🔢 mathematics

Limitations on Joint Coherence Transfer in Quantum Thermodynamics

This paper demonstrates that coherence transfers which are individually optimal cannot generally be achieved simultaneously by a single thermodynamic process due to fundamental compatibility constraints arising from energy conservation, trace preservation, and Gibbs preservation, a conclusion supported by both analytic bounds and computer-assisted semidefinite programming.

Original authors: Piotr Ćwikliński, Michał Studziński

Published 2026-08-13

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Piotr Ćwikliński, Michał Studziński

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Quantum Dance Floor: Why You Can't Have Your Cake and Eat It Too

Imagine a world where energy isn't just a number, but a strict bouncer at a club. In the realm of quantum thermodynamics, this bouncer is the Second Law of Thermodynamics, but with a twist: it cares deeply about "coherence." Think of coherence as a perfectly synchronized dance move. If a group of quantum particles (like electrons or atoms) are dancing in perfect unison, they hold a special kind of energy potential that can be used to do work, run engines, or power future computers. However, the universe has a rule: you can't just create this perfect dance out of thin air. You have to trade it for heat, and the "bouncer" (thermodynamics) decides exactly how much dance you can keep and how much you must lose.

Now, imagine you have a complex dance routine with several different pairs of dancers. You might ask: "If I can make Pair A dance perfectly, and I can make Pair B dance perfectly, can I make both pairs dance perfectly at the same time using the same music?" This is the big question scientists are asking. In the quantum world, the "music" is a thermodynamic process, and the "dancers" are energy levels. For a long time, physicists thought that if you could optimize one pair of dancers, you could probably optimize the others too, as long as you had the right setup. But what if the rules of the dance floor itself prevent everyone from shining at once? This paper dives into that exact problem, exploring whether the universe allows us to have our quantum cake and eat it too, or if there's a hidden limit that forces us to choose.

The Paper's Big Discovery: The "One Unitary" Bottleneck

The authors, Piotr Ćwikliński and Michał Studziński, tackle a tricky puzzle: Can multiple quantum coherence transfers be optimized simultaneously? In simpler terms, if you have a machine that can move energy and "dance moves" (coherence) around, can you tune it to get the absolute best result for three different tasks at the exact same time?

Their answer is a resounding "No, not always."

They discovered that while you can often find a perfect solution for one specific task (like moving coherence from energy level A to B) or another (moving it from C to D), you generally cannot find a single physical process that does both perfectly at the same time. It's like trying to tune a radio to two different stations at full volume simultaneously; the physics of the machine just won't allow it.

How They Figured It Out: The "Shell" Analogy

To understand why, the authors used a clever mathematical trick. They imagined the quantum system interacting with a "bath" (a giant reservoir of heat) as a series of energy shells. Think of these shells as different floors in a building. The system and the bath can only swap energy if they land on the same floor.

When the system tries to move a "dance move" (coherence) from one state to another, it's like a dancer jumping from one floor to another. The paper shows that the "amplitude" (the strength) of this jump is determined by a vector (a mathematical arrow) associated with that specific energy floor.

The Catch: To get the perfect jump (maximum coherence), these arrows must point in the exact same direction (they must be "collinear").
The Problem: If you want to maximize three different jumps at once, you need three different sets of arrows to all line up perfectly at the same time. The authors proved that because all these jumps must come from one single, global interaction (one big dance move between the system and the bath), these arrows often get stuck pointing in different directions. You can't force them to align without breaking the rules of the dance.

The "Polygon" Obstacle

The paper also introduces a fun geometric rule called the "Polygon Obstruction." Imagine you have a set of arrows representing the strength of different transfers. If you try to make them all as long as possible, they might form a shape that doesn't close.

In math, if you add up a bunch of vectors and they must sum to zero (because of conservation laws), they have to form a closed loop, like a polygon.
If one arrow is too long compared to the others, you can't close the loop.
The authors showed that for certain setups, the "perfect" lengths required for individual tasks are so unbalanced that they can't form a closed polygon. Therefore, a single process cannot achieve all those perfect lengths simultaneously.

The Computer Proof: A Rigorous "No"

To prove this wasn't just a guess, the authors ran a rigorous computer simulation on a specific quantum system with a Hamiltonian (energy setup) of H = diag(0, 1, 1, 3). This system has a "mixed degeneracy," meaning one energy level is unique, but another level is shared by two states (like having two dancers on the same floor).

They asked a computer to find a "magic channel" (a thermodynamic process) that could maximize three specific coherence transfers at once.

The Result: The computer proved that while you can get the best possible result for Transfer A, and the best possible for Transfer B, and the best possible for Transfer C individually, no single channel can achieve all three best results at the same time.
The Evidence: They didn't just rely on a standard computer guess. They used "machine-certified" math (using rigorous ball arithmetic) to provide a "witness" that proves the gap is real. They showed that the sum of the individual best scores is strictly higher than the best score you can get for the combined task.
The Conclusion: Since every "Thermal Operation" (a standard thermodynamic process) is a type of "Covariant Gibbs-Preserving Channel" (the broader class they tested), this incompatibility applies to real thermal operations too.

What This Means for the Future

The paper doesn't just say "it's hard"; it says "it's fundamentally impossible to have it all."

What is proven: For the specific system they tested, there is a strict mathematical gap. You cannot simultaneously optimize three specific coherence transfers.
What is still open: The authors note that while they proved the impossibility for the broader class of channels, they haven't yet proven that the individual optimal values for thermal operations specifically are incompatible. However, their findings strongly suggest that the "incompatibility" is a real feature of quantum thermodynamics, not just a quirk of their specific math model.

