Gettering and defect engineering techniques for low cost and high efficiency industrial crystalline silicon solar cells
- Autores: Ana Peral Boiza
- Directores de la Tesis: Carlos del Cañizo Nadal (dir. tes.)
- Lectura: En la Universidad Politécnica de Madrid ( España ) en 2017
- Idioma: español
- Tribunal Calificador de la Tesis: Juan Carlos Jimeno Cuesta (presid.), David Fuertes Marrón (secret.), Laura Mendez Gimenez (voc.), Jean Francois Lelievre (voc.), Ignacio Tobias Galicia (voc.)
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
Defects in silicon crystalline materials determine the efficiency of the industrial silicon solar cells, which account by far as the largest share in the photovoltaic market. This thesis aim is to identify and manipulate these defects in order to reduce losses for increasing efficiency and reducing costs of silicon solar cells. Techniques to optimize the existing steps of the solar cell fabrication process are proposed, for reducing SRH extrinsic recombination, improving performance in the emitter region and in the bulk of the wafer, while maintaining the optimum solar cell parameters.
Solutions for reducing recombination in the emitter have been studied first. Through simulations and experiments heretofore under-appreciated parameters as the inactive phosphorus concentration have been demonstrated to strongly affect the final solar cell performance. Strategies for reducing the effect of inactive phosphorus concentration in the emitter have been investigated.
Low temperature annealing step has been demonstrated to have an effect on total phosphorus distribution and chemical state. The behavior of electrically-active phosphorus has been widely studied and its variation during low temperatures processes is negligible. However, a modification of total phosphorus distribution is observed during a low temperature annealing step, observing an increase of total phosphorus concentration in the superficial region. It exists a correlation of dose increase with initial phosphorus dose present. It has been calculated that electrically-active P is reduced during annealing and that the increase in total P occupy electrically-inactive positions. In addition, micro-characterization of the total phosphorus using Atom-Probe Tomography (APT) experiments have been run to investigate phosphorus precipitates, obtaining a direct measurement of precipitates with higher size after low temperature annealing step in agreement with Glow Discharge Optical Emission Spectroscopy (GD-OES) total phosphorus concentration measurements.
Furthermore, the effect of this change of total phosphorus during low temperature annealing step on SRH recombination in the emitter layer has been studied. For short-term temperature treatments well below the POCl3 diffusion temperature, a reduction of emitter saturation current density j0e up to -60 fA/cm2 has been achieved without driving the emitter further into the silicon substrate.
This technique provides a suitable method for differentiating managing emitter recombination of other emitter parameters.
In addition, defect engineering strategies for decreasing bulk recombination have been developed focusing on its application during two steps of the solar cell fabrication process, the emitter formation step and the contact formation step.
The design of an emitter formation step that allows the reduction of SRH recombination should take into account two parameters: initial wafer impurity content and emitter phosphorus compounds formation. Efficacy of different gettering strategies for reducing bulk impurities have been studied, varying separately both parameters, first the bulk impurity content and second the emitter phosphorus distribution. On this way, a study about the efficacy of different gettering strategies has been presented first, in order to reduce bulk impurities using different bulk materials, in terms of their initial impurity content, coming from low-cost solar grade silicon ingots. Using a variation of the usual emitter formation process has shown beneficial effects in segregating iron impurities for wafers with initial distribution and concentration of iron inside a specific range, e.g. materials with high concentration of big Fe precipitates; while for other cases, standard process is efficient enough . An analysis based on the comparison of measured lifetime and dissolved iron concentration with theoretical calculations helps to infer the initial iron distribution and concentration into the as-grown wafers and, according to that, choose the more effective type of gettering to reach higher final lifetime. A second study has been presented, showing that the efficacy of gettering step vary as a function of the emitter phosphorus content, concretely the electrically-inactive phosphorus.
Electrically-active phosphorus is traditionally considered in the PV community as the cause of metal gettering. This thesis contributes to consider the different compounds formed in the emitter, not only the electrically-active phosphorus, and demonstrate their effectiveness in segregating bulk impurities. Relation between electrically-inactive phosphorus concentration and gettering of detrimental impurities from the bulk of the wafer is experimentally proven. Extra gettering effect has been measured after the annealing with the presence of the PSG layer.
Finally, the effects of contact formation step degrading the impurity gettering benefits obtained during emitter formation are studied by means of both simulations and experiments in an industrial belt furnace. The defect engineering tool, named textitextended co-firing step is proposed and verified as a means to maintain the gettering effects of previous fabrication steps or further gettering. Detailed experiments on solar cell level and micro-characterization are implemented to evaluate the limits and the impact of the novel technique in industrial processing, so that the co-firing process can be extended without impacting solar cell performance while obtaining the optimal contact quality. To probe contact formation after extended co-firing step, specific contact resistance measurements using TLM as macro characterization technique and SEM as micro characterization technique were used, particularly on FIB cross sectional samples.
