Publication date: 17th July 2025
Battery innovation relies on developing new electroactive materials [1]. To timely respond to the increasing demand for energy storage solutions, the European Battery 2030+ Initiative targets accelerating by 5-to-10 fold the current rate of battery materials discovery within the next 5-10 years. Taking up this challenge requires disruptive approaches that allow rethinking the traditional experimentation process (based on researcher’s chemical intuition and trial-error scheme), which is inherently slow and economically expensive. Indeed, the crystal-chemical space offered by the periodic table for the search for new battery materials is huge and still far from being exhaustively explored.
Accelerating the exploration of broad chemical spaces requires a mind change in our approach to materials research, but also building new lab infrastructures and analytical tools, which include automated high-throughput synthesis modules, automated data analysis programs able to handle large amounts of data, as well as AI-aided experimental planners.
In this presentation, we will present several strategies explored at CIC energiGUNE to speed up the different stages of the development of new materials for Li-ion and Na-ion batteries. Such approaches include:
(i) the use of automated screenings of materials databases in search for new families of compounds that can be converted into electroactive materials [2];
(ii) the development of MAITENA, a Materials Acceleration and Innovation plaTform for ENergy Applications, which include semi-automated modules for high-throughput inorganic syntheses (e.g. co-precipitation, solvothermal, sol-gel) and characterizations (e.g. XRD, electrochemistry) [3];
(iii) the development of analysis tools for automated data treatment and analysis, including a Machine-Learning experimental planner, chemometrics approaches for data analysis of large XAS data sets, or the new FullProfAPP that enables automated Rietveld refinements of large series of data, in particular those generated from operando experiments [4,5].
This work was in particular supported by the Spanish MCIN/AEI/10.13039/501100011033 and ERDF/EU (project ref. PID2019-106519RB-I00, PID2022-140823OB-I00, PhD grant PRE2020-092978), the Ministerio de Industria, Comercio y Turismo and Next Generation EU (VEC-020100–2022–127/PP27), the Basque Government (PhD grants ref. PRE-2021-2-011) and the European Commission (G.A. No 957189).