R.
S. DaBell (839), P. M. Chu (839), R. D. Suenram (844)
Objective: To establish the feasibility of Fourier-transform microwave
(FTMW) spectroscopy as a reliable technique for quantifying trace-gas analytes.
Problem: There is a continuing need for improved analytical techniques to
measure the concentration of trace gases for monitoring hazardous air
pollutants, industrial emissions, chemical warfare agent release, etc. In particular, a method that can
conclusively identify and quantify multiple analytes in ambient air samples is
critical for addressing pressing issues such as global climate change. Towards this end, the use of
Fourier-transform microwave (FTMW) spectroscopy as a quantitative analytical
technique is being investigated. The
high spectral resolution and high sensitivity of FTMW spectroscopy suggest that
the technique can provide near real-time response and unambiguous
identification of analytes in a gas sample.
The principal goal of the present program is to evaluate the technique’s
potential as a reliable and robust tool for quantitative measurements of trace
gases in ambient air samples. To achieve these goals requires: 1) evaluation of
the instrument function; 2) consideration of sample delivery issues; 3)
comparison with other established analytical techniques; and 4) blind sample
testing.
Results: Initial evaluation of the instrument revealed that measured peak
positions for a given analyte were reproducible to within 5 kHz providing powerful
and reliable information for species identification. The signal intensities, however, varied substantially over long
measurement periods (12 h), with worst-case scenarios showing complete loss of
signal. To regain signal intensity, it
was necessary to adjust or modify a number of instrument parameters. To improve
the reliability of signal intensity, the performance of each individual
component was evaluated and several instrument modifications have been
incorporated. Currently, the remaining
drift is ≈10 % over 12 h and the chief source of drift has been traced to
fluctuations in the resonant frequency of the Fabry-Perot cavity caused by
temperature induced changes of the optical cavity length.
Concomitant
with our effort to improve the instrument performance, we have also undertaken
preliminary tests to examine the signal intensity as a function of the sample
concentration and determine the detection limits using NIST standard reference
samples of sulfur dioxide in a nitrogen matrix and ethanol in a blended air
matrix. For sulfur dioxide, using a
sample concentration range of 50 mmol/mol to1000 mmol/mol, a 0.14 mmol/mol detection limit was
obtained. For ethanol, using a
concentration range of 200 mmol/mol to 1400 mmol/mol, the detection limit
was 1.3 mmol/mol. In both cases, the
signal intensity was linear with the sample concentration over the entire
ranges. These measurements support the
potential of FTMW spectroscopy to quantitatively measure analytes in nitrogen
or blended air matrices.
Future Work: As work to optimize the
instrument performance continues, an effort will also focus on evaluating the
capability of FTMW spectroscopy to measure analytes in gas matrices which more
closely match ambient air. Since this
technique is based on probing rotational transitions and requires a supersonic
expansion of the gas sample to concentrate the population distribution,
different sample matrices will strongly impact the properties of the expansion
and ultimately affect the detection limits. To date, quantitative measurements
have not been the major driving force of FTMW studies and most measurements are
made using an idealized balance gas of helium and neon. An effort will be focused on characterizing
the signal intensities for samples with high concentrations of nitrogen,
oxygen, carbon dioxide, and water, which are at high concentrations in
analytically relevant samples.
Publications: DaBell, R. S., Chu, P. M., Fraser, G. T.,
Suenram, R. D., “Evaluation of Fourier-transform microwave spectroscopy as a
tool for quantitative analysis: Signal stability considerations”, Proc. SPIE,
4574, to be published (2001).