Lithium-ion batteries were long considered free of memory effect, but that blanket claim turned out to be incomplete. In 2013, researchers demonstrated a genuine memory effect in lithium iron phosphate (LiFePO₄), one of the most widely used lithium-ion cathode chemistries, showing it could appear after just a single cycle of partial charge and discharge. The effect is subtle compared to the dramatic capacity loss seen in older nickel-cadmium cells, but the story is more nuanced than the simple “no, lithium-ion batteries don’t have memory” answer that has circulated for years.
What the Memory Effect Actually Is
The memory effect refers to a battery appearing to “remember” how much capacity was used in previous cycles and then delivering less energy on subsequent full discharges. In older rechargeable battery chemistries, if you routinely drained the cell to only about half its capacity before recharging, the battery would eventually behave as though it only had half its capacity. The voltage would dip prematurely at the point where you had been stopping, making the battery seem depleted even though stored energy remained.
This phenomenon was first studied in detail in nickel-cadmium (NiCd) batteries and later in nickel-metal hydride (NiMH) cells. Researchers found the root cause in the nickel electrode: repeated shallow discharge cycles caused a different crystalline phase to form in the electrode material. Specifically, a compound called γ-NiOOH would build up alongside the normal β-NiOOH, starting at the current collector side of the electrode and gradually spreading outward through the material during shallow cycling.1Journal of The Electrochemical Society. Possible Cause of the Memory Effect Observed in Nickel‐Cadmium Secondary Batteries This structural change lowered the working voltage of the cell, creating the characteristic voltage dip that defined the memory effect.2Bulletin of the Chemical Society of Japan. Cause of the Memory Effect Observed in Alkaline Secondary Batteries Using Nickel Electrode
Because lithium-ion batteries use completely different electrode chemistries, the assumption for nearly two decades was that they were immune to this problem. That assumption shaped how manufacturers marketed them and how consumers treated them. Unlike NiCd cells, nobody told you to “fully drain your lithium-ion battery before charging.” And for good reason: deep discharges actually stress lithium-ion cells. But the story changed when researchers looked more carefully at specific lithium-ion chemistries.
The 2013 Discovery in Lithium Iron Phosphate
The turning point came from a team publishing in Nature Materials who demonstrated that LiFePO₄, one of the most common cathode materials in lithium-ion batteries, exhibits a measurable memory effect after just one cycle of partial charge followed by partial discharge.3PubMed. Memory effect in a lithium-ion battery The effect manifests as a small but real voltage bump during subsequent charging or discharging. The battery’s voltage curve develops a kink at the state of charge where the previous partial cycle ended, as if the cell has recorded where you stopped.
The mechanism is different from what happens in NiCd batteries. LiFePO₄ is what scientists call a two-phase material: during charging and discharging, individual particles in the electrode flip between a lithium-rich phase and a lithium-poor phase. When you partially cycle the battery, you leave some particles in one phase and some in the other, and the distribution of those phases across all the particles in the electrode creates an asymmetry. On the next cycle, the battery “remembers” this distribution, producing a detectable voltage anomaly.
Follow-up research showed that the memory effect in LiFePO₄ depends heavily on how long the battery sits after partial cycling. The voltage bump is more pronounced when the cell rests for a while between the memory-writing partial cycle and the next full cycle, suggesting the effect is tied to the particles reaching a relaxation equilibrium during rest periods.4PubMed. Relaxation-Induced Memory Effect of LiFePO(4) Electrodes in Li-Ion Batteries The size of the electrode particles also matters: researchers found that the memory effect varies with particle size, especially after extended rest periods.5PubMed. Size-Dependent Memory Effect of the LiFePO(4) Electrode in Li-Ion Batteries
Why It Matters for State-of-Charge Estimation
You might wonder whether a small voltage bump really matters. After all, nobody is going to notice a few millivolts of change when using their phone. The practical concern is not about what you feel as a user but about what the battery management system (BMS) calculates behind the scenes. Modern devices estimate how much charge remains in the battery largely by reading voltage. Even a slight, unexpected shift in the voltage curve can throw off that estimate, causing the battery gauge to read higher or lower than reality.
The original Nature Materials paper flagged this explicitly: the slight voltage change from the memory effect can lead to substantial miscalculations in estimating the state of charge.6PubMed. Memory effect in a lithium-ion battery A study on large-format LiFePO₄ batteries, the kind used in electric vehicles and grid storage, confirmed that the memory effect becomes more prominent when the battery is repeatedly discharged to about 80% depth of discharge, which is a very common operating pattern in real-world applications.7Journal of Power Sources. Influence of memory effect on the state-of-charge estimation of large-format Li-ion batteries based on LiFePO4 cathode That finding is worth pausing on: the very usage pattern most people follow, charging before the battery is fully empty, is the pattern most likely to trigger the memory effect in these cells.
