Instrument to instrument variability: Unless they have been “standardized” all spectrometers will characterize a reference emission spectrum differently. This is a function of many issues including variations in detector efficiency, bandpass and various system optics
Uncertainties in wavelength ratios: If a sample spectrum covers an extended spectral range, wavelength ratios may vary between instruments due to instrumental component variables. If measuring accurate wavelength ratios is necessary then the data delivered by an instrument needs to be “normalized” to a standard.
Without correction, data will be highly reproducible, yet wavelength ratios will be “precisely inaccurate.” The main wavelength dependent culprits include:
- The quantum efficiency (QE) of the spectrum detector such as a CCD or CMOS camera or photomultiplier tube (PMT)
- The wavelength versus efficiency profile of a diffraction grating or prism
- Variations between electronic bandpass filters such as Acousto Optic Tunable Filters (AOTF), Liquid Crystal Tunable Filters (LCTF) or interferometers.
- Optical coatings, lenses, mirrors, filters, polarizers and any other optic that is located between the sample and the detector
- Aliasing when the system fails to accurately reconstruct an analog spectrum into a digital format.
The impact of all these influences are eliminated when a spectrum is acquired in %transmission, absorption, or %reflection. In these cases, the sample spectrum is divided by a spectrum of the illumination source that has also passed through the same system.
However, in emission spectroscopy including LED, OLED photo-luminescance fluorescence, and Raman there is no illuminant spectrum to divide by. Therefore, in order to accurately reconstruct wavelength ratio intensities, it is necessary to apply a correction factor at each wavelength.