Isotope splitting of the zero-phonon line of Fe2+ in cubic III–V semiconductors

Abstract

A theoretical study of the isotopic-mass dependence of the internal transitions of Fe 2+ at a cation site in a cubic zinc-blende semiconductor is presented. The model used is based on crystal-field theory and includes the spinorbit interaction and a weak dynamic Jahn-Teller coupling between the 5 Γ 5 excited manifold of Fe 2+ and a local vibrational mode (LVM) of Γ 5 symmetry. The mass dependence of the LVM frequency is described, in the harmonic approximation, within two different limits: the rigid-cage model and a molecular model. In the rigidcage model, the Fe 2+ ion undergoes a displacement but the rest of the lattice is fixed. In this case, a simple M -1l2 dependence of the frequency is obtained and the Jahn-Teller energy, E JT is independent of the mass. In the molecular model, the four nearest neighbors of the magnetic ion are allowed to move and the LVM then behaves as the Γ 5 mode of a MX 4 tetrahedral molecule leading to a more complicated dependence of the frequency on the isotopic mass and to a mass-dependence of E JT . The theoretical results obtained with these two models are compared with the observed isotopic shifts of the zero-phonon lines in InP:Fe and GaP:Fe corresponding to an optical transition between the vibronic Γ 1 ground state and the lowest Γ 5 state originating from the 5 Γ 5 excited orbital multiplet. A prediction of the isotopic shifts of the zero-phonon line in GaAs:Fe is also presented.

The effect of electron-phonon interaction in iron-doped III-V cubic semiconductors

“…We proposed a new interpretation of both the emission and absorption spectra of Fe 2+ in InP and GaP, which also accounts for the isotopic shifts of the zero-phonon line observed in this compound. Finally, the spin-orbit coupling constants deduced from our calculations show that the covalent character of the bonds is much more important in III-V compounds than in II-VI compounds such as ZnS [24].…”

“…Experimental Calculated Relative Transition energy (cm −1 ) energy (cm −1 ) intensity λ 1 /λ free ion = 0.56 (InP:Fe) and 0.62 (GaP:Fe) is much smaller than in the case of II-VI semiconductors [24]. This ratio is a measure of the degree of covalency of the Fe-P bonds and, following the theory of Vallin and Watkins [21], the ratio 0.56 (0.62) means that the d electrons of the Fe ion spend 56% (62%) of their time in the vicinity of the magnetic impurity and 44% (38%) of the time in the vicinity of the phosphorus atoms.…”

“…These two figures can be used to get an idea of the approximate values of E JT andhω 5 that need to be used in the full numerical calculation. In order to explain the isotopic shifts of line I, it is most favourable to have as large a JT energy as possible [13]. At the same time, one must also reproduce the intensity ratio of line III to line II.…”

“…They depend on the isotopic mass of the Fe 2+ impurity and will contribute to the isotopic shift of the zero-phonon line. To describe the mass dependence of the gap mode frequency we use the MX 4 molecular model [13,23]. This model assumes a central atom of mass M (i.e., Fe) surrounded by a tetrahedral cage of nearest-neighbour cations of mass M − (i.e., P).…”

“…A consequence of the more detailed experimental data is that it became clear that pure crystal-field theory cannot fully explain the spacings of zero-phonon lines and the isotopic fine structure. The existence of a Jahn-Teller coupling between the electronic states of the Fe 2+ substitutional impurities and phonons or local vibrational modes of the host III-V semiconductor was investigated by a number of people [10][11][12][13]. All these papers focused on only some of the experimental data but did not attempt to explain with a single model the optical absorption, photoluminescence and isotopic measurements.…”