Effect of electrically inactive posphorus versus electrically active phosphorus oniron gettering

In this study we investigate the efficacy of iron gettering as a function of electrically inactive phosphorus in the emitter in combination with low temperature annealing steps. To achieve different amounts of electrically inactive phosphorus in the emitter a highly doped PSG produced emitter with a large plateau depth of electrical active phosphorus is etched back stepwise by a wet-chemical procedure. Therewith we achieve a gradual reduction in electrically inactive phosphorus with small changes in electrically active phosphorus (?Rsh < 4 ?/sq). After this step, the wafers with different emitters have been annealed at 700 °C for 30 min and the content of Feiin the bulk has been measured using QSS-PC. The results show, (i) that for higher amounts of electrically inactive phosphorusa stronger iron gettering effect can be observed and (ii) that an additional annealing step leads to a significant change of Fei. This means, (i) that anelectricallyinactive phosphorus concentration dependence for iron gettering is observed and (ii) additional annealing steps, below the usual diffusion temperature of phosphorus, can be used to reduce interstitial iron in highly contaminated wafers further.