Thermal Gradient Effects on Active Material Loss in Prismatic Formats
Internal thermal gradients accelerate prismatic cell active material loss by driving localized current crowding, high-temperature SEI growth, and particle cracking.

Core
Operating prismatic lithium-ion cells at elevated currents creates internal heat patterns shaped directly by how the electrode sheets and current collector foils are stacked. In large-format hard-case enclosures from 50 Ah to 320 Ah, volumetric heat generation combines with directional thermal conductivity to create distinct temperature variations through the jelly-roll or z-folded stack. Tab connections and aluminum-copper foil resistance produce localized Joule heating near the header, whereas reaction entropy and polarization resistance spread heat across the active layers.
Because of the planar aspect ratio of prismatic cells, heat extraction through the broad faces, narrow sides, or bottom plate sets up multidirectional thermal fields that diverge from isothermal lab baselines.
Internal thermal gradients split into through-plane and in-plane directions. Through-plane transport works perpendicular to the stack, crossing alternating active coatings, polymeric separators, metallic foils, and liquid electrolyte films. In-plane transport runs parallel along the continuous copper and aluminum current collectors.
While metallic foils have high bulk thermal conductivity, the active coating composites and porous separators introduce substantial contact resistance and poor thermal transport. As a result, in-plane thermal conductivity frequently exceeds through-plane conductivity by more than twenty to one, forcing heat along the foil pathways toward the perimeter before it can conduct into external cooling surfaces.
Temperature variations across the casing disrupt local thermodynamic equilibrium and kinetic rates through the stack. Charge transfer resistance at the solid electrolyte interphase follows an Arrhenius relationship with temperature: warmer zones see faster charge-transfer kinetics, lower solid-state diffusion resistance, and reduced electrolyte viscosity, while cooler zones face higher overpotentials and lower ion mobility. When external terminals draw current from the parallel coating layers, current splits unevenly across the foils.
Local current density spikes in warmer sections of the jelly-roll, accelerating material consumption and capacity loss along specific spatial bands.
Bottom-plate liquid cooling on a 280 Ah cell under continuous 1C cycling generates an internal through-plane temperature difference of 8.2 Kelvin between the internal electrode center and the cooling plate interface.
Physical constraints in prismatic metallic cans worsen these transport mismatches. Aluminum enclosures with wall thicknesses from 0.6 mm to 1.5 mm offer structural containment and outer heat spreading, but internal jelly-rolls are separated from the can walls by insulating polymeric wraps and electrolyte reservoirs. Void spaces along lateral edges and top headspace form insulating gas pockets that hinder uniform peripheral heat dissipation.
The internal electrode stack therefore operates under continuous thermal tension during high-rate cycling, driving microstructural changes that permanently reduce available electrochemical capacity.
Internal thermal variations are routinely treated as minor boundary artifacts absorbed by pack-level safety margins.

Flux
The heat dissipation path chosen in mechanical pack design sets the size of internal thermal deltas across prismatic cells. Bottom-plate liquid cooling is standard in commercial energy storage and electric vehicles because it simplifies integration and sealing. However, this layout forces heat to travel down the full cell height ~ from 100 mm to over 220 mm in standard form factors.
Conducting heat along the entire vertical length of thin copper and aluminum foils to reach the cold bottom plate creates a steep temperature gradient between the bottom corners and the uncooled terminal header.

