Tesla’s Tape Tests: Optimizing Small Things

By Kyle Proffitt

April 1, 2024 | Steve He, Manager, Cell Development, Tesla Motors, wanted to talk about adhesive tape. He addressed the audience of the 2024 International Battery Seminar to explore the “little things” that can impact battery life and efficiency. We often talk about batteries as though they’re really just 4 components—cathode, separator, anode, and electrolyte, but He reminded us, “there’s a lot more parts in an actual cell than those 4 materials”.

Like Jeff Dahn, another Florida Battery plenary speaker, He brought a focus to the jelly roll architecture of cylindrical cells, but he pointed to some blue adhesive tape used to hold it together and said the composition of that tape can have surprising effects.

The basis for his concern stems from a 2023 Nature Materials paper that studied polyethylene terephthalate (PET) tape undergoing depolymerization to release dimethyl terephthalate (DMT). DMT is problematic because it can act as a redox shuttle and cause substantial self-discharge. This is of course undesirable for a battery that is intended to hold charge.

The reported work, from Michael Metzger’s at Dalhousie University, showed that PET decomposition requires methanol and lithium methoxide, byproducts of the common electrolyte solvent dimethyl carbonate formed in the presence of water and reduction at the anode, respectively. “Without these reagents, you can’t have this reaction,” He said.

He also pointed to an email from Jeff Dahn expressing concern about whether Tesla was using PET in its designs. They apparently were not, but the discussion prompted a deeper exploration. He’s group wondered if polybutylene terephthalate (PBT), a common engineering polymer that Tesla uses in some cell designs, could undergo a similar reaction. He expected it would, as PBT is very similar to PET, having only 2 extra carbons separating the terephthalate groups.

He’s group consulted with Anu Adamson, lead researcher in the Nature Materials study, and adopted the group’s methods. This involved mixing PBT with dimethyl carbonate, spiking in methanol and lithium methoxide, and heating to accelerate the process. They then analyzed the mixture using GC-MS. DMT formation was detected, and the PBT dissolved, discoloring the solution. However, when they repeated the experiment using a full-formulation electrolyte including salts and additives, the DMT was surprisingly no longer formed. He was not yet satisfied, however, desiring to understand the mechanism of PBT protection. After “a lot of additional testing,” their group was able to pinpoint two additives that independently prevent PBT depolymerization. He said these additives (which he did not disclose) work by consuming the methanol or the lithium methoxide.

Electrolyte Matters

He took a slight turn at this point to discuss the merits of electrolytes. “We still think there’s plenty of opportunities in developing conventional electrolytes,” He said. Starting with the salt, he showed the more common variants such as LiPF6 and LiBOB, but pointed to LiFSi as a salt that Tesla really likes. According to He, LiFSi has allowed a 10% reduction in fast-charge time, an increase in thermal runaway temperature, and increased capacity retention. Electrolyte solvents were also discussed, and He expressed interest in the use of esters, saying fast-charge time can be reduced up to 40% compared with carbonate solutions. However, esters suffer from capacity retention loss due to side reactions. He is hopeful that further exploration of different electrolyte solvents will greatly assist rapid charging without capacity sacrifice.

He then turned back to the additives, small molecules usually added in less than one percent by weight. “They have enormous impact on the performance of cells. Our management tends to call additives magic pixie dust, because you just sprinkle some in and it does completely amazing things depending on which additive package you choose,” He said. A report from Jeff Dahn’s group was shown in which additives can drastically affect battery longevity, increasing from 200 to 600 cycles with these simple additions.

He showed a table of 13 different additives (again, structures undisclosed) that Tesla uses to enhance performance, including in the areas of reliability, fast-charge, and longevity, through mechanisms such as metal chelation and electrode passivation.

He also discussed electrolyte cost breakdown: 50% lithium salts, 25% solvent, and 25% additives. Some of those additives come with surcharges related to intellectual property, and this may not be desirable going forward, He said. “As we scale to terawatt hours of production, being single-source in terms of IP-locked additives is going to be a roadblock.” He called on the community to continue work in this area to help in the collective goal of advancing a renewable energy economy.

Finally, we returned to the tape. He said that Tesla conducted several more tests involving actual batteries, both Tesla-developed cells and vendor-supplied cells used in Tesla products (some of which do use PET tape), including in aged conditions. He also highlighted the cells that use LFSi, saying this also reacts with DMT. In all cases thus far (some aging continues), no degradation products have been observed. “Thank goodness,” He said. His thesis point, if not already apparent, was explicitly stated to close his presentation: “small things matter.”