Battery Aging — Fewer Marbles, Cracked Shelves

Why a Two-Year-Old Phone Wears Out First

A new phone, when filled up in the morning, lasts until night. After about two years, even with the same usage, you find yourself looking for a charger before evening comes. The screen brightness is the same, the apps you use are the same, but the day grows short.

Meanwhile, inside the battery, two things happen together. The amount you can take out and use when fully charged decreases, and taking out the same amount becomes harder (DOI: 10.1016/j.est.2022.105634). Where these two changes come from, what changes their speed, and who estimates that state and how are the trunk of this article.

The subject is the lithium-ion battery that goes into today's phones and electric cars. We do not cover how to charge it, but translate what happens strictly based on what the evidence has measured.

Aging Reveals Itself in Two Numbers

Saying a battery has aged translates to two numbers. A battery that has been used for a long time can hold less and has higher resistance (DOI: 10.1016/j.etran.2025.100410).

Capacity fade is when the total amount of energy you can get out of a fully charged battery drops compared to when it was new. This is what you notice first in a phone whose day has gotten shorter. What unit this amount is written in is covered in another part of this series, the one dealing with reading battery numbers.

The second number requires three terms used in earlier pieces covering electricity. Current is the flow of electricity along a wire, measured in amperes as to how much passes a point in one second. Voltage is the difference between two points that makes electricity flow, measured in volts. Resistance is how much an appliance or wire blocks the flow of current, measured in ohms.

Internal resistance is how much the battery itself blocks the flow of current inside. When you draw current, part of the voltage is trimmed away inside the battery because of this resistance, and the trimmed portion becomes heat. When internal resistance grows, drawing the same current makes more heat, and the limit of power the battery can push out all at once—its output—also shrinks.

As mechanisms that make a battery age during normal use, five major ones and thirteen minor ones are counted. The ways these manifest converge into five, and the representative outcomes visible on the outside are capacity reduction and output reduction (DOI: 10.1039/d1cp00359c). Because there are many causes but only two representative numbers are seen, you cannot tell what happened underneath by looking at the numbers alone. Separately devised measurements are used to sort out the causes (DOI: 10.1039/d1cp00359c).

Beads Moving Back and Forth Inside a Single Compartment

A cell is the smallest compartment that makes up a battery, containing a positive electrode, a negative electrode, an electrolyte filling the space between them, and a separator dividing the two, all inside a single container. Aging happens inside this single compartment. The main character traveling back and forth inside is the lithium ion. A lithium ion is a particle that takes on a + electrical charge when a lithium atom gives up one electron.

The positive electrode is the + electrode of the cell, where lithium ions return and settle during discharge. The negative electrode is the − electrode of the cell, where lithium ions cross over and stay during charging. The negative electrode of cells used today is usually graphite, and when charged at a normal speed, lithium ions go inside the graphite and settle down (DOI: 10.1021/acsnano.1c02942).

The electrolyte is a substance filling the space between the two electrodes that lets lithium ions pass through while preventing electrons from crossing. The separator is a thin membrane dividing the positive and negative electrodes so they do not touch each other, with many tiny pores through which lithium ions pass.

This article pictures the cell as two shelves where beads move back and forth. The beads are lithium ions, and the shelves are the two electrodes with sites where the lithium ions can enter and sit. When charging, the beads cross over to sit on the negative electrode shelf, and when discharging, they return to the positive electrode shelf. While the beads cross through the internal electrolyte, electrons travel along the external wire, and that flow is the electric current we use. When one bead crosses inside, one electron travels outside, so the two always move in pairs. However, real lithium ions do not sit on top of the shelves, but squeeze in between the layers of the electrode material. If they fail to squeeze in and pile up on the surface, that is lithium plating, which we will see later (DOI: 10.1021/acsenergylett.0c00859). The article covering cell structure follows these two paths in detail.

In this picture, there are two paths of aging. One is the path where the beads available to travel back and forth decrease, and the other is the path where the shelves for the beads to sit on are damaged. Research dealing with aging counts these separately as losing lithium that can travel back and forth and losing electrode material where lithium can sit (DOI: 10.1039/d2cp00417h). The worsening of contact between electrode materials, which increases resistance, happens together as well (DOI: 10.1002/admt.202502455).

The First Path to Losing Marbles — The Film on the Negative Electrode Surface

In the final step of making a cell, after letting the electrolyte soak into the electrodes, the cell is slowly charged and discharged a few times. One purpose of this step is to grow a film on the surface of the negative electrode (DOI: 10.1016/j.carbon.2016.04.008).