Can Bottom Cooling Induce Asymmetric Particle Cracking?
Direct mechanical coupling to a base chiller forces an uneven vertical reaction profile across the cathode stack. High thermal flux at the base keeps local temperatures lower, increasing interfacial charge-transfer resistance and solid-state diffusion overpotentials in the bottom third of the active layers. Meanwhile, the upper third runs hot, heated by tab Joule losses and hindered by poor convective shedding into the pack headspace.
During high-rate discharge, lithium extracts rapidly from upper electrode regions where kinetic barriers are lower, driving localized high state-of-charge swings while the cold bottom remains underutilized. This cyclic spatial strain causes mechanical microcracking in upper-segment layered transition metal oxide particles long before the lower material reaches its fatigue limit.
Tab heating reinforces this vertical imbalance. Prismatic cells use internal foil tabs welded in ultrasonic or laser bundles to connect the stack to external terminals. Current crowding at these constricted joints generates intense local Joule heating.
Under continuous 1.5C or 2C loads, the tab assembly acts as an internal heat source perched directly above the jelly-roll. The top of the stack absorbs this heat flux, widening the thermal spread along the cell height.
| Component Layer | In-Plane Thermal Conductivity (W/m·K) | Through-Plane Thermal Conductivity (W/m·K) | Thickness (micrometers) | Characteristic Biot Number Range |
|---|---|---|---|---|
| Positive Foil (Aluminum) | 210.0 | 210.0 | 13 | 0.001 – 0.004 |
| Cathode Coating (LFP/PVDF) | 1.8 | 0.6 | 75 | 0.120 – 0.350 |
| Separator (Polyethylene/Ceramic) | 0.4 | 0.15 | 16 | 0.450 – 1.100 |
| Anode Coating (Graphite/CMC-SBR) | 2.5 | 0.9 | 85 | 0.090 – 0.280 |
| Negative Foil (Copper) | 385.0 | 385.0 | 8 | 0.0005 – 0.002 |
| Prismatic Shell Wall (Aluminum 3003) | 160.0 | 160.0 | 800 | 0.010 – 0.030 |
Side cooling mitigates vertical thermal gradients by pressing cooling plates against the large planar faces of adjacent cells, cutting the conduction path length to half the cell thickness ~ typically 25 mm to 36 mm. Face cooling introduces its own challenges, however. Compressible thermal interface materials must maintain uniform contact pressure across broad aluminum surfaces undergoing cyclic breathing and permanent thickness expansion.
If interface pressure relaxes or surface waviness forms micro-gaps, thermal contact resistance spikes, creating hot zones under areas of poor contact that disrupt planar current distribution.
Conductive heat transport models quantify the temperature profile across a standard 72173204 format cell. The Biot number across the through-plane stack exceeds unity under high convection or direct liquid chilling, showing that internal conduction resistance dominates over surface heat transfer. During fast-charge transients, the core of the jelly-roll stays thermally isolated from external cooling circuits.
Ignoring directional heat flux disparities during pack integration accelerates localized particle isolation and invalidates module cycle-life warranties.

