By Kyle Proffitt
October 9, 2024 | A common concern with solid-state batteries is the need to maintain tight contacts between layers, as there is no liquid that can access voids and ensure conductivity; volume changes associated with lithium deposition further compound this issue. A common solution is the application of external stack pressure, but many consider this a stepping stone or crutch that is primarily applicable in laboratory settings.
Particularly in automotives, adding significant pressure creates unique design challenges and almost certainly increases cost while decreasing total energy density. For this reason, many researchers are aiming to create solid-state designs that can operate effectively with either very low or no external pressure application. At the 2024 Solid-State Battery Summit in Chicago, a series of presentations presented designs that address the pressure issue and provide clues about which solid-state technologies we will encounter first.
Manufacturing Challenges and Porous Electrolyte
Michael Tucker, from Lawrence-Berkeley National Lab, set the stage for this track by detailing manufacturing techniques that can pay off in improving laceface contacts and moving toward reduced pressure. These efforts converge in designs that exploit pores in a bilayer solid state electrolyte (SSE). Specifically, Tucker discussed work with lithium lanthanum zirconium oxide (LLZO), a ceramic material commonly referred to as garnet because the crystal structure is very similar to that of the silicate gem mineral. LLZO is promising as an SSE because it is low-cost, has high ionic conductivity, and is stable with most cathode and anode materials. Tucker rehearsed his academic slog through different optimization techniques for this material.
Tucker is interested in overcoming some of the challenges that ceramics present in terms of processability, which he defined to include cost, throughput, and compatibility with existing equipment. Because the SSE detracts from total energy density, making it thin is desirable. Tucker’s group was trying to create very thin (20 µm) layers of LLZO but having difficulty. He explained that methods such as thermal barrier coating exist for precisely controlling ceramic structure and properties, but they are very expensive and only make sense for very high-value, critical scenarios such as coating airplane wings.
For their work, Tucker’s group began with tape casting, a much less expensive method. Here, a slurry of LLZO and pore formers, suspended in solvents, is extruded through a small “blade gap” onto a moving surface, laying down a thin, continuous layer. Solvents are dried off, and then the material is heated, or sintered, burning off binders and solvents to achieve desired compactness and other properties. Tucker’s group decided to play with texturing the material to increase surface contacts. After tape casting, a film with a patterned structure could be pressed into the material to give it texture, which was retained after sintering. “What this really does is enhance the surface area contact between the electrode and the electrolyte,” Tucker said. That move increased critical current density in cells—meaning the more textured interface withstood higher current.
LLZO showed some unique properties in the tape casting process, though, “evolving or evaporating lithium the entire time that you’re trying to sinter it,” Tucker explained. His group circumvented this problem by adding sacrificial lithium carbonate and found that the time and temperature of sintering had to be carefully controlled to achieve LLZO with proper cubic microstructure. Ultimately with tape casting, they could make thin sheets that were good for coin cells, but scaling up did not work, as the sheets became too brittle.
The next idea was to create thick, porous LLZO as a support structure with a layer of thin, dense LLZO. The porous structure gives mechanical support, and active cathode or anode material can infiltrate these pores so space is not wasted. As with the textured surfaces, the surface area of contacts is greatly increased. Although their pores are amenable to cathode or anode material infiltration, Tucker’s group has focused on the cathode side.
In different experiments, they successfully filled pores with lawsone, a liquid organic lithium carrier, or with a polymer-based low-viscosity slurry of NMC particles. Most recently, his group designed coin cells using a composite of succinonitrile (SN) mixed with lithium salts for catholyte, NMC811 cathode material, and carbon black as conductor. SN is an organic ionic plastic crystal (OIPC) that melts well at 80 °C and can then wick into pores or surface imperfections in LLZO, but it solidifies at room temperature to make a true solid-state cell. These cells used a thin, dense LLZO, and provided ~125 mAh/g energy density at room temperature while requiring no external pressure. This energy density is low, but it provides an example of SSBs that do not require pressure to operate.
