In Electron Microscope (EM), X-ray fluorescence occurs as a result of photoelectric absorption of primary X-rays as the target atom relaxes from an excited state. These secondary X-rays may be produced both by characteristic or continuum x-rays. Since X-rays can penetrate much longer distances than incident electrons in materials, the range of fluoresced X-rays is greater (~10 times) than the interaction range of the primary electron at the same energy. The secondary X-ray interaction volume may be as much as 1000 times greater than the interaction volume of the primary X-rays.

For a K-shell ionization, the X-ray yield ϖ_K is defined as the fraction of atomic decays by emitting a K-shell X-ray. The rate of X-ray emission is evaluated for an electric dipole transition between the two atomic states of the atom, |i> and |j>, and is the ratio of the X-ray rate to the total rate, which includes the X-ray rate and Auger electron rate. For a K-shell emission, ϖ_K empirically depends on atomic number, Z, given by,

        ϖ_K = Z⁴/(10⁶+Z⁴) ----------------------------- [4637]

Therefore, heavy elements tend to emit X-rays and light elements tend to emit Auger electrons.

Figure 4637a shows X-ray (fluorescence) yield ω as a function of atomic number for electron ionization within the K, L, and M electron shells. The yield is defined by emitted photons per ionization, which is the fraction of ionizations that results in X-ray emission. For each shell, we have ω_K > ω_L > ω_M, where ω_K, ω_L, and ω_M are yields for K, L, and M shells, respectively. For atoms with low atomic numbers, e.g. Z < 26, there are no ω_L and ω_M yields because there are less electrons. Detailed information on yields, sometimes called weights and refered to intensities, can be found in the Section Photon Energies and Relative Intensities of K-, L-, and M-Shell Lines.

As an example, Figure 4637b shows the X-ray and Auger electron yields per K vacancy as a function of atomic number as well as from all vacancies generated by accelerated incident electrons. For low-Z material Auger emission is dominant, while for high-Z material X-ray emission dominates. The two emission probabilities are approximately equal at Z ≈ 30 for K shell ionization.

The relationship between the yields of K X-ray lines (ω_K) and K Auger electrons (α_K) per K vacancy is given by,
         ω_K + α_K = 1 ------------------------------------- [4637a]
And, the relationship between the yields of all X-ray lines (ω_all) and all Auger electrons (α_all) per vacancy is given by,
         ω_all + α_all = 1 ------------------------------------- [4637b]

Furthermore, X-ray emission also depends on the energy of incident beam. For incident electrons at low energies, it is very important to know that a useful X-ray signal can only be generated with a beam energy which is at least 1.3 x the ionization energy for the relevant characteristic X-rays. X-ray yields, in this condition, are also relative low and depend strongly on overvoltage  (especially when U<3) and the X-ray signals are relatively weak. For effective generation of X-ray by ionization processed by electron interaction, the overvoltage U should be greater than 1.3 because the X-ray peak to background ratio (P/B) of X-ray signal is reduced substantially as E₀ (energy of incident beam) approaches E_i (ionization energy). It is necessary to mention that the low energy X-ray lines, e.g. Cu(L) at 0.93 keV instead of Cu(K) at 8.04 keV, are used at low accelerating voltage of incident electrons.