enEnglish

Fundamentals of Lithium Ion Battery Technology

Nov 11, 2025

Leave a message

 

Lithium ion batteries don't really "generate" power in the way most people think about it. What they do is store electrical energy through reversible electrochemical reactions, then release it when the external circuit demands current. The confusion about this comes up a lot in design meetings, especially when someone's trying to size a battery pack for the first time.

Two things happen during discharge. First, lithium ions migrate from the negative electrode (anode) through the electrolyte and separator to the positive electrode (cathode). Second, electrons flow through the external circuit from anode to cathode, doing useful work. During charging, you reverse the process by applying an external voltage that forces ions and electrons to move in the opposite direction.

The voltage a cell produces depends entirely on the electrode materials you pick and their respective electrochemical potentials. A fresh cell sitting on the shelf with no load will show its open-circuit voltage-typically around 3.6 to 3.7V for most lithium ion chemistries, though this number moves around depending on state of charge and temperature. Once you connect a load and start drawing current, the voltage drops because of internal resistance. How much it drops tells you a lot about the cell's health.

 

Lithium Ion Battery Technology

 

Cell Chemistry Fundamentals

 

All lithium ion cells share the same basic operating principle, but chemistries vary widely. The cathode material largely determines the cell's performance characteristics-energy density, power capability, cycle life, thermal stability, and cost.

Layered oxide cathodes were the first commercial chemistry. Sony introduced them back in 1991 with LiCoO₂ (lithium cobalt oxide), which is still used in consumer electronics where energy density matters more than cost or safety margins. These cells pack roughly 150-200 Wh/kg at the cell level. Cobalt's expensive though, and the chemistry gets unstable above 150°C. We've seen thermal runaway start at temperatures as low as 130°C in abused cells.

The drive for better safety and lower cost led to LiMn₂O₄ (lithium manganese oxide) in the mid-1990s. Manganese is dirt cheap and the spinel structure is inherently more stable. These cells won't run away until you get past 250°C typically. The tradeoff? Energy density drops to 100-120 Wh/kg, and manganese dissolves into the electrolyte over time, particularly at elevated temperatures. Cycle life suffers-you're looking at maybe 300-700 cycles before capacity drops below 80%.

LiFePO₄ (lithium iron phosphate) showed up around 2001 and changed the safety conversation. The olivine structure is rock-solid thermally; thermal runaway doesn't occur until above 270°C, and even then it's less violent. Cycle life is outstanding-2,000+ cycles to 80% capacity is standard, and some cells have been tested past 5,000 cycles. The downside is voltage: only 3.2V nominal, and energy density is limited to 90-120 Wh/kg. Also, the phosphate patent situation was messy for years.

NMC (lithium nickel manganese cobalt oxide) and NCA (lithium nickel cobalt aluminum oxide) emerged as the "balanced" chemistries. By mixing nickel, manganese, and cobalt in various ratios-common ones are NMC 111, 532, 622, and 811 where the numbers indicate relative metal content-you can tune the performance. Higher nickel content pushes energy density up toward 200-250 Wh/kg but at the cost of thermal stability and cycle life. NMC 811 cells can hit 250 Wh/kg but need much more careful thermal management.

On the anode side, graphite has been the standard since day one. Theoretical capacity is 372 mAh/g, and commercial cells typically achieve 340-360 mAh/g. The lithium intercalates between graphene layers during charging, expanding the graphite volume by roughly 10%. This mechanical stress contributes to capacity fade over cycling.

Silicon anodes have been "the next big thing" for about fifteen years now. Silicon's theoretical capacity is 4,200 mAh/g-more than ten times graphite. Problem is, silicon expands by 300% when it absorbs lithium. This breaks the anode apart after a few cycles. Current approaches use silicon-graphite blends with silicon content typically below 10% to keep expansion manageable. Even so, first-cycle irreversible capacity loss runs 15-25% in silicon-containing anodes versus 5-10% for pure graphite.

 

Cell Construction and Formats

 

Cylindrical cells are probably what most people picture when they think "battery." The 18650 format (18mm diameter, 65mm length) became ubiquitous after laptop manufacturers standardized on it in the early 2000s. Tesla famously used thousands of them in the original Roadster. Typical 18650 capacity runs 2,000-3,500 mAh depending on chemistry and whether you optimize for energy or power.

