Doped ZnS Thin Films for Intermediate Band Solar Cells - Deposition and Characterization
MetadataVis full innførsel
- Institutt for fysikk 
Intermediate band solar cells (IBSCs) has the potential of increasing the efficiency and reducing the cost of commercial solar cells. The realization of such cells relies on the development of intermediate band (IB) materials. The IB materials are semiconductors with one or more energy level in the band gap. Introducing additional IBs can increase photon absorption and improve the utilization of the absorbed energy. In addition, a p-type emitter with an appropriate band gap, low sub-band gap absorption, high carrier concentration and good change transport is needed to prevent significant energy losses in the IBSCs. The first goal was to investigate if spectroscopic ellipsometry (SE) could be used in the characterization of doped materials for IBSCs, such as the crystal structure and doping level. Particularly has the dielectric function (DF) of chromium doped zinc sulfide (Cr:ZnS) films deposited with PLD been studied in detail. The second goal of this work was to deposit titanium, copper, and indium doped ZnS films with room temperature thermal evaporation, to measure their compositional, optical and electrical properties and to discuss their promise as an IB material or as a $p$-type emitter in ZnS based IBSCs. The DF of zinc blende phase (Zb) ZnS and wurtzite phase (Wz) ZnS from literature were used to simulate the ordinary and extraordinary components of the DF for different phase volume fractions. The results suggested that the SE measurement of the ZnS films will be almost insensitive to the extraordinary component of the DF. When studying Cr:ZnS films with SE, it was also found that it is very challenging to determine if the changes in the DF is induced by changes in crystal structure, grain size, defects or additional states related to the dopant. The analysis is further complicated by the limited availability of reliable reference DFs for the Zb and Wz phases. A new oscillator model for undoped ZnS was developed, and the model was found to be in reasonable agreement with the corresponding dispersion models for ZnS found in literature. Furthermore, a Cr:ZnS oscillator model was developed from the ZnS model for undoped films and used to analyze the Cr:ZnS films deposited with PLD. The Cr:ZnS films were then analyzed using critical point analysis. Finally, an alternative oscillator model was proposed for modeling Cr:ZnS films, by the assumption that the ZnS-related transition changes does not change significantly compared to the ZnS model. The alternative model was found to give a more reasonable description of the changes seen in the DF due to doping, particularly around the band gap. By employing the models on the SE data from the Cr:ZnS films, it was found that the band gap transition (E_0) was broadened, increased in amplitude and shifted to lower energy for increased doping. Furthermore, the amplitude of the E1 peak was decreased and the sub-band gap absorption was increased. %In summary, it is challenging to use oscillator models for such complex DFs, due to the parameter correlations. Ti, Cu and In doped ZnS films were deposited with thermal evaporation. The energy-dispersive spectroscopy (EDS) results indicated substitutional doping only at low doping levels. For strongly doped samples the optical properties were severely degraded, particularly for the Cu doped samples. The sub-band gap absorption was around 0.1E5/cm for all the dopants, but Ti:ZnS and Cu:ZnS appeared to have strong absorption states immediately below the band gap. The resistivity of the Cu doped samples reduced exponentially from 5E-2 Ohm-cm for 26% to 4.0E-4 Ohm-cm for 34% doping. The optical measurements, however, indicated that the structure was severely degraded at such high doping levels and one must expect composite materials containing Cu clusters and sulfides must be expected. The minimum resistivity of the In doped samples were 0.75 Ohm-cm at 14.6% doping. The optical properties indicated that the film structure was well maintained for doping levels up to at least 5%.