Gerbrand Ceder

A high capacity thiospinel cathode for Mg batteries

Magnesium batteries are energy storage systems that potentially offer high energy density owing to their ability to employ magnesium metal as a negative electrode. Their development, however, has been thwarted by a paucity of functional positive electrode materials after the seminal discovery of the Mo6S8 Chevrel phase over 15 years ago. Herein, we report the second such material – a thiospinel – and demonstrate fully reversible Mg2+ electrochemical cycling vs. a Mg anode, which is complemented by diffraction and first principles calculations. The capacity approaches 80% of the theoretical value at a practical rate (C/5) at 60 °C, and yields a specific energy of 230 Wh kg−1, twice that of the Chevrel benchmark. Our results emphasize the advantage in employing “soft” anions to achieve practical divalent cation mobility.

Constructing first-principles phase diagrams of amorphous LixSi using machine-learning-assisted sampling with an evolutionary algorithm

The atomistic modeling of amorphous materials requires structure sizes and sampling statistics that are challenging to achieve with first-principles methods. Here, we propose a methodology to speed up the sampling of amorphous and disordered materials using a combination of a genetic algorithm and a specialized machine-learning potential based on artificial neural networks (ANNs). We show for the example of the amorphous LiSi alloy that around 1000 first-principles calculations are sufficient for the ANN-potential assisted sampling of low-energy atomic configurations in the entire amorphous LixSi phase space. The obtained phase diagram is validated by comparison with the results from an extensive sampling of LixSi configurations using molecular dynamics simulations and a general ANN potential trained to ∼45 000 first-principles calculations. This demonstrates the utility of the approach for the first-principles modeling of amorphous materials.