Simulation of Near Infrared Interference Bandpass Filters for Spectroscopic Applications

Food quality can be characterized by noninvasive techniques such as spectroscopy. Near Infrared spectroscopy in the similar to 930-1450 nm wavelength range are of great importance to detect the quality and to differentiate between meat samples. Miniaturized NIR spectrometers are useful for quick and mobile characterization of food samples. Such devices can be fabricated using Fabry-Perot (FP) filters consisting of two highly reflecting mirrors with a central cavity in between. Distributed Bragg Reflections (DBRs) consisting of alternating high and low refractive index material pairs are the most commonly used mirrors in FP filters, due to their high reflectivity. However, DBRs have high reflectivity for a selected range of wavelengths known as the stopband of the DBR. This range is usually much smaller than the sensitivity range of the spectrometer detector. Therefore, a bandpass filter is required to restrict wavelengths outside the stopband of the FP DBRs. Three bandpass filters are designed and optimized in this work for three different wavelength ranges using the needle synthesis method. The passbands are selected based on optimal wavelengths required for meat classification and/or meat quality determination. All three filters show high quality with average transmissions of 96%, 90% and 90% within the passbands. The transmission outside the passband is less than 3% in all three cases. Special attention has been given to keep the thickness of the filters within economic limits.