Efficient energy transfer in ethynyl bridged corrole–BODIPY dyads

Abstract

A series of new corrole–BODIPY dyads bridged by ethynyl linker moieties have been synthesized in high yields and fully characterized. The direction of energy transfer upon electronic excitation has been explored, and was found to be dependent on the number of corrole rings and their connection position on the BODIPY core. Intense bands in the absorption spectrum cover most of the visible region, which is potentially advantageous for capturing solar energy. Studies on the excitation spectra and lifetimes suggest that the energy transfer efficiency between the BODIPY and corrole moieties reaches almost 85%, which appears to be efficient in the context of energy transfer within the singlet manifold.

---

Introduction

In recent years, a wide range of compounds have been used to construct artificial photosynthetic systems that have the ability to convert incident sunlight into chemical energy. The development of single-molecule panchromatic donor–π–accepter (D–π–A) sensitizers as part of the efforts to improve the overall power conversion efficiency (PCE) of dye-sensitized solar cells (DSCs) devices remains a molecular engineering challenge.¹ Recently, the favorable photophysical properties of porphyrin–BODIPY (BODIPY = boron dipyrromethene) conjugates have gained attention in this regard.² BODIPYs are excellent antenna molecules that can be used as light harvesting chromophores to absorb light in the green region of the spectrum, where porphyrins lack significant absorption intensity. Various porphyrin derivatives have been explored to serve as the principal building blocks for the construction of a wide variety of synthetic multicomponent light-harvesting arrays.³ Single-molecule interfacial electron transfer dynamics have also been reported for porphyrins on TiO₂ nanoparticles,⁴ which provide a detailed molecular level understanding of the inhomogeneous interfacial electron transfer (ET) reactivity and the factors that influence it. One group of compounds that have been studied are the corroles, which are tetrapyrrolic ligands with three meso-carbons and one direct pyrrole–pyrrole bond.⁵ Corroles have similar optical properties to porphyrins with higher fluorescence quantum yields.⁶ During the last two decades, they have been studied extensively for use in numerous applications such as solar cells, photochemical sensors, artificial photosynthesis,⁷ and biomedical applications.⁸

Recently, there has been a strong research focus on the synthesis of corrole–BODIPY dyads, which can act as light-harvesting antennas for the photochemical conversion of solar energy. A number of corrole–BODIPY dyads have been reported with various covalent linkers, such as direct meso-β bonds⁹ and triazole moieties.¹⁰ In this study, the synthesis and spectroscopic characterization of new covalently linked corrole derivatives bridged by ethynyl linkage is reported. As anticipated, these corrole–BODIPY conjugates absorb strongly in the 400–700 nm region, across almost the entire visible region. To the best of our knowledge, no previous examples of ethynyl-linked corrole–BODIPY conjugates have been reported. Energy-transfer can occur when a dyad compound consisting of two or more fluorophores are connected with each other.¹¹ One unit acts as an energy donor and the other as an acceptor. The energy absorbed from an incident photon of light by the donor is quickly transferred to the acceptor, resulting in emission at a longer wavelength. Many factors can influence the process of energy-transfer, such as differences in the linker groups that are used,¹² the spectral overlap of the donor emission with the acceptor absorbance, the effectiveness of other relaxation pathways¹³ and the relative orientation of the donor and acceptor moieties. In this work, the photophysical properties of a series of novel corrole–BODIPY conjugates with ethynyl linkers are systematically investigated in order to elucidate the energy transfer process and interactions between the donor and acceptor chromophores.

Results and discussion

Synthesis

Corrole 4 was synthesized using the general literature method for A₂B corroles.¹³ The preparation involves the condensation of an aldehyde (1 equiv.) and 5-(pentafluorophenyl)dipyrromethane (2 equiv.) in the presence of catalytic amount of aqueous HCl (36%) in MeOH/H₂O (1 : 1). This method is particularly advantageous for the separation of the corrole, since it prevents the formation of by-products. BODIPY 1 was obtained using literature procedures,¹⁴ and BODIPYs 2 and 3 were synthesized using a Knoevenagel condensation reaction¹⁵ between a meso-phenyl-substituted 1,3,5,7-tetramethyl-BODIPY and 1 equiv. of 4-iodobenzaldehyde in toluene in the presence of p-toluenesulfonic acid (PTSA) and piperidine. Although Sonogashira coupling reactions are generally co-catalyzed by CuI and palladium, with an amine present as a base and a phosphine as a ligand,¹⁶ dyads 5–7 were synthesized by using Pd₂(dba)₃ as the catalyst (Scheme 1).¹⁷ The alkynylation reactions (details provided as ESI†) readily occur and provide 5–7, respectively, in yields of 74, 85 and 77%.

