The density-matrix renormalization group applied to transfer matrices: Static and dynamical properties of one-dimensional quantum systems at finite temperature
The density-matrix renormalization group (DMRG) applied to transfer matrices allows it to calculate static as well as dynamical properties of one-dimensional quantum systems at finite temperature in the thermodynamic limit. To this end the quantum system is mapped onto a two-dimensional classical system by a Trotter-Suzuki decomposition. Here we discuss two different mappings: The standard mapping onto a two-dimensional lattice with checkerboard structure as well as an alternative mapping introduced by two of us. For the classical system an appropriate quantum transfer matrix is defined which is then treated using a DMRG scheme. As applications, the calculation of thermodynamic properties for a spin-1/2 Heisenberg chain in a staggered magnetic field and the calculation of boundary contributions for open spin chains are discussed. Finally, we show how to obtain real time dynamics from a classical system with complex Boltzmann weights and present results for the autocorrelation function of the XXZ-chain.
The density-matrix renormalization group (DMRG) applied to transfer matrices allows it to calculate static as well as dynamical properties of one-dimensional quantum systems at finite temperature in the thermodynamic limit. To this end the quantum system is mapped onto a two-dimensional classical system by a Trotter-Suzuki decomposition. Here we discuss two different mappings: The standard mapping onto a two-dimensional lattice with checkerboard structure as well as an alternative mapping introduced by two of us. For the classical system an appropriate quantum transfer matrix is defined which is then treated using a DMRG scheme. As applications, the calculation of thermodynamic properties for a spin-1/2 Heisenberg chain in a staggered magnetic field and the calculation of boundary contributions for open spin chains are discussed. Finally, we show how to obtain real time dynamics from a classical system with complex Boltzmann weights and present results for the autocorrelation function of the XXZ-chain.
The present work applies a numerical method, namely the transfer-matrix density-matrix renormalization group (TMRG), to two seemingly different types of models. In a first part the TMRG is used to investigate the thermodynamics of one-dimensional fermionic models. A second part deals with a novel TMRG method for one-dimensional stochastic models, whose development is an integral part of the thesis. First, the traditional TMRG algorithm for quantum systems is outlined in its historical context. Two different variants are presented, following works of Xiang et al. and Sirker and Klümper, respectively. The basic idea of the method is to map the thermodynamics of a one-dimensional quantum model by Trotter-Suzuki decomposition onto a two-dimensional statistical one. The latter is then solved by a transfer-matrix approach combined with the iterative numerical procedure of White's density-matrix renormalization-group (DMRG) algorithm. Thereby precise computations of various thermodynamic properties, such as thermodynamic potentials, susceptibilities, thermal expectation values and correlation functions are possible. As the first part of the thesis deals with fermionic models, we next review some basics about the theory of strongly correlated fermions in one dimension. Thereupon we elucidate the so-called Hirsch model, which recently gained a lot of theoretic interest in respect to high-temperature superconductivity. It extends the well-studied Hubbard model by an off-diagonal bond-charge interaction term. The current state of research is briefly summarized and mainly refers to ground state properties. Showing numerical TMRG results we then investigate and discuss the almost unknown thermodynamics of the Hirsch model. Various phases are identified and characterized in terms of Tomonaga-Luttinger and Luther-Emery liquid properties, in accordance with previous studies of the ground state. As an important result, superconducting singlet-pair correlation lenths are observed to dominate the physics at finite temperatures in a certain spin-gaped phase. Subsequent to our thermodynamic studies, we turn to the second part of the thesis and outline some theoretic basics of stochastic models. Most notably is the important formal analogy of the master equation, that describes the dynamics of the model, to a Schrödinger equation in imaginary time. This analogy is used to construct a stochastic TMRG algorithm almost similar to the quantum case, that facilitates the computation of dynamic properties, e.g. the local density of particles. We intensively focus on interesting mathematical properties of the stochastic transfer-matrix. As an astonishing result it is found, that the temporal evolution of the non-equilibrium process is reflected by a certain causal structure of the stochastic TMRG. But even if this new approach seems to be promising at first glance, severe numerical problems limit significantly its practical use. In order to solve these instabilities we propose a completely new variant of the algorithm, which we call stochastic light-cone corner-transfermatrix DMRG (LCTMRG). As suggested by its name, the LCTMRG makes use of the causal structure mentioned above and combines it with the stochastic TMRG algorithm. Applications of the LCTMRG onto various reaction-diffusion models verify highly precise numerical data and a great improve compared to the old algorithm by several orders of magnitude. Additionally it is stressed, that the newly proposed analysis tool provides some considerable advantages to common simulation techniques.
