We next sought to interrogate the mechanism of this reaction. Addition of TEMPO inhibited the reaction, and addition of 1,1-diphenylethylene successfully captured chlorine radical, affording the radical adduct
4 (
Schemes 5a and
b). To probe the generation of chlorine radical, a series of UV–vis spectra were carried out. UV–vis spectroscopy of CuCl
2/CH
3CN and CuCl
2/HCl/CH
3CN solution exhibited typical peaks of [(MeCN)
2CuCl
2] and [(MeCN)CuCl
3]
−, respectively (
Fig. 1a) [
65], indicating [(MeCN)
2CuCl
2] would react with HCl to form [(MeCN)CuCl
3]
−. According to literature [
30], [(MeCN)CuCl
3]
− would readily undergo LMCT to generate chlorine radicals. This was further confirmed by the irradiation of CuCl
2/HCl/CH
3CN solutions, UV–vis spectroscopy of which showed [(MeCN)CuCl
3]
− vanished (
Fig. 1a). To further confirm the change of cupric oxidation state in the catalytic cycle, the X-ray photoelectron spectroscopy (XPS) measurement of the reaction mixture using the CuCl
2/HCl system was carried out (
Fig. 1b).
Fig. 1b shows the high resolution XPS scans over Cu 2p
3/2 peak. The peak at 932.5 eV was known as the characteristic of Cu
+ [
66], while the peak at 934.3 eV together with shake-up satellite peaks on the higher binding energy side, 942.4, and 944.6 eV, indicated the presence of an unfilled Cu 3d shell and thus confirmed the existence of Cu
2+ [
67,
68]. The results of UV–vis spectroscopy and XPS suggest that the
β-chlorination reaction may proceed through a Cu(Ⅰ)–Cu(Ⅱ) involving LMCT mechanism. To investigate the possibility of 1,4-HAT processes, we subjected substrate
5 to the standard conditions (
Scheme 5c). Notably,
β-halogenated product
6 was not detected, but 27% of product
1b and 9% of product
2b were obtained. It would be attributed to the lack of N-H bond in
5, which is essential to trigger 1,4-HAT for
β-halogenation by converting N-H bond into N−Cl bond with chlorine radical from LMCT. Under the standard condition,
1s provided
2s with 6% of ring-opening product
7, a typical radical clock product (
Scheme 5d) Quantum yield and light on/off experiments suggest that the transformation needed continuous irradiation of visible light and is not a radical chain processes (see Supporting information).