Dihydrooxazine Byproduct of a McMurry–Melton Reaction en Route to a Synthetic Bacteriochlorin

Abstract

A synthetic route to gem-dimethyl-substituted bacteriochlorins—models of native bacteriochlorophylls—relies on the formation of a dihydrodipyrrin precursor via a series of established reactions: van Leusen pyrrole formation, Vilsmeier formylation, Henry reaction, borohydride reduction, Michael addition, and McMurry–Melton pyrroline formation. The latter is the least known of the series. Here, the McMurry–Melton reaction of a 2-(6-oxo-2-nitrohexyl)pyrrole in the presence of TiCl₃ and an ammonium acetate buffer formed the expected Δ¹-pyrroline, as well as an unexpected polar, cyclic byproduct (a 5,6-dihydro-4H-1,2-oxazin-6-ol), each attached to the 2-methylpyrrole unit. Both species were characterized by single-crystal X-ray diffraction. The McMurry–Melton reaction is a type of intercepted Nef reaction (the transformation of a nitroalkyl motif into a carbonyl group), where both the Δ¹-pyrroline and the dihydrooxazine derive from the reaction of the nitrogen derived from the nitro group upon complete or partial reductive deoxygenation, respectively, with the γ-keto group. The report also considers competing Nef and McMurry–Melton reactions, the nature of available TiCl₃ reagents, and the use of ammonium acetate for buffering the TiCl₃/HCl reagent.

1. Introduction

Bacteriochlorophyll a constitutes nature’s chief light absorber in the near-infrared (NIR) region and underpins anoxygenic bacterial photosynthesis (Chart 1) [1]. The ability to harvest NIR light offers many research and practical applications spanning energy sciences, photonics, and medicine. The core chromophore of bacteriochlorophyll a is a bacteriochlorin. To gain access to synthetically malleable bacteriochlorins, we have been working to develop routes to gem-dimethyl-substituted macrocycles that contain the bacteriochlorin chromophore. The rationale for the use of gem-dimethyl substituents is twofold: (1) enforce saturation at the β-positions of the pyrrolic rings; and (2) achieve more direct synthesis without a requirement for the control of stereochemical configuration, as would be the case for the trans-dialkyl substituted native bacteriochlorophyll a.

The synthesis of bacteriochlorins can be achieved by self-condensation of a dihydrodipyrrin-acetal (5a), as illustrated for bacteriochlorin BC-1 (Chart 1). During the course of the synthesis of BC-1, a polar byproduct was isolated in the final step of dihydrodipyrrin formation. The byproduct arises during the McMurry–Melton reaction wherein a Δ¹-pyrroline is formed by TiCl₃-mediated reductive deoxygenation of a 1-nitro-4-oxo-substituted alkane. In this paper, we first report the synthesis of BC-1 and then pivot to focus on the McMurry–Melton reaction byproduct.

2. Results and Discussion

A. Bacteriochlorin synthesis. The synthesis of bacteriochlorin BC-1 follows the established pathway [2,3] shown in Scheme 1. The van Leusen reaction [4,5] of methyl cinnamate with toluenesulfonylmethyl isocyanide (TosMIC) afforded the corresponding pyrrole as the crude product in estimated yield of 62%. The subsequent Vilsmeier formylation [6] gave aldehyde 1a in 64% yield, accompanied by the positional isomer 1b in 6.4% yield. Aldehyde 1a was subjected to the Henry reaction [7] with nitromethane to give the nitrovinyl product, which upon reduction with NaBH₄ gave the 2-(2-nitroethyl)pyrrole 2 in 50% yield. The Michael addition [8] of 2 with the α,β-unsaturated ketone 3 [9] under solventless conditions [3] in the presence of DBU provided Michael adduct 4 in 68% yield. The McMurry–Melton reaction was carried out by treatment of 4 with NaOMe followed by TiCl₃ in a buffer of saturated aqueous ammonium acetate. The expected dihydrodipyrrin 5a was obtained in 41% yield, accompanied by the unexpected and more polar byproduct 5b, a 5,6-dihydro-4H-1,2-oxazine, in 23% yield. The difference in polarity of 5a and 5b is substantial, with R_f values on silica TLC of 0.5 and 0.1, respectively. The Eastern–Western self-condensation [2,3] of dihydrodipyrrin 5a in the presence of BF₃·O(Et)₂ in CH₂Cl₂ gave 5-unsubstituted bacteriochlorin BC-1 in 37% yield and the 5-methoxybacteriochlorin MeOBC-1 in 6.1% yield.

