Research ArticleMATERIALS SCIENCE

Machine learning–assisted molecular design and efficiency prediction for high-performance organic photovoltaic materials

See allHide authors and affiliations

Science Advances  08 Nov 2019:
Vol. 5, no. 11, eaay4275
DOI: 10.1126/sciadv.aay4275
  • Fig. 1 Information about our database of OPV donor materials.

    (A) Distribution of PCE values of the 1719 molecules in our database. (B) Schematics of expressions of a molecule, including image, simplified molecular-input line-entry system (SMILES), and fingerprints.

  • Fig. 2 Testing results of ML models.

    (A) Testing of the deep learning model using images as input. (B to D) Testing results of different ML models using (B) SMILES, (C) PaDEL, and (D) RDKIt descriptors as input.

  • Fig. 3 Performance of ML models.

    (A to D) The testing results of (A) BPNN, (B) DNN, (C) RF, and (D) SVM using different types of fingerprints as input.

  • Fig. 4 Verification of ML models with experiment.

    (A) Comparison of the results from four different models. (B) Schematic diagram of the cell architecture used in this study. (C) J-V curve of the solar cell with the active layer using the predicted donor material. (D) Prediction results versus experimental data for the predicted donor materials with the RF algorithm and Daylight fingerprints.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/5/11/eaay4275/DC1

    Section S1. ML methods and machine language expressions of molecule

    Section S2. Process of experiment and proof of the reliability of the ML model

    Fig. S1. Introduction to different ML algorithms.

    Fig. S2. Chemical structures of the 10 new donor materials.

    Fig. S3. Prediction results versus experimental data for the 10 new donor materials.

    Table S1. Details of PaDEL descriptors.

    Table S2. Details of RDKIt descriptors.

    Table S3. Complete MACCS fingerprint of P3HT and PTB7.

    Table S4. Photovoltaic parameters of OPV devices fabricated with different donor materials.

    Table S5. Prediction results from DNN, RF, and SVM using Hybridization and FP2 fingerprints as inputs, as well as DNN and RF using Daylight fingerprints.

    Table S6. Prediction results from BPNN using Daylight fingerprints when classification threshold is 10%.

    References (48, 49)

  • Supplementary Materials

    This PDF file includes:

    • Section S1. ML methods and machine language expressions of molecule
    • Section S2. Process of experiment and proof of the reliability of the ML model
    • Fig. S1. Introduction to different ML algorithms.
    • Fig. S2. Chemical structures of the 10 new donor materials.
    • Fig. S3. Prediction results versus experimental data for the 10 new donor materials.
    • Table S1. Details of PaDEL descriptors.
    • Table S2. Details of RDKIt descriptors.
    • Table S3. Complete MACCS fingerprint of P3HT and PTB7.
    • Table S4. Photovoltaic parameters of OPV devices fabricated with different donor materials.
    • Table S5. Prediction results from DNN, RF, and SVM using Hybridization and FP2 fingerprints as inputs, as well as DNN and RF using Daylight fingerprints.
    • Table S6. Prediction results from BPNN using Daylight fingerprints when classification threshold is 10%.
    • References (48, 49)

    Download PDF

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article