Trace-distance-based complementarity relations in a multipath interferometer

AI summary

Overview

Research area: Quantum foundations and quantum information theory — specifically wave–particle duality and complementarity relations in multi-path interferometers, quantified with the trace distance.

Technical level: Advanced. The paper assumes familiarity with density operators, trace norm, strictly incoherent operations, Kraus representations, Schur-convexity, Weyl's inequality, and quantum Pinsker inequality.

Scope (one sentence): The paper develops a general recipe for building wave and particle measures from a class of distance-like functions between quantum states, then specializes to the trace distance to derive normalized wave/particle measures and a family of duality and triality relations in a d-path interferometer.

What This Paper Is About

In a multi-path interferometer a quantum particle shows both wave behavior (interference) and particle behavior (which-path information), and these two aspects trade off against each other. Existing quantitative treatments often build wave measures out of coherence measures, but the authors point out that coherence and "waveness" are not the same thing — for instance the l₂ norm (and more generally the l_p norm) is a common wave measure yet is not monotonic under incoherent operations. The goal is to characterize wave and particle behavior using functions that need not be coherence measures, and then to establish the resulting complementarity (duality and triality) relations, with the trace distance as the central case study.

Key Contributions

1. A general construction principle. The authors define a class of functions D on pairs of quantum states satisfying nonnegativity, monotonicity under strictly incoherent operations, and joint convexity, and prove that W(ρ) = D(ρ, ρ_diag) is a valid wave measure and P(ρ) = D(ρ_diag, I/d) is a valid particle measure, verified against the four standard criteria (W1)–(W4) and (P1)–(P4).

2. Affirmative answer to an open question. The paper states that Ref. [22] asked whether a function satisfying only non-negativity, monotonicity, and convexity can characterize wave–particle duality; the authors answer this affirmatively, and note that monotonicity can be restricted to strictly incoherent operations rather than all incoherent operations, extending the reach of wave measures.

3. Normalized trace-distance wave and particle measures. Using two lemmas on the maximal trace norm of ρ − ρ_diag and of the deviation of a probability vector from uniform, the authors obtain W_tr(ρ) = d/(2(d−1))‖ρ − ρ_diag‖_tr and P_tr(ρ) = d/(2(d−1))‖ρ_diag − I/d‖_tr, both bounded in [0, 1].

4. Four dualities, three trialities, one generalized triality. The paper establishes a tight duality for qubit and qutrit systems, four upper-bound dualities for arbitrary dimension (based on rank, purity, linear entropy, and relative entropy), and complementarity relations involving mixedness — including a qubit equality and a general linear/quadratic triality framework.

Main Findings

• Trace distance qualifies as both a wave and a particle measure. Although the trace distance is generally not a coherence measure, it is contractive under completely positive and trace-preserving maps and satisfies all requirements for the function D, so it can quantify both waveness and particleness.

• Tight low-dimensional duality. For any qubit and qutrit state, W_tr(ρ)² + P_tr(ρ)² ≤ 1. For qubits the inequality becomes an equality if and only if the state is pure, matching the no-detector case of Ref. [3]. For qutrits equality holds if and only if ρ = |ψ⟩⟨ψ| with |ψ⟩ = (1/√3) Σᵢ e^{iθᵢ}|i⟩, or ρ = |i⟩⟨i| for i = 1, 2, 3.

• The quarter-circle bound fails beyond low dimensions. Using 2000 randomly generated density matrices, the authors find that at d = 6 the range of coordinates (P_tr, W_tr) exceeds the region P_tr² + W_tr² ≤ 1 (shown as yellow triangles in subfigure (c) of Fig. 2). The paper notes that Englert et al. had already discussed violation of the standard duality in Ref. [10], and attributes the effect to the shape of the boundary depending on the particular choice of wave and particle measures.

• Exact qubit triality. For a two-path interferometer the qubit relation can be rewritten as the complete wave–particle–mixedness equality W_tr(ρ)² + P_tr(ρ)² + 2(1 − Tr(ρ²)) = 1, where 2(1 − Tr(ρ²)) is the mixedness supplying the missing part of the qubit duality.

• Four high-dimensional upper bounds, all for W_tr(ρ)² + P_tr(ρ)²:

  • rd²/(4(d−1)²) · (Tr(ρ²) − 1/d), with r = max{r₁, r₂}, r₁ = r(ρ − ρ_diag), r₂ = r(ρ_diag − I/d) (Theorem 6, rank-based).
  • R′d²/(d−1)² · (Tr(ρ²) − 1/d), with R′ = max{R₁, R₂}, R₁ = r(ρ)r(ρ_diag)/(r(ρ) + r(ρ_diag)) and R₂ = r(ρ_diag)d/(r(ρ_diag) + d) (Theorem 7).
  • d²/(d−1)² · [2 − 1/d − Tr(ρ_diag²)] (Theorem 8, linear-entropy based).
  • (ln 2)d²/(2(d−1)²) · (log d − S(ρ)) (Theorem 9, relative-entropy based).

• Bound behavior is dimension- and mixing-dependent. For the family ρ = (1 − ε)|0⟩⟨0| + εσ with 50 random density matrices sampled per dimension for d = 2, 3, 4, 5, 6, 7: at d = 2 the Theorem 6 bound coincides with the exact value; at d = 3 Theorem 6 is best for ε < 0.098236 and ε > 0.707827 while Theorem 9 is best for 0.098236 ≤ ε ≤ 0.707827, and Theorem 8 beats Theorem 7 for 0 < ε ≤ 0.104738; at d = 4 Theorem 9 is best for 0 < ε ≤ 0.943534 and Theorem 6 for ε ≥ 0.943534, with Theorem 8 beating Theorem 7 for 0 < ε ≤ 0.186393; at d = 5 Theorem 9 is best on (0, 0.497688], Theorem 6 on (0.497688, 1], and Theorem 8 beats Theorem 6 on [0, 0.052661) and Theorem 7 on [0, 0.228316); at d = 6 Theorem 8 becomes tightest on (0, 0.014759). The general conclusion is that the optimal bound should be chosen adaptively based on dimension and mixing degree.

