An investigation of the allylation cascade reactions of substituted indigos

In a continuation of the exploration of indigo cascade reactions, a series of –OMe, –Ph, –Br and –NO2 substituted indigos 1a–i were synthesised to probe electronic effects upon the outcome of allylation cascade reactions. When indigos 1a–i in the presence of base were reacted with allyl bromide, spiroindolinepyridoindolones 17–25 (36–75%) were obtained as the major products in each case, marking a shift in outcome relative to that previously reported for unsubstituted indigo. In electron-rich derivatives (–OMe, –Ph), C-allylspiroindolinepyridoindolediones 26–29 (3–11%) were also isolated, which are most likely formed via a Claisen rearrangement of the respective spiroindolinepyridoindolones 18–21. Additionally, the isolation of diallylbiindolone 16, oxazinobiindole 30 and N,N′-diallyl-3,3′-bis(allyloxy)biindole 31 each represented novel polyheterocyclic derivatives, providing intriguing new mechanistic insights, reaction pathways and in the case of 30 the first common heterocyclic skeletal outcome shared in both allylation and propargylation cascade reactions of indigo.


The Claisen rearrangement of dimethoxyspiroindolinepyridoindolone 17
A

dione 33a
All

S4.1 Structural elucidation of 3-acetoxy-5-nitroindole 10
The synthesis of 3-acetoxy-5-nitroindole 10 was reported previously by Huang et al., 18 however comparison between the 1 H NMR and 13 C NMR spectra reported here revealed discrepancies relative to spectra reported by Huang et al. 18 for 10 (Table S1) presence of a broad singlet in the 1 H NMR spectrum at 8.27 ppm and the observation of a stretch at 3284 cm -1 in the IR spectrum were attributed to NH proton H1. Analysis of the HMBC spectrum revealed a strong correlation between protons H4 and H6 and a 13 C resonance at 135.5 ppm, assigned as C7a, which further correlated to an aromatic 1 H NMR resonance at 7.56 ppm, assigned as H2 ( Figure S62, red). Protons H2, H4 and H7 were also observed to correlated to a carbon resonance at 132.0 ppm, assigned as C3 ( Figure S62, blue).
A weak correlation between C3 and a singlet integrating to three protons at 2.41 ppm, assigned as acetoxy methyl protons H2' (Figure S62

S4.3 Structural elucidation of diallylbiindolone 16
The correlations to proton resonances at 2.89 ppm, 2.43 ppm and 5.42 ppm, assigned as diastereotopic protons H8'a and H8'b and alkenyl proton H7' (Figure S69, blue). The HSQC spectrum of 17 showed both H8'a and H8'b exhibited strong correlations to a carbon at 31.9 ppm, assigned as C8'. Prominent HMBC correlations were observed from C8' to proton H7' and a resonance at 7.18 -7.11 ppm, with the remaining proton in this multiplet assigned as deshielded alkenyl proton H6' (Figure S69, magenta). Two strong HMBC correlations were observed from protons H6' and H8'b to a carbon resonance at 121.7 ppm, assigned as quaternary carbon C9a' (Figure S69, teal).    Analysis of the NMR spectra of C-allylspiroindolinepyridoindolediones 26, 28 -29 and 33a showed identical spectral characteristics to 27, and were assigned using the key COSY, HMBC and NOESY correlations with the HRMS data ( Figure S74).

