Physical mechanisms: how temperature gradients create SOC heterogeneity

Temperature governs every rate-limiting electrochemical process inside a lithium-ion cell: solid-state lithium diffusivity in both electrodes, charge-transfer kinetics at the electrode–electrolyte interface, ionic conductivity in the electrolyte, and the ohmic resistance of current collectors and separator. When a spatial temperature gradient exists across the active area of a large-format cell, these parameters vary continuously from the cooler region to the warmer region, producing corresponding gradients in local reaction current density and lithium insertion rate. Different portions of the electrode stack therefore charge at different effective rates even when a single bulk current is applied — the definition of local state-of-charge (SOC) nonuniformity.

The coupling between thermal gradients and SOC distribution was quantified explicitly in a two-dimensional electro-thermal model developed at Imperial College London (2018) that resolved not only the electrode stack but also tab weld points and non-core components. The study showed that surface cooling keeps average cell temperature lower while simultaneously imposing a larger through-plane thermal gradient compared with tab cooling — demonstrating that the choice of cooling topology fundamentally reshapes internal SOC distribution, independent of bulk temperature.

The directionality of the thermal gradient matters beyond its magnitude. Research from the Hawaii Natural Energy Institute (2021) applied intentional inter-electrode thermal gradients of as little as ±2 °C and documented dramatically accelerated capacity fade: 77% fade over 20 cycles when the negative electrode was warmer, and 100% fade when the positive electrode was warmer. Incremental capacity analysis confirmed that each gradient polarity produced a chemically distinct failure mode, underscoring that even modest thermal nonuniformity inside a large cell can redirect SOC distribution in ways that selectively overstress one electrode — a finding consistent with the standards and safety frameworks maintained by bodies such as IEC.

“A ±2 °C inter-electrode temperature gradient is sufficient to shift the dominant degradation mechanism from positive to negative electrode — with the colder electrode reaching local SOC limits first and plating lithium.”

Cell format and cooling topology: why larger cells face steeper SOC gradients

Large-format cells exhibit fundamentally more severe surface temperature distributions during fast charging than small-format cells, as demonstrated by research from the University of Bath (2022). That study found that larger cells show considerably greater surface temperature distributions during internal heating, implying that any heat generation heterogeneity — whether from tab resistance, electrode thickness variation, or cooling asymmetry — is amplified by cell size. This makes large-format pouch and prismatic cells used in EV modules significantly more vulnerable to SOC nonuniformity under fast charging compared with 18650 or 21700 formats.

For cylindrical large-format cells, the interaction between current collector design and cooling strategy is decisive. Research from the Technical University of Munich (2022) used a validated multiphysics model to compare 18650, 21700, and Tesla 4680 cell formats. Mantle (surface) cooling was most efficient for segmented tab designs, while tab cooling performed equally well and even outperformed mantle cooling for the tabless 4680 format. Critically, the tabless design reduced polarization drops by approximately 250 mV at 3C and reduced heat generation in current collectors by up to 99%, dramatically homogenizing the internal temperature distribution and thereby reducing the local SOC spread across the electrode area.

For pouch cells under high-power charging, the thermal management of the cooling circuit directly controls temperature distribution and electrochemical nonuniformity. Research from Tsinghua University (2020) modeled a 30 Ah LiFePO4 pouch cell using a coupled 1D electrochemical and 3D heat generation approach, demonstrating that both inlet flow rate and cooling channel geometry have a significant influence on temperature distribution during high-rate charging. Inadequate cooling creates temperature hot spots that locally advance SOC faster than the cell average, shortening usable capacity and threatening lithium plating in adjacent cooler regions — a risk factor highlighted in technical guidance from NREL.

The degradation impact of sustained thermal gradients was rigorously quantified using a prismatic 60 Ah graphite/NCM-LMO cell by BMW Peugeot Citroën Electrification GmbH (2012). A transient 3D thermal FVM model combined with a 3D impedance-based finite network model was validated using eight internal thermocouples inside the jelly roll. The study directly linked localized temperature gradients from active cooling geometries to inhomogeneous aging, with hotter zones consuming capacity faster — a direct consequence of accelerated local SOC cycling.