The newer 21700 format (21mm × 70mm) that Tesla and Panasonic jointly developed offers about 50% more energy per cell-4,000-5,000 mAh is common now. The larger diameter increases the ratio of active material to inactive components (current collectors, can, safety devices), improving energy density at the pack level. Manufacturing lines had to be retooled, which is part of why adoption took a while.

Prismatic cells came from the automotive industry's desire for better space utilization. Instead of filling a box with cylinders and leaving all that void space, you make rectangular cells that stack efficiently. Automotive-grade prismatic cells range from 20Ah to over 100Ah capacity. They're easier to thermal-manage from a packaging standpoint since you can put cooling plates directly against the flat sides. The downside is you've got all your eggs in fewer baskets-if one large prismatic cell fails, you lose more capacity than if one small cylindrical cell fails.

Pouch cells take the space efficiency idea further by eliminating the metal can entirely. The cell is sealed in a flexible aluminum-laminate pouch. This saves maybe 10-15% weight versus a prismatic can, and the format is extremely flexible-you can make them any size or shape the application requires. EV manufacturers like them because you can stack them directly in cooling plates. The weakness is mechanical: they need external compression to prevent electrode delamination during cycling, and they're more vulnerable to puncture damage.

 

Lithium Ion Battery Technology

 

Separator Technology

 

The separator doesn't get much attention, but it's arguably the most critical safety component. It's a thin (16-25 μm typically) porous membrane that keeps the anode and cathode from touching while allowing lithium ions to pass through. Early separators were single-layer polyethylene (PE) or polypropylene (PP).

Modern high-performance separators use trilayer structures, typically PP/PE/PP. The PE layer has a lower melting point (135°C) than PP (165°C). If the cell starts overheating, the PE melts and fills in the pores, shutting down ionic transport before the temperature reaches dangerous levels. This is called thermal shutdown, and it's your last line of defense before thermal runaway.

Ceramic-coated separators add another safety margin. A thin (2-4 μm) coating of alumina or other ceramic particles on one or both sides of the separator maintains structural integrity even if the polymer melts. The coating is porous enough that ionic transport continues, but it prevents the electrode from short-circuiting even at temperatures above 150°C. The downside is cost-ceramic-coated separators run 2-3× the price of standard separators-and slightly higher impedance.

Porosity typically runs 40-50%. Too low and ionic resistance goes up, limiting power capability. Too high and mechanical strength suffers. Pore size distribution matters too; the Gurley number (air permeability test) is a standard spec. Most EV-grade separators target 200-400 seconds/100cc.

 

Electrolyte Composition and Additives

 

The electrolyte in a lithium ion cell is more complex than you might think. Base formulation is typically a lithium salt-LiPF₆ (lithium hexafluorophosphate) in 95%+ of cells-dissolved in a mixture of organic carbonates. Common solvents include ethylene carbonate (EC), dimethyl carbonate (DMC), diethyl carbonate (DEC), and ethyl methyl carbonate (EMC).

LiPF₆ concentration is usually around 1.0 to 1.2 M (molar). Higher concentration improves ionic conductivity up to a point, but beyond 1.3 M or so you start getting salt precipitation at low temperatures. LiPF₆ has issues-it's moisture-sensitive and starts decomposing above 60°C-but alternatives like LiBOB or LiFSI haven't displaced it yet due to cost or other tradeoffs.

The carbonate solvent blend gets tuned for the application. EC has high dielectric constant and good SEI-forming properties, but it freezes at 36°C. You need to blend it with lower-viscosity carbonates like DMC or EMC to maintain low-temperature performance. A typical formulation might be EC:DMC 1:1 by volume, or EC:EMC 3:7. The exact ratios are proprietary and closely guarded.

Additives are where the real chemistry magic happens. Modern electrolytes contain 2-5% by weight of various additives that modify SEI formation, prevent overcharge, suppress gas generation, or improve high-temperature stability. Vinylene carbonate (VC) at 1-2% is nearly universal for improving SEI quality on graphite anodes. Fluoroethylene carbonate (FEC) works better for silicon-containing anodes. These compounds preferentially reduce during the initial charging cycles, forming a protective layer on the anode that's ionically conductive but electronically insulating.

