On effective surface recombination parameters

Abstract

This paper examines two effective surface recombination parameters: the effective surface recombination velocity S eff and the surface saturation current density J 0s. The dependence of S eff and J 0s on surface charge Q, surface dopant concentration N s , and interface parameters is derived. It is shown that for crystalline silicon at 300 K in low-injection, S eff is independent of N s only when Q 2 /N s < 1900 cm in accumulation and Q 2 /N s < 1600 cm in depletion; otherwise S eff increases with N s. These conditions are rarely satisfied in undiffused wafers but sometimes satisfied in heavily diffused wafers when coated with lowly charged films. Under the same conditions, J 0s is independent of N s when Q 2 /N s > 1.5 Â 10 7 cm for accumulation and Q 1.85 /N s > 1.5 Â 10 6 cm for inversion. These conditions are commonly satisfied in undiffused wafers but rarely in diffused wafers. We conclude that for undiffused silicon, J 0s is superior to the conventional S eff as a metric for quantifying the surface passivation, whereas for diffused silicon, the merit in using J 0s or S eff (or neither) depends on the sample. Experimental examples are given that illustrate the merits and flaws of J 0s and S eff .

Degradation of Surface Passivation on Crystalline Silicon and Its Impact on Light-Induced Degradation Experiments

“…1, bottom). Comparing different fit ranges between Δn = 5·10 15 and 1.5·10 16 cm −3 results in an estimated uncertainty of the absolute value of J 0s of ∼ 20%, increasing up to ∼ 50% when J 0s < 5 fA/cm 2 . However, as will be shown later, relative changes of J 0s during sample treatment exceed this uncertainty by far, making relative changes in J 0s well visible.…”

“…We use this approach on samples without emitter, and a linear fit of 1/τ corr , therefore, yields the surface saturation current density J 0s [16] as shown in two examples in Fig. 1.…”

“…The range of Δn for the fit of J 0s is set from 8·10 15 to 1.5·10 16 cm −3 with the exception of one 200 Ω·cm sample where Δn ranges from 5·10 15 to 8·10 15 cm −3 due to a limited injection range of PCD measurement data. The upper limit makes sure that measurement artifacts at the beginning of a PCD measurement (equaling highest Δn) do not influence the J 0s fit while the overall restriction to measurement data to rather high Δn further reduces influences of changes in τ b on the determination of J 0s .…”

“…First, it will be shown that a time-and injectionresolved visualization of lifetime data and a time-resolved calculation of the surface saturation current density J 0s [16] are useful tools for assessing the degree of degradation in surface passivation, even in the case of changes of τ b during an ongoing treatment. This method is then applied to compare different passivation layers with regard to severity and timescale of surface-related degradation.…”

“Zero-charge” SiO2/Al2O3 stacks for the simultaneous passivation of n+ and p+ doped silicon surfaces by atomic layer deposition

“…Using the device simulation package Atlas [51], the J 0n þ values belonging to Gaussian-shaped n þ doping profiles having different surface doping concentrations and a fixed depth of 0.1 mm were calculated. Note that the commonly reported effective surface recombination velocity S eff typically varies with surface doping concentration [52]. However, S eff was not required as input parameter for the Atlas simulations.…”