Thermal ideality factor of hydrogenated amorphous silicon p-i-n solar cells

Abstract

The performance of hydrogenated amorphous silicon (a-Si:H) p-i-n solar cells is limited, as they contain a relatively high concentration of defects. The dark current voltage (JV) characteristics at low forward voltages of these devices are dominated by recombination processes. The recombination rate depends on the concentration of active recombination centers and the recombination efficacy of each of these centers. The first factor causes the ideality factor of the devices to be non-integer and to vary with voltage. The temperature dependence of the dark current can be expressed by its activation energy. For microcrystalline silicon solar cells the activation energy varies with voltage with a so-called thermal ideality factor of 2. This value was derived for devices with a spatially uniform defect distribution and reflects the recombination efficacy. Here we present results of a thickness series of a-Si:H p-i-n solar cells. We have matched the experimental curves with computer simulations, and show that the voltage-dependent ideality factor curve can be used to extract information on the cross sections for electron and hole capture. Also, the activation energy is used as a measure for the mobility gap, resulting in a mobility gap for a-Si:H of 1.69 eV. We find a thermal ideality factor close to 2 for all samples. This is explained with a theoretical derivation, followed by a comparison between the internal electric field strength and the spatial variation of the defect density in the intrinsic layer. The thermal ideality factor is shown to be insensitive to the defect distribution and the recombination profile in the device. It is, therefore, an appropriate parameter to characterize a-Si:H p-i-n devices, providing direct insight on the recombination efficacy.

Recombination efficacy in a-Si:H p-i-n devices

“…The error margin for m(V) was obtained by analysis of propagation of the measurement inaccuracies, and we found it to be about 3%. As shown by Kind et al [9], m(V) increases with voltage. They also observed that the m(V) curves are slightly dependent on the i-layer thickness.…”

“…Apparently, in this region the activation energy is independent of the i-layer thickness and the presence of a buffer layer. Kind et al [9] showed that the thermal ideality factor for this series was 2.09 ± 0.04, close to the value of 2. This value for the activation energy implies that the recombination efficacy peaks at the position where R σ n = p.…”

“…This equation was derived under the following assumptions: (i) abrupt mobility edges; (ii) a uniform electric field in the device; (iii) the quasi-Fermi levels are constant in the device with respect to the potentials at the contacts; (iv) the separation between the quasi-Fermi levels is equal to the applied voltage multiplied by the elementary charge q; (v) the temperature dependence of the concentration of active recombination centers is small compared to that of the efficacy; and (vi) in the region where R σ n = p the separations between the quasi-Fermi levels and their respective band edges are approximately equal. When m th = 2 the applied voltage results in equal shifts of the quasi-Fermi levels for electrons and holes at the position where the recombination efficacy of the SRH process is maximal, i.e., where R σ n = p. The factor of 2 was found for microcrystalline silicon p-i-n solar cells [8] as well as for a-Si:H p-i-n solar cells [9]. The latter is quite surprising, considering Eq.…”

“…(4) was derived assuming a spatially homogeneous density of states distribution. However, Kind et al [9] showed that Eq. (4) also holds for a-Si:H solar cells having a spatially inhomogeneous density of states distribution, provided the internal electric field strength is strong.…”

Impact of implementing the Meyer-Neldel behavior of carrier emission pre-factors in solar cell and optical detector modeling

“…Samples were deposited in a multi-chamber RF-PECVD facility under identical conditions and characterized with appropriate equipment at Delft University of Technology. 19 Each device structure is as follows: TCO/p-aSiC:H/i-a-Si:H/n-a-Si:H/Al. The front contact is an Asahi U-(SnO 2 :F) type substrate with a textured surface.…”