Minimal Long-Term Intrinsic Degradation Observed in a Polymer Solar Cell Illuminated in an Oxygen-Free Environment

P solar cells (PSCs) have attracted substantial attention over the past decade, and research devoted to increasing the starting power conversion efficiency (PCE) has driven it above 10%. While promising, this value only represents the initial performance of the PSCshow the PCE of the solar cell degrades with time is also of critical importance. Commercial silicon modules are typically warrantied for 25 years, which sets the standard for the photovoltaic community. PSCs that are exposed to continuous illumination and monitored over time are observed to degrade on two time scales. First, there is an initial fast degradation, called burn-in, that slows down with time. After burn-in ends, there is a second period of linear degradation. The slope of this linear portion determines a PSC’s lifetime, which is chosen to be the time over which it takes the solar cell to reach 80% (T80) of postburn-in efficiency. If linear degradation proceeds slowly and T80 is not reached while the solar cells are monitored, then the lifetime is determined from a linear extrapolation of the performance after burn-in ends. Oxygen and water, extrinsic to PSCs, are known to affect their lifetimes; PSCs without any packaging degrade in air in minutes. While the performance of flexible plastic barrier materials has improved, PSCs packaged with PET films demonstrate lifetimes of only several hundred to a thousand hours. Glass-on-glass packaging has improved the observed lifetimes of several types of PSCs. When solar cells made with the polymer poly(3-hexylthiophene-2,5-diyl) (P3HT) are encapsulated in glass-on-glass packaging and illuminated for several thousand hours, extrapolated lifetimes of 5000−6000 h are observed. PSC’s with the polymer poly[9′hepta-decanyl-2,7-carbazole-alt-5,5-(4′,7′-di-2-thienyl-2′,1′,3′benzothiadiazole) (PCDTBT) have demonstrated even better extrapolated lifetimes, on the order of 12 000−18 000 h. Assuming there are on average 5.5 h of sunlight in 1 day, 2007 h of constant illumination correspond to 1 year of operational lifetime. Thus, PSCs in flexible packaging might be expected to last 6 months, while PCDTBT solar cells in glass-on-glass packaging might last 6−9 years. It is encouraging that improvements in packaging translate into improved observed lifetimes, but important questions remain. When using encapsulated PSCs to evaluate lifetimes, there is always the question of whether the degradation rate is determined by the packaging leak rate, however small it may be. This may be from residual water in the packaging materials outgassing over time or in glass-on-glass encapsulation, the leak rate of the epoxy used to seal the devices. Most importantly, it is unknown whether polymer materials, and completed PSCs, are intrinsically stable enough to withstand repeated photo-excitation for the span of 25 years. To test the intrinsic stability of PSCs in a controlled environment and to minimize extrinsic degradation through any uncertainty of package leaking, we constructed a portable environmental chamber that can hold PSCs at their maximum power point and simultaneously measure their current−voltage (IV) characteristics at fixed time intervals in an atmosphere with less than 0.1 ppm water and oxygen. We fabricated PSCs in both the standard and the inverted architectures, loaded them into the environmental chamber, and exposed them to continuous illumination for 7700 h (>9 months). We observed a slower burn-in for inverted architecture devices, though ultimately both standard and inverted devices lost ∼40% PCE. Once burn-in ended, the linear degradation proceeded slowly. In fact, the degradation rate for many of the monitored devices was so slow that accurately assigning device lifetimes became difficult. By minimizing oxygen and water content in the atmosphere for the duration of the lifetime test, we observed that, on average, PSCs can operate with minimal intrinsic degradation for thousands of hours with extrapolated lifetimes beyond 15 years. We fabricated PSCs using the polymer PCDTBT and the fullerene [6,6]-phenyl C70-butyric acid methyl ester (PC71BM) as the semiconducting materials (Figure 1). We chose PCDTBT because it is well studied, so it offers a point of comparison, and has previously shown promising stability. Furthermore, it has a sufficiently high glass transition temperature (∼135 °C), such that temperature induced degradation is minimal and the primary degradation is photoinduced. We chose PC71BM to avoid the photodimerization that occurs in PC61BM. 20−22 In addition to the standard architecture, which uses PEDOT:PSS to collect holes and calcium to collect electrons, we fabricated inverted devices that use zinc oxide nanoparticles as the electron collector and molybdenum oxide as hole collector to eliminate low work function metals from the PSC. Aluminum and silver were used as reflective electrodes for the standard and inverted devices, respectively (Figure 1). Before loading the environmental chamber, IV curves of all devices were taken. The average PCE of the standard and inverted devices monitored in the lifetime test was 5.3% and 5.1%, respectively.