The absorption spectra of the two bacteriochlorins are displayed in Figure 1. The spectra are typical of bacteriochlorins [10], with strong transitions in the NIR and near-ultraviolet region and a weak transition in the visible region. For BC-1, the long-wavelength (Q_y) band is at 767 nm, whereas that for MeOBC-1 is at 748 nm, consistent with the known hypsochromic effect [2] of the 5-methoxy group.

B. The dihydrooxazine product of the McMurry–Melton reaction. The structure of 5b was first thought to be an acyclic oxime. Single-crystal X-ray determination, however, revealed the presence of the 5,6-dihydro-4H-1,2-oxazine motif (Figure 2). In 5b, the pyrrole and dihydrooxazine rings exhibit a butterfly arrangement, tilted by ~60° with respect to each other due to the intervening methylene unit. Slight disorder is observed for the phenyl substituent. The distance between the pyrrolic NH and the dihydrooxazine N (2.955 Å) is consistent with moderate intramolecular hydrogen bonding. The dihydrooxazine presents as a somewhat flattened chair structure due to the sp²-hybridized imine carbon, where the C=N and O–C bonds exhibit a dihedral angle of 22.96° (versus 60° in chair cyclohexane). The hydroxy group of the hemi-ketal is positioned axial, whereas the dimethoxymethyl group is equatorial; the H···O distance between the OH group and the two methoxy oxygen atoms is 2.539 and 3.103 Å. In short, the structure of the dihydrooxazine unit somewhat resembles that of fructopyranose, which is also a cyclic hemiketal. The crystal structure of 5b contains both R and S stereoisomers at the C3 position, confirming that 5b exists as a pair of enantiomers.

The structure of dihydrodipyrrin 5a is ordinary, containing nearly coplanar pyrrole and pyrroline rings, a partially rotated phenyl group with respect to the pyrrole plane (52.7°), and the ester carbonyl unit projected away from the adjacent phenyl moiety.

Although the crystal structure of 5b was unambiguous, we wanted to be able to rely on NMR spectra going forward to examine crude dihydrodipyrrin-forming reaction mixtures for the presence of analogous pyrrole–dihydrooxazines. The distinct structures of the precursor 4 and products 5a and 5b are readily seen in the ¹H NMR spectra (Figure 3). There are four key differences between compounds 4 and 5b. (1) An apparent (but partially unresolved) doublet of doublets is observed for the proton (labeled “d”) attached to the stereogenic carbon bearing the nitro group of compound 4, a feature that is missing in 5b due to the C=N bond of the dihydrooxazine. (2) An intense singlet for the two methoxy groups of the acetal is observed for 4, whereas the two methoxy groups in 5b resonate with Δδ of ~0.15 ppm, consistent with the distinct environments (where one but not the other is likely hydrogen-bonded with the hydroxyl group) observed upon X-ray structural analysis. (3) The methylene protons (“a”) adjacent to the pyrrole in 4 are diastereotopic and resonate as an apparent doublet of doublets at 3.19 ppm (largely resolved) and 2.94 ppm (unresolved), whereas in 5b the “a” protons again are diastereotopic but resonate downfield (Δδ of ~0.34 and ~0.55 ppm) as a partially obscured doublet. (4) The methylene protons flanked by the gem-dimethyl group and the ketone (“b”) in 4 resonate as a multiplet centered at ~2.5 ppm, versus ~1.77 ppm in 5b due to the flanking hemiketal motif rather than the ketone. Both 4 and 5b are racemic.

The ¹H NMR spectrum of dihydrodipyrrin 5a is simpler, with singlets due to the methine proton “a” and the methylene protons “b”. In all three compounds, the chemical shift of the dimethyl acetal methine proton (4.2 in 4, 4.0 in 5b, and 5.2 ppm in 5a) reflects differences in structure at the adjacent carbon. On the other hand, the resonance of the protons of the carbomethoxy group are nearly constant across the three compounds. The gem-dimethyl groups also give distinct resonances (not shown) for the three compounds, including 4 (0.98, 1.00 ppm), 5b (0.89, 1.07 ppm), and 5a (1.14 ppm). The observed distinctions in ¹H NMR spectra should provide a convenient diagnostic for the future assessment of crude samples from dihydrodipyrrin-forming reactions.

