J.W. Metselaar

Optical modeling of a-Si:H solar cells with rough interfaces: Effect of back contact and interface roughness

An approach to study the optical behavior of hydrogenated amorphous silicon solar cells with rough interfaces using computer modeling is presented. In this approach the descriptive haze parameters of a light scattering interface are related to the root mean square roughness of the interface. Using this approach we investigated the effect of front window contact roughness and back contact material on the optical properties of a single junction a-Si:H superstrate solar cell. The simulation results for a-Si:H solar cells with SnO2:F as a front contact and ideal Ag, ZnO/Ag, and Al/Ag as a back contact are shown. For cells with an absorber layer thickness of 150–600 nm the simulations demonstrate that the gain in photogenerated current density due to the use of a textured superstrate is around 2.3 mA cm−2 in comparison to solar cells with flat interfaces. The effect of the front and back contact roughness on the external quantum efficiency (QE) of the solar cell for different parts of the light spectrum was de...

Computer modelling of current matching in a-Si : H/a-Si : H tandem solar cells on textured TCO substrates

Abstract Computer modelling is used as a tool for optimising a-Si : H/a-Si : H tandem cells on textured substrate in order to achieve current matching between the top and bottom cell. To take light scattering at the textured interfaces of the cell into account, we developed a multirough-interface optical model which was used for calculating the absorption profiles in the tandem cells. In order to simulate multi-junction solar cell as a complete device we implemented a novel model for tunnel/recombination junction (TRJ), which combines the trap-assisted tunnelling and enhanced carrier transport in the high-field region of the TRJ. We investigated the influence of light scattering and thickness of the intrinsic layer of the bottom cell on the optimal ratio i2/i1 between the thicknesses of the bottom (i2) and top (i1) intrinsic layers in the current-matched cell. The simulation results show that increasing amount of scattering at the textured interfaces leads to a lower ratio i2/i1 in the current-matched cell. This ratio depends on the thickness of the intrinsic layer of the bottom cell. The simulation results demonstrate that a-Si : H/a-Si : H tandem cell with 300 nm thick intrinsic layer in the bottom cell exhibits higher efficiency than the cell with 500 nm thick bottom intrinsic layer.

Accurate generation rate profiles in a-Si :H solar cells with textured TCO substrates

Abstract With a computer program (GENPRO2D) based on incoherent multi-interface scattering, we are able to calculate the carrier generation profile (light absorption profile) in an a-Si:H solar cell with several rough interfaces, provided that the light scattering at every interface is known and the spatial distribution of the incoming light flux is known. In this paper, we show for the first time the absorption profile in a-Si :H solar cells grown on rough TCO.

Extraction of amorphous silicon solar cell parameters by inverse modelling

Abstract A novel and unique numerical method of extraction of physical parameters from the measured characteristics was applied for the first time to single junction p-i-n amorphous silicon solar cells to determine several important input parameters used for their modelling. A set of realistic parameters, which describe the solar cells, has been determined from the fits of simulated behaviour to the measured one. The single junction p-i-n solar cells were deposited at the Utrecht University. A set of input parameters that closely describes the solar cells behaviour is an important step for their further optimisation.

Device Modeling of a-Si:H Alloy Solar Cells: Calibration Procedure for Determination of Model Input Parameters

A calibration procedure for determining the model input parameters of standard a-Si:H layers, which comprise a single junction a-Si:H solar cell, is presented. The calibration procedure consists of: i) deposition of the separate layers, ii) measurement of the material properties, iii) fitting the model parameters to match the measured properties, iv) simulation of test devices and comparison with experimental results. The inverse modeling procedure was used to extract values of the most influential model parameters by fitting the simulated material properties to the measured ones. In case of doped layers the extracted values of the characteristic energies of exponentially decaying tail states are much higher than the values reported in literature. Using the extracted values of model parameters a good agreement between the measured and calculated characteristics of a reference solar cell was reached. The presented procedure could not solve directly an important issue concerning a value of the mobility gap in a-Si:H alloys.

Optical modeling of a-Si:H based solar cells on textured substrates

The authors have developed a new GENPRO2 computer program package to meet the needs of optical modeling of a-Si:H based solar cells on textured substrates. The new program takes into account the light scattering effects at the rough interfaces, allowing the accurate optical modeling of a-Si:H based solar cells with textured substrates. The authors present the state-of-the-art of optical modeling and their recent results on textured solar cells with different structures, such as: glass/TCO/p-i-n(a-Si:H)/Ag, ITO/p-i-n(a-Si:H)/Ag/SS, tandem cells and cells with TCO/Ag back contacts.

Photoresist stripping in afterglow of Ar‐O2 microwave plasma

The experimental investigation of the photoresist etching rate in the afterglow of Ar‐O2 microwave plasma has been performed. It was found that up to 50% Ar can be added to O2 without etching rate decrease. A high photoresist etching rate of 2.9 μm/min has been achieved. A mathematical model based on the one‐dimensional mass continuity equation for atomic oxygen, taken to be responsible for the etching process, has been proposed. Modeling of the Ar‐O2 afterglow microwave photoresist stripping process was performed. The simulated behavior of the atomic oxygen flux shows a similar dependence with Ar dilution and pressure as the experimentally observed photoresist etching rate. It indicates that the photoresist etching rate is determined by the atomic oxygen flux to the substrate. The proposed model explains the measured dependencies.

Computer modeling of amorphous silicon tandem cells

We have used a new modeling approach for simulation of a-Si:H tandem cells. The trap assisted tunneling model and an enhanced effective extended state mobility were used to model the recombination and transport in the high field region of the tunnel recombination junction. With this approach, realistic tandem cell characteristics could be obtained. We illustrate how modeling can be used for the analysis of multi-junction cells.

Plasma deposition of hydrogenated amorphous silicon: Effect of rf power

The experimentally determined growth rate power dependence of plasma‐enhanced chemical‐vapor‐deposited amorphous silicon was compared with the predicted dependence for the case when the SiH3 radicals are the dominant contributors to film growth. The higher experimental growth rate compared to its calculated value can be explained by the influence of Si2H6 produced in the discharge. The maximum of the growth rate was found to be around 50 mW/cm3 . The decrease of the growth rate at powers exceeding 50 mW/cm3 can be explained by the increased influence of hydrogen, which is consistent with the measurements of the refractive index of the deposited films.

Defect re-distribution in amorphous silicon below equilibration temperature

Abstract The effect of a Fermi level shift in an intrinsic energy distribution of gap states of amorphous silicon prepared by glow discharge is investigated with the aim of simulating the energy distribution of gap states near the i–n (p–i) interface in a p–i–n (n–i–p) device. Therefore we carried out experiments in which the Fermi level is moved to either the conduction or valence band edge in a ‘programmed’ intrinsic energy distribution of gap states by subjecting a metal/insulator/amorphous-silicon structure to n-type or p-type bias stress, respectively. Its effect on the energy distribution of gap states is measured by the charge version of deep level transient spectrometry. We observe that upon n-type (p-type) bias stress the energy distribution of gap states does not immediately adjust to the applied Fermi level shift, but that first an intermediate distribution is formed with a larger neutral dangling bond state contribution. In addition, it appears that the negatively charged dangling bond states are more resistant to p-type stress than the positively charged dangling bond to n-type stress.