Two-parameter families of matrix product operator integrals of motion in Heisenberg spin chains

This paper presents the discovery of two-parameter families of matrix product operator integrals of motion for various anisotropic Heisenberg spin chains, offering a symbolic algebra approach to their construction and extending previous one-parameter results to include Floquet charges for specific protocols.

The Gist

Imagine you are watching a massive, complex dance floor filled with thousands of tiny dancers (the atoms in a magnetic chain). In the world of quantum physics, these dancers are "spins," and they are constantly interacting with their neighbors, spinning, flipping, and swapping energy.

Usually, when you have a system this complex, it's chaotic. If you nudge one dancer, the ripple effect spreads everywhere, and you can't predict what will happen next. This is called "chaos."

However, some special systems are "integrable." This is a fancy word meaning they are perfectly ordered. Even though they are moving, they follow strict rules that never change. They have hidden "conserved quantities" (like a secret scorecard that stays the same no matter how wild the dance gets). Finding these secret scorecards is the holy grail of understanding these systems.

The Problem: Finding the Secret Scorecards

For decades, physicists have been trying to write down the formulas for these secret scorecards for different types of magnetic chains (called Heisenberg models).

• The Old Way: It was like trying to solve a giant jigsaw puzzle by looking at one tiny piece at a time. You could find the scorecards, but the formulas were messy, complicated, and hard to generalize.
• The New Idea: Recently, a group of scientists found a shortcut. Instead of looking at the whole dance floor at once, they proposed building the scorecards out of small, repeating Lego blocks (called Matrix Product Operators or MPOs). If you can find the right shape for one Lego block, you can snap them together to build the whole scorecard.

What This Paper Does: The Master Key

The author of this paper, Vsevolod Yashin, took that Lego idea and supercharged it.

1. The "Two-Parameter" Magic
Previous attempts found scorecards that depended on just one knob you could turn. Yashin discovered a way to build scorecards that depend on two knobs.

• The Analogy: Imagine you have a map. A one-knob map only lets you move North or South. A two-knob map lets you move North/South and East/West. Yashin found a "two-dimensional map" of solutions.
• The Result: This map covers all the different types of magnetic dances (XXX, XXZ, XY, XYZ). If you turn the knobs to specific settings, you get the solution for one type of chain; turn them slightly differently, and you get the solution for another. It's like having one master key that opens every door in the building.

2. The "Sphere" Connection
For the most common types of chains, the author realized these two knobs correspond to points on a sphere.

• The Analogy: Imagine the Earth. You can describe any location using Latitude and Longitude. Yashin found that the "secret scorecards" for these quantum chains are just like locations on a globe. If you pick a spot on the globe (a specific angle), you get a specific, valid scorecard. This makes the math much more beautiful and easier to visualize.

3. The "Time-Traveling" Dance (Floquet Systems)
The paper also looks at what happens if you make the dancers switch partners in a rhythmic, repeating pattern (like a "brick-wall" pattern in a dance routine). This is called a "Floquet" system.

• The Analogy: Imagine a dance where the music stops and starts, and the dancers switch partners every beat. Usually, this breaks the order. But Yashin showed that even in this choppy, stop-start rhythm, the secret scorecards still exist and stay stable. He wrote down the exact formulas for these "rhythmic" scorecards.

How Did He Do It? (The Detective Work)

The author didn't just guess. He used a mix of:

• Symmetry: He looked at the rules of the dance (what happens if you flip the spins? What if you rotate them?). He knew the scorecards had to respect these rules, which narrowed down the search space massively.
• Computer Algebra: He turned the problem into a giant system of algebraic equations (like a massive Sudoku puzzle) and used a computer to solve it.
• The "Error Term" Trick: He used a clever mathematical trick where he allowed for a small "error" in the equations, as long as those errors canceled each other out when you put the whole chain together.

Why Should You Care?

You might think, "So what? It's just math about spinning atoms." But here is why it matters:

• Quantum Computers: Understanding these "perfectly ordered" systems helps us build better quantum computers. If we know the rules, we can prevent errors from ruining our calculations.
• New Materials: This helps us design new materials that conduct electricity or magnetism in weird, useful ways.
• The "Universal" Method: The author suggests this method (building things out of Lego blocks and checking symmetries) could be used to solve other difficult problems in physics, not just these spin chains.

In a Nutshell

This paper is like finding a universal remote control for a specific family of quantum dances. Before, you needed a different remote for every type of dance, and they were hard to program. Now, we have one remote with two dials. Turn the dials to the right spot (like picking a location on a globe), and it instantly programs the perfect "secret scorecard" for that dance, keeping the system perfectly ordered even when the music changes. It's a beautiful, unifying discovery that simplifies a very complex part of quantum physics.

Technical Summary

1. Problem Statement

The paper addresses the challenge of explicitly constructing local integrals of motion (charges) for one-dimensional Heisenberg spin chain models. While the integrability of these models (XXX, XXZ, XYZ) is well-established via the Bethe Ansatz and the Algebraic Bethe Ansatz, these methods do not always provide a direct, concrete description of local conserved quantities (higher Hamiltonians).

Recent work by Fendley et al. (2025) demonstrated a simplified method to prove integrability by constructing a one-parameter family of Matrix Product Operator (MPO) charges with bond dimension 4 for the XYZ model. Yashin's work seeks to extend this framework to discover two-parameter families of MPO integrals of motion applicable to a broader range of anisotropies (XXX, XXZ, XX, XY, and XYZ) and to investigate their stability under Trotterized (Floquet) dynamics.

2. Methodology

The author employs a combination of symmetry analysis, symbolic algebra, and tensor network techniques to solve the problem.

