Preparation and selection of samples

The NIR spectrum depends not only on the chemical composition of the sample, but also on some physical properties such as the size, form, particle distribution and degree of compression of the sample. This is useful in some cases as it enables the spectroscopic determination of some physical parameters of the sample. However, physical differences can lead to multiplicative effects in the spectrum, which, together with other additive effects such as baseline shift or chemical absorption, can complicate calibration models and detract from the quality of the results of quantitative analyses if not properly understood and accounted for. 

These effects can be modeled by ensuring that the calibration set encompasses aU possible sources of variability in the samples, which in turn can be ensured by using samples from an adequate number of production batches.  

Different production batches of the same pharmaceutical preparation can have very similar concentrations of API (+5% around the nominal value) and excipients close to the nominal value. It is difficult to obtain a set of production samples of the target product spanning the concentration range required to develop a calibration model that will allow the pharmacopoeial recommendations to be met and different enough values ( 20% around the nominal one) of potency, content uniformity, etc., accommodated. 

Incorporating all sources of variability in the production process with a known or presumed influence on the spectra for the samples entails expanding the calibration set with new samples obtained from the pilot plant or prepared in the laboratory. Preparing these additional samples can be more or less complicated depending on their physical state because no unique methodology exists for this purpose, this section describes the most widely used and efficient choices. 

The samples used to construct the models should be similar to the production samples also, their spectra should be recorded in the same mode (reflectance, transmission or transflectance) as those for the samples to be subsequently predicted, and include all potential sources of variability. Although such sources are relatively limited - pharmaceutical samples usually have a well-defined qualitative and quantitative composition from raw material to end product, and production processes are solidly established and reproducible - their 

---

As far as is practically possible, the selection and preparation of samples must take into account all possible variations in the matrix of the material to be analysed. The applicability of the method should be studied using various samples ranging from pure standards to mixtures with complex matrices as these may contain substances that interfere to a greater or lesser extent with the quantitative determination of an analyte or the accurate measurement of a parameter. Matrix effects can both reduce and enhance analytical signals and may also act as a barrier to recovery of the analyte from a sample. 

---