materials · energy · manufacturing · family: it worked in the lab
one pinhole ruinsthe solid-state battery
Thin ceramic separators for solid-state batteries cannot be manufactured defect-free at automotive scale
Problem statement
Solid-state batteries promise to replace flammable liquid electrolytes with solid ceramic or glass separators, enabling higher energy densities (>400 Wh/kg vs. about 280 Wh/kg for the best commercial Li-ion), faster charging, and inherent safety. QuantumScape (ten years and $300M spent to reach prototype cells) has sent sample cells with its ceramic separator to an automaker for testing, but manufacturing defect-free thin ceramic sheets at the sizes, volumes, and speeds required for automotive battery production has not been achieved after at least ten years of development. Sakti3 (acquired by Dyson for $90M in 2015) was written down by Dyson by more than $60M. As of 2025, several manufacturers were qualifying cells and starting low-volume manufacturing lines. Toyota has said it will offer cars with solid-state batteries as early as 2027, but most researchers believe commercialization is still years away.
Why this matters
Solid-state batteries could enable EVs with 500+ mile range, 15-minute charging, and reduced fire risk. Toyota, Volkswagen (through its battery arm PowerCo), Nio, and IM Motors have announced solid-state battery plans, but most researchers expect commercialization to take years. If the ceramic manufacturing barrier is not resolved, the EV industry will remain dependent on incremental lithium-ion improvements, leaving range anxiety and charging speed as persistent adoption barriers.
What’s been tried and why it hasn’t worked
QuantumScape's approach uses a proprietary ceramic separator with a lithium-metal anode. At small scale, the ceramic suppresses lithium dendrite formation (the needle-like growths that short-circuit batteries). At large format (automotive cell sizes), pinholes, cracks, or uneven surface topography in the solid electrolyte can become starting points for dendrites. It took QuantumScape two years just to develop a continuous ceramic manufacturing process (rather than batch-baking individual pieces like pottery), and cells stack tens to hundreds of layers with no redundancy, so every layer has to be reliable. Dyson bought Sakti3 and later wrote down the investment; the cited report does not say what went wrong technically. Sulfide-based solid electrolytes (Toyota, Solid Power) conduct ions well at room temperature but face interface problems and lithium dendrite growth.
What would unlock progress
Ceramic processing innovations from other industries (MLCC multilayer ceramic capacitors, SOFC solid oxide fuel cells) could provide manufacturing approaches for thin, defect-free ceramic sheets at scale. Composite electrolytes (ceramic particles in polymer matrices) could tolerate defects better than pure ceramics while retaining most of the performance benefits. In-line defect detection systems that can identify pinholes and cracks in ceramic separators at production speed (meters per minute) would enable the inline quality control that manufacturers still need to adopt. Novel ceramic compositions with higher fracture toughness (garnet-type, NASICON-type) could reduce sensitivity to mechanical defects.
Entry points for student teams
A team could prototype a quality inspection system for thin ceramic sheets using optical coherence tomography or other non-destructive techniques, testing detection limits for pinhole and crack defects relevant to battery separator performance. A materials-focused team could characterize the defect tolerance of composite electrolytes (ceramic-in-polymer) compared to pure ceramic separators, quantifying how defect density affects dendrite suppression. Relevant disciplines: materials science, ceramics engineering, manufacturing engineering, quality systems.
Genes
Structural cousins: stuck the same way, in other fields
Sources
PNAS, "Solid-state batteries could revolutionize EVs and more—if they can surmount technical and financial hurdles," 2024, Battery Technology Online, "VW, QuantumScape, and the Challenges of Scaling Solid-State Batteries," 2024; KLA, "Resolving Production Challenges that Hinder Advancement in Solid-State Batteries," 2025; InsideEVs, "Dyson Dumps Solid-State Battery Developer Sakti3," 2018; accessed 2026-02-20 go to source ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
QuantumScape ($1.8B+ invested) and Sakti3/Dyson ($105M invested, written off) represent two fundamentally different approaches (thick ceramic separator vs. thin-film deposition) that both failed at scale. The challenge is universal: Solid Power, ProLogium, Samsung SDI, and Toyota all face variants of the same ceramic/glass manufacturing scalability problem. Related: energy-ev-battery-cold-climate-performance (different battery challenge — performance rather than manufacturing), manufacturing-am-metal-part-qualification-barrier (shares the theme of qualification barriers when scaling a new manufacturing process). The almost-cluster signal `materials` + `process` + `sensing` is relevant (materials challenge, process innovation needed, in-line sensing for quality control).