Mismatch
Spatial temperature variations within a single prismatic cell create parallel electrochemical domains operating at different kinetic rates. The open-circuit potential of lithium transition metal oxides and graphite carries a temperature coefficient that causes slight equilibrium voltage variations across the cell plane. More critically, exchange current density depends exponentially on absolute temperature per classical activation energy barriers, as expressed by the Butler-Volmer relation:
Local current density scales with local exchange current density and surface overpotential. Warmer active material domains offer faster desolvation kinetics and higher lithium-ion conductivity through the surface film. Under a given terminal load, current lines bend toward warmer regions of the jelly-roll: local rate in a hot core or top header zone can reach 1.8C while the chilled bottom operates at 0.6C during a nominal 1C cycle.
This current redistribution balances total terminal potential by forcing warmer zones to carry disproportionate electrochemical throughput.
Localized current crowding drives rapid phase transformations within the active materials. In lithium iron phosphate, the flat two-phase voltage plateau conceals local state-of-charge divergence. In layered nickel-manganese-cobalt oxides, continuous solid-solution delithiation means high local current density drives warmer particles to deeper discharge levels and higher local cutoff potentials.
Active material in hot zones undergoes excessive lithium extraction, crossing structural phase boundaries into irreversible lattice collapse.
- Spatial Overpotential Distribution alters local particle surface charge states, creating internal circulating currents during rest periods that consume cyclable inventory.
- Phase Transition Asymmetry forces high-temperature crystal domains through volumetric lattice transitions earlier than adjacent low-temperature zones.
- Electrolyte Conductivity Variance produces uneven ohmic drop along the porous separator pathways, concentrating current near the separator-electrode boundary in warmer regions.
- Localized State Of Charge Drift leads to premature electrochemical end-of-charge cutoffs while cold active materials remain incompletely utilized.
Multi-channel sensing arrays embedded in developmental cell hardware track localized current density divergence. Measurements confirm that a steady-state thermal gradient of 10 Kelvin across the cell height induces a 35 to 45 percent current density imbalance between the warmest and coolest active material segments. This redistribution operates as a self-reinforcing degradation mechanism: the hotter region carries excess current, generating additional Joule and polarization heat that sustains the local temperature elevation until terminal cutoff voltage ends the half-cycle.
Thermal gradients exceeding 7 Kelvin across a prismatic jelly-roll cause local current density in the upper electrode zone to exceed the nominal cell C-rate by 40 percent during 2C constant-current discharge.
Uneven active material utilization leaves persistent state-of-charge differences across the electrode area. When external current stops, these gradients equalize via internal circulating currents across the continuous current collector foils, undetected by the battery management system. These parasitic currents cause low-rate cycling during idle periods, accumulating extra equivalent full cycles in local regions of the cell.
| Spatial Stack Zone | Local Operating Temperature (Celsius) | Effective Exchange Current Density (A/m²) | Electrolyte Conductivity (mS/cm) | Observed Local Capacity Loss Rate (percent per 500 cycles) |
|---|---|---|---|---|
| Upper Tab Header Zone | 42.5 | 14.2 | 11.8 | 4.8 |
| Core Jelly-Roll Center | 38.0 | 11.6 | 10.5 | 3.6 |
| Lateral Edge Boundary | 34.5 | 9.8 | 9.4 | 2.7 |
| Bottom Chilled Interface | 32.5 | 8.7 | 8.7 | 2.1 |
Electrochemical disparities between stacked layers degrade pack-level state-of-charge estimation. External busbar voltage reflects a mixed potential across parallel electrochemical paths with differing internal resistances and local states of charge. Management systems relying on standard open-circuit voltage lookup tables misjudge actual energy content and power limits, allowing localized over-charge or over-discharge to occur undetected inside the metallic housing.
It remains uncertain whether dynamic pulse balancing protocols can interrupt current crowding before microstructural material damage becomes irreversible.

Depletion
Degradation from internal thermal gradients splits into two primary mechanisms: loss of lithium inventory and loss of active material. Lithium inventory loss dominates at temperature extremes, driven by accelerated side reactions in hot zones and metallic lithium plating in cold zones. Active material loss stems mainly from mechanical lattice strain, particle pulverization, transition metal dissolution, and binder breakdown within the electrode matrix.

