Monitoring defect-induced perturbations of the ideal crystal structure of ZnO and Cu2O by Raman spectroscopy
Deviations from the ideal crystal structure, i.e., intrinsic or extrinsic defects, modify the dynamics of a crystalline lattice among many other material properties. As a consequence, additional modes may occur in Raman spectra, Raman-forbidden modes may become Raman active, and the shape as well as the position in the frequency of the Raman ... signals may change. The impacts of perturbations of the ideal crystal structure on Raman spectra are addressed within the framework of this thesis.Crystalline ZnO samples are investigated using Raman spectroscopy. It is found that all experimental results are consistent with the laws of conservation as well as the Raman activity and the symmetry selection-rules established for an ideal crystal structure. With this in mind, the samples studied should be free of defects and have an ideal hexagonal crystal lattice. In reality, there are no perfect crystals as entropy always leads to the formation of defects . Furthermore, it is well known that nominally undoped ZnO is often intrinsically n-type conducting due to intrinsic and/or, more likely, due to extrinsic defects [7-10]. Therefore, Raman spectroscopy is not sensitive to such kinds of defects at the concentration level present in the ZnO samples studied and in the frequency range considered here in experiment.The situation is different when ZnO is intentionally doped with nitrogen. Raman spectra of nitrogen doped ZnO exhibit five additional modes independent of the growth technique applied. Various attempts to explain the appearance of the nitrogen-related modes are reviewed [11-19]. A promising approach is the assignment of the N-related modes to silent ZnO phonons induced by the breakdown of the translational symmetry of the lattice . The Raman active ZnO phonons and the N-related modes are analyzed with respect to the amount of nitrogen incorporated in ZnO as well as their behavior upon hydrostatic pressure applied and sample rotation. The findings reveal discrepancies for the silent B modes between the outcome of ab initio calculations and the nitrogen-related modes observed in experiments. In addition, their symmetry characteristics cannot be represented in a conclusive picture by silent B modes of wurtzite ZnO if the assumption is valid that the perturbation introduced by nitrogen on an anion site is large enough to manipulate the Raman activity and the symmetry character of the extended lattice modes. Thus, the assignment of the additional signals to silent B modes of wurtzite ZnO becomes disputable. Consistent with all experimental findings, the formation of clusters containing N atoms/ions, in particular, of Zn3N2-like clusters that are formed when the solubility limit of N on O sites is exceeded, are proposed as the origin of the additional signals. Raman spectra of crystalline Cu2O are often dominated by Raman-forbidden lattice vibrations rather than by the single Raman-active phonon predicted by group theory for the ideal crystal structure . A group theoretical analysis of the symmetry reduction due to the presence of defects demonstrates that the copper split vacancy, a point defect particular to Cu2O, introduces Raman activity for all phonons that are Raman forbidden. The symmetry characteristics of the lattice vibrations observed in Raman spectra corroborate this assumption. These findings are surprising: Raman spectra of most crystalline materials are consistent with the laws of conservation as well as the Raman activity and the symmetry selection-rules established for an ideal crystal structure, despite the considerable amount of defects present in crystalline materials as shown for pure ZnO. The special role of the split vacancy may be due to the strong perturbation introduced by it compared to simple point defects, which is indicated by DFT calculations for supercell defect structures. However, resonant Raman scattering mechanisms cannot entirely be ruled out to alter the Raman activity and symmetry selection-rules such that all phonons become Raman active, since the analyzed Raman spectra are dependent on the excitation frequency. The theoretical discussions and experimental investigations within the present thesis demonstrate that Raman scattering is influenced in many different ways by defects in the material under study. A thoughtful analysis of the corresponding Raman spectra may give insight into occurring defects and their properties. In particular, Raman experiments upon sample or plane of polarization rotation of extended lattice vibrations and defect-related modes may reveal their symmetry character and accordingly information about their origin.