SEI is a very thin film that forms on the surface of the negative electrode, created when part of the electrolyte breaks down during the first few charges, and it continues to grow a little bit after that. The electrolyte does not let electrons pass across, but at the surface of the negative electrode, it accepts electrons and breaks down, and the buildup of these breakdown products is this film (DOI: 10.1002/aenm.202203307). In one model, this film grew mainly while charging (DOI: 10.1021/acs.energyfuels.4c03927, model).

This film makes the cell work properly. Without the film, electrons would keep passing from the negative electrode into the electrolyte, causing the cell to wear out quickly on its own and lose performance, but the film blocks that transfer (DOI: 10.1149/2.1441707jes). Lithium ions pass back and forth through this film to enter and leave the negative electrode. The film is essential for the long-term performance of the cell, and it also affects the initial capacity lost, the degree of self-wearing, the ability to release power quickly, and safety (DOI: 10.1016/j.carbon.2016.04.008).

The problem is the material used to make the film. Lithium goes into the film, and the lithium trapped in the film cannot travel back and forth anymore. As the film grows, the lithium that can travel back and forth decreases (DOI: 10.1149/2.040310jes, model; DOI: 10.1149/2.1111702jes, lab cell). Visually, some marbles get stuck in the film on the shelf surface and cannot move back and forth anymore, and the amount that can be moved decreases by as many marbles as are stuck.

There is a model that explains the gradual loss of capacity over thousands of cycles in commercial cells using a graphite negative electrode quite accurately with a single film that grows slowly while competing with the reaction of lithium entering and leaving the graphite (DOI: 10.1149/2.044302jes). In a model built with a graphite negative electrode and a manganese oxide positive electrode, as the film got thicker, the growth speed slowed down (DOI: 10.1149/2.040310jes, model). There are also results in the opposite direction. In a lab cell where manganese dissolved from the positive electrode got mixed into the film, when the upper voltage at which charging stops was set high, the speed at which lithium was trapped did not slow down and continued evenly (DOI: 10.1149/2.1111702jes). There was also a cell where the repeating cycle of the film growing, breaking, and filling back in occurred over and over (DOI: 10.1002/aenm.202404997).

The resistance of the film itself also behaved differently from cell to cell. In a lab cell made with a nickel-rich positive electrode and a graphite negative electrode, the film resistance stayed almost the same after the initial few cycles, and what grew steadily was the resistance of the contact between particles and the side where lithium spreads through the solid (DOI: 10.1002/admt.202502455). In another lab cell aged by fast charging, the film resistance grew steadily until the capacity dropped by 10% (DOI: 10.1016/j.cej.2024.155588). The reason internal resistance grows does not lie in the film alone.

What this film is made of and how it changes have not yet been fully uncovered. A 2009 review called this film the "most important yet least understood" part of a cell (DOI: 10.1524/zpch.2009.6086). SEI is a film that lets the cell work, while also trapping a little bit of the mobile lithium as it grows.

The Second Path of Losing Beads — Lithium Solidifying on the Surface

The faster you charge, the more lithium ions rush all at once to the negative electrode. If that speed exceeds the rate at which they enter into the graphite, something else happens.

Lithium plating is when lithium ions fail to enter inside the negative electrode and instead accumulate as metallic lithium on the electrode's surface. Visually, it is like beads failing to fit into the shelf slots and clumping together on the outside. The accumulated lithium does not all return during discharge. Because the process of plating and then dissolving back out has poor reversibility, plating leads directly to capacity loss (DOI: 10.1021/acsenergylett.0c00859). According to the value modeled by the same study under those experimental conditions, the share of plated lithium that dissolved back out was 65%.

Plating does not form evenly across the entire negative electrode. When you look inside a rapidly charged laboratory cell using X-ray computed tomography, a moss-like tangled layer of metallic lithium accumulates on the face where the negative electrode meets the separator. That layer blocks the path, so less lithium went into the graphite particles directly underneath (DOI: 10.1021/acsnano.1c02942). This suggests that plating once formed on the surface can obstruct the normal charging underneath it.

When plated lithium grows like thorns and then breaks off, it becomes "dead lithium" that is connected nowhere (DOI: 10.1021/acsenergylett.4c02898). When plating is severe, a mass of metallic lithium can grow inside the electrode, swelling the graphite electrode and causing it to crack. This event observed in a coin-shaped laboratory cell suggests that the path of losing beads can spread into a path that damages the shelf (DOI: 10.1021/acsami.0c07803).

