The Effect of In-Chain Impurities on 1D Antiferromagnets: An NMR Study on Doped Cuprate Spin Chains

The thesis is devoted to the study of in-chain impurities in spin 1/2 antiferromagnetic Heisenberg chains (S=1/2 aHC's)---a model which accompanies the research on magnetism since the early days of quantum theory and which is one of the few integrable spin systems. With respect to impurities it is special insofar as an impurity perturbs the system strongly due to its topology: there is no way around the defect. To what extend the one-dimensional picture stays a good basis for the description of real materials even if the chains are disturbed by in-chain impurities is an interesting question which is addressed in this work. For this purpose, Cu Nuclear Magnetic Resonance (NMR) measurements on the cuprate spin chain compounds SrCuO2 and Sr2CuO3 intentionally doped with nickel (Ni), zinc (Zn) and palladium (Pd) are presented. These materials are well known to be among the best realizations of the S=1/2 aHC model and their large exchange coupling constants allow the investigation of the low-energy dynamics within experimentally easily feasible temperatures. NMR provides the unique ability to study the static and dynamic magnetic properties of the spin chains locally which is important since randomly placed impurities break the translational invariance. Because copper is the magnetically active ion in those materials and the copper nuclear spin is most directly coupled to its electron spin, the NMR measurements have been performed on the copper site. The measurements show in all cases that there are changes in the results of these measurements as compared to the pure compounds which indicate the opening of gaps in the excitation spectra of the spin chains and the emergence of oscillations of the local susceptibility close to the impurities. These experimental observations are compared to theoretical predictions to clarify if and to what extend the already proposed model for these doped systems---the finite spin chain---is suitable to predict the behavior of real materials. Thereby, each impurity shows peculiarities. While Zn and Pd are know to be spin 0 impurities, it is not clear if Ni carries spin 1. To shed some light on this issue is another scope of this work. For Zn impurities, there are indications that they avoid to occupy copper sites, other than in the layered cuprate compounds. Also this matter is considered