Certifying ergotropy under partial information

This paper introduces a general framework for certifying lower bounds on quantum ergotropy using only partial information from a limited set of observables, providing robust, confidence-certified estimates that account for statistical noise and demonstrating its effectiveness through both synthetic and experimental data on an IBM quantum processor.

The Gist

Imagine you have a quantum battery. This isn't a battery you plug into a wall; it's a tiny, invisible machine made of atoms that stores energy. The big question in quantum physics is: How much work can we actually get out of it?

In the perfect world of theory, if you know the battery's exact internal blueprint (its "state") and the rules of the game (its "Hamiltonian"), you can calculate exactly how much energy you can extract. This maximum extractable energy is called Ergotropy.

But here's the real-world problem: We rarely know the full blueprint.

In a real lab, measuring a quantum system is like trying to guess the contents of a sealed, complex box by only peeking through a few tiny holes. You get some clues (data), but you don't have the whole picture. If you try to calculate the energy using your partial clues, you might get it wrong, or worse, you might think there's no energy left when there actually is.

This paper introduces a clever safety net to solve this problem. Here is how it works, broken down into simple concepts:

1. The "Worst-Case" Guarantee

Instead of guessing the exact amount of energy (which is impossible with partial data), the authors ask a different question: "What is the minimum amount of energy we can guarantee is there, based on the clues we have?"

Think of it like a treasure hunt. You don't know exactly where the gold is buried, but you have a map with a few landmarks. Instead of saying, "I bet the gold is right here," you say, "Based on these landmarks, I can guarantee that there is at least this much gold hidden in this general area."

If your calculation says "at least 5 joules," you know for a fact that the battery has 5 joules or more. It might have 100, but you are 100% sure it has at least 5. This is a certified lower bound.

2. The Two-Step "Detective" Protocol

The authors created a mathematical recipe (using a tool called Semidefinite Programming, which is like a super-smart calculator for optimization) to find this guarantee. They break it down into two steps:

• Step 1: Pick a "Test Drive" Strategy.
  Imagine you have a list of possible ways to rearrange the atoms in the battery to squeeze out energy. Since you don't know the exact state, you pick a "best guess" state that fits your current clues. You then figure out the perfect way to rearrange that specific guess to get energy out. Let's call this your "Extraction Plan."

• Step 2: The "Devil's Advocate" Check.
  Now, you keep that "Extraction Plan" fixed. You ask the computer: "Okay, given my clues, what is the worst possible version of the battery that still fits the clues? If I use my Extraction Plan on this worst-case battery, how much energy do I get?"

  The answer to this question is your Certified Lower Bound. Even if the battery is the "worst-case" version compatible with your data, you are guaranteed to get at least this much energy.

3. Dealing with "Static" (Noise)

In the real world, measurements aren't perfect. If you flip a coin 10 times, you might get 7 heads. If you flip it 1,000 times, you'll get closer to 500. This is called shot noise or finite statistics.

The authors' method is robust against this noise. They add a "fuzziness" factor to their calculation.

• Analogy: Imagine you are trying to fit a square peg into a round hole. If your measurements are shaky, you don't just draw a perfect circle; you draw a slightly larger, fuzzy circle to account for the shaking.
• The math ensures that even with this fuzziness, your guarantee holds true with a very high probability (e.g., 99.7% confidence). It tells you: "I'm not 100% sure, but I'm almost certain there is at least this much energy."

4. Why This Matters

This is a game-changer for Quantum Batteries and Quantum Thermal Machines.

• Before: If you couldn't measure the whole system perfectly, you couldn't prove your battery was working. You were flying blind.
• Now: You can run an experiment, measure just a few things (like checking a few specific properties of the atoms), and immediately say, "Yes, this battery is charged and ready to work, and here is the minimum power it can deliver."

The Real-World Test

The authors didn't just do this on paper. They tested their method on:

1. Fake data: Simulated scenarios to see if the math held up.
2. Real data: They used a real quantum computer from IBM (a physical machine in a lab) to measure a 4-qubit system. Even with the machine's natural errors and limited measurements, their method successfully certified that the system contained extractable work.

Summary

Think of this paper as a quality control inspector for quantum energy. You don't need to take the battery apart to know it works. You just need to check a few key features, run them through this special "safety calculator," and it will give you a guaranteed minimum of how much power you can pull out, even if the data is noisy and incomplete. It turns a vague guess into a solid, certified promise.

Technical Summary

1. Problem Statement

Ergotropy is defined as the maximum work extractable from a quantum system via unitary operations. It is a fundamental resource in quantum thermodynamics, particularly for quantum batteries and thermal machines.

• The Challenge: Calculating ergotropy analytically requires full knowledge of the system's quantum state ($\rho$) and its Hamiltonian ($H$). However, in realistic experimental settings (e.g., many-body systems), full state tomography is infeasible due to exponential scaling, noise, and finite sampling.
• The Gap: Existing methods for estimating ergotropy under incomplete information are either limited to specific cases (e.g., only measuring the Hamiltonian) or lack rigorous certification. There is a need for a general framework that provides a certified lower bound on ergotropy using only a limited set of expectation values $\{ \langle O_i \rangle \}$, while accounting for statistical noise (shot noise) and finite data.

2. Methodology

The authors propose a two-step semidefinite programming (SDP) protocol to certify a lower bound on ergotropy ($E_{LB}$) given partial information.

