Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:1103.1793 · Physical Review Letters

Spin Gap in the Zigzag Spin- 1/2 Chain Cuprate Sr0.9Ca0.1CuO2

Franziska Hammerath, Satoshi Nishimoto, H.‐J. Grafe, Anja U. B. Wolter, V. Kataev, Patrick Ribeiro, Christian Heß, S.‐L. Drechsler, Bernd Büchner

Published Jun 30, 2011• 33 citations
arXivHTMLPDF

Abstract

We report a comparative study of $^{63}\mathrm{Cu}$ nuclear magnetic resonance spin lattice relaxation rates ${T}_{1}^{\ensuremath{-}1}$ on undoped ${\mathrm{SrCuO}}_{2}$ and Ca-doped ${\mathrm{Sr}}_{0.9}{\mathrm{Ca}}_{0.1}{\mathrm{CuO}}_{2}$ spin chain compounds. A temperature independent ${T}_{1}^{\ensuremath{-}1}$ is observed for ${\mathrm{SrCuO}}_{2}$ as expected for an $S=1/2$ Heisenberg chain. Surprisingly, we observe an exponential decrease of ${T}_{1}^{\ensuremath{-}1}$ for $T<90\text{ }\text{ }\mathrm{K}$ in the Ca-doped sample evidencing the opening of a spin gap. The data analysis within the ${J}_{1}\mathrm{\text{\ensuremath{-}}}{J}_{2}$ Heisenberg model employing density-matrix renormalization group calculations suggests an impurity driven small alternation of the ${J}_{2}$-exchange coupling as a possible cause of the spin gap.

DOI
10.1103/physrevlett.107.017203

Related papers

Spin gap in the single spin- 12 chain cuprate Sr1.9 Ca0.1 CuO3

Franziska Hammerath, E. M. Brüning, Samuele Sanna +7 more

We report $^{63}\mathrm{Cu}$ nuclear magnetic resonance and muon spin rotation measurements on the $S=1/2$ antiferromagnetic Heisenberg spin chain compound ${\mathrm{Sr}}_{1.9}$${\mathrm{Ca}}_{0.1}$${\mathrm{CuO}}_{3}$. An exponentially decreasing spin-lattice relaxation rate ${T}_{1}^{\ensuremath{-}1}$ indicates the opening of a spin gap. This behavior is very similar to what has been observed for the cognate zigzag spin chain compound ${\mathrm{Sr}}_{0.9}$${\mathrm{Ca}}_{0.1}$${\mathrm{CuO}}_{2}$, and it confirms that the occurrence of a spin gap upon Ca doping is independent of the interchain exchange coupling ${J}^{\ensuremath{'}}$. Our results therefore suggest that the appearance of a spin gap in an antiferromagnetic Heisenberg spin chain is induced by a local bond disorder of the intrachain exchange coupling $J$. A low-temperature upturn of ${T}_{1}^{\ensuremath{-}1}$ evidences growing magnetic correlations. However, zero-field muon spin rotation measurements down to 1.5 K confirm the absence of magnetic order in this compound, which is most likely suppressed by the opening of the spin gap.

arXiv:1403.2890•Physical Review B•May 16, 2014•19 citations

Spin diffusion and the anisotropic spin- 12 Heisenberg chain

Jesko Sirker

Measurements of the spin-lattice relaxation rate $1∕{T}_{1}$ by nuclear magnetic resonance for the one-dimensional Heisenberg antiferromagnet ${\mathrm{Sr}}_{2}\mathrm{Cu}{\mathrm{O}}_{3}$ have provided evidence for a diffusionlike contribution at finite temperature and small wave vector. By analyzing real-time data for the auto- and nearest-neighbor spin-spin correlation functions obtained by the density-matrix renormalization group I show that such a contribution indeed exists for temperatures $T>J$, where $J$ is the coupling constant, but that it becomes exponentially suppressed for $T⪡J$. I present evidence that the frequency dependence of $1∕{T}_{1}$ in the Heisenberg case is smoothly connected to that in the free fermion case where the exponential suppression of the diffusionlike contribution is easily understood.

