No difference in local structure about a Zn dopant for congruent and stoichiometricLiNbO3

Abstract

We compare EXAFS (extended x-ray absorption fine structure) data at the Zn K edge for a low concentration of Zn (0.7 mol%) in a stoichiometric crystal with that for higher Zn concentrations (nominally 5 and 9 mol%) in congruent LiNbO3 (LNO). Note that stoichiometric and congruent LNO have significantly different optical properties. We find no significant difference in the local structure about Zn out to 4Å for the two types of crystals and different dopant levels. Although some earlier theoretical models suggest a self compensation model with 75 % of Zn on a Li site and 25 % Zn on Nb, we find no clear evidence for a significant fraction of Zn on the Nb site, and estimate at most 2-3 % of Zn might be Zn Nb .

EXAFS simulations in Zn-doped LiNbO3based on defect calculations

“…Figure 1 shows the simulated Fourrier transform of the EXAFS oscillations (red curves) for the model where Zn substitutes for the Li site ( figure 1(a)) and substitutes at the Nb site ( figure 1(b)) but considering the perfect lattice site with no distortion. Figure 1 also reproduces for comparison the experimental result of LiNbO 3 :Zn sample (blue curves) extracted from reference [5]. For all simulated curves, the k space used in the Fourier transform of the EXAFS oscillations is compatible with the ones used in the experiment, i.e., k in the range from 3.8 to 13.2Å -1 .…”

“…The first attempt would be to include the simplest possible charge compensating that is a Li vacancy, although the defect simulations presented in ref [2] did show that this is not the defect with the lower solution energy. One can argue that since the concentration of Zn in the samples measured in ref [5] is low, the Zn ions should be far apart from each other and the neutral defect that would keep the Zn isolated is the !" − !"…”

“…The final simulated EXAFS was taken to be the average of the four individual simulations. The Debye-Waller factor was set to 0.003Å -2 that is a typical value for low temperature measurements and is close to the experimental values fitted to the EXAFS data in refs [4] and [5]. Figure 1 shows the simulated Fourrier transform of the EXAFS oscillations (red curves) for the model where Zn substitutes for the Li site ( figure 1(a)) and substitutes at the Nb site ( figure 1(b)) but considering the perfect lattice site with no distortion.…”

“…The clusters, thus, already include the lattice distortion around the Zn ions due to the defect generated in the LiNbO3 lattice and since both the dopant and the charge compensating defects were included in the defect calculations, the distortion represents more accurately the system. The defect modelling calculations were done at 0K and this temperature was chosen due to the fact that the experimental data from refs [4] and [5] were measured at low temperature.…”

“…Recent EXAFS studies ( [4], [5]) suggest that Zn occupies the Li site, and the model was based on assuming as …”