Learning objectives
By the end of this lesson you will be able to:
- name six aging mechanisms and say where in the cell each one happens;
- link each mechanism to impedance rise, capacity fade or self-discharge;
- use the Arrhenius factor to estimate how temperature changes a degradation rate;
- explain why charging a cold cell makes the temperature–aging curve U-shaped.
Before you start: Part 01B · Cell architecture for the electrode sandwich, and Part 02 · Capacity and C-rate for what “2C” means.
The equations here are teaching models — a √t film, Arrhenius factors, one-line stress terms. They reproduce the direction, the shape and the rough relative size of each effect. They are not a state-of-health estimator, and no number on this page should be quoted for a real cell: quote the paper it came from.
01The whole chain on one page
Six mechanisms. Three effects. One outcome.
The mechanisms are side reactions and physical changes, and not every chemistry suffers from all of them. Their observable consequences fall into three useful signals: impedance rises, capacity falls and self-discharge rises. Other diagnostic quantities change too; these three are the simplified outputs the teaching model uses. Select a mechanism to trace where it ends up.
Where this comes from — Vetter et al. 2005 (the canonical taxonomy of these mechanisms) · Edge et al. 2021 (how they interact) · Birkl et al. 2017 (why they collapse into three measurable effects).
02Where it happens
The panel shows a zoomed cross-section of one repeating electrode sandwich: copper current collector, porous graphite negative electrode, separator, porous NMC positive electrode, aluminium current collector. Electrolyte fills every pore of the three middle layers — the structure from Part 01B.
The most common misconception is that these mechanisms happen in separate places. They do not. They happen on particle surfaces, inside particles and at material interfaces, distributed through the whole cell. In a pouch cell these sandwiches are stacked; in a cylindrical cell they are wound.
Where this comes from — Plett 2015, Vol. I for the porous-electrode picture · Vetter et al. 2005 for where each mechanism sits · geometry after Argonne’s battery fundamentals.
The rate law underneath all of it
Almost every mechanism here is a chemical reaction, and chemical reaction rates follow the Arrhenius law:
For a typical activation energy near 50 kJ/mol this gives the familiar rule of thumb: the rate roughly doubles for every +10 °C near room temperature. Within one thermally activated mechanism, heat usually accelerates the same reaction; across a real cell, temperature can also change which mechanism dominates.
| Temperature | Ea = 45 kJ/mol (corrosion) | 50 kJ/mol (SEI) | 62 kJ/mol (gassing) |
|---|---|---|---|
| −10 °C | 0.09× | 0.07× | 0.04× |
| 0 °C | 0.19× | 0.16× | 0.10× |
| 10 °C | 0.38× | 0.34× | 0.27× |
| 25 °C | 1.00× | 1.00× | 1.00× |
| 35 °C | 1.80× | 1.92× | 2.25× |
| 45 °C | 3.13× | 3.55× | 4.82× |
| 60 °C | 6.73× | 8.32× | 13.84× |
Where this comes from — Broussely et al. 2005 measured calendar-aging rates that follow Arrhenius behaviour · Waldmann et al. 2014 mapped how the dominant mechanism itself changes with temperature.
Corrosion — the catch-all mechanism
Corrosion is deterioration by chemical reaction with the surroundings, and inside a cell it is a catch-all for unwanted side reactions: the electrolyte solvent attacking the current collectors, or reacting with electrode materials. The products can themselves be corrosive, so the process can feed itself.
The cost is mostly impedance: less conductive area and poorer contact raise the internal resistance, so the cell loses power before it loses much charge. Like the other chemical mechanisms it speeds up with temperature — the teaching model uses an activation energy of 45 kJ/mol. Not every chemistry suffers every one of these reactions; the panels show the aluminium collector as one example. The next mechanism, passivation, is a special case of corrosion.
03Passivation — the SEI film
Negative electrodes that work at low potential — graphite above all, but also silicon and lithium metal — suffer a particular form of corrosion called passivation. The electrolyte solvent is not chemically stable at the potentials found in most lithium-ion cells, so it reacts with the electrode particles. The reaction products form a layer on the particle surfaces: the solid–electrolyte interphase (SEI).
