First attempts to employ the four-component coupling-isomerization-Stetter-Paal-Knorr pyrrole synthesis (Braun et al., 2001; Braun and Müller, 2004) to introduce the second chromophore in a one-pot fashion failed, predominantly due to solubility issues. Therefore, we envisioned that a phenol containing pyrrole might ideally serve for a post MCR ligation by etherification or esterification. Therefore, performing the four-component pyrrole synthesis with THP-protected para-hydroxybenzaldehyde gave rise to the formation of 4-(1-benzyl-2-(4-hydroxy-phenyl)-5-phenyl-1H-pyrrol-3-yl)benzonitrile (1) after isolation and chromatographic purification in 26% (Scheme 1). Under the acidic conditions of the terminal Paal-Knorr cyclocondensation the THP group is cleaved to give the free phenol.

Absorption spectroscopy of phenol 1 reveals a longest wavelength absorption maximum at 319 nm and an additional absorption band at 275 nm. Upon excitation at the longest wavelength maximum a broad intense emission at 442 nm is found with a relative fluorescence quantum yield Φ_F of 0.11, i.e., in a comparable magnitude as other benzonitrile substituted pyrroles (Braun and Müller, 2004) (vide infra). With this intensively blue emissive building block in hand the stage was set for the synthesis of luminescent bichromophores in the sense of a level-2 functionalization.

The reference luminophore 2, a methyl ether derivative, was synthesized by etherification of phenol 1 with methyl iodide in 68% yield. Halide functionalized luminophores 3 (Aathimanikandan et al., 2005) and 5 (Kucherak et al., 2010), 6 (Lord et al., 2010), and 7 (Lord et al., 2010) (some also containing alkyl spacers) were prepared according to literature protocols (dansyl chloride was purchased and used as received) and submitted to base mediated Williamson ether synthesis or sulfonylation with phenol 1 to give bichromophores 8–12 in moderate to excellent yield (Scheme 2). While the bromides 3 and 5, dansyl chloride 4 and methyl iodide react smoothly with potassium carbonate as a base, for the chlorides 6 and 7, Finkelstein conditions with potassium iodide have to applied, which work best in these cases with sodium hydride as a base.

The ¹H and ¹³C NMR and mass spectra (MALDI-TOF and HRMS) unambiguously confirm the successful ligation of the 1 with the second chromophores 3–7 as well as methyl iodide, and thereby the structures of pyrrole reference chromophore 2 and the bichromophores 8–12.

Upon excitation with a handheld UV lamp the pyrrole reference chromophore 2 and all bichromophores 8–12 luminesce, indicating that the photonic excitation is neither completely quenched by internal conversion nor by electron transfer into long-lived charge separated states arising from photoinduced intramolecular electron transfer (PIET) (Kavarnos, 1993). This prompted us to study the absorption and emission spectra of the luminophores 2 and 8–12 in more detail (Table 1).

The absorption spectra of all bichromophores 8–12 behave essentially additively with respect to the underlying subchromophores. This is quantitatively demonstrated by comparison of the UV/Vis spectrum of the pyrrole-anthracene bichromophore 8 and the sum spectra of the reference chromophores 2 (pyrrole) and 9-methyl anthracene (anthracene) (for spectral details, see Supplementary Material). As expected, in the electronic ground state, from where photonic excitation starts, the two subchromophores essentially neither interact nor form aggregates at the concentrations investigated. However, in the excited state, as studied by fluorescence spectroscopy, in some cases a cooperative behavior exists, which is also dependent on the excited subchromophore.

Upon excitation of the pyrrole-anthracene bichromophore 8 at the absorption bands of the anthracene chromophore at 260, 367, and 387 nm clearly the emission bands of anthracene with vibrational resolution are detected (for spectral details, see Supplementary Material). However, upon excitation at the pyrrole absorption bands at 317 and 330 nm exclusively the pyrrole typical structureless emission band at 443.5 nm appears (Figure 2). The separate excitations of the discrete absorption bands of the subchromophores in bichromophore 8 are not accompanied by energy transfer from the donor (pyrrole) to the acceptor (anthracene). Although the fluorescence quantum yield Φ_f of 0.09 accounts for a significant energy dissipation upon excitation of the anthracene moiety, the lack of overlap of the absorption band of the acceptor with the emission band of the donor and, hence, the absence of energy transfer suggests that in bichromophore 8 are not electronically coupled.

In the other bichromophores 9–12 bearing redshifted absorptions of the acceptor chromophores the situation of the emission characteristics changes (for spectral details, see Supplementary Material).

