Solid-State Batteries: Why They're So Hard to Build

Solid-State Batteries: Why Swapping a Liquid for a Solid Turned Out to Be So Hard

Toyota first said it would put a solid-state battery into a production car in 2020. The date slipped to 2023, then to 2026, and the company's current target is 2027 or 2028, most likely on a Lexus. Nissan is aiming for fiscal year 2028. Samsung SDI and CATL are both circling 2027. Every one of these companies has real laboratory results behind its claims. None of them has a solid-state battery on a showroom floor. That gap between demonstrated science and delivered product is the actual story here, more than any single number about range or charging speed.

The reason the gap exists starts with something almost boring: the liquid sitting inside every lithium-ion cell on the road today.That liquid, the electrolyte, is a mixture of organic solvents and lithium salts whose only job is to let lithium ions shuttle between the positive electrode (the cathode) and the negative electrode (the anode) while charging and discharging. It does that job well, which is why lithium-ion has dominated the battery industry since Sony commercialized it in 1991. It is also flammable, and under the wrong conditions  a puncture, an internal short, sustained overheating  it can ignite and drive the runaway chain reaction known as thermal runaway.

Replacing that liquid with a solid material sounds, on paper, like changing one ingredient in a recipe. Instead of a wet electrolyte and a porous plastic separator soaked in it, you use a single solid layer that conducts lithium ions on its own. Researchers have known this was theoretically possible for decades. What nobody quite appreciated, until they tried to build it at scale, is that a battery isn't really a material problem. It's an assembly problem, and solids are far less forgiving of imperfect assembly than liquids ever were.

Inside a conventional cell, electrons and lithium ions travel by different routes for the same reason a couple splitting a hotel bill might take different paths to the front desk. During discharge, lithium atoms in the anode give up an electron, becoming lithium ions. The ions move through the electrolyte toward the cathode. The electrons can't follow them through that electrolyte  it's built specifically to block electronic conduction while allowing ionic conduction  so they take the external circuit instead, through the wire, the device, and back. That detour is where the energy actually gets used: it's what lights a screen or turns a motor. During charging, an outside power source reverses the whole sequence, pushing ions back into the anode and re-forming metallic or intercalated lithium there. A separator, typically a thin polymer membrane, keeps the electrodes from touching directly and short-circuiting; current collectors, thin sheets of copper and aluminum, carry electrons to and from the external circuit.

A solid-state battery keeps this same basic choreography. What changes is the electrolyte's physical form, and "solid-state battery" is not one design so much as a label covering several very different chemistries that share only that one property. Some use ceramic oxides. Some use sulfide glasses. Some use polymers, or halide compounds, or blends of more than one. They conduct ions and block electrons the same way a liquid electrolyte does, in principle. In practice, getting a rigid solid to do this reliably, cheaply, and for thousands of cycles has turned into one of the more stubborn problems in applied materials science.

Safety is the argument used most often to justify the effort, and it deserves a precise treatment rather than a marketing one. Removing a flammable liquid electrolyte can reduce certain failure pathways, particularly the kind of vapor-phase ignition that follows a puncture or a severe short. It does not make the surrounding cell immune to thermal or mechanical failure. A solid-state cell still contains a high-energy-density cathode, still generates heat under abuse conditions, and in sulfide chemistries can release toxic hydrogen sulfide gas if the electrolyte reacts with moisture. Safety in these cells is better described as a shift in which failure modes matter most, not as their elimination.

Energy density is the more concrete prize, and it's the one actually driving billions of dollars in investment. A conventional lithium-ion cell today typically delivers somewhere in the range of 250 to 300 watt-hours per kilogram at the cell level. Solid electrolytes open the door to using lithium metal as the anode instead of the graphite or silicon-graphite blends used now, and lithium metal packs far more energy into far less mass, because you're no longer carrying inactive host material for the lithium to sit inside. That's the real reason solid-state and lithium-metal keep appearing together in the same sentence: the solid electrolyte's chemical and mechanical stability is what finally makes a lithium-metal anode survivable across many cycles, something liquid electrolytes struggle with because they react continuously with bare lithium metal's surface.

The catch is that a laboratory pouch cell reporting 450 or 500 watt-hours per kilogram is not the same thing as a battery pack in a car. CATL has reported lab-scale sulfide cells reaching roughly 500 Wh/kg, but the company itself places that technology at an early readiness level, still being validated outside full vehicle integration. Cell-level numbers shrink once you add packaging, cooling, structural support, and the electronics that manage hundreds of cells working together. Pack-level energy density, the number that actually determines a vehicle's range, is consistently and substantially lower than the headline lab figure, for solid-state chemistries exactly as it always has been for lithium-ion.

The deepest problem, though, and the one that explains most of the delayed timelines, lives somewhere less obvious than the electrolyte material itself: the interface where two solids touch.

