[Paper Review] Quantum entanglement and geometry of determinantal varieties

Hao Chen|ArXiv.org|Oct 17, 2001

Quantum Information and Cryptography13 references4 citations

TL;DR

This paper introduces determinantal varieties in complex projective spaces as algebraic invariants for mixed quantum states under local unitary transformations, linking quantum entanglement to algebraic geometry. It establishes a geometric separability criterion: separable states correspond to determinantal varieties that are sums of linear subspaces, enabling new criteria for entanglement, Schmidt number bounds, Hamiltonian simulation, and systematic construction of PPT bound entangled states.

ABSTRACT

Quantum entanglement was first recognized as a feature of quantum mechanics in the famous paper of Einstein, Podolsky and Rosen [18]. Recently it has been realized that quantum entanglement is a key ingredient in quantum computation, quantum communication and quantum cryptography ([16],[17],[6]). In this paper, we introduce algebraic sets, which are determinantal varieties in the complex projective spaces or the products of complex projective spaces, for the mixed states in bipartite or multipartite quantum systems as their invariants under local unitary transformations. These invariants are naturally arised from the physical consideration of measuring mixed states by separable pure states. In this way algebraic geometry and complex differential geometry of these algebraic sets turn to be powerful tools for the understanding of quantum enatanglement. Our construction has applications in the following important topics in quantum information theory: 1) separability criterion, it is proved the algebraic sets have to be the sum of the linear subspaces if the mixed states are separable; 2) lower bound of Schmidt numbers, that is, generic low rank bipartite mixed states are entangled in many degrees of freedom; 3) simulation of Hamiltonians, it is proved the simulation of semi-positive Hamiltonians of the same rank implies the projective isomorphisms of the corresponding algebraic sets; 4) construction of bound enatanglement, examples of the entangled mixed states which are invariant under partial transpositions (thus PPT bound entanglement) are constructed systematically from our new separability criterion. On the other hand many examples of entangled mixed states with rich algebraic-geometric structure in their associated determinantal varieties are constructed and studied from this point of view.

Motivation & Objective

• To establish a geometric framework for analyzing quantum entanglement using algebraic invariants derived from mixed states.
• To develop a separability criterion based on the algebraic structure of determinantal varieties associated with mixed states.
• To apply algebraic geometry to solve key problems in quantum information theory, including Schmidt number bounds and Hamiltonian simulation.
• To systematically construct PPT bound entangled states using the new geometric separability criterion.
• To demonstrate that generic low-rank mixed states in bipartite systems are entangled by analyzing the non-linear structure of their associated determinantal varieties.

Proposed method

• Define determinantal varieties as the zero loci of minors of matrices constructed from the density matrix under local basis choices.
• Use the rank condition $\text{rank}(\sum r_i A_i) \leq n-1$ to define the determinantal variety associated with a mixed state $\rho$.
• Apply complex algebraic geometry and differential geometry to analyze the structure of these varieties in $\mathbb{C}^m$ or $\mathbb{C}^m \times \mathbb{C}^n$.
• Establish that separable mixed states correspond to determinantal varieties that are unions of affine linear subspaces.
• Use projective isomorphisms of these varieties to characterize equivalence under Hamiltonian simulation.
• Construct explicit examples of PPT entangled states by ensuring the associated determinantal variety is non-linear, even when the state is invariant under partial transposition.

Experimental results

Research questions

• RQ1 Can the separability of a mixed quantum state be characterized by the algebraic structure of its associated determinantal variety?
• RQ2 What geometric conditions on determinantal varieties imply that a mixed state is entangled or bound entangled?
• RQ3 How does the Schmidt number of a mixed state relate to the dimension and irreducibility of its determinantal variety?
• RQ4 Can the simulation of semi-positive Hamiltonians be characterized via projective isomorphisms of determinantal varieties?
• RQ5 What conditions ensure that a mixed state invariant under partial transposition is entangled (i.e., bound entangled)?

Key findings

• Separable mixed states in bipartite systems correspond to determinantal varieties that are unions of affine linear subspaces.
• Generic low-rank bipartite mixed states are entangled because their associated determinantal varieties are not unions of linear subspaces.
• The simulation of semi-positive Hamiltonians of the same rank implies projective isomorphism of the corresponding determinantal varieties.
• A new systematic method is developed to construct PPT bound entangled mixed states by ensuring the determinantal variety is non-linear while the state is invariant under partial transposition.
• The state $\rho_{0,0,1}$ is the first example of a PPT entangled mixed state with separable pure states in its range, constructed via this geometric criterion.
• For generic parameters $e_1, e_2, e_3$, the determinantal variety $V_A^5(\rho_{e_1,e_2,e_3}) \cap C^3$ contains a curve, not a line, proving entanglement via non-linear structure.