🔬 physics

Coherence-to-Entanglement Conversion under Faulty CNOT Gates and Matched Idle Noise

This paper provides closed-form analytic benchmarks quantifying how four distinct classes of CNOT gate faults and matched idle noise degrade the conversion of single-qubit coherence into two-qubit entanglement, revealing that while coherent errors are mild and control-phase errors are invisible to this metric, stochastic faults cause proportional degradation with the bit-flip class exhibiting a unique angle-dependent signature.

Original authors: Asad Ali

Published 2026-09-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Asad Ali

Original paper licensed under CC BY 4.0 (https://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

In the quiet, ultra-cold laboratories where quantum computers are built, two fragile properties hold the key to their power: coherence and entanglement. Coherence is the ability of a single quantum bit, or qubit, to exist in a superposition of states, effectively being in two places at once. Entanglement is the even stranger phenomenon where two qubits become so deeply linked that the state of one instantly defines the state of the other, no matter the distance between them. Turning the first into the second is one of the most basic operations a quantum processor must perform. It is the act of taking a single qubit's delicate superposition and using a logic gate to weave it into a two-qubit bond. In a perfect world, this conversion is flawless and immediate. But in the real world, the hardware is imperfect, and the environment is noisy. The moment a gate acts, it can introduce errors, and the moment the entangled pair is created, it begins to decay as it waits to be used. Understanding exactly how these imperfections destroy the connection is vital for building machines that can actually work.

A researcher at Hamad Bin Khalifa University has mapped out this destruction with surgical precision, creating a clear, mathematical guide for how four specific types of gate errors interact with the environment to kill entanglement. The study focuses on a standard operation where a control qubit flips a target qubit, a process known as a controlled-NOT gate. The researcher asked a simple but difficult question: if the gate is slightly broken, and the resulting entangled pair sits idle in a warm, noisy environment for a short time, how much of the connection survives? By treating the gate faults and the environmental noise as a matched pair, the study derives exact formulas that predict the fate of the entanglement for four distinct failure modes. These modes include the gate rotating too far or not far enough, a phase error shifting the timing of the flip, a random flip of the target bit, and a random shift in the phase of the target.

The findings reveal that not all errors are created equal. Some mistakes are surprisingly gentle. If the gate is miscalibrated so that it rotates the qubit slightly too much or too little, the damage to the entanglement is minimal, growing only with the square of the error. This means that small calibration slips are almost harmless. Similarly, if the gate introduces a coherent phase shift, the entanglement measure used in this study remains completely unchanged, making this specific type of error invisible to this particular diagnostic tool. However, random, stochastic errors are far more destructive. When the gate randomly flips the target bit or randomly shifts its phase, the entanglement degrades in direct proportion to the probability of the error occurring. The study shows that a random bit-flip error leaves a unique signature: the amount of damage depends heavily on the angle of the input qubit, becoming more severe as the input moves away from the most balanced state. This angle-dependent slope allows researchers to distinguish a bit-flip error from a phase-flip error, something that would be impossible if they only looked at a single, standard input.

Perhaps the most striking result concerns the survival of the entangled pair after it leaves the gate. The study confirms that in any environment with a finite temperature, entanglement does not just fade away slowly; it can vanish abruptly at a specific moment, a phenomenon known as entanglement sudden death. The researcher calculates the exact boundary where this happens, showing that even a perfect gate cannot save the entanglement if the qubits sit idle for too long in a warm environment. The heat of the environment eventually forces the qubits into a state where they are no longer linked. The study also identifies the optimal way to prepare the input qubit to survive this noise. Counterintuitively, the best starting point is not the most balanced superposition, but a state biased slightly toward the ground state. This bias helps the system resist the specific way the environment drains the connection, trading a small loss in initial potential for a much longer survival time.

The paper does not stop at theory; it validates every formula against independent computer simulations, confirming that the mathematical predictions match the behavior of the simulated quantum systems to the limits of numerical precision. It also proposes a practical diagnostic protocol. By measuring the entanglement at several different input angles, an experimentalist can infer which type of fault is dominating their machine. The study shows that this method is highly reliable for identifying strong random errors but struggles to distinguish weak calibration errors from a perfect gate, because those small errors are so subtle they disappear into the noise. The research also highlights a limitation: the method cannot detect coherent phase errors at all, a fact the author presents plainly as a constraint of the tool rather than a flaw. Ultimately, this work provides a transparent, analytical benchmark for a specific task, offering a clear window into how specific errors degrade specific resources, without the need for complex, general-purpose characterization tools. It serves as a precise map for navigating the treacherous transition from a single qubit's potential to a pair's shared reality.

Technical Summary: Coherence-to-Entanglement Conversion under Faulty CNOT Gates and Matched Idle Noise

Problem Statement
Quantum coherence and entanglement are fundamental resources in quantum information processing. A minimal protocol for converting single-qubit coherence into two-qubit entanglement involves applying a coherent superposition $|\psi_A\rangle = \cos\theta|0\rangle + e^{i\phi}\sin\theta|1\rangle$ to the control qubit of a CNOT gate, with the target prepared in $|0\rangle$ . Ideally, this produces a maximally entangled state with negativity $N_0 = \sin(2\theta)/2$ , exactly half the input $\ell_1$ -coherence. However, laboratory implementations deviate from this ideal due to two primary factors: (1) the CNOT gate itself is faulty (due to calibration errors, residual couplings, or stochastic noise), and (2) the generated entangled state is not consumed instantly but "idles" in a finite-temperature environment characterized by relaxation, thermal excitation, and dephasing. This post-gate evolution can lead to Entanglement Sudden Death (ESD), where entanglement vanishes at a finite time.

