The degradation of perovskite solar cells due to reverse bias (RB) is one of the remaining challenges hindering the commercialization of the technology. To overcome this challenge, a thorough understanding of and control over the breakdown (BD) voltage are crucial. A prerequisite for this is that the community “speaks the same language,” that is, that the reported BD voltages are comparable. A review of literature data shows that the impact of measurement parameters is often unknown and seems to depend strongly on sample properties. It follows that standardization is the only way to reach comparability. Here, a set of measurement parameters to fill this gap is proposed. Additionally, various definitions of a “BD voltage” are used in parallel without any way of relating them to each other; this metric and its determination need to be considered as well. After a thorough discussion of the available definitions, the use of the point of maximum curvature is introduced. Its main advantage is the possible connection to an analytical description of the BD mechanism. In this way, a starting point for scientists new to the field of RB stability is provided, and the ground for a broader discussion in the community is prepared.

Partial shading of CIGS modules can lead to permanent damage of the module in the shaded area. This is caused by harmful reverse bias voltages in the shaded area which lead to reverse bias induced defects, also known as wormlike defects. A lot is already known about the origin and propagation of wormlike defects. However, the fundamental question; why is CIGS so sensitive to reverse bias damage? has not yet been answered. In this study we show that CIGS semiconductor material in the presence of an electric field will spontaneously decompose.

Two types of Cu(In,Ga)Se₂ (CIGS) solar cells, both designed for implementation in CIGS modules, were subjected to temperatures between 25C and 105C. Simultaneous exposure to AM1.5 illumination allowed the measurement of their electrical parameters at these temperatures. These two types of solar cells, produced with different deposition routes on soda lime glass (SLG) and polyimide (PI) substrates, showed large variations in the temperature dependency of their electrical parameters. It was shown that the temperature dependency of the open circuit voltage (V_oc) was dependent on its room temperature value: a high V_oc at 25 °C led to a slower loss of V_oc when the temperature was increased. For the V_oc, the normalised temperature dependency varied between -0.28%/°C and -0.47%/°C, which is in agreement with the literature. The temperature dependency of the short circuit current density (J_sc) showed more surprising results: while the PI samples had the expected positive temperature dependency (0.03 to 0.32%/°C), the SLG samples showed a small negative impact of increasing temperature (-0.01 to -0.05%/°C). A correlation between the temperature dependencies of the J_sc and the ideality factor n was observed. Therefore, this difference in the temperature dependence of the J_sc could be caused by increased recombination for the SLG samples. Furthermore, the temperature coefficients of the fill factor (negative), efficiency (negative), and the series (slightly negative) and shunt (negative) resistances were calculated.

The degradation behavior of Mo/MoSe₂ layers have been investigated using damp heat exposure. The two studied molybdenum based films with different densities and microstructures were obtained by lifting off Cu(In,Ga)Se₂ layers from a bilayer molybdenum stack on soda lime glass. Hereby, a glass/Mo/MoSe₂ was obtained, which resembles the back contact as present in Cu(In,Ga)Se₂ solar cells. The samples were degraded for 150 h under standard damp heat conditions and analyzed before, during and after degradation. It was observed that the degradation resulted in the formation of needles and molybdenum oxide layers near the glass/Mo and the Mo/Cu(In,Ga)Se₂ interfaces. X-ray Photoelectron Spectroscopy measurements have shown that the sodium was also present at the surface of the degraded material and it is proposed that the degraded material consists mostly of MoO₃ with intercalated Na⁺. This element has likely migrated from the soda lime glass. This intercalation process could have led to the formation of Na_xMoO₃ ‘molybdenum bronze’ following this redox reaction: xNa⁺ + MoO₃ + xe⁻ ↔ Na_xMoO₃ It is proposed that the formed oxide layer contains Na_xMoO₃ with different Na⁺ contents and different grades of conductivity. This intercalation process can also explain the high mobility of Na⁺ via the grain boundaries in molybdenum. It was also observed that the molybdenum film with a top layer deposited at a high pressure is more susceptible for damp heat degradation.

Unencapsulated CIGS solar cells were simultaneously exposed to damp heat and illumination. In-situ monitoring of their electrical parameters demonstrated a rapid decrease of the efficiency, mainly driven by changes in the series and shunt resistances. The non-degraded and degraded solar cells were studied by SIMS and XRD to investigate the material changes leading to efficiency loss. SIMS showed the migration of sodium and potassium, likely leading to changes in the shunt resistance and output voltage. Extensive XRD measurements showed that molybdenum oxide was formed and that the in-plane stress in the ZnO:Al film increased. The stress increase is most likely due to the incorporation of species like hydroxide in the grain boundaries. These phenomena could lead to the observed increased series resistance in the solar cells.

Degradation rates and mechanisms for molybdenum back contacts and ZnO:Al front contacts exposed to damp heat were obtained from literature and experiments. It was found that molybdenum back contacts with a higher density and covered by a MoSe₂ film are more stable than their low density, bare counterparts. For ZnO:Al front contacts, various material changes, including thickness, deposition temperatures and doping concentration increase as well as post-deposition treatments can lead to a more stable material. Moreover, the degradation rate of encapsulated and non-encapsulated CIGS solar cells and modules exposed to damp heat were determined It was observed that in general, exposure to damp heat led to a reduction in efficiency, mostly caused by a reduction in Voc and FF. However, it was found that very large differences exist between samples.

Non-encapsulated CIGS solar cells with different contents of sodium (Na) and potassium (K) were simultaneously exposed to damp heat and illumination. The solar cells with higher alkali (Na, K) content exhibited higher initial conversion efficiencies, but degraded severely within 100 hours, while samples with a lower alkali content retained their efficiency longer. The degradation was likely caused by the migration of alkali atoms, leading to shunting of the solar cells and the formation of alkali rich spots.

A 'hybrid' degradation setup, which allows the use of humidity, temperature and illumination as loads in order to accelerate degradation of solar cells and modules, has been designed and constructed. In this setup, the current voltage output of photovoltaic samples is automatically logged and the electrical parameters are calculated. This allows the study of the impact of illumination and damp heat induced degradation by in-situ monitoring. Additionally, this setup also allows the determination of temperature dependency of solar cells by a simple procedure.

CIGS solar cells were exposed to liquid water purged with the atmospheric gases carbon dioxide (CO₂), oxygen (O₂), nitrogen (N₂) and air in order to investigate their chemical degradation behavior. The samples were analyzed by electrical, compositional and optical measurements before, during and after exposure in order to follow the degradation behavior of these solar cells in time. The solar cells showed a rapid decrease in conversion efficiency when exposed to water purged with a combination of CO₂ and N₂ as well as to water purged with air, while their efficiency was slowly reduced in unpurged water and water purged with N₂ or O₂. Cross-section SEM showed that the exposure of samples to H₂O with large concentrations of CO₂ led to the dissolution of the ZnO:Al layer, likely starting from the grain boundaries. This resulted in an increased series resistance, which is likely related to an increase in resistivity of the ZnO:Al layer. It also led to a very rapid decrease of the short-circuit current of these samples. Therefore, the conversion efficiency was rapidly lost.