Lithium plating does not only reduce lifespan; it also affects safety (DOI: 10.1149/2.0621506jes). That story belongs to another piece in this series, the one covering Thermal Runaway — A Staircase Climbed on Its Own Heat.

The Path Where a Shelf Cracks — How Electrodes Decay

Every time lithium leaves and enters, the crystals of the cathode material shrink and swell. When nickel-rich cathode material cycled between 3.0 and 4.3 volts relative to lithium metal, the volume of a single crystal changed by about 5.1%, compared to about 1.2% for nickel-poor cathode material. Microscopic cracks formed more frequently in the nickel-rich particles, and this rapid aging was linked to particles fracturing due to stress at the boundaries where tiny crystals meet (DOI: 10.1021/acs.jpcc.6b12885, laboratory cell). When a crack forms on the shelf, the space where a bead can sit shrinks. When measuring the resistance of an aged cell at various states of charge, the increased resistance came mainly from the cathode when measured at high states of charge, and mainly from the anode when measured at low states of charge (DOI: 10.1002/admt.202502455, nickel-rich cathode/graphite laboratory cell).

Metal ions dissolved from the cathode cross the electrolyte and mix into the anode's SEI. In laboratory cells using nickel-cobalt-manganese cathodes, this process sped up noticeably when the upper voltage where charging stops, namely the upper charge voltage, exceeded 4.30 volts, and above 4.4 volts cathode particles fractured further, with the fractured faces becoming new surfaces that increased electrolyte decomposition and metal leaching. The authors estimate that a single manganese ion trapped inside the film traps roughly 100 more lithium ions. Even so, the main share of capacity loss in these cells was lithium lost at the anode film (DOI: 10.1149/2.1111702jes). This is a scene where the path of damaging the shelf accelerates the path of losing beads. The various mechanisms of aging are intertwined like this through feedback loops (DOI: 10.1039/d1cp00359c).

Even when the shelf is damaged, it sometimes does not show up right away in the capacity numbers. When laboratory cells made of nickel-manganese-cobalt cathodes and graphite anodes were run for 2.5 years, the cells lost a fair amount of cathode material. Because the anode was what determined the end of discharge, this loss did not show up in the capacity. However, the cathode material loss was tightly linked to cathode swelling and the drying up of electrolyte lost from the electrodes, and the authors note that this will eventually invite sudden failure (DOI: 10.1149/1945-7111/ac4b82). Even if the capacity numbers look fine, not all damage inside the cell is captured in those numbers.

It Ages When Left Alone, and It Ages When Used

A spare battery left in a drawer also ages. Calendar aging is the aging that proceeds as time passes even when left unused, and cycle aging is the aging that proceeds as charging and discharging are repeated. To speak of the speeds of the two types of aging, one more term is needed. The state of charge is the amount currently remaining in the cell compared as a ratio to the amount when fully charged, written as a percentage. The two types of aging are sometimes treated with separately calculated models (DOI: 10.1016/j.apenergy.2023.121531, LFP cell model). Factors that change the speed of aging include temperature, state of charge, speed of charge and discharge, and number of times used (DOI: 10.1016/j.est.2022.105634), with the depth traveled in one go being another among them (DOI: 10.1016/j.apenergy.2023.121531).

Heat advances aging, and cold reduces the efficiency of charging and discharging (DOI: 10.1039/d5ra00934k). When hot, the reaction in which metals dissolve from the positive electrode and a new film forms accelerates, reducing capacity and increasing internal resistance (DOI: 10.1002/smll.202407740). In laboratory studies where LFP positive electrode and graphite negative electrode cells were aged quickly at various temperatures, the process of film growing, cracking, and refilling was the main cause of aging, and its share was greater the higher the temperature (DOI: 10.1002/aenm.202404997). The path by which cold leads to aging lies on the charging side. Charging when cold makes it difficult for lithium ions to enter the negative electrode, causing lithium plating to appear earlier (DOI: 10.1021/acs.energyfuels.4c03927, model). When hot, films grow quickly; when filled quickly while cold, lithium easily hardens on the surface.

Aging when left alone varies depending on storage temperature and state of charge, and relationships are modeled by measuring cells under various storage conditions (DOI: 10.1109/tii.2019.2941747; DOI: 10.3390/en14061732, LFP cells stored for 27 to 43 months). The abstracts of both studies do not state whether higher or lower states of charge age faster.