3. Discussion

A key objective over the years has been to refine reaction pathways so as to achieve streamlined implementation and increased yields of bacteriochlorins, for which BC-1 and MeOBC-1 are prototypical. Here, the identification of an unexpected byproduct in the McMurry–Melton reaction, leading to the dihydrodipyrrin, prompts focus on this transformation. In this section, we describe the original report by McMurry and Melton, then overview extension of the reaction to form dihydrodipyrrins with emphasis on observations that point to the complexities of this ostensibly simple transformation. A goal of the following discussion is to develop a framework in support of future, more extensive studies.

The treatment of a nitroalkane with a base followed by aqueous acid constitutes the Nef reaction [11,12,13,14,15], first reported in 1894 (Scheme 2). The base causes formation of the nitronate, whereas the aqueous acid causes hydrolysis of the nitronate [14], forming the corresponding ketone or aldehyde. McMurry and Melton (1973) reported that a nitroalkane gave the ketone directly upon treatment with TiCl₃ in aqueous acid (pH < 1) at room temperature [16]. This modification of the Nef reaction has been referred to previously as the McMurry method [14,15] or the McMurry–Melton method [17], to be distinguished from the better-known McMurry reaction [18] (low-valent titanium-induced deoxygenative dimerization of ketones to form an alkene), which was reported one year later [19].

In an effort to broaden the scope, McMurry and Melton also examined reactions at more neutral pH: the nitroalkane was first treated with one equivalent of NaOMe in methanol followed by an aqueous solution of TiCl₃ buffered at pH 6 (with ammonium acetate) [16]. They stated [16], “Much to our surprise, however, several nitro compounds gave worse results under these neutral conditions than under acidic conditions. For example, nitro ketone [I] gave none of the expected diketone [heptane-2,5-dione, II] but gave rather, as the only isolable organic compound, the pyrroline [III] (20%).” The inadvertent formation of a Δ¹-pyrroline (III, albeit without reported characterization data; Scheme 2) [16], was extended by Battersby and coworkers (1981) for the conversion of a 6-oxo-2-nitrohexylpyrrole (IV) to a tetrahydrodipyrrin (V) for use in the synthesis of an isobacteriochlorin [20]. Battersby and coworkers (1988) also employed the McMurry–Melton reaction to convert a 6-oxo-2-nitrohexylpyrrole (VI) to a dihydrodipyrrin (VII) for use in the synthesis of a chlorin (Scheme 2) [21]. We followed in the same manner to synthesize diverse dihydrodipyrrins, with the conversion of 4 to 5a being illustrative. The same route has been exploited to prepare Δ¹-pyrrolines unrelated to tetrapyrrole macrocycles [22].

Although the McMurry–Melton reaction constitutes a cornerstone of the synthesis of chlorins and bacteriochlorins [2,3,9,23,24,25,26,27,28,29,30,31,32], anomalies in this domain are known. During the synthesis of ¹⁵N isotopologues of bacteriochlorins, we observed partial loss of the ¹⁵N label during the McMurry–Melton reaction upon reaction of VIII-[¹⁵N]-NO₂ [28] (Scheme 3). The dihydrodipyrrin mixture was comprised of intact ¹⁵N-dihydrodipyrrin isotopologue IX-[¹⁵N] and the natural abundance dihydrodipyrrin IX-NA. The latter was proposed to originate from competing Nef and McMurry–Melton reactions. Thus, the Nef reaction of nitronate VIII-[¹⁵N]-nitronate yields the diketone X-keto along with the loss of a ¹⁵N species. Condensation with ammonia (from the ammonium acetate buffer) affords the imine or enamine X-enamine, which then undergoes intramolecular condensation in a Paal–Knorr-like process with the second carbonyl group to form IX-NA.

The dihydrooxazine 5b contains an oxime motif. Reduced nitro species, including nitroso and oxime groups, have been proposed as reaction intermediates by McMurry and Melton [16], and the oxime is considered an intermediate (if not a product) of the Nef reaction [13,14,15]. 5,6-Dihydro-4H-1,2-oxazines have been prepared previously [33,34,35], typically by reduction of the corresponding and well-studied 1,2-oxazine-N-oxides (for which multiple synthetic entry points are known [36]). To our knowledge, 5,6-dihydro-4H-1,2-oxazines have heretofore not been identified in a McMurry–Melton reaction. The mechanistic question of whether a dihydrooxazine N-oxide is formed here and then reduced, or the reduction occurs prior to cyclization to form the dihydrooxazine directly, requires additional study. Extensive work remains to understand the role of various reaction parameters on the McMurry–Melton reaction course and the yields of the desired Δ¹-pyrroline versus the dihydrooxazine byproduct. Further discussion of the various McMurry–Melton reaction conditions is provided in Appendix A.

