro-5,6-dicyano1,4-benzoquinone, 2.0 mmol) was added and the solution was Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

stirred for 30 min. Excess boron trifluoride diethyl etherate (BF3OEt2, 4 mL) and triethylamine (NEt3, 4 mL) were added and stirring was continued for 30 min. The intense fluorescence of the reaction mixture was observed at that stage. The formation of intermediates and BODIPY products at every stage was monitored by UV/Vis absorption spectroscopy. After that, the reaction mixture was washed with water (100 mL V 3), and the organic layer was dried over anhydrous MgSO4, filtered, and evaporated. The crude product was purified by silica gel chromatography (eluent CH2Cl2/ n-hexane = 2:1 v/v) to afford pure samples. Ph-TMBDP (4,4-difluoro-8-phenyl-1,3,5,7-tetramethyl-4-boron3 a,4 a–diaza-s-indacene): Ph-TMBDP was prepared according to the general procedure by using benzaldehyde as a precursor. The product is dark-red crystals. Yield: 154 mg, 20 %. M.p.: 127–130 8C; IR (KBr): n˜ = 723, 978, 1072, 1155, 1196, 1471, 1508, 1543 (nBODIPY ring); 1508, 1543 (nB-F); 1308, 2854, 2924 cm@1 (nCH3); UV/Vis (CH2Cl2) max: 501 nm; 1H NMR (600 MHz, CDCl3): d = 7.54–7.47 (m, 3 H), 7.31–7.28 (m, 2 H), 6.00 (s, 2 H), 2.58 (s, 6 H), 1.39 ppm (s, 6 H); HRMS (APCI): m/z calcd for C19H20BF2N2 : 325.1682 [M+ +H] + ; found: 325.1682; HRMS (APCI): m/z calcd for C19H19BFN2 : 305.1625 [M@F] + ; found: 305.1623. NP-TMBDP (4,4-difluoro-8-naphthyl-1,3,5,7-tetramethyl-4-boron3 a,4 a–diaza-s-indacene): NP-TMBDP was synthesized according to the general synthesis procedure by using 1-naphthaldehyde as a precursor. The product is dark-red crystals (yield = 16 %). 1H NMR (600 MHz, CDCl3): d = 7.99 (d, J = 8.3 Hz, 1 H), 7.92 (d, J = 8.2 Hz, 1 H), 7.83 (d, J = 8.4 Hz, 1 H), 7.61–7.58 (m, 1 H), 7.54 (t, J = 7.2 Hz, 1 H), 7.46 (t, J = 7.6 Hz, 1 H), 7.42 (d, J = 6.6 Hz, 1 H), 5.97 (s, 2 H), 2.62 (s, 6 H), 1.08 ppm (s, 6 H); 13C NMR (151 MHz, CDCl3): d = 155.60, 143.03, 140.15, 133.53, 132.43, 131.95, 131.67, 129.27, 128.20, 127.27, 126.64, 125.96, 125.81, 124.92, 121.19, 14.70, 13.91 ppm; HRMS (APCI): m/z calcd for C23H21BF2N2 : 355.1782 [M@F] + , 375.1844 [M+ +H] + ; found: 355.1771, 375.18429. AN-TMBDP (4,4-difluoro-8-antryl-1,3,5,7-tetramethyl-4-boron3 a,4 a–diaza-s-indacene): Anthracene-9-carbaldehyde (0.21 g, 1.0 mmol) and 2,4-dimethylpyrrole (0.20 g, 2.1 mmol) were dissolved in 10 mL of anhydrous dichloromethane. One drop of trifluoroacetic acid was added, and the solution was stirred at room temperature for 10 h. When TLC monitoring showed complete consumption of starting materials, a solution of tetrachlorobenzoquinone (0.25 g, 1.0 mmol) in dichloromethane was added, and stirring was continued for 15 min. The reaction mixture was washed with saturated NaCl aqueous solution, dried with MgSO4, filtered, and evaporated. The crude product was purified by column chromatography (300–400 mesh silica, dichloromethane) to afford a brown-orange solid. The solid and N,N-diisopropylethylamine (5 mL) were dissolved in absolute dichloromethane (50 mL) under an argon atmosphere and boron trifluoride etherate (8 mL) was added. Stirring was continued for 2 h. The reaction mixture was washed with water and dried with Na2SO4, filtered, and evaporated. The crude product was purified by silica gel chromatography (dichloromethane/hexane = 1:1 v/v) to afford the pure sample. Yield: 15 %. 1H NMR (600 MHz, CDCl3): d = 8.58 (s, 1 H), 8.04 (d, J = 8.5 Hz, 2 H), 7.91 (d, J = 8.8 Hz, 2 H), 7.49 (t, J = 7.3 Hz, 2 H), 7.43 (m, 2 H), 5.89 (s, 2 H), 1.58 (s, 6 H), 1.25 ppm (s, 6 H); 13C NMR (150 MHz, CDCl3): d = 155.80, 142.94, 138.98, 132.38, 131.33, 129.69, 128.37, 128.30, 128.25, 126.99, 125.77, 125.14, 121.20, 22.72, 14.75, 14.17, 13.37 ppm; HRMS (APCI): m/z calcd: 405.1931 [M@F] + , 425.1995 [M+ +H] + ; found: 405.1933, 425.1991. Pyrene-TMBDP (4,4-difluoro-8-pyrryl-1,3,5,7-tetramethyl-4boron-3 a,4 a–diaza-s-indacene): Pyrene-TMBDP was synthesized according to the general procedure by using 1-pyrenecarboxaldehyde as a precursor. The product is bright-green crystals (yield =

