Crystal fields and exchange parameters in LaAl2:Eu2+and LaAl2:Gd3+single crystals

Reported are 10- and 35-GHz ESR measurements on single crystals of ${\mathrm{Gd}}_x{\mathrm{La}}_{1-x}{\mathrm{Al}}_2$ and ${\mathrm{Eu}}_x{\mathrm{La}}_{1-x}{\mathrm{Al}}_2$, $x=100\mathrm{}-1000$ ppm. For the Gd alloys no effects of fine structure could be detected, which, under the prevalent experimental conditions implies $b_4<10\mathrm{}$ G. In contrast the Eu alloys exhibit a large, $b_4=+112\mathrm{}$ G, $b_6=-8\mathrm{}$ G, crystal field, the largest in a metal to date and corresponding to an overall crystal-field splitting of 480 mK. This system is the first to show resolved fine structure and a bottleneck. A computer analysis using the full Barnes-Plefka theory shows excellent agreement between theory and experiment. As is usually the case, the exchange obtained from the $g$ shift does not agree with that from the linewidth. We show that this cannot be explained in terms of the current "spherically symmetric partial-wave-expansion" approach but is reconcilable with the crystal-field split-band theory of Narath. Finally, we show that, including the ferromagnetic Kondo effect, our $g$-shift exchange parameter is consistent with the theoretical value of Harmon and Freeman for ${\mathrm{Gd}}^{3+}$ scaled to ${\mathrm{Eu}}^{2+}$.