Could Tab Geometry Suppress Current Crowding?
Altering tab geometry modifies current entry paths and reduces localized ohmic heating near the top of the cell. Designs using dual-ended tabs or continuous multi-tab ultrasonic welding spread current entry across larger foil areas, easing local current concentration. This flattens internal Joule heating profiles and lowers the top-to-bottom thermal gradient during high-rate operation.
Distributing current symmetrically across electrode foils prevents severe localized overpotentials, curbing side reactions.
In hot zones, cyclable lithium is continuously consumed as the solid electrolyte interphase rapidly reforms. Carbonate solvent molecules and hexafluorophosphate salt anions decompose faster according to Arrhenius kinetics. The passive film thickens unevenly, trapping active lithium ions in insoluble inorganic compounds like lithium carbonate, lithium fluoride, and lithium alkyl carbonates.
This growth clogs separator pores and increases interfacial resistance, restricting ionic flux to adjacent active material.
- Solid Electrolyte Interphase Thickening proceeds rapidly in warm stack regions, consuming cyclable lithium ions and generating resistive surface deposits that isolate active particle facets.
- Mechanical Particle Microcracking results from repeated anisotropic lattice contraction and expansion under non-uniform current loads, fracturing primary grains and breaking electronic percolation pathways to the current collector foil.
- Lithium Plating Deposition occurs on graphite particle surfaces in chilled electrode segments when localized low temperature elevates charge transfer resistance and drives the anode potential below zero volts versus metallic lithium.
- Transition Metal Dissolution accelerates at high local potentials and elevated temperatures, causing manganese, cobalt, or iron ions to migrate across the separator and contaminate the negative electrode interphase.
Particle cracking separates active cathode material from the conductive carbon matrix. Layered nickel-rich oxides like NMC811 undergo large anisotropic volume shifts along the c-axis during delithiation. In hot zones where local state-of-charge swings are wide, repeated anisotropic strain opens intergranular cracks along primary particle boundaries.
Electrolyte penetrates these fresh microcracks, consuming lithium to form passivating films on new surfaces while isolating particle fragments electronically.
Post-mortem cross-sectional scanning electron microscopy reveals that active material particle isolation in the top 20 percent of a bottom-cooled prismatic cell is 3.2 times greater than in the bottom 20 percent after 1,200 full cycles.
Cold regions in the cell face lithium plating during fast charging. Lower exchange current density drops the anode polarization potential; if the local negative electrode potential falls below 0 V relative to Li/Li+, metallic lithium plates onto the graphite surface instead of intercalating. This plated lithium reacts with electrolyte solvents to form dead lithium, permanently consuming inventory and creating internal short-circuit hazards.
| Internal Thermal Condition | Total Capacity Loss (percent) | Attributed Loss of Lithium Inventory (percent) | Attributed Positive Active Material Loss (percent) | Attributed Negative Active Material Loss (percent) |
|---|---|---|---|---|
| Isothermal Reference (25°C) | 11.4 | 7.2 | 2.4 | 1.8 |
| Controlled 5K Gradient | 16.8 | 10.5 | 3.8 | 2.5 |
| Controlled 10K Gradient | 24.3 | 14.8 | 5.7 | 3.8 |
| Controlled 15K Gradient | 35.6 | 21.2 | 8.9 | 5.5 |
Differential voltage and incremental capacity analysis resolve coexisting degradation mechanisms from terminal electrical measurements. Phase-transition peaks shift to higher potentials and shrink in area as active material breaks down and lithium inventory drops. Under internal thermal gradients, these peaks broaden and skew asymmetrically, reflecting multiple out-of-phase electrochemical domains operating simultaneously inside the single casing.
Procurement contracts based on IEC 62660-1 cycle-life ratings become difficult to enforce when pack integration creates thermal gradients exceeding the manufacturer’s 3 Kelvin test boundary.

Settlement
Gradient-induced degradation directly impacts the financial performance and warranty coverage of large-scale battery assets. Evaluating prismatic cells solely on datasheet cycle-life figures means relying on isothermal performance measured in tightly controlled lab settings. Standard factory tests use environmental chambers with mild convective airflow or clamped aluminum cooling blocks that keep surface temperature spreads within 2 Kelvin.
Commercial battery packs rarely reproduce such ideal thermal conditions.
In utility storage systems or heavy commercial transport packs, thermal management constraints create sustained internal gradients. A cell rated for 6,000 cycles to 80 percent state of health under isothermal 25°C testing may deliver fewer than 3,800 cycles under a persistent 8 Kelvin gradient. This cycle-life compression accelerates replacement schedules, reduces internal rates of return, and shifts replacement liabilities onto system integrators.
Factory cell grading protocols must account for internal impedance variations that worsen thermal gradients. Prismatic cells with higher direct-current resistance generate heavier heat loads under dynamic operation. Acceptance sampling that checks only static capacity and nominal 1 kHz alternating-current resistance misses internal layer misalignments and tab weld resistance flaws that magnify thermal gradients in the field.
Technical specifications in cell purchase agreements need clear operational boundary limits, requiring multi-point thermal characterization data alongside standard capacity ratings. Mandating accelerated degradation testing under forced thermal gradient boundaries of 5 Kelvin and 10 Kelvin prior to volume supply commitments protects capital investments against premature capacity fade and ensures realistic lifecycle costing models.
Thermal gradients remain manageable as long as cooling system conduction matches internal foil heat transport rates across the operational window.


