At the beginning of our studies, we investigated the synthesis of mono- and dipyrrole-substituted corroles via the condensation reaction of pyrrole-2-carboxaldehyde with 5-phenyldipyrromethane (Scheme 1a) and the reaction between tris(2-pyrrolyl)methane with benzaldehyde (Scheme 1b). Although we tried many reaction conditions and catalysts, unidentified product mixtures were obtained instead of corrole products in both reactions.

Scheme 1: Synthetic studies to obtain mono- and dipyrrole-substituted compounds.

Then, the reaction of 1,9-diformyl-5-phenyldipyrromethane (1a) in an excess amount of pyrrole was tested to obtain pyrrole-substituted metal-free corrole through the oxidation of the bilane intermediate by using DDQ (Scheme 2). Pyrrole was used as both reagent and solvent in these reactions. The desired product was not observed in the reaction medium when various catalysts (TFA, I2, AlCl3, InCl3, FeCl3, H2SO4, p-TsOH, Mont. KSF, Mont. K-10, and AgOTf) were used at different temperatures (Supporting Information File 1, Table S1). However, the copper complex of the desired product 2a was obtained in 5% yield in the presence of Cu(OTf)2 catalyst at room temperature.

Scheme 2: The reaction of 5-phenyl-1,9-diformyldipyrromethane (1a) with pyrrole.

When the synthetic methods in the literature are examined to obtain corrole compounds, it is observed that temperature, pyrrole ratio, reaction time, catalyst type, and oxidant are important parameters on the yields of the reactions [28-32]. For this reason, optimization studies were carried out on these parameters. Based on the results of the preliminary studies, optimization studies were carried out in the presence of 10 mol % Cu(OTf)2 catalyst, and the effect of temperature on the synthesis of pyrrole-substituted trans-A2B corrole compounds was investigated in 40 equivalents of pyrrole using a reaction time of 2 hours (Table 1, entries 1–4). No product was formed as a result of increasing the reaction temperature to 40 °C. It was observed that the yield increased gradually when the reaction temperature was decreased. The yield of the product, which was obtained with 5% efficiency at room conditions and 6% at 0 °C, increased to 9% by reducing the temperature to −20 °C. The pyrrole/1a ratio played little role in improving the yield of the product. The yield of 2a decreased to 4% yield when the reaction was carried out in 20 equivalents of pyrrole (Table 1, entry 5). Increasing the amount of pyrrole above 40 equivalents did not affect the reaction yield (Table 1, entries 6 and 7). Then, the effect of the reaction time before the oxidant addition on the yield of product was investigated at −20 °C in 40 equivalents of pyrrole. When the reaction time was 1 hour, the yield decreased to 4% (Table 1, entry 8). If the reaction time exceeded 2 hours, unexpectedly no desired product was found at all (Table 1, entries 9 and 10). This situation can be explained by the instability of the bilane intermediate formed in the reaction medium and its decomposition during long reaction times. It was also investigated whether the product yield would increase with the amount of catalyst since only the copper complex of the expected product could be isolated at the end of the reaction. The reaction was repeated using 20 mol %, 50 mol %, and equimolar amounts of copper triflate under previously optimized conditions. In the case of using 20 mol % copper triflate, the reaction efficiency increased to 12%, while a further increase in the amount of catalyst did not affect the yield (Table 1, entries 11–13). In order to determine the effect of the oxidant type and the oxidant amount, reactions were carried out with 3 and 4 equivalents of DDQ and p-chloranil. While more than 2 equivalents of DDQ did not have a positive effect on the reaction yield (Table 1, entries 14 and 15), p-chloranil formed a product with a lower yield than DDQ (Table 1, entries 16–18). The activities of different copper catalysts were also tested in the model reaction. Only CuCl2 formed the product in 5% yield and the other salts did not catalyze the reaction (Table 1, entries 19–22).

