The design principles of NASICON synthetic accesibility based on 3881 potential NASICON type oxides over periodic table
The stability of multicomponent ceramic materials is difficult to predict as the phase stability is a non-local effect that depends on the energy difference between the material and many possible competing phases. Due to the complex bond topology and composition, a simple physical model is not straightforward to be constructed, which greatly limits the discovery of new materials among the enormous combinatorial space. Sodium (Na) Super Ionic Conductor (NASICON) materials are a good example of such materials. Despite the great combinatorial space for exploration, there is no stability rule for guiding the experimental trial and error. In this work, we developed a computational framework to understand the stability origin of NASICON-type materials. We present a high throughput phase diagram dataset that consist of 3881 computed NASICON materials, develop a physical interpretation of the stability origin, as well as propose a phenomenological “tolerance factor” for predicting new NASICON materials without more DFT calculations. This work not only provide tools to understand synthetic accessibility of NASICON-type materials, but also demonstrate an efficient paradigm for discovering new materials with complicate composition and atomic structure.
Click here to play with the interactive synthesis map.
The preprint of how to make a synthetic accesible NASICON across the periodic table can be found in Nature Communications as “Synthetic accessibility and stability rules of NASICONs”.
The previous understanding of NASICONs are mainly on Zr-based compounds, majorly because of their superior performance since the first discovery. However, our dataset reveals that there are much larger synthesis space for NASICONs. My collegue Jingyang Wang has tried to make some of the most promising ones as predicted by my high-throughput phase diagram calculations and end up with a 5/6 succesful rate, while the only failure forms a NASICON phase but with smore amount of impurity.
The origin of the stability comes from two facts. One is the bond compatibility, as the chemical freedom of metal site is bounded by the covalency between Na-O (lower bound) and A-O (upper bound, polyanionic bond). The other would be site miscibility, while the mixing of cation and polyanions site are largely determined by the lattice distortion. A mild lattice distortion can be passed from polyanion site to cation site to stablize the structure. However larger lattice distortion would lead to significant driving force for decomposition.
On base of the dataset we have. We have developed a physical intuited “tolerance factor” than enables one to predict NASICON stability. You will only need to use the expression 0.273*t1+t2≤0.421 to estimate if one NASICON is synthetic accessible. In this formula, t1 and t2 are two features machine learnt from basic physical properties (t1 being x axis, t2 being y axis in the figue below).
We are currently developing a website for convient predicting of NASICON stability, especially for experimentalist who need a handy tool for supporting their synthesis. A screenshot of the website in construction is given below. Please feel free to keep track on it if you are interested.