Weak anharmonic effects inMgB2: A comparative inelastic x-ray scattering and Raman study

Abstract

We study anharmonic effects in MgB2 by comparing Inelastic X-ray and Raman scattering together with ab-initio calculations. Using high statistics and high q resolution measurements we show that the E2g mode linewidth is independent of temperature along ΓA. We show, contrary to previous claims, that the Raman-peak energy decreases as a function of increasing temperature, a behaviour inconsistent with all the anharmonic ab-initio calculations of the E2g mode at Γ available in literature. These findings and the excellent agreement between the X-ray measured and ab initio calculated phonon spectra suggest that anharmonicity is not the main mechanism determining the temperature behaviour of the Raman-peak energy. The Raman E2g peak position and linewidth can be explained by large dynamical effects in the phonon self-energy. In light of the present findings, the commonly accepted explanation of the reduced isotope effect in terms of anharmonic effects needs to be reconsidered.

Vibrational thermodynamics of materials

“…Early work indicated that superconductivity in MgB 2 [389] originated with a strong EPI caused by anharmonic E 2g phonons. Later this interpretation has been called into question, and it has been argued that MgB 2 is not particularly anharmonic [390]. Furthermore, the adiabatic approximation itself has been questioned because the electrons may not respond instantly to nuclear displacements [390].…”

Raman study on the effects of sintering temperature on the Jc(H) performance of MgB2 superconductor

“…26 This is consistent with the later harmonic phonon dispersion results 27,28 obtained from inelastic x-ray scattering. 27,29,30 As for the MgB 2 synthesized at the different temperatures, there are a number of defects both in and between the crystals. Eisterer et al have researched the influence of disorder on the superconducting properties of MgB 2 wires and extracted the mean free path of the charge carriers in the band from the Goŕkov-Goodman relation.…”

Ab initiocalculation of low-energy collective charge-density excitations inMgB2

“…However, despite the tremendous progress achieved in understanding of superconductivity in MgB 2 , many of its fundamental properties related to phonons and electron-phonon (e − ph) coupling are still puzzling and remain under intense debate. Thus such important issues as, e.g., the reduced isotope effect, 2,3,4 the nature of the strong broad continuum in Raman spectra, the unusually broad linewidth of the E 2g phonon mode seen in Raman spectra, 5,6,7 the inconsistency between the results of Raman scattering and x-ray measurements, 8,9,10 and the role of non-adiabatic effects 8,11,12,13,14 have not yet found satisfactory explanations.As for the electronic part of the e−ph picture of superconductivity in MgB 2 , it has received substantially less attention than the phonon part. In particular, in all evaluations of the e−ph interaction the adiabatic approximation has been used, and even in ab initio calculations 15 of T c the dynamical Coulomb interaction has been considered in its static form.…”

“…4 we show the calculated dispersion of the bare σπ mode along with the bare E 2g phonon mode 37 compared to the experimental data. 8,13 One can see how the two bare curves cross each other close to the Γ point. Due to its localization in the boron plane the σπ plasmon strongly interacts with the boron E 2g phonon mode resulting in strong hybridization of these boson modes.…”

“…The σπ plasmon can dramatically affect the dispersion of the optical phonon modes in the "pathological" region of small q, 13 offering an unexpected explanation of widely-discussed discrepancies between the Raman and x-ray measurements of the phonon structure in MgB 2 . The ∼ 77 meV peak observed in Raman-scattering experiments and commonly attributed to the boron E 2g phonon mode is always strongly renormalized in comparison with a bare phonon dispersion, whereas the x-ray experiments performed for large q c do not see any appreciable renormalization.…”