arXiv subjects

Y. Ohta

Publications and source records attributed to Y. Ohta.

At least 73 records · Page 4Linked to original sources

arXiv1999

Singularity confinement and algebraic entropy: the case of the discrete Painlevé equations

We examine the validity of the results obtained with the singularity confinement integrability criterion in the case of discrete Painlevé equations. The method used is based on the requirement of non-exponential growth of the homogeneous degree of the iterate of the mapping. We show that when we start from an integrable autonomous mapping and deautonomise it using singularity confinement the degrees of growth of the nonautonomous mapping and of the autonomous one are identical. Thus this low-growth based approach is compatible with the integrability of the results obtained through singularity confinement. The origin of the singularity confinement property and its necessary character for integrability are also analysed.

solv-int↗

arXiv1998

On Discrete Painleve Equations Associated with the Lattice KdV Systems and the Painleve VI Equation

A new integrable nonautonomous nonlinear ordinary difference equation is presented which can be considered to be a discrete analogue of the Painleve V equation. Its derivation is based on the similarity reduction on the two-dimensional lattice of integrable partial difference equations of KdV type. The new equation which is referred to as GDP (generalised discrete Painleve equation) contains various ``discrete Painleve equations'' as subcases for special values/limits of the parameters, some of which were already given before in the literature. The general solution of the GDP can be expressed in terms of Painleve VI (PVI) transcendents. In fact, continuous PVI emerges as the equation obeyed by the solutions of the discrete equation in terms of the lattice parameters rather than the lattice variables that label the lattice sites. We show that the bilinear form of PVI is embedded naturally in the lattice systems leading to the GDP. Further results include the establishment of Baecklund and Schlesinger transformations for the GDP, the corresponding isomonodromic deformation problem, and the self-duality of its bilinear scheme.

solv-int↗

arXiv1998

On Integrability and Chaos in Discrete Systems

The scalar nonlinear Schrodinger (NLS) equation and a suitable discretization are well known integrable systems which exhibit the phenomena of ``effective'' chaos. Vector generalizations of both the continuous and discrete system are discussed. Some attention is directed upon the issue of the integrability of a discrete version of the vector NLS equation.

solv-int↗

arXiv1998

On Discretizations of the Vector Nonlinear Schrodinger Equation

Two discretizations of the vector nonlinear Schrodinger (NLS) equation are studied. One of these discretizations, referred to as the symmetric system, is a natural vector extension of the scalar integrable discrete NLS equation. The other discretization, referred to as the asymmetric system, has an associated linear scattering pair. General formulae for soliton solutions of the asymmetric system are presented. Formulae for a constrained class of solutions of the symmetric system may be obtained. Numerical studies support the hypothesis that the symmetric system has general soliton solutions.

solv-int↗

arXiv1998

Optical Conductivity of the Trellis-Lattice t-J Model: Charge Fluctuations in NaV_2O_5

Optical conductivity of the trellis lattice t-J model at quarter filling is calculated by an exact-diagonalization technique on small clusters, whereby the valence state of V ions of NaV_2O_5 is considered. We show that the experimental features at \sim 1 eV, including peak positions, presence of shoulders, and anisotropic spectral weight, can be reproduced in reasonable range of parameter values, only by assuming that the system is in the charge disproportionated ground state. Possible reconciliation with experimental data suggesting the presence of uniform ladders at T>T_c is discussed.

cond-mat.str-el↗

arXiv1998

A Model Study of the Low-Energy Charge Dynamics of NaV_2O_5

An exact-diagonalization technique on small clusters is used to calculate the dynamical density correlation functions of the dimerized t-J chain and coupled anisotropic t-J ladders (trellis lattice) at quarter filling, i.e., the systems regarded as a network of pairs (dimers or rungs) of sites coupled weakly via the hopping and exchange interactions. We thereby demonstrate that the intersite Coulomb repulsions between the pairs induce a low-energy collective mode in the charge excitations of the systems where the internal charge degrees of freedom of the pairs play an essential role. Implications to the electronic states of NaV_2O_5, i.e., fluctuations of the valence state of V ions and phase transition as a charge ordering, are discussed.

