Magnetic Excitations in the Spin-1 Anisotropic Heisenberg Antiferromagnetic Chain SystemNiCl2−4SC(NH2)2

Abstract

NiCl2-4SC(NH2)2 (DTN) is a quantum S = 1 chain system with strong easy-pane anisotropy and a new candidate for the Bose-Einstein condensation of the spin degrees of freedom. ESR studies of magnetic excitations in DTN in fields up to 25 T are presented. Based on analysis of the single-magnon excitation mode in the high-field spin-polarized phase and previous experimental results [ Phys. Rev. Lett. 96, 077204 (2006)], a revised set of spin-Hamiltonian parameters is obtained. Our results yield D = 8.9 K, Jc = 2.2 K, and J a,b = 0.18 K for the anisotropy, intrachain, and interchain exchange interactions, respectively. These values are used to calculate the antiferromagnetic phase boundary, magnetization and the frequency-field dependence of two-magnon bound-state excitations predicted by theory and observed in DTN for the first time. Excellent quantitative agreement with experimental data is obtained. PACS numbers: 75.40.Gb, 75.10.Jm Antiferromagnetic (AFM) quantum spin-1 chains have been the subject of intensive theoretical and experimental studies, fostered especially by the Haldane conjecture [1]. Due to quantum fluctuations, an isotropic spin-1 chain has a spin-singlet ground state separated from the first excited state by a gap ∆ ∼ 0.41J [2], where J is the exchange interaction. As shown by Golinelli et al. [3], the presence of a strong easy-plane anisotropy D can significantly modify the excitation spectrum, so that the gap size is not determined by the strength of the AFM quantum fluctuations exclusively, but depends on the dimensionless parameter ρ = D/J. The Haldane phase is predicted to survive up to ρ c = 0.93 [4], where the system undergoes a quantum phase transition. For ρ > ρ c the gap reopens, but its origin is dominated by the anisotropy D, and the system is in the so-called large-D regime. While the underlying physics of Haldane chains is fairly well understood, relatively little is known about the magnetic properties (and particularly the elementary excitation spectrum) of nonHaldane S = 1 AFM chains in the large-D phase. Intense theoretical work and numerous predictions [3,4,5,6,7,8,9,10] make the experimental investigation of large-D spin-1 chains a topical problem in low-dimensional magnetism.Recently, weakly-coupled spin-1 chains have attracted renewed interest due to their possible relevance to the fieldinduced Bose-Einstein condensation (BEC) of magnons. When the field H, applied perpendicular to the easy plane, exceeds a critical value H c1 (defined at T = 0), the gap closes and the system undergoes a transition into an XY -like AFM phase with a finite magnetization and AFM magnon excitations. If the spin Hamiltonian has axial symmetry with respect to the applied field, the AFM ordering can be described as BEC of magnons by mapping the spin-1 system into a gas of semi-hard-core bosons [11]. The applied field plays the role of a chemical potential, changing the boson population. In accordance with mean-field BEC theory [12,13,14], the phasediagram boundary for a three-dimensional system sh...

Thermal Transport and Strong Mass Renormalization inNiCl2−4SC(NH2)2

“…We will show that the large asymmetry between the peaks in the low-temperature specific heat, C v (H), in the vicinity of H c1 and H c2 is closely described by analytical and Quantum Monte Carlo (QMC) calculations. The mass renormalization also explains similar asymmetries observed in other properties of DTN, such as magnetization [7], electron spin resonance [8], sound velocity [9,10], and magnetostriction [11]. In a remarkable contrast to these properties, peaks in the low-temperature thermal conductivity, κ, near H c1 and H c2 do not show any substantial asymmetry.…”

“…The dominant single-ion uniaxial anisotropy D = 8.9 K splits the Ni S = 1 triplet into an S z = 0 ground state and an S z = ±1 excited doublet. The antiferromagnetic exchange coupling between Ni ions is J c = 2.2 K along the c-axis and J a = 0.18 K along the a-and b -axes, while the gyromagnetic factor along the c-axis is g = 2.26 [8]. A magnetic field applied along the c-axis lowers the energy of the S z = 1 state producing a T = 0 BEC transition at H c1 = 2.1 T. The long-range order occurs in a domeshaped region of the T −H phase diagram between H c1 and H c2 = 12.5 T and below the maximum ordering temperature T max 1.2K [12].…”

“…The T 3/2 dependence of the critical field expected for a BEC-QCP has been established via direct measurements of the phase boundary with ac susceptibility down to 1 mK [14], and by magnetization measurements [16]. The asymmetry between H c1 and H c2 [7][8][9][10][11][12] can also be seen directly in the skewed shape of the phase diagram [12].The Hamiltonian describing the S = 1 spin degrees of freedom of DTN in external field is given by [8]where e ν are the primitive vectors of the lattice, ν = {a, b, c}, and h = gµ B H. We introduce Schwinger bosons associated arXiv:1009.0053v2 [cond-mat.str-el] …”

“…In a remarkable contrast to these properties, peaks in the low-temperature thermal conductivity, κ, near H c1 and H c2 do not show any substantial asymmetry. We provide an explanation to this dichotomy by demonstrating that the leading boson-impurity scattering amplitude is also renormalized by quantum fluctuations, effectively canceling mass renormalization effect in κ.DTN is a quantum magnet with tetragonal crystal symmetry that exhibits a field-induced BEC [7,8,[12][13][14]] to a very good approximation [15]. The dominant single-ion uniaxial anisotropy D = 8.9 K splits the Ni S = 1 triplet into an S z = 0 ground state and an S z = ±1 excited doublet.…”

Antiferromagnetism in a Family of S = 1 Square Lattice Coordination Polymers NiX2(pyz)2 (X = Cl, Br, I, NCS; pyz = Pyrazine)

“…Since Haldane 1,2 predicted that an antiferromagnetic Heisenberg chain has a singlet ground state and a finite gap to the lowest excited state for integer spins, this conjecture has inspired numerous studies of S = 1 antiferromagnets in lowdimensions. While most of the work done so far is related to one-dimensional (1D) models or quasi-one-dimensional (Q1D) compounds [3][4][5][6][7][8][9][10][11] , less work has been performed on two-dimensional models (2D) or quasi-two-dimensional (Q2D) compounds [12][13][14][15] partially due to the difficulty of applying theoretical/numerical techniques to these models. In low-dimensional S = 1 antiferromagnets, the nature of the ground state can be strongly modified by the spatial dimensionality as well as the zero-field splitting (ZFS) of Ni(II), 16 both of which can be tuned by chemical synthesis.…”