CROMOS: A cryogenic near-infrared, multi-object spectrometer for the VLT

We discuss a cryogenic, multi-object near-infrared spectrometer as a second generation instrument for the VLT. The spectrometer combines 20 to 40 independent integral field units (IFUs), which can be positioned by a cryogenic robot over the entire unvignetted field of the VLT (∼7’). Each IFU consists of a contiguous cluster of 20 to 30 pixels (0.15 to 0.25” per pixel). The individual IFUs have cold fore-optics and couple into the spectrograph with integrated fibers-microlenses. The spectrometer has resolving power of λ/∆λ∼4000 and simultaneously covers the J-, H-, and K-bands with three HAWAII 2 detectors. The system is designed for operation both in seeing limited and MCAO modes. Its speed is approximately 3500 times greater than that of ISAAC and 60 times greater than NIRMOS (in H-band). The proposed instrument aims at a wide range of science, ranging from studies of galaxies/clusters in the high-z Universe (dynamics and star formation in z>1 galaxies, evolution of ellipticals, properties of distant, obscured far-IR and X-ray sources), to investigations of nearby starbursts, star clusters and properties of young low mass stars and brown dwarfs. 1 Motivation and Science Drivers Optical photometry/spectroscopy is the easiest and most commonly used technique for studying stellar and interstellar components in galaxies. Yet extinction/reddening by dust and cosmological redshift are two of the reasons that make observations in the near-infrared (NIR: λ ∼ 1–2.4 μm) necessary and/or highly attractive. The much better performance of adaptive optics at infrared wavelengths is another. Furthermore the NIR emission traces older stellar populations that are better measures of stellar mass in galaxies, and there are a number of NIR spectral features (e.g. H2, CH4, H2O etc.) that uniquely trace cool interstellar and circumstellar gas. Because of poor detector performance and size, low instrument transmission and high sky brightness, however, infrared observers until a few years ago had to pay a price of >3 mag in sensitivity compared to optical spectroscopy, making NIR spectroscopy of distant or faint sources challenging or impossible (Figure 2). With the advent of new high quality detectors and spectrometers, such as ISAAC on the VLT and NIRSPEC on the Keck telescope [10][11], combined with software suppression of the OH sky emission lines, NIR long-slit spectroscopy has become competitive with optical spectroscopy. Adaptive optics assisted, integral field spectroscopy (e.g. with SINFONI on the VLT [14]) will soon open up sensitive, near-diffraction limited, NIR