A problem with their first-generation porous LLZO layers was that the pores were poorly connected, making it difficult to fill them with active material. The group moved to freeze-tape casting. In this approach, water is mixed in with the slurry before casting onto a freezing bed, causing the water to freeze and create larger, connected pores that remain after the ice is freeze-dried out. The group is also using phase inversion tape casting, where solvent is mixed into the extruded slurry, and then the cast film is submerged in lithium-saturated water, causing a solvent-water exchange that creates connected surface pores. In both approaches, the goal is to create connected, low tortuosity pores, maximize interface contacts, and move toward pressure-free SSBs.
Porous Ceramic, Fill it with Lithium Metal
Gregory Hitz, founder and CTO of ION Storage Systems, showed how his company is using a similar approach to make industrial-grade, pressure-free, solid-state cells. ION is also using pores in their SSE, but the main difference is that they eventually fill those pores with the anode material, lithium metal. This approach overcomes the problems associated with swelling during charging, because the lithium just plates into empty pores. ION also uses a bilayer of ceramic LLZO—a thin layer of non-porous material with a thicker porous section on top. The thin layer operates as a separator and SSE, and Hitz says the porous structure results in about 50x the surface area contact compared with planar surfaces.
Their design is initially anode-free. “All of the lithium in our system comes from the cathode,” Hitz explained. It also works with a range of cathode materials, including NMC, LFP, and Li-S systems—Hitz says the design blocks the sulfide shuttle that can otherwise degrade Li-S batteries. He highlighted an important feature of their cathode-agnostic approach. “If you project out the world’s use of cobalt, it’s something like 3x the amount of cobalt that the world has.” Having different options makes sense.
Hitz says ION is taking a customer-first approach, prioritizing the requests of clients. They designed their system to work with lithium ion equipment, “creating a pouch cell that works and looks just like customers are used to, except it’s safe, it has more energy, it’s more temperature-tolerant,” Hitz said. And although the EV market represents perhaps the greatest opportunity, it’s not where ION has focused first. They have agreements with the Department of Defense, developing specialized batteries such as a variant of the BB-2590 (a standardized military battery pack) and fighter jet helmets with integrated batteries. “There are pack-level advantages of solid-state with this approach, even at the very small pack level,” Hitz said, meaning that consumer electronics are also within their purview.
EVs are a major goal, of course. “We have an investor through Toyota,” Hitz reported. He highlighted the different strengths that a solid-state battery needing no compression could provide, including limited thermal management, no need for fire barriers, and higher overall energy density with no pressure jigs. Another advantage, Hitz said, is that “this technology fundamentally enables prismatic packaging,” which will make it amenable to OEMs accustomed to this design. Grid storage is another implementation; “Tenaska is also an investor,” Hitz said.
Partnerships have been important as they try and scale up to TWh-level manufacturing. Saint-Gobain produces a very high purity LLZO powder for them through proprietary methods and is working on an LLZO recycling process. “We expect to be able to take an end of life cell and recapture the garnet out of that cell and put it back into our system,” Hitz said.
There was only one real data slide, a cell operating for 800 C/3 charge and discharge cycles with 99% capacity retention, “without any pressure at all.” Those metrics exceed customer requests. “We’ve actually explicitly told our scientists, please stop doing any dedicated efforts to improving cycle life. We’ve de-prioritized cycle life and pretty much all other performance metrics,” Hitz said. ION has not publicly disclosed energy density for their cells, although they are expected to exceed state-of-the-art lithium-ion batteries.
No More Dendrites, Faster Charge
Eric Wachsman, Maryland University, who co-founded ION Storage Systems with Hitz, echoed much of the same explanation for how a bilayer dense/porous LLZO SSE structure enables pressure-free, high performance batteries. However, he talked about different modifications that push the envelope for fast charging. Wachsman referred to the Department of Energy’s Vehicle Technologies Office goals for EV batteries. Currently, they are setting a goal of 10 mA/cm2 charge density, equating to about a 15-minute charge. Wachsman says their group can accomplish this using the ceramic bilayer electrolyte, anode-free setup.
