In this paper we describe a first quantitative search for several molecules in Titan's stratosphere in Cassini CIRS infrared spectra. These are: ammonia (NH3), methanol (CH3OH), formaldehyde (H2CO), and acetonitrile (CH3CN), all of which are predicted by photochemical models but only the last of which observed, and not in the infrared. We find non-detections in all cases, but derive upper limits on the abundances from low-noise observations at 25{\deg}S and 75{\deg}N. Comparing these constraints to model predictions, we conclude that CIRS is highly unlikely to see NH3 or CH3OH emissions. However, CH3CN and H2CO are closer to CIRS detectability, and we suggest ways in which the sensitivity threshold may be lowered towards this goal.

In this paper we select large spectral averages of data from the Cassini Composite Infrared Spectrometer (CIRS) obtained in limb-viewing mode at low latitudes (30S--30N), greatly increasing the path length and hence signal-to-noise ratio for optically thin trace species such as propane. By modeling and subtracting the emissions of other gas species, we demonstrate that at least six infrared bands of propane are detected by CIRS, including two not previously identified in Titan spectra. Using a new line list for the range 1300-1400cm -1, along with an existing GEISA list, we retrieve propane abundances from two bands at 748 and 1376 cm-1. At 748 cm-1 we retrieve 4.2 +/- 0.5 x 10(-7) (1-sigma error) at 2 mbar, in good agreement with previous studies, although lack of hotbands in the present spectral atlas remains a problem. We also determine 5.7 +/- 0.8 x 10(-7) at 2 mbar from the 1376 cm-1 band - a value that is probably affected by systematic errors including continuum gradients due to haze and also an imperfect model of the n6 band of ethane. This study clearly shows for the first time the ubiquity of propane's emission bands across the thermal infrared spectrum of Titan, and points to an urgent need for further laboratory spectroscopy work, both to provide the line positions and intensities needed to model these bands, and also to further characterize haze spectral opacity. The present lack of accurate modeling capability for propane is an impediment not only for the measurement of propane itself, but also for the search for the emissions of new molecules in many spectral regions.