Photophysical Properties of II-VI Semiconductor Nanocrystals
- Author(s): Gong, Ke
- Advisor(s): Kelley, David F
- et al.
As it is well known, semiconductor nanocrystals (also called quantum dots, QDs) are being actively pursued for use in many different types of luminescent optical materials. These materials include the active media for luminescence downconversion in artificial lighting, lasers, luminescent solar concentrators and many other applications. Chapter 1 gives general introduction of QDs, which describe the basic physical properties and optical properties. Based on the experimental spectroscopic study, a semiquantitative method-effective mass model is employed to give theoretical prediction and guide. The following chapters will talks about several topics respectively. A predictive understanding of the radiative lifetimes is therefore a starting point for the understanding of the use of QDs for these applications. Absorption intensities and radiative lifetimes are fundamental properties of any luminescent material. Meantime, achievement of high efficiency with high working temperature and heterostructure fabrication with manipulation of lattice strain are not easy and need systematic investigation.
To make accurate connections between extinction coefficients and radiative recombination rates, chapter 2 will consider three closely related aspects of the size dependent spectroscopy of II-VI QDs. First, it will consider the existing literature on cadmium selenide (CdSe) QD absorption spectra and extinction coefficients. From these results and fine structure considerations Boltzmann weighted radiative lifetimes are calculated. These lifetimes are compared to values measured on very high quality CdSe and CdSe coated with zinc selenide (ZnSe) shells. Second, analogous literature data are analyzed for cadmium telluride (CdTe) nanocrystals and compared to lifetimes measured for very high quality QDs. Furthermore, studies of the absorption and excitation spectra and measured radiative lifetimes for CdTe/CdSe Type-II core/shell QDs are reported. These results are also analyzed in terms of a Boltzmann population of exciton sublevels and calculated electron and hole wave functions. Much of the absorption data and fine structure calculations are already in the literature. These results are combined with new measurements of radiative lifetimes and electron-hole overlap calculations to produce an integrated picture of the II-VI QD spectroscopic fundamentals. Finally, we adopt recent synthetic advances to make very monodisperse zincblende CdSe/CdS quantum dots having near-unity photoluminescence quantum yields (PLQYs). Due the absence of nonradiative decay pathways, accurate values of the radiative lifetimes can be obtained from time resolved PL measurements. Radiative lifetimes can also be obtained from the Einstein relations, using the static absorption spectra and the relative thermal populations in the angular momentum sublevels. One of the inputs into these calculations is the shell thickness, and it is useful to be able to determine shell thickness from spectroscopic measurements. We use an empirically corrected effective mass model to produce a “map” of exciton wavelength as a function of core size and shell thickness. These calculations use an elastic continuum model and the known lattice and elastic constants to include the effect of lattice strain on the band gap energy. Radiative lifetimes calculated both experimentally and theoretically are checked and the size dependence is compared to previous studied Type-I, II and single component particles.
However, it is not enough to just understanding these basic photophysics of absorption and emission. The emission intensities (related to QYs) also change with changes of the temperature. The temperature dependent PLs of II-VI QDs is extensively studied, but most of this work is at low temperatures. Temperatures well above ambient are of interest to lighting applications and in this regime both the reversible and irreversible loss of quantum yield (thermal quenching) are serious impediments to the implementation of QDs in commercial devices. Chapter 3 will elucidate the mechanism of static thermal quenching, in which the reduction of QYs does not affect the PL decay kinetics, on CdSe, CdTe and CdSe/ZnSe QDs as a function of particle sizes/shapes, surface composition and surface ligands. Through systematic experiments, this part of the dissertation discusses several possible mechanisms (e.g. structural, activated excited state, and electronic charging) and examines which the dominant cause for loss of QY at high temperature is. The more practical step is to develop the synthetic method of highly luminescent and stable core/shell QDs with minimum thermal quenching, which greatly enhance the energy efficiency of light emitting and photovoltaic devices.
