How Are Lithium-Ion Batteries Manufactured?
A lithium-ion cell is not assembled the way a mechanical part is assembled. Its electrochemical behaviour is largely fixed before the cell is ever closed, during electrode coating, drying and formation. That is why two cells with identical datasheets can behave very differently after a few hundred cycles: the datasheet describes the design, not the process discipline that produced it.
This guide walks through the production sequence stage by stage, explains the failure mode that each stage can introduce, and ends with the process questions a buyer can realistically put to a cell or pack supplier. It is written for engineering and procurement readers evaluating cells for industrial duty cycles rather than for laboratory work.

The Five Stages, End to End
Almost every lithium-ion production line, whatever the cell format, follows the same sequence:
- Electrode manufacturing – mixing active materials into a slurry, coating it onto metal foil, drying it and compacting it.
- Cell assembly – cutting and stacking or winding electrodes with a separator, then housing the resulting jelly roll or stack.
- Electrolyte filling and sealing – introducing the electrolyte under controlled humidity and letting it wet the electrode porosity.
- Formation – the first controlled charge, which builds the passivation layer that governs how the cell ages.
- Degassing, aging, testing and grading – letting the cell stabilise, screening out defective units and sorting the rest into matched groups.
Cell format changes the details. Cylindrical cells are wound, prismatic and pouch cells are usually stacked, and each format has its own sealing and degassing method. The sequence and the underlying risks stay the same.
Stage 1: Electrode Manufacturing

This is the stage that sets the electrochemistry. Nothing downstream can repair a badly made electrode - it can only be detected, or missed.
Slurry Mixing
Cathode slurry combines an active material - commonly lithium iron phosphate (LFP), nickel-manganese-cobalt oxide (NMC) or lithium cobalt oxide - with a conductive additive such as carbon black and a binder. In the most widely used cathode route the binder is PVDF dissolved in NMP solvent. Anode slurry is typically graphite in a water-based system using a cellulose thickener and a styrene-butadiene binder.
Mixing is usually staged rather than done in one pass: a dry phase to break up agglomerates, then a wet phase to disperse the solids evenly. Many lines mix under vacuum, because entrained air becomes a coating defect once the slurry reaches the coater.
Binder content is a genuine trade-off rather than a number to copy. Less binder means more room for active material and higher energy density; too little and the coating loses adhesion to the foil as the electrode expands and contracts during cycling. The workable range depends on the particle size and surface area of the specific active material, which is why an experienced line re-qualifies the formulation when it changes material supplier and an inexperienced one does not.
Coating and Drying
Slot-die coating is the dominant method for power cells because it gives good control of coat weight across the web. Coating uniformity matters directly: local variation in electrode loading becomes local variation in capacity, and local variation in capacity becomes uneven current distribution inside a finished cell.
Drying is where two opposite mistakes are possible, and both are worth understanding because neither one announces itself at end-of-line testing.
- Drying too aggressively can drive binder toward the coating surface as solvent evaporates. Adhesion to the current collector weakens, and the coating separates progressively under cycling stress.
- Drying too gently leaves residual moisture in the electrode. Once the cell is filled, that water participates in electrolyte side reactions, generating gas and, in pouch and prismatic formats, swelling.
The practical consequence for a buyer: both defects tend to pass factory acceptance testing and surface months later as capacity fade, impedance rise or swelling. Neither is visible on a specification sheet.
Calendering
The dried electrode is compacted between rollers. The target is a porosity window, not a maximum density: enough void volume for the electrolyte to wet the coating and for lithium ions to move, but enough compaction to get useful energy per unit volume. Over-calendering closes the pore network and drives up impedance; under-calendering wastes volume and can leave the coating poorly bonded.
Stage 2: Cell Assembly
Assembly brings the electrodes and separator together. The dominant risks here are geometric rather than chemical.
Alignment and the N/P Ratio
The ratio of negative to positive electrode capacity - the N/P ratio - must stay above unity at every point of the facing area, not merely on average. Where the anode is locally under-sized or misaligned, arriving lithium has nowhere to intercalate during charging and plates as metallic lithium on the anode surface instead. That costs capacity immediately and raises internal short-circuit risk over time.