That film is not purely bad news. It protects the graphite from further reaction, which is why the process slows so sharply after the first few cycles. But it never stops: the SEI keeps growing slowly for the whole life of the cell, and every nanometre of growth consumes lithium that can no longer be cycled. Two ideas do most of the work:
- √t growth. The film is its own diffusion barrier, so its thickness grows like the square root of time, not linearly.
- Arrhenius. Warmer means faster — roughly double per +10 °C.
Here is in years. In the teaching model, δ1 is 12 nm of film after one year at 25 °C, and capacity loss is proportional to film thickness. Set the panel to 45 °C and watch what a hot summer does to a cell that is doing nothing at all.
| Years | 0 °C | 25 °C | 45 °C |
|---|---|---|---|
| 1 | 97.9 % | 94.8 % | 90.3 % |
| 2 | 97.1 % | 92.7 % | 86.2 % |
| 5 | 95.4 % | 88.5 % | 78.2 % |
| 10 | 93.5 % | 83.7 % | 69.2 % |
Four times the time gives twice the film: 12 nm after one year at 25 °C, 24 nm after four.
Where this comes from — Pinson & Bazant 2013 derive the √t law and its lifetime consequences · Broussely et al. 2005 for the storage measurements · Vetter et al. 2005 for the film chemistry.
04Gas generation
Some chemistries produce gas while charging that, ideally, returns to its earlier aqueous state on discharge — lead-acid is the classic example. In lithium-ion cells, gas comes mainly from irreversible electrolyte decomposition, so the recombination term in the model below is a simplification. Either way, two quite different failure paths follow:
- The gas escapes. Through a breach in the enclosure, that capacity is gone for good — and in some cells the released gases are flammable or explosive. Overcharged lead-acid cells, for example, give off oxygen and hydrogen.
- The gas stays. In a sealed cell, pressure builds until the cell swells or ruptures unless it has a release vent. A vent makes the failure survivable, not free: venting loses electrolyte too.
The model balances generation against recombination, so the gas held in the cell approaches a plateau:
Within specification the plateau stays low. Push generation up with heat or overcharge and the plateau climbs past swelling, then past the vent. Move the overcharge slider and watch the two thresholds arrive in that order.
Where this comes from — Rowden & Garcia-Araez 2020, a review of which gases form, when, and what they do · Vetter et al. 2005 for electrolyte decomposition.
05Crystal formation
Material that leaves an electrode during discharge does not generally return to the same place when the cell is recharged. Instead, crystal structures tend to build up on the electrode surfaces. This is a very general mechanism in redox cells, and its consequence is geometric rather than chemical.
A porous electrode works because it has an enormous internal surface area — that is where the reaction happens. As crystals accumulate, the active area shrinks and the reaction has to squeeze through a smaller doorway. Resistance rises, so the cell loses its ability to deliver high power. This is a power problem before it is a capacity problem.
| Cycles | Active area | Resistance | Peak power | Max current |
|---|---|---|---|---|
| 0 | 100 % | 30 mΩ | 84 W | 30 A |
| 250 | 85 % | 35 mΩ | 72 W | 26 A |
| 500 | 76 % | 39 mΩ | 64 W | 23 A |
| 1000 | 64 % | 47 mΩ | 54 W | 19 A |
| 2000 | 49 % | 61 mΩ | 42 W | 15 A |
Where this comes from — Vetter et al. 2005 on surface-film and morphology changes · Han et al. 2019 on how loss of active surface shows up as power fade.
06Dendrites — crystal growth that becomes a safety problem
Tree-like metal structures can grow through the separator, raising the self-discharge rate or even causing a short circuit. In lithium-ion cells the trigger is specific and well understood: charging at low temperature or with excessive current can deposit lithium metal on the negative-electrode particles instead of inserting it into them — lithium plating. Some deposits develop filament-like structures that threaten the separator.
Why cold makes it worse is worth stating plainly. Intercalation is a kinetic process, so its rate falls with temperature like every other reaction. The charging current, however, is whatever the charger decides to push. When lithium arrives faster than the graphite can take it in, it has nowhere to go but onto the surface, as metal.
The heat map is the practical summary: the danger is the top-left corner — high current, low temperature. That is why a well-designed battery management system refuses to fast-charge a cold pack, and warms it first if it can.
| Temperature | Intercalation kinetics | Will it plate | Plating drive |
|---|---|---|---|
| −10 °C | 0.15× | 90 % | 12.5 |
| 0 °C | 0.27× | 84 % | 6.7 |
| 10 °C | 0.47× | 74 % | 3.6 |
| 25 °C | 1.00× | 55 % | 1.4 |
| 40 °C | 1.97× | 33 % | 0.4 |
Where this comes from — Waldmann, Hogg & Wohlfahrt-Mehrens 2018, the review this section is built on · Waldmann et al. 2014 for post-mortem evidence that plating dominates below about 25 °C.