In cycle experiments repeating charge and discharge, the voltage at which charging stops also touches upon aging. When laboratory cells made with an NMC positive electrode and a graphite negative electrode were continuously cycled, the dissolution of metal from the positive electrode into the negative electrode film noticeably accelerated past an upper charge voltage limit of 4.30 volts, and capacity decrease moved along with it (DOI: 10.1149/2.1111702jes). Laboratory cells made with a very nickel-rich positive electrode yielded slightly more initial capacity when filled up to 4.5 volts than when filled up to 4.3 volts. Even after being cycled 100 times, they still retained 87% (4.3 volts) and 81% (4.5 volts) of their initial capacity, with the two percentages compared against different initial amounts (DOI: 10.1021/acs.chemmater.7b03268).

Depth of discharge is a value indicating what percentage of the fully charged amount was drawn and used in a single use. If operating between a state of charge of 0% and 50%, the depth of discharge is 50%. When laboratory cells made with an NMC positive electrode and a natural graphite negative electrode were cycled for 2.5 years across seven state-of-charge ranges—namely four starting from the bottom (0–25%, 0–50%, 0–75%, 0–100%) and three ending at the top (75–100%, 50–100%, 25–100%)—along with three speeds and two temperatures, the degrees of lithium loss, resistance increase, and electrode material loss differed across the ranges (DOI: 10.1149/1945-7111/ac4b82). The abstract of this test does not state which range aged faster. The direction of depth comes from a single laboratory study that cycled commercial cells in 13 patterns mimicking actual car driving and back-calculated values inside the cells using a physical model. In this interpretation, using cells deeply at once appeared to spur cracking of positive electrode particles and loss of positive electrode material, and the results of dismantling the cells supported that interpretation (DOI: 10.1016/j.etran.2025.100410, model interpretation).

These numbers were all obtained by cycling cells made with specific electrode materials under conditions set in a laboratory. They cannot be directly transferred to the charging settings of a mobile phone or an electric vehicle.

What Changes When You Fill It Fast

Fast charging is filling a battery in a short time with a large current. If we take 1C as the current that drains a full battery in one hour (the scale called C-rate, covered in the primer on reading numbers), one laboratory cell made with a lithium iron phosphate cathode and a graphite anode showed no lithium plating when filled at 1C and 2C under those test conditions. At 3C, metallic lithium formed and began to grow, and at 4C to 6C, it grew branching out like thorns. In aged cells, the plated lithium and the organic components of the solid-electrolyte interphase layer grew exponentially as the charging got faster, while the inorganic components of the layer grew slowly (DOI: 10.1021/acsenergylett.4c02898). This threshold is the value for that cell and those conditions. There is also a model showing that the colder it is, the earlier plating appears (DOI: 10.1021/acs.energyfuels.4c03927, model). In cells with thick electrodes designed to hold more energy, two bottlenecks block fast charging: the electrolyte fails to supply lithium ions deep into the anode in time, and plating occurs quickly on the side of the anode that touches the separator (DOI: 10.1149/2.0451908jes, model).

The results also change depending on how the charging is done. In a laboratory test using a commercial cell made with a nickel-cobalt-manganese cathode and a graphite anode, researchers divided the state of charge into five sections and designed several methods applying different currents to each section. Among them, one method had a charging time 20% shorter than the standard charging method compared in the study, yet lasted 200 cycles longer, measured by the number of full charge-and-discharge cycles. Cells aged with this method showed less graphite damage and less growth of the anode film (DOI: 10.1016/j.jechem.2023.01.016, laboratory test). The authors of this study also start from the premise that it is inevitable for large currents to accelerate aging. Filling faster increases the risk of lithium hardening on the surface, but the magnitude varies depending on the temperature, cell design, and current delivery method. How much of the reduced lifespan in fast-charging tests is due to speed versus the temperature that rose during the test is covered in the primer on reading battery numbers.

Where the Curve Bends and "End of Life"

The aging curve sometimes bends at a certain point. There are cells where the capacity decreased evenly like a straight line until the first 10% was lost, and then grew steep after that (DOI: 10.1016/j.cej.2024.155588, laboratory test aging a nickel-manganese-cobalt cathode and graphite cell with fast charging). The straight section was driven by the growth of the negative electrode film, and after the bend, defects in the electrodes caused by long-term mechanical stress took over (DOI: 10.1016/j.cej.2024.155588).

