The QualitySpec Pro NIR spectrophotometer scanning the spectral region 350-2500nm has been shown capable of providing an accurate measure of the moisture content in corn starch food products
The QualitySpec Pro NIR spectrophotometer scanning the spectral region 350-2500nm was the instrument used to demonstrate the feasibility of using NIR technology as an accurate means of predicting the moisture content in corn starch food products. It was coupled to a bi-furcated reflectance probe via a fibre optic cable and industry standard SMA 905 connectors. 70 different starch samples were obtained, and three spectral reflectance measurements were collected on each sample.

A spectrum averaging of 25 was used to optimise the signal to noise ratio: however, since the QualitySpec Pro has a 100 millisecond scan time, each spectral acquisition was completed in just 2.5 seconds.

For a background subtraction, a reference spectrum was collected using a Spectralon reflectance standard disc.

Once spectral data collection was completed for the samples, they were then analysed for moisture content via the oven drying and weighing method.

Primary oven dry results were then correlated to the relevant NIR spectra already measured on the QualitySpec Pro.

Grams/AI chemometric software package with PLS/PlusIQ was used for the multivariate analysis and calibration model construction.

The correlation coefficient (R2) was calculated at 0.958 and the standard error of cross validation (SECV) was calculated to be only 0.081% at 6 recommended factors or principal components.

Hence it was concluded that the technology employed by the QualitySpec Pro NIR analyser is ideal for making accurate predictions of the moisture content in corn starch.

Results are obtained in just seconds from instrumentation which can be placed at-line and/or used in a portable manner, thus providing answers when and where they are needed most.