Search

37th European Photovoltaic Solar Energy Conference and Exhibition; 474-477, Half-cell modules are a preferred technological option to produce high efficiency photovoltaic modules. In this work, we present results on different half-cell processes applied to state-of-the-art heterojunction solar cells. We show that for heterojunction solar cells, a well optimized cell separation process is necessary for reduced losses at the half-cell edge. The cell separation leads to additional depletion region recombination currents, which can be imaged and quantified by dark lock-in thermography (DLIT). To this end, we present a ‘Local I-V’ analysis that quantifies local electrical losses for different half-cell processes.

This book discusses lock-in thermography (LIT) as a dynamic variant of the widely known IR thermography. It focuses on applications to electronic devices and materials, but also includes chapters addressing non-destructive evaluation. Periodically modulating heat sources allows a much-improved signal-to-noise ratio (up to 1000x) and a far better lateral resolution compared to steady-state thermography. Reviewing various experimental approaches to LIT, particularly the commercial LIT systems available, this 3rd edition introduces new LIT applications, such as illuminated LIT applied to solar cells, non-thermal LIT lifetime mapping and LIT application to spin caloritronics problems. Numerous LIT investigation case studies are also included.

Potential-induced degradation (PID) can cause massive power losses in photovoltaic power plants with serially connected modules when high electrical potential between solar cells and the module front glass surface leads to shunting of Si solar cells. The shunting is reversible to a large extend through elevated temperatures and/or by application of reversed voltages. It was previously shown that PID-related shunting (PID-s) of the p-n junction occurs due to Na-decorated stacking faults in {111} planes. Defects at former PID-shunt positions that have been thermally 'healed' (recovered) do not show a Na decoration anymore. In this contribution we investigate stacking faults with Na decoration and with vanished Na decoration after thermal recovery. HAADF STEM imaging is used for principal investigation of the atomic structure of PID-s defects. The Na decorated stacking fault appears as a dark region in the HAADF STEM image. By contrast, the atomic structure of the recovered stacking fault can be clearly imaged. The orientation of Si dumbbells indicates an extrinsic stacking fault that remains after recovery by Na out-diffusion.

Multicrystalline solar cells break down strongly at reverse voltages well below the theoretical limit. Previous explanations were based on assuming a constant depth of the junction below the surface. In this work, preferred phosphorous diffusion at special line defects in grain boundaries is shown to lead to spikes in the p-n junctions even below flat surfaces. The curvature radii of the spherical p-n junction bending are measured by electron beam-induced current to be in the range of 300-500nm, leading to the observed type III avalanche breakdown voltages.

28th European Photovoltaic Solar Energy Conference and Exhibition; 1538-1544, The local efficiency parameters of readily processed solar cells can be investigated both by applying dark lock-in thermography (DLIT) and Solar Cell Local characterization (CELLO), which is a special bias-dependent lock-in variant of light beam-induced current (LBIC) imaging. While the DLIT investigation directly leads to local two-diode data, until now results of CELLO investigations were displayed in a different way. The main purpose of this work is to develop a transformation of CELLO-based maps into local dark power and current density maps, which allow a direct comparison of the two approaches for local efficiency analysis. The comparison on one and the same cell shows a very good qualitative correspondence. Both DLIT and CELLO detect basically the same local inhomogeneities in J01 and J02. However, there are still differences in the quantitative values. The origin of these differences is discussed.

28th European Photovoltaic Solar Energy Conference and Exhibition; 1659-1664, The surface recombination velocity at grain boundaries can be evaluated by means of Dark Lock-in Thermography (DLIT) or Light Beam Induced Current measurements (LBIC). In this contribution for a first time we apply both methods to the same grain boundaries and compare the results to each other. In spite of the quite similar models, which are the basis of both evaluations, there are quite strong differences in the required interpretation of both approaches. A detailed discussion presents the main issues in comparing recombination parameters of grain boundaries obtained from the different measurements and exposes that the used model represents the LBIC-based method better, but the DLIT-based evaluation proposes a better estimation of the influence of the GB on the solar cell performance. This is possible since DLIT provides a direct measurement of the dark current density caused by the recombination at the grain boundary, whereas the LBIC-based method is mainly sensitive to the recombination in the bulk. For this reason both methods may give similar results under strict limitations, but in some cases the DLIT method cannot be used to calculate a surface recombination velocity at the GB as it is defined in the underlying model.

It had been shown already earlier by X-ray microanalysis that, in positions of defect-induced junction breakdown in industrial multicrystalline (mc) silicon solar cells, iron-containing precipitates may exist. However, the nature of these precipitates was unknown so far. Here, in such positions, scanning transmission electron microscopy was performed after defect-controlled focused ion beam preparation. First of all, the defect site was localized by microscopic reverse-bias electroluminescence imaging. The high accuracy of following FIB target preparation (