Reconciliation 2026-09-29: Fact repair after a site review read the whole write-up against all four cited sources. Each was fetched and read in full on 2026-09-29. The PNAS article (by John Carey, published 19 December 2024) was read from the open copy at https://europepmc.org/articles/PMC11670104 because the PNAS address blocks automated access, and its citation record was read from https://api.crossref.org/works/10.1073/pnas.2425219121. The KLA article (8 August 2025) was read at https://www.kla.com/advance/innovation/resolving-production-challenges-that-hinder-advancement-in-solid-state-batteries. The Battery Technology Online article (12 July 2024) and the InsideEVs article (27 September 2018) were read from Internet Archive copies of their original addresses, https://www.batterytechonline.com/ev-batteries/vw-quantumscape-and-the-challenges-of-scaling-solid-state-batteries and https://insideevs.com/news/339941/dyson-dumps-solid-state-battery-developer-sakti3/. (1) Source line: the PNAS title was missing the words "and more" and the year was given as 2025; the record says 2024. The KLA title was shortened; it now matches the page. (2) The text gave "~250 Wh/kg for Li-ion." The PNAS article says today's best commercial lithium-ion batteries reach about 280 Wh/kg. (3) The text said QuantumScape had "$1.8B+ raised, stock down >95% from peak" and "15 years of development." None of the sources gives these. The PNAS article says the company spent five years and $100 million choosing its electrolyte material and another five years and $200 million building prototypes; the text now gives those totals. (4) The text said QuantumScape's separator "works at postage-stamp scale." In the PNAS article the postage-stamp remark is a general comment by an ARPA-E program director. The article says QuantumScape has sent sample cells, each equivalent to one EV battery cell, to an automaker for testing; the text now says that. (5) The text gave Sakti3 as "$105M." The InsideEVs article says Dyson bought Sakti3 in 2015 for $90 million and recorded a charge of £46 million (over $60 million). The account of why Sakti3 failed (thin-film deposition borrowed from semiconductor fabrication) is in none of the sources and was removed from the body. (6) "No solid-state battery company has reached gigawatt-hour scale production" is not stated in any source; the KLA article says several manufacturers are qualifying cells and starting low-volume lines, and the text now says that. (7) Toyota's delays "from 2020 to 2023 to 2026 to 2028" are in none of the sources; the PNAS article says Toyota has claimed cars as early as 2027 and that most researchers think commercialization is years away. (8) Removed as unsupported: the $400B market projection for 2030, the statement that every major automaker has a program (BMW, Samsung SDI and CATL are not named in any source), and the claim about the three primary consumer barriers. "Elimination" of fire risk now reads "reduced fire risk," which is what the sources say. (9) The text said every pinhole or crack becomes a dendrite site and that such defects are undetectable in the lab and inevitable in production. The KLA article says dendrites can be caused by pinholes, cracks or uneven topography; the text now says that. "2+ years" for the continuous ceramic process now reads "two years," as in the PNAS article. The claim about impedance and stress building with each layer was replaced with the KLA statement that cells have tens to hundreds of layers with no redundancy. (10) The text said sulfide electrolytes are air-sensitive and need moisture-free factories, and named Samsung SDI. None of the sources says this. The PNAS article names Toyota and Solid Power as sulfide users, and the KLA article says sulfides face interface problems and dendrite growth; the text now follows them. (11) In "What Would Unlock Progress," the statement that capacitor and fuel-cell makers work at smaller areas and lower throughput was removed, and "quality control that currently does not exist" was changed, because the KLA article says KLA is working with manufacturers on optical inspection of solid-electrolyte separators. Note on the working notes above: the first paragraph of Source Notes is kept as first written, and its figures ($1.8B+, $105M), its phrase "thick ceramic separator," its statement that both companies "failed at scale," and its mention of ProLogium and Samsung SDI are not supported by the sources; read items 3, 5 and 10 in their place. Confirmed against the sources and left as written: energy density above 400 Wh/kg for lithium-metal designs; QuantumScape's ceramic separator with a lithium-metal anode; the continuous process replacing batch baking "like pottery"; range of 500 miles or more (the PNAS article reports Toyota's claim of 750 miles) and 15-minute charging as targets; the titles and years of the Battery Technology Online and InsideEVs articles. NOT checked: the proposals in "What Would Unlock Progress" (capacitor and fuel-cell methods, composite electrolytes, garnet and NASICON compositions) and the student entry points are suggestions that the sources do not make; they were left as written. The H1 and the short title were not changed. Genes untouched.