Román-Ospino, Andrés D.

Profile Picture

Filters

20131
Reset filters

Settings

Citations

Publication Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    Publication
    Development of calibration models for the in-line prediction of active pharmaceutical ingredient concentration in the range 0-100% W/W by near infrared and Raman spectroscopy
    (2013) Román-Ospino, Andrés D.; Romañach, Rodolfo J.; College of Arts and Sciences - Sciences; Rivera Portalatin, Nilka; Mendez Román, Rafael; Department of Chemistry; Perales, Oscar
    The development of quantitative calibration models by near infrared (NIR) and Raman spectroscopy with the capability of predicting concentration in a wider range than the currently used approach will be described. Current quantitative calibration models in the pharmaceutical industry have been developed around a small range of the nominal value of concentration of the active principal ingredient (API). NIR and Raman Partial least squares regression (PLSR) calibration models were compared to polynomial PLSR. The target concentration was 10% w/w of API and to challenge the calibration models different concentrations in a wide range of concentration were evaluated (10, 52, 82% w/w of API). Polynomials of grade 0-5 were evaluated by PolyPLS calibration models (NIR & Raman spectroscopy). Precision was calculated for each set through a repeatability study. More precise results were obtained by Raman spectroscopy (comparing the spectroscopic technique). Raman calibration models were one magnitude order more precise than NIR calibration models (SD = 0.032, 1 factor and grade 1 polynomial grade for Raman versus SD = 0.34, 4 Factors and grade 1 polynomial for NIR spectroscopy. The most accurate prediction for the target concentration (10% w/w) in the range 0-100% was obtained by Raman PolyPLS (RMSEP = 1.46) followed by Raman PLS (RMSEP = 2.38) calibration model.