S4.8 Structural elucidation of transoid C-allylspiroindolinepyridoindoledione 33b
Analysis of the LRMS (ESI + ) of 33b revealed a base peak at m/z 443, assigned as the [M+H] + ion. Analysis of the COSY spectrum revealed an aromatic resonance at 7.10 ppm, which exhibited a strong and a weak correlation to a resonance at 6.78 ppm and a multiplet at 6.86 -6.80 ppm, assigned as H6, H4 and H7, respectively ( Figure S79, blue, black). Further examination revealed a resonance at 7.16 ppm exhibited a strong and a weak correlation to resonances at 7.00 ppm and the multiplet at 6.86 -6.80 ppm, assigned to aromatic protons H3', H4' and H1', respectively ( Figure S79, red, purple). Analysis of the HMBC spectrum of 33b showed a strong correlation from protons H4 and H6 to a carbon resonance at 156.6 ppm, assigned as C7a, which correlated to proton resonances at 4.39 ppm and 4.04 ppm, assigned as diastereotopic protons H1''a and H1''b ( Figure S80, blue). Proton H4 was also observed to correlate strongly to a deshielded carbon resonance at 197.7 ppm, assigned as carbonyl C3, which also correlated strongly to proton resonances at 2.81 -2.75 ppm and 2.05 ppm, assigned as H8'a and H8'b ( Figure S80, red) S85 exhibited a strong correlation to a carbon resonance at 69.5 ppm, assigned as C9a', which also correlated strongly to a proton resonance at 6.90 ppm, assigned as H6' (Figure S80, magenta).
Examination of the HSQC spectrum revealed a correlation between H8' and a carbon resonance at 29.9 ppm, assigned as C8', which also exhibited a strong HMBC correlation to H6' (Figure S80, black). Proton H1' exhibited a strong HMBC correlation to a carbon resonance at 197.4 ppm, assigned as carbonyl C10', which also correlated to proton resonances at 2.81 -2.75 ppm and 2.66 ppm, assigned as H1'''a and H1'''b ( Figure S80, teal). A carbon resonance at 69.1 ppm was observed to correlate to protons H1''' and H8', assigned as spirocyclic carbon

S4.9 Assignment of C-allylspiroindolinepyridoindoledione 33a-b relative stereochemistry
To assign the relative stereochemistry of the cisoid and transoid Callylspiroindolinepyridoindoledione 33a-b, analysis of the NOESY spectra was attempted, however this was inconclusive due to multiplet resonances hindering the clear assignment of key correlations. Analysis of the 1 H NMR chemical shifts of allyl methylene moieties was identified as one indicator of relative stereochemistry as wider peak splittings were expected in the cisoid isomer due to extra steric hindrance caused by clashing allyl moieties. Analysis of the 1 H NMR spectrum of the major C-allylspiroindolinepyridoindoledione isomer 33a revealed peak splittings of 0.35 ppm, 0.55 ppm and 0.84 ppm for methylene protons H1'', H8' and H3''', respectively, while these protons in minor isomer 33b showed peak splittings of 0.35 ppm, 0.13 ppm and 0.73 ppm ( Figure S81a-b). The significantly reduced peak splitting observed for protons H1''' in 33b suggest rotation is less restricted than in the case of 33a, and therefore that 219b possesses transoid relative stereochemistry.
Further analysis of the 1 H NMR spectrum of 33b revealed that the N-allyl substituent exhibited significantly deshielded chemical shifts relative to 33a, particularly the alkenyl protons H2'' and H3''a and H3''b, which were observed at 5.95 ppm, 5.65 ppm and 5.41 -5.28 ppm, respectively, in 33b compared to 5.39 -5.14 ppm (H2'') and 4.91 -4.78 ppm (H3'') in 33a (Figure S81c). Analysis of the 3D structures of the relative stereoisomers of 33 reveal that in the transoid form, the N-allyl (teal) substituent is located adjacent to the pyridoindolone phenyl moiety (yellow), while in the cisoid form, these substituents are oriented in opposite directions ( Figure S82). As a measure of the difference in distance between the N-allyl substituent and the pyridoindolone phenyl moieties in each isomer, the distance between methylene C1'' and bridgehead C10a' was measured, equalling 5.2 Å and 3.9 Å in the cisoid and transoid forms, respectively ( Figure S82). The closer proximity of the N-allyl substituent to the pyridoindolone phenyl moiety in the transoid could cause the deshielding observed in 33b, suggesting that 33a is the cisoid isomer and 33b is the transoid isomer.