In addition to the complex ionized (and non-thermal) sources just discussed, regions of massive star formation are rich in molecular emissions. The NH molecule is one of the most useful for the study of high density molecular material, because of series of inversion transitions just above 23 GHz. These transitions have hyperfine components and, thus, allow direct estimates of density and clumping (filling factor).
For example, the NH (22) inversion transition has five components: a main line, two inner satellites separated by 116.55 km/s, and two outer satellites 125.7 km/s from the main line. To resolve the lines, sometimes as narrow as 0.25 km/s, a channel separation of 0.05 km/s is desirable. To allow for enough channels outside the outer satellites in order to measure the baseline, a total bandwidth of 100 km/s (8 MHz) needed with about 2000 channels. If the source has gas motions which are of order 16 km/s, every one of these channels could contain line signal. (If the motions are more typically 5 km/s or less, some savings could be had by positioning independent groups of channels on each of the lines in which case only 1000 channels are necessary.) For greater sensitivity, one would like to obtain the LCP and RCP data simultaneously. Using the present correlator, most observers are forced to ignore the outer satellites (although they can help constrain LTE models and provide and independent column density estimate). Typically only the main line and the inner satellites are observed in a 3.125 MHz band with 16 or 32 channels. The main line and the satellites are often blended in the channels which have only 1.25 km/s resolution.
A new broadband data transmission system and correlator are needed to make a major advance and provide adequate spectral resolution and capabilities to do multiple lines simultaneously. For example, in the 22.5 GHz band, one could make T and density maps using the many NH transitions. For hot cores near H II regions one would like to use the (11) and (22) transitions to get the extended, lower temperature structures; use the (44) and (55) transitions for moderate temperatures, and higher metastable line to get the higher density, hotter features; also the non-metastable lines such as the (21) can be used to study the highest densities (N); and finally the (33) and (66) ortho lines often mase. Clearly at least 4, and preferably 8, independent channels would be desirable. The frequency range from (11) to (66) is 23-25 GHz and spectral channels of width of about 8 MHz per line and 8 kHz spectral resolution. Also, a simultaneously, broad-band continuum measurement would provide valuable complementary information regarding the physical setting and sources of excitation.
The CS molecule has its fundamental J=10 transition at 49 GHz, in the new 40-50 GHz system of the VLA. Recent observations of the IRc2 region in Orion using the existing 45 GHz system have approximately twice the spatial resolution of any existing image made with a thermally excited molecular line. The outflow of IRc2 is evident as well as a velocity gradient running along the length of the Orion molecular ridge. The enhanced VLA could produce an image with approximately 30 times better sensitivity, making it possible to map the dynamics of the outflow as traced by the weak emission in the line wings.
Maser observations, whether OH at 18cm or CHOH at 7mm, are some of the most demanding spectral line projects. High spectral resolution is required to measure the very narrow lines ( km/s), a wide velocity range is needed because the masers in star forming regions can be present over a wide range in velocity, wide fields may need to be imaged because star formation is often highly clustered, and polarization measurements are needed if one is trying to use the Zeeman effect to measure the in situ magnetic field. One may wish to cover 40-80 km/s with channels as narrow as 0.1 (even less for Zeeman experiments). Thus 800-1000 channels are needed with circular polarization information. Masers are important sign posts of new star formation, they a valuable probes of the interstellar gas dynamics, they are used in proper motions studies to determine the distances, and they are valuable tools for self-calibration of continuum or other line measurements because they are bright, unresolved point sources.
There are additional molecules of interest. Massive organic molecules (e.g., carbon chains and possible rings), are building blocks of molecules crucial to formation of living organisms. There are reports of mm-wavelength detection of amino acids. These observations were not possible with single dishes owing to line crowding, but were successful with an interferometer. Thus, large molecules with transitions at centimeter -wavelengths are probably best detected with the VLA, over (say) the Green Bank Telescope, because of the excellent spectral baselines achieved with an interferometer and the higher duty cycle on-source. Also, since the VLA antennas are much smaller than a large single dish such as the Green Bank Telescope, the increase in noise from the source (e.g., Sgr B2, Cas A) is much smaller for the VLA. The combination of all these advantages (for the enhanced VLA versus the Green Bank Telescope we estimate factors of 1.5 for collecting area, 1.5 for increased duty cycle, and a factor of order 3 (for a 100 Jy source) from system temperatures) yields a sensitivity increase of 7, or an integration time of 50-favoring the VLA! The wider field of view at the VLA might also be valuable if the position of the line emission/absorption is not precisely known.
Formaldehyde has been a staple molecule for centimeter-wavelength telescopes for many years. Unfortunately, only two lines have been available for study-at 6cm and at 2cm. These lines are affected by the well known excitation anomaly which causes them to be seen in absorption against the 2.7K background over much of a cloud's extent. Fortunately, the VLA will make available several very interesting new transitions with completion of the 45 GHz system. Formaldehyde may turn out to be the ``molecule of choice" for probing physical conditions in clouds with the VLA. In addition to the 6cm and 2cm lines, the 3(1,2)3(1,3) line at 28.975 GHz, and the 4(1,3)4(1,4) line at 48.3 GHz could be available on the VLA. These lines are very useful densitometers. Mangum (1990) noted in his dissertation that these K-doublet transitions have a spatial density filter built into their excitation; for densities below the formaldehyde lines are in absorption for J of 4 or below, while for densities above this value all are in emission. At J 5, all lines will be in emission where the excitation is sufficiently robust. Mangum also showed that formaldehyde is less susceptible to abundance peculiarities than some other molecules, notably ammonia, which render interpretation of the observations difficult. Formaldehyde observations at higher frequencies will provide an excellent densitometer, free from the complex cloud-envelope effects which plague interpretation of lower-lying formaldehyde lines. Formaldehyde has been detected in several external galaxies; the additional higher frequency lines should be detectable in several galaxies with the VLA, providing valuable density information.