Raman polarization and symmetry in semiconductor nanocrystals

 

Raman spectroscopy is usually performed with polarized light from a laser, and the Raman scattered light having polarization either parallel or perpendicular to that of the laser can be detected. The perpendicular to parallel intensity ration is known as the Raman depolarization ratio and is typically measured in Raman studies of molecules, but rarely in semiconductor nanocrystals. Semiconductor quantum dots are often cartooned as perfect spheres, and a spherical structure has a Raman depolarization ratio of zero: since the polarizability is isotropic, the induced polarization is always along the same direction as the laser polarization. However, depolarization ratios of zero are not observed experimentally, whether for CdSe QDs in the uniaxial wurtzite crystal structure or for ZnSe QDs in the isotropic zincblende structure. The optical asymmetry may be a result of shape asymmetry, charged defects on the surface, asymmetric ligand coverage, and/or different crystal facets on the surface (metal-rich versus chalcogenide-rich).

We are undertaking a systematic study of resonance Raman polarization in II-VI nanocrystals, both in nominally highly symmetric shapes (dot, core/shell dot, cubes) and shapes in which the anisotropy can be varied systematically (rods, dot-in-rods, platelets). These measurements will also be made as a function of excitation wavelength, from the lowest excitonic transitions to higher energies where the nature of the states is not well understood. In particular, we seek to understand the origin of the “continuum” absorption superimposed on the excitonic transitions predicted by effective mass models. To what extent does this continuum absorption arise from the breakdown of the symmetry selection rules predicted for simple, symmetric structures?

Nanocrystal shapes to be studied (not to scale).