Scheme 1 Synthesis of corrole–BODIPY dyads 5–7.

Electronic absorption spectra

Fig. 1(A)–(C) contains the UV-visible absorption spectra of 1–7 in toluene, and the related data are summarized in Table 1. The spectrum of 5 has two intense bands at 427 and 505 nm, which is almost identical to what would be expected based on the sum of the spectra of BODIPY 1 and corrole 4. In marked contrast, the styryl–ethynyl linker clearly influences the absorption spectrum of 6 and 7. The spectrum of 6 has two intense bands at 427 and 578 nm, since the latter exhibits a red-shift of ca. 11 nm when compared to the main spectral band in the spectrum of BODIPY 2. The spectrum of 7 contains two intense bands at 427 and 658 nm, with a red-shift of about 21 nm for the latter compared to the analogous band in the spectrum of BODIPY 3. The absorption spectra of conjugates 6 and 7 cover most of the visible region (ca. 300–700 nm), in a manner that is potentially advantageous for capturing solar energy.

Fig. 1 UV-visible absorption spectra of 1–7 in toluene at concentration of 1 × 10⁻⁵ M at 298 K.

Table 1 Absorption, luminescence data and fluorescence quantum yield (ϕ) values for 1–7 measured at 298 K in toluene

	λabs (ε/10^5 M^−1 cm^−1)			λem(max) (nm)		ϕ^a
	Corrole		BODIPY	Corrole	BODIPY
	B band	Q band
a Tetraphenylporphyrin was used as the standard (ϕF = 0.11), λex = 422 nm.b Total quantum yield (BODIPY and corrole).
1			505 (1.06)		522	0.64
2			567 (0.95)		582	0.61
3			637 (1.46)		654	0.19
4	422 (1.37)	564 (0.22)		664		0.17
		615 (0.13)
5	427 (1.60)	566 (0.24)	505 (1.23)	664	521	0.15^b
		615 (0.14)
6	424 (1.12)	613 (0.17)	578 (1.47)	664	594	0.16^b
7	427 (2.58)	610 (0.64)	658 (1.29)		676	0.36^b

---

Theoretical calculations

DFT and TD-DFT calculations were carried out for 1–7, so that trends in the electronic structures could be identified (Fig. 2 and 3). The B3LYP functional of the Gaussian 09 software package¹⁸ was used with 6-31G(d) basis sets for the geometry optimizations, and the CAM-B3LYP functional was used for the TD-DFT calculation, since it includes a long-range correction making it suitable for compounds where there is scope for transitions with charge transfer character.

Fig. 2 The angular nodal patterns and MO energies of the HOMO and LUMO of 1 and the a, s, −a and −s MOs of 4 at an isosurface value of 0.04 a.u. Trends in the MO energies of 1–7. The HOMO and LUMO of the BODIPY moiety are highlighted with gray circles, while crosses and black triangles are used for the a/−a and s/−s MOs of Michl's perimeter model,¹⁷ respectively. The HOMO–LUMO gap for the BODIPY moiety and the average HOMO–LUMO gap taking into account the a, s, −a and −s MOs are highlighted with blue and red diamonds, respectively, and are plotted against a secondary axis. Structures with the non-protonated pyrrole nitrogens adjacent to the direct pyrrole–pyrrole ring are denoted as 4a, 5a, 6a, 7a, while those with the non-protonated pyrrole nitrogens adjacent to the y-axis meso-phenyl ring are denoted as 4b, 5b, 6b and 7c. 7b is a structure with both types of NH-tautomers.