There are also cells where the same bend is explained by a different cause. The sudden drop seen in the accelerated aging of an iron-phosphate cathode and graphite anode cell came from the additives put into the electrolyte running out, shifting decomposition toward the solvent and lithium salts. Internal resistance increase, lithium plating, electrolyte drying, and mechanical deformation were not the main causes (DOI: 10.1002/aenm.202404997). There is a place where the curve bends, but why it bends is reported differently depending on the cell type and conditions.

"End of life" is a line drawn by humans on top of this curve. Commonly, 80% of the initial capacity is used as that line, which is a customary definition. Even for the same cell, the path to reaching that line varies depending on how it is used, and at least five such paths emerged within a single model. However, fitting these model values to real cells still remains a major challenge, the authors write (DOI: 10.1039/d2cp00417h).

State of Health is an Estimated Value

The device settings screen shows a number telling you how much battery life remains compared to when it was new. That number has a name.

State of Health is a value representing how much performance the cell retains compared to when it was new, estimated by extrapolating from measurable values such as voltage, current, and temperature. There are three main paths for estimation: a model that translates the chemistry inside the cell into equations, a model that mimics the cell with a circuit of resistors and electrical components, and a method of learning relationships from accumulated data (DOI: 10.3390/wevj14090247). Accurate State of Health estimation is the foundation of battery management (DOI: 10.3390/en18061463).

A battery management system is a device that monitors the cell's voltage, current, and temperature to estimate State of Charge and State of Health, protects the cell, and balances the cells evenly. Voltage and current monitoring, charge and discharge estimation, protection, cell balancing, and thermal management are cited as the responsibilities of this device (DOI: 10.3390/batteries9030152). The authors of a study evaluating cell-to-cell misalignment using electric taxi operation data note that unmanaged imbalances between cells can accelerate aging (DOI: 10.1109/tte.2020.3018143). Measuring the returned electricity against the put-in electricity with extreme precision reveals even minor plating, and this measurement can be used for management that varies the fastest charging speed depending on temperature to avoid plating (DOI: 10.1149/2.0621506jes). A method to capture the moment plating begins using only voltage changes during rest has also emerged from room-temperature laboratory tests, and the authors view this as a method with commercialization potential (DOI: 10.1021/acsenergylett.0c00831).

Accurately measuring State of Health in an actual car is said to be "still difficult, and perhaps impossible." Departing from that problem, one study verified a method linking cell tests and pack calculations with data from laboratory modules, a pack, and an electric taxi in operation, and reports that both cases accurately estimated pack-level State of Health. The cells inside the pack aged along roughly similar paths (DOI: 10.1109/jestpe.2021.3112754).

Separately, many studies refining charging methods are based on fixed cycles of full charging and emptying, but actual usage is erratic and often stops midway (DOI: 10.1016/j.etran.2026.100561). In laboratory tests where cells were aged using a fixed scenario mimicking commute patterns and a scenario that changes according to State of Health, the authors note that adapting the charging method to match the State of Health extended lifespan while reducing total charging time (DOI: 10.1016/j.etran.2026.100561, 실험실 시험·모형). The State of Health shown by the device is an estimate derived from values the battery management system can measure.

Looking Back at a Phone Used for Two Years

If we go back to our first phone and lay out what we saw in this article, it looks like this. The charging and discharging repeated every day for two years piled up cycling aging, and the two years that elapsed whether used or resting piled up calendar aging. Most of the time spent plugged into the charger overnight was time spent sitting fully charged, and calendar aging ticked on during that time too. How much time spent sitting fully charged alters aging is not written in the abstract of the storage study this article looked at.

Meanwhile, the film on the negative electrode surface may have grown little by little and trapped the lithium moving back and forth. On days filled up quickly, especially cold days, some lithium may have solidified on the surface depending on the cell's design and charging method, and the positive electrode particles may have cracked.

How large each pathway grew depends on the type of cell inside that phone and how it was used. The representative numbers visible from the outside are just the two: the reduced capacity and the increased internal resistance, so the causes cannot be sorted out from those numbers alone. The health status number appearing on the device is also an estimate produced by the battery management system.

Apart from growing old, on a cold day the same cell has a larger internal resistance at that moment (DOI: 10.1021/acs.energyfuels.4c03927, model). This is a temporary effect depending on that day's temperature, so it is distinguished from aging. Inside a phone whose day has grown short, beads diminishing and shelves cracking may have been proceeding together, and the two pathways are intertwined (DOI: 10.1039/d1cp00359c).

