Bures Entanglement in Quantum Systems

• Bures entanglement is a quantum measure defined by the minimum Bures distance between a bipartite state and the set of separable states.
• It links geometric concepts such as Uhlmann fidelity and Riemannian metrics to provide explicit formulas, including a connection to concurrence for two-qubit systems.
• The measure is applicable across diverse areas including random-matrix theory, holography, and quantum algorithm design, highlighting its broad utility in quantum information research.

Bures entanglement is an entanglement measure defined from the Bures distance to the set of separable states of a bipartite quantum system. In the resource-theoretic formulation, it quantifies how far a state lies from separability in the geometry induced by Uhlmann fidelity; in random-state theory, closely related usage concerns typical entanglement properties of states distributed according to the Bures–Hall measure. Across these settings, the Bures construction is tied to a monotone Riemannian metric, quantum Fisher information, ambiguous state discrimination, exact two-qubit formulas, and analytically tractable random-matrix ensembles (Bromley et al., 2014, Wei, 2020).

1. Metric foundation and formal definition

The starting point is the Bures distance between density operators $\rho$ and $\sigma$,

$$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$

with Uhlmann fidelity

$$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$

Within the formulation used for bipartite correlations, $D_{Bu}$ is a bona fide metric: it is nonnegative, symmetric, satisfies the triangle inequality, and vanishes iff $\rho=\sigma$. It is also contractive under completely positive trace-preserving maps,

$$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$

and is locally Riemannian, with associated metric coinciding with the quantum Fisher information. These properties are the basis for monotonicity and geometric robustness of the associated correlation measures (Bromley et al., 2014).

For a bipartite system $AB$, let $\mathcal S$ denote the separable states,

$$\mathcal{S} = \left\{\sigma = \sum_i p_i\, \rho_i^A \otimes \rho_i^B \right\}.$$

The Bures entanglement of $\sigma$0 is the minimum Bures distance from $\sigma$1 to $\sigma$2,

$\sigma$3

Equivalently, if $\sigma$4 attains the infimum,

$\sigma$5

A second convention, common in distance-based entanglement literature and variational formulations, uses the squared Bures distance,

$\sigma$6

The two conventions differ only by whether the metric itself or its square is taken as the entanglement functional (Bromley et al., 2014, Gao et al., 2020).

Because the set of classical-quantum states is contained in $\sigma$7, Bures-discord-type quantum correlations always upper-bound Bures entanglement. This inclusion is one element of a broader geometric hierarchy in which total, classical, quantum, and entanglement correlations are all defined as distances to nested sets of states (Bromley et al., 2014).

2. Exact formulas for pure states, two qubits, and Bell-diagonal families

For pure bipartite states, Bures entanglement reduces to a simple function of the Schmidt spectrum. If

$\sigma$8

with $\sigma$9, then

$$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$0

in the squared-distance convention. In this regime, Bures entanglement, Bures geometric discord, and the Bures distance to the set of fully classical states coincide; if $$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$1 is nondegenerate, the closest separable state is the pure product state built from the Schmidt vectors associated with $$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$2 (Spehner et al., 2013).

For arbitrary two-qubit states, the entanglement measure admits a closed form in terms of concurrence $$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$3. The fidelity with the nearest separable state is

$$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$4

hence

$$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$5

In the squared-distance convention this same relation is often written as

$$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$6

so two-qubit Bures entanglement is an explicit monotone of concurrence (Bromley et al., 2014, Gao et al., 2020).

For two-qubit Bell-diagonal states,

$$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$7

the concurrence simplifies to

$$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$8

where $$D_{Bu}(\rho,\sigma)\equiv \sqrt{2\left(1-\sqrt{F(\rho,\sigma)}\right)},$$9 are the Bell-basis eigenvalues. This yields explicit Bures entanglement for the full Bell-diagonal class. Two subfamilies are especially prominent. For Werner states,

$$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$0

Bures entanglement vanishes exactly for $$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$1, becomes nonzero for $$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$2, and is non-analytic at $$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$3. For rank-2 Bell-diagonal states,

$$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$4

one finds

$$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$5

so Bures entanglement equals the Bures discord-type quantum correlation throughout this mixed-state family (Bromley et al., 2014).

3. Position within the hierarchy of quantum correlations

The Bures formalism is typically presented as a unified geometric hierarchy. Product states $$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$6, classical-quantum states $$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$7, and separable states $$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$8 define the targets for total, quantum, and entanglement correlations, respectively: $$F(\rho,\sigma)\equiv\left[\mathrm{Tr}\left(\sqrt{\sqrt{\rho}\,\sigma\,\sqrt{\rho}}\right)\right]^2.$$9 For any bipartite state,

$D_{Bu}$0

For Bell-diagonal families studied explicitly, total correlations are subadditive with respect to the sum of classical and quantum ones,

$D_{Bu}$1

with strict inequality in general (Bromley et al., 2014).

Bures geometric discord gives the closest parallel notion to Bures entanglement. It is defined as the Bures distance to the set of zero-discord, $D_{Bu}$2-classical states. For pure states it is identical to the geometric measure of entanglement, while for mixed states it coincides with the optimal success probability of an ambiguous quantum state discrimination task; moreover, the closest zero-discord states are obtained from the corresponding optimal measurements. This operational link supplies an information-theoretic interpretation for the same fidelity geometry that underlies Bures entanglement (Spehner et al., 2013).

Later work sharpened the relation between discord and entanglement by optimizing discord over state extensions. In the asymmetric formulation,

$D_{Bu}$3

and in the symmetric formulation,

$D_{Bu}$4

both minimized extension-based Bures discords are exactly equal to the Bures distance of entanglement,

$D_{Bu}$5

A related result states that the minimal Bures distance of discord over cross-symmetric extensions is equivalent to the Bures distance of entanglement. These equivalences place discord in a more primitive position conceptually: entanglement becomes the irreducible part of Bures discord that survives minimization over extensions (Xiong et al., 2020, Xiong et al., 2023).

