Nuclear Magnetic Resonance and Relaxation of<mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"><mml:mrow><mml:mmultiscripts><mml:mrow><mml:mi mathvariant="normal">P</mml:mi></mml:mrow><mml:mprescripts/><mml:mrow/><mml:mrow><mml:mn>31</mml:mn></mml:mrow><mml:mrow/><mml:mrow/></mml:mmultiscripts></mml:mrow></mml:math>in the Paramagnetic State of the UP-US Solid Solutions

Measurements by continuous wave and pulsed NMR were made on $^{31}\mathrm{P}$ in the paramagnetic state of the UP-US solid solutions. All the $\mathrm{U}{\mathrm{P}}_{1\ensuremath{-}x}{\mathrm{S}}_{x}$ compositions used have a NaCl-type structure, are good conductors of electricity, and show systematic changes in the paramagnetic Curie temperature $\ensuremath{\theta}$ and in the magnetic ordering, from type-I antiferromagnetic in UP to ferromagnetic in the US-rich compositions. Measurements were made of the Knight shift $K$, the linewidth $\ensuremath{\Delta}H$, the Bloch decay time constant ${{T}_{2}}^{*}$, the echo decay time constant ${T}_{2}$, and the spin-lattice relaxation time ${T}_{1}$ as functions of temperature $T$ and sulfur concentration $x$. $K$ is linear in the molar susceptibility, $K={K}_{0}+(4.0\ifmmode\pm\else\textpm\fi{}0.5){\ensuremath{\chi}}_{M}$, with $|{K}_{0}|\ensuremath{\ll}K$, and with a slope independent of $x$ and similar to the slopes of $K$ versus ${\ensuremath{\chi}}_{M}$ for other U compounds with magnetic ordering. The changes in $\ensuremath{\theta}$ and in magnetic ordering along the $\mathrm{U}{\mathrm{P}}_{1\ensuremath{-}x}{\mathrm{S}}_{x}$ system that could be accounted for by RKKY-type (Rudermann-Kittel-Kasuya-Yosida) interactions are not reflected in $K$, thereby disqualifying the RKKY model for the Knight shift. The broad NMR lines are due to inhomogeneous effects in the powder samples (demagnetization and composition variation), and both $\ensuremath{\Delta}H$ and ${{T}_{2}}^{*}$ are dominated by inhomogeneous magnetic fields. The spin-spin relaxation time ${{T}_{2}}^{\ensuremath{'}}$, deduced from ${T}_{2}$, is due mainly to homogeneous internuclear dipolar interactions. ${T}_{1}$ increases with temperature for all compositions, regardless of the magnetic ordering. ${T}_{1}T$ is empirically related to $K$ differently in the antiferromagnetic compositions (${T}_{1}T{K}^{2}\ensuremath{\cong}\mathrm{const}$) and in the ferromagnetic compositions (${T}_{1}\mathrm{TK}\ensuremath{\cong}\mathrm{const}$). The ${T}_{1}$ results are analyzed in terms of recently proposed models for nuclear relaxation due to indirect interactions with localized magnetic moments via the conduction electrons.