Mpemba Effect Field Guide: When “Hotter” Can Reach Equilibrium Faster

Date: 2026-03-04
Category: knowledge

Why this is interesting

The Mpemba effect is the claim that, under some conditions, a hotter system can cool (or even start freezing) faster than a cooler one.

That sounds impossible at first glance. But the modern lesson is not “thermodynamics is wrong” — it is:

• nonequilibrium paths matter,
• initial conditions can suppress slow relaxation modes,
• and “closer in temperature” does not always mean “closer in dynamical state.”

First, what the effect is not

Not a violation of the first law

No energy law is broken. A hotter sample still contains more energy. The puzzle is about time to a target state under specific dynamics, not energy conservation.

Not guaranteed in your freezer

“Hot water always freezes faster” is false. In kitchen-scale water experiments, results are highly sensitive to setup details (container geometry, evaporation, supercooling behavior, dissolved gases, contact with freezer surface, air flow).

Not one single mechanism

For water, several coupled mechanisms can contribute. There is no universal one-line explanation that works for every experiment.

Why freezing experiments are so messy

A key source of confusion is mixing two stages:

1. Cooling to near 0°C
2. Nucleation/freezing onset (often after supercooling)

Because supercooling and nucleation are stochastic and history-dependent, “time to first ice” can vary a lot run-to-run.

This is exactly why some studies emphasize that if two water samples are truly identical except initial temperature and are cooled under the same conditions to a specified temperature, the hotter one should not beat the colder one in a trivial way — unless other path-dependent effects intervene.

The modern framework: relaxation modes

Recent theory reframes Mpemba-like behavior as a relaxation-geometry problem:

• Any relaxing system can be decomposed into decaying modes.
• Long-time behavior is dominated by the slowest decaying mode (SDM).
• If an initially hotter state overlaps less with the SDM than a cooler state, it can approach equilibrium faster.

In the strong Mpemba effect, overlap with the SDM can vanish, giving an exponential speedup governed by the next-fastest mode.

This perspective explains why the effect shows up beyond water: colloids, spin systems, granular media, and now quantum systems.

The experiment that cleaned up the debate (for generic systems)

In 2020, Kumar & Bechhoefer showed reproducible Mpemba behavior in a controlled colloidal setup (Nature):

• engineered potential landscape,
• well-defined quench protocol,
• quantitative agreement with theory,
• and observation of a strong effect (exponentially faster relaxation under tuned conditions).

This mattered because it moved discussion from “did your ice tray do a weird thing?” to a precise nonequilibrium phenomenon that can be designed and measured.

Inverse and quantum versions

Inverse Mpemba effect

In 2022 (PNAS), the same line of work reported anomalous heating: under specific conditions, an initially colder state heated up faster than a less-cold state when both were coupled to the same bath.

Quantum strong Mpemba effect

In 2025 (Nature Communications), experiments with a trapped-ion open quantum system reported observation of quantum strong Mpemba behavior, where tailored initial superposition states suppress the slowest decay mode and accelerate relaxation.

So the “Mpemba family” is now broader than water freezing: it is a toolkit for relaxation speed engineering.

Practical intuition (one-liner mental model)

Equilibrium distance is not a straight thermometer line.

A “hotter” start can sit on a faster lane in state space, while a “cooler” start can be stuck on a slow manifold.

Why this matters beyond curiosity

If you can shape initial conditions and dissipation channels, Mpemba-like effects may help with:

• faster thermal reset protocols,
• accelerated state preparation,
• improved control of stochastic/quantum devices,
• and more generally, shortcut strategies in nonequilibrium control.

Bottom line

“Hotter freezes faster” is not a universal kitchen law.
But Mpemba-like behavior is a real and now experimentally grounded nonequilibrium phenomenon: relaxation speed depends on path geometry, not just starting temperature rank.

References

1. Jeng, M. (2006). The Mpemba effect: When can hot water freeze faster than cold? American Journal of Physics, 74, 514–522. DOI: 10.1119/1.2186331
   https://ui.adsabs.harvard.edu/abs/2006AmJPh..74..514J/abstract
2. Burridge, H. C., & Linden, P. F. (2016). Questioning the Mpemba effect: hot water does not cool more quickly than cold. Scientific Reports, 6, 37665.
   https://www.nature.com/articles/srep37665
3. Lu, Z., & Raz, O. (2017). Nonequilibrium thermodynamics of the Markovian Mpemba effect and its inverse. PNAS, 114(20), 5083–5088.
   https://www.pnas.org/doi/10.1073/pnas.1701264114
4. Kumar, A., & Bechhoefer, J. (2020). Exponentially faster cooling in a colloidal system. Nature, 584, 64–68.
   https://www.nature.com/articles/s41586-020-2560-x
5. Kumar, A., Chétrite, R., & Bechhoefer, J. (2022). Anomalous heating in a colloidal system. PNAS, 119(5), e2118484119.
   https://www.pnas.org/doi/10.1073/pnas.2118484119
6. Ma, Y. et al. (2025). Observation of quantum strong Mpemba effect. Nature Communications, 16, 318.
   https://www.nature.com/articles/s41467-024-54303-0