Existing characterization methods like randomized benchmarking target average gate fidelity or full error structures but do not directly address how specific fault classes degrade this specific resource-conversion task. This paper aims to provide a transparent, analytically tractable benchmark for this specific protocol, analyzing how four representative fault classes interact with a matched idle environment.

Methodology
The author develops a "matched" noise model where the gate faults and the post-gate idle noise are parametrized by a common set of device parameters (e.g., $T_1$ , $T_2$ , equilibrium populations $p_{eq}$ ). The model assumes:

Two qubits (control A, target B) with the target initially in $|0\rangle$ .
The gate is an ideal CNOT composed with exactly one phenomenological fault class at a time.
Idle noise is local, uncorrelated, Markovian, and phase-covariant.
The output state belongs to the "X-state" class (non-zero elements only on the diagonal and anti-diagonal), which allows for exact calculation of negativity.

The study treats four fault classes:

Coherent Controlled- $R_x$ Calibration Error: Under- or over-rotation of the conditional flip.
Coherent Controlled-Phase Error: Residual $ZZ$ coupling or AC-Stark shifts.
Stochastic Phase-Flip ( $Z$ ) Fault: Dephasing during the gate.
Stochastic Bit-Flip ( $X$ ) Fault: Bit-flip-like processes on the target.

A "commutation argument" is used to determine when intra-gate noise processes reduce exactly to these four classes. For instance, target bit-flip and control dephasing commute with the gate generator and reduce exactly; target dephasing reduces only to first order; and intra-gate relaxation does not reduce to a single Pauli class and is therefore excluded from the gate-fault set, assigned instead to the idle channel.

Key Contributions and Results

Exact Gate-Level Benchmarks:
The author derives closed-form expressions for the conversion efficiency $\chi = N_{gate}/N_0$ for each fault class:
- Coherent $R_x$ errors: The efficiency is $\chi_{Rx} = |\sin(\lambda\pi/2)|$ . Near the ideal point, the error is "quadratically harmless" (first-order sensitivity vanishes), meaning small calibration errors have minimal impact.
- Coherent Phase Errors: These are invisible to the negativity benchmark ( $\chi_\delta = 1$ ) because the measure depends only on the magnitude of coherence, not its phase. This is a structural limitation of the protocol, not a strength.
- Stochastic $Z$ Errors: The efficiency is $\chi_Z = |1 - 2q_Z|$ , degrading linearly with fault probability $q_Z$ .
- Stochastic $X$ Errors: The efficiency follows a complex two-block formula. Crucially, the $X$ -fault class exhibits a unique angle-dependent signature: $\chi_X \approx 1 - q_X(1 + \csc 2\theta)$ . The slope steepens near separable endpoints ( $\theta \to 0, \pi/2$ ), distinguishing it from $Z$ -faults which scale uniformly.
Post-Idle Survival and Sudden Death:
The paper derives a "master formula" (Eq. 29) for the final negativity after the matched idle channel.
- Sudden Death Boundaries: Closed-form boundaries for Entanglement Sudden Death (ESD) are derived for each class. For $Z$ -faults, the boundary is an explicit function of $q_Z$ and $\theta$ (Eq. 32). For $X$ -faults, it is the root of a quadratic equation (Eq. 33).
- Finite-Temperature ESD: The analysis confirms that at any finite temperature, entanglement sudden death is unavoidable for all input angles and gate efficiencies, recovering the known Yu–Eberly threshold as a zero-temperature limit.
- Optimal Input Angles: The Bell angle ( $\theta = \pi/4$ ) is not optimal under noise. The author derives a noise-adapted optimal angle $\theta_{opt}$ that shifts toward the ground state ( $\theta < \pi/4$ ) to suppress population leakage caused by thermal excitation and relaxation. The shift is inversely proportional to the surviving coherence.
Diagnostic Protocol and Robustness:
A diagnostic protocol is proposed to infer the dominant fault class from angle-resolved negativity measurements using nonlinear least squares.
- Identifiability: Phase faults are unidentifiable (indistinguishable from ideal). $Z$ and $X$ faults are distinguishable by their angle-dependent slopes ( $\csc 2\theta$ for $X$ ). Coherent $R_x$ and stochastic $Z$ faults are degenerate at the gate level but distinguishable at second order in the post-gate evolution.
- Robustness: The protocol is robust to modest SPAM-like contrast loss and idle parameter miscalibration. However, it cannot distinguish fault mixtures (which are absorbed into a single effective class) and fails to detect weak coherent errors if they fall below the noise floor, as their "quadratically harmless" nature makes them statistically invisible.

Significance and Claims
The paper positions itself not as a general device characterization tool (like randomized benchmarking) but as a transparent analytic benchmark for a specific resource-conversion task. Its significance lies in:

Providing exact, closed-form expressions for conversion efficiency, survival conditions, and sudden-death boundaries, verified against double-precision numerical simulations.
Establishing distinct "fingerprints" for different fault classes, particularly the angle-dependent slope of bit-flip errors versus the uniform scaling of phase-flip errors.
Highlighting a structural limitation: coherent phase errors are completely invisible to negativity-based benchmarks, a fact the author states plainly rather than presenting as a feature.
Offering a minimal model where the pairing of gate faults and idle noise is physically justified via commutation criteria, allowing for precise quantification of where the model holds and where it requires higher-order corrections.

The results are intended to serve as a baseline for understanding how specific noise mechanisms degrade entanglement generation and storage, with the caveat that the diagnostic power is limited by the specific assumptions of the model (local Markovian noise, X-state outputs, single dominant fault).

Technical Summary: Coherence-to-Entanglement Conversion under Faulty CNOT Gates and Matched Idle Noise

More like this