4. Conclusions

The isolation of dihydrooxazine 5b sheds light on the little-studied McMurry–Melton reaction. An unknown is whether analogous dihydrooxazine byproducts are generally formed in McMurry–Melton reactions and have heretofore not been identified given far greater polarity compared with the Δ¹-pyrroline. The use of the ¹H NMR spectral comparisons outlined in Figure 3 should facilitate the examination of crude reaction mixtures and thereby afford a broader understanding of the McMurry–Melton reaction.

5. Experimental Section

General. Commercial compounds were used as received. All solvents (anhydrous or reagent-grade) were used as received unless noted otherwise. THF was freshly distilled from sodium/benzophenone. Silica (40 μm average particle size) was used for column chromatography. Electrospray ionization mass spectrometry (ESI-MS) data generally enable accurate mass measurements, were obtained in the positive ion mode (unless noted otherwise) and are reported for the protonated molecular ion. ¹H NMR and proton-decoupled ¹³C NMR spectra were collected at room temperature in CDCl₃ unless noted otherwise and using instruments as indicated for each compound.

4-Methoxycarbonyl-3-phenylpyrrole-2-carboxaldehyde (1a). A solution of methyl cinnamate (5.00 g, 30.8 mmol) and TosMIC (7.22 g, 37.0 mmol) in dry DMF (100 mL) was treated with NaOH (1.85 g, 46.3 mmol), and stirring was continued under argon for 20 h at room temperature. The reaction mixture was quenched by the addition of saturated aqueous NH₄Cl (200 mL), and then stirred for 30 min at room temperature. After H₂O (100 mL) was added, the mixture was extracted with ethyl acetate (100 mL × 3). The combined organic extract was washed with H₂O (100 mL) and brine (100 mL), dried over Na₂SO₄, and concentrated to dryness in vacuo. The residue was suspended in hexanes/CH₂Cl₂ (50 mL, 9:1). The suspension was sonicated and filtered, and the precipitate was rinsed with hexanes/CH₂Cl₂ (50 mL 9:1 × 3) to give 3-methoxycarbonyl-4-phenylpyrrole (3.83 g, 62%) as a brown solid. The Vilsmeier reagent was prepared by treatment of dry DMF (16 mL) with POCl₃ (4.01 g, 26.2 mmol) at room temperature and stirring of the resulting mixture for 1 h. In a separate flask, a solution of crude 3-methoxycarbonyl-4-phenylpyrrole (3.83 g, 19.0 mmol) in dry DMF (162 mL) was treated with the freshly prepared Vilsmeier reagent at room temperature and stirring was continued for 3 h under argon. The reaction mixture was quenched by the addition of saturated aqueous NaOAc (100 mL), and then stirred for 30 min at room temperature. After H₂O (200 mL) was added, the mixture was extracted with ethyl acetate (100 mL × 3). The combined organic extract was washed with H₂O (300 mL × 2) and brine (100 mL), dried over Na₂SO₄, and concentrated to dryness in vacuo. Purification by column chromatography [silica gel, 100 g; hexanes/ethyl acetate (4:1)] afforded the title compound (2.81 g, 64%, R_f = 0.25 (hexanes/ethyl acetate = 2:1)) followed by the positional isomer 1b (0.281 mg, 6.4%, R_f = 0.30 (hexanes/ethyl acetate = 2:1)).

Data for 1a: white solid; mp 184–186 °C; ¹H NMR (500 MHz, CDCl₃) δ 3.74 (3H, s), 7.41–7.47 (5H, m), 7.72 (1H, dd, J = 3.5 Hz, J = 1 Hz), 9.38 (1H, d, J = 1 Hz), 9.79 (1H, brs); ¹³C-{¹H} NMR (175 MHz, CDCl₃) δ 51.4, 116.5, 128.0, 128.4, 130.0, 130.89, 130.90, 131.1, 137.3, 163.9, 181.0. HRMS (ESI) for C₁₃H₁₁NO₃ (M + H⁺), calcd 230.0812, found 230.0811.