Electrochemical modeling: resolving the local anode SOC constraint during fast charging

The connection between thermal conditions and local SOC state during fast charging is most precisely captured through electrochemical-thermal (ECT) coupled models. Research from Chongqing University (2023) explicitly studied a 12-cell module configured in a 4S3P arrangement, using a polynomial approximation pseudo-2D model. The mechanism was mapped in detail: cells experiencing higher heat dissipation rates developed cooler local temperatures, lower ionic conductivity, and higher polarization, causing their local SOC to lag behind thermally advantaged neighbors. This internal imbalance persisted even after the bulk current was removed.

An optimal fast-charging strategy based on an electrochemical-thermal model incorporating intercalation-induced stresses and SEI film growth was developed at Zhejiang University (2020). Dynamic programming was used to minimise charging time while respecting constraints on electrode stress, SEI growth, and temperature. The backstepping technique was applied to control temperature and prevent overheating — recognising that thermal excursions both degrade the cell and invalidate the SOC estimation on which the optimal current profile depends.

The trade-off between energy density and fast-charge capability in cells with thick electrodes was systematically analysed using validated pseudo-2D (P2D) models (2022). The study showed that low anode potentials during fast charging — the primary precursor to lithium plating — can be effectively avoided by elevated cell temperature or by a CC/constant-potential/CV protocol that includes a constant anode potential phase. This result directly links the temperature field to the local anode SOC limit: colder regions of a large cell will have locally lower anode potentials at the same bulk current, and will plate lithium before the bulk SOC-based current limit is triggered. Such findings align with electrochemical safety frameworks published by The Electrochemical Society.

NXP Semiconductors’ model-based fast-charging patent (CN, 2022) controls charging current to maintain the anode lithium surface concentration below the saturation threshold, explicitly acknowledging that diffusion time can vary by a factor of ten between 15 °C and 35 °C. In a large-format cell with a 10–20 °C internal temperature gradient, the local diffusion-limited SOC ceiling therefore varies dramatically across the electrode area — a single bulk current setpoint will push cooler zones toward lithium plating while undercharging warmer zones.

BMS innovations: zone-resolved sensing and gradient-aware current control

Zone-resolved temperature sensing within individual cells — not just pack-level sensing — is required to detect and mitigate thermal-gradient-driven SOC nonuniformity. The most extensive patent portfolio in this space belongs to Samsung SDI, whose multi-zone BMS architecture (KR, 2025) computes zone-specific maximum temperature deviations for each battery cell and calculates the maximum inter-cell temperature deviation by subtracting the smallest from the largest zonal deviation. When this inter-cell deviation exceeds a specified lower threshold temperature, charge current control is applied to the “vulnerable cell” — the cell exhibiting the largest internal temperature gradient.

Samsung SDI’s gradient-slope-based variant computes maximum slope averages of temperature over zone-surrounding areas, linking temporal thermal gradient dynamics to charge current adjustment. This approach explicitly acknowledges that intra-cell temperature gradients indicate current concentration and internal electrochemical nonuniformity, and that derating charge current is the primary mitigation.

Hyundai Motor applied a reference electrode approach to directly measure negative electrode potential in individual cells during fast charging (KR, 2025). The system determines per-cell negative electrode potential using a reference electrode and controls per-cell charge current based on the minimum observed negative electrode potential, directly preventing lithium deposition on colder or higher-SOC zones within the module — an approach that aligns with electrochemical measurement standards documented by NIST.

Zhejiang Wanma New Energy patented a thermal equalization method specifically for ultra-fast charging (10–12 minute) applications (CN, 2021). The method recognises that in ultra-fast charging, large heat generation exacerbates inter-cell temperature imbalance, the hottest cell ages fastest, and capacity inconsistency is progressively amplified. A pulse charging modality is employed within an SOC window bracketed between minimum and maximum thresholds to reduce polarisation resistance-driven heat accumulation and temperature rise.

