Finding local integrals of motion in quantum lattice models in the thermodynamic limit
J. Pawłowski, J. Herbrych, M. Mierzejewski · 2025
Read Finding local integrals of motion in quantum lattice models in the thermodynamic limit by J. Pawłowski, J. Herbrych, M. Mierzejewski (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-05882.txt.
Introduction
Pawłowski, Herbrych, and Mierzejewski find hidden conserved traits in quantum lattice models. Their paper, from twenty twenty-five, deals with local integrals of motion. These are quantities that never change as a system runs. They shape how a closed system behaves over long times.
Old routes to find them need huge matrix work, which caps the system size. The new method cuts the task down to one with an exact answer even for endless chains. A simple algorithm does the job fast. Tests on spin chains and quantum circuits recover known results. The method also spots slow modes in near-ordered models and times how fast they fade.
Catalog record
| CREATOR | J. Pawłowski, J. Herbrych, M. Mierzejewski |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 37 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-05882
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Finding local integrals of motion in quantum lattice models in the thermodynamic limit J. Pawłowski, J. Herbrych, M. Mierzejewski Abstract Local integrals of motion (LIOMs) play a key role in understanding the long-time properties of closed macroscopic systems. They were found for selected integrable systems via complex analytical calculations. The existence of LIOMs and their structure can also be studied via numerical methods, which, however, involve exact diagonalization of Hamiltonians, posing a bottleneck for such studies. We show that finding LIOMs in translationally invariant lattice models or unitary quantum circuits can be reduced to a problem for which one may numerically find an exact solution in the thermodynamic limit. We develop a simple algorithm and demonstrate its efficiency by calculating LIOMs and bounds on correlations (the Mazur bounds) for infinite integrable spin chains and unitary circuits. Finally, we demonstrate that this approach identifies slow modes in nearly integrable spin models and estimates their relaxation times. I Introduction Closed macroscopic quantum systems undergo a unitary evolution. However, a reduced density matrix of a subsystem is expected to reach a stationary state. The latter is determined by local conservation laws or, equivalently, by local integrals of motion (LIOMs). In particular, the expectation values of local observables are equal to averages over an appropriate ensemble [ 1 ] . Generic systems have only a few conserved quantities (like Hamiltonian, particle number, magnetization) and the relevant ensemble has the form of the thermal Gibbs state [ 2 , 3 , 4 ] . Interacting integrable systems have extensive number of LIOMs and the stationary states (of subsystems) have the form that is consistent with the generalized Gibbs ensemble [ 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 ] . Identification of LIOMs is also the starting point for construction of the hydrodynamics or the generalized hydrodynamics [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 ] in case of integrable models. Moreover, anomalous transport (or its anticipated absence) in various quantum systems has recently been linked to the presence of LIOMs. Most prominent examples concern strongly disordered systems [ 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 ] , tilted chains with emergent conservation of the dipole moment [ 45 ] and systems exhibiting the Hilbert space fragmentation [ 46 , 47 , 45 ] . While the importance of LIOMs for the long-time dynamics appears unquestionable, in many cases their construction remains a challenging problem. Even for the most studied integrable interacting model, i.e., the XXZ chain, the complete set of LIOMs (that includes also quasilocal LIOMs) has been established analytically only recently [ 48 , 49 , 50 , 51 , 14 , 52 , 53 ] . In many cases, LIOMs are identified through complex calculations that are specific to the investigated model [ 54 , 55 , 56 , 57 , 58 , 59 , 60 , 61 ] . Although there are model-independent approaches which allow to find LIOMs in integrable models [ 62 , 63 ] or approximate LIOMs in nearly integrable systems [ 64 ] , they are based on exact diagonalization of Hamiltonians. The latter poses serious limitations on the studied models and raises questions concerning the reliability of the finite-size scaling. In this work, we show that finding LIOMs