Where to Look Next

To look at current, voltage, and resistance from the beginning, you can go to "Voltage, Current, Power — What Goes All the Way Around Undiminished". How to measure the amount you can take out in energy is covered by "Watts and Watt-Hours — Reading Electricity's Two Numbers Like a Speedometer and an Odometer".

Among the other pieces in this series, "The Battery Cell — A Rocking Chair Between Two Electrodes", which covers cell structure, details the beads and shelves of this article, namely the two paths lithium ions travel along and the electrode materials. "Ampere-hours and Watt-hours — How Much It Holds, How Fast It Gives It Back", which deals with reading battery numbers, covers the units for writing capacity, the point that state of charge is an estimate, and how the amount you can take out changes depending on how fast you take it out. "Thermal Runaway — A Staircase Climbed on Its Own Heat", which deals with thermal runaway, follows how lithium deposition and membranes connect to safety.

Source Papers

• "Aging aware operation of lithium-ion battery energy storage systems: A review" (2022) — DOI: 10.1016/j.est.2022.105634 — Results of aging are capacity fade, resistance increase, and safety impacts; state of charge, charge/discharge speed, cycle count, and temperature are aging factors.
• "Degradation of lithium-ion batteries under automotive-like conditions: P2D model-based understanding and ex-situ validation" (2025) — DOI: 10.1016/j.etran.2025.100410 — Capacity fade and internal resistance increase; in interpreting 13 automotive-like cycles retrofitted through a P2D model (confirmed by post-mortem analysis), cycle depth appears to drive cathode cracking and active material loss.
• "Numerical Research on Electrochemical Behavior, Thermal Characteristics, and Aging Formation of Lithium-Ion Cell at Different Ambient Temperatures and Charge/Discharge Rates" (2024) — DOI: 10.1021/acs.energyfuels.4c03927 — Model. Internal resistance and heat generation increase with higher charge/discharge speeds and lower ambient temperatures; lithium plating appears earlier in low-temperature, fast charging.
• "Lithium ion battery degradation: what you need to know" (2021) — DOI: 10.1039/d1cp00359c — 5 primary mechanisms, 13 secondary mechanisms, 5 observation modes, results of capacity and power fade, and feedback loops between mechanisms.
• "3D Detection of Lithiation and Lithium Plating in Graphite Anodes during Fast Charging" (2021) — DOI: 10.1021/acsnano.1c02942 — At moderate speeds, lithium is intercalated into graphite; during fast charging, moss-like lithium forms on the separator side, while particles beneath are less charged.
• "Lithium-ion battery degradation: how to model it" (2022) — DOI: 10.1039/d2cp00417h — Loss of lithium inventory and loss of active material; even identical cells reach the end of life through five distinct pathways depending on use, leaving model parameterization a major challenge.
• "A Formal Analysis of Internal Resistances of Li‐Ion Cells Using EIS and DRT to Simultaneously Probe Both Working Electrodes" (2026) — DOI: 10.1002/admt.202502455 — NMC811/graphite. SEI resistance stabilizes after an initial rise, contact and diffusion resistances steadily increase, and cathode resistance dominates at high states of charge.
• "The state of understanding of the lithium-ion-battery graphite solid electrolyte interphase (SEI) and its relationship to formation cycling" (2016) — DOI: 10.1016/j.carbon.2016.04.008 — SEI forms during initial operation and is essential for long-term performance, grown through slow charge/discharge in the formation stage.
• "Lithium Batteries and the Solid Electrolyte Interphase (SEI)—Progress and Outlook" (2023) — DOI: 10.1002/aenm.202203307 — SEI is formed by electrolyte decomposition and the accumulation of reduction products on the anode surface.
• "Review—SEI: Past, Present and Future" (2017) — DOI: 10.1149/2.1441707jes — Electron transfer from electrode to solution causes rapid self-discharge and performance degradation; preventing this is the core of SEI models.
• "A Comprehensive Capacity Fade Model and Analysis for Li-Ion Batteries" (2013) — DOI: 10.1149/2.040310jes — Model. SEI growth drives cyclable lithium loss, and growth slows down as it thickens.