A liquid electrolyte is forgiving because it's a liquid. It flows into every microscopic crevice on an electrode's surface and maintains full physical contact no matter how rough that surface is. Two rigid solids pressed together cannot do this. Even electrode and electrolyte surfaces polished to a mirror finish leave microscopic gaps, voids, and points of uneven contact when stacked. Lithium ions moving across that boundary encounter resistance at every gap, and as the electrode expands and contracts during cycling  which all electrodes do, some by a lot  those gaps grow, cracks propagate, and the electrolyte can react chemically with the electrode material at the boundary, forming resistive byproducts that choke ion transport further. An electrolyte that looks excellent when its ionic conductivity is measured in isolation can perform poorly once it's sandwiched against a real cathode or anode, because the number that matters in a working cell isn't bulk conductivity, it's interfacial resistance, and that's a property of the junction, not of either material alone.

Dendrites make this worse in a way that undercuts one of the field's oldest assumptions. For years, the working theory was that a rigid ceramic or glass electrolyte would simply be too mechanically strong for lithium filaments to force their way through, unlike the soft polymer separators used with liquid electrolytes. That theory has not held up. Lithium metal, it turns out, can still penetrate solid electrolytes, sometimes at current densities well under what a commercial cell would need to charge at a useful rate. A team from Shanghai Jiao Tong University and the Max Planck Institute for Sustainable Materials, publishing in Nature in April 2026, used cryogenic electron microscopy to watch lithium fill microscopic cracks at the tips of defects inside a garnet-type ceramic electrolyte, growing outward from flaws that were already there rather than nucleating cleanly at a surface. A separate Nature paper published a few months later, using a purpose-built compression method on the same class of garnet material, found that dendrites can originate from inside the electrolyte itself during long cycling, not only at its surface, contradicting an assumption that had shaped a decade of mitigation strategies built around surface treatments alone. Exactly which mechanism dominates under which conditions, in which electrolyte chemistry, remains an active and unsettled question  which matters, because you can't reliably engineer around a failure mode whose origin you don't fully agree on.

None of the electrolyte families on offer sidesteps every version of this problem, which is why the field hasn't converged on one winner. Sulfide glasses conduct ions almost as well as a liquid, often in the range of one thousandth to one hundredth of a siemens per centimeter, and they're soft enough to press into good physical contact with electrodes. They're also unstable in humid air, reacting with trace moisture to release hydrogen sulfide, which means manufacturing has to happen in dry rooms far stricter than a normal battery gigafactory. Oxide ceramics like the garnet LLZO are chemically stable against lithium metal and don't share that moisture problem, but they require sintering at temperatures above a thousand degrees Celsius and are brittle enough that handling large, thin sheets of the stuff without cracking it is its own manufacturing discipline. Halide electrolytes have emerged more recently as a compromise: reasonably good conductivity, strong stability against high-voltage cathode materials, and better mechanical give than oxides, but many are themselves moisture-sensitive and thermodynamically unstable in direct contact with lithium metal, requiring a thin protective layer between the two. Polymer electrolytes are the easiest to process into thin, flexible sheets using equipment the industry already understands, but their room-temperature ionic conductivity lags well behind the others, and their electrochemical stability window is narrower, limiting the cathode chemistries they can pair with. Composite designs that blend two of these families are an increasingly common attempt to borrow one material's strength to cover another's weakness, though every hybrid adds a second, third, or fourth internal interface, which reopens the interfacial resistance problem the whole approach was meant to solve.

Lithium-ion, meanwhile, is not standing still while all of this gets sorted out, and it's worth being honest about how large an advantage that gives it. It has three decades of manufacturing refinement behind it, gigafactory-scale supply chains for graphite, electrolyte solvents, and cathode materials, mature battery-management electronics, established recycling pathways, and steadily falling costs driven by volume rather than breakthrough. A new chemistry doesn't just need to beat lithium-ion on a spec sheet. It needs to out-compete an enormous, well-optimized industrial system that has had thirty years to work the yield problems out of its own process. That's a different and harder competition than the one implied by a lab result.

This is where the distinction between semi-solid and all-solid-state stops being pedantic and starts being the difference between what's actually on the road and what's still on a pilot line. Several vehicles already sold in China, including NIO's 150 kWh pack and the IM L6, are marketed using solid-state language, but they use a semi-solid or "condensed" design that still contains a small amount of liquid or gel electrolyte alongside solid components. That's a legitimate and useful engineering step, and it's shipping today. It is not the same product as an all-solid-state cell with no liquid at all, which is still confined to pilot lines and prototype packs. China's regulators are expected to publish a national technical standard for solid-state batteries around mid-2026, partly to formalize this distinction after enough products blurred it in marketing that the terminology needed policing.