For electric vehicles, an inaccurate state-of-charge reading is more than an annoyance. If the car’s computer thinks the battery has 15% remaining when it actually has 10%, the driver could be stranded. If the system thinks the battery has more headroom for regenerative braking than it truly does, it might overcharge certain cells. Battery management systems in LiFePO₄-based vehicles and storage systems need to account for this subtle voltage distortion, and researchers have recommended incorporating memory-effect corrections into BMS algorithms.8Journal of Power Sources. Influence of memory effect on the state-of-charge estimation of large-format Li-ion batteries based on LiFePO4 cathode
Does This Affect Your Phone or Laptop?
Most smartphones and laptops today use lithium-ion batteries with different cathode chemistries, typically lithium cobalt oxide (LCO), lithium nickel manganese cobalt oxide (NMC), or lithium nickel cobalt aluminum oxide (NCA). The memory effect discovered in 2013 was specific to LiFePO₄. No equivalent memory effect has been documented in LCO, NMC, or NCA cathodes. So the conventional advice still holds for the vast majority of consumer electronics: you do not need to fully discharge your phone before charging it, and doing so would actually be worse for battery longevity.
Where LiFePO₄ does show up in consumer-facing products is in some electric vehicles, power tools, solar energy storage systems, and certain portable power stations. Tesla uses LiFePO₄ cells in its Standard Range vehicles. BYD’s Blade Battery is LiFePO₄. Many home solar storage units rely on it because of its long cycle life, thermal stability, and lower cost compared to NMC. If you own a device or vehicle using LiFePO₄ cells, the memory effect is at least theoretically relevant to you, though the battery management system should be handling the compensation without any action on your part.
That said, battery management systems in consumer devices already use strategies that minimize the practical impact. Most modern BMS designs combine voltage-based estimation with coulomb counting, which tracks the charge flowing in and out of the battery in real time. Some systems also recalibrate by referencing the open-circuit voltage when the device is in standby mode, which helps correct for accumulated errors in charge estimation.9Computers & Electrical Engineering. Programmable logic controller based lithium-ion battery management system for accurate state of charge estimation Periodic full charge-discharge cycles, which many EV manufacturers recommend once every few months, also help recalibrate the gauge and would incidentally erase any accumulated memory effect.
Memory Effects on the Anode Side
The research that first made headlines focused on the cathode, the LiFePO₄ electrode. But more recent work has shown that the anode side of a lithium-ion battery can develop its own form of memory behavior. A study examining graphite/silicon oxide blend anodes, a type increasingly used in high-energy-density cells, found that repeated partial cycling without a full discharge caused lithium to become trapped in a crystalline phase within the silicon oxide component. The trapped lithium effectively reduced the usable capacity of the anode, creating a memory-like degradation pattern.10Journal of Power Sources. Lithium trapping induced memory effect of Gr/SiOx blend anodes in lithium-ion batteries subjected to repeated partial cycling
The encouraging finding from that research is that a single deep discharge was enough to reverse the effect. When the cells were discharged to below about 5 to 10% state of charge, the phase transition in the silicon oxide reversed and the trapped lithium was released.11Journal of Power Sources. Lithium trapping induced memory effect of Gr/SiOx blend anodes in lithium-ion batteries subjected to repeated partial cycling This mirrors the old advice for NiCd batteries, “drain it fully once in a while,” except with a more specific threshold. It also suggests that as silicon-containing anodes become more common in next-generation lithium-ion cells designed for higher energy density, this particular form of memory effect could become more practically relevant.
A Broader Phenomenon Than Originally Thought
Researchers have mapped out a surprisingly long list of factors that influence how strong or weak the memory effect is in cells that exhibit it. A comprehensive study examined the impacts of average particle size, rest time, discharge cutoff voltage during the memory-writing cycle, lithium mobility, current rate, state of charge, depth of discharge, and the boundaries of the charge-discharge curve.12PubMed. Memory Effects’ Mechanism in the Intercalation Batteries: The Particles’ Bipolarization All of these variables modulated the memory effect’s strength, suggesting this is a complex, multi-parameter phenomenon rather than a simple on-off switch.