As the nonradiative Auger processed are induced by surface charging described in chapter 3, static and time-resolved fluorescence and high and low power transient absorption results on CdSe/CdS and CdSe/ZnSe core/shell particles are presented in chapter 4. Two CdS shell thicknesses were examined and all of the particles had either octadecylamine (ODA) and tributylphosphine (TBP) or just ODA ligands. The results can be understood in terms of a mechanism in which there is a thermal equilibrium between electrons being in the valence band or in chalcogenide localized surface states. Thermal promotion of a valence band electron to a surface state leaves the particle core positively charged. Photon absorption when the particle is in this state results in a positive trion, which undergoes a fast Auger recombination, making the particle nonluminescent. A lack of TBP ligands results in more empty surface orbitals and therefore shifts the equilibrium toward surface trapped electrons and hence trion formation. Low- and high-power transient absorption measurements give the trion and biexciton lifetimes and the ratio of the trion to biexciton Auger lifetimes are examined and compared to the degeneracies of Auger pathways. We also study the shell thickness and composition dependence of Auger times, which is compared to the scaling factors of effective volume and electron-hole overlap considerations.
Core/shell QDs often exhibit much higher luminescence quantum yields (QYs), more stability, and are depicted as having a nearly spherical core and a shell of very nearly uniform thickness, which results in a very simple picture of surface passivation. The uniformity of the shell is crucial in obtaining QDs with well passivated surfaces. However, transmission electron microscope (TEM) images disprove the ideal situation. Defects and thickness inhomogeneity in shell materials are treated qualitatively as an analog to film thickness inhomogeneity in epitaxially grown thin films. More quantitatively, the extent to which the shell thickness of core/shell particles is constant can be determined by time-resolved PL studies that measure the dynamics of hole tunneling to acceptors that are adsorbed on the shell surface due that tunneling rates varies strongly with core-acceptor separation. Careful analysis of the hole transfer kinetics reveals the extent of shell thickness inhomogeneity, however, it may be complicated by the distribution of numbers of adsorbed acceptors. All the considerations can be incorporated into a model we establish in Chapter 5for the distribution of measured hole tunneling rates. From this analysis the distribution of shell thicknesses can be extracted from the luminescence kinetic results. This approach is therefore a sensitive measure of the distribution of tunneling distances. Thus, any defects or structural irregularities that allow the hole acceptors to adsorb closer to the particle core increases the hole tunneling rate and can be detected and quantified.
A quantitative treatment of the lattice strain energy in determining the shell morphology of CdSe/CdS core/shell nanoparticles is presented in chapter 5. We use the inhomogeneity in hole tunneling rates through the shell to adsorbed hole acceptors to quantify the extent of shell thickness inhomogeneity. The results can be understood in terms of a model based on elastic continuum calculations, which indicate that the lattice strain energy depends on both core size and shell thickness. This model assumes thermodynamic equilibrium, i.e., that the shell morphology corresponds to a minimum total (lattice strain plus surface) energy. Comparison with the experimental results indicates that CdSe/CdS nanoparticles undergo an abrupt transition from smooth to rough shells when the total lattice strain energy exceeds about 27eV or the strain energy density exceeds 0.59 eV/nm2.
The predictions of this model are not followed for CdSe/CdS nanoparticles when the shell is deposited at very low temperature and therefore equilibrium is not established. The effects of lattice strain on the spectroscopy and photoluminescence quantum yields of zincblende CdSe/CdS core/shell quantum dots are examined. The quantum yields are measured as a function of core size and shell thickness. High quantum yields are achieved as long as the lattice strain energy density is below ~0.85 eV/nm2, which is considerably greater than the limiting value of 0.59 eV/nm2for thermodynamicstability of a smooth, defect free shell, as previously reported in chapter 5. Thus, core/shell quantum dots having strain energy densities between 0.59 and 0.85 eV/nm2 can have very high PL QYs, but are metastable with respect to surface defect formation. Such metastable core/shell QDs can be produced by shell deposition at comparatively low temperatures (< 140 °C).Annealing of these particles causes partial loss of core pressure, and a red shift of the spectrum.