arXiv:cond-mat/0604327•Physical Review B•Jun 16, 2006•34 citations

Spin gap and dynamics in Sr14−xCaxCu24O41 comprising hole-doped two-leg spin ladders: Cu NMR study on single crystals

K. Magishi, Shinji Matsumoto, Yoshitaka KITAOKA +5 more

We report comprehensive Cu NMR studies on single crystals of ${\mathrm{Sr}}_{14\ensuremath{-}x}{\mathrm{Ca}}_{x}{\mathrm{Cu}}_{24}{\mathrm{O}}_{41},$ which contain simple ${\mathrm{CuO}}_{2}$ chains and two-leg ${\mathrm{Cu}}_{2}{\mathrm{O}}_{3}$ ladders. From measurements of the ${}^{63}\mathrm{Cu}$ NMR shift, it is clear that the spin gap in the ladders decreases with isovalent Ca substitution from $\ensuremath{\Delta}=550\ifmmode\pm\else\textpm\fi{}30$ K for ${\mathrm{Sr}}_{14}{\mathrm{Cu}}_{24}{\mathrm{O}}_{41}$ (Sr14) to $350\ifmmode\pm\else\textpm\fi{}30$ K, $280\ifmmode\pm\else\textpm\fi{}30$ K, and $270\ifmmode\pm\else\textpm\fi{}30$ K for $x=6$ (Ca6), $x=9$ (Ca9), and $x=11.5$ (Ca11.5), respectively. The exponential decrease of the nuclear spin-lattice relaxation rate ${1/T}_{1}$ below \ensuremath{\sim}130 K is consistent with the presence of the spin gap in the spin excitation spectrum. In the $T$ range higher than $\ensuremath{\sim}200$ K, we observed the following dependences: ${1/T}_{1}=\mathrm{const}$ and the square of Gaussian spin-echo decay time, ${T}_{2G}^{2}\ensuremath{\propto}T$ which are consistent with the scaling theory for the $S=1/2$ one-dimensional (1D) Heisenberg model. The value of ${T}_{2G}{/T}_{1}\sqrt{T}$ is compatible with the theoretical prediction of an exchange constant along the leg ${J}_{\ensuremath{\parallel}}\ensuremath{\sim}1800$ K for Ca6 and ${J}_{\ensuremath{\parallel}}\ensuremath{\sim}1500$ K for Ca9 and Ca11.5. A notable finding is that the magnitude of the spin gap remains nearly constant and characteristics of novel 1D-like spin dynamics are maintained in the content varying from Ca9 to Ca11.5. On the other hand, the charge transport changes with increasing Ca content so that the more conductive Ca11.5 exhibits pressure-induced superconductivity exceeding 3.5 GPa. We have found that ${T}_{2G}^{2}$, which is proportional to the inverse spin correlation length ${\ensuremath{\xi}}^{\ensuremath{-}1}$, deviates from a linear $T$ dependence upon cooling and is described by $A+BT\mathrm{exp}(\ensuremath{-}\ensuremath{\Delta}/T)$, regardless of the Ca substitution. We point out that the value of ${T}_{2G}^{2}(T=0)=A$ is proportional to the finite value of ${\ensuremath{\xi}}_{0}^{\ensuremath{-}1}=\ensuremath{\Delta}{/c}_{1D},$ where ${c}_{1D}=(\ensuremath{\pi}{/2)J}_{\ensuremath{\parallel}}$ is the spin-wave velocity. From the result that the values of ${A}^{\ensuremath{-}1}\ensuremath{\sim}{\ensuremath{\xi}}_{\mathrm{eff}}$ for Ca6, Ca9, and Ca11.5 are significantly reduced compared to that for Sr14, it is suggested that ${\ensuremath{\xi}}_{\mathrm{eff}}$ is dominated at low $T$ by an average distance $d$ among mobile holes obeying the relation ${\ensuremath{\xi}}_{\mathrm{eff}}\ensuremath{\sim}d={\ensuremath{\xi}}_{h}.$ From an estimate of ${\ensuremath{\xi}}_{0}/a\ensuremath{\sim}$ 5.2 for Sr14, where $a$ is the Cu-Cu distance along the leg, ${\ensuremath{\xi}}_{h}/a$ is obtained as $\ensuremath{\sim}3.5$, 2.3, and 2.0, and hole content $x$ as $\ensuremath{\sim}0.14$, 0.22, and 0.25 per ${\mathrm{Cu}}_{2}{\mathrm{O}}_{3}$ ladder for Ca6, Ca9, and Ca11.5, respectively. These values were consistent with $x=$ 0.14, 0.2, and 0.22 for Ca6, Ca9, and Ca11 estimated from the optical conductivity experiment by Osafune et al. [Phys. Rev. Lett. 78, 1980 (1997)]. The ${\mathrm{Sr}}_{14\ensuremath{-}x}{\mathrm{Ca}}_{x}{\mathrm{Cu}}_{24}{\mathrm{O}}_{41}$ compounds are thus hole-doped two-leg spin-ladder systems which reveal the metallic behavior dominated by the 1D-like spin dynamics at high $T$ and accompanied by the spin gap formation at low $T$. For Ca11.5, as the spin gap is formed upon cooling below $\ensuremath{\sim}180$ K, the resistivity increases in the direction perpendicular to the ladder, whereas the conductivity along the ladder remains metallic, followed by the localization of mobile holes in both directions below ${T}_{L}\ensuremath{\sim}60$ K. We point out that preformed pairs are confined in each ladder and localized below $\ensuremath{\sim}60$ K at an ambient pressure.