2454

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper 23 %). 1H NMR (600 MHz, CDCl3): d = 8.32 (d, J = 7.8 Hz, 1 H), 8.29 (d, J = 7.6 Hz, 1 H), 8.24 (d, J = 7.5 Hz, 1 H), 8.21–8.15 (m, 2 H), 8.10–8.04 (m, 3 H), 7.92 (d, J = 7.8 Hz, 1 H), 5.97 (s, 2 H), 2.65 (s, 6 H), 0.91 ppm (s, 6 H); 13C NMR (151 MHz, CDCl3): d = 155.71, 143.24, 140.55, 132.27, 131.72, 131.25, 131.02, 129.34, 129.29, 128.94, 128.34, 127.31, 126.49, 125.81, 125.78, 125.70, 125.37, 124.67, 124.51, 124.12, 121.29, 14.73, 13.92 ppm; HRMS (APCI) m/z calcd for C29H23BF2N2 : 429.1938 [M@F] + ; found: 429.1924. TMBDP (4,4-difluoro-1,3,5,7-tetramethyl-4-boron-3 a,4 a–diaza-sindacene): TMBDP was synthesized by modifying a literature procedure.[26] 1,2-Dichloroethane (250 mL) was deaerated by bubbling N2. 2,4-Dimethyl pyrrole (1 mL, 11.37 mmol), triethylorthoformate (0.95 mL, 5.69 mmol), and POCl3 (0.58 mL, 6.25 mmol) were added to the deaerated solvent. The reaction was allowed to stir for 2 h at room temperature. Then, NEt3 (11.5 mL) and BF3-etherate (11.5 mL) were added. After 1 h, the reaction was washed with water (3 V 250 mL), the organic layer separated, dried with anhydrous NaSO4, and evaporated under vacuum. Column chromatography with CHCl3 as the eluent yielded the pure product as a reddish solid (400 mg, 28 %). 1H NMR (600 MHz, CDCl3): d = 7.07 (s, 1 H), 6.07 (s, 2 H), 2.55 (s, 6 H), 2.27 ppm (s, 6 H); 13C NMR (151 MHz, CDCl3): d = 156.69, 133.35, 120.06, 118.99, 14.67, 11.29 ppm; HRMS (APCI) calcd: 249.1369 [M+ +H] + , 271.1189 [M+ +Na] + , 519.2485 [2 M+ +Na] + ; found: 249.1356, 271.1174, 519.2455.

The details of photophysical measurements were similar to our previous reports.[27, 28] The absorption and fluorescence spectra, fluorescence quantum yields, and excited singlet-state lifetimes, as well as triplet properties were investigated at room temperature (ca. 24 8C). Steady-state fluorescence spectra were acquired with a FLS 920 instrument. All spectra were corrected for the sensitivity of the photo-multiplier tube. The fluorescence quantum yield (Ff) was measured by using Equation (2),[29] Fs A0 n2s F0 AS n20