• "Theory of SEI Formation in Rechargeable Batteries: Capacity Fade, Accelerated Aging and Lifetime Prediction" (2012) — DOI: 10.1149/2.044302jes — Model. Explains capacity fade over thousands of cycles in commercial graphite cells as SEI growth competing with reversible intercalation, predicted using high-temperature accelerated data.
• "The Solid Electrolyte Interphase – The Most Important and the Least Understood Solid Electrolyte in Rechargeable Li Batteries" (2009) — DOI: 10.1524/zpch.2009.6086 — SEI properties and composition are difficult to understand and analyze.
• "Quantification of Inactive Lithium and Solid–Electrolyte Interphase Species on Graphite Electrodes after Fast Charging" (2020) — DOI: 10.1021/acsenergylett.0c00859 — Lithium plating reduces to metal on the surface instead of between graphite layers, leading to immediate capacity loss due to low reversibility, with a stripping efficiency of 65% found by model.
• "Quantitative Analysis of the Coupled Mechanisms of Lithium Plating, SEI Growth, and Electrolyte Decomposition in Fast Charging Battery" (2024) — DOI: 10.1021/acsenergylett.4c02898 — LFP/graphite. No plating at 1C and 2C, nucleation and growth at 3C, dendritic growth at 4C to 6C, exponential increase of plated lithium and organic SEI with charging speed (inorganic SEI grows moderately), and detached "dead" lithium.
• "In Operando Detection of the Onset and Mapping of Lithium Plating Regimes during Fast Charging of Lithium-Ion Batteries" (2020) — DOI: 10.1021/acsami.0c07803 — Coin cell. Severe plating can swell and crack the graphite electrode, leading to active material loss.
• "In-Situ Detection of Lithium Plating Using High Precision Coulometry" (2015) — DOI: 10.1149/2.0621506jes — Plating significantly shortens lifespan and impacts safety, detectable via high-precision coulombic efficiency measurements, and usable for controlling maximum charging speed by temperature.
• "Anisotropic Lattice Strain and Mechanical Degradation of High- and Low-Nickel NCM Cathode Materials for Li-Ion Batteries" (2017) — DOI: 10.1021/acs.jpcc.6b12885 — Unit cell volume change between 3.0V and 4.3V is about 5.1% for NCM811 and about 1.2% for NCM111; accelerated aging of NCM811 is secondary particle collapse driven by grain boundary fracture.
• "Transition Metal Dissolution, Ion Migration, Electrocatalytic Reduction and Capacity Loss in Lithium-Ion Full Cells" (2017) — DOI: 10.1149/2.1111702jes — NCM523 full cell. Transition metal dissolution and SEI incorporation accelerate above an upper voltage limit of 4.30V; roughly 10² lithium ions are estimated to be additionally trapped per Mn ion, while the primary cause of capacity fade remains lithium loss at the anode SEI.
• "How do Depth of Discharge, C-rate and Calendar Age Affect Capacity Retention, Impedance Growth, the Electrodes, and the Electrolyte in Li-Ion Cells?" (2022) — DOI: 10.1149/1945-7111/ac4b82 — NMC622/natural graphite pouch. 7 state-of-charge ranges, 3 speeds, and 2 temperatures over 2.5 years show different degrees of loss per range; cathode material loss was anode-limited so it did not appear in capacity, but combined with swelling and electrolyte "dewetting" it will ultimately cause dramatic failure.
• "Increasing the lifetime profitability of battery energy storage systems through aging aware operation" (2023) — DOI: 10.1016/j.apenergy.2023.121531 — A model that separately linearizes calendar and cycle capacity loss for LFP cells.
• "A review of the combined effects of environmental and operational factors on lithium-ion battery performance: temperature, vibration, and charging/discharging cycles" (2025) — DOI: 10.1039/d5ra00934k — High temperatures accelerate aging, while low temperatures reduce charge/discharge efficiency (described by contrast).
• "Electrochemo‐Mechanical Degradation and Failure of Active Particles in High Energy Density Batteries: A Review" (2025) — DOI: 10.1002/smll.202407740 — High temperatures accelerate transition metal dissolution and new SEI formation, causing capacity fade and internal resistance increase.
• "Quantitative Analysis of Aging and Rollover Failure Mechanisms of Lithium‐Ion Batteries at Accelerated Aging Conditions" (2025) — DOI: 10.1002/aenm.202404997 — LFP/graphite. SEI growth, rupture, and repair are the primary causes of aging, and their share increases with temperature; the sharp drop is due to the depletion of the additive VC (not resistance, plating, electrolyte drying, or deformation).
