The infrared spectrum of ortho-enstatite from reflectance experiments and first-principle simulations

Infrared (IR) spectra provide a rich amount of information concerning chemical composition, lattice structure, size and shape of circumstellar dust. Accurate reference data are then required for the analysis of the various detected components. This study provides the IR characterization of one of th...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Monthly notices of the Royal Astronomical Society 2012-02, Vol.420 (1), p.147-154
Hauptverfasser: Demichelis, Raffaella, Suto, Hiroshi, Noël, Yves, Sogawa, Hisato, Naoi, Takahiro, Koike, Chiyoe, Chihara, Hiroki, Shimobayashi, Norimasa, Ferrabone, Matteo, Dovesi, Roberto
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Infrared (IR) spectra provide a rich amount of information concerning chemical composition, lattice structure, size and shape of circumstellar dust. Accurate reference data are then required for the analysis of the various detected components. This study provides the IR characterization of one of the most frequently observed compounds, MgSiO3 ortho-enstatite, and shows that IR experiments and ab initio techniques can be used synergically to obtain high-quality data concerning crystalline materials. The IR reflectance spectrum of synthetic ortho-enstatite was collected and compared to ab initio results (Gaussian-type basis set, PBE0 hybrid density functional theory (DFT) functional). An excellent agreement is observed both for vibrational frequencies (ν) and intensities, the latter estimated through the oscillator strength. The mean absolute difference between experimental and calculated ν is of the order of 7-8 cm−1 (43 out of 65 peaks differ by less than 10 cm−1, only four peaks differ by 15-19 cm−1). The static dielectric tensor and its components (electronic and ionic contributions) were measured and compared to calculated data: differences are in the 2-5 per cent range.
ISSN:0035-8711
1365-2966
DOI:10.1111/j.1365-2966.2011.20018.x