Direct monitoring of minority carrier density during light induced degradation in Czochralski silicon by photoluminescence imaging

Abstract

In this paper, we present a new method for studying the light induced degradation process, in which the minority carrier density is monitored directly during light soaking by photoluminescence imaging. We show experimentally that above a certain minority carrier concentration limit, Δnlim, the boron oxygen (B-O) defect generation rate is fully independent of the injected carrier concentration. By simulation, we determine Δnlim for a range of p-type Czochralski silicon samples with different boron concentrations. The normalized defect concentrations, Nt*, are determined for the same samples by time-resolved Quasi Steady State Photoconductance measurements. After 10 min of light degradation, no correlation between Δnlim, and Nt* is observed. These results indicate that the role of the excess carriers during the rapid decay is to first change the charge state of the defects by shifting the electron quasi-Fermi level across the energy level of the defect centre in its passive state (Elat = EV + (635 ± 18) meV) and that, subsequently, another rate-determining step proceeds before the defect centre becomes recombination active.

Studying Light-Induced Degradation by Lifetime Decay Analysis: Excellent Fit to Solution of Simple Second-Order Rate Equation

“…This cannot be the rate determining step [4], but it also means that none of the earlier steps can be rate determining since the reaction does not start until the excess carriers are present.…”

“…Step 1 (6) can under any circumstances not be rate-determining since the rate of the defect reaction is independent of Δn [4].…”

“…For the fast defect the reaction steps 1, 2 and 3, described in (2)-(4) can all be excluded as rate determining steps because of step 3; In this step the presence of excess electrons cause a shift of the electron quasi-Fermi level across the energy level of the defect center in its passive state [4], [10]. This cannot be the rate determining step [4], but it also means that none of the earlier steps can be rate determining since the reaction does not start until the excess carriers are present.…”

“…where t is time, τ 0 is the initial carrier lifetime before LID and τ(t) is the carrier lifetime after time t. Upon prolonged illumination at a sufficiently high carrier injection level [4], N t * approaches a saturated value N t * sat that has been found to be proportional to the boron concentration and the oxygen concentration squared ( , * ∝ [ ][ ] 2 ) [5]. A model by Schmidt et al [5], where fast diffusing oxygen dimers capture substitutional boron atoms to form B s -O 2i recombination centers, was the first complete model that took this into account and for years this was the prevailing model.…”

Preventing light-induced degradation in multicrystalline silicon

“…19 Naerland et al 28 have also shown that increasing the illumination intensity increases the degraded area due to excess carrier diffusion. Figure 1 presents the high-injection level lifetime as a function of illumination time at intensities 1 and 18 W/cm 2 in an oxidized top mc-Si sample with positive corona charge (þ4 lC/cm 2 ), while Figure 2 shows the corresponding medium injection lifetime maps before and after illumination.…”