Dataset: The photocurrent-composition dependence of binary bulk heterojunction organic solar cells-combining high throughput experimentation and artificial intelligence models

This dataset corresponds to the article: "Predicting the photocurrent-composition dependence in organic solar cells" Xabier Rodríguez-Martínez, Enrique Pascual-San-José, Zhuping Fei, Martin Heeney, Roger Guimerà and Mariano Campoy-Quiles The article was first published on 07 Jan 2021 Energ...

Descripción completa

Detalles Bibliográficos
Autores: Rodríguez Martínez, Xabier, Pascual San José, Enrique, Guimerà, Roger, Garriga Bacardi, Miquel, Campoy Quiles, Mariano
Tipo de recurso: conjunto de datos
Fecha de publicación:2020
País:España
Institución:Consejo Superior de Investigaciones Científicas (CSIC)
Repositorio:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/223231
Acceso en línea:http://hdl.handle.net/10261/223231
Access Level:acceso abierto
Palabra clave:Raman spectroscopy
High-throughput
Combinatorial screening
Organic photovoltaics
Machine learning
Descripción
Sumario:This dataset corresponds to the article: "Predicting the photocurrent-composition dependence in organic solar cells" Xabier Rodríguez-Martínez, Enrique Pascual-San-José, Zhuping Fei, Martin Heeney, Roger Guimerà and Mariano Campoy-Quiles The article was first published on 07 Jan 2021 Energy Environ. Sci., 2021 DOI: 10.1039/D0EE02958K The dataset is divided into five main folders, which are briefly described below. - The "Descriptors" folder contains the detailed lists of optoelectronic descriptors used for the different training procedures performed throughout the work, i.e., a random selection of parameters, a hand-picked selection of parameters and a median of their corresponding distributions. The database of HOMO/LUMO levels and mobilities as retrieved from literature is included as well in an Excel spreadsheet. - The "Discrete devices" folder contains the photovoltaic figures-of-merit (open-circuit voltage, short-circuit current density, fill factor and power conversion efficiency) of the devices prepared at controlled donor:acceptor ratios. In these samples, the active layer thickness was screened in a high-throughput fashion using lateral gradients. Therefore, the variations observed at a given donor:acceptor ratio are solely due to the change in active layer thickness. - The "Graded devices" folder contains Raman mapping and Light-Beam Induced Current (LBIC) data for the 2D combinatorial devices, i.e. those including and orthogonal arrangement of active layer thickness and donor:acceptor ratio gradients on a single large area substrate. - The "Figures" folder contains a Jupyter Notebook file and, alternatively, a Python script, whose full execution in a row generates the figures of the main text and its Electronic Supplementary Information. These codes are designed to run with Python 3.7.3 and Scikit-Learn v0.22.2, while accessing data from other main folders (including "Descriptors", "Discrete devices" and "Graded devices") and subfolders ("Pools" and "Supplementary data"). For this reason, either code must be executed right from their original location within the "Figures" folder while keeping the original folder labels as well. - Finally, the "Raw data" folder contains the binary files generated as per the optoelectronic characterization of the 2D combinatorial devices. These were extracted employing our WITec alpha 300 RA+ confocal Raman setup in combination with LBIC and White-Beam Induced Current (WhiteBIC) measurements. We include as well scripts compatible with GNU Octave (*.m) to display the raw data.