Importance of Singlet Oxygen in Photocatalytic Reactions of 2-Aryl-1,2,3,4-tetrahydroisoquinolines Using Chalcogenorosamine Photocatalysts

Aerobic oxidation of 2-aryl-1,2,3,4-tetrahydroisoquinolines was achieved photocatalytically using chalcogenorosamine photocatalysts and LED irradiation. The photocatalytic aza-Henry reaction between these substrates and nitromethane was more efficient with selenorosamine and tellurorosamine photocatalysts than with thiorosamine and rosamine photocatalysts, corresponding to the propensity of the photocatalysts to generate singlet oxygen (¹O₂). Appropriately, yields for the photocatalytic aza-Henry reaction were greatly reduced when the reactions were conducted under a nitrogen atmosphere. The 2-aryl-1,2,3,4-tetrahydroisoquinolines were oxidized to the corresponding 2-aryl-3,4-dihydroisoquinolones 13a-13c with selenorosamine and tellurorosamine photocatalysts in 2% aqueous acetonitrile. Di-2-aryl-1,2,3,4-tetrahydroisoquinolin-1-yl peroxides 14a and 14b were shown to be intermediates in this reaction. Thiorosamine photocatalysts, which do generate ¹O₂ upon irradiation, did not give 2-aryl-3,4-dihydroisoquinolones. These results suggested that the exciplex between ¹O₂ and the chalcogen atom of the chalcogenorosamines (the corresponding pertelluoxide, perselenoxide, or persulfoxide) and/or the hydrated perchalcogenoxide [hydroxy (perhydroxy)tellurane, -selenane, or -thiane] might be an active oxidant in the formation of 13a-13c. Computational methods were employed to provide support for the observed photocatalytic reactivity of the tellurorhodamine and selenorhodamine chromophores compared to the thiorosamine chromophores. δG values were determined for the oxidation and hydration of 10-Te, 10-Se, and 10-S for formation of perchalcogenoxides and hydroxyl(perhydroxy)chalcogenanes, respectively. Calculations indicate formation of the pertelluroxide perselenoxide, and persulfoxide exciplex intermediates are energetically favorable. Hydration of the exciplexes of 10-Te and 10-Se have similarly small δG of -3.49 and 4.51 kcal/mol, respectively. However, a significantly higher δG value of +22.4 kcal/mol is observed for the hydration of 10-S, which suggests that this reactive intermediate is not readily formed.