In short, the universe has a limit on how much "perfect coordination" you can squeeze out of a single thermodynamic process. If you want to optimize one part of your quantum dance, you might have to sacrifice the perfection of another. It's a beautiful, if slightly frustrating, reminder that in the quantum world, you can't have your cake and eat it too.

Technical Summary: Limitations on Joint Coherence Transfer in Quantum Thermodynamics

Problem Statement
In quantum thermodynamics, thermal operations (TO) are constrained by energy conservation, time-translation covariance, and Gibbs preservation. While individual coherence transfers between specific energy levels are known to be bounded by scalar damping factors (derived from the Damping-Matrix-Positivity or DMP analysis), a fundamental question remains: if several coherence transfers can be individually optimized to their theoretical limits, can they be simultaneously optimized by a single thermodynamic process? The authors investigate whether "joint sharpness" is possible when multiple coherence transfers are targeted simultaneously, particularly in systems with mixed energy-level degeneracy (where eigenspaces have unequal dimensions).

Methodology
The paper employs a multi-faceted approach combining analytic derivations, geometric constraints, and rigorous computer-assisted optimization:

Shell-Amplitude Representation: The authors utilize an energy-resolved Stinespring representation of thermal operations. They express elementary coherence-transfer amplitudes as inner products of "shell-amplitude vectors" ( $|w_{r|i}\rangle$ ) associated with total-energy shells. These vectors are bath-weighted and arise from a single global energy-preserving unitary.
Cauchy–Schwarz and Collinearity: They establish that saturating an elementary coherence-transfer bound is equivalent to saturating a Cauchy–Schwarz inequality, which forces the corresponding shell-amplitude vectors to be collinear. Joint optimization requires these collinearity conditions to be simultaneously satisfiable within one common dilation.
Polygon Obstructions: The authors derive "polygon inequalities" based on trace and Gibbs preservation. These conditions require that complex transfer amplitudes, whose moduli are fixed by saturation, must sum to zero (forming a closed polygon). If the fixed moduli violate the triangle inequality (i.e., one length exceeds the sum of the others), simultaneous saturation is impossible.
Choi-Sector Gram Matrices: The problem is reformulated using positive semidefinite Gram matrices (Choi/DMP representations). The authors analyze how rank-deficient directional blocks (resulting from saturation) impose range-inclusion constraints on coupled data in higher-dimensional sectors.
Semidefinite Programming (SDP) Certification: To provide a rigorous counterexample, the authors analyze a mixed-degenerate Hamiltonian $H = \text{diag}(0, 1, 1, 3)$ . They optimize over the larger class of Covariant Gibbs-Preserving (CGP) channels using SDP. They employ machine-certified primal and dual witnesses verified in rigorous ball arithmetic to prove incompatibility, avoiding reliance on floating-point solver outputs.

Key Contributions and Results

Incompatibility of Joint Saturation: The paper demonstrates that coherence transfers which are individually optimal are generally not compatible with one another under a single physical process. Specifically, for the Hamiltonian $H = \text{diag}(0, 1, 1, 3)$ , three directional coherence transfers ( $0 \to 1a$ , $1b \to 3$ , and $0 \to 3$ ) are shown to be individually optimal within the class of CGP channels but cannot be simultaneously attained by any single CGP channel.
Rigorous Separation: Using certified SDP bounds, the authors prove a strict separation between the sum of individual optima and the maximum attainable value for a joint optimization. The certified gap is $> 0.014882$ . Since every thermal operation is a CGP channel, this result implies that no thermal operation can simultaneously attain these three specific CGP-optimal values.
Structural Mechanisms:
- Collinearity Constraints: Joint saturation imposes a "saturation graph" where vectors from different sectors must align. In mixed-degenerate systems, this forces alignments between vectors of unequal dimensions, which is structurally restrictive.
- Polygon Obstructions: Trace and Gibbs preservation impose phase-closure conditions. If the diagonal transition probabilities are fixed, the resulting polygon inequalities can rule out simultaneous saturation even if individual bounds are theoretically reachable.
- Range-Inclusion Constraints: In the Choi-sector Gram description, saturation in one direction (creating a rank-one block) forces coupled data in other directions to lie within the range of that rank-one vector. This leads to polynomial alignment equations (determinantal conditions) that restrict joint saturation to a measure-zero set in the parameter space.
Distinction between TO and CGP: The paper carefully distinguishes between the incompatibility of CGP optima and TO optima. While the certified example proves that the CGP optima are incompatible, it does not yet prove that the TO optima (which may be lower than CGP optima) are incompatible. The authors note that proving the latter requires showing that the individual TO optima coincide with the CGP optima or establishing a direct TO-specific contradiction.

Significance and Claims
The paper claims to provide the first rigorous demonstration that optimal coherence transfers, when considered separately, need not be compatible with a single physical thermodynamic process. The significance lies in:

Moving Beyond Scalar Bounds: It highlights that in degenerate systems, coherence is not merely a collection of independent scalar modes but a structured object where joint optimization is constrained by global unitarity and conservation laws.
Rigorous Certification: By using machine-certified witnesses, the authors provide a mathematically rigorous proof of incompatibility, moving beyond numerical approximations.
Open Questions: The authors modestly conclude that while they have established incompatibility for the broader class of CGP channels, a "TO-sharp" result (proving incompatibility specifically for the individual optima of thermal operations) remains an open problem. They suggest that the shell-amplitude representation offers a natural pathway to address this by combining local saturation conditions with global unitarity constraints.

The work does not propose new experimental protocols or applications but rather refines the theoretical understanding of the structural limitations of quantum thermodynamic processes, specifically regarding the trade-offs between optimizing multiple coherence transfers simultaneously.