Table 1: Optimization of reaction conditions.a

Entry Catalyst Catalyst amount (%) Temp (°C) Pyrrole/1a Time (h) Oxidant
(oxidant/1a) Yield (%)b 1 Cu(OTf)2 10 40 40 2 DDQ (2) – 2 Cu(OTf)2 10 rt 40 2 DDQ (2) 5 3 Cu(OTf)2 10 0 40 2 DDQ (2) 6 4 Cu(OTf)2 10 −20 40 2 DDQ (2) 9 5 Cu(OTf)2 10 −20 20 2 DDQ (2) 4 6 Cu(OTf)2 10 −20 60 2 DDQ (2) 9 7 Cu(OTf)2 10 −20 80 2 DDQ (2) 9 8 Cu(OTf)2 10 −20 40 1 DDQ (2) 4 9 Cu(OTf)2 10 −20 40 4 DDQ (2) – 10 Cu(OTf)2 10 −20 40 6 DDQ (2) – 11 Cu(OTf)2 20 −20 40 2 DDQ (2) 12 12 Cu(OTf)2 50 −20 40 2 DDQ (2) 12 13 Cu(OTf)2 100 −20 40 2 DDQ (2) 12 14 Cu(OTf)2 20 −20 40 2 DDQ (3) 12 15 Cu(OTf)2 20 −20 40 2 DDQ (4) 12 16 Cu(OTf)2 20 −20 40 2 p-chloranil (2) 10 17 Cu(OTf)2 20 −20 40 2 p-chloranil (3) 10 18 Cu(OTf)2 20 −20 40 2 p-chloranil (4) 10 19 CuCl2 100 −20 40 2 DDQ (2) 5 20 CuCl 100 −20 40 2 DDQ (2) – 21 Cu(OAc)2 100 −20 40 2 DDQ (2) – 22 Cu(NO3)2 100 −20 40 2 DDQ (2) –

aReaction conditions: 1a (0.36 mmol, 0.10 g), pyrrole (14.4 mmol, 0.97 g, 1 mL), CHCl3 (2 mL). bIsolated yield.

With the best conditions in our hands (Table 1, entry 11), different diformylated dipyrromethanes were subjected to condensation reactions. Electron-withdrawing 4-chlorophenyl, pentafluorophenyl, and 4-nitrophenyl-substituted corrole compounds were isolated in 13% yields (Table 2, entries 2–4). While electron-donating 4-methoxyphenyl (2e) and p-tolyl-substituted corrole (2f) were isolated in 8% yield and 12% yields respectively, a p-bromophenyl substituent resulted in a mixture of undefined products after the reaction. This might be due to scrambling, which is an acid-catalyzed rearrangement of the substituent in intermediates of the condensation reaction.

Table 2: Synthesis of pyrrole-substituted corroles.a

Entry R 2 Yield (%)b 1 C6H5 2a 12 2 4-ClC6H4 2b 13 3 C6F5 2c 13 4 4-NO2C6H4 2d 13 5 4-CH3OC6H4 2e 8 6 4-CH3C6H4 2f 12 7 4-BrC6H4 2g –

aReaction conditions: 1a–g (0.36 mmol), pyrrole (14.4 mmol, 0.97 g, 1 mL), Cu(OTf)2 (0.072 mmol, 0.026 g), DDQ (0.72 mmol, 0.16 g ), CHCl3 (2 mL). bIsolated yield.

The structures of the meso-pyrrole substituted corroles were identified by using 1H NMR, 1H,1H-COSY NMR and HRMS techniques (see Supporting Information File 1). The 1H NMR spectrum of 2a is shown in Figure 1. As expected, pyrrole C4, C3, C5 and NH protons appeared at 6.51, 7.05, 7.27 and 8.90 ppm, respectively. Coupling of all pyrrole protons can be seen in the 1H,1H-COSY NMR spectrum (Supporting Information File 1, Figure S3). The β-protons of the corrole macrocycle and the phenyl group gave signals between 7.40–8.20 ppm.

Figure 1: 1H NMR spectrum of 2a in THF-d8.

Electronic absorption spectra of corroles 2a–g were recorded in CHCl3 at 2.0 × 10−5 M. The Soret bands of all compounds are located between 410–420 nm (see Supporting Information File 1). The Q bands of the compounds are seen as broad absorptions in the 500–700 nm region. Figure 2 shows the absorption spectrum of 2a with a strong Soret band at 412 nm and weak Q-bands at 542 and 611 nm.

Figure 2: Electronic absorption spectrum of 2a in CHCl3.

After the synthesis of corrole compounds, we tried to extent our studies to obtain meso-pyrrole-substituted porphyrin compounds. For this purpose, the MacDonald [2 + 2] porphyrin macrocyclization reaction of 1,9-diformyl-5-phenyl dipyrromethane (1a) with tris(2-pyrrolyl)methane was investigated by changing the reaction conditions in the presence of various acids such as acetic acid, hydrochloric acid and p-toluenesulfonic acid (Scheme 3). Among these reactions, the reaction in acetic acid resulted in 5,15-diphenylporphyrin (3) and 5-phenylporphyrin (4) in 4% and 1% yields, respectively. No pyrrole-substituted porphyrin product was detected. The structures of compounds 3 [33] and 4 [34] are in agreement with the literature data.

Scheme 3: [2 + 2] Mac Donald type condensation reaction.