Photochemical [2+2] cycloaddition reaction of enone derivatives with 2-siloxy-1 H -pyrrole derivatives

The regio-and stereo-selectivity of the photochemical [2+2] cycloaddition reaction of 2-( t -butyldimethyl- silyloxy)-1 H -pyrrole with enone derivatives are investigated. Regioselective formation of cyclobutanes was found in the intermolecular photochemical [2+2] cycloaddition reaction of the pyrrole derivative with cyclopentenone. In contrast, an oxetane rather than a cyclobutane product was selectively formed when 1,4-naphthoquinone was used as the enone. The reaction conditions were optimized to obtain the cycloadducts in high yields.


Results and Discussion
The pyrrole 2-(t-butyldimethylsilyloxy)-N-t-butoxycarbonylpyrrole (1) was prepared from the corresponding lactam according to the literature. 43Firstly, the reaction mixture of 1 (0.51 M) and 4 (0.34 M) was irradiated by 365 nm using light-emitting diode (LED) at room temperature in a sealed Pyrex NMR tube in acetone-d 6 after the deoxygenation by 20 minutes nitrogen bubbling.According to the NMR spectroscopic analysis of the reaction mixture after 24 hours irradiation (Figure 1e), newly appeared signals at δ 6.8, 3.5-2.0were assigned to the products 6 and 7.After removing the solvent under reduced pressure, the products 6 and 7 were isolated in 12% yield with 64 : 36 isomer ratio by silica gel chromatography (entry 1 in Table 1).The structures of 6 and 7 were confirmed by 1D-NMR and 2D-NMR spectroscopic analyses, i.e. 1 H, 13 C, H-H COSY, H-C HSQC, H-C HMBC, H-H NOESY, and high-resolution mass spectroscopic analysis.The formation of further regioisomers 9 and 10 was excluded by the observed two vinyl protons H a and H b at 6.75, 6.60 ppm and 5.15, 5.07 ppm, respectively, in the formed cycloadducts (Figures 2 and 3).Because the two protons, H a (δ 6.75, 6.60) and H b (δ 5.16, 5.09), were correlated with the sp 2 carbons (δ 132.34, 132.32, 107.4,107.0), the two protons at ~6.7 and ~5.1 ppm were assigned to the vinylic protons (Figure 3).In the H-C HMBC spectrum (Figure 4), the correlation of H d (δ 2.98) and the cyclic carbonyl carbon (δ 219.7) was observed in the major isomer of the products, indicating that the major product was the tricyclic cyclobutane 6.The stereochemistry of compounds 6 and 7, exo versus endo, was determined by their 2D-NOESY spectroscopic analyses (Figure 5).Thus, the correlations between H b (δ 5.20, 5.13) and H c (δ 2.38, 2.37) were observed in the H-H NOESY spectrum, indicating that both compounds 6 and 7 possess the exo-configuration as shown in Figure 5.The reaction conditions were optimized to obtain 6 and 7 in high yields, because the large amount of enone 4 was recovered under the conditions in entry 1 (Table 1).The ratio of compound 1 and 4 dramatically influenced the conversion of compound 4 to give high yields of compounds 6 and 7 (entries 1-4, Table 1).When a large excess of compound 1 (15 equiv) was used for the photochemical reaction with 4, 72% of compounds 6 and 7 were obtained after 24 h using 365 nm-LED in acetone (entry 4).The yield of adducts 6 and 7 decreased when acetonitrile or benzene was used as a solvent.(entries 5 and 6, Table 1), indicating the reaction occurs in the triplet excited state.The product ratio of 6 and 7, ~65:35, was not affected by the concentration or solvent (Table 1).Next, the photochemical reaction of naphthoquinone ( 5) with 1 was conducted under the similar reaction conditions (Scheme 1, Table 2).5][46][47][48][49][50] However, on irradiation of 1 with 5 the formation of the oxetane product 8 was observed, and the alternative possible cyclobutane products were not observed in the photo-reaction.
The high pressure Hg lamp (300W) through Pyrex filter was found to be better light source than the 365 nm LED lamp for the reaction (entries 1,2, Table 2).The chemical yield of compound 8 was largely dependent upon the compound ratio of 1/5 and solvent.When the excess amount of compound 1 was used for the reaction (entries 2-4), the conversion yield of 8 was higher than that for the reaction with excess amount of compound 5 (entries 5,6).Acetone was inappropriate for this PB reaction of 1 and 5 (entry 8).The structure of oxetane 8 was confirmed by the measurement of 1D, 2D NMR and HRMS spectra.Except the protons of the Boc and silyloxy groups, only one aliphatic proton H 5 at δ 3.80 was observed in the isolated product, which was correlated to the sp 3 carbon at δ 64.19 according to H-C HSQC spectrum (Figure 6).In addition, the findings of correlation of H 1 and H 2 and correlation of H 1 and H 5 in H-H COSY spectrum could also exclude the possibilities of regioisomers (Figure 7).The observation excludes the possibility of the formation of cyclobutanes and the regioisomeric oxetane shown in Figure .6.In addition to the spectroscopic evidence, only one carbonyl carbon at δ 183.9 was found in the photochemical product which is correlated to H 16 (δ 8.09) and H 12 (δ 7.30) that also could exclude the possibility of a cyclobutane adduct (Figure 8).The stereochemistry of compound 8 was determined to be exo, because the correlation of H 5 (δ 3.80) and H 13 (δ 8.39) was observed in the H-H NOESY spectrum (Figure 9).

