The copper Kα doublet, comprising Kα1 and Kα2 X-ray lines, is generated in an X-ray tube through a series of steps. Electrons, emitted by thermionic emission from a heated cathode, are accelerated towards the copper anode. Upon collision with copper atoms in the anode, these high-energy electrons cause inner-shell ionization, particularly in the K-shell. This process results in the creation of a vacancy in the K-shell, leading to an excited state of the atom. To regain stability, outer-shell electrons move to fill the vacancy, releasing energy in the form of X-rays. Kα corresponds to the transition from the L shell to the K shell, Kβ corresponds to the M shell to K shell transition, and Kγ corresponds to the N shell to K shell transition, each resulting in the emission of X-rays with distinct energy levels. The closely spaced Kα1 and Kα2 X-ray lines are a result of different electron transitions within the inner and outer electron shells. These two X-ray lines are closely spaced because generally the energy levels of electrons residing at a higher shell is not identical. They have slightly different binding energies, resulting in two X-ray lines with slightly different wavelengths.
sorry, but let me clearify: there is only one K-shell level, but there are three L shell levels; i.e. LI, LII and LIII.
The electron transition from the LIII shell will result in the release of a K-alpha1 fluorescence photon and the transition from the LII shell will release the K-alpha2 photon. They are 'closely spaced' in energy and/or wavelength as already mentioned by Souvik Bhattacharjee.
So far the production of the Cu K-alpha doublet in the x-ray tube spectrum is described.
The split of the diffraction peaks into K-alpha1 and K-alpha2 doublets is due to the Bragg law (lambda = 2*d*sin[theta]), which for fixed lattice plane spacings d, assigns each wavelength lambdaK-alpha1 and lambdaK-alpha2 to a special theta (or 2theta) position the XRD pattern. In most of the cases the doublet peaks overlap strongly; see for example page 8 of
As @Gerhard show, there is also a weak K-alfa 3, but it is covered by two other strong lines. Sometimes, as in Topas, it has been considered in Refinement.
Gerhard Martens , in fact there are indeed Cu Kalpha3,4 lines, and even some hypersatellite lines, which can be seen in XRD patterns of good monocrystals. You can find details here: https://nvlpubs.nist.gov/nistpubs/jres/109/1/j91deu.pdf