Fig. 3 The calculated electronic absorption spectra calculated for the B3LYP optimized geometries of 1–7 by using the CAM-B3LYP functional with 6-31G(d) basis sets. The observed spectra are plotted against a secondary axis. The Q and B bands of the corrole rings are highlighted with red diamonds, while blue diamonds are used to denote the main spectral band of the BODIPY chromophore. The different possible NH-tautomers for the corrole rings are determined. Structures with the non-protonated pyrrole nitrogens adjacent to the direct pyrrole–pyrrole ring are denoted as 4a, 5a, 6a, 7a, while those with the non-protonated pyrrole nitrogens adjacent to the y-axis meso-phenyl ring are denoted as 4b, 5b, 6b and 7c. 7b is a structure with both types of NH-tautomers. The details of the calculations are provided as ESI.†

When 4-iodo-styryl groups are added at the 3,5-positions of the BODIPY core to form 2 and 3, there is a significant narrowing of the HOMO–LUMO gap due to the extension of the π-conjugation system, and this results in a marked red-shift of the main BODIPY spectral band (Fig. 2 and 3). Similar trends are observed in the energies of the MOs derived from the HOMO and LUMO of the BODIPY core in the electronic structures of 6 and 7 and this accounts for the observed red shift of the intense absorption band that is associated with the BODIPY moiety from 505 nm for 5, to 577 and 658 nm for 6 and 7 (Fig. 1 and 3), respectively.

In contrast, the introduction of the BODIPY core and ethynyl linker has a smaller effect on the energies of the frontier π-MOs of corrole ring relative to those of corrole 4, and this means the main spectral bands associated with the corrole ring remain largely unchanged in the spectra of dyads 5–7. The electronic structures of corroles result from perturbations to an M_L = 0, ±1, ±2, ±3, ±4, ±5, ±6, ±7 sequence of the MOs of a parent D_15h symmetry C₁₅H₁₅³⁻ perimeter corresponding to the inner ligand perimeter.¹⁹ The HOMO and LUMO of the parent C₁₅H₁₅³⁻ perimeter have M_L = ±4 and ±5 nodal properties, respectively. In the context of Gouterman's 4-orbital model²⁰ this results in allowed B (or Soret) and forbidden Q bands based on ΔM_L = ±1 and ±9 transitions, respectively, in the 400–450 and 500–650 nm regions. Michl introduced an a, s, −a and −s nomenclature for MOs whether there is a π-system (Fig. 2).²¹ Since the ethynyl linker moiety is attached to a meso-aryl ring that does not lie coplanar with the corrole ring (Fig. 4), there is an inductive rather than a mesomeric effect on the frontier π-MOs of the corrole ring, which tends to affect the energy of each MO in a similar manner regardless of the size of the MO coefficients at the point of attachment, in marked contrast with what happens with the HOMO and LUMO of the BODIPY core. As a result, the wavelengths of the main corrole bands remain largely unchanged from those of 4. For example the B band of 4 at 421 nm shifts to 426, 423 and 428 nm (Fig. 1 and 3), respectively. The possible effect of NH-tautomerism was taken into consideration, based on whether the nonprotonated pyrrole nitrogen is adjacent to the non-fluorinated meso-aryl moiety. The possible tautomers were found to have energy differences of less than 1 kcal mol⁻¹. Significant differences are predicted in the energies of the Q and B bands (Fig. 3), as has been reported previously for free base corroles.²²

Fig. 4 The B3LYP optimized geometries for one of the NH-tautomers of dyads 5–7. The 5a and 6a structures have non-protonated pyrrole nitrogens adjacent to the direct pyrrole–pyrrole ring, while the 7c structure has both non-protonated pyrrole nitrogens adjacent to the y-axis meso-phenyl ring.

