(Ga,In)(N,As,Sb) Solar Cells: N Incorporation using Novel Precursor in MOVPE Growth and Solar Cell Characteristics
This work presents a full investigation cycle of a new precursor, beginning with the synthesis of the precursors of interest. The first MOVPE investigations in different material systems like GaAs, (GaIn)As and Ga(NAsSb) were performed and finally the realization and investigation of several solar c...
Keine Tags, Fügen Sie den ersten Tag hinzu!
|Zusammenfassung:||This work presents a full investigation cycle of a new precursor, beginning with the synthesis of the precursors of interest. The first MOVPE investigations in different material systems like GaAs, (GaIn)As and Ga(NAsSb) were performed and finally the realization and investigation of several solar cell sets was achieved. First experiments with Ga(NAs) material showed an extremely high N incorporation of DTBAA in comparison with the UDMHy grown material. Consequently, experiments on (GaIn)(NAs) material followed. In contrast to UDMHy grown (GaIn)(NAs) material, no N incorporation reduction was observed for the DTBAA grown (GaIn)(NAs) layer. This led to an even higher N incorporation efficiency, 60 - 80 times higher than for UDMHy. Higher purification of the DTBAA precursor reduced the O content in the Ga(NAs) and (GaIn)(NAs) layer, however, no C incorporation reduction was observed with increasing DTBAA purity. The comparison between the C incorporation between Ga(NAs) and (GaIn)(NAs) revealed TMIn to be one main C source. Therefore, additional C incorporation investigations for Ga(NAs) and (GaIn)(NAs) were performed. A significant C incorporation reduction in DTBAA based (GaIn)(NAs) was found by either additional TBAs supply or by substituting TMIn with TIPIn. While N incorporation was reduced with additional TBAs, the TIPIn growth was not trivial, due to the fluctuation of In content. In UDMHy based (GaIn)(NAs) material, the C incorporation reduction was attempted by the usage of Sb as a surfactant. The results show no beneficial interaction between Sb treated and untreated (GaIn)(NAs) material. In fact, Sb even increased the challenges of the (GaIn)(NAs) growth due to strong N incorporation reduction with additional Sb supply. The interaction between TESb and DTBAA was investigated: it was found that the N incorporation decreased with additional TESb. However, the effect on the UDMHy based material is significantly larger. While the growth of a 1 eV Ga(NAsSb) material with MOVPE was not possible as of yet, the low negative interactions observed with the DTBAA precursor made the realization of the first 1 eV Ga(NAsSb) material grown by MOVPE possible. SIMS investigations revealed a high amount of O, which originates from the DTBAA precursor and must be dealt with and resolved. The knowledge obtained in the first part of this work was utilized for solar cell growth: five (GaIn)(NAs) solar cell sets, three of which were based on the DTBAA and two on the UDMHy precursor, were realized. The results revealed the UDMHy based material to be of higher quality. The C incorporation reduction led only to small solar cell quality improvement, which is probably related to the high O content that drastically reduces the material quality. The simulation of the best EQE measurements showed an active (GaIn)(NAs) thickness of around 400 nm for the UDMHy based material. DTBAA based material, however, revealed an active (GaIn)(NAs) material of only 180nm. Among with the active doping densities obtained from test structures, the depletion region was calculated, and an indication for the diffusion lengths was found. For the UDMHy based material, the diffusion length was estimated to be around 300 nm while the DTBAA based material was estimated to have a diffusion length of around only 60 nm. The poor solar cell characteristics of DTBAA grown material is related to the high O content, which deteriorates the solar cell quality drastically. Additionally, TIPIn was found to increase the open circuit voltage of the DTBAA as well as UDMHy based (GaIn)(NAs) solar cell. However, the current of the same solar cell material was significantly lower than the TMIn grown pendant. This, however, is probably connected to the In incorporation fluctuation during the growth. The results summarized above show that the DTBAA precursor has a huge potential to improve the (GaIn)(NAs) material growth and therefore to replace the conventional N precursor UDMHy. However, the DTBAA purity must be improved. Furthermore, TIPIn was found to reduce C incorporation in comparison with TMIn. However, the In incorporation fluctuation when using TIPIn is an issue that must be resolved. As the Sb-N interaction is drastically lower for DTBAA based dilute N materials, (GaIn)(NAs):Sb growth with DTBAA needs to be investigated in terms of potential C incorporation reduction. Additionally, Ga(NAsSb) as well as (GaIn)(NAsSb) solar cells need to be grown in future experiments with highly purified DTBAA, as this kind of solar cells grown with MBE showed the highest solar cell qualities observed so far.|