Physical review. B, Condensed matter•May 1, 1998•135 citations

Consequences of lattice imperfections and interchain couplings for the critical properties of spin-1/2 chain compounds

Sirker

To allow for a comparison of theoretical predictions for spin chains with experimental data, it is often important to take impurity effects as well as interchain couplings into account.Here we present the field theory for finite spin chains at finite temperature and calculate experimentally measurable quantities like susceptibilities and nuclear magnetic resonance spectra.For the interchain couplings we concentrate on geometries relevant for cuprate spin chains like Sr 2 CuO 3 and SrCuO 2 .The field theoretical results are compared to experimental as well as numerical data obtained by the density matrix renormalization group.

arXiv:0905.1075•Condensed Matter Physics•Jan 1, 2009•2 citations

Nuclear relaxation in the spin- 1/2 antiferromagnetic chain compound Sr2CuO3: Comparison between theories and experiments

Masashi Takigawa, Oleg A. Starykh, Anders W. Sandvik +1 more

The NMR relaxation data on ${\mathrm{Sr}}_{2}{\mathrm{CuO}}_{3}$ [Phys. Rev. Lett. 76, 4612 (1996)] are reexamined and compared with the analytic theory of the dynamic susceptibility in the $S=1/2$ antiferromagnetic Heisenberg chain including multiplicative logarithmic corrections [Phys. Rev. Lett. 78, 539 (1997)]. Comparisons of the spin-lattice and the Gaussian spin-echo decay rates (${1/T}_{1}$ and ${1/T}_{2G}$) and their ratio all show good quantitative agreement. Our results demonstrate the importance of the logarithmic corrections in the analysis of experimental data for quasi-one-dimensional systems and indicate that the dynamics of ${\mathrm{Sr}}_{2}{\mathrm{CuO}}_{3}$ is well described by an $S=1/2$ one-dimensional Heisenberg model with a nearest-neighbor exchange.

arXiv:cond-mat/9706177•Physical review. B, Condensed matter•Dec 1, 1997•66 citations

NMR relaxation in the spin-1 Heisenberg chain

Sylvain Capponi, Maxime Dupont, Anders W. Sandvik +1 more

We consider the isotropic $S=1$ Heisenberg chain with a finite Haldane gap $\mathrm{\ensuremath{\Delta}}$ and use state-of-the-art numerical techniques to investigate its dynamical properties at finite temperature, focusing on the nuclear spin-lattice relaxation rate $1/{T}_{1}$ measured in nuclear magnetic resonance (NMR) experiments, for instance. In particular, we analyze the contributions from modes with momenta close to $q\ensuremath{\approx}0$ and $q\ensuremath{\approx}\ensuremath{\pi}$ as a function of temperature. At high-temperature we observe spin diffusion, while at low-temperature we argue that a simple activated behavior $1/{T}_{1}\ensuremath{\propto}exp(\ensuremath{-}\mathrm{\ensuremath{\Delta}}/T)$ can be observed only at temperatures much smaller than the gap $\mathrm{\ensuremath{\Delta}}$.