Data for 1b: white solid; mp 118–121 °C; ¹H NMR (500 MHz, CDCl₃) δ 3.80 (3H, s), 7.07 (1H, dd, J = 3 Hz, J = 1 Hz), 7.32–7.35 (1H, m), 7.36–7.42 (1H, m), 10.13 (1H, brs), 10.17 (1H, d, J = 1 Hz); ¹³C-{¹H} NMR (175 MHz, CDCl₃) δ 51.8, 119.5, 1234.0, 127.5, 128.1, 129.4, 130.0, 133.6, 133.7, 164.3, 182.3. HRMS (ESI) for C₁₃H₁₁NO₃ (M + H⁺), calcd 230.08117, found 230.08069.

4-Methoxycarbonyl-2-(2-nitroethyl)-3-phenylpyrrole (2). A solution of aldehyde 1a (1.03 g, 4.49 mmol) in EtOH (45 mL) was treated with potassium acetate (0.610 g, 8.99 mmol) and methylamine hydrochloride (0.441 g, 4.49 mmol), and stirring was continued under argon for 40 min at room temperature. The solution was treated with nitromethane (1.00 mL, 18.6 mmol). After stirring for 24 h at room temperature, saturated aqueous NH₄Cl (50 mL) was added. Stirring was continued for 30 min at room temperature, then H₂O (200 mL) was added. The mixture was extracted with ethyl acetate (100 mL × 3). The combined organic extract was washed with H₂O (50 mL) and brine (50 mL), dried over Na₂SO₄, and concentrated to dryness in vacuo. The residue (1.37 g) was dissolved in MeOH (50 mL), and the solution was cooled at 0 °C. The solution was treated portionwise with NaBH₄ (205 mg, 5.42 mmol). The reaction mixture was stirred for 1 h at 0 °C, and then quenched by the addition of saturated aqueous NH₄Cl (100 mL). After H₂O was added, the mixture was extracted with ethyl acetate (100 mL × 3). The combined organic extract was washed with H₂O (50 mL) and brine (50 mL), dried over Na₂SO₄, and concentrated to dryness in vacuo. Purification by column chromatography [silica gel 50 g; hexanes/ethyl acetate (2:1)] afforded a colorless film (0.61 g, 50% in 2 steps). ¹H NMR (500 MHz, CDCl₃) δ 3.20 (2H, t, J = 6.5 Hz), 3.67 (3H, s), 4.42 (2H, t, J = 6.5 Hz), 7.24–7.27 (2H, m), 7.29–7.34 (1H, m), 7.37–7.40 (3H, m), 8.61 (1H, brs); ¹³C-{¹H} NMR (125 MHz, CDCl₃) δ 23.5, 51.0, 75.2, 115.0, 124.0, 124.6, 125.3, 127.2, 128.1, 130.3, 134.3, 164.9. HRMS (ESI) for C₁₄H₁₄N₂O₄ (M + H⁺), calcd 275.1026, found 275.1023.

4-Methoxycarbonyl-2-(6,6-dimethoxy-5-oxo-3,3-dimethyl-2-nitrohexyl)-3-phenylpyrrole (4). A solution of pyrrole 2 (610. mg, 2.22 mmol) and Michael acceptor 3 [9] (520. mg, 3.29 mmol) in DBU (25 mL) was stirred for 22 h at room temperature under argon. Saturated aqueous NH₄Cl (100 mL) was added, and the mixture was extracted with ethyl acetate (20 mL × 3). The combined organic extract was washed with H₂O (10 mL) and brine (10 mL), dried over Na₂SO₄, and concentrated to dryness in vacuo. Purification by column chromatography [silica gel 25 g; hexanes/ethyl acetate (2:1)] afforded a pale-yellow film (652.1 mg, 68%). ¹H NMR (600 MHz, CDCl₃) δ 0.98 (3H, s), 1.00 (3H, s), 2.42 (1H, d, J = 18.6 Hz), 2.60 (1H, d, J = 18.6 Hz), 2.94 (1H, dd, estimated J~15.6 Hz, J~2.4 Hz), 3.19 (1H, dd, J = 15.6 Hz, J = 12 Hz), 3.37 (6H, s), 3.63 (3H, s), 4.26 (1H, s), 4.98 (1H, dd, estimated J~12 Hz, J~2.4 Hz), 7.23–7.25 (2H, m), 7.28–7.30 (1H, m), 7.32 (1H, d, J = 3 Hz), 7.35–7.38 (2H, m), 8.43 (1H, brs); ¹³C-{¹H} NMR (150 MHz, CDCl₃) δ 23.9, 24.1, 24.9, 36.6, 44.8, 50.9, 55.20, 55.24, 95.0, 104.7, 114.9, 124.0, 124.8, 125.3, 127.0, 128.0, 130.4, 134.3, 164.9, 203.4. HRMS (ESI) for C₂₂H₂₈N₂O₇ (M + H⁺), calcd 433.1969, found 433.1965.

