In this thesis, we studied spin dynamics, optical nonlinearity and the optical Stark effect in bulk GaSe, mono- and few-layer GaSe, and colloidal CdSe nanocrystal quantum dots (NQDs), respectively.Control of the spin has been a long-term goal due to its potential applications in quantum information processing. Candidates for spintronics should have a long spin lifetime and allow for generation of a high initial spin polarization. GaSe caught our attention due to its orbitally nondegenerate valence bands, which are in contrast to the degenerate heavy and light hole valence bands in conventional III-V and II-VI semiconductors, like GaAs and CdSe. With time- and polarization-resolved photoluminescence, we demonstrated the generation of initial spin polarization as high as 0.9 followed by bi-exponential spin relaxation at 10 K ($\sim$30 ps and $\geq$300 ps), owing to such orbitally nondegenerate valence bands in GaSe. We also directly revealed the initial spin and population relaxation as transitions from triplet excitons to singlet excitions via spin-flip of the electron or hole.Contrary to semiconductor transition metal dichalcogenides, MX$_2$(M=Mo,W; X=S,Se,Te), GaSe is a direct band gap semiconductor in bulk, but transforms to an indirect band gap semiconductor in a monolayer as the maximum of the valence band is shifted away from the $\Gamma$ point. Associated with such a valence band in monolayer GaSe, ferromagnetism has been predicted upon hole doping due to a strong electronic exchange field. To study the electronic structure of GaSe in mono- and few-layer GaSe, we measured layer- and frequency-dependent second-harmonic generation (SHG) in GaSe from monolayer to $\geq$100 layers and determined a second-order optical nonlinearity $\chi^{(2)}$ in the multi-slab system. We found reduced a $\chi^{(2)}$ in GaSe with thickness$\lesssim$ 7 layers, tentatively attributed to the predicted increase in the band gap.How quantum confinement affects the light-matter interaction in colloidal CdSe nanocrystal quantum dots (NQDs) has been a long-standing puzzle. The effective mass approximation (EMA) with $k\cdot p$ method has been successful in describing the energies of size-dependent optical transitions and the fine structures of the lowest-energy excitons. However, in contrast to the EMA's prediction of a nearly size-independent oscillator strength of the lowest-energy exciton, numerous experimental studies reported strongly reduced oscillator strength in CdSe NQDs as the size decreases, and failed to give an explanation. We developed a method based on the optical Stark effect to extract the oscillator strength of the lowest-energy excitons in semiconductor nanocrystals, and demonstrated a nearly constant oscillator strength from 6.7 nm to 2.5 nm, consistent with the prediction of the EMA. The optical Stark effect has been used as a method of transiently manipulating the spin. With circularly polarized off-resonant pump, we achieved an energy splitting of $\sim$9 meV between transitions with different helicities, corresponding to an effective magnetic field of 110 T.In this thesis, I will first introduce the background for all the optical studies, and then talk about the optical techniques before discussing the three projects mentioned above.
Read
