Abstract
Copper is a common element in the environment and hence a difficult contaminant to control on silicon device manufacture lines. In p-type Si devices (e.g. solar cells), even very low copper concentrations can lead to copper-related light-induced degradation (Cu-LID), that is, degradation of the bulk minority carrier lifetime under excess carrier injection. As copper concentrations even below the detection limits of analytical methods can cause Cu-LID, it is necessary to identify its presence directly from device-level effects, which can be perturbed by the simultaneous occurrence of other light-induced degradation (LID) mechanisms. Hence, the first aim of this work is to clarify the properties of Cu-LID that enable its distinction from other LID mechanisms at the solar cell level. Thus, strong LID observed in industrial passivated emitter and rear contact (PERC) solar cells was characterized, and complemented with an analysis of the effects of corona charging on lifetime sample wafers; a method that has earlier been used in the detection of Cu-LID. The results reveal that Cu-LID in solar cells can be recognized based on its relatively fast degradation rate and laterally heavy degradation patterns on extended defects. On the other hand, Cu-LID and another LID mechanism called Sponge-LID showed mutually similar properties, and further investigations possibly involving lifetime spectroscopic methods are necessary to clarify their relationship. The second objective of the thesis is to deepen the theoretical understanding of Cu-LID. Hence, a physical model based on a theory of electrostatically limited copper precipitation was derived, which together with a previously published Schottky junction model of metal precipitates enables the modeling of Cu-LID directly at the minority carrier lifetime level. Agreement between the model and experiments was obtained in most of the investigated cases. Consequently, the different material properties and environmental conditions that affect the strength of Cu-LID were identified. Theoretically, the strength of Cu-LID is mostly affected by the Cu and the doping concentrations, and the density of heterogeneous nucleation sites, all of which influence the final precipitate size that was found to reside between few to few tens of nanometers in radius. These results provide confirmation for the precipitation theory of Cu-LID, and provide insights for mitigating its effects in silicon devices.
Translated title of the contribution | Kuparin erkautumisen aiheuttama valodegradaatio kiteisessä piissä: Mallinnus ja vaikutukset aurinkokennoihin |
---|---|
Original language | English |
Qualification | Doctor's degree |
Awarding Institution |
|
Supervisors/Advisors |
|
Publisher | |
Print ISBNs | 978-952-60-7921-9 |
Electronic ISBNs | 978-952-60-7922-6 |
Publication status | Published - 2018 |
MoE publication type | G5 Doctoral dissertation (article) |
Keywords
- silicon
- solar cell
- copper
- LID
- precipitation