This is why the anode is deliberately made slightly larger than the cathode, with a controlled overhang, and why separator wrinkling and electrode misregistration during stacking or winding are treated as serious defects rather than cosmetic ones. Like the drying defects above, misalignment is typically introduced during assembly, survives initial testing and manifests during cycling.
Electrolyte Filling and Dry Rooms
The common electrolyte salt, LiPF₆, reacts with water to produce hydrofluoric acid, which attacks cell internals. Electrolyte handling therefore takes place in dry rooms held at a far lower dewpoint than ordinary cleanrooms - conditions strict enough that human breath is a meaningful moisture source, which is why operators in filling areas are gowned and masked accordingly.
Filling itself is usually assisted by vacuum to pull electrolyte into the electrode porosity. Complete wetting is not instantaneous: it can take hours or longer depending on porosity, electrode thickness and cell geometry, and some lines hold cells at slightly elevated temperature to lower electrolyte viscosity and shorten the wait. A line that shortens wetting time without validating it ships cells with dry regions, which behave as locally higher-impedance zones for the rest of their life.
Stages 3–5: Formation, Degassing and Aging

Cell finishing is the slowest part of the line and, in most plants, the bottleneck that determines throughput. It is also the stage most exposed to schedule pressure.
What Formation Actually Does
During the first controlled charge, electrolyte components reduce at the anode surface and build the solid electrolyte interphase (SEI): a passivating film that must be stable enough to stop continuous electrolyte decomposition while still conducting lithium ions. The character of that film depends on the current, temperature, voltage window and rest steps used during formation, which is why formation is treated as intellectual property by serious cell makers rather than as a fixed recipe.
Formation protocols are an active research area precisely because they trade production time against cell lifetime. Weng and co-authors published a controlled comparison of fast and baseline formation protocols on NMC/graphite pouch cells in Joule (2021), and subsequent studies have reported that cycle life can differ substantially between protocols on otherwise identical cells. Reported magnitudes vary with chemistry, cell format and protocol, so treat any single figure as specific to the study that produced it rather than as an industry constant.
Why Aging Cannot Be Compressed Away
After formation, cells rest - in many plants for days to weeks - before final grading. Three things happen during that rest: the SEI stabilises, generated gas is removed or redistributed, and, most usefully for the buyer, internal defects reveal themselves as abnormal self-discharge. Open-circuit voltage is tracked across the rest period, and cells that drift are pulled.
Aging is expensive: it is working capital sitting in a warehouse and floor space that produces nothing. That makes it the most tempting step to shorten when a plant is behind schedule or competing on price. Shortening it does not create defects, but it removes the screen that would have caught them, so the defects ship instead. If you take one process question away from this article, this is the one worth asking.
What Today's Price Environment Does to Process Discipline
Cell pricing has been falling steeply, and understanding by how much helps explain the pressure on the production side. According to BloombergNEF's 2025 battery price survey, published in December 2025, the global volume-weighted average lithium-ion pack price fell 8% year on year to $108/kWh, a record low, even though lithium and cobalt prices rose during the year. BNEF attributed the decline to persistent cell overcapacity, aggressive competition and continued migration to LFP.
The same survey reports meaningful spread beneath that average, which is more useful to a buyer than the headline:
| Segment or region (BNEF, 2025) | Average pack price |
|---|---|
| All segments, global average | $108/kWh |
| LFP packs, all segments | $81/kWh |
| NMC packs, all segments | $128/kWh |
| Stationary storage packs | $70/kWh |
| Battery electric vehicle packs | $99/kWh |
| China average | $84/kWh |
BNEF also reported that average pack prices in North America and Europe ran 44% and 56% above the Chinese average, which it linked to higher local production costs and reliance on imports. These are market-wide averages across many applications and are not quotes for industrial traction packs; use them to understand the direction of pressure, not to benchmark a specific offer.
The relevance to manufacturing is straightforward. When margin compresses, the steps that cost time and floor space without producing visible output - wetting, formation, aging - are the ones under pressure. A quotation well below the market band is not automatically a warning sign, because genuine scale and process maturity produce genuine cost advantage. But it does mean the difference between efficiency and corner-cutting has to be established by asking, not assumed.