Within dynamical models, the distinction between discord-type and entanglement-type Bures measures is sharp. For two independent qubits in local pure-dephasing channels, Bures quantum and classical correlations can exhibit freezing, whereas Bures entanglement never freezes, decays monotonically, and can show sudden death. This difference is one of the clearest operational manifestations of the hierarchy $D_{Bu}$6 (Bromley et al., 2014).

4. Multipartite structure and monogamy

Distance-based entanglement measures satisfying the usual invariance and monotonicity conditions are maximal on pure states under one-sided operations. In particular, if a mixed state $D_{Bu}$7 can be obtained from a pure state by a channel on one party, the corresponding distance-based entanglement cannot exceed that of the pure precursor. This general statement applies to the Bures measure of entanglement and frames it as a genuinely geometric pure-state benchmark (Gao et al., 2020).

For pure multiqubit states, the Bures measure satisfies CKW-type monogamy. Because for two-qubit states

$D_{Bu}$8

and because the function $D_{Bu}$9 obeys

$\rho=\sigma$0

one obtains for every pure $\rho=\sigma$1-qubit state $\rho=\sigma$2,

$\rho=\sigma$3

In particular,

$\rho=\sigma$4

The same work derives a class of tighter monogamy relations by weighting the pairwise terms according to their ordered magnitudes (Gao et al., 2020).

These results situate Bures entanglement among the few geometrically defined measures for which nontrivial multipartite sharing constraints are available in closed analytic form. They also reinforce the role of the concurrence-to-Bures map as the bridge from bipartite exact formulas to multiqubit inequalities (Gao et al., 2020).

5. Statistical Bures entanglement and the Bures–Hall ensemble

In random-matrix theory, “Bures entanglement” often refers to the typical entanglement of bipartite pure states whose reduced density matrices are distributed according to the Bures–Hall measure. For subsystem dimensions $\rho=\sigma$5, the joint eigenvalue density of the reduced state $\rho=\sigma$6 is

$\rho=\sigma$7

The additional factor $\rho=\sigma$8 distinguishes Bures–Hall from the Hilbert–Schmidt ensemble and drives much of the analytic difficulty (Wei, 2020).

Two exact finite-size formulas are central. The mean purity is

$\rho=\sigma$9

and the mean von Neumann entropy is

$$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$0

These formulas prove conjectures of Sarkar and Kumar and show that for large $$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$1 with fixed $$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$2, the average entropy approaches $$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$3 while the average purity approaches $$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$4, so Bures-typical bipartite states are highly entangled (Wei, 2020).

The fluctuation theory has also been worked out. The exact variance is

$$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$5

and later work derived an exact closed form for the third cumulant,

$$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$6

leading to a skewness-corrected approximation of the entropy distribution. The standardized skewness decays as $$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$7, supporting Gaussian asymptotics in the large-dimension regime (Wei, 2020, Wei et al., 7 Jun 2025).

A further development replaces large summation manipulations by a recurrence for real-valued spectral moments of the unconstrained Bures–Hall ensemble. This recurrence re-derives the mean entropy and purity and suggests a systematic route to higher-order entanglement statistics through Christoffel–Darboux formulas for the correlation kernels (Wei et al., 1 Feb 2026).

6. Algorithms, holography, and Bures-distance applications

A recent variational quantum algorithm, EvQA, uses Uhlmann’s theorem to estimate Bures entanglement of general quantum states. For a target state $$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$8, one fixes a purification $$D_{Bu}(\Lambda(\rho),\Lambda(\sigma)) \le D_{Bu}(\rho,\sigma),$$9, parametrizes purifications of separable states,

$AB$0

and minimizes

$AB$1

The overlap between purifications is obtained with a SWAP test, and the optimized ansatz reconstructs the closest free state as well as the Bures entanglement value. Reported applications include two-qubit Werner states, three-qubit dephased cluster states, and four-qubit noisy Smolin states, including bound-entangled regimes (Friedrich et al., 28 Jul 2025).

In holography, the Bures metric on reduced density matrices has been used as an entanglement-sensitive information metric rather than as a strict entanglement monotone. For single intervals in holographic $AB$2 CFTs, the Bures metric of locally excited reduced states reproduces the expected entanglement wedge: inside the wedge it is proportional to the AdS metric on a time slice, whereas outside the wedge it vanishes. Closely related work derives a bulk observable dual to the Bures metric, localized in the entanglement wedge, and compares the resulting Bures-metric complexity with subregion complexity in the complexity$AB$3volume paradigm (Suzuki et al., 2019, Gerbershagen et al., 2024).

A different caution appears in curved-spacetime applications. In the Dirac-field analysis around a charged Reissner–Nordström black hole, the authors use the Bures distance between Hawking-evolved reduced states and the initial entangled state as a geometric diagnostic. They explicitly note that this is not the distance to the nearest separable state and therefore is not a genuine entanglement monotone in the strict resource-theoretic sense. The same distinction applies in finite matter-field phase-diagram studies where Bures distance is used to compare neighboring ground states and characterize abrupt reorganizations of matter–field entanglement, rather than to minimize over separable states (Chhieb et al., 6 May 2026, Cordero et al., 2020).

The resulting terminological boundary is important. In its strict sense, Bures entanglement is the Bures distance to the separable set. In broader contemporary usage, Bures geometry also underlies diagnostic notions of entanglement, complexity, distinguishability, and typicality. The literature surveyed here supports both senses, but it also makes clear that only the distance-to-separable-state construction defines the entanglement monotone proper (Bromley et al., 2014, Xiong et al., 2020).