Overcharge protection additives like biphenyl or cyclohexylbenzene start polymerizing around 4.5V, creating an internal shunt that prevents voltage from climbing further. This gives you some protection if the BMS fails, though relying on it is obviously not a design best practice.

 

Solid Electrolyte Interface Formation

 

The SEI is probably the least understood but most important aspect of lithium ion battery operation. During the first few charging cycles, electrolyte components react with the anode surface, forming a passivation layer. This layer is critical: it must be ionically conductive (to allow lithium ions through) but electronically insulating (to prevent further electrolyte decomposition). SEI composition is a mess-dozens of lithium salts, organic compounds, and polymers all mixed together in a layer 10-100 nm thick.

Good SEI formation is the difference between a cell that cycles 500 times and one that cycles 3,000 times. The problem is that the SEI isn't static. It cracks during volume changes in the anode, exposing fresh surface that consumes more electrolyte and lithium to repair the damage. This is why capacity fades over cycling even when you're being gentle with the cell.

Formation cycling is a critical manufacturing step. Cells undergo one or more slow charge-discharge cycles at controlled temperatures to establish the initial SEI. Formation protocols are proprietary, but typical first-cycle charging rates are C/20 to C/10, and the process can take 24-48 hours. Manufacturers optimize formation voltage limits, temperature, rest periods, and cycling patterns to produce the most stable SEI possible. Getting this wrong costs you cycle life.

Calendar aging-capacity loss even when the cell is just sitting there-is largely an SEI phenomenon too. The SEI continues growing slowly at open circuit, consuming cyclable lithium. Storage at high state of charge and high temperature accelerates this. A cell stored at 100% SOC and 60°C might lose 20% capacity in a year, while the same cell at 50% SOC and 25°C might lose 3%.

 

Charging Protocols and Battery Management

 

Lithium ion cells are sensitive to overcharge, over-discharge, and charging at inappropriate temperatures. This is why every multi-cell battery pack needs a BMS (battery management system).

The standard charging method is constant current/constant voltage (CC-CV). During the CC phase, you push current into the cell at a fixed rate-typically 0.5C to 1C for most cells, though some high-power cells can handle 3C or more. Voltage rises as the cell charges. When voltage reaches the upper limit (4.2V for most chemistries, 3.65V for LFP, 4.3V or 4.35V for some high-energy NMC variants), you switch to CV mode. Current tapers off as the cell approaches full charge, typically cutting off when current drops below C/20 or C/50.

Fast charging is more complicated. Higher charge rates accelerate lithium plating on the anode, which is dangerous-metallic lithium is highly reactive and can lead to internal shorts or dendrite formation that penetrates the separator. To fast-charge safely, you need to understand how voltage, current, and temperature interact with lithium plating onset conditions.

The issue is that you can't measure lithium plating directly in a sealed cell. You have to infer it from other signals. One approach is to track anode potential versus lithium metal reference. If anode potential goes below 0V versus Li/Li⁺, plating is occurring. Problem is, most commercial cells don't have reference electrodes.

Temperature rise during fast charging also matters. A cell charging at 2C might see its internal temperature rise 15-20°C above ambient even with active cooling. At cold temperatures, this is actually helpful-a cold cell (say -10°C) has very poor power capability, but if you can warm it up by charging at moderate rates (0.5C), performance improves. Some EVs actually do this on purpose: in cold weather, they'll run a brief high-current charge pulse to warm the battery before the driver demands high power for acceleration.

Cell balancing is necessary because cells in series never stay perfectly matched. Manufacturing tolerances, small differences in self-discharge rates, and thermal gradients across the pack cause voltage drift. If you charge a series string without balancing, some cells hit the upper voltage limit before others. The strong cells are undercharged, the weak cells are overcharged, and performance suffers.

Passive balancing uses resistors to bleed off energy from higher-voltage cells. It's simple and cheap but wastes energy as heat. Active balancing uses DC-DC converters or capacitors to transfer energy from high cells to low cells. More efficient, more complex, more expensive. For a 400V EV pack, passive balancing might waste 50-100W continuously, which is negligible compared to driving power but adds up over time.

 

Lithium Ion Battery Technology

 

Thermal Management Considerations

 

Heat generation in a lithium ion cell comes from three sources: irreversible heat (Joule heating from internal resistance), reversible heat (entropy change of the electrochemical reaction), and heat from side reactions. At low to moderate C-rates, reversible heat dominates. At high C-rates, irreversible heat takes over.