cond-mat.str-el↗

arXiv1998

NaV_2O_5 as an Anisotropic t-J Ladder at Quarter Filling

Based on recent experimental evidences that the electronic charge degrees of freedom plays an essential role in the spin-Peierls--like phase transition of NaV$_2$O$_5$, we first make the mapping of low-energy electronic states of the $d$$-$$p$ model for NaV$_2$O$_5$ to the quarter-filled $t$$-$$J$ ladder with anisotropic parameter values between legs and rungs, and then show that this anisotropic $t$$-$$J$ ladder is in the Mott insulating state, of which lowest-energy states can be modeled by the one-dimensional Heisenberg antiferromagnet with the effective exchange interaction $J_{eff}$ whose value is consistent with experimental estimates. We furthermore examine the coupling between the ladders as the trellis lattice model and show that the nearest-neighbor Coulomb repulsion on the zigzag-chain bonds can lead to the instability in the charge degrees of freedom of the ladders.

cond-mat.str-el↗

arXiv1998

Critical Dimerization Strength of the Quarter-Filled t-J Model

An exact-diagonalization technique on small clusters is used to study the ground state of the dimerized t-J model at quarter filling. The equal-time charge and spin correlations, charge and spin gaps, and binding energy are calculated for the two-dimensional lattice with a spatial dimerization pattern corresponding to organic conductors $κ-(BEDT-TTF)_2X$. We show that competition between the effects of two types of dimerization, i.e., dimerization in the hopping integral and dimerization in the exchange interaction, leads to the insulator-superconductor transition in the ground state of the model. The phase diagram is thereby presented.

cond-mat.str-el↗

arXiv1998

Ground-state phase diagram of the one-dimensional dimerized t-J model at quarter filling

The ground state of the one-dimensional dimerized $t$$-$$J$ model at quarter filling is studied by a Lanczos exact-diagonalization technique on small clusters. We calculate the charge gap, spin gap, binding energy, Tomonaga-Luttinger-liquid parameter, Drude weight, anomalous flux quantization, etc. We thereby show that the two types of dimerization, i.e., a dimerization of hopping integral and a dimerization of exchange interaction play a mutually competing role in controlling the ground state of the system and this leads to the emergence of various phases including the Mott insulating, Tomonaga-Luttinger-liquid, and Luther-Emery-liquid phases. The ground-state phase diagram of the model is given on the parameter space of the dimerizations.

cond-mat.str-el↗

arXiv1997

Bogoliubov quasiparticle spectra of the effective d-wave model for cuprate superconductivity

An exact-diagonalization technique on finite-size clusters is used to study the ground state and excitation spectra of the two-dimensional effective fermion model, a fictious model of hole quasiparticles derived from numerical studies of the two-dimensional t-J model at low doping. We show that there is actually a reasonable range of parameter values where the $d_{x^2-y^2}$-wave pairing of holes occurs and the low-lying excitation can be described by the picture of Bogoliubov quasiparticles in the BCS pairing theory. The gap parameter of a size $Δ_d\simeq 0.13|V|$ (where $V$ is the attractive interaction between holes) is estimated at low doping levels. The paired state gives way to the state of clustering of holes for some stronger attractions.

cond-mat.str-el↗

arXiv1997

The 1D t-J model with next-nearest neighbor hopping - breakdown of the Luttinger liquid?

We investigate the effect of a next-nearest neighbor hopping integral t' in the 1D t-J model, using Lanczos diagonalization of finite chains. Surprisingly, even moderate values of t' have a quite dramatic effect on the dynamical correlation functions and Fermi surface volume. The `holon' becomes diffuse and overdamped, the band structure near the Fermi energy is dominated by t'. Most importantly, however, with increasing hole concentration the system undergoes a phase transition from a `hole pocket' Fermi surface, where k_F corresponds to the number of holes rather than electrons, to a high doping phase where the doped holes correspond to spinless Fermions. Neither phase thus has a Fermi momentum compatible with the `standard' Luttinger liquid.

cond-mat.str-el↗

arXiv1997

Landau mapping and Fermi liquid parameters of the 2D t-J model

We study the momentum distribution function n(k) in the 2D t-J model on small clusters by exact diagonalization. We show that n(k) can be decomposed systematically into two components with Bosonic and Fermionic doping dependence. The Bosonic component originates from the incoherent motion of holes and has no significance for the low energy physics. For the Fermionic component we exlicitely perform the one-to-one Landau mapping between the low lying eigenstates of the t-J model clusters and those of an equivalent system of spin-1/2 quasiparticles. This mapping allows to extract the quasiparticle dispersion, statistics, and Landau parameters. The results show conclusively that the 2D t-J model for small doping is a Fermi liquid with a `small' Fermi surface and a moderately strong attractive interaction between the quasiparticles.

