Triphenylamine-Based Dyes in DSSCs | PCE % Predictor Web App

We trained deep neural network (DNN) models to predict these properties directly from molecular structures represented by SMILES strings.

  • Input: SMILES string
  • Algorithm: R² of 0.90, 91, and 90 for Jsc, Voc, and FF, respectively
Model Info Launch App | Demo

Quick Guide: Input Options for PCE Prediction

SMILES Input Example
Option 1: SMILES Format

Draw your dye in a chemical editor (e.g., ChemDraw, RDKit) and copy its SMILES notation. Paste the SMILES string of your dye molecule directly into the input box. Example: O=C(O)C1=CC=CN=C1

Batch Screening Example
Option 2: Batch Screening

Import multiple SMILES strings from a .txt file or .csf file. This allows efficient high-throughput screening of large molecular databases.