Relation between copperLx-ray fluorescence and 2px-ray photoelectron spectroscopies

Lα1,2,β₁(L_3,2-V) x-ray fluorescence spectra (XRF) and 2p1/2,3/2 x-ray photoelectron spectra (XPS) of various copper compounds are measured. It is found that the intensity of the high-energy hump of the Cu L{α} XRF has a correlation with that of the high-binding-energy satellite (corresponding to the poorly screened 2p^-1 final state) of the Cu 2p3/2 XPS. While both the poorly screened peak in the 2p3/2 XPS and the high-energy hump in the L{α} XRF are strong for ionic divalent copper compounds, both of them are very weak for covalent divalent copper compounds, and they exist for neither monovalent nor metallic copper compounds. It was believed that the high-energy hump of the L{α} XRF originated from the electron transition between L₃{M}_{4,5{{-}}}M4,5² multiple-hole states, where the L₃{M}4,5$ double-hole state was created by the ${L}1,2L₃{M}4,5$ Coster-Kronig transition prior to the x-ray transition. In this context, the L{α} line shape, except for the high-energy hump, was believed to represent the Cu 3d electron density of states (DOS). Our results, however, exclude the possibility of the multiple vacancy satellite 2p⁵3d⁸{→}3d⁷ for the origin of the high-energy hump of the L{α} XRF of the divalent copper compounds. It is concluded that the major portion of the high-energy hump of the L{α} XRF of the divalent copper compounds is due to the transition between the poorly screened states 2p⁵3d⁹{→}3d⁸. Consequently, it is also concluded that the L{α} line shape does not directly represent the 3d DOS but the high-energy hump hidden in the L{α} main line represents the 3d DOS. We also conclude that the L{α} main line originates from the charge-transfer effect.