The absorption bands of the dansyl-pyrrole (9), the Nile red-pyrrole (10), and the quinoxalinyl-styryl-pyrrole (12) bichromophores clearly possess significant energy transfer characteristics, where the selective excitation of the pyrrole donor chromophore with its emission band more (10, 12) or less (9) overlaps with the absorption bands of the corresponding acceptor chromophores. The efficiency of the energy transfer Φ_EnT can be estimated according to the formula (Qing et al., 2011) ΦEnT=1- ΦDonorΦDonor0, where Φ_donor is the measured quantum yield of the (residual) donor emission band of the bichromophore and Φdonor0 is the quantum yield of the donor chromophore. Determination of Φ_Donor of the residual pyrrole emission at 445 nm furnishes for bichromophore 12 a quantum yield of the energy transfer Φ_EnT of over 0.97. For the dansyl bichromophore 9 Φ_EnT can be estimated to be >0.98, while for the Nile red bichromophore 10 no residual emission around 450 nm can be detected, indicating an Φ_EnT of unity. However, it has to be kept in mind that an exergonic photoinduced intramolecular electron transfer (PIET) cannot be fully excluded, in particular, since the determined fluorescence quantum yields are relatively low. For the anthryl-pyrrole bichromophore 8 the PIET can be estimated to be endothermic.

Most remarkable, however, are the emission characteristics of the 3-cyano-5,5-dimethylfuran-2(5H)-ylidene)malononitrile-styryl bichromophore 11, where a strong dependence on the excitation wavelength λ_exc can be detected. The dichloromethane solution of bichromophore 11 does not show the typical red emission of the acceptor chromophore upon eyesight at excitation with 254 or 365 nm by a handheld UV lamp, but rather a magenta-rose emission with significant intensity (Figure 3, top). This mixing emission color between violet (λ_{max, em} = 442 nm) and orange red (λ_{max, em} = 618 nm) almost matches with the line of purples (Westland, 2003; Broadbent, 2004; Schanda, 2007). The emission spectra at various excitation wavelength clearly reveal that an excitation at 365 nm not only causes an energy transfer from the pyrrole donor to the 3-cyano-5,5-dimethylfuran-2(5H)-ylidene)malononitrile-styryl acceptor (Figure 3, bottom), which emits at 618 nm, but also an emission of the donor itself at 442 nm. This peculiar behavior can be interpreted in the sense of a partial energy transfer (for representative small molecule emitters operated by frustrated and partial energy transfer by ESIPT, see e.g., Park et al., 2009; Kwon et al., 2013; Benelhadj et al., 2014), i.e., a dual emission as a consequence of an excited state communication between donor and acceptor.

For a deeper understanding of the observed chromophore-chromophore interactions TD-DFT calculations were performed on the pyrrole 2 and the bichromophores 8–12. The geometries of the electronic ground-state and excited structures were optimized by using Gaussian09 (Frisch et al., 2009) with the B3LYP functional (Lee et al., 1988; Becke, 1993; Kim and Jordan, 1994; Stephens et al., 1994) and the Pople 6-311G^** basis set (Krishnan et al., 1980). Since absorption properties were measured in dichloromethane solutions, the polarizable continuum model (PCM) with dichloromethane as a solvent was utilized (Scalmani and Frisch, 2010). All minimum structures were unambiguously assigned by analytical frequency analysis. The optimized structures were then submitted to TD-DFT calculations employing the gradient-corrected exchange and correlation Perdew-Burke-Ernzerhof functionals PBE₁PBE (Adamo and Barone, 1999)/6-311^** (Krishnan et al., 1980) with dichloromethane (IEFPCM) (Scalmani and Frisch, 2010) as a solvent.

The blue emissive pyrrole 2 was considered as the model donor chromophore in the bichromophore systems. The TD-DFT calculation of structure 2 revealed, in reasonably good agreement with the experimentally determined longest wavelength absorption band at 316 nm (for detailed calculated transitions, see Table S1), a lowest energy transition at 347 nm for the S₁ Franck-Condon absorption, which is represented to 99% as a HOMO → LUMO transition with considerable charge transfer character from the anisyl and phenyl moieties of the pyrrole to the p-cyanophenyl acceptor (Figure 4). The excitation from the vibrationally excited ground state S0* to the relaxed first excited state S₁ translates to the process of fluorescence. The involved HOMO → LUMO transition (99%, f = 0.2298, λ_calc = 443 nm, λ_{max, exp} = 446 nm) almost exactly reverts the electron coefficient density distribution of the absorption.

For further discussion only the absorption characteristics of the bichromophores 8-12 were considered (for Jablonski diagrams of the dominant absorption bands, see Figures S12–S16). The calculations support the additive nature of the absorption bands of the constituting chromophores, i.e., all chromophore-chromophore interactions occur in the excited state. As observed experimentally, energy transfer as an envisioned interaction requires sufficient overlap of the emission band of the pyrrole and the absorption band of the corresponding ligated second chromophore. If a dual emission is intended as observed for bichromophore 11, for creating dual emission color mixing, it becomes obvious that partial energy transfer is more favorable than complete Förster resonance energy transfer.