Steady-state and time-resolved fluorescence spectroscopy

The details of the emission spectroscopy are provided as ESI.† Fig. S7 and S8† provide the absorption and fluorescence spectra of BODIPY 1–3, corrole 4 and dyads 5–7. These spectra demonstrate that the introduction of π-conjugated styryl substituents induce a bathochromic shift of the absorption and fluorescence bands (Fig. S9†). Each dyad displays a fluorescence band near 664 nm that is readily attributable to the fluorescence of the corrole ring along with a band that is attributable to the BODIPY moiety, which lies either at higher (550 nm for 5, and 594 nm for 6) or at lower energy (676 nm for 7). It is noteworthy that only one strong band is observed at 676 nm, which is attributed to both the BODIPY 3 and corrole 4 units. On the basis of the fluorescence positions, it can be deduced that the BODIPY chromophore can probably act as either an energy donor (in 5 and 6) or acceptor (in 7).

Transient fluorescence spectra were recorded by the method described in the Experimental section (provided as ESI†). Fluorescence quantum yields were determined using free base tetraphenylporphyrin as a standard (ϕ_F = 0.11).²³ The fluorescence emission maxima (λ_em), quantum yields (ϕ) and lifetime (τ) values in toluene are summarized in Tables 1 and 2. The model BODIPY compounds 1 and 2 have high fluorescence quantum yields (ϕ) of over 0.6, while in contrast BODIPY 3 and corrole 4 display low ϕ values. Although the ϕ values of dyads are total quantum yield values that include both the BODIPY and corrole moieties, the values are lower than that of their BODIPY and corrole precursors, indicating that additional non-radiative processes have occurred.²⁴

Table 2 Fluorescence lifetimes (τ) for 1–7 measured at 298 K in toluene and the energy transfer efficiencies (ET_eff) and rates (K_ET) values

Dye	λexc	λem	τ (ns)	D–A	ETeff	KET (10^8 s^−1)
1	480	522	3.7	—	—	—
2	540	582	5.8	—	—	—
3	620	654	5.4	—	—	—
4	422	664	4.0	—	—	—
5	480	522	1.1	BDP to Cor	69.3%	0.63
		664	3.4
6	540	594	0.9	BDP to Cor	84.5%	0.93
		664	3.2
7	422	676	2.2	Cor to BDP	58.1%	0.27

---

The fluorescence spectra of BODIPY 1 and corrole 4 in toluene exhibit the characteristic emission peaks for these fluorophores at 522 and 664 nm, respectively. The emission spectrum of conjugate 5, in which BODIPY is linked at the meso-phenyl ring of the corrole 4, exhibits wavelength dependent emission spectra. When a comparison is made with the absorption spectrum, the corrole ring of 5 exhibits an intense band at 422 nm and the BODIPY ring can be expected to have minimal absorbance at this wavelength. In a similar manner, the corrole ring can be expected to lack any significant absorbance at 480 nm, where the BODIPY moiety absorbs strongly. Thus allowing for selective excitation of the chromophore of BODIPY 1 for the examination of energy transfer (ET) processes from the BODIPY moiety to the corrole ring. To study the energy transfer process between the BODIPY and corrole moieties of dyad 5, 4 and 5 are excited at 422 nm and 480 nm. The related data are summarized in Fig. 5(A), S9(A)† and Table 1. Upon excitation at 422 nm (in the B band of the corrole π-system), toluene solution of the fluorescence spectrum of 5 contains only one emission band at 664 nm, which is consistent with the emission from the corrole moiety. It should be noted, that corrole 4 has almost the same fluorescence intensity as dyad 5 (Fig. S9(A)†). So there is no evidence for ET to the BODIPY core from corrole in this context. In contrast, upon excitation at 480 nm, two emission bands are observed at 522 and 664 nm, due to emission from the BODIPY and corrole fluorophores, respectively. The fluorescence emission intensity observed at 664 nm is more intense than that obtained by direct excitation of corrole 4 at 422 nm. These observations clearly demonstrate that ET occurs at this excitation wavelength. Since only part of the energy absorbed by the BODIPY donor moiety is transferred to the corrole unit, there is a weak residual emission band that can be attributed to the donor. Reverse energy transfer from the corrole ring (accepter) to the BODIPY (donor) core is not feasible. Fig. 6(A) overlays the fluorescence spectra of the donor (BODIPY 1) with the absorption spectra of the acceptor (corrole 4), and this provides an explanation for the direction of ET.

