Skip to Content

Pushing Chirped-Pulse Fourier Transform Spectroscopy to Millimeter-Wave Frequencies

Researchers Brandon Carroll, Justin Neill & Dan Zaleski



Chirped-pulse Fourier transform spectroscopy, which builds on recent technological advances in arbitrary waveform generation and detection, is a powerful technique for broad-bandwidth molecular rotational spectroscopy.  The coming availability of sensitive radio telescopes with high detection bandwidths in the microwave and millimeter-wave regions provides the driving force for pushing chirped-pulse spectrometers to higher frequency.


In developing chirped-pulse spectrometers at millimeter-wave frequencies, the primary challenges are producing phase-stable chirped pulses, overcoming the reduced power levels available as frequency increases, and sensitively detecting molecular emissions against zero background.  Chirped-pulse millimeter wave spectrometers have been constructed to date at MIT (R. Field) and NIST (D.F. Plusquellic, K.O. Douglass).  Here we present the design of a Ka-band (26-40 GHz) chirped-pulse spectrometer that can collect a 15 GHz spectrum in 10 microseconds, with high phase stability (permitting extensive signal averaging) and accurate relative intensities.  This spectrometer can detect transient species, such as radicals or ions, that are challenging to synthesize in large abundances.  Also, this spectrometer overlaps in frequency with both the Green Bank Telescope and Expanded Very Large Array (above).

The primary concern of the current spectrometer design is the presence of spurious signals and “ghosts”. Ghosts are molecular signals that are the result of sub-harmonic mixing. This is a problem because ghosts can occupy a large number of detection channels for a dense spectrum. Fortunately, the main ghosts that remain come from sub-harmonic mixing of twice the PDRO frequency, and are thus predictable and able to be cut. Ultimately the final design will utilize a 44 GHz PDRO. Down-converting from above the molecular signals will greatly reduce the effects of sub-harmonic mixing. A filter for a 44 GHz PDRO is currently being prototyped (K&L Microwave).


The measurement of the pure rotational spectrum of OCS (0.4% dissolved in neon), shown above, is used to demonstrate the sensitivity and intensity accuracy of the Ka-Band CP-FTMW spectrometer. After 315,000 signal acquisitions, the signal to noise of the J=3-2 rotational transition of the normal species of OCS was approximately 200,000:1. The isotopomer O13C33S can clearly be seen in natural abundance. Across all measured isotopomers, the relative intensities are on the order of 27.64% of the literature isotope abundance ratios. Accurate relative intensities allow relative species populations and relative dipole moments to be derived.


Identified species:


CH3CCH (methylacetylene)*


CH2CHCCH (vinylacetylene)


HCCCHO (2-propynal)*


H3CCCCCH (methyldiacetylene)*




H2CCO (ketene)*

H2CCHCHO (propenal)*

* Indicates that the molecule is interstellar

A major advantage of this spectrometer is in the ability to analyze complex mixtures, especially for experiments in which a pulsed-discharge nozzle (PDN) introduces countless new molecular carriers. These new signals could be the result of synthesis, fragmentation, ions, radicals, or high energy conformers. Many of these discharge products may also be found in the interstellar medium (ISM). This ability is demonstrated above. Allene (0.1% in Neon carrier gas) was discharged at 1.1 kV. Because allene has no rotational spectrum, all observed molecular species are due to synthesis in the discharge nozzle. A total of 159 molecular transitions were identified, 60 by using


G.G. Brown et al. Rev. Sci. Instrum. 79 (2008) 053103.

G.B. Park et al., 64th International Symposium on Molecular Spectroscopy, 2009.

K.O. Douglass et al., 65th International Symposium on Molecular Spectroscopy, 2010.