Direct comparison of low- and mid-frequency Raman spectroscopy for quantitative solid-state pharmaceutical analysis

[Display omitted] •Low- and mid-frequency Raman spectra were compared for quantitative analysis.•Ternary mixtures of piroxicam solid-state forms β, α2 and monohydrate were used.•The crystal forms could easily be distinguished based on low-frequency spectra.•The low-frequency spectral region was bett...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of pharmaceutical and biomedical analysis 2018-02, Vol.149, p.343-350
Hauptverfasser: Lipiäinen, Tiina, Fraser-Miller, Sara J., Gordon, Keith C., Strachan, Clare J.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:[Display omitted] •Low- and mid-frequency Raman spectra were compared for quantitative analysis.•Ternary mixtures of piroxicam solid-state forms β, α2 and monohydrate were used.•The crystal forms could easily be distinguished based on low-frequency spectra.•The low-frequency spectral region was better for quantitative analysis. This study considers the potential of low-frequency (terahertz) Raman spectroscopy in the quantitative analysis of ternary mixtures of solid-state forms. Direct comparison between low-frequency and mid-frequency spectral regions for quantitative analysis of crystal form mixtures, without confounding sampling and instrumental variations, is reported for the first time. Piroxicam was used as a model drug, and the low-frequency spectra of piroxicam forms β, α2 and monohydrate are presented for the first time. These forms show clear spectral differences in both the low- and mid-frequency regions. Both spectral regions provided quantitative models suitable for predicting the mixture compositions using partial least squares regression (PLSR), but the low-frequency data gave better models, based on lower errors of prediction (2.7, 3.1 and 3.2% root-mean-square errors of prediction [RMSEP] values for the β, α2 and monohydrate forms, respectively) than the mid-frequency data (6.3, 5.4 and 4.8%, for the β, α2 and monohydrate forms, respectively). The better performance of low-frequency Raman analysis was attributed to larger spectral differences between the solid-state forms, combined with a higher signal-to-noise ratio.
ISSN:0731-7085
1873-264X
DOI:10.1016/j.jpba.2017.11.013