NIR Chemical Imaging as a Tool for Understanding blending of Pharmaceutical Materials and Optimization of Blending Process Control


Hua Ma

Defense Date


Graduation Date

Fall 1-1-2008


Campus Only

Submission Type


Degree Name





School of Pharmacy

Committee Chair

James K. Drennen, Carl A. Anderson

Committee Member

Peter Wildfong

Committee Member

Mitch Johnson


NIR, NIR chemical imaging, powder blending, blending mechanisms


NIR chemical imaging as a valid tool for powder blend characterization was established after comparing blend end-point determination via NIR imaging technique with that obtained using conventional UV and single-point NIR spectroscopy (NIRS) methods in the exploratory phase of the dissertation.

An optimal objective magnification level, a critical parameter in evaluating pharmaceutical solid dosage forms using NIR imaging, was investigated next. Three magnification levels of NIR chemical imaging objectives were evaluated for their effects on imaging a model binary pharmaceutical powder mixture (salicylic acid and lactose). High, medium and low objective magnification levels were investigated by comparing the resulting blend surface images and spectra extracted from respective images. It was concluded that a change to the magnification level did not impose variations in the characteristics of the image or spectral data from the blend surface; and the primary factors of obtaining maximum information while minimizing data collection time can be balanced by using low magnification objective for the system being studied.

Powder blend characterizations via NIR chemical imaging using the determined optimal magnification level were carried out in a small-scale powder blending study using a model ternary pharmaceutical powder system consisting of acetaminophen (APAP, the model API), microcrystalline cellulose (MCC) and lactose monohydrate; and characterization of the powder blending process (of the model mixture) was achieved in both radial and axial mixing directions. In either direction, mixtures were blended for different time periods for a total of ten time points (ten blending trials). After each blend was compressed into a compact, images were collected from multiple locations of each compact for a detailed characterization. By employing multivariate qualitative and quantitative image analyses and chemometric tools, blending mechanisms in radial and axial mixing directions were delineated; blend end-point and blending efficiency along the two directions determined from NIR images were compared with those obtained by conventional UV method; optimal sampling locations were identified; and appropriate beam size for NIR sensors in powder blending process monitoring was determined.

Blending process understanding gained from this dissertation research is valuable for future blending process control optimization and blending method development and validation.





This document is currently not available here.