Design Principles for Wurtzite-type Ferroelectrics
Low-energy compute-in-memory architectures promise to reduce global energy demand for computation and data storage. Wurtzite-type ferroelectrics such as (Al,Sc)N alloy offer potential advantages in both performance and integration with existing semiconductor processes, but at present, all known wurtzite-type ferroelectrics require excessively large operating voltages for polarization reversal. We report a large-scale computational search among multinary compounds for new materials with polarization-switching barriers lower than AlN while maintaining large breakdown fields.
Geoff Brennecka and Prashun Gorai (Colorado School of Mines)
![](https://imagedelivery.net/qXYabcYg0JPsn0bBVWE4SA/74ea68a6-2df0-4230-aa6e-60a29e768403/Design_Principles_for_Wurtzite-type_Ferroelectrics.png/raw)
•Ferroelectrics (FEs) based on tetrahedral bonding such as wurtzite (Al,Sc)N, (Al,B)N, and (Zn,Mg)O alloys promise performance and semiconductor integration advantages over classic FEs.
•Challenge: Known tetrahedral FEs require large electric fields for polarization reversal (coercive fields).
•Approach: We performed a large-scale computational search to identify candidates for tetrahedral FEs; four compounds emerged as targets for further investigation and potential alloying.
•Materials Design Rules: Our findings highlight the fundamental materials properties—bond ionicity and strength–that affect the magnitude of coercive fields in tetrahedral ferroelectrics and showed that the popular design rule (c/a lattice constant ratio) applies only within a given alloy, not across different compounds