• Operational meaning of the wave measure. By the Holevo–Helstrom theorem, T(ρ, ρ_diag) upper bounds the advantage of any strategy trying to discriminate ρ from ρ_diag in a one-shot scenario; ρ_diag is prepared by making all paths perfectly distinguishable via the maximally decohering map Δ. The coherence-measurement method of Ref. [11] is a special case that only compares maximum intensities from ρ and ρ_diag.

• Triality framework. Defining M(ρ) = C − D(ρ, I/d) with C = max_ρ D(ρ, I/d) as a mixedness quantifier, the authors obtain the linear relation W(ρ) + P(ρ) + M(ρ) ≥ C (when D also satisfies the triangle inequality) and the quadratic relation W²(ρ) + P²(ρ) ≥ ½(C − M(ρ))². The supplied text is truncated within this section, so the explicit three triality relations and the single generalized triality announced in the introduction are not fully reproduced here.

Methodology in Plain English

The authors model a d-path interferometer by a d-dimensional density matrix whose diagonal entries are the probabilities of the particle taking each path and whose off-diagonal entries encode the superposition responsible for interference. The diagonal part — what remains if all path coherence is destroyed — is written ρ_diag.

Their strategy is axiomatic: instead of picking a specific coherence measure, they list three mild properties a state-to-state function D should have (it vanishes only on identical states, it does not increase under operations that cannot create path coherence, and it is jointly convex). Any such D immediately yields a wave measure by comparing ρ with ρ_diag, and a particle measure by comparing ρ_diag with the maximally mixed state I/d, because those comparisons capture, respectively, "how far the state is from being classical" and "how far the path distribution is from uniform."

They then adopt the trace distance as the concrete choice of D and normalize both measures so they lie between 0 and 1, using lemmas that pin down the largest possible values. Duality relations are derived by rewriting the trace norms in terms of the Hilbert–Schmidt norm, rank, linear entropy, or relative entropy, each comparison producing a different upper bound. For qubits the algebra collapses to an exact equality that includes a mixedness term; for qutrits and higher dimensions the bounds are inequalities. Finally, numerical sampling of random states (2000 states for the dimension scan in Fig. 2, and 50 states per dimension for the ε-scan in Fig. 3) is used to check where each bound is tight and where the naive quarter-circle picture breaks down.

Why This Matters

The work reframes wave–particle duality away from coherence measures, showing that a broader class of geometric distances — including the trace distance — can serve as legitimate quantifiers of waveness and particleness. It also answers an open question from prior literature about how weak the axioms for such quantifiers can be, and it connects the interferometric picture to the trace distance's well-established operational interpretation in state discrimination.

Because the trace distance is already a workhorse in quantum information, expressing duality in its terms may link wave–particle complementarity to other physical scenarios. The paper explicitly lists four application areas for the trace distance:

• Quantum cryptographic protocols — security analyses that bound an adversary's distinguishing advantage.
• Quantum complexity theory — distance-based measures used in complexity arguments.
• Quantum state discrimination — via the Holevo–Helstrom theorem, giving the wave measure a direct operational reading.
• Entanglement measures — where trace-distance-based quantities already appear.

Industry relevance: The results are foundational rather than device-oriented, but they touch on technologies built on interferometry and on distinguishing quantum states — quantum key distribution and secure communication, benchmarking and verification of quantum processors, and precision measurement with multi-path interferometers. The explicit dimension-dependent bounds (d from 2 to 7 in the numerics) matter for practitioners who need quantitative statements in systems larger than a qubit.

Future Directions

• Complete and generalize the triality relations. The paper announces three wave–particle–mixedness trialities and one generalized triality; extending these beyond the qubit equality W² + P² + 2(1 − Tr(ρ²)) = 1 to arbitrary d with tight, closed-form conditions remains the natural next step.
• Characterize the exact feasible region of (P_tr, W_tr). The random sampling at d = 6 shows states outside the quarter-circle, but the true boundary for general dimension is not derived analytically; identifying it, and how it depends on the choice of measures, is an open question the authors themselves flag.
• Replace incoherent operations with strictly incoherent operations more broadly. Since the authors show monotonicity can be confined to strictly incoherent operations, it is natural to ask whether other candidate distance functions, or other operational classes, yield still more general wave measures.
• Apply the framework in other distance-based settings. Given the listed roles of trace distance in cryptography, complexity theory, state discrimination, and entanglement, an obvious direction is to see whether the duality/triality structure carries over to those contexts and provides useful constraints.

Target Audience

Researchers in quantum foundations and quantum information theory working on wave–particle duality, coherence theory, and complementarity relations; graduate students comfortable with density-matrix formalism and distance measures on state space; and mathematically inclined readers interested in Schur-convexity, majorization, and norm inequalities applied to quantum states. Readers seeking experimental protocols or device implementations will find little here, since the paper is theoretical and numerical.

Authors’ abstract

The complementarity relations in an interferometer reflect an important phenomenon in quantum mechanics: wave-particle duality. Here, we develop a method to quantify both wave and particle behaviors in a multi-path interferometer. In particular, we find that the trace distance is a good candidate for wave and particle measures. As a result, some duality relations and triality ralations are established respectively. This work not only extends the application of the trace distance to the interferometer, but also opens up new perspectives on the quantification of waveness and particleness.