arXiv:1905.12697•Physical review. B./Physical review. B•Sep 9, 2019•18 citations
We report a comparative study of 63Cu^{63}\mathrm{Cu}63Cu nuclear magnetic resonance spin lattice relaxation rates T1\ensuremath−1{T}_{1}^{\ensuremath{-}1}T1\ensuremath−1​ on undoped SrCuO2{\mathrm{SrCuO}}_{2}SrCuO2​ and Ca-doped Sr0.9Ca0.1CuO2{\mathrm{Sr}}_{0.9}{\mathrm{Ca}}_{0.1}{\mathrm{CuO}}_{2}Sr0.9​Ca0.1​CuO2​ spin chain compounds. A temperature independent T1\ensuremath−1{T}_{1}^{\ensuremath{-}1}T1\ensuremath−1​ is observed for SrCuO2{\mathrm{SrCuO}}_{2}SrCuO2​ as expected for an S=1/2S=1/2S=1/2 Heisenberg chain. Surprisingly, we observe an exponential decrease of T1\ensuremath−1{T}_{1}^{\ensuremath{-}1}T1\ensuremath−1​ for T<90  KT<90\text{ }\text{ }\mathrm{K}T<90  K in the Ca-doped sample evidencing the opening of a spin gap. The data analysis within the J1\ensuremath-J2{J}_{1}\mathrm{\text{\ensuremath{-}}}{J}_{2}J1​\ensuremath-J2​ Heisenberg model employing density-matrix renormalization group calculations suggests an impurity driven small alternation of the J2{J}_{2}J2​-exchange coupling as a possible cause of the spin gap.
We report 63Cu^{63}\mathrm{Cu}63Cu nuclear magnetic resonance and muon spin rotation measurements on the S=1/2S=1/2S=1/2 antiferromagnetic Heisenberg spin chain compound Sr1.9{\mathrm{Sr}}_{1.9}Sr1.9​Ca0.1{\mathrm{Ca}}_{0.1}Ca0.1​CuO3{\mathrm{CuO}}_{3}CuO3​. An exponentially decreasing spin-lattice relaxation rate ${T}_{1}^{\ensuremath{-}1}$ indicates the opening of a spin gap. This behavior is very similar to what has been observed for the cognate zigzag spin chain compound Sr0.9{\mathrm{Sr}}_{0.9}Sr0.9​Ca0.1{\mathrm{Ca}}_{0.1}Ca0.1​CuO2{\mathrm{CuO}}_{2}CuO2​, and it confirms that the occurrence of a spin gap upon Ca doping is independent of the interchain exchange coupling ${J}^{\ensuremath{'}}$. Our results therefore suggest that the appearance of a spin gap in an antiferromagnetic Heisenberg spin chain is induced by a local bond disorder of the intrachain exchange coupling JJJ. A low-temperature upturn of ${T}_{1}^{\ensuremath{-}1}$ evidences growing magnetic correlations. However, zero-field muon spin rotation measurements down to 1.5 K confirm the absence of magnetic order in this compound, which is most likely suppressed by the opening of the spin gap.
Measurements of the spin-lattice relaxation rate 1∕T11∕{T}_{1}1∕T1​ by nuclear magnetic resonance for the one-dimensional Heisenberg antiferromagnet Sr2CuO3{\mathrm{Sr}}_{2}\mathrm{Cu}{\mathrm{O}}_{3}Sr2​CuO3​ have provided evidence for a diffusionlike contribution at finite temperature and small wave vector. By analyzing real-time data for the auto- and nearest-neighbor spin-spin correlation functions obtained by the density-matrix renormalization group I show that such a contribution indeed exists for temperatures T>JT>JT>J, where JJJ is the coupling constant, but that it becomes exponentially suppressed for T⪡JT⪡JT⪡J. I present evidence that the frequency dependence of 1∕T11∕{T}_{1}1∕T1​ in the Heisenberg case is smoothly connected to that in the free fermion case where the exponential suppression of the diffusionlike contribution is easily understood.
