Nd. Rezgui et al., QUANTITATIVE MILLIMETER-WAVE SPECTROSCOPY .2. DETERMINATION OF WORKING-CONDITIONS IN AN OPEN FABRY-PEROT CAVITY, Analytica chimica acta, 312(1), 1995, pp. 115-125
A confocal Fabry-Perot frequency modulated cavity spectrometer of qual
ity factor 1.25 X 10(5) operating inside a chamber maintained at ambie
nt temperature and pressure of 1 Pa to 1 KPa was employed for spectrom
etric measurements in the region of 72 GHz and 140 to 160 GHz. The spe
ctrometer used a spatial filter to suppress unwanted, non-axial modes.
The solid state microwave source frequency was derived from a phase-l
ocked frequency synthesizer and detection was by a liquid helium coole
d bolometer. Transitions in acrylonitrile, formaldehyde, and sulphur d
ioxide were studied demonstrating parts per million sensitivity for th
ese species in atmospheric samples, whilst carbonyl sulphide samples w
ere detected at sub-parts per million concentration. The effect of pre
ssure on line intensities was studied in order to determine the optimu
m operating regime. It was found that the technique was not restricted
to the 5-50 Pa region characteristic of centimetric wave spectroscopy
, but was able also to function in the 0.1 to 1 KPa regime. Furthermor
e the intensities in this latter region were found to be not criticall
y dependent on sample pressure. A treatment of the effect of pressure
and depth of frequency modulation on absorption signals was carried ou
t and the resulting theory applied to the observed intensity-pressure
relationships. There was good quantitative agreement between the frequ
ency modulation depth and cavity response characteristics and qualitat
ive agreement between the pressure, frequency modulation and spectral
line intensity characteristics. It became clear that the possibility o
f power saturation, coupled with the non-uniform power distribution wi
thin the cavity, was affecting the fits of theoretical curves to the o
bserved data, and that taking this into account produced marked improv
ement in the fits. Nonetheless the treatment permitted some practicall
y useful conclusions: at low modulation depths and pressures, the shar
p spectral absorption peak makes identification of the target species
easy, but extraction of quantitative information more difficult, as th
e intensity will depend critically on the power level in the cavity. A
t pressures in the 100 Pa region however, the signal obtained is maxim
al, power broadening minimal and comparative intensity measurements po
ssible over a range of sample species and concentrations.