Solid-state cell designs don’t converge on one negative electrode — manufacturers are actively pursuing at least four distinct architectures, trading off energy density, manufacturing complexity, and dendrite/plating risk against each other. This page works through the spectrum from conventional graphite through to fully anode-free designs.
Graphite / silicon composites. The most conservative path, and the one used by several first-generation solid-state designs: keep a conventional intercalation-type graphite or graphite-silicon composite negative electrode, and get the solid-state benefit purely from the separator and cathode side. This sidesteps the hardest problems — lithium metal’s dendrite risk and the volume change of a plating/stripping anode — at the cost of giving up most of the energy-density upside that motivates solid-state development in the first place, since the negative electrode still needs its usual share of inactive mass and volume.
Bulk thin lithium foil (20–50 µm). Using a thin lithium metal foil captures most of the energy-density benefit without the full complexity of an anode-free design, but foil below about 50 µm is difficult to handle: thin extruded lithium tends to have poor structural integrity, a rough surface, and a resistive native passivation layer straight out of manufacturing (thermally-evaporated lithium thickness study). The harder problem shows up during cycling rather than manufacturing: stripping lithium from the foil faster than it can be replenished by diffusion opens interfacial voids at the lithium/electrolyte contact, and those voids — not dendrite growth through the bulk — are what drives a lot of solid-state lithium-metal cell failure (Science Advances, void formation study). Maintaining enough stack pressure to keep the interface in contact as it strips is critical; one study found around 20 MPa to be the optimum cycling pressure for maintaining good lithium/electrolyte contact over life.
Anode-free / zero-excess lithium. The most aggressive path removes the lithium reservoir entirely at assembly: the cell is built with a bare current collector on the negative side, and the lithium metal anode forms in situ, plated out of the cathode’s own lithium inventory during the first formation charge. Samsung SDI’s landmark demonstration used a 5 µm silver-carbon (Ag–C) nanocomposite interlayer on the current collector, which regulates where and how uniformly that first-charge lithium plates; the resulting pouch cell reached roughly 900 Wh/L and cycled past 1,000 times without a large loss of capacity (Samsung SDI / SAIT, Nature Energy 2020). QuantumScape’s separator-contact architecture follows the same anode-free principle from the other side: rather than an interlayer, it relies on its ceramic solid separator being stable enough in direct contact with plated lithium metal to control plating uniformity without a dedicated host layer (QuantumScape). Either way, that first formation charge is the highest-risk single event in the cell’s life: it plates the entire anode inventory at once, from nothing, and any non-uniformity in current distribution at that moment becomes locked-in non-uniform lithium morphology for every subsequent cycle — which is why formation protocol (rate, pressure, temperature) matters more for anode-free designs than for almost any other cell architecture.
A related but more incremental version of zero-excess design pairs an anode-free negative electrode with LMFP rather than a nickel-rich cathode, trading peak energy density for lower cost and an easier-to-manage voltage window. This configuration has the same basic advantages as any anode-free design — higher energy density through reduced stack mass, a simplified manufacturing process, and reduced cost — but shares the same formation-charge risk described above, and the pilot-line work on it to date is a single scalable low-cost demonstration rather than an established commercial track record (Bree et al., ChemRxiv 2025).
Silicon-dominant micro-silicon anodes. A fourth path skips lithium metal altogether and instead pushes silicon as far as it will go: a team led by Shirley Meng at UC San Diego demonstrated a 99.9 wt% micro-silicon anode — no carbon, no binder — cycled stably against a sulfide solid electrolyte (Tan, Chen, Meng et al., Science 2021). The key finding was that the sulfide electrolyte’s own interfacial passivation behaviour suppresses the continuous interfacial growth and irreversible lithium loss that normally makes carbon-free, binder-free silicon impractical in a liquid-electrolyte cell, and using cheaper, less-processed micron-scale silicon rather than nanosilicon removes a major cost barrier to silicon-dominant anodes generally.
See also: Solid State Battery, Grain Boundary Dendrite Growth in Solid-State Electrolytes, Cathode-Electrolyte Interface & Catholyte Engineering