background spectrum. The following conditions must remain constant or be accounted for in quantitative FTIR-PAS analyses:

1. FTIR mirror velocity
2. FTIR and PAS detector amplifier gain settings
3. PAS detector optical alignment
4. FTIR source intensity
5. FTIR interferometer alignment
6. Helium gas concentration in the PAS detector
7. Sample volume
9. Sample matrix
10. Microphone sensitivity
11. Carbon black standard response

The FTIR mirror velocity must be set at the same value when spectra of the calibration standards (learning set) and of unknowns are acquired. If the mirror velocity setting is not maintained, changes in the PAS sampling depth will occur which are analogous to changing sample thickness or concentration in transmission spectroscopy.
All amplifier gain settings should be at the same values when sample spectra are acquired in order to have constant measurement conditions.
Good quantitative analyses also require that the detector positioning in the FTIR is reproducible when the accessory is put in and taken out of the FTIR and that its position remains fixed when the accessory is installed. Properly designed kinematic mount and baseplate registration accomplish these requirements.
Scale variations are caused in FTIR-PAS spectra if changes occur in source intensity, interferometer alignment, helium concentration; sample volume, morphology, and matrix; and microphone sensitivity and carbon black response. In some instances, if proper care is taken, these changes can be held small enough to yield satisfactory quantitative analyses. But in most instances, adding a standardization procedure that addresses all of these potential changes is the best way to insure that quantitative results will have the degree of reproducibility necessary for a particular application.
FTIR-PAS spectra are usually best standardized by one of two methods. The first is applicable to situations where the spectral changes over the full analyte range are small relative to the whole spectrum. In this case the spectra can be scaled so that the area under the whole spectrum from 400-4000 cm is held constant.
The second method exploits the fact that highly saturated bands in the FTIR-PAS spectra vary much slower with changes in analyte concentration than do unsaturated bands. Consequently, FTIR-PAS spectra can be standardized by scaling spectra so that a particular intense band's height or area remains constant.