Conclusions
In the present study, regio-and stereoselectivity of photochemical reaction of the pyrrole derivative 1 with two types of enone derivatives 4 or 5 were investigated in detail.In the reaction with cyclopentenone 4, exoselective formation of tricyclic cyclobutanes 5 and 6 was observed.In contrast to the selective formation of cyclubutane formation, the selective formation of oxetane was found in the reaction with naphthoquinone (5), although the reaction conversion was rather low.The mechanistic study based on the quantum chemical and spectroscopic investigation is now ongoing to clarify the dramatic effect of the enone derivative on the product selectivity.

Experimental Section
General.All the reagents and solvents were obtained at reagent grade and used without further purification.Thin layer chromatography (TLC) analysis was performed on silica gel plates and imagined under ultraviolet light. 1 H NMR and 13 C NMR data were recorded with an 400 MHz NMR spectrometer.CDCl 3 (0.03% TMS) was used as deuterated solvents.Chemical shifts were described in parts per million (ppm) relative to trimethylsilane (δ = 0.00 ppm), and the coupling constants (J) was stated in Hertz (Hz).High-resolution Mass (HRMS) spectroscopic analyses were conducted using a Orbitrap XL instrument using the positive ion mode.
Starting material 1 was prepared by the reported method. 43eneral procedure of the photoreaction of 1 and 4, 5 a. [4] or [5] was 340 mM The solution of starting material in Pyrex NMR tube was bubbled with nitrogen for 20 minutes.After irradiation (high pressure Hg lamp or 365 nm light-emitting diode), the reaction mixture was concentrated under reduced pressure.The desired product was purified by silica gel column chromatography (elution of EtOAc : Hexane = 1:8).b. [4] or [5] was 42.5 mM The solution of starting material in test tube was degassed by freeze/pump/thaw 5 times.After irradiation (365 nm light-emitting diode), the reaction mixture was concentrated under reduced pressure.The desired product was purified by silica gel column chromatography (elution of EtOAc : Hexane = 1:8).

Supplementary Material
1 H, 13 C and HRMS spectra of 6, 7 and 8 and UV spectra of 4 and 5.

Table 1 .
Photochemical reaction of 1 with cyclopentenone (4) a a Acetone was used for the reaction solvent.A light-emitting diode lamp at 365 nm was used for 24 hours irradiation; b Isolation yield after silica gel chromatography; c Acetonitrile was used for solvent; d Benzene was used for solvent.

Table 2 .
Optimization of the reaction conditions of 1 and 1,4-naphthoquinone (5) a bIsolation yield after silica gel chromatography; c The yield was calculated based on the consumed naphthoquinone by 1 H NMR analysis using Ph 3 CH as internal standard; d Light-emitting diode lamp at 365 nm was used for light resource; e Acetone was used as solvent.