Rethinking Lead Halide Perovskite Nanomaterials Literature with Knowledge Graphs
Lidiia Varhanik a, Alexander Belikov b, Dmitry Baranov a
a Division of Chemical Physics and NanoLund, Lund University, Sweden
b GrowGraph, Paris, France
Proceedings of MATSUS Fall 2025 Conference (MATSUSFall25)
A4 Fundamental understanding of halide perovskite materials and devices - #PeroFun
València, Spain, 2025 October 20th - 24th
Organizers: Krishanu Dey, Iván Mora-Seró and Yana Vaynzof
Poster, Lidiia Varhanik, 467
Publication date: 21st July 2025

Colloidal metal halide perovskite nanocrystals (NCs) are versatile nanomaterials with strong potential for use in light-emitting diodes, solar cells, quantum light sources, and other optoelectronic devices. The broad compositional tunability and simple fabrication of these materials have garnered significant academic attention, with over 1500 publications appearing each year. This rapid growth in publications complicates the process of information search and validation, leading to difficulties in reproducible synthesis and optimization of the perovskite NCs. Here, we apply OntoCast — an LLM-powered knowledge graph (KG) generation framework — to analyze scientific papers in the perovskite NCs research.[1] The tool combines ontology management, triple generation, critique, entity disambiguation, and serialization, converting unstructured data (raw text, tables, etc) into structured, interconnected knowledge[2]. To demonstrate its applicability, we collected papers on perovskite nanomaterials published between 2015 and 2024, chose a subset of papers, and used OntoCast to capture relationships among materials and compositions, synthesis methods (e.g., precursors, ligands, solvents, temperatures, and routes), and report outcomes (e.g., optical properties). As a result, we developed a framework for automated generation of semantic triples in the domain of perovskite studies, stored in a Fuseki triple store, and queried using SPARQL (the standard query language for triples stores). The KG representation enables the aggregation of experimental results, facilitating reproducibility analysis and the identification of gaps in the semantic space of perovskite synthesis and characterization. We expect our tool to address the challenge of the growing volume of primary experimental data relative to derivative results, thereby enhancing reproducibility across research groups [3], and optimizing complex synthesis parameters to accelerate material discovery.

The work of L.V. and D.B. was funded by the European Union (ERC Starting Grant PROMETHEUS, project no. 101039683) and Swedish Research Council (Reg. no. 2024-04967). Views and opinions expressed are, however, those of the authors only and do not necessarily reflect those of the European Union or the European Research Council Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.

© FUNDACIO DE LA COMUNITAT VALENCIANA SCITO
We use our own and third party cookies for analysing and measuring usage of our website to improve our services. If you continue browsing, we consider accepting its use. You can check our Cookies Policy in which you will also find how to configure your web browser for the use of cookies. More info