The gap between a promising pilot cell and a reliable product built by the million is where most of the remaining difficulty actually sits, and it's less glamorous than chemistry but arguably harder. CATL has described its own path from 20 amp-hour laboratory samples to 60 amp-hour automotive-grade cells as an engineering challenge in its own right, separate from getting the chemistry to work at all; the company currently rates its all-solid-state technology at an early readiness level and is targeting a roughly 2-gigawatt-hour pilot line at its Yibin facility for commissioning in the second half of 2026, with small-batch trial production not expected before 2027. In July 2026, multiple outlets reported that CATL had launched a 15-gigawatt-hour mass production line for all-solid-state cells; battery-market analysts flagged the claim as very likely false within days, since it wildly outpaced the company's own stated pilot-scale timeline. That episode is a reasonably clean illustration of how easily an ambitious roadmap curdles into an inflated headline once it leaves the industry press and hits a wider audience.

Pressure deserves its own mention, because it's one of those engineering details that sounds trivial until you scale it. Many sulfide and oxide cell designs need sustained mechanical compression, applied continuously, to keep their rigid layers in close enough contact to conduct ions efficiently. In a lab, that's a clamp on a test rig. In a vehicle, it means the pack has to carry hardware capable of maintaining that pressure evenly across every cell, for years, through vibration, temperature swings, and crash-safety loads, without the stack shifting or cracking. Engineers building sulfide stacks have described the ceramic layers themselves as behaving less like a flexible sheet and more like thin, glazed stone: dense, hard, and prone to cracking rather than bending under uneven load. Whatever hardware solves that problem adds weight and cost, which eats into the very energy-density advantage the solid electrolyte was supposed to deliver in the first place.

Even where solid-state cells clear every one of these hurdles, it isn't obvious the environmental case is as clean as the branding suggests. A life-cycle analysis published by researchers at Purdue University in March 2026 modeled a full recycling process for an all-solid-state cell built around a nickel-cobalt-manganese cathode, a lithium-metal anode, and a sulfide electrolyte, and compared it against an equivalent lithium-ion cell. On a per-cell basis, the solid-state design showed higher energy use and climate impact during recycling, not lower, and the study found no significant life-cycle advantage over lithium-ion at the cell level unless the technology's weight savings were substantial enough to meaningfully cut a vehicle's energy use over its lifetime. Lithium-metal anodes also don't fit into existing recycling infrastructure built around graphite and liquid electrolytes, which is part of why Factorial Energy, a solid-state battery developer, began working with a South Korean recycler in 2023 specifically to build a lithium-metal recovery process from scratch. Solid-state batteries may end up more sustainable than lithium-ion in the vehicles that benefit most from lighter packs. They are not automatically more sustainable simply by virtue of containing no liquid.

Beyond cars, the same properties that make solid-state attractive for EVs, high energy density per unit of volume and reduced flammability risk, matter even more in applications where space and mass are tightly constrained and the cost premium is easier to absorb. Samsung SDI showed pouch-format solid-state cells built for humanoid robots at South Korea's InterBattery exhibition in March 2026, framing robotics as a market that could adopt the technology before mainstream EVs do, precisely because robot manufacturers care less about matching lithium-ion's price per kilowatt-hour and more about squeezing energy into a small, oddly shaped chassis. Drones, implantable and wearable medical devices, and certain aerospace and defense applications share that same calculus, and are likely to see solid-state cells in real products well before an affordable family EV does.

None of this converges on a single metric that determines who wins. A cell with extraordinary energy density but a low manufacturing yield is not commercially better than a slightly less exotic cell that a factory can actually build a million times over without a high defect rate. The companies furthest along, Toyota with Idemitsu Kosan's solid-electrolyte pilot plant now under construction, QuantumScape running its Eagle Line and shipping sample cells to Volkswagen for real-world testing, Samsung SDI running roll-to-roll pilot production instead of slower batch processing, are all, in effect, racing to solve a manufacturing problem that happens to be dressed up as a chemistry problem. Cost, defect rates, cycle life under real driving conditions, and the ability to maintain interfacial contact for a decade of use will decide the outcome far more than any single lab-reported energy density figure.

Which brings the question back to where it started. Replacing a liquid electrolyte with a solid one is not difficult because nobody can find a material that conducts lithium ions; several exist, and have for years. It's difficult because a battery is a system of things touching other things under stress, repeatedly, for a decade, and solids are catastrophically less tolerant of imperfect contact than the liquid they're meant to replace. That tolerance gap is still visible at the most fundamental level available to science right now: researchers using some of the best cryogenic imaging tools that exist still disagree, based on genuinely new 2026 data, about whether the lithium filaments that eventually short-circuit these cells start growing from defects buried inside a "perfect" ceramic or from its surface. Until that argument settles, engineers are, in a real sense, still trying to defect-proof a material against a failure mode whose birthplace inside the crystal they haven't fully located.

YOU MAY ALSO LIKE

Scientists Are Building Harry Potter's Invisibility Cloak

Meet the Computer That Has Never Touched an Electron