We report comprehensive Cu NMR studies on single crystals of ${\mathrm{Sr}}_{14\ensuremath{-}x}{\mathrm{Ca}}_{x}{\mathrm{Cu}}_{24}{\mathrm{O}}_{41},$ which contain simple CuO2{\mathrm{CuO}}_{2}CuO2​ chains and two-leg Cu2O3{\mathrm{Cu}}_{2}{\mathrm{O}}_{3}Cu2​O3​ ladders. From measurements of the 63Cu{}^{63}\mathrm{Cu}63Cu NMR shift, it is clear that the spin gap in the ladders decreases with isovalent Ca substitution from $\ensuremath{\Delta}=550\ifmmode\pm\else\textpm\fi{}30$ K for Sr14Cu24O41{\mathrm{Sr}}_{14}{\mathrm{Cu}}_{24}{\mathrm{O}}_{41}Sr14​Cu24​O41​ (Sr14) to $350\ifmmode\pm\else\textpm\fi{}30$ K, $280\ifmmode\pm\else\textpm\fi{}30$ K, and $270\ifmmode\pm\else\textpm\fi{}30$ K for x=6x=6x=6 (Ca6), x=9x=9x=9 (Ca9), and x=11.5x=11.5x=11.5 (Ca11.5), respectively. The exponential decrease of the nuclear spin-lattice relaxation rate 1/T1{1/T}_{1}1/T1​ below \ensuremath{\sim}130 K is consistent with the presence of the spin gap in the spin excitation spectrum. In the TTT range higher than $\ensuremath{\sim}200$ K, we observed the following dependences: 1/T1=const{1/T}_{1}=\mathrm{const}1/T1​=const and the square of Gaussian spin-echo decay time, ${T}_{2G}^{2}\ensuremath{\propto}T$ which are consistent with the scaling theory for the S=1/2S=1/2S=1/2 one-dimensional (1D) Heisenberg model. The value of T2G/T1T{T}_{2G}{/T}_{1}\sqrt{T}T2G​/T1​T​ is compatible with the theoretical prediction of an exchange constant along the leg ${J}_{\ensuremath{\parallel}}\ensuremath{\sim}1800$ K for Ca6 and ${J}_{\ensuremath{\parallel}}\ensuremath{\sim}1500$ K for Ca9 and Ca11.5. A notable finding is that the magnitude of the spin gap remains nearly constant and characteristics of novel 1D-like spin dynamics are maintained in the content varying from Ca9 to Ca11.5. On the other hand, the charge transport changes with increasing Ca content so that the more conductive Ca11.5 exhibits pressure-induced superconductivity exceeding 3.5 GPa. We have found that T2G2{T}_{2G}^{2}T2G2​, which is proportional to the inverse spin correlation length ${\ensuremath{\xi}}^{\ensuremath{-}1}$, deviates from a linear TTT dependence upon cooling and is described by $A+BT\mathrm{exp}(\ensuremath{-}\ensuremath{\Delta}/T)$, regardless of the Ca substitution. We point out that the value of T2G2(T=0)=A{T}_{2G}^{2}(T=0)=AT2G2​(T=0)=A is proportional to the finite value of ${\ensuremath{\xi}}_{0}^{\ensuremath{-}1}=\ensuremath{\Delta}{/c}_{1D},$ where ${c}_{1D}=(\ensuremath{\pi}{/2)J}_{\ensuremath{\parallel}}$ is the spin-wave velocity. From the result that the values of ${A}^{\ensuremath{-}1}\ensuremath{\sim}{\ensuremath{\xi}}_{\mathrm{eff}}$ for Ca6, Ca9, and Ca11.5 are significantly reduced compared to that for Sr14, it is suggested that ${\ensuremath{\xi}}_{\mathrm{eff}}$ is dominated at low TTT by an average distance ddd among mobile holes obeying the relation ${\ensuremath{\xi}}_{\mathrm{eff}}\ensuremath{\sim}d={\ensuremath{\xi}}_{h}.