A. Theoretical Framework

• Feasible Set ($\Omega_I$): Given a known Hamiltonian $H$ and a set of measured observables $\{O_i\}$ with expectation values $\{o_i\}$, the set of compatible states is defined as:
  $$\Omega_I = \{ X : X \succeq 0, \text{tr}(X)=1, \text{tr}(X O_i) = o_i \quad \forall i \in I \}$$
  This set is convex and guaranteed to contain the true state $\rho$.
• The Optimization Problem: The goal is to find the minimum ergotropy compatible with the data:
  $$\bar{E}_{LB} = \min_{X \in \Omega_I} \max_{U \in U(d)} \left[ \text{tr}(HX) - \text{tr}(H U X U^\dagger) \right]$$
  Since the maximization over unitaries makes the problem non-convex, the authors relax it into a tractable two-step procedure.

B. The Two-Step Protocol

1. Step (i): State Selection and Unitary Construction
   • Solve an SDP to select a specific state $\tilde{\rho} \in \Omega_I$. The authors suggest minimizing state purity ($\text{tr}(X^2)$) or maximizing linear entropy to find a "worst-case" (most disordered) state within the feasible set.
   • Compute the spectral decomposition of $\tilde{\rho} = \sum \tilde{r}_j | \tilde{r}_j \rangle \langle \tilde{r}_j |$.
   • Construct the optimal unitary $U^*$ for this specific state: $U^* = \sum_j |E_j\rangle \langle \tilde{r}_j|$, where $|E_j\rangle$ are the energy eigenstates of $H$.
2. Step (ii): Lower Bound Certification
   • Fix the unitary $U^*$ obtained in Step (i).
   • Solve a second SDP to minimize the work extraction over the entire feasible set $\Omega_I$ using this fixed unitary:
     $$E_{LB} = \min_{X \in \Omega_I} \left[ \text{tr}(HX) - \text{tr}(H U^* X U^{*\dagger}) \right]$$
   • Result: Since $U^*$ is a valid unitary, the result $E_{LB}$ is a rigorous lower bound for the true ergotropy ($E_{LB} \leq E(\rho, H)$).

C. Handling Finite Statistics

To address experimental shot noise, the protocol incorporates confidence regions.

• Instead of exact constraints $\text{tr}(X O_i) = o_i$, the feasible set is expanded to include states where the estimated expectation values deviate within a confidence interval $\epsilon_i$:
  $$|\text{tr}(X O_i) - o_i^{(est)}| \leq \epsilon_i$$
• The deviation $\epsilon_i$ is derived using Hoeffding's inequality and the union bound, ensuring the true state lies within $\Omega_I$ with a probability of at least $1-\delta$ (where $\delta$ is the failure probability).
• This yields probabilistic bounds that hold with a prescribed confidence level.

D. Monotonicity Enforcement

To ensure the bound improves (or stays the same) as more observables are added, the protocol includes a conditional update rule: the unitary from Step (i) is only updated if the new bound is strictly tighter than the previous iteration; otherwise, the previous unitary is retained.

3. Key Contributions

1. General Certification Framework: A method to lower-bound ergotropy using any arbitrary set of measured observables, without requiring full state tomography or assumptions about the state's structure (e.g., low rank).
2. Finite-Statistics Robustness: The integration of statistical confidence intervals directly into the SDP constraints, providing bounds that are valid with high probability in realistic experimental regimes.
3. Computational Efficiency: The use of Semidefinite Programming ensures the problem is solvable in polynomial time, making it scalable for moderate-sized quantum systems.
4. Experimental Validation: The method is benchmarked on both synthetic data and real experimental data from an IBM Quantum Processor (ibm_perth).

4. Results

The authors validated the framework using 4- and 5-qubit systems with various Hamiltonians (XXZ, ANNNI, Mixed Field Ising) and true states (GHZ, W, Product, Negative-temperature Gibbs).

• Synthetic Data (Fig. 3):
  • As the number of measured Pauli strings ($K$) increases, the certified lower bound ($E_{LB}$) monotonically tightens and converges to the true ergotropy.
  • For pure states (GHZ, W), convergence is rapid. For full-rank states (negative temperature), convergence is smoother but still effective with few measurements.
• Finite Statistics (Fig. 4a):
  • Simulated shot noise (binomial statistics) was introduced. The protocol successfully produced non-trivial lower bounds even with low shot counts ($10^4$ shots), demonstrating robustness against noise.
  • As the number of shots increased, the bounds concentrated around the ideal-statistics curve.
• Experimental Data (Fig. 4b):
  • Applied to a 4-qubit GHZ state prepared on an IBM quantum processor.
  • Using only 60 Pauli constraints (out of 256 possible) and $2^{14}$ shots per observable, the protocol certified a substantial fraction of the true ergotropy.
  • The bound remained monotonic and tightened steadily despite device imperfections and noise.

5. Significance and Implications

• Practical Quantum Thermodynamics: This work provides the first experimentally accessible tool to certify that a quantum system contains extractable work without needing to fully characterize the state. This is crucial for verifying the performance of quantum batteries and thermal machines.
• Resource Efficiency: It drastically reduces the measurement overhead compared to full tomography, making it feasible for current noisy intermediate-scale quantum (NISQ) devices.
• Theoretical Insight: The framework highlights that non-commuting observables are essential for certifying coherent ergotropy (work extractable from quantum coherence), distinguishing it from incoherent (population-based) work.
• Future Directions: The authors suggest extending the method to learn the Hamiltonian simultaneously (relaxing the assumption of perfect $H$ knowledge) and applying it to certify local ergotropy in scalable spin-chain architectures (e.g., Rydberg atoms, trapped ions).

In summary, the paper bridges the gap between theoretical quantum thermodynamics and experimental reality by providing a rigorous, noise-resilient, and computationally efficient method to certify work extraction capabilities in partially known quantum systems.