cond-mat.str-el↗

arXiv1997

Photoemission spectra of the t-J model in 1 and 2D: similarities and differences

We present an exact diagonalization study of the single particle spectral function in the 1D and 2D t-J model. By studying the scaling properties with J and t we find a simple building pattern in 1D and show that every spectral feature can be uniquely assigned by a spinon and holon momentum. We find two types of low energy excitations: a band with energy scale J and high spectral weight disperses upwards in the interior part of the Brillouin zone and reaches E_F at k_F, and a band with energy scale t and low spectral weight disperses downwards in the outer part of the zone, touching E_F at 3k_F. An analogous analysis of the 2D case at half filling shows that the t-band exists also in this case, but is diffuse and never reaches the Fermi energy. For the doped case in 2D the picture is more reminiscent of 1D, in particular the `main-band' with a dispersion proportional J and the `shadow band' with energy scale t can be identified also in this case. This leads us to propose that the shadow bands discovered by Aebi et al. in Bi2212 are the 2D analogue of the 3k_F singularity in 1D systems and unrelated to antiferromagnetic spin fluctuations.

cond-mat.str-el↗

arXiv1996

Doping dependent quasiparticle band structure in cuprate superconductors

We present an exact diagonalization study of the single particle spectral function in the so-called t-t'-t''-J model in 2D. As a key result, we find that unlike the `pure' t-J model, hole doping leads to a major reconstruction of the quasiparticle band structure near (pi,0): whereas for the undoped system the quasiparticle states near (pi,0) are deep below the top of the band at (pi/2,pi/2), hole doping shifts these states up to E_F, resulting in extended flat band regions close to E_F and around (pi,0). This strong doping-induced deformation can be directly compared to angle resolved photoemission results on Sr_2 Cu Cl_2 O_2, underdoped Bi2212 and optimally doped Bi2212. We propose the interplay of long range hopping and decreasing spin correlations as the mechanism of this deformation.

cond-mat.supr-con↗

arXiv1996

Optical conductivity of strongly correlated electron systems

We present an exact diagonalization study of the frequency and wave vector dependent conductivity $σ(q,ω)$ in small clusters of $2D$ $t$$-$$J$ model. Unlike the related dynamical density correlation function, $σ(q=0,ω)$ in the underdoped regime has the exchange constant $J$ as its characteristic energy scale and is dominated by a resonance-like excitation with frequency $\sim 1.7J$. We interpret this as transition to a $p$-like excited state of a spin-bag type quasiparticle (or, alternatively, a tightly bound spinon-holon pair) and show that a simple calculation based on the string picture explains the numerical results semiquantitatively. For doping levels $\ge 25% t$ remains the only energy scale of $σ(q=0,ω)$.

cond-mat↗

arXiv1996

Low energy excitations of the t-J model in 1D and 2D

We present an exact diagonalization study of the low energy singlet and triplet states for both 1D and 2D t-J model. A scan of the parameter ratio J/t shows that for nearly all low energy states in both 1D and 2D the excitation energy takes the linear form E(t,J) = a t + b J. In 1D this is the natural consequence of the factorization of the low energy wave functions, i.e. spin-charge separation. Examination of the low energy eigenstates in 2D shows that in the singlet sector the lowest states are collective modes, which for larger J correspond to `dynamical hole columns', i.e. a periodic modulation of the hole density. This modulation is well reproduced by treating holes as hard-core Bosons with an attractive interaction.

cond-mat↗

arXiv1996

From Discrete to Continuous Painlevé Equations: A Bilinear Approach

We present the bilinear forms of the (continuous) Painlevé equations obtained from the continuous limit of the analogous expresssions for the discrete ones. The advantage of this method is that it leads to very symmetrical results. A new and interesting result is the bilinearization of the P$_{\rm VI}$ equation, something that was missing till now.

solv-int↗

arXiv1996

Cellular Automata and Ultra-Discrete Painlevé Equations

Starting from integrable cellular automata we present a novel form of Painlevé equations. These equations are discrete in both the independent variable and the dependent one. We show that they capture the essence of the behavior of the Painlevé equations organizing themselves into a coalescence cascade and possessing special solutions. A necessary condition for the integrability of cellular automata is also presented.

solv-int↗