Polynomial Invariants for Mixed States
- Polynomial invariants for mixed states are functions of density operator elements that remain unchanged under local unitary and SLOCC transformations.
- They are constructed using trace monomials and index contractions, with explicit degree bounds and algebraic structures derived from invariant theory.
- These invariants facilitate entanglement classification and practical computation in two-qubit, multi-qudit, and qubit–qutrit systems.
Polynomial invariants for mixed states are polynomial functions of the matrix elements of quantum density operators that remain unchanged under specified classes of local transformations. They form a fundamental tool in the classification, comparison, and quantification of multipartite quantum correlations and entanglement in mixed-state quantum systems. The invariants are particularly critical for the study of local unitary (LU) equivalence and, in a broader framework, stochastic local operations and classical communication (SLOCC) equivalence for both pure and mixed states. This article provides a comprehensive exposition of the construction, algebraic structure, completeness, and key results related to polynomial invariants for mixed quantum states, drawing on algebraic geometry, invariant theory, and representation theory.
1. Invariant Theory for Mixed State Quantum Systems
Let be the Hilbert space of an -partite quantum system, with each . A mixed state is described by a density operator , i.e., a positive semidefinite Hermitian matrix of trace one. The local unitary group acts on by conjugation: Two mixed states are LU-equivalent if they lie in the same orbit of this action. Polynomial invariants under this group action are those such that for all (Turner et al., 2015, Vrana, 2011).
In the SLOCC framework, the relevant group or its projective restriction acts similarly, and the corresponding invariants possess weighted transformation properties under determinant scaling.
2. Construction and Generation of Polynomial Invariants
The algebra of polynomial LU invariants for mixed states is generated by so-called "trace monomials." For the operator space (matrix elements ), the coordinate ring is , and one constructs invariants via the multilinear trace procedure: where each runs over the symmetric group on letters, and is the block on (Turner et al., 2015, Szalay, 2011). By polarization, the invariant ring is generated as a finite polynomial algebra in such trace-monomials and their partial traces.
For mixed multipartite qudit systems, polynomial invariants can be systematically constructed via index contractions, graph-theoretic methods, and representations of the local unitary group. The algebraic independence and explicit form of generators are fully understood in low degrees via combinatorial enumeration of index contractions and the application of the Cayley-Hamilton theorem for bounded degrees (Vrana, 2011, Szalay, 2011).
3. Algebraic Structure and Degree Bounds
The ring of polynomial LU invariants for mixed states is finitely generated and Noetherian. Explicit degree bounds on generators are available via the Derksen–Kemper bound: where is the maximal degree among a set of homogeneous invariants generating the null cone (Turner et al., 2015). In practice, the generating set may require trace-monomials up to degrees scaling as .
The "girth" of an invariant—the maximal cycle-length in the corresponding permutation —is bounded by in general and when .
In the inverse-limit regime of large local dimensions, the Hilbert (Molien) series and the conjectural freeness of the algebra emerge. The number of algebraically independent generators of degree is given by the number of index- subgroup conjugacy classes in a free group on generators (for -partite systems), with the explicit Hilbert series: where counts index- subgroups of (Vrana, 2011).
4. Completeness and Separation of Orbits
A central theorem, established by Turner and Morton, is that these polynomial invariants uniquely classify mixed states up to LU-equivalence: two density operators are in the same LU-orbit if and only if all invariants agree (Turner et al., 2015). This completeness follows from (i) the equivalence of - and -orbit closures for normal (in particular Hermitian) matrices and (ii) the existence of polynomial invariants that separate closed orbits by classical results from geometric invariant theory.
Thus, a finite set of trace-monomials, up to explicit degree bounds, provides a complete system of invariants for LU equivalence of density operators in finite dimensions.
5. Special Cases and Explicit Structures
5.1. Two-Qubit Mixed States
For , explicit invariants include , , , and , as well as polynomials in partial traces over subsystems. In the Bloch tensor formalism, the LU-invariant algebra for generic two-qubit states is generated by 9 algebraically independent polynomials expressed via the Bloch parameters, cross and dot-products of local Bloch vectors and the correlation matrix, as established using Zariski-dense slices and Weyl group techniques (Candelori et al., 2023). In practical computational terms, LU-equivalence of two generic two-qubit states may be checked by comparing these invariants.
For "X-states"—a restricted 7-dimensional family—Gerdt et al. show that the LU-invariant ring modulo syzygies maps injectively to a free ring generated by five invariants of degrees (1,1,1,2,2), which are sufficient for entanglement classification within this submanifold (Gerdt et al., 2016).
5.2. Multi-Qudit and Qubit-Qutrit Systems
For bipartite systems of type , the algebra of LU-invariants is characterized via the Molien function and Poincaré series. The number of algebraically independent invariants at low degrees is explicitly determined (e.g., 3 degree-2, 4 degree-3, etc.), and minimal generating sets can be formed from traces of noncommutative monomials in the blocks of the density matrix (Gerdt et al., 2011). Casimir invariants of the corresponding Lie algebra further provide an alternative generating set, subject to tight positivity inequalities.
5.3. Degree Classification and Graph-Theoretic Methods
Szalay and others provide explicit, index-free constructions of all degree-2, 4, and 6 LU-invariants for -partite systems by enumerating index contractions via color-labeled graphs, corresponding to simultaneous action of permutations on tensor indices (Szalay, 2011). The algebraic independence of these generators in the "inverse limit" is established via group and invariant-theoretic arguments.
6. Connections to SLOCC Invariants and Pure-State Limits
For pure states, the theory of local unitary polynomial invariants is closely related but the algebraic structure of SLOCC invariants diverges: generally, polynomial SLOCC invariants are incomplete except in very special cases. In mixed states, SLOCC invariants can be constructed using hyperdeterminants of the Bloch tensor representation; for even-partite, -dimensional systems, algebraically independent invariants can be extracted from the coefficients of the hyper-characteristic polynomial of the hypermatrix representing (Jing et al., 2014).
A key mathematical equivalence is that the algebra of LU invariants of -partite mixed states is reflected, via purification, in the LU invariant algebra of -partite pure states, with degree shifts (Vrana, 2011).
7. Practical Computation and Applications
Polynomial invariants are crucial for:
- Deciding LU-equivalence of mixed states (are two states interconvertible by local unitaries?)
- Classifying entanglement types in multipartite systems
- Constructing entanglement monotones and measures (especially in low dimensions, e.g., concurrence for two qubits (Aniello et al., 2010))
- Characterizing state spaces and their positivity cones via explicit inequalities among invariants (Gerdt et al., 2011)
For many physically relevant scenarios (e.g., two- and three-qubit systems), the invariants reduce to polynomials of low degree in Bloch vector components and correlation tensors, or traces and determinants of matrix products. Algorithmically, computation of these invariants is polynomial in system size for fixed , and practical implementations rely on explicit enumeration and evaluation of prescribed index contractions (Jing et al., 2015).
In summary, the theory provides a finite, explicit, and complete algebraic framework for distinguishing mixed-state quantum orbits under local operations using polynomial invariants, with degree and number growing rapidly with system size and local dimensions. All construction principles, degree bounds, and completeness results trace to deep structures in classical and modern invariant theory, algebraic geometry, and representation theory (Turner et al., 2015, Vrana, 2011, Szalay, 2011, Candelori et al., 2023).