$ From an estimate of ${\ensuremath{\xi}}_{0}/a\ensuremath{\sim}$ 5.2 for Sr14, where aaa is the Cu-Cu distance along the leg, ${\ensuremath{\xi}}_{h}/a$ is obtained as $\ensuremath{\sim}3.5$, 2.3, and 2.0, and hole content xxx as $\ensuremath{\sim}0.14$, 0.22, and 0.25 per Cu2O3{\mathrm{Cu}}_{2}{\mathrm{O}}_{3}Cu2​O3​ ladder for Ca6, Ca9, and Ca11.5, respectively. These values were consistent with x=x=x= 0.14, 0.2, and 0.22 for Ca6, Ca9, and Ca11 estimated from the optical conductivity experiment by Osafune et al. [Phys. Rev. Lett. 78, 1980 (1997)]. The ${\mathrm{Sr}}_{14\ensuremath{-}x}{\mathrm{Ca}}_{x}{\mathrm{Cu}}_{24}{\mathrm{O}}_{41}$ compounds are thus hole-doped two-leg spin-ladder systems which reveal the metallic behavior dominated by the 1D-like spin dynamics at high TTT and accompanied by the spin gap formation at low TTT. For Ca11.5, as the spin gap is formed upon cooling below $\ensuremath{\sim}180$ K, the resistivity increases in the direction perpendicular to the ladder, whereas the conductivity along the ladder remains metallic, followed by the localization of mobile holes in both directions below ${T}_{L}\ensuremath{\sim}60$ K. We point out that preformed pairs are confined in each ladder and localized below $\ensuremath{\sim}60$ K at an ambient pressure.
To allow for a comparison of theoretical predictions for spin chains with experimental data, it is often important to take impurity effects as well as interchain couplings into account.Here we present the field theory for finite spin chains at finite temperature and calculate experimentally measurable quantities like susceptibilities and nuclear magnetic resonance spectra.For the interchain couplings we concentrate on geometries relevant for cuprate spin chains like Sr 2 CuO 3 and SrCuO 2 .The field theoretical results are compared to experimental as well as numerical data obtained by the density matrix renormalization group.
The NMR relaxation data on Sr2CuO3{\mathrm{Sr}}_{2}{\mathrm{CuO}}_{3}Sr2​CuO3​ [Phys. Rev. Lett. 76, 4612 (1996)] are reexamined and compared with the analytic theory of the dynamic susceptibility in the S=1/2S=1/2S=1/2 antiferromagnetic Heisenberg chain including multiplicative logarithmic corrections [Phys. Rev. Lett. 78, 539 (1997)]. Comparisons of the spin-lattice and the Gaussian spin-echo decay rates (1/T1{1/T}_{1}1/T1​ and 1/T2G{1/T}_{2G}1/T2G​) and their ratio all show good quantitative agreement. Our results demonstrate the importance of the logarithmic corrections in the analysis of experimental data for quasi-one-dimensional systems and indicate that the dynamics of Sr2CuO3{\mathrm{Sr}}_{2}{\mathrm{CuO}}_{3}Sr2​CuO3​ is well described by an S=1/2S=1/2S=1/2 one-dimensional Heisenberg model with a nearest-neighbor exchange.
We consider the isotropic S=1S=1S=1 Heisenberg chain with a finite Haldane gap $\mathrm{\ensuremath{\Delta}}$ and use state-of-the-art numerical techniques to investigate its dynamical properties at finite temperature, focusing on the nuclear spin-lattice relaxation rate 1/T11/{T}_{1}1/T1​ measured in nuclear magnetic resonance (NMR) experiments, for instance. In particular, we analyze the contributions from modes with momenta close to $q\ensuremath{\approx}0$ and $q\ensuremath{\approx}\ensuremath{\pi}$ as a function of temperature. At high-temperature we observe spin diffusion, while at low-temperature we argue that a simple activated behavior $1/{T}_{1}\ensuremath{\propto}exp(\ensuremath{-}\mathrm{\ensuremath{\Delta}}/T)$ can be observed only at temperatures much smaller than the gap $\mathrm{\ensuremath{\Delta}}$.