Separate from the memory effect itself, lithium-ion batteries also exhibit voltage hysteresis, a phenomenon where the voltage during charging follows a slightly different path than during discharging, even at the same state of charge. This is not memory in the traditional sense, since it doesn’t depend on cycling history, but it creates a similar practical challenge for state-of-charge estimation. Research has shown that hysteresis increases as the state of charge decreases, meaning the gap between the charge voltage and discharge voltage widens when the battery is more depleted.13Journal of Energy Storage. Study of hysteresis voltage state dependence in lithium-ion battery and a novel asymmetric hysteresis modeling For battery management systems, hysteresis and memory effect together create a layered challenge in knowing exactly how much energy is left.
The interaction between these phenomena matters because engineers designing BMS algorithms need to distinguish between a voltage anomaly caused by the memory effect (which depends on cycling history and goes away with a full cycle) and one caused by hysteresis (which is an inherent property of the cell chemistry). Lumping them together leads to inaccurate models; ignoring one while correcting for the other still leaves errors in the gauge reading.
How Charging Patterns Interact with the Effect
Real-world charging is rarely a simple constant-current affair. Electric vehicles, for instance, typically use staged constant-current charging, where the current steps down in stages as the battery fills up. Researchers studying how these real-world charging methods interact with the memory effect found that LiFePO₄ batteries that had long been assumed memory-free could indeed develop measurable memory artifacts under staged charging protocols.14IOP Conference Series: Earth and Environmental Science. Analysis of the memory effect of lithium iron phosphate batteries charged with stage constant current The specifics of the charging profile, how many stages are used, at what currents, and when the transitions happen, all influence whether and how strongly the memory signature appears.
This is relevant because it means the memory effect is not just a laboratory curiosity produced under idealized constant-current conditions. It can emerge under the exact charging protocols used in commercial applications. For engineers working on EV charging infrastructure or grid-scale storage systems that use LiFePO₄ cells, this interaction between charging strategy and memory behavior adds another variable to optimize.
Memory-Like Behavior Beyond Lithium-Ion
The discovery that lithium-ion batteries can exhibit memory raised an obvious question: do next-generation battery chemistries have similar issues? Early research on sodium-ion batteries suggests the answer is yes, at least in certain electrode materials. A study of tin-carbon composite electrodes for sodium-ion batteries found that specific voltage plateaus during desodiation (the equivalent of discharging in a sodium cell) disappeared depending on how fast the preceding sodiation (charging) had occurred, a behavior the researchers explicitly linked to a memory effect.15Advanced Functional Materials. Sodium Storage and Electrode Dynamics of Tin–Carbon Composite Electrodes from Bulk Precursors for Sodium‐Ion Batteries
Sodium-ion batteries are being developed as a lower-cost alternative to lithium-ion for grid storage and some vehicle applications, using abundant sodium instead of relatively scarce lithium. Finding memory-like behavior in these cells early in their development cycle is useful because it gives researchers the chance to design around the problem before the technology reaches mass production, rather than discovering it after billions of cells are already in the field, which is essentially what happened with LiFePO₄.
Practical Advice for Battery Users
Given everything above, here is what actually matters if you are trying to take care of your batteries:
- Standard consumer electronics: Your phone, laptop, or tablet almost certainly uses an NMC or LCO cathode. No documented memory effect applies. Charge whenever you want. The old “drain it completely first” advice from the NiCd era does not apply and will actually shorten your battery’s life.
- LiFePO₄ devices and vehicles: A memory effect exists but is subtle. If your device or vehicle recommends an occasional full charge-discharge cycle for calibration, follow that advice. It helps the battery management system maintain accurate gauge readings and incidentally clears any accumulated memory signature.
- Devices with silicon-containing anodes: Some newer high-capacity batteries blend silicon into the anode. If you exclusively top off between, say, 40% and 80% and never let the battery go below 10%, there is a theoretical risk of lithium trapping over many cycles. An occasional deeper discharge can reverse the effect.
- Solar storage and grid batteries: LiFePO₄ is the dominant chemistry in stationary storage. Operators managing large installations should ensure their BMS firmware accounts for memory-effect voltage distortions, especially if the system operates within a narrow state-of-charge window for extended periods.
The irony is that the best practices for lithium-ion longevity, avoiding deep discharges and keeping the battery in a moderate charge range, are the very conditions that can trigger the memory effect in susceptible chemistries. The good news is that the magnitude of the memory effect in lithium-ion cells is small enough that it mainly matters for estimation accuracy rather than usable capacity. You will not wake up one morning to find your EV has lost 30% of its range the way a NiCd power tool might have in the 1990s. But if you have ever noticed your battery gauge jump unexpectedly or your EV’s range estimate seem inconsistent from one day to the next, a subtle memory-related voltage distortion could be one piece of the puzzle.