Quantum Confinement Effects in II-Vi Semiconductor Nanocrystals

The quantum size effect and other associated properties of CdS_{rm x}Se _{rm 1-x} nanocrystals (quantum dots) embedded in a glass matrix have been extensively studied. By varying annealing and growth conditions, we obtained a series of samples with crystallite sizes from 18 A to 124 A (diameter). Raman experiments and transmission electron microscopy were performed to determine the composition and dot size. Absorption and electromodulation spectra were measured to identify the quantized energy levels. Careful and extensive analysis of the data shows that it is necessary to consider the effect of a finite well-depth of the conduction band. Also the magnitude of the quantum-confined Stark effect in CdS_{rm x}Se _{rm 1-x} nanocrystals has been measured. After normalization to the semiconductor volume, the magnitude of the Stark effect is only weakly dependent on particle size, contrary to the predictions of simple models. We suggest that it is necessary to include the effects of finite quantum well depth on the wavefunctions of electron and hole to explain this weak dependence. We also compare low-power steady-state photoinduced transmission modulation, photoluminescence and electromodulation measurements on a series of well-characterized samples. We propose that the photomodulation effect is dominated by coulombic interaction of trapped carriers for small (< 50 A) particles. We also propose that a combination of state-filling and coulombic interaction is responsible for the observed photomodulation effect in larger ( ~100 A) crystallites. Photoluminescence has been used to study surface states in the system. A relatively strong photoluminescence band, from well below the absorption edge, is observed for smaller particles. This suggests the importance of surface states when particle size reduced. In the theoretical modeling, we present a calculation for the first quantized energy level for optical transitions in CdS_{rm x}Se _{rm 1-x} quantum dots (radius from 10 A to 100 A) with a finite potential well depth model. It is shown that the energy level is significantly lower than predicted by infinite well depth model for dot radius smaller than its bulk exciton radius (~30 A). The importance of finite potential well depth is emphasized and comparison with experimental data is discussed.