METRICS FOR THE PERFORMANCE OF CHIRPED-PULSE FOURIER TRANSFORM MICROWAVE SPECTROMETERS

We have been pursuing several applications of molecular rotational resonance (MRR) spectroscopy in quantitative analysis relevant to the pharmaceutical industry. These applications include chiral analysis (absolute configuration and enantiomeric excess), diastereomer identification, regioisomer identification, and isotopologue/isotopomer analysis. Because this work requires quantitative determination of species abundances, we have examined issues related to signal-to-noise determination and quantitation in the low signal regime. Models for the noise distribution in broadband, chirped-pulse Fourier transform microwave (CP-FTMW) spectrometers where the spectra are reported as the magnitude Fourier transform will be presented. The noise distribution is also found to be the same in cavity FTMW spectrometers. We have also modeled the accuracy of signal levels observed in the magnitude Fourier transform as a function of signal-to-noise ratio. This modeling shows that measured transition intensities are accurate down to the root-mean-squared noise level of the instrument without the need to correct for noise power in the signal channel. Finally, using the proper definitions for the noise in magnitude FTMW spectra, we have determined the performance metric introduced by Porterfield et al.\footnote{J.P. Porterfield, L. Satterthwaite, S. Eibenberger, D. Patterson, M.C. McCarthy, \textit{Rev. Sci. Instrum. } \textbf{2019}, \textit{90}, 053104.} for the UVA 6-18 GHz broadband spectrometer. The CP-FTMW instrument speed metric is S = 36,000,000 MHz/min using OCS as the test sample. The reported value for the cryogenic buffer gas cell instrument described in Ref. 1 is S = 4,400,000 MHz/min. These results are in strong disagreement with the CP-FTMW speed metric reported in Ref. 1, S = 60,360 MHz/min, that was the basis for claiming that buffer gas cooling spectrometers have almost a factor of 100 better performance than CP-FTMW instruments