This study is focused on the development of liquid organic hydrogen carriers (LOHC) based on N-heterocyclic compounds. These LOHC-substrates are attractive for their lower hydrogen extraction temperature compared to cycloalkanes, which is caused by the low enthalpy of the dehydrogenation reaction of the N-heterocyclic compounds. The low hydrogen extraction temperature, as well as the low volatility of the heterocycles, provide high purity hydrogen from the reaction. Under similar reaction conditions, the comparison of the efficacy of three promising heterocycles (1-methyl-octahydroindole (8HMI), tetradecahydrophenazine and decahydroquinoline) was carried out in the presence of palladium-containing catalysts. As a result, the advantages of using catalysts supported by alumina, and the high perspectivity of the 8MHI application as a LOHC-substrate, were shown. The dehydrogenation of 8HMI in the presence of 1 wt.% Pd/Al₂O₃ allowed for reaching a 100% yield in hydrogen under the conditions of the standard catalytic test (1 h, 240 °C). In order to study the high reactivity of 8HMI, thermodynamic dehydrogenation reaction profiles were computationally evaluated, which showed that 8HMI was the most energetically preferred in the field of hydrogen storage from the studied heterocyclic compounds.

Conjugated organic polymers are the cornerstone of modern organic electronics, yet accurate prediction of their properties remains a challenging task due to their synthetic complexity and high computational cost of quantum-chemical methods. Here, we develop a graph neural network based on the DimeNet++ direct message passing architecture to predict HOMO, LUMO and energy gaps of conjugated polymers directly from their 3D monomer structure. The model was pre-trained on TD-DFT-extrapolated data and trained on a limited dataset of experimentally measured properties. As a result, pre-training had significantly improved model’s accuracy compared to direct training (MAEs ~0.3 eV vs. 0.074 eV, 0.141 and 0.172 for HOMO/LUMO and energy gap, respectively). Pre-training on monomer DFT data did not provide comparable gains. The results demonstrate that polymer-relevant pre-training is critical for capturing structure–property relationships and enables accurate predictions without delta-learning or prior quantum-chemical calculations, facilitating efficient screening and rational design of conjugated polymers for organic optoelectronics.