Vrije Universiteit Brussel developed MPC-based charging and balancing algorithms using a reduced-order electrochemical Equivalent Hydraulic Model (2020) that simultaneously maximises charging current and configures shunting to balance SOC across cells, with degradation phenomena explicitly included. Guangzhou Xiaopeng Motors’ EP patent (2023) divides charging into SOC-based stages and for each stage separately determines negative electrode rebound potential, polarisation potential, and the allowable minimum negative electrode potential limit, implicitly accommodating the local anode potential variation that arises from thermally induced SOC nonuniformity.

The positive feedback loop: why SOC imbalance compounds over hundreds of cycles

SOC nonuniformity and thermal nonuniformity form a positive feedback loop that progressively amplifies capacity divergence across cells and within individual large-format cells. Research from Vrije Universiteit Brussel (2015) documented that cells with a non-zero initial depth-of-discharge show up to 5% higher temperature increase compared with reference cells during high-rate discharge. This means that SOC nonuniformity causes thermal nonuniformity, which in turn further advances local SOC in already-hotter zones — accelerating the divergence of aging trajectories.

The pack-level consequences were quantified by WMG, University of Warwick (2021): a temperature gradient of 10 °C introduced across four parallel cells resulted in measurable current imbalance during fast charging, which amplified over 1,400 cycles into diverging aging trajectories — with the current-loaded cell exhibiting lithium plating and accelerated degradation. This result demonstrates that cell-level thermal gradients and pack-level SOC nonuniformity are not independent phenomena; they mutually reinforce one another, making pack-level thermal management a direct control lever for the rate of SOC divergence in parallel strings.

“A 10 °C inter-cell temperature gradient produces measurable current imbalance during fast charging that amplifies over 1,400 cycles into diverging aging trajectories — with the current-loaded cell exhibiting lithium plating and accelerated degradation.”

The mechanical dimension of this feedback loop was revealed by CT-Lab Stuttgart (2021), which documented using laser measurement and X-ray CT that cylindrical cells expand non-uniformly around their circumference as a function of SOC change — the so-called “Potato Effect.” This internal SOC nonuniformity produces localised mechanical stress fields that further distort thermal conductivity and reaction pathways in subsequent cycles, adding a structural dimension to the electrochemical feedback loop.

The dataset examined for this analysis comprises over 50 patent filings and peer-reviewed publications spanning academic institutions, OEMs, and technology companies active between 2009 and 2025. Key contributing assignees include Samsung SDI, Tsinghua University, Technical University of Munich, Imperial College London, BMW Peugeot Citroën Electrification GmbH, the University of Warwick, Zhejiang Wanma New Energy, and Guangzhou Xiaopeng Motors. A recurring theme across all sources is that thermal gradients across large-format cells — whether pouch, prismatic, or large-diameter cylindrical formats — are qualitatively more severe than in small-format cells, and that the resulting electrochemical nonuniformity during fast charging represents one of the principal pathways to accelerated aging, lithium plating, and safety hazards. This conclusion is consistent with the battery safety research agenda maintained by the U.S. Department of Energy.

Taken together, the evidence from patent filings and academic research establishes a clear hierarchy of interventions: (1) cell design choices — particularly tabless current collector geometries — that minimise internal heat generation gradients at source; (2) cooling topology selection that matches the cell format to avoid imposing large through-plane thermal gradients; (3) model-based charging algorithms that incorporate spatially resolved temperature to correctly compute the local anode SOC constraint; and (4) zone-resolved BMS architectures that detect gradient slope dynamics and derate charging current before local lithium plating conditions are reached. No single intervention is sufficient in isolation; effective thermal gradient management for large-format fast charging requires all four layers to be co-designed.