07Volume change, cracking and loss of contact
Charging and discharging an intercalation electrode changes its volume, and that stress can crack the active material. This is a primarily mechanical mechanism, and one that responds strongly to how you cycle; temperature, materials, particle design and electrode construction matter too. Every insertion and removal of lithium swells and shrinks the host lattice. Repeat it a few thousand times and the particle behaves like any other fatigued material.
It damages the cell in two ways at once:
- Fracture and structural collapse take active material out of service → capacity fade. In some positive-electrode materials the structure collapses so far that it can no longer intercalate lithium at all.
- Broken binder and conductive additive electrically orphan particles that are still chemically fine → impedance rise.
| DOD | Cycles to 80 % | Equivalent full cycles |
|---|---|---|
| 20 % | 7,980 | 1,596 |
| 40 % | 2,640 | 1,056 |
| 60 % | 1,380 | 828 |
| 80 % | 870 | 696 |
| 100 % | 610 | 610 |
In this illustrative model, shallow cycles are gentler than deep ones, and halving the depth of discharge buys far more than halving the number of cycles. Limiting depth of discharge often extends cycle life in practice, but the best state-of-charge window is cell- and duty-specific and must be established from validated data.
Where this comes from — Christensen & Newman 2006 on stress generation and fracture during intercalation · Edge et al. 2021 on particle cracking and contact loss as coupled degradation modes.
08The whole cell — all six mechanisms at once
Everything above, running together. The stacked bars are the interesting part: change the duty and watch which mechanism is actually to blame change with it.
| Duty | Temperature | Capacity | Impedance | Self-discharge | Years to 80 % | Main cause |
|---|---|---|---|---|---|---|
| Shelf storage | 25 °C | 87.6 % | 1.59× | 2.0 %/mo | 12.1 | SEI |
| Hot operation | 45 °C | 65.1 % | 3.23× | 2.9 %/mo | 2.0 | SEI |
| Cold / fast charge | 0 °C | 63.2 % | 1.74× | 6.2 %/mo | 1.7 | Plating (dendrites) |
| Deep cycling | 30 °C | 70.2 % | 2.47× | 2.4 %/mo | 2.6 | SEI |
| Overcharge | 40 °C | 65.4 % | 3.60× | 3.7 %/mo | 2.2 | SEI |
Three results are worth pulling out by hand:
- Shelf storage at 25 °C is dominated by passivation, and takes over a decade to reach 80 %.
- Heat alone shortens it: storage at 45 °C reaches 80 % in about 3.5 years, and cycling at 45 °C (hot operation) in about two.
- Cold fast-charging hands the blame to plating instead — a different mechanism, a different warning sign, and a safety problem rather than a performance one.
One caveat the panel cannot show: measured capacity can appear to recover slightly after rest, temperature equilibration or a different diagnostic procedure. That does not reverse the underlying degradation, and deep discharge should not be used as a general rejuvenation method.
Where this comes from — Birkl et al. 2017 on separating the modes from measurable behaviour · Han et al. 2019 for whole-life-cycle behaviour · Attia et al. 2022 on why the trajectory can bend sharply downward (“knees”).
09Temperature shifts the aging balance
Higher temperature generally accelerates aging, so keeping cells inside an acceptable temperature range is the most effective single way to extend their life. It is the most actionable idea in the whole topic, and the panel below shows why.
