Discrepancies in NMRON knight shifts of impurity nuclei in ferromagnets

Precision NMRON field shift studies have been carried out up to 8 T on a 54MnNi single crystal along a hard [100] direction and on a 125SbFe single crystal along a hard [110] direction. For both systems, high field (Bapp≥1 T) and low field (0.3 T<Bapp<0.8 T) data sets are obtained. The analysis reveals a significant discrepancy between the apparent Knight Shift dependent on whether the low field or high field data set is utilised. For both systems, consideration of the high field data sets yield a zero Knight Shift K(54MnNi)=+0.0(0.2)% and K(125SbFe)=+0.2(1.4)%, whereas the low field data sets yield K(54MnNi)=+7.5(3.9)% and K(125SbFe)=-5.4(3.3)% respectively. The field range dependence of K suggests that only Knight shifts measured in large fields (≥1 T) are meaningful for establishing systematics. This casts some doubt on the greater bulk of the literature's NMRON Knight shift studies, where predominantly low fields have been used to determine K. (orig.)