Manufacturing Defects and How They Show Up in the Field
The table below maps the defects introduced in each stage to what a buyer eventually observes. The middle column is the one that matters for qualification: most of these defects are not reliably caught by end-of-line testing.
| Defect | Where it originates | How it appears in service |
|---|---|---|
| Metallic particle contamination | Equipment wear, incoming material impurities | Internal soft or hard shorts; abnormal self-discharge |
| Binder migration | Drying profile too aggressive | Coating delamination and accelerating capacity fade |
| Residual moisture | Under-drying, or humidity excursions before filling | Gas generation and cell swelling |
| Coating cracks | Drying and calendering stress | Rising internal impedance |
| Electrode misalignment or separator wrinkling | Stacking or winding | Localised lithium plating; short-circuit risk |
| Incomplete electrolyte wetting | Filling and soak time cut short | Uneven impedance and cell-to-cell spread within a pack |
| Unstable SEI | Unvalidated formation protocol | Faster than expected calendar and cycle ageing |
| Undetected self-discharge outliers | Aging period shortened | A minority of cells drifting out of balance, pulling down the pack |
What to Ask Before You Qualify a Cell Supply

These questions are answerable by any manufacturer with real process control, and the quality of the answer tells you more than the specification sheet does. Ask for evidence, not assurances.
- What is your formation protocol, and what data supports it? A supplier that treats formation as a validated parameter set will describe how it was qualified. One that cannot describe it at all is buying cells rather than making them - which is fine, provided you know it.
- How long do cells age before grading, and what OCV drift limit pulls a cell? This is the single most informative throughput-versus-quality question.
- How is cell-to-cell matching done? Ask which parameters are used for grading and how tight the bands are. Pack lifetime is driven by spread, not by the average cell.
- What dry room conditions apply at filling, and how are excursions logged? Ask to see the monitoring record, not the design specification.
- What is the incoming inspection regime for active materials and foil? Metallic contamination is largely a supply-chain and equipment-wear problem.
- Is production traceable to cell serial number? Without batch and line traceability, a field failure cannot be root-caused and a containment cannot be scoped.
- Has the cell line or formulation changed since the qualification samples? A silent material or line change invalidates the qualification you paid for.
- What testing accompanies the shipment? Capacity, internal resistance and OCV records per batch are reasonable to request.
Cell Manufacturing Versus Pack Building
These are two different businesses, and it is worth being clear about which one you are talking to, because it determines which questions are even answerable.
A cell manufacturer controls everything described above: formulation, coating, formation, aging and grading. A pack builder selects and sources cells, then controls what happens around them - series and parallel configuration, busbar and interconnect design, the BMS and its balancing and protection logic, thermal design, enclosure sealing and mechanical robustness, communication interfaces, and the acceptance testing applied to incoming cells.
Both layers can cause a field failure and they fail in different ways. Cell-level process defects show up as spread and premature fade; pack-level problems show up as connection heating, balancing faults, ingress or vibration damage. When a battery underperforms, the first diagnostic question is which layer it came from.
Polinovel operates at the pack level: our custom battery pack design service covers voltage and capacity configuration, BMS and communication selection, enclosure and IP rating, and mechanical fit to an existing battery compartment. For duty cycles such as multi-shift material handling, where cell-to-cell consistency is what determines whether a pack finishes the third shift, that means incoming cell selection and acceptance criteria are part of the pack specification, not an afterthought - the same logic applies across our electric forklift battery packs.
Certifications and What They Cover
Certification is evidence of conformity to a defined test regime, not a general quality rating. Three references come up most often for lithium cells and batteries:
- UN 38.3 - the transport test series from the UN Manual of Tests and Criteria, required for shipping lithium cells and batteries. It addresses transport safety, not service life.
- IEC 62133 - safety requirements for portable sealed secondary cells and batteries.
- UL 1642 - a safety standard for lithium cells.
When you review a certificate, check the specific model and configuration it covers, the issuing laboratory and the edition of the standard applied. A certificate issued against a different cell or a superseded edition does not cover the product you are buying. Confirm the current edition of each standard with the issuing body before relying on it in a specification.
The Practical Takeaway
The manufacturing story matters to a buyer for one narrow reason: most of the process defects that determine whether a cell meets its rated life are invisible at delivery and only appear after the warranty conversation has become expensive. You cannot audit a cell by reading its datasheet, but you can ask about formation, aging, grading and traceability, and you can judge a supplier by whether those questions produce data or deflection.