Fig. 5 Emission spectra of 3–7 in aerated toluene (1 μM) at room temperature. (A) Emission spectra of 4 and 5 (λ_ex = 480 nm). (B) Emission spectra of 4 and 6 (λ_ex = 540 nm). (C) Emission spectra of 3, 4 and 7 (λ_ex = 422 nm).

Fig. 6 Superposition of the normalized absorption and fluorescence spectra in aerated toluene at 298 K of (A) 4 (black) with 1 (red), (B) 4 (black) with 2 (red), (C) 3 (black) with 4 (red). The absorption and fluorescence spectra of the BODIPYs and corrole are in red and black, respectively. The overlaps between the spectra are shaded in gray-blue.

Toluene solutions of 4 and 6 were excited at 540 nm at room temperature to study the ET between the corrole and BODIPY moieties. The related data are summarized in Fig. 5(B) and Table 1. The fluorescence spectrum of 6 contains a band at 664 nm. In contrast with the spectra of 5, when different excitation wavelengths are selected, there is no indication of any other emission bands, suggesting that there is efficient fluorescence quenching of the BODIPY moiety. It is noteworthy that the fluorescence intensity of 6 is stronger than that obtained by direct excitation of corrole 4. According to Fig. 6(B), the fluorescence spectra of the donor (similar to BODIPY 2) overlays part of the absorption spectra of the acceptor (similar to corrole 4), providing a strong indication of the direction of energy transfer. As a result, an efficient ET process between the BODIPY moiety and corrole ring can be anticipated to exist, and the obvious explanation is that all of the energy absorbed by the BODIPY donor moiety is transferred to the corrole ring.

To explore the ET process between two corrole rings and the BODIPY core of 7, toluene solutions of 3, 4 and 7 were excited at 422 nm. The related data are summarized in Fig. 5(C) and Table 1. In contrast with the typical corrole emission that is observed in the spectra of 5 and 6 at 664 nm, the fluorescence spectrum of 7 has a single emission maximum at 676 nm, due to the superposition of the BODIPY and corrole fluorescence bands.

The fluorescence intensity of 7 is stronger than that obtained by direct excitation of corrole 4 (Fig. 5(C)). These observations indicate that part of the energy absorbed by the corrole donor moiety is transferred to the BODIPY unit. It should be noted that although 7 has almost the same fluorescence intensity as 3, upon excitation at 620 nm (in the absorption band of the BODIPY 3 chromophore), there is a hypsochromatic shift of ca. 23 nm compared to BODIPY 3 (Fig. S10(C)†), due to the π-conjugated-styryl substituent (Fig. S9(B) in ESI† and Table 1). The fluorescence spectra of the donor (similar to that of corrole 4) overlays part of the absorption spectra of the acceptor (similar to that of BODIPY 3) from the Fig. 6(C). The fluorescence excitation and emission data provide strong evidence for efficient ET between the corrole ring and the BODIPY core.

The fluorescence lifetimes of each dyad were measured separately at the fluorescence maxima of the corrole and the BODIPY chromophores. Dyads 5 and 6 exhibit two emissions from the BODIPY (donor, 522 nm for 5 and 594 nm for 6) and the corrole (acceptor, 664 nm), in comparison with the model compound corrole 4, the slight decrease in the lifetime values indicates the presence of a small nonradiative process associated with the incorporation of the flexible chain.²⁴ The ET efficiencies were obtained from the lifetime values according to ET_eff = ((1/τ) − (1/τ⁰))/(1/τ) (where τ⁰ and τ are the fluorescence lifetimes of the donor in the absence and presence of the acceptor, respectively).²⁵ The values obtained were 69.3% for 5, 84.5% for 6 and 58.1% for 7, Table 2. This method is more accurate than measuring the I_F values for strongly overlapping fluorescence spectra, since the fluorescence intensity cannot be readily obtained in each case. The high ET_eff values demonstrate that corrole–BODIPY dyads 5–7 provide an effective way to construct energy transfer systems. The rates of energy transfer (K_ET) were also evaluated, since K_ET = (1/τ) − (1/τ⁰),²⁵ Table 2. Qualitative analyses of the ET efficiency and rate data are the standard criteria for the Förster resonance energy transfer (FRET) mechanism, and are highly dependent on the center-to-center donor–acceptor separation, and the spectral overlap of the donor emission and the acceptor absorption.