• "Gaussian Process Regression With Automatic Relevance Determination Kernel for Calendar Aging Prediction of Lithium-Ion Batteries" (2019) — DOI: 10.1109/tii.2019.2941747 — Predicts calendar aging as a relationship between storage temperature, state of charge, and capacity across 9 storage conditions.
• "The Degradation Behavior of LiFePO4/C Batteries during Long-Term Calendar Aging" (2021) — DOI: 10.3390/en14061732 — LFP calendar aging over 27 to 43 months, modeling the effects of time, temperature, and state of charge.
• "Impact of Microcrack Generation and Surface Degradation on a Nickel-Rich Layered Li[Ni 0.9 Co 0.05 Mn 0.05 ]O 2 Cathode for Lithium-Ion Batteries" (2017) — DOI: 10.1021/acs.chemmater.7b03268 — Ni90 cathode shows higher initial capacity at a 4.5V upper limit (238 vs 227 mAh/g) and "still maintains" 87% at 4.3V and 81% at 4.5V after 100 cycles; reduction is caused by anisotropic volume change, microcracks, and surface rock-salt phases (worse at 4.5V).
• "Investigation of multi-step fast charging protocol and aging mechanism for commercial NMC/graphite lithium-ion batteries" (2023) — DOI: 10.1016/j.jechem.2023.01.016 — Commercial NMC532/graphite cell. Among a multi-step charging protocol across 5 state-of-charge segments, a designed method reduced charging time by 20% and extended equivalent full cycles by 200 compared to standard charging, reducing graphite damage and SEI growth, under the premise that large currents inevitably accelerate aging.
• "Requirements for Enabling Extreme Fast Charging of High Energy Density Li-Ion Cells while Avoiding Lithium Plating" (2019) — DOI: 10.1149/2.0451908jes — Model. Fast charging limits in high-energy cells are salt depletion inside the anode and plating at the graphite/separator interface.
• "Evolution of aging mechanisms and performance degradation of lithium-ion battery from moderate to severe capacity loss scenarios" (2024) — DOI: 10.1016/j.cej.2024.155588 — NMC/graphite, fast charging protocols. Transition from linear to non-linear starting at 10% loss; the linear phase is SEI growth, and the non-linear phase is electrode defects caused by mechanical stress.
• "A Systematic Literature Review of State of Health and State of Charge Estimation Methods for Batteries Used in Electric Vehicle Applications" (2023) — DOI: 10.3390/wevj14090247 — State estimation branches into electrochemical models, equivalent circuit models, and data-driven methods.
• "Deep Learning for State of Health Estimation of Lithium-Ion Batteries in Electric Vehicles: A Systematic Review" (2025) — DOI: 10.3390/en18061463 — Accurate SOH estimation is the foundation of the BMS.
• "Lithium-Ion Battery Management System for Electric Vehicles: Constraints, Challenges, and Recommendations" (2023) — DOI: 10.3390/batteries9030152 — BMS handles voltage and current monitoring, charge/discharge estimation, protection, balancing, and thermal management.
• "A Novel Consistency Evaluation Method for Series-Connected Battery Systems Based on Real-World Operation Data" (2020) — DOI: 10.1109/tte.2020.3018143 — Electric taxi operation data. Proposes an evaluation method under the premise that unmanaged cell imbalances can accelerate aging or cause accidents.
• "Detecting the Onset of Lithium Plating and Monitoring Fast Charging Performance with Voltage Relaxation" (2020) — DOI: 10.1021/acsenergylett.0c00831 — Detects the onset of plating via relaxation voltage analysis during room-temperature fast charging, with the authors foreseeing potential for commercial online detection.
• "Battery State-of-Health Estimation Based on Incremental Capacity Analysis Method: Synthesizing From Cell-Level Test to Real-World Application" (2021) — DOI: 10.1109/jestpe.2021.3112754 — Precise real-world SOH measurement is "still difficult, and perhaps impossible"; starting from this problem, pack-level SOH is accurately estimated in both lab packs and real-world electric taxis (RMSE of 0.0204, etc.), with cells inside the pack sharing similar degradation pathways.
• "Degradation analysis and tanks-in-series modeling of lithium-ion batteries with state of health-adaptive charging strategies" (2026) — DOI: 10.1016/j.etran.2026.100561 — Real-world use involves random, partial charge/discharge unlike fixed cycles; switching charges adapted to SOH during commuting pattern tests extends lifespan and shortens charging time.