In a screw-cap Schlenk vessel with a magnetic stir bar were placed dry THF (10mL), p-bromo benzonitrile (910mg, 5.00 mmol), 1-phenylprop-2-yn-1-ol (660 mg, 5.00 mmol), PdCl₂(PPh₃)₂ (70mg, 0.10 mmol), CuI (40mg, 0.20 mmol), and triethylamine (1.7mL, 12 mmol) under nitrogen and the mixture was stirred at 90°C (oil bath) for 10 h. Then, after cooling to room temp, 4-(tetrahydro-2H-pyran-2-yloxy)benzaldehyde (1.10 g, 5.25mmol), 3,4-dimethyl-5-(2-hydroxyethyl)-thiazolium iodide (385mg, 1.35 mmol), and triethylamine (2.5mL, 18mmol) were added and the mixture was stirred at 80°C for 4 d. After cooling to room temp acetic acid (10mL) and benzyl amine (2.6 g, 25mmol) were added and the reaction mixture was stirred at 80°C for 3 d. After cooling to room temp a saturated aqueous sodium carbonate solution was added, the phases were separated and the aqueous phase was extracted with diethyl ether (3 × 50mL). The combined organic layers were dried (anhydrous sodium sulfate), filtered and the filtrate was adsorbed on celite^®; and purified by flash chromatography on silica gel (hexane/ethyl acetate 20:1) to give pyrrole 1 as a colorless solid (554mg, 26%), Mp 202°C.

¹H NMR (500 MHz, CDCl₃): δ 5.15 (s, 2 H), 6.7 (m, 3 H), 6.82 (d, ³J = 8.6 Hz, 2 H), 7.07 (d, ³J = 8.6 Hz, 2 H), 7.12 (m, 2 H), 7.29 (m, 1 H), 7.36 (m, 3 H), 7.41 (m, 2 H), 7.46 (m, 2 H), 7.52 (m, 2 H), 8.61 (s, 1 H). ¹³C NMR (126 MHz, CDCl₃): δ 49.4 (CH₂), 108.7 (C_quat), 109.8 (CH), 116.5 (C_quat), 116.6 (CH), 119.8 (C_quat), 121.9 (C_quat), 124.3 (C_quat), 126.7 (CH), 127.7 (CH), 128.2 (CH), 128.3 (CH), 129.1 (CH), 129.4 (CH), 129.7 (CH), 132.7 (CH), 133.3 (CH), 134.2 (C_quat), 134.9 (C_quat), 136.6 (C_quat), 139.9 (C_quat), 142.5 (C_quat). MALDI-TOF [m/z (%)]: 426.0 ([M]⁺, 100). IR (KBr): ν~ [cm⁻¹] = 3,342 (w), 2,229 (m), 1,601 (s), 1,516 (m), 1,450 (m), 1,402 (w), 1,338 (m), 1,267 (s), 1,228 (m), 1,172 (m), 1,072 (w), 939 (w), 839 (s). Anal. calcd. for C₃₀H₂₂N₂O + H₂O (426.5 + 18.01): C 81.06, H 5.44, N 6.30; Found: C 81.01, H 5.01, N 6.14. HRMS (ESI) calcd. for [C₃₀H₂₂N₂O – H]⁺: 425.16484; Found: 425.16485.

Variation A: Pyrrole 1 (1.00 equiv), halide 2–4 or methyl iodide (1.00 equiv), K₂CO₃ (2.00 equivs) and dry DMF (5 mL) were placed in a screw-cap Schlenk vessel with a magnetic stir bar under nitrogen (for details, see Table 2). The reaction mixture was heated at 100°C for 5 h. After cooling to room temp celite^®; was added to the reaction mixture and the solvents were removed in vacuo. The residue was purified by flash chromatography on silica gel (hexane/ethyl acetate) to give the pyrrole reference chromophore 2 or bichromophores 8–10.

Variation B: Pyrrole 1 (1.00 equiv), KI (1.00 equiv), and dry DMF (6 mL) were placed in a screw-cap Schlenk vessel with a magnetic stir bar under nitrogen. The mixture was cooled to 0°C (ice/water bath) and sodium hydride (2.00 equivs) was added. Then, halide 5 or 6 (1.00 equiv) was added and the mixture was allowed to come to room temp. Then, the mixture was stirred at 90°C for 3.5 h (for details, see Table 2). After cooling to room temp water (2 mL) was carefully added and the mixture was extracted with ethyl acetate (3 × 25 mL). The combined organic phases were dried (anhydrous sodium sulfate) and the solvents were removed in vacuo. The residue was purified by flash chromatography on silica gel (hexane/ethyl acetate 4:1) to give the bichromophores 11 or 12.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The handling editor declared a past co-authorship with one of the authors TM.