First-order magnetoelastic transitions usually involve mechanisms unique to each family of materials. For (Mn, Fe) ₂ (P, Si) compounds, it is generally predicted that the unit cell distortion occurring at the ferromagnetic transition leads to a strong electronic reconstruction of the Fe d states accompanied by a notable change in magnetic moment. However, there is no experimental consensus on this mechanism. Here, we use x-ray emission spectroscopy (XES) complemented by first-principles calculations, high-energy resolution fluorescence detected x-ray absorption (HERFD-XAS), and resonant inelastic x-ray scattering (RIXS) experiments, to clarify the nature of the first-order transition in a Mn₀. ₇₄Fe₁. ₂₃P₀. ₇₁Si₀. ₃₂ crystal. HERFD-XAS and RIXS data show a minor evolution of the spectral features in the upper part of the K-edge for Mn and Fe, consistent with the calculated 4p density of states and fingerprinting the transition. In contrast, no significant evolution of the XES spectra is observed when the transition is crossed. In Fe-rich compositions, the calculations indicate that Fe at the 3g site develops a magnetic moment (2. 43 ₁) that is smaller than that of Mn at the 3g site (2. 93 ₁), but larger than that of Fe at the 3f site (1. 48 ₁). Quantitative XES analysis using the IAD method gives a reasonable agreement with the magnetic moments for different (Mn, Fe) ₂ (P, Si) compositions. However, the reduction of the Fe moment predicted by theory (approx. -0. 60. 16em{0ex}₁) is not observed around the transition. This study indicates that the Fe moment collapse at the transition may be weaker than the theoretically predicted value or more gradual in temperature, suggesting a secondary role for this moment instability in the giant magnetocaloric effect of (Mn, Fe) ₂ (P, Si) compounds.
Yibole et al. (Mon,) studied this question.