The Signal
RIKEN Team Derives Entropy Increase from Quantum Mechanics
A RIKEN-led group derives the thermodynamic entropy-increase law from quantum mechanics using a new definition of equilibrium and thermodynamic operations. The proof includes pure quantum states but has a stated limit for operations whose duration grows with system size.

A RIKEN-led group has derived two distinct forms of the thermodynamic second law from quantum mechanics, including the rule that entropy does not decrease during an adiabatic operation. The researchers’ framework treats thermal equilibrium, operations and entropy through macroscopic physical quantities, and includes pure quantum states among the states that can represent equilibrium.
The work, by researchers from RIKEN, the University of Tokyo and Osaka University, appeared online in PRX Quantum on November 3, Japan time November 4. The release says the result addresses how thermodynamic laws can arise from quantum mechanics, whose equations allow reversible time evolution.
iMATE gives equilibrium a quantum definition
The difficulty is that a quantum state can correspond to thermal equilibrium even when it is pure. Yet allowing every mathematically possible unitary operation on such a state would permit transitions that conflict with the second law. The standard von Neumann entropy also fails to match thermodynamic entropy for a pure state: it is zero for a pure state, even when that state represents thermal equilibrium.
The team defines equilibrium using additive observables—quantities such as energy, particle number and total magnetization whose values come from adding contributions across parts of a system. A state qualifies when all additive observables have macroscopic values matching those of the corresponding thermal-equilibrium state. The researchers call this framework iMATE, short for “infinite-observable macroscopic thermal equilibrium”; pure states can satisfy its conditions.
Only macroscopic controls count as thermodynamic operations
The framework distinguishes operations that manipulate macroscopic controls from the full range of operations allowed by quantum mechanics. In thermodynamics, controls include variables such as a magnetic field or volume. The group defines a macroscopic operation as time evolution caused by changing a finite number of external control variables that act on additive observables.
That definition excludes finely controlled interventions on individual degrees of freedom, which the release says are not thermodynamic operations. With the operation restricted in this way, the researchers prove Planck’s principle: returning the external controls to their starting values cannot lower the system’s energy by a macroscopically significant amount. For a closed system, the energy decrease would correspond to work extracted from it.
A quantum macro-entropy matches thermodynamic entropy
To address entropy increase, the team introduces a quantum macro-entropy density. It is defined from the von Neumann entropy of a state averaged over a region of fixed size. For every state in iMATE, including pure states, the researchers show that this quantity equals the thermodynamic entropy density.
They then study what happens after a macroscopic operation, while the controls are held fixed and the system relaxes. Under conditions the release describes as physically natural, the quantum macro-entropy density does not decrease. When the final state of relaxation is again in thermal equilibrium, the result becomes the thermodynamic rule that entropy cannot decrease under an adiabatic operation.
The proof has a limit on operation time
The release qualifies the result: the proof holds when the duration of the macroscopic operation does not increase as the system grows. Some familiar thermodynamic processes, such as moving a piston slowly, can require longer operation times for larger systems, so they fall outside that stated condition.
The researchers also find that applying the theory unchanged to such long operations can produce counterexamples to the second law. The release points to quantum simulators as a way to test those counterexamples and says the next task is to identify what additional conditions on states or operations would preserve the second law during long operations.
Source
Related stories
- Southeast University Team Demonstrates Quantum Mpemba Effect in a Qubit
- A Dissipative Controller Adds a Hidden Cost to Quantum Gates
- Perimeter Team Finds Multiple Peaks in Noisy Quantum Circuits
- Photonic Quantum-Walk Chip Shows Integer Mixed-State Uhlmann Winding Numbers
- Twist-angle-dependent quantum phases mapped in bilayer MoTe₂