Bridging Indoor and Outdoor Perovskite Degradation Using Physics Modeling and Bayesian Analysis

In this work, we use physics-based modelling with a probabilistic Bayesian framework, relying solely on current-density-voltage (J-V) curves recorded throughout aging in both indoor and outdoor degradation tests. We investigate whether the same degradation mechanisms occur in both outdoor tests and indoor accelerated aging tests — an important step toward identifying the dominant degradation mechanisms from commonly available data and establishing the validity of accelerated aging tests.
The data provided here includes these J-V curves as well as the outdoor maximum power point tracking (MPPT) data with corresponding irradiance and cell temperature recordings. For the indoor accelerated tests, the J-V curves were recorded during the aging process and therefore under the same aging conditions of the cells: ~1.3 suns and fixed temperature levels (25, 45, 65 and 85 °C). The aging setup and measurement routine were detailed in Köbler (2022).
The indoor_jvs.csv file includes the following information: cell_id: such as D1_P1, represents substrate 1 (D1), cell 1 (P1), where each substrate contains six electrically isolated perovskite solar cells. Note that the naming of substrates restarts at each temperature; temp: the aging temperature in °C; meas_time: the time in hours since the start of aging; scan: scan direction of J-V (FOR: forward; REV: reverse); v and j: the raw voltage and current density data (as lists).
For the outdoor tests, J-V curves were recorded under a sun simulator at standard test conditions (1 sun, 25 °C). The outdoor_jvs.csv file follows the same structure as the indoor data, with two exceptions: the temp column is omitted, and meas_time is given as a date.
The outdoor MPPT data (outdoor_mpp_) is provided separately for three cells and includes voltage, current, and power at MPP as a timeseries. Irradiance and cell temperature data are provided in a separate merged file (irradiance_and_cell_temp.csv).