in translationally-invariant lattice models belongs to the same class of problems as the high-temperature expansion [ 65 , 66 , 67 ] . Namely, one can numerically find exact LIOMs and the corresponding Mazur bounds [ 68 , 69 ] in the thermodynamic limit without diagonalizing the Hamiltonian. This holds true for quantum systems where the dynamics is set either by a Hamiltonian or by a unitary transformation, as is the case in unitary circuits. The geometry of the system is not important, and the only essential assumption concerns the dynamics that needs to be set by local transformations. II General approach II.1 Algorithm We consider a system containing L sites, which are enumerated by the index l=1,\ldots,L , and study extensive Hermitian operators, \hat{O}^{s}=\sum_{l}\hat{O}^{s}_{l} , where the upper index, s , distinguishes between various operators. The densities, \hat{O}^{s}_{l} , are local and act on up to M sites adjacent to l . We introduce the Hilbert-Schmidt (HS) product \langle\hat{O}^{s}\hat{O}^{p}\rangle={\rm Tr}(\hat{O}^{s\dagger}\hat{O}^{p})/(ZL) and the HS norm of operators, \lVert\hat{O}^{s}\lVert^{2}=\langle\hat{O}^{s}\hat{O}^{s}\rangle , where Z is the dimension of the Hilbert space. We recall that the HS product of the local densities, \langle\hat{O}^{s}_{l}\hat{O}^{s^{\prime}}_{l^{\prime}}\rangle , can be exactly calculated for infinite lattice systems using just the properties of the operator algebra, similarly to the technique employed in the high-temperature expansion. Our main assumption is that the dynamics is set either by a local Hamiltonian, \hat{H} , or by a repeated sequence of local unitary transformations \hat{U} . In the former case, we take \hat{H}=\sum_{l}\hat{H}_{l} , where \hat{H}_{l} acts on M_{H} sites. We look for local operators, \hat{Q} , which commute with the Hamiltonian, [\hat{Q},\hat{H}]=0 or, equivalently, we require vanishing of the norm of a Hermitian operator \hat{Q}^{\prime}=i[\hat{Q},\hat{H}] , i.e., \lVert\hat{Q}^{\prime}\lVert^{2}=\langle\hat{Q}^{\prime}\hat{Q}^{\prime}\rangle=0 . We note that \hat{Q}^{\prime} is still local, \hat{Q}^{\prime}=\sum_{l}\hat{Q}^{\prime}_{l} , where \hat{Q}^{\prime}_{l} acts on up to M+M_{H}-1 sites. For this reason, \langle\hat{Q}^{\prime}\hat{Q}^{\prime}\rangle can also be calculated exactly in the thermodynamic limit. In case of circuits built out of repeated sequence of local unitary transformations, the same reasoning can be applied for \hat{Q}^{\prime}=\hat{U}^{\dagger}\hat{Q}\hat{U}-\hat{Q} . I.e., \hat{Q}^{\prime} is a local operator and vanishing norm of \hat{Q}^{\prime} implies that \hat{Q} is conserved. Both cases can be comprehended within a unified formulation. Namely, we introduce a function f(\hat{O}^{s}) which has the following properties: (i) f is linear, (ii) vanishing of \lVert f(\hat{O}^{s})\lVert is equivalent to conservation of \hat{O}^{s} ; (iii) f(\hat{O}^{s}) is a local and Hermitian operator so that one can exactly calculate \langle f(\hat{O}^{s})f(\hat{O}^{p})\rangle . If the dynamics is set by a Hamiltonian then one may choose f(\hat{O}^{s})=i[\hat{O}^{s},\hat{H}] , whereas f(\hat{O}^{s})=\hat{U}^{\dagger}\hat{O}^{s}\hat{U}-\hat{O}^{s} can be used for unitary circuits. The specific choice of f is irrelevant and can be adapted to the desired setup, as long as it is consistent with conditions (i) - (iii) . As a next step, we consider a set containing N_{O} orthonormal operators, \langle\hat{O}^{s}\hat{O}^{p}\rangle=\delta_{sp} . We aim to find all LIOMs which can be expressed as linear combinations of these operators. To this end, we solve the eigenproblem \sum_{sp}V_{\alpha s}F_{sp}\;(V^{T})_{p\beta}=\lambda_{\alpha}\delta_{\alpha\beta}, (1) where V is an orthogonal matrix and F is a matrix of HS products, F_{sp}=\langle f(\hat{O}^{s})f(\hat{O}^{p})\rangle , constructed in the basis (of size N_{O} ) containing all \hat{O}^{s} . Since f is a linear function, one may introduce rotated operators \hat{A}^{\alpha}=\sum_{s}V_{\alpha s}\hat{O}^{s}, (2) and rewrite ( 1 ) as \langle f(\hat{A}^{\alpha})f(\hat{A}^{\beta})\rangle=\lambda_{\alpha}\delta_{\alpha\beta}, (3) from which one finds the norm \lVert f(\hat{A}^{\alpha})\lVert=\sqrt{\lambda_{\alpha}} . The rotated operators are local and orthogonal, \langle\hat{A}^{\alpha}\hat{A}^{\beta}\rangle=\delta_{\alpha\beta} . Consequently, all \hat{A}^{\alpha} corresponding to the eigenvalue \lambda_{\alpha}=0 represent orthogonal LIOMs, identified here via the norm \lVert f(\hat{A}^{\alpha})\lVert=0 . To distinguish these LIOMs from other operators introduced in ( 2 ), we denote LIOMs as \hat{Q}^{\alpha} where