The density matrix renormalization group (DMRG) method and its applications to finite temperatures and two-dimensional systems are reviewed. The basic idea of the original DMRG method, which allows precise study of the ground-state properties and low-energy excitations, is presented for models which include long-range interactions. The DMRG scheme is then applied to the diagonalization of the quantum transfer matrix for one-dimensional systems, and a reliable algorithm at finite temperatures is formulated. Dynamic correlation functions at finite temperatures are calculated from the eigenvectors of the quantum transfer matrix with analytical continuation to the real frequency axis. An application of the DMRG method to two-dimensional quantum systems in a magnetic field is demonstrated and reliable results for quantum Hall systems are presented.
The density matrix renormalization group (DMRG) is a numerical method for studying low dimensional strongly correlated models. The study of these models is one of the most active areas of condensed matter physics today and has been spurred not only by the discovery of high temperature superconductivity but also by experiments on new compounds. In this thesis we study fundamentals of the density matrix renormalization group method as well as applications of the method to Heisenberg spin chains. The introductory text is meant as a brief overview of the DMRG as well as Heisenberg spin systems and provides an introduction to the papers. In papers I and II we investigate the density matrix renormalization group with the aim to better understand the underlying structure of the method. We find that if the renormalization converges to a fixed point in the thermodynamic limit, the resulting basis states can be written as a matrix product form. We also show an explicit relation of the DMRG to variational calculations using the fact that the matrix product states can be rederived through a variational ansatz without any reference to the DMRG calculation. It is then shown how the matrix product states can be used to define a set of Bloch waves for describing elementary excitations of the Heisenberg spin-1 chain. The low lying energy spectrum is calculated. In papers III and IV we investigate the thermodynamics of impurities in a spin-1/2 chain. For this purpose we have shown that a recently developed transfer matrix DMRG method can be used to calculate impurity properties. Both impurity corrections to thermodynamic properties and local properties at finite temperature can be obtained. The method gives quicker and more accurate results than quantum Monte Carlo. The numerical results show good agreement with field theory calculations for a number of impurity configurations and the cross-over function for the magnetic susceptibility from high- to low-temperature behavior shows the expected data collapse. Our results also give quantitative predictions for experiments on doped spin-1/2 chains.
We study the one-dimensional S = 1/2 Heisenberg model with an uniform and a staggered magnetic fields, using the dynamical density-matrix renormalization group (DDMRG) technique. The DDMRG enables us to investigate the dynamical properties of chain with lengths up to a few hundreds, and the results are numerically exact in the same sense as 'exact diagonalization' results are. Thus, we can analyze the low-energy spectrum almost in the thermodynamic limit. In this work, we calculate the dynamical spin structure factor and demonstrate the performance of the DDMRG method applying the open-end boundary conditions as well as the periodic boundary conditions.
The density matrix renormalization group is one of the most powerful numerical methods for computing ground-state properties of two-dimensional (2D) quantum lattice systems. Here we show its finite-temperature extensions are also viable for 2D, using the following strategy: At high temperatures, we combine density-matrix purification and numerical linked-cluster expansions to extract static observables directly in the thermodynamic limit. At low temperatures inaccessible to purification, we use the minimally entangled typical thermal state (METTS) algorithm on cylinders. We consider the triangular Heisenberg antiferromagnet as a first application, finding excellent agreement with other state of the art methods. In addition, we present a METTS-based approach that successfully extracts critical temperatures, and apply it to a frustrated lattice model. On a technical level, we compare two different schemes for performing imaginary-time evolution of 2D clusters, finding that a Suzuki-Trotter decomposition with swap gates is currently the most accurate and efficient.
Ulrich Schollwo¨ck Arnold Sommerfeld Center for Theoretical Physics, University of Munich, Germany Since its invention in 1992 by White [1,2], the density matrix renormalization group (DMRG) has firmly established itself as the currently most powerful method to calculate static and more recently also dynamic [3-5] properties of low-lying eigenstates of strongly correlated Hamiltonians defined on onedimensional quantum lattices. At the same time, its extension to the analysis of two-dimensional classical [6] and one-dimensional quantum [7,8] transfer matrices has given access to highly precise thermodynamics in classical two-dimensional and quantum one-dimensional systems. It has even been extended to the numerically much more demanding study of non-Hermitian (pseudo-)Hamiltonians emerging in the analysis of classical steady states in one-dimensional systems [9-11]. However, it also turned out quite quickly that extending DMRG to the study of two-dimensional systems did not lead to similarly huge successes.