Wachsman says they can go faster, with some tweaks. At the outset, it makes sense that your separator (which the SSE also functions as) must allow charge-carrying lithium ions to pass through, but not electricity, or the cell would short circuit. However, Wachsman explained that a problem with lithium plating into voids is that the lithium plates first at the interface between SSE and current collector and does not reach into further recesses. Increasing the electronic conductivity of the SSE at this interface, however, enables the lithium to plate more evenly throughout voids. “Without it, it’s only going to form at the triple-phase boundary,” Wachsman said, referring to the convergence of current collector, lithium ions moving through the electrolyte, and anode material.
To overcome this, they designed cells that include a mixed ionic/electronic conductor (MIEC) with significantly increased electronic conductivity, but they only use it in the porous portion of the ceramic. The dense portion maintains a lack of electronic conductivity. This design allowed them to reach a critical current density of at least 100 mA/cm2. “We actually could have gone higher, but my student stopped at that point,” Wachsman said, and based on this, “at least on the lithium side, we can do a 1.5-minute charge.” That statement actually reveals an obstacle. They’re cycling up to 30 mAh/cm2 of lithium, and Wachsman says this is about 6 times state-of-the-art cathode capacities. Therefore to fully take advantage of the design and see sub-2-minute charges, newer cathode materials must be developed.
Wachsman is ebullient about what this architecture will mean for batteries. “My opinion: you use our structure and our materials, dendrites are a thing of the past,” he said, and he sees this creating exceptional battery lifetimes. He said that 18.5 Ah/cm2 lithium cycling, which they could do repeatedly with these materials, corresponds to 3700 full charge/discharge cycles, or one full cycle per day for more than 10 years. Wachsman is doubtful anyone wants to drive this much. The designs really help safety and thermal management, too, because the lithium oxidation rate is limited by being buried in pores. He showed an image of “a full operating battery, in air, with a torch on it.” And based on his batteries’ performance at both higher and lower temperatures, he suggested we go ahead and “imagine not requiring temperature control between -20 and 150 °C.”
Finally, similar to work from Shirley Meng’s group, Wachsman’s group has created sodium-ion SSBs using this same layering of dense and porous SSE portions, but with NASICON instead of LLZO. They solved a problem of low conductivity for NASICON by doping it with magnesium and zinc. “This results in extremely high sodium conduction, enabling room-temperature sodium batteries,” Wachsman explained. And unlike the work of Meng, these batteries operate without external pressure. “We’re cycling 30 mA/cm2 of sodium, at room temperature, no applied pressure,” he said.
Let it Swell: High-Density, Fast-Charging
Cheng-Chieh Chao, VP of QuantumScape, also talked about their solid-state design improvements enabling low- or no-pressure operation. Their design uses a dense ceramic SSE—widely speculated to be LLZO—and an anode-free design. Chao said QuantumScape has put a lot of effort into optimizing multilayer stacking of bilayer cells around a central cathode current collector, such as the 24-layer design in their early A0 design. Volkswagen’s Group’s PowerCo testing lab has cycled these A0 cells more than 1000 times, using only light pressure of about 3.4 atmospheres.
Their newest platform, QSE-5, is an approximately 5-Ah cell that uses a FlexFrame format. This format essentially includes a pre-indented area in the frame of the cell that allows expansion during cycling without swelling beyond the external dimensions. Chao said they have been working in the last couple of years on improving energy density, increasing cathode loadings and packaging efficiency, leading to their newest unit cell, Alpha-2, which will eventually go into the QSE-5 platform. Right now, these cells are going through a range of safety tests, such as nail puncture and thermal stability tests. In one example, Chao showed a prototype cell that was in good shape at 200 °C at the same time that a 2170 cell burst into flames, and the prototype did not ignite even as the temperature increased another 100 degrees.
Chao showed that the Alpha-2 cells could charge 10-80% in under 14 mins with 0.7 atmospheres of pressure—less than atmospheric pressure. One forward-looking slide included projections that their cells will achieve up to 1,000 Wh/L energy density with ~15-minute 10-80% charge times. He said these charging speeds could be even faster with lower energy density designs. He also showed data for fast discharge (50% capacity was usable at 10C discharge) and for low temperature (>70% capacity was available at -25 °C) operation.
