Imaging Characterization of Solar Cells
Imaging methods are fast and highly versatile tools for solar cell characterization with spatial resolution in the micrometer range. In our labs the distribution of luminescence radiation emitted from the solar cell under electrical excitation (= Electro-Luminescence Imaging, ELI) and/or under illumination (= Photo-Luminescence Imaging, PLI) is detected either by a Si-CCD camera in the UV-vis or by a InGaAs-CMOS camera in the NIR spectral range. If the detection is shifted further into the infrared part of the electromagnetic spectrum, information about device heat emission - similarly to black body radiation - can be gained. This is employed in normal or - with higher sensitivity - in lock-in thermography, where in the latter case periodic excitation of the sample, either by application of a voltage (Dark Lock-In Thermography, DLIT) or by illumination (Illuminated Lock-In Thermography, ILIT). Another method for studying solar cells in operation is Light-Beam Induced Current (LBIC), where the solar cell is locally excited by a focussed laser beam and the resulting local short-circuit current is measured. Although a scanning and not an imaging method by definition, it allows to map the distribution of the photocurrent at high spatial as well as electric resolution. Figure 1 depicts the complementary information obtained by these imaging methods on one and the same sample.
Figure 1: Imaging methods yield complementary information when applied to a polymer solar cell. ELI: high intensities indicate large injection currents; PLI: low intensities indicate thin layers or stronger excitation quenching; LBIC: large photocurrents are confirmed for regions of low thickness and large injection currents; DLIT: high intensities display heat emission and large injection currents.
In general these imaging methods are highly useful for in-line process or post-production quality control, e.g. in the fabrication of polymer solar modules.
Figure 2 demonstrates the high lateral resolution as well as dynamic range of the obtained LBIC image measured on a partly degraded solar cell, where part of the back-contact was oxidized and thus hampered charge extraction from the photoactive layer by large extent. In addition several lateral inhomogeneities are clearly visible, leading to a variation in the monochromatic (@ 445 nm), local external quantum efficiency (EQE).
Figure 2: LBIC image of a partly degraded solar cell. Variations in the monochromatic external quantum efficiency are clearly visible, and the smallest signals can be assigned to regions where the back-contact is oxidized - thereby hampering charge extraction by large extend.
When going from simple single solar cells to solar modules, the serial connection of solar cells within the solar module generally leads to some complications. This is demonstrated by LBIC imaging of a simple polymer solar module consisting of 3 serially connected solar cells (compare with Figure 3). In the first case (left) the solar module is depicted as obtained in simple LBIC scanning mode. In the second case (right) the same solar module was measured under background illumination, ensuring a background current that carries away also the local photocurent generated by the laser beam. The complication that arises for this technique is to separate the local photocurrent (=alternating current, AC) from the background photocurrent (=direct current, DC), which is done by application of the lock-in technique in combination with a periodic laser power variation by a function generator.
Figure 3: LBIC images of one and the same polymer solar module: obtained under constant laser light intensity (left) and under periodic variation of the laser light in combination with a background illumination (right). Clearly the left image, obtained without background illumination, contains artifacts, whereas in the other case a lock-in technique is required for proper sensing the AC-signal.
Furthermore imaging characterization can be highly useful for contributing to undestand degradation phenomena (→Stability). Therefore we are often combining accelerated ageing experiments for studying the degradation behaviour of polymer solar cells with these imaging characterization techniques. Comparing e.g. luminescence images obtained by different excitation mechanisms, i.e. electrical and optical excitation, allows to discern between degradation at the electrical contact layers (by electroluminescence imaging, ELI) and within the active layer (by photoluminescence, PLI). Already a qualitative comparison of measurements obtained by the imaging techniques mentioned above allows distinguishing between different degradation pathways and gives useful information on the local device behaviour (Figure 4).
Figure 4: Combination of information obtained by LBIC (a), ELI (b), PLI (c) and forward (d) and reverse (e) DLIT imaging on a solar cell prepared by the NREL OPV group after degradation. LBIC clearly shows the photovoltaically active regions under operation, while ELI depicts a similar behavior under reverse operation, respectively. PLI allows better discrimination between photoactive layer degradation through pinholes (black circular regions) and electrode delamination (darker grey regions). Finally, DLIT forward and reverse operations point out 4–5 local shunts, differently active in both bias directions. Taken from Ref. .
For solar cell modules also illuminated lock-in thermography (ILIT) can be applied to investigate the overall device performance. It exhibits the great advantage that in principal no electrical contacts are necessary. There are two modi posibble. ILIT under open circuit conditions and under short circuit conditions. While under open circuit all current will run through shunts - if there are any at all, the current recombines through the outer circuit under short circuit and only "heavy" shunts will highlight. Furthemore delaminations within the layer stack are easily recognised, since they alter the heat collection and emission to a large extend.
Figure 5: ILIT image of a commercial organic solar cell module under open circuit condition. Small red dots represent shunts, while larger red areas are delaminations of the substrate foil.
A further scope of our work is the quantitative analysis of luminescence images by development of suitable discrete models as well as image processing procedures (compare with Figure 6). An electric network simulation called Micro-Diode-Model (MDM) has been developed to extract the pristine (non-falsified) current-voltage characteristics of the active layer from experimental light emission measurements. Thereby the finite conductance of the electrodes (e.g. specifically that of ITO) is explicitely considered and helps to resolve the current and voltage distributions across the solar cell.
Figure 6: Left: Distribution of electroluminescence intensity in a laterally homogeneous polymer solar cell. Right: Extraction of the local current and junction voltage allows determination of the transparent electrodes sheet-resistance and extraction of the pristine (non-falsified) current-voltage characteristics of the active layer. Taken from Ref. .
The above evaluation method is restricted to laterally homogeneous solar cells, however, we also work on the quantitative evaluation of electroluminescence images of laterally inhomogeneous solar cells by decoupling local equivalent circuit parameters inside respective numerical frameworks (Figures 7-9).
Figure 7: Electroluminescence image (a) of a polymer solar cell and calculated parameter images: b) voltage, c) series resistance and d) saturation current-density. Taken from Ref. .
Figure 8: Current-density, saturation current-density and series resistance image calculated from ELI measurements (right) of a microcrystalline solar cell.
Figure 9: Calculated distribution of the saturation current-distribution (left) and the series resistance (right) of a polycrystalline silicon solar cell.
 D. M. Tanenbaum, H. F. Dam, R. Rosch, et al., Sol. Energy Mater. Sol. Cells, submitted (2011).
 M. Seeland, R. Rösch and H. Hoppe, Imaging Techniques for Studying OPV Stability and Degradation, in Stability and Degradation of Organic and Polymer Solar Cells, edited by F. C. Krebs (John Wiley & Sons, 2012).
 M. Seeland, R. Roesch and H. Hoppe, J. Appl. Phys. 109 (6) (2011).
 R. Rösch, D. M. Tanenbaum, M. Jorgensen, et al., Energy Environ. Sci. 5 (4), 6521-6540 (2012).
 M. Seeland, R. Rösch and H. Hoppe, J. Appl. Phys. 111 (2), 024505 (2012).
 M. Seeland, C. Kästner and H. Hoppe, Appl. Phys. Lett. 107, 073302 (2015).