Accurate and reliable prediction of the optical and photophysical properties of organic compounds is important in various research fields. Here, we developed deep learning (DL) optical spectroscopy using a DL model and experimental database to predict seven optical and photophysical properties of organic compounds, namely, the absorption peak position and bandwidth, extinction coefficient, emission peak position and bandwidth, photoluminescence quantum yield (PLQY), and emission lifetime. Our DL model included the chromophore–solvent interaction to account for the effect of local environments on the optical and photophysical properties of organic compounds and was trained using an experimental database of 30 094 chromophore/solvent combinations. Our DL optical spectroscopy made it possible to reliably and quickly predict the aforementioned properties of organic compounds in solution, gas phase, film, and powder with the root mean squared errors of 26.6 and 28.0 nm for absorption and emission peak positions, 603 and 532 cm –1 for absorption and emission bandwidths, and 0.209, 0.371, and 0.262 for the logarithm of the extinction coefficient, PLQY, and emission lifetime, respectively. Finally, we demonstrated how a blue emitter with desired optical and photophysical properties could be efficiently virtually screened and developed by DL optical spectroscopy. DL optical spectroscopy can be efficiently used for developing chromophores and fluorophores in various research areas.

Experimental databases on the optical properties of organic chromophores are important for the implementation of data-driven chemistry using machine learning. Herein, we present a series of experimental data including various optical properties such as the first absorption and emission maximum wavelengths and their bandwidths (full width at half maximum), extinction coefficient, photoluminescence quantum yield, and fluorescence lifetime. A database of 20,236 data points was developed by collecting the optical properties of organic compounds already reported in the literature. A dataset of 7,016 unique organic chromophores in 365 solvents or in solid state is available in CSV format.

The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energies, which are key factors in optoelectronic devices, must be accurately estimated for newly designed materials. Here, we developed a deep learning (DL) model that was trained with an experimental database containing the HOMO and LUMO energies of 3026 organic molecules in solvents or solids and was capable of predicting the HOMO and LUMO energies of molecules with the mean absolute errors of 0.058 eV. Additionally, we demonstrated that our DL model was efficiently used to virtually screen optimal host and emitter molecules for organic light-emitting diodes (OLEDs). Deep-blue fluorescent OLEDs, which were fabricated with emitter and host molecules selected via DL prediction, exhibited narrow emission (bandwidth = 36 nm) at 412 nm and an external quantum efficiency of 6.58%. Our DL-assisted virtual screening method can be further applied to the development of component materials in optoelectronics.

Singlet fission (SF) is an intriguing process in which a singlet exciton produces two triplet excitons in molecular aggregates. Perylenediimide (PDI) derivatives are promising materials for SF-based photovoltaics, and the SF process in PDI aggregates is important to investigate for their applications. In this work, we studied the entire SF process occurring in the colloidal nanoparticles of a PDI derivative in solutions by using time-resolved fluorescence and transient absorption (TA) experiments. PE–PDI was found to form the colloidal nanoparticles of H- and J-aggregates in polar solvents. The TA signals of PE–PDI aggregates in solutions were selectively measured by wavelength-dependent excitation. The TA signals were analyzed by using a global fitting analysis, and all kinetic parameters involved in the entire SF process were determined. Our current investigation has confirmed that fast SF occurs on the surface of the colloidal nanoparticles of PDI aggregates via the charge transfer mediated mechanism, giving a high quantum yield of triplet excitons.

Radiation therapy (RT) is an important and potentially curative modality for head and neck squamous cell carcinoma (HNSCC). Locoregional recurrence (LR) of HNSCC after RT is ranging from 15% to 50% depending on the primary site and stage. In addition, the 5-year survival rate of patients with LR is low. To classify high-risk patients who might develop LR, a deep learning model for predicting LR needs to be established. In this work, 157 patients with HNSCC who underwent RT were analyzed. Based on the National Cancer Institute’s multi-institutional TCIA data set containing FDG-PET/CT/dose, a 3D deep learning model was proposed to predict LR without time-consuming segmentation or feature extraction. Our model achieved an averaged area under the curve (AUC) of 0.856. Adding clinical factors into the model improved the AUC to an average of 0.892 with the highest AUC of up to 0.974. The 3D deep learning model could perform individualized risk quantification of LR in patients with HNSCC without time-consuming tumor segmentation.