粒子数不确定状态下量子纠缠与可分性的应用机制探究

Great question—this is a really interesting bridge between the discrete qubit entanglement we often learn first and the wilder world of continuous-variable quantum systems. Let’s break this down in plain terms.

First: A Quick Refresher on Discrete Particle-Number Entanglement #

When we talk about Alice and Bob’s two qubits, entanglement is straightforward to frame with tensor products:

• A state is separable if it can be written as $|\psi\rangle_A \otimes |\phi\rangle_B$ (a product of Alice’s qubit state and Bob’s qubit state).
• A state is entangled if it can’t be split this way (think Bell states like $\frac{1}{\sqrt{2}}(|01\rangle + |10\rangle)$).

This relies on having well-defined individual particles to assign states to—something we don’t have with particle-number-indefinite states like coherent oscillators.

Extending to Particle-Number-Indefinite States #

Here’s the key shift: we can’t use single-particle tensor products anymore, so we fall back on the general definition of separability for composite quantum systems:

A state $\rho$ of a composite system (say, two harmonic oscillators A and B) is separable if and only if it can be written as:
$$\rho = \sum_i p_i \rho_A^{(i)} \otimes \rho_B^{(i)}$$
Where:

• $p_i \geq 0$ and $\sum_i p_i = 1$ (these are classical probabilities for the mixture),
• $\rho_A^{(i)}$ and $\rho_B^{(i)}$ are valid density matrices for systems A and B individually.

If you can’t write the state this way? It’s entangled.

How Entanglement Operates in These States #

The core idea of entanglement—non-classical correlations between spatially separated systems—still holds, but it manifests in continuous observables (like position $x$ or momentum $p$) instead of discrete particle states. Here’s how it plays out:

• Correlation in Continuous Observables: Instead of correlating discrete qubit states, entanglement ties together continuous properties of the subsystems. Measure one subsystem’s position, and you instantly fix the statistical distribution of the other’s position—even if they’re light-years apart.
• Concrete Example: Two-Mode Squeezed Vacuum: This is a particle-number-indefinite state (a superposition of equal particle-number pairs: $\sum_{n=0}^\infty \frac{\tanh^n r}{\cosh r}|n,n\rangle$). It enforces $x_A = x_B$ and $p_A = -p_B$ (in appropriate units). This strict, non-classical correlation can’t be explained by classical statistics—it’s pure entanglement.
• Entangled Coherent States: Even single-mode coherent states (which are "classical-like" on their own) can form entangled states. For example, $\frac{1}{\sqrt{2}}(|\alpha,\beta\rangle + |-\alpha,-\beta\rangle)$ can’t be split into individual oscillator states, making it entangled—while still having no definite particle number.

Applying Separability Criteria #

To check if a particle-number-indefinite state is entangled, we use tools tailored for continuous-variable systems:

• Peres-Horodecki Criterion: For a composite density matrix $\rho$, compute its partial transpose (transpose the density matrix for just one subsystem, say A). If this partial transpose has negative eigenvalues, the state is entangled. This works for both pure and mixed states.
• Entropy Criterion for Pure States: For a pure composite state, entanglement is equivalent to the subsystem having non-zero von Neumann entropy. For the two-mode squeezed vacuum state, the entropy of either oscillator is positive—proof that it’s entangled.

Wrapping Up #

At the end of the day, entanglement in particle-number-indefinite states is just a generalization of the discrete qubit case. The key difference is we stop tying separability to individual particle states and use the broader definition of subsystem density matrix mixtures. The entanglement still shows up as non-classical correlations between observables—just in continuous degrees of freedom rather than discrete qubit states.

内容的提问来源于stack exchange，提问作者user184773