Our latest paper in Nature Materials with Brett Helms’ group at Lawrence Berkeley National Laboratory, is the result of a five-year effort in understanding electrodeposition instabilities at solid-solid interfaces, leading to high-performing lithium metal based batteries. In this article, we aim to dig deeper into the journey of how we got here.
In the fall of 2015, we began exploring the role of mechanical properties in stabilizing lithium electrodeposition at solid-solid interfaces in solid state batteries. Previous results from an elegant linear stability analysis performed by Monroe and Newman suggested that solids with sufficiently large moduli could block dendrite growth due to the stabilizing role of the hydrostatic part of the stress. Its contribution is of the form, given below:
where V refers to the volume and p refers to the pressure. Inorganic solids have mechanical moduli much larger than polymers, so one would naively expect them to stabilize electrodeposition, however this was not the case in practice.
Our analysis suggested otherwise, that inorganic solids with high moduli are destabilizing while those with low moduli are stabilizing. This made us realize that the other key ingredient in this analysis is the molar volume of lithium, i.e. the volume of lithium when it is in the electrode and in the electrolyte. Monroe and Newman’s analysis covered the regime where the molar volume of lithium-ion in the electrolyte is larger than the molar volume of lithium in lithium metal, i.e. volume reduction. In inorganic solids, the molar volume of the lithium ion is much smaller than the molar volume of lithium metal which makes the hydrostatic part of the stress destabilizing due to volume expansion during electrodeposition. From these two ingredients in the theory, we were able to obtain the following stability diagram.
We were surprised by this result until we found a similar stability diagram in geology. In their paper, they note, “In the majority of solid-solid phase transformation processes, the propagation of the interface is accompanied by a change in density. For this reason the density is an important order parameter that quantitatively characterizes the difference between the two phases.” This convinced us about our own surprising result. We termed these two regions of stability: pressure-driven stability corresponding to high modulus electrolytes and density-driven stability corresponding to low modulus electrolytes.
This density-driven dendrite suppression mechanism was one of the core ideas of our successful proposal to ARPA-E as part of the IONICS (Integration and Optimization of Novel Ion-Conducting Solids) program, started by Dr. Paul Albertus. The goals of the IONICS program are well-described in this excellent perspective. Our team, consisting of Brett Helms’ group at LBNL, Sepion Technologies (a spin-out from LBNL) and 24M Technologies, got to work on developing rechargeable lithium metal anodes. The goal was to develop advanced separators that would lead to lithium electrode subassemblies that could be seamlessly integrated with current and next-generation cathodes. In particular, we were developing polymer-ceramic composites with unique ion transport properties along with dendrite-suppressing capability.
After a few iterations, we settled on LiF being the inorganic phase. Bulk LiF does not conduct lithium ions, so we had to engineer ionic conductivity in this system. Using a variety of computational techniques, we explored the energetics of ion motion pathways at the interface between LiF and polymers with intrinsic microporosity (PIM). Surprisingly, we found interfacial ion transport to be much more favorable than in the bulk components of the composite, showing that the composite’s ionic conductivity was high enough for its use as a separator. Brett Helms’ group showed experimental validation of our predicted values for the energetic barrier for ion motion, leading to the conclusion that ionic motion path was through the interface. Based on this conclusion, we were able to calculate the average molar volume of the mobile lithium ion in the composite separator and compare it to the corresponding bulk lithium value, which placed us in the dendrite-suppressing region of the stability diagram shown above. Further (unreported) simulations had given some evidence of the composite’s low shear modulus, but experimental work in Brett Helms’ group showed the dependency between shear modulus and LiF loading in the composite, adding the finishing touches to the work: we had confirmed that the LiF@PIM composite falls in the density-driven stability regime and does indeed suppress dendrite growth.
Beyond this paper, our team, led by 24M Technologies, has been able to develop these batteries into large format (80 cm²), with these custom polymer-ceramic composites. As this project was on-going, my group was working on battery needs for various markets, and in-particular, electric vertical take-off and landing aircraft. eVTOLs, popularly called as flying cars, promise to offer a new form of urban mobility. There are some unique power needs for eVTOLs that make them different from electric vehicles. In a paper written with Airbus A³, we showed that the landing segment which requires about as much power as take-off, is the challenging segment for the battery to handle. In a popular piece, we discuss the importance of the specific energy of the battery in leading to a feasible design.
In the beginning of 2019, the two pieces of research came together. We realized that our high specific energy lithium metal batteries could be ideal to deliver the needs of the mission profiles of eVTOLs. We have come a long way in taking these batteries to the sky. Finally, thanks to Advanced Research Projects Agency-Energy (ARPA-E) for supporting this work and allowing this amazing team to come together. In the spirit of ARPA-E, this team has truly “changed what’s possible”!