The reversible heat term is interesting because it changes sign depending on SOC. For most lithium ion chemistries, charging generates heat at low SOC but absorbs heat at high SOC. Discharging does the opposite. The crossover point is around 50-60% SOC typically. This is why you might see cell temperature actually drop during the final phase of charging if the current is low enough.

Internal resistance varies with temperature, SOC, and aging. At 25°C, a fresh 18650 cell might have 40-60 milliohms DC resistance. At -20°C, that can jump to 200-300 milliohms. This is why cold-weather EV range drops so dramatically. Not only is chemistry slower at low temperatures, but the internal resistance increase means more of the battery's energy is wasted as heat inside the cell.

The thermal runaway threshold depends on chemistry. For NMC cells, exothermic decomposition reactions start around 180-220°C. Once started, temperature can rise at 10-50°C per second, reaching 800°C or higher. LFP is much safer; thermal runaway onset is 270°C+ and the maximum temperature reached is lower.

Propagation between cells in a pack is the real danger. If one cell goes into thermal runaway, it heats its neighbors. Whether the neighboring cells also run away depends on cooling capability, cell spacing, and insulation. UL 9540A propagation testing simulates this by forcing one cell into thermal runaway and monitoring whether adjacent cells follow. Good pack design contains the failure to one cell or at most a small module.

Cooling strategies vary. Air cooling is simplest-blow air over the cells or pack. Works OK for low power density applications like PHEVs or energy storage systems. Liquid cooling is necessary for high-performance EVs. Most designs use a 50:50 water-glycol mix at 10-25 liters per minute through cold plates or cooling channels. Inlet temperature is typically controlled to 20-35°C. Battery pack temperature gradients should stay below 5°C max to min to avoid accelerated aging of the hottest cells.

Some experimental designs use refrigerant cooling, immersion cooling in dielectric fluid, or phase-change materials. Refrigerant cooling can pull more heat out but requires a more complex AC system. Immersion cooling has excellent heat transfer coefficients (500-2,000 W/m²K versus 50-150 W/m²K for indirect liquid cooling) but sealing and fluid compatibility are challenges. PCMs work passively but need to reject the stored heat eventually, so they mainly help with transient cooling during fast charging or hard acceleration.

 

Performance Degradation and Failure Modes

 

Capacity fade and impedance growth are the two main degradation mechanisms. They're caused by several different physical and chemical processes happening simultaneously.

On the anode side, SEI growth consumes cyclable lithium and electrolyte, increasing resistance. Graphite exfoliation can occur if the cell is charged at low temperatures-lithium plates on the graphite surface instead of intercalating, and when it does eventually intercalate, it breaks the graphite structure apart. This is often irreversible. Binder decomposition at elevated temperatures causes loss of electrical contact between particles.

Cathode degradation includes transition metal dissolution (especially manganese in LMO or manganese-containing NMC), structural changes from repeated lithium insertion/extraction, and surface reconstruction in high-nickel cathodes. The dissolved transition metals migrate to the anode where they catalyze SEI growth, so cathode degradation indirectly accelerates anode degradation.

Electrolyte decomposition and gas generation are bigger problems at high voltages and high temperatures. Common gases include CO₂, CO, and various hydrocarbons from carbonate decomposition. In pouch cells, you'll see the pouch swell visibly. In cylindrical or prismatic cells with hard cases, pressure builds up until the safety vent opens (typically at 10-15 bar).

Loss of lithium inventory is a major fade mechanism. Every time the SEI grows or lithium plates irreversibly on the anode, some lithium is taken out of the pool of cyclable lithium. Eventually you run out and capacity drops.

Sudden failures can occur from internal shorts. Most shorts start small-a tiny metal particle punctures the separator, or a lithium dendrite grows through. The short creates a hot spot, which accelerates degradation locally, which makes the short worse, and you get a positive feedback loop. Sometimes the cell self-heals if the short melts itself open. Other times it progresses to thermal runaway.

Nail penetration tests (forcing a steel nail through a charged cell) are a standard abuse test. LFP cells typically don't go into thermal runaway from nail penetration. NMC cells often do, though designs with better separators and lower specific energy can sometimes pass.

