As described in Section 2.1, hydrogenation by FGA after
SiNx layer capping is performed on p-type and n-type
doped CSPCs to reintroduce the hydrogen that effused after high
temperature annealing. Figure 4 shows the comparison in passivation of
p-type and n-type doped poly-SiOx symmetric samples
after thermal annealing and after hydrogenation. The optimum thermal
oxidation and annealing conditions, as described in Figure 3, have been
chosen for each type of CSPC. We observe that p-type and n-type doped
poly-SiOx CSPCs applied on DSP and DST symmetric
samples, respectively, gave the same iVoc of 690 mV
after high-temperature annealing which improved to 710 mV after
hydrogenation. Using the 2-step annealing technique, the symmetric
p-type doped poly-SiOx applied on DST wafer exhibited an
iVoc of 687 mV after hydrogenation. Applying the same
2-step annealing technique to symmetric n-type doped
poly-SiOx on DST wafer (including the thermally grown
tunnelling SiOx prepared at 675 °C for 3 minutes as in
the p-type case), an iVoc of 690 mV was found after
hydrogenation, resulting in lower passivation quality than the single
step annealing case. Here, as the intrinsic poly-Si layer resulting from
the first annealing got denser [73], we speculate that the
phosphorus doping atoms do not easily reach the tunnelling
SiOx/c-Si bulk interface to establish an effective
electric field. In addition, as shown in Figure 3(b), the tunnelling
SiOx prepared at 675 °C for 3 minutes is not the best
condition for the n-type doped poly-SiOx on a textured
surface. Still, this case is investigated (and later put forward in
solar cell fabrication) to realize a neat flow chart in which both
n-type and p-type doped poly-SiOx layers essentially
undergo the same thermal processes at the same time.