ð2Þ

in which F is the integrated fluorescence intensity, A is the absorbance at the excitation wavelength, n is the refractive index of the solvent used, the subscript 0 stands for a reference compound and s represents samples. Fluorescein was used as the reference (Ff = 0.92 in 0.1 m NaOH aq. solution).[30, 31] Excitation wavelengths of 475 nm corresponding to the vibronic band of S0 to S1 transitions were employed. The sample and reference solutions were prepared with the same absorbance (Ai) at the excitation wavelength (near 0.09 per cm). All solutions were air saturated for Ff measurements. The fluorescence lifetime of S1 was measured by a time-correlated single photon counting method (Edinburgh FLS920 spectrophotometer) with excitation at 509 nm with a diode laser (169 ps full width at half maximum (FWHM)) and the emission was monitored at the emission maximum. Fluorescein was used as the reference (tf = 4.16 ns in 0.1 m NaOH aq. solution).[30, 31] Transient absorption spectra were recorded in degassed solution (prepared by bubbling with Argon for 20 min) with an Edinburgh LP920 laser flash photolysis system. An Nd:YAG laser (Brio, 355 nm and 4 ns FWHM) was used as the excitation source. The analyzing light was from a pulsed xenon lamp. The laser and analyzing light beams perpendicularly passed through a quartz cell with an optical path length of 1 cm. The signal was displayed and recorded with a Tektronix TDS 3012B oscilloscope and an R928B detector. The laser energy incident at the sample was attenuated to ca. 10 mJ per Chem. Asian J. 2017, 12, 2447 – 2456

Singlet oxygen quantum yield (FD) determinations in other solvents were carried out by using the chemical trapping method.[6] Typically, a 3 mL portion of the respective PS solutions that contained diphenylisobenzofuran (DPBF) was irradiated at 540 nm in an air-saturated solvent. The FD value was obtained by the relative method by using Equation (3):[32] FD ¼ Fref D

k Iaref kref Ia

ð3Þ

where FDref is the singlet oxygen quantum yield for the standard (8-methylthio-2,6-diiodoBODIPY, FDR = 0.85, practically independent of the solvent, for excitation at 505 nm),[32] k and kref are the DPBF photo-bleaching rate constants in the presence of the respective samples and standard, respectively; Ia and Iaref are the rates of light absorption at the irradiation wavelength of 540 nm by the samples and standard, respectively. Their ratio can be obtained from Equation (4).

Iaref 1 @ 10@Aref ¼ Ia 1 @ 10@A

Photophysical measurements

Ff ¼ F0f

pulse. Time profiles at a series of wavelengths from which pointby-point spectra were assembled were recorded with the aid of a PC-controlled kinetic absorption spectrometer. The concentrations of the target compounds were typically 20 mm, providing A355 = 0.25 in 10 mm cuvettes.

www.chemasianj.org

ð4Þ

To avoid chain reactions induced by DPBF in the presence of singlet oxygen, the concentration of DPBF was lowered to approximately 3 V 10@5 mol dm@3. A solution of sensitizer (absorbance & 0.80 at the irradiation wavelength) that contained DPBF was prepared in the dark and irradiated in the Q-band region. DPBF degradation was monitored by UV/Vis absorption spectroscopy. The error in the determination of FD was & 10 % (determined from several FD values).

Acknowledgments We thank the financial support from Hebei Provincial Hundred Talents Plan (Contract E2013100005), Hebei Provincial Science Foundation (Contract B2014407080), and HBUST (Contract CXTD2012-05).

Conflict of interest The authors declare no conflict of interest. Keywords: aryl-BODIPY · electron transfer · fluorescence · photosensitizers · singlet oxygen · solvent effects

