Nuclear-Magnetic-Resonance Measurements in the Rare-Earth Group-<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mi mathvariant="normal">V</mml:mi><mml:mi>A</mml:mi></mml:math>Intermetallic Compounds

Abstract

The temperature dependences, signs, and magnitudes of the phosphorus, arsenic, antimony, and bismuth nuclear-magnetic-resonance (NMR) Knight shifts in the NaCl-type rare-earth group-$\mathrm{V}A$ intermetallic compounds have been studied in detail for the temperature range of 1.5-600\ifmmode^\circ\else\textdegree\fi{}K. The Knight shifts in the nonmagnetic compounds, such as LaP, were found to be temperature-independent. For most of the magnetic compounds, such as GdP, the temperature-dependent part of the Knight shift was found to obey a Curie-Weiss law. However, for the praseodymium, samarium, and thulium compounds, the effect of the sixfold cubic crystal-field interaction on the rare-earth ion could not be neglected. An analysis, showing the relationship between the Knight shift and the rare-earth $4f$ paramagnetic susceptibility ${\ensuremath{\chi}}_{f}(T)$, is presented. In this manner, it is shown that detailed information about ${\ensuremath{\chi}}_{f}(T)$ can be derived from an analysis of the NMR data. Furthermore, it is shown that the hyperfine field at the nonmagnetic site in these compounds arises from the spin component S of the total angular momentum J of the rare-earth ion, and not from the magnetization at the rare-earth site. The $s\ensuremath{-}f$ exchange energies $\ensuremath{\Gamma}$ for these compounds were calculated using the uniform conduction-electron spin-polarization model for the Knight shift. For the rare-earth Group-$\mathrm{V}A$ intermetallic compounds, $\ensuremath{\Gamma}$ was found to be negative (antiferromagnetic) and relatively constant (\ensuremath{\simeq} -0.3 eV) for all of these compounds. Comparisons are made for the $s\ensuremath{-}f$ exchange energies for all of the various rare-earth intermetallic compounds for which NMR measurements are available. It is again found that $\ensuremath{\Gamma}$ is negative and relatively constant in magnitude. For all of these compounds, the total hyperfine field per unit spin $S$ is found to be of the order of -50 kOe. No explanation for this phenomenon is advanced. It is shown that a consistent analysis of the Knight-shift data in terms of the Ruderman-Kittel-Kasuya-Yosida theory is probably impossible. The NMR linewidths $\ensuremath{\delta}H$ in the nonmagnetic compounds were found to be in agreement with the calculated dipolar contribution, while for the magnetic compounds, the NMR linewidth $\ensuremath{\delta}H$ was found to be due to the presence of demagnetization fields arising from the nonspherical shape of the particles in the powdered samples.