8-Methoxycarbonyl-1-(1,1-dimethoxymethyl)-3,3-dimethyl-7-phenyl-2,3-dihydrodipyrrin (5a). In a first flask, a solution of the Michael adduct 4 (498 mg, 1.15 mmol) in distilled THF (25 mL) was treated with NaOMe (88.4 mg, 1.64 mmol), and stirring was continued under argon for 30 min at room temperature. In a second flask, TiCl₃ (10.0 mL, 12 wt % TiCl₃ in aqueous HCl solution, 15.6 mmol TiCl₃) was treated with saturated aqueous NH₄OAc solution until a pH of 6 was obtained (pH paper; ultimately 60 mL saturated aqueous NH₄OAc solution was added), and stirring of the resulting cloudy, heterogeneous mixture was continued for 10 min at room temperature under argon. (Note that saturated aqueous NH₄OAc solution is reported to contain 148 g NH₄OAc/100 mL at 4 °C [37]). The contents of the first flask were added to the second flask. The reaction mixture under argon was stirred for 24 h at room temperature. After saturated aqueous NH₄Cl (100 mL) was added, the mixture was extracted with ethyl acetate (30 mL × 3). The combined organic extract was washed with H₂O (20 mL) and brine (20 mL), dried over Na₂SO₄, and concentrated to dryness in vacuo. Purification by column chromatography [silica gel 25 g; hexanes/ethyl acetate (2:1)] afforded the title compound (179 mg, 41%, R_f = 0.50 (hexanes/ethyl acetate = 1:1)) and byproduct 5b (111 mg, 23%, R_f = 0.10 (hexanes/ethyl acetate = 1:1)), each as a pale-yellow film.

Data for 5a: ¹H NMR (700 MHz, CDCl₃) δ 1.14 (6H, s), 2.62 (2H, s), 3.47 (6H, s), 3.69 (3H, s), 5.05 (1H, s), 5.78 (1H, s), 7.31–7.33 (1H, m), 7.37–7.41 (4H, m), 7.53 (1H, d, J = 3.5 Hz), 11.22 (1H, brs); ¹³C-{¹H} NMR (175 MHz, CDCl₃) δ 29.1, 40.5, 48.5, 50.9, 54.7, 102.7, 105.3, 114.1, 124.7, 125.5, 126.6, 127.7, 129.7, 131.0, 134.6, 161.6, 165.3, 175.6. HRMS (ESI) for C₂₂H₂₆N₂O₄ (M + H⁺), calcd 383.1965, found 383.1970.

Data for the byproduct 4-methoxycarbonyl-2-(6-(dimethoxymethyl)-(6-hydroxy-4,4-dimethyl-5,6-dihydro-4H-1,2-oxazin-3-yl)methyl)-3-phenylpyrrole (5b): ¹H NMR (600 MHz, CDCl₃) δ 0.90 (3H, s), 1.07 (3H, s), 1.71 (1H, d, J = 12 Hz), 1.84 (1H, d, J = 12 Hz), 3.49 (3H, s), 3.51 (1H, d, J = 16 Hz), 3.56 (1H, d, J = 16 Hz), 3.63 (3H, s), 3.66 (3H, s), 4.18 (1H, s), 7.29–7.30 (3H, m), 7.37–7.40 (3H, m), 9.16 (1H, brs); ¹³C-{¹H} NMR (150 MHz, CDCl₃) δ 16.7, 26.5, 26.7, 28.1, 30.7, 38.8, 50.8, 55.8, 58.2, 96.8, 105.6, 114.4, 122.8, 123.8, 126.1, 126.7, 130.6, 134.8, 164.9, 165.2. HRMS (ESI) for C₂₂H₂₈N₂O₆ (M + H⁺), calcd 417.2020, found 417.2016.