Figure 5 plots capacity retention versus cycle number for several chemistries under moderate cycling conditions (1C charge/discharge, 25°C, 100% DOD).

 

State of Charge and State of Health Estimation

 

You can't directly measure how much energy is in a lithium ion cell. You have to estimate it from other measurements: voltage, current, and temperature.

The simplest SOC estimation method is voltage-based. Each chemistry has a characteristic open-circuit voltage versus SOC curve. Measure the voltage after the cell has rested for a while (to let transient voltage drops from internal resistance decay), look it up on the OCV curve, and you know SOC. Problem is, you rarely have time for the cell to rest in real applications.

Coulomb counting is the standard approach. You integrate current over time to track charge in and out. If you start at a known SOC, you can calculate the new SOC at any time. Accuracy depends on your current sensor (±0.5% is typical) and knowing the true capacity. Errors accumulate over time, so you need to periodically recalibrate by doing a full charge or discharge cycle.

Model-based methods use an equivalent circuit model or electrochemical model of the cell. You measure terminal voltage and current, run them through your model, and extract internal states including SOC. Extended Kalman filters or similar state observers are common. These approaches can be very accurate (±2% SOC error) but require good models and significant computational resources.

SOH estimation is harder because you're trying to quantify degradation, which is slow and gradual. Capacity fade and impedance growth don't necessarily correlate linearly with each other or with cycle count. A cell that's been fast-charged a lot might have high impedance but only moderate capacity fade. A cell that's been stored at high SOC/temperature might have significant capacity fade but relatively low impedance growth.

Industry practice is to define SOH based on capacity: a cell at 80% of its original capacity is at 80% SOH, and this is often considered end-of-life for EV applications. The cell still works, but range has dropped 20%. For energy storage applications, cells might be used down to 60-70% SOH.

Some BMSs do periodic capacity checks-fully discharge the battery at low rate and measure how much energy comes out. This is accurate but intrusive (the battery is unavailable during the test) and takes hours. Other approaches try to estimate capacity indirectly from voltage curves, impedance measurements, or coulombic efficiency.

Internal resistance can be measured by applying a current pulse and measuring voltage response, or by injecting a small AC signal at various frequencies (electrochemical impedance spectroscopy). EIS gives much more information but requires specialized hardware that's rarely present in commercial BMSs.

 

Lithium Ion Battery Technology

 

Second-Life Applications and Recycling

 

When an EV battery reaches end-of-life (typically 70-80% of original capacity), it's still perfectly functional for less demanding applications. Second-life battery use is gaining traction for stationary energy storage.

The economics are tricky. You have to test the retired pack, potentially remanufacture it (replace the BMS, cooling system, or damaged modules), certify it for the new application, and provide a warranty. All this costs money. For second-life to make sense, the refurbished pack needs to cost significantly less than a new pack designed for the stationary application. Costs break even if refurbishment runs less than 40-50% of new pack cost, depending on whose analysis you believe.

Testing retired cells is non-trivial. A module might contain hundreds of cells in series-parallel. You can't easily test them individually. You can test the module as a unit, but one bad cell can mask itself. Some degradation modes are hard to detect without destructive testing. There's also the liability question: if a second-life battery catches fire, who's responsible?

Recycling is the ultimate end-of-life path. Current large-scale recycling uses pyrometallurgy (smelting) or hydrometallurgy (chemical leaching). Pyrometallurgy is simpler but less selective-you get mixed metal alloys that need further refining. Hydrometallurgy can recover individual metals at higher purity but requires more steps and generates chemical waste.

The economics of recycling depend heavily on metal prices. Cobalt is valuable (roughly $30-40/kg historically, though prices swing wildly), so recycling cobalt-rich chemistries is economically viable. Nickel is worth recycling at scale. Manganese, iron, and aluminum are low-value metals, so recycling makes sense mainly to keep them out of landfills. Lithium is interesting-it's not particularly valuable per kilogram, but supply constraints make recovery attractive.

Direct recycling-disassembling the battery and reusing cathode or anode material directly without breaking it down to metal salts-is a hot research area. If you could recover cathode powder in usable form, you'd save the energy and cost of cathode synthesis. Challenges include separating the active material from current collectors and binders, and dealing with the fact that recycled material is a mixture of cells from different manufacturers, ages, and chemistries.

 

 

Send Inquiry