The catch — and the reason the total curve is U-shaped rather than simply rising — is lithium plating. Cooling a cell slows the side reactions, but it also slows the intercalation you want, so a cold cell that is being charged plates lithium instead. Cold is only free if no current is flowing. Switch between shelf storage and cold / fast charge to see both halves. Both curves come from the model in section 08, so the U-shape is a result, not a second opinion. In this teaching model the slowest-aging temperature for the cycling duties sits between about 4 and 13 °C (around 20 °C for cold fast charging); the measurements cited below put the minimum nearer room temperature, a reminder that the model shows shape, not calibrated values.
| Duty | −10 °C | 0 °C | 10 °C | 25 °C | 40 °C | 55 °C |
|---|---|---|---|---|---|---|
| Shelf storage | >30 | >30 | >30 | 12.1 | 4.8 | 2.0 |
| Hot operation | 5.3 | 7.0 | 7.0 | 4.8 | 2.5 | 1.2 |
| Cold / fast charge | 0.8 | 1.7 | 2.6 | 2.9 | 2.0 | 1.1 |
| Deep cycling | 2.1 | 3.2 | 3.6 | 3.0 | 1.8 | 1.0 |
| Overcharge | 4.1 | 6.1 | 6.5 | 4.4 | 2.2 | 1.0 |
Where this comes from — Waldmann et al. 2014 is the key result: post-mortem analysis of cells aged from −20 to +70 °C, showing two competing regimes with a minimum near room temperature · Broussely et al. 2005 for the thermal half.
10Recap
| Mechanism | Where it lives | Feeds mainly | Accelerated by | Caveat |
|---|---|---|---|---|
| Corrosion | Collectors, active and inactive materials | Impedance ↑ | Temperature | A catch-all term; not every chemistry suffers every reaction |
| Passivation (SEI) | Graphite particle surfaces | Capacity ↓ | Temperature, time | Forms on low-potential negative electrodes (graphite, silicon, lithium metal) |
| Gas generation | Pores, then the cell’s free volume | Capacity ↓, safety | Temperature, overcharge | Species and rates depend on chemistry, SOC and abuse history |
| Crystal formation | Electrode surfaces | Impedance ↑ → power fade | Cycling | Very general to redox cells; morphology varies enormously |
| Dendrites / plating | Negative electrode → through the separator | Self-discharge ↑, shorts | Cold charging, high current | A risk condition, not normal operation |
| Volume change | Inside particles, binder, additives | Capacity ↓ and impedance ↑ | Deep cycling | Severity depends on chemistry, particle size, electrode design |
- Aging is gradual and essentially irreversible, and it ends in failure. Rest or a different diagnostic procedure can show an apparent short-term recovery, but it does not undo degradation.
- This teaching model summarises aging through impedance ↑, capacity ↓ and self-discharge ↑.
- Temperature strongly changes degradation rates, while cold charging adds a distinct plating risk.
11Check yourself
Work these out from the panels above before opening the answers.
The SEI film is a degradation product. Why is a cell with no SEI at all worse off?
Because the film is what stops the reaction. Without a passivating layer the solvent would keep reacting with the graphite at its initial, very fast rate. The SEI is the reason the loss curve bends over into √t instead of continuing as a straight line — a self-limiting side reaction is enormously better than an unlimited one. In section 03, compare the film after one year with the film after four.
A pack is fast-charged at 0 °C every day. Which mechanism dominates, and which effect shows up first?
Lithium plating. In section 06, set 0 °C and 2–3C: the teaching model shows a much larger plating drive than at 25 °C. Irreversible capacity loss is a common early consequence when plated lithium becomes electrically isolated. Increased self-discharge is a more serious possible symptom if deposits create leakage paths or a soft internal short; it is not guaranteed to appear first.
Two identical cells sit on a shelf, one at 25 °C and one at 45 °C. Roughly how much faster does the hot one age?
About 3.5× faster, from the Arrhenius factor for a reaction near 50 kJ/mol — a little under two doublings (1.92 × 1.85 ≈ 3.5). In section 08, set shelf storage and compare: time to 80 % goes from about 12 years to about 3.5, with the cell doing no work at all.
Why is a rising self-discharge rate treated as more alarming than a slow capacity fade?
Because an unexpected rise can indicate a leakage path or soft internal short and therefore deserves investigation. Self-discharge can also come from parasitic reactions or electronic leakage without separator failure, so it is a warning signal, not proof of one specific mechanism.
Your pack lives in a hot climate and is fast-charged. You can fix one thing. Which?
Keep the cells inside their validated operating window and manage charge power with temperature. For the illustrative cold / fast charge duty, the model favours around 20 °C over 45 °C. For a pack that is already cold, reduce the charge current or pre-heat it — extra cooling would raise the plating risk.
12Practice quiz
Four questions on this lesson, marked in your browser — including Arrhenius and film-growth drills with fresh numbers every time.
