Abstract
All near infrared instruments are equipped with a “window” to protect the detectors from the atmosphere. Many types of sample cells are also fitted with a window between the sample surface and the detector. These windows may be composed of special optical glass or quartz. This paper describes variability of wheat flour spectra when using a series of glass and quartz windows. Differences among the windows of both glass and quartz exhibited differences of up to respectively 40% and 14% in the spectral data when they were used in the development of calibration models for the prediction of protein content in wheat flour. The differences among the windows did not affect the precision of spectral data when the same window was used for replicate testing, but induced absolute biases of up to over 3% (glass) and 1% (quartz) to the predicted protein results. Biases could be countered by use of derivatives or by combining spectral data from flours by using the different windows in the development of the calibration models. An anomaly was noted in the flour spectra obtained using quartz window number 4. The anomaly was found to cause significant biases in the prediction of protein content. The biases could be removed if the calibration model was developed using a wavelength range that avoided the anomaly.
© 2006 NIR Publications
PDF Article
More Like This
Cited By
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access Optica Member Subscription