Local disorder in the oxygen environment around praseodymium in Y1−x Prx Ba2 Cu3 O7 from x-ray-absorption fine structure

${\mathrm{PrBa}}_{2}$${\mathrm{Cu}}_{3}$${\mathrm{O}}_{7}$ (PBCO) is unique in the R${\mathrm{Ba}}_{2}$${\mathrm{Cu}}_{3}$${\mathrm{O}}_{7}$ (R=rare earth) series because it is not superconducting. In fact, for ${\mathrm{Y}}_{1\mathrm{\ensuremath{-}}\mathit{x}}$${\mathrm{Pr}}_{\mathit{x}}$${\mathrm{Ba}}_{2}$${\mathrm{Cu}}_{3}$${\mathrm{O}}_{7}$, ${\mathit{T}}_{\mathit{c}}$ drops monotonically with Pr concentration, with ${\mathit{T}}_{\mathit{c}}$ going to zero at \ensuremath{\sim}55% Pr. There have been many studies of this material with the hope that an explanation for the lack of superconductivity in PBCO might help explain why ${\mathrm{YBa}}_{2}$${\mathrm{CU}}_{3}$${\mathrm{O}}_{7}$ (YBCO) is superconducting. Explanations center around hole localization, requiring an extra hole on the Y (Pr) site or a localized hole in the O 2p shell, at the expense of a mobile hole in the Cu-O planes. To help provide clues that could point to a particular model, and to search for anomalies in the local structure, we present K-edge x-ray-absorption fine structure (XAFS) data for various concentrations of Pr in ${\mathrm{Y}}_{1\mathrm{\ensuremath{-}}\mathit{x}}$${\mathrm{Pr}}_{\mathit{x}}$${\mathrm{Ba}}_{2}$${\mathrm{Cu}}_{3}$${\mathrm{O}}_{7}$. The character of the Pr K-edge XAFS data indicates that most of the Pr substitutes onto the Y site and is well ordered with respect to the unit cell. These data also show that the amplitude of the first Pr-O peak is greatly reduced when compared to the first Y-O peak in pure YBCO, and decreases with increasing Pr concentration.In contrast, the Y K-edge data for these alloys show little if any change in the oxygen environment, while the Cu K-edge data show a 10% reduction in the first Cu-O peak. Fits to the Pr data suggest that some oxygen atoms about the Pr become disordered and/or distorted; most of the Pr-O nearest-neighbor distances are 2.45 \AA{}, but about 15--40 % of them are in a possibly broadened peak at 2.${27}_{\mathrm{\ensuremath{-}}0.12}^{+0.03}$ \AA{}. The Cu K-edge XAFS data show a slight broadening but no loss of oxygens, which is consistent with a radial distortion of the Pr-O bond. The existence and the size of these two bond lengths is consistent with a mixture of ${\mathrm{Pr}}^{3+}$ and ${\mathrm{Pr}}^{4+}$ bonds, and to a formal valence of +3.${33}_{\mathrm{\ensuremath{-}}0.18}^{+0.07}$ for the Pr ion.