SIBUR’s scientists test AI-powered hypotheses for catalysts and new materials development

3 February 2026. SIBUR has set out to accelerate its R&D processes through digital solutions and is building collaborations with leading universities into a unified platform for joint artificial intelligence research. This was announced by Ilnaz Zaripov, CEO of the SIBUR Innovations R&D Centre, during the Digital Solutions for New Materials and Chemistry session at the 3^rd International Forum of New Materials, Chemistry, and Technologies AMTEXPO.

One of the core workstreams involves applying machine learning methods to predict structure-property relations, conduct virtual screening of catalysts and formulations, and identify optimal synthesis conditions. SIBUR has developed a mathematical approach to generating experimental conditions that can significantly reduce the number of laboratory and pilot experiments, thereby accelerating new product development. The Company’s scientists are currently working on an algorithm for the digital discovery of novel post-metallocene catalysts, as well as projects focused on the development and production of specialty polymers and high-performance plastics. In this area, SIBUR collaborates with teams from leading universities through public-private partnership programmes.

Another workstream focuses on creating digital twins of industrial facilities and AI-driven optimisation of process parameters. This includes predictive control, reducing energy intensity, improving process selectivity, and lowering environmental impact. To achieve this, SIBUR's researchers work closely with physicists, chemists, and mathematicians to achieve a multiple-fold reduction in the computational complexity of models while improving prediction accuracy. Such solutions are expected to shorten the time required for the design, commissioning, and management of complex production facilities by orders of magnitude.

The third and most ambitious workstream involves analytics of large datasets on the quality and performance of polymer materials in customer applications. These data span a wide range of uses, from packaging and car parts to medical devices. The goal is to build a foundation that will enable AI to independently design molecules with targeted properties.

Ilnaz Zaripov, PhD in Chemistry, CEO of SIBUR Innovations R&D Centre:

"For us, artificial intelligence is a way to fundamentally transform the very logic of scientific research in chemistry. Collaborative projects with partners allow us to work at the intersection of chemistry, mathematics, and AI, tackling challenges that cannot be addressed within a single discipline or team. This approach opens new horizons – both in the development of catalysts and advanced polymer materials, and in managing the processes of scientific discovery and subsequent scaling of resulting technologies."

Accelerating R&D processes is one of SIBUR's strategic priorities. The Company's extensive network of research and innovation centres enables it to effectively adapt its product portfolio to the needs of key industries, as well as to develop and scale its own knowledge-intensive technologies in specialty, medium-, and small-tonnage chemistry for industrial application. This contributes to advancing Russia's technological sovereignty, enhancing the resilience of the Company's and its customers' production chains, and achieving the goals of the national project New Materials and Chemistry."