1 Introduction

Our view of the outer Solar System has changed dramatically in the last decade, with the discovery of a large number of bodies in orbit beyond Neptune (see Jewitt & Luu 2000, for a review). These so-called Kuiper Belt objects (KBO) form three distinct dynamical classes: the Plutinos near the 3:2 mean-motion resonance with Neptune, with dynamical similarities with Pluto; the scattered objects (SKBO) with large, eccentric and inclined orbits; and the classical KBOs with modest eccentricities and semimajor axes $a\geq 42$ AU. It is believed that the inner edge of the Kuiper Belt, gravitationally influenced by Neptune and Uranus, has supplied the Centaurs population (Levison & Duncan 1997). Indeed, these objects, with semimajor axes 8 AU <a< 35 AU, between those of Jupiter and Neptune, have strongly chaotic and short-lived orbits, so that a source region from which the Centaurs' population is continuously replenished must exist. In turn, Centaurs can be a source of short-period comets (Levison & Duncan 1997). At the time of writing, $\sim$400 KBOs, $\sim$40 SKBOs and $\sim$30 Centaurs have been catalogued.

The study of the compositional properties of KBOs and Centaurs is of strong interest because they are likely to consist of the most unaltered matter from the Solar Nebula. Photometric measurements show evidence for a wide dispersion in the colors of KBOs and Centaurs, from neutral to very red (e.g. Barucci et al. 2000a; Jewitt & Luu 2000, for a review), with some data sets arguing for a bimodal color distribution (Tegler & Romanishin 1998, 2000). The color diversity has been interpreted as due to diverse degrees of impact rejuvenation of the surface altered by cosmic rays (Luu & Jewitt 1996). Near-infrared spectroscopy revealed water ice at the surface of the Centaurs 10199 Chariklo, 5145 Pholus, 2060 Chiron, and possibly 8405 Asbolus (Brown & Koresko 1998; McBride et al. 1999; Cruikshank et al. 1998; Foster et al. 2001; Luu et al. 2000; Kern et al. 2000; Barucci et al. 2000b). The depth of the water absorption features varies among objects and with observational date. Methanol ice (or another light hydrocarbon or oxidized derivative) is seen in the near-infrared spectrum of Pholus (Cruikshank et al. 1998). Water ice absorption features were detected in the spectrum of the KBO 1996 TO66 (Brown et al. 1999). Spectra obtained in a few other Centaurs and KBOs show no evidence for ice signatures (e.g. Luu & Jewitt 1998; Brown et al. 2000a, 2000b). This diversity is intriguing and, up to now, no correlation between surface and dynamical properties has been established. In addition, Chiron presents cometary-like activity, while 8405 Asbolus, which has similar color and orbital characteristics, does not. The detection of CO outgassing in 29P/Schwassmann-Wachmann 1 (Senay & Jewitt 1994; Crovisier et al. 1995; Festou et al. 2001), a comet sometimes classified as a Centaur which orbits at $\sim$6 AU from the Sun, and of CO and CO₂ in the coma of comet C/1995 O1 (Hale-Bopp) at large distances from the Sun (Biver et al. 1999a; Biver et al. 1999c; Crovisier et al. 1997), shows that the outgassing of very volatile species is at the origin of distant activity. Evolutionary thermal models of Centaurs and KBOs, assuming that these bodies are made of different ices and dust, have been developed to investigate their differentiation and outgassing (Prialnik et al. 1995; Prialnik & Podolak 1995; Capria et al. 2000; De Sanctis et al. 2000, 2001).

Pluto is the only transneptunian object for which a bound atmosphere has been positively identified. Observationally, most of our knowledge of Pluto's atmosphere was acquired from the 1988 stellar occultation (see e.g. Millis et al. 1993; Stansberry et al. 1994; Yelle & Elliot 1997), which revealed a tenuous (several $\mu$bar) atmosphere exhibiting a strong thermal inversion with the atmospheric temperature rising from $\sim$40 K to $\sim$100 K in the first $\sim$20 km. The subsequent discovery in the near-infrared of N₂, CO, and H₂O ices on Pluto's surface (Owen et al. 1993; Cruikshank et al. 1997) in addition to the already detected CH₄ ice, measurements of their relative abundances, and considerations on vapor pressure equilibria, established that N₂ is the major gas of Pluto's atmosphere and that CH₄ and CO must be secondary compounds. From near-IR spectroscopy, Young et al. (1997) reported the detection of methane gas at the percent level, but this detection has not yet been confirmed. Regarding CO, Barnes (1993) reported an upper limit on the CO J(1-0) rotational emission, but this upper limit was not stringent enough to constrain the CO abundance. More recently, from high-resolution spectroscopy at 2.3 $\mu$m, Young et al. (2001) obtained upper limits of ( $1.2 \times 10^{21}{-}9.2 \times 10^{24}$) molcm^-2 for the CO column density. The corresponding CO/N₂ mixing ratio limit is 0.06-0.23 in the best case (i.e. for a N₂ surface pressure of 58 $\mu$bar).

However, theoretical expectations suggest a much lower CO/N₂ mixing ratio in Pluto's atmosphere. Based on their estimate of the CO abundance in the ice phase (0.5% of N₂), and assuming that the N₂-CO-CH₄system is ideal (i.e. applying Raoult's law), Owen et al. (1993) estimated CO/N₂ = (0.02-0.2)% in Pluto's atmosphere for an equilibrium surface temperature in the 34-58 K range. Lellouch (1994) and Strobel et al. (1996) adopted similar values (0.075% and 0.046%) in their thermal structure models of Pluto's atmosphere. With the lower CO abundance in the ice (0.1-0.2%) inferred by Douté et al. (1999) from a refined analysis of the near-IR observations, the CO atmospheric abundance should be accordingly lower. We note, however, that applying the same approach to Triton, which has a CO abundance of 0.05% in the ice phase at 38 K (Quirico et al. 1999), leads to an expected atmospheric CO mixing ratio of $4 \times 10^{-5}$, significantly below the $(0.25{-}5) \times 10^{-3}$ value suggested by Strobel et al. (1996) from thermal structure modelling.

As demonstrated by Lellouch (1994) and Strobel et al. (1996), the CO abundance impacts the thermal structure of Pluto's atmosphere, especially at sub-microbar pressure levels, through efficient cooling in pure rotational lines. In what follows, we report our attempt to detect CO in two of these lines, with the increased sensitivity of the 30-m telescope of the Institut de Radioastronomie Millimétrique (IRAM). We also present upper limits on the CO outgassing in a few Centaurs and KBOs obtained from CO observations at the James Clerk Maxwell Telescope (JCMT) and Caltech Submillimeter Observatory (CSO).