2455

[1] J. F. Lovell, T. W. B. Liu, J. Chen, G. Zheng, Chem. Rev. 2010, 110, 2839 – 2857. [2] R. R. Allison, G. H. Downie, R. Cuenca, X.-H. Hu, C. J. Childs, C. H. Sibata, Photodiagn. Photodyn. Ther. 2004, 1, 27 – 42. [3] M. C. DeRosa, R. J. Crutchley, Coord. Chem. Rev. 2002, 233 – 234, 351 – 371. [4] L. M. Moreira, F. V. dos Santos, J. P. Lyon, M. Maftoum-Costa, C. PachecoSoares, N. S. da Silva, Aust. J. Chem. 2008, 61, 741 – 754. [5] E. A. Lukyanets, J. Porphyrins Phthalocyanines 1999, 3, 424 – 432. [6] T. Nyokong, Coord. Chem. Rev. 2007, 251, 1707 – 1722. T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper [7] C. Tanielian, R. Mechin, R. Seghrouchni, C. Schweitzer, Photochem. Photobiol. 2000, 71, 12 – 19. [8] K. Plaetzer, B. Krammer, J. Berlanda, F. Berr, T. Kiesslich, Lasers Med. Sci. 2009, 24, 259 – 268. [9] C. N. Lunardi, A. C. Tedesco, Curr. Org. Chem. 2005, 9, 813 – 821. [10] N. Mehraban, H. S. Freeman, Materials 2015, 8, 4421 – 4456. [11] J. Zhao, W. Wu, J. Sun, S. Guo, Chem. Soc. Rev. 2013, 42, 5323 – 5351. [12] a) W. Pang, X.-F. Zhang, J. Zhou, C. Yu, E. Hao, L. Jiao, Chem. Commun. 2012, 48, 5437 – 5439; b) X.-F. Zhang, X. Yang, J. Phys. Chem. B 2013, 117, 9050 – 9055. [13] X.-J. Jiang, C.-L. Wong, P.-C. Lo, D. K. P. Ng, Dalton Trans. 2012, 41, 1801 – 1807. [14] T. Wang, Y. Hou, Y. Chen, K. Li, X. Cheng, Q. Zhou, X. Wang, Dalton Trans. 2015, 44, 12726 – 12734. [15] M. Gorbe, A. Barba-Bon, C. de la Torre, S. Gil, A. M. Costero, F. Sancenjn, J. R. Murgu&a, R. Mart&nez-M#Çez, Chem. Asian J. 2015, 10, 2121 – 2125. [16] T.-I. Kim, S. Park, Y. Choi, Y. Kim, Chem. Asian J. 2011, 6, 1358 – 1361. [17] S. V. Mulay, T. Yudhistira, M. Choi, Y. Kim, J. Kim, Y. J. Jang, S. Jon, D. G. Churchill, Chem. Asian J. 2016, 11, 3598 – 3605. [18] J. Zhao, K. Xu, W. Yang, Z. Wang, F. Zhong, Chem. Soc. Rev. 2015, 44, 8904 – 8939. [19] X.-F. Zhang, X. Yang, J. Phys. Chem. B 2013, 117, 5533 – 5539. [20] K. Tram, H. Yan, H. A. Jenkins, S. Vassiliev, D. Bruce, Dyes Pigm. 2009, 82, 392 – 395. [21] R. L. Watley, S. G. Awuah, M. Bio, R. Cantu, H. B. Gobeze, V. N. Nesterov, S. K. Das, F. D’Souza, Y. You, Chem. Asian J. 2015, 10, 1335 – 1343.

Chem. Asian J. 2017, 12, 2447 – 2456

www.chemasianj.org

[22] N. Epelde-Elezcano, E. Palao, H. Manzano, A. Prieto-CastaÇeda, A. R. Agarrabeitia, A. Tabero, A. Villanueva, S. de la Moya, 2. Ljpez-Arbeloa, V. Mart&nez-Mart&nez, M. J. Ortiz, Chem. Eur. J. 2017, 23, 4837 – 4848. [23] E. S. Pysh, N. C. Yang, J. Am. Chem. Soc. 1963, 85, 2124 – 2130. [24] L. Meites, P. Zuman, CRC Handbook Series in Organic Electrochemistry, CRC Press, Boca Raton, FL, 1977 – 1982. [25] C.-W. Wan, A. Burghart, J. Chen, F. Bergstroem, L. B.-A. Johansson, M. F. Wolford, T. G. Kim, M. R. Topp, R. M. Hochstrasser, K. Burgess, Chem. Eur. J. 2003, 9, 4430 – 4441. [26] K. Umezawa, Y. Nakamura, H. Makino, D. Citterio, K. B. Suzuki, J. Am. Chem. Soc. 2008, 130, 1550 – 1551. [27] X.-F. Zhang, J. Fluoresc. 2011, 21, 1559 – 1564. [28] X.-F. Zhang, W. Guo, J. Phys. Chem. A 2012, 116, 7651 – 7657. [29] J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer, Singapore, 2006, p. 954. [30] D. Magde, R. Wong, P. G. Seybold, Photochem. Photobiol. 2002, 75, 327 – 334. [31] X.-F. Zhang, J. Zhang, L. Liu, J. Fluoresc. 2014, 24, 819 – 826. [32] N. Epelde-Elezcano, V. Mart&nez-Mart&nez, E. PeÇa-Cabrera, C. F. A. Gjmez-Dur#n, I. L. Arbeloa, S. Lacombe, RSC Adv. 2016, 6, 41991 – 41998. Manuscript received: June 7, 2017 Revised manuscript received: July 1, 2017 Accepted manuscript online: July 13, 2017 Version of record online: August 30, 2017

2456

T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim