Tim Wiirger¹³, Linqian Wang¹, Darya Snihirova¹, Min Deng¹, Sviatlana Lamaka¹, David Winkler⁴,
Daniel Hoche¹, Robert MeiBner¹³, Christian Feiler¹, Mikhail Zheludkeyich¹²*

¹ Institute of Surface Science, Helmholtz-Zentrum Hereon, Geesthacht, Germany;

²Institute for Materials Science, Kiel University, Kiel, Germany;

³ School of Biochemistry and Genetics, La Trobe University, Bundoora, Australia

⁴ Institute of Polymers and Composites, Hamburg University of Technology, Hamburg, Germany

EXTENDED ABSTRACT: As the lightest structural engineering metal, magnesium (Mg) is a promising base material for advanced technology. However, to unlock the full potential of Mg-based materials, precise control over the corrosion rate is important whereas it was demonstrated that its degradation behaviour can be affected by small organic molecules. Recent research has discovered new, effective magnesium corrosion inhibitors,] 1] and electrolyte additives that boost the efficiency of magnesium-air primary batteries. [2] However, as small molecule chemistry space is essentially infinite, efficiently searching it to find small molecules with superior dissolution modulating properties (inhibitors or accelerators) using time- and resource-consuming experimental discovery methods is intractable. Consequently, computer-assisted selection of the most promising candidates prior to experimental investigation is of great benefit in the search for effective corrosion modulating additives for Mg-based materials.[3-5] Apart from a sufficiently large, diverse and reliable training data set and a suitable modelling framework (usually based on one or more machine learning algorithms), relevant molecular descriptors are a prerequisite for the development of predictive quantitative structure-property relationship models. The latter can either be selected by chemical intuition or based on statistical methods. Aqueous primary Mg-air batteries have considerable potential as energy sources for sea applications and portable devices. Apart from aspects regarding the air cathode, mainly two challenges have to be solved at the anode-electrolyte interface. The first refers to the power profile stability (discharge potential) and respective voltage drop, which is triggered by self-corrosion (anode fouling) and aging of the electrolyte. The second is related to the mass of anode material that can be utilized for discharge (utilization efficiency), which is also triggered by previous aspects, but additionally by the chunk-effect susceptibility of the anode material. Aside from alloying, an effective strategy to mitigate self-corrosion and battery failxire is the use of electrolyte additives. The talk outlines our recent activities in the systematic search for effective battery electrolyte additives for primary Mg-air batteries based on quantitative structure-activity relationship models and the active design of experiments. [6]

Keywords: magnesium battery; machine learning; QSPR; active learning; REFERENCES

[1] S. V. Lamaka, B. Vaghefinazari, D. Mei, R.P. Petrauskas, D. Hoche, M. L. Zheludkevich, Corros. Sci. 2017, 128, 224-240. ⑵ B. Vaghefinazari, D. Hoche, S. V Lamaka D. Snihirova, M.L. Zheludkevich, J. Power Sources 2020,453, 227880

[3]   C. Feiler, D. Mei, B. Vaghefinazari, T. Wurger, R. H. MeiBner, B. J. C. Luthringer-Feyerabend, D.A. Winkler, M. L. Zheludkevich, S. V. Lamaka, Corros. Sci. 2020,163, 108245.

[4]   T. Wurger, D. Mei, B. Vaghefinazari, D. A. Winkler, S.V Lamaka, M. L. Zheludkevich, R. H. MeiBner, C. Feiler, npj Mater. Degrad. 2021, 5, 2.

[5]   E. J. Schiessler, Tim Wiirger, S. V Lamaka, R. H. MeiBner, C. J. Cyron, M. L. Zheludkevich, C. Feiler, R. C. Aydin, npj Comp. Mater. 2021, 7, 193.

[6]   T Wurger, L Wang, D Snihirova, M Deng, S Lamaka, D. A. Winkler, D. Hoche, M. L. Zheludkevich, R. H. MeiBner, C. Feiler, Data-driven Selection of Electrolyte Additives for Aqueous Magnesium Batteries, Journal of Materials Chemistry A, 2022, accepted.