13References
The narrative follows the normal-aging treatment in the course text. Everything else is drawn from the reviews and primary papers below.
Course text
- G. L. Plett, Battery Management Systems, Volume I: Battery Modeling, Artech House, 2015. ISBN 978-1-63081-023-8.
Mechanism taxonomy and overviews
- J. Vetter, P. Novák, M. R. Wagner, C. Veit, K.-C. Möller, J. O. Besenhard, M. Winter, M. Wohlfahrt-Mehrens, C. Vogler, A. Hammouche, “Ageing mechanisms in lithium-ion batteries,” Journal of Power Sources 147, 269–281 (2005). doi:10.1016/j.jpowsour.2005.01.006
- M. Broussely, Ph. Biensan, F. Bonhomme, Ph. Blanchard, S. Herreyre, K. Nechev, R. J. Staniewicz, “Main aging mechanisms in Li ion batteries,” Journal of Power Sources 146, 90–96 (2005). doi:10.1016/j.jpowsour.2005.03.172
- X. Han, L. Lu, Y. Zheng, X. Feng, Z. Li, J. Li, M. Ouyang, “A review on the key issues of the lithium ion battery degradation among the whole life cycle,” eTransportation 1, 100005 (2019). doi:10.1016/j.etran.2019.100005
- J. S. Edge, S. O’Kane, R. Prosser, N. D. Kirkaldy, A. N. Patel, A. Hales, A. Ghosh, W. Ai, J. Chen, J. Yang, S. Li, M.-C. Pang, L. Bravo Diaz, A. Tomaszewska, M. W. Marzook, K. N. Radhakrishnan, H. Wang, Y. Patel, B. Wu, G. J. Offer, “Lithium ion battery degradation: what you need to know,” Physical Chemistry Chemical Physics 23, 8200–8221 (2021). doi:10.1039/D1CP00359C
Individual mechanisms
- M. B. Pinson, M. Z. Bazant, “Theory of SEI formation in rechargeable batteries: capacity fade, accelerated aging and lifetime prediction,” Journal of The Electrochemical Society 160, A243–A250 (2013). doi:10.1149/2.044302jes — the √t law used in section 03.
- B. Rowden, N. Garcia-Araez, “A review of gas evolution in lithium ion batteries,” Energy Reports 6, 10–18 (2020). doi:10.1016/j.egyr.2020.02.022 — section 04.
- T. Waldmann, B.-I. Hogg, M. Wohlfahrt-Mehrens, “Li plating as unwanted side reaction in commercial Li-ion cells — a review,” Journal of Power Sources 384, 107–124 (2018). doi:10.1016/j.jpowsour.2018.02.063 — section 06.
- J. Christensen, J. Newman, “Stress generation and fracture in lithium insertion materials,” Journal of Solid State Electrochemistry 10, 293–319 (2006). doi:10.1007/s10008-006-0095-1 — section 07.
Diagnosis, trajectories and temperature
- C. R. Birkl, M. R. Roberts, E. McTurk, P. G. Bruce, D. A. Howey, “Degradation diagnostics for lithium ion cells,” Journal of Power Sources 341, 373–386 (2017). doi:10.1016/j.jpowsour.2016.12.011
- T. Waldmann, M. Wilka, M. Kasper, M. Fleischhammer, M. Wohlfahrt-Mehrens, “Temperature dependent ageing mechanisms in lithium-ion batteries — a post-mortem study,” Journal of Power Sources 262, 129–135 (2014). doi:10.1016/j.jpowsour.2014.03.112 — the U-shape in section 09.
- P. M. Attia, A. Bills, F. Brosa Planella, P. Dechent, G. dos Reis, M. Dubarry, P. Gasper, R. Gilchrist, S. Greenbank, D. Howey, O. Liu, E. Khoo, Y. Preger, A. Soni, S. Sripad, A. G. Stefanopoulou, V. Sulzer, “Review — ‘Knees’ in lithium-ion battery aging trajectories,” Journal of The Electrochemical Society 169, 060517 (2022). doi:10.1149/1945-7111/ac6d13
Schematic geometry
- Argonne National Laboratory, Battery fundamentals.
- NREL, SEI, plating and cracking degradation models.
- DOE / Idaho National Laboratory, Degradation during fast charge.
Prefer to change the equations yourself? Run the Part 04 notebook in your browser — the same panels plus the Python model behind every number on this page.