• MPO Formalism: The search is conducted for operators $M$ in the form of a Matrix Product Operator:
  $$M = \text{tr}[A_1 A_2 \cdots A_L]$$
  where $A_j$ is a $4 \times 4$ matrix of local operators acting on site $j$. The bond dimension is fixed at 4.
• The Main Equation (Commutator Condition): Instead of solving $[H, M] = 0$ directly, the author uses a sufficient condition proposed by Fendley et al. This involves finding local tensors $A$ and "error terms" $E$ such that for every local Hamiltonian term $H_{j,j+1}$:
  $$i [H_{j,j+1}, A_j A_{j+1}] = E_j A_{j+1} - A_j E_{j+1}$$
  Summing over all sites causes the error terms to telescope and cancel, ensuring $[H, M] = 0$.
• Symmetry Constraints: To reduce the complexity of the algebraic system, the author imposes symmetry requirements on the tensor $A$ based on the specific model's symmetries (e.g., $Z_2 \times Z_2$ for XYZ, $U(1)$ for XXZ, $SU(2)$ for XXX). This restricts the form of the $4 \times 4$ matrix $A$, reducing the number of free parameters.
• Symbolic Algebra & Gröbner Bases: The commutator condition is substituted with the symmetry-constrained parametrization of $A$ and $E$. This generates a system of homogeneous quadratic algebraic equations. The author uses computer algebra systems (Mathematica) and Gröbner basis methods to solve these systems.
• Trotterization: For Floquet systems, the condition is generalized to the Yang-Baxter-like equation:
  $$e^{i\tau H_{j,j+1}} A_o A_e e^{-i\tau H_{j,j+1}} = A_e A_o$$
  where $A_o$ and $A_e$ are alternating tensors for odd and even bonds.

3. Key Contributions and Results

A. Two-Parameter Families for Various Models

The primary contribution is the explicit construction of two-parameter families of MPO charges for five distinct Heisenberg models. These solutions depend on parameters that can be geometrically interpreted (e.g., points on a sphere or projective plane).

1. XXX Model (Isotropic):

   • Found a solution depending on parameters $(\theta, \phi)$, corresponding to points on a sphere.
   • The MPO tensor $A(\theta, \phi)$ generalizes the one-parameter solution found by Katsura.
   • Completeness Conjecture: Expanding the MPO near the zero point generates local charges of all weights. The author conjectures this family contains information about all local integrals of motion for the XXX model.

2. XXZ Model (Anisotropic):

   • Derived a two-parameter solution generalizing the XXX case.
   • Provided two parametrizations: one using spherical coordinates and another using diagonal elements $(\omega_x, \omega_z)$, which is useful for taking limits.
   • The solution reduces to the XXX case when the anisotropy $\Delta \to 1$.

3. XX and XY Models:

   • Obtained solutions by taking the limit $J_z \to 0$ (or $\Delta \to 0$) from the XXZ solution.
   • These solutions are consistent with known free-fermion descriptions (Onsager strings).

4. XYZ Model (General Anisotropic):

   • Main Result: Discovery of a two-parameter family of MPO charges depending on homogeneous coordinates $[\pi_x : \pi_y : \pi_z]$ (points on a complex projective plane).
   • This solution unifies all previous cases. The XXX, XXZ, XX, and XY solutions are obtained as specific limiting cases of the XYZ solution.
   • The solution is expressed using determinants involving the coupling constants $J_x, J_y, J_z$ and the parameters.

B. Floquet Integrals of Motion (Trotterized Dynamics)

The paper explicitly constructs two-parameter families of Floquet charges for the two-step brick-wall Trotterized protocols of the XXX and XXZ models.

• The author provides explicit forms for the alternating tensors $A_o$ and $A_e$ that satisfy the discrete-time evolution condition.
• In the limit of the time step $\tau \to 0$, these Floquet charges reduce to the continuous-time MPO charges derived earlier, demonstrating the stability of these integrals under discretization.

C. Connection to Known Solutions

• The paper demonstrates that the previously known one-parameter solutions by Fendley et al. and Fukai & Yamada are specific limiting cases of the new two-parameter families found here.
• Unlike Baxter's solution for the eight-vertex model, these MPO solutions do not require special functions (like elliptic functions) for their definition, relying instead on algebraic structures.

4. Significance and Outlook

• Unified Framework: The work provides a unified algebraic framework for generating local charges across the entire hierarchy of Heisenberg models (from isotropic to fully anisotropic) using a single MPO ansatz with bond dimension 4.
• Completeness Conjecture: The author conjectures that these MPO families are complete for the class of local charges. If proven, this would mean that any local conserved quantity in these integrable models can be generated as a linear combination of the coefficients in the series expansion of these MPOs.
• Applications to Dynamics:
  • Generalized Gibbs Ensembles (GGE): Having an analytical characterization of all charges allows for the description of non-equilibrium steady states using generalized temperatures.
  • Dynamical Freezing: The solutions are relevant for studying dynamical freezing and non-chaotic behavior in periodically driven systems.
  • Quantum Circuits: The Floquet solutions are directly applicable to the study of integrable quantum circuits and dual-unitary circuits.
• Methodological Impact: The paper outlines a general "Step 0 to Step 3" methodology for finding tensor network observables in integrable systems, which can be extended to other models (e.g., Hubbard model, Kitaev honeycomb model) and higher dimensions (PEPOs).
• Numerical Potential: The reduction of the problem to quadratic algebraic equations suggests that numerical methods (semidefinite programming) could be used to find approximate integrals of motion in non-integrable or chaotic systems to detect "near-integrability."

In summary, Yashin's work significantly advances the constructive theory of integrable spin chains by providing explicit, parameter-rich MPO representations of conserved quantities, bridging the gap between abstract integrability and concrete local operator structures, and extending these results to driven (Floquet) systems.