3,13-Dimethoxycarbonyl-8,8,18,18-tetramethyl-2,12-diphenylbacteriochlorin (BC-1). A solution of dihydropyrrin 5a (156.4 mg, 408.9 μmol) in dry CH₂Cl₂ (25 mL) was treated with BF₃·OEt₂ (0.16 mL, 1.5 mmol), and stirring under argon was continued for 24 h at room temperature. After saturated aqueous NaHCO₃ (50 mL) was added, the mixture was extracted with CH₂Cl₂ (10 mL × 3). The combined organic extract was washed with H₂O (10 mL) and brine (10 mL), dried over Na₂SO₄, and concentrated to dryness in vacuo. Purification by column chromatography [silica gel 50 g; hexanes/ethyl acetate (4:1)] afforded the title bacteriochlorin (48.2 mg, 37%) followed by the 5-methoxybacteriochlorin MeOBC-1 (8.3 mg, 6.1%).

Data for BC-1: purple solid; ¹H NMR (500 MHz, CDCl₃) δ –1.14 (2H, s), 1.84 (12H, s), 4.05 (6H, s), 4.42 (4H, s), 7.65–7.68 (2H, m), 7.72–7.75 (4H, m), 7.93–7.95 (4H, m), 8.45 (2H, s), 9.63 (2H, s). ¹³C-{¹H} NMR (125 MHz, CDCl₃) δ 29.9, 30.9, 46.1, 52.0, 91.1, 97.5, 120.0, 128.0, 132.1, 134.3, 134.6, 135.6, 138.6, 161.3, 166.8, 172.2. HRMS (ESI) for C₄₀H₃₈N₄O₄ (M + H⁺), calcd 639.2966, found 639.2955. λ_abs (toluene) 357, 382, 526, 767 nm.

Data for 5-methoxy-3,13-dimethoxycarbonyl-8,8,18,18-tetramethyl-2,12-diphenylbacteriochlorin (MeOBC-1): purple film; ¹H NMR (700 MHz, CDCl₃) δ –1.54 (1H, s), –1.25 (1H, s), 1.81 (6H, s), 1.86 (6H, s), 4.04 (3H, s), 4.13 (3H, s), 4.26 (3H, s), 4.36 (2H, s), 4.42 (2H, s), 7.63–7.67 (2H, m), 7.71–7.75 (4H, m), 7.92–7.93 (2H, m), 8.08-8.09 (2H, m), 8.44 (1H, s), 8.58 (1H, s), 9.57 (1H, s); ¹³C-{¹H} NMR (175 MHz, CDCl₃) δ 29.9, 30.8, 30.9, 45.6, 46.2, 48.0, 51.7, 51.9, 53.0, 64.5, 96.2, 98.1, 98.2, 119.1, 124.5, 127.9, 128.19, 128.23, 128.9, 131.8, 132.2, 133.0, 133.6, 134.3, 134.4, 134.9, 135.4, 135.7, 138.7, 156.5, 161.3, 166.8, 169.2, 172.7. HRMS (ESI) for C₄₁H₄₀N₄O₅ (M + H⁺), calcd 669.3072, found 668.3000. λ_abs (toluene) 361, 373, 528, 748 nm.

X-ray crystallography Single-crystal X-ray diffraction (SCXRD) data were collected using a Bruker D8 Venture single-crystal diffractometer at 100 K via the Psi and Omega scan strategy (MoKα = 0.71073 Å). The crystal data acquisition and refinement were achieved with use of APEX4 [38], SHELX2018 [39], and OLEX2 [40] programs. Suitable crystals of 5a and 5b for SCXRD analysis were obtained upon recrystallization in CH₂Cl₂/hexane. Compounds 5a and 5b are found in monoclinic systems with the P2₁/c and P2₁ space group, respectively. The crystallographic data and molecular packing are described in the Supplementary Materials (see Table S1).

pH simulation The simulation was carried out using fsolve from SciPy [41]. The parameters employed were as follows: pK_a for NH₄⁺ = 9.25; pK_b for AcO^– = 9.25 (note that the pK_a for AcOH is 4.75, thus the pK_b = 14–4.75 = 9.25); and K_w = 10⁻¹⁴ (autoprotolysis of water). The script for the calculations is provided in the Supplementary Materials.