Monitoring powder blending by NIR spectroscopy

Critical evaluation of methods for end-point determination in pharmaceutical blending processes

“…Another algorithm used calculates the difference between the moving block average as compared to a spectrum considered to represent homogeneity (the target spectrum) (DMBA-TS); 1,5,6 to this end, it calculates the difference in the average of a block of 5 consecutive spectra recorded during the monitoring process that is moved one spectrum at a time over the spectral range against a target spectrum (viz. the spectrum for a mixture assumed to be homogeneous and containing the nominal concentration of API).…”

“…This method provides a measure of dissimilarity between an individual spectrum and an average set of scans for a mixture assumed to have reached homogeneity (averaged target spectra) (DIS-ATS). 1,5,6 We also estimated the dissimilarity between consecutive spectra (DIS) 8 by calculating the difference between two consecutive spectra at each wavelength and then the standard deviation of the differences at all wavelengths.…”

“…Qualitative methods (e.g. those based on intensity changes in the NIR band for one component, 27,42 dissimilarity 1,5,6,8 or the Euclidean distance between spectra 6 ) estimate the difference between two spectra recorded during the process. Worth special Based on reduction of variables where the matrix spectral decomposition is performed simultaneously with the matrix of the concentration of component to determine 9,10,12,14,15,16,20,21,23,24,32,33,34,35,36,37,38,39,40,43 note in this category are the methods based on moving block standard deviation (MBSD), which usually rely on calculating the standard deviation for consecutive blocks of spectra moving over time: [3][4][5]7,16,18,[33][34][35] either the average standard deviation of difference of spectra 6,9 between PC scores 35 or a statistic such as chi-square.…”

Polarization Effects of Confocal Raman Microspectrometry of Crystal Powders Using Interactive Self-Modeling Analysis

“…Simplisma has been described for a variety of applications including: (i) Raman spectra of a time-resolved reaction of tetramethyl orthosilicate, , hydrogen peroxide activation by nitriles, photooxidation products of lead sulfide crystals, imaging of dust particles emitted by smelters, , and aromatic compounds occluded in zeolites; − (ii) FTIR microscopy of a polymer laminate; ,− (iii) pyrolysis mass spectral data of plant materials; (iv) time-resolved mass spectral data of photographic color-coupling compounds; , (v) infrared spectra of mixtures of solvents, , polymers, , monitoring of powder blending, and water; (vi) diffuse reflectance UV−vis spectra of zeolites, , metal oxide catalysts, and polynucleotides; (vii) time-resolved absorption spectra of short-lived transient species of aromatic compounds; (viii) fluorescence spectra of fulvic acids, − leaf litter, and natural organic matter from water; (ix) ion mobility spectra; , and (x) X-ray photoelectron spectra of blends of poly(vinyl chloride) and poly(methyl methacrylate) …”

Characterization of granular mixing in vertical axis bladed mixers using X-ray imaging systems

“…Direct observation of the IR-spectra lends to the qualitative measurement of mixture homogeneity while statistical analysis of the variations in the spectra over time provides objective estimation of mixing end points (i.e., points at which maximum or acceptable levels of mixture homogeneity have been achieved) [108]. For homogeneity measurements, the difference in the obtained spectra and the spectra from an "ideally mixed state" can be compared [117,118] whereas the standard deviation of the moving block standard deviation of the collected spectra can also be calculated to quantify mixture homogeneity and identify mixing end points [104,105].…”