Conclusion

Several novel corrole–BODIPY conjugates with ethynyl linkers have been prepared in high yields through Sonogashira coupling reactions, which have absorption spectra with intense bands that cover most of the visible region (300–700 nm). The mechanism and direction of energy transfer has been explored, and is found to depend on the number of corrole rings and their linkage position on the BODIPY core. In dyads 5 and 6, there is efficient excited state ET from the BODIPY antenna chromophore to the corrole ring, while the reverse process is observed for dyad 7. A combination of the typical emission peaks of the BODIPY core and the corrole ring is observed in the spectrum of 5, but only one emission peak is observed for 6 at 664 nm, due to efficient quenching of the BODIPY fluorescence. An absorption band lies at 578 nm, which is red-shifted by ca. 11 nm compared with BODIPY 2. Dyad 7 has only one emission band at 676 nm, which is red-shifted by ca. 23 nm compared with that of BODIPY 3. These energy transfers are extremely efficient with respect to other corrole–BODIPY conjugates that have been reported, and this demonstrates that ethynyl linkages provide an effective way to construct energy transfer systems. These properties make the dyads potentially suitable for applications in solar cells and OLEDs.

Acknowledgements

Financial support was provided by the Major State Basic Research Development Program of China (Grant No. 2013CB922101 & 2011CB808704), the National Natural Science Foundation of China (No. 21371090), the Natural Science Foundation of Jiangsu Province (BK20130054) to ZS, the China-South Africa joint research program (CS08-L07 and UID: 95421) to ZS and JM, and an NRF of South Africa CSUR grant (93627) to JM. The theoretical calculations were carried out at the Centre for High Performance Computing in Cape Town.

Notes and references

1. S. Mathew, A. Yella, P. Gao, R. Humphry-Baker, B. F. E. Curchod, N. Ashari-Astani, I. Tavernelli, U. Rothlisberger, M. K. Nazeeruddin and M. Gratzel, Nat. Chem., 2014, 6, 242–247.
2. C. Y. Lee and J. T. Hupp, Langmuir, 2009, 26, 3760–3765.
3. A. K. Burrell, D. L. Officer, P. G. Plieger and D. C. Reid, Chem. Rev., 2001, 101, 2751–2796.
4. (a) V. G. Rao, B. Dhital, Y. F. He and H. P. Lu, J. Phys. Chem. C, 2014, 118, 20209–20221; (b) V. G. Rao, B. Dhital and H. P. Lu, Chem. Commun., 2015, 51, 16821–16824.
5. (a) C. Erben, S. Will and K. M. Kadish, in The Porphyrin Handbook, ed. K. M. Kadish, K. M. Smith and R. Guilard, Academic Press, Inc., London, 2000, vol. 2, pp. 232–300; (b) R. Guilard, J. M. Barbe, C. Stern and K. M. Kadish, in The Porphyrin Handbook, ed. K. M.Kadish, K. M. Smith and R. Guilard, Academic Press, Inc., London, 2003, vol. 18, pp. 303–351.
6. (a) C. L. He, F. L. Ren, X. B. Zhang and Z. X. Han, Talanta, 2006, 70, 364–369; (b) T. Ding, E. A. Aleman, D. A. Modarelli and C. J. Ziegler, J. Phys. Chem. A, 2005, 109, 7411–7417.
7. (a) B. Brizet, N. Desbois, A. Bonnot, A. Langlois, A. Dubois, J.-M. Barbe, C. P. Gros, C. Goze, F. Denat and P. D. Harvey, Inorg. Chem., 2014, 53, 3392–3403; (b) M. Tasior, D. T. Gryko, M. Cembor, J. S. Jaworski, B. Ventura and L. Flamigni, New J. Chem., 2007, 31, 247–259.
8. (a) D. Aviezer, S. Cotton, M. David, A. Segev, N. Khaselev, N. Galili, Z. Gross and A. Yayon, Cancer Res., 2000, 60, 2973–2980; (b) A. Mahammed, H. B. Gray, J. J. Weaver, K. Sorasaenee and Z. Gross, Bioconjugate Chem., 2004, 15, 738–746.
9. M. Murugavel, R. R. Reddy and J. Sankar, RSC Adv., 2014, 4, 13669–13672.
10. (a) Q. Q. Hu, Y. Z. Zhu, S. C. Zhang, Y. Z. Tong and J. Y. Zheng, Dalton Trans., 2015, 44, 15523–15530; (b) B. Basumatary, A. Raja Sekhar, R. Ramana Reddy and J. Sankar, Inorg. Chem., 2015, 54, 4257–4267.
11. (a) L. Stryer and R. P. Haugland, Proc. Natl. Acad. Sci. U. S. A., 1967, 58, 719–726; (b) G. S. Engel, T. R. Calhoun, E. L. Read, T.-K. Ahn, T. Mančal, Y. C. Cheng, R. E. Blankenship and G. R. Fleming, Nature, 2007, 446, 782–786.
12. M. Inokuti and F. Hirayama, J. Chem. Phys., 1965, 43, 1978–1989.
13. (a) M. J. Plater, S. Aiken and G. Bourhill, Tetrahedron, 2002, 58, 2405–2413; (b) D. T. Gryko and K. Jadach, J. Org. Chem., 2001, 66, 4267–4275.
14. S. Mula, S. Frein, V. Russo, G. Ulrich, R. Ziessel, J. Barberá and R. Deschenaux, Chem. Mater., 2015, 27, 2332–2342.
15. R. Ziessel, T. Bura and J. H. Olivier, Synlett, 2010, 2304–2310.
16. L. H. Tong, S. I. Pascu, T. Jarrosson and J. K. Sanders, Chem. Commun., 2006, 1085–1087.
17. (a) R. R. Tykwinski, Angew. Chem., Int. Ed., 2003, 42, 1566–1568; (b) E. I. Negishi, J. Organomet. Chem., 2002, 653, 34–40.
18. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford, CT, 2009.
19. (a) X. Liang, J. Mack, L. Zheng, Z. Shen and N. Kobayashi, Inorg. Chem., 2014, 53, 2397; (b) J. Mack, M. Bunya, D. Lansky, D. P. Goldberg and N. Kobayashi, Heterocycles, 2008, 76, 1369.
20. M. Gouterman, in The Porphyrins, ed. D. Dolphin, Academic Press, New York, 1978, vol. III, Part A, pp. 1–165.
21. (a) J. Michl, J. Am. Chem. Soc., 1978, 100, 6801; (b) J. Michl, J. Am. Chem. Soc., 1978, 100, 6812; (c) J. Michl, Pure Appl. Chem., 1980, 52, 1549; (d) J. Michl, Tetrahedron, 1984, 40, 3845.
22. (a) C. J. Ziegler, J. R. Sabin, G. R. Geier and V. N. Nemykin, Chem. Commun., 2012, 48, 4743; (b) M. Li, X. Liang, J. Mack, S. Mkhize, T. Nyokong and W. Zhu, J. Porphyrins Phthalocyanines, 2016 , submitted for publication.
23. Y. Kurabayashi, K. Kikuchi, H. Kokubun, Y. Kaizu and H. Kobayashi, J. Phys. Chem., 1984, 88, 1308–1310.
24. B. Brizet, N. Desbois, A. Bonnot, A. Langlois, A. Dubois, J. M. Barbe, C. P. Gros, C. Goze, F. Denat and P. D. Harvey, Inorg. Chem., 2014, 53, 3392–3403.
25. B. Brizet, N. Desbois, A. Bonnot, A. Langlois, A. Dubois, J. M. Barbe, C. P. Gros, C. Goze, F. Denat and P. D. Harvey, Inorg. Chem., 2014, 53, 3392–3403.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental details, ¹H- and ¹³C-NMR spectra, fluorescence spectra and details of theoretical calculations. See DOI: 10.1039/c6ra12271j

‡ These authors contributed equally to this work.

---