On the absence of molecular absorption in high-redshift millimetre-band searches

S. J. Curran 3 M. T. Whiting 1 3 F. Combes 0 N. Kuno 7 P. Francis 6 N. Nakai 5 J. K. Webb 3 M. T. Murphy 3 4 T. Wiklind 2 8 9 0 LERMA, Observatoire de Paris, 77 Avenue Denfert-Rochereau, 75014 Paris, France 1 CSIRO Australia Telescope National Facility , PO Box 76, Epping, NSW 1710, Australia 2 Joint ALMA Observatory, Santiago, Chile 3 School of Physics, University of New South Wales , Sydney, NSW 2052, Australia 4 Centre for Astrophysics and Supercomputing, Swinburne University of Technology , PO Box 218, Hawthorn, VIC 3122, Australia 5 Institute of Physics, University of Tsukuba , Ten-noudai, Tsukuba, Ibaraki 305-8571, Japan 6 Australian National University , Canberra, Australia 7 Nobeyama Radio Observatory, Nagano 384-1305, Japan 8 Onsala Space Observatory, S-439 92 Onsala, Sweden 9 Space Telescope Science Institute , Baltimore, MD 21218, USA A B S T R A C T We have undertaken a search for millimetre-waveband absorption (through the CO and HCO+ rotational transitions) in the host galaxies of reddened radio sources (z = 0.405-1.802). Despite the colour selection (optical-near-infrared colours of V K 5 in all but one source), no absorption was found in any of the eight quasars for which the background continuum flux was detected. On the basis of the previous (mostly intervening) H2 and OH detections, the limits reached here and in some previous surveys should be deep enough to detect molecular absorption according to their V K colours. However, our survey makes the assumption that the reddening is associated with dust close to the emission redshift of the quasar and that the narrow millimetre component of this emission is intercepted by the compact molecular cores. By using the known millimetre absorbers to define the colour depth and comparing this with the ultraviolet luminosities of the sources, we find that, even if these assumptions are valid, only 12 of the 40 objects (mainly from this work) are potentially detectable. This is assuming an excitation temperature of T x = 10 K at z = 0, with the number decreasing with increasing temperatures (to zero detectable at T x 100 K). 1 I N T R O D U C T I O N Millimetre-wave observations of molecular absorption systems along the sightlines to distant quasars provide a powerful probe of the cold, dense, star-forming gas in the distant Universe. Furthermore, a comparison of the redshifts of the rotational transitions of the molecules with those of the spin-flip transition of H I, as well as the electronic optical/UV transitions of metal ions, can be used to determine high-redshift values of the fundamental constants, to at least an order of magnitude the sensitivity of purely optical data (see Curran, Kanekar & Darling 2004a). However, despite much searching, only four such systems are currently known (Wiklind & Combes 1995; 1996ab; 1997), the highest redshift being at zabs = 0.89. Of these, two are intervening systems (gravitational lenses towards more distant quasars), with the other two systems arising through absorption within the host galaxy of the quasar. Subsequent searches at the redshifts of known high column density H I absorption systems, intervening the sightlines to more distant quasi-stellar objects (QSOs), have also failed to detect molecular absorption in the millimetreband (Curran et al. 2004b and references therein), despite the possibility that these so-called damped Lyman systems (DLAs)1 1 These have neutral hydrogen column densities of NH I 2 1020 cm2 and are usually detected at zabs 1.8, where the Lyman transition is redshifted into the optical band. 2 One of which, J1439 + 113, has also been detected in the CO A X UV band (Srianand et al. 2008). 1070.3 and V K 4 (Fig. 1, middle), i.e. in the same range as a typical QSO (Fig. 1, bottom),3 whereas the millimetre- and decimetre-band absorbers have molecular fractions of F 0.61 and opticalnear-infrared colours of V K 5. The correlations in Fig. 1 present strong evidence that the quasar light is reddened by dust in the foreground absorber: since the presence of the dust is necessary to prevent the dissociation of the molecular gas by the ambient ultraviolet field, the molecular fraction is expected to be correlated with the dust abundance, as observed. The paucity of millimetre-waveband absorption can therefore be attributed to the traditional optical selection of targets biasing towards absorbers of low dust content and therefore low molecular fractions. The fact that intervening absorbers are usually found through optical spectroscopy, yielding a redshift but also giving the above bias against dusty objects, means that millimetre-band searches of known intervening absorbers have generally been unsuccessful (Curran et al. 2004b and references therein). An alternative target for molecular absorption is towards the fainter red quasars, where the red colour may indicate an intervening column of dust. However, due to the relatively narrow bandwidths in the millimetre band (see Section 4.2), such an approach is currently only practical at longer (decimetre) wavelengths (see Curran et al. 2011b). In the absence of any known intervening absorbers, selecting the quasar itself gives a redshift (zem) to which to tune the receiver. Naturally, such a selection of targets prevents any useful comparison with the optical redshifts, in order to measure the values of the fundamental constants, although any detections could be followed up in 21 cm, giving the redshift of the spin-flip transition of H I. In Curran et al. (2006, 2008, 2011a), we presented the results from our decimetre-wave searches for such associated (OH and H I) absorption and here we present the results of our millimetrewave survey for the associated absorption. 2 O B S E RVAT I O N S 2.1 Target selection As per Curran et al. (2006), our sources were selected from the Parkes Half-Jansky Flat-spectrum Sample (PHFS, Drinkwater et al. 1997),4 on the basis of their opticalnear-IR photometry (Francis, Whiting & Webster 2000). From these, we selected the 30 reddest sources (which correspond to an extinction of AV 4.1), in which the emission redshift of the quasar (zem) would place a strong absorption line (CO or HCO+) into the 3-mm band. After culling these further, by selecting those of > 30 (thus being observable from northern latitudes)5 and with 3-mm flux densities expected to be 100 mJy, the 10 objects listed in Table 1 remained. 2.2 The IRAM 30-m observations From 2003 December to 2004 February we observed three of the targets with the IRAM (Institut de Radio Astronomie Millimetrique) 30-m telescope at Pico Veleta, Spain. We used two 3 V K = 2.88 1.04 in general and 3.05 0.97 if radio-loud. 4 With the addition of 0500 + 019, included since it has been detected in 21-cm absorption (Carilli et al. 1998). We also included J0906 + 4952 and J1341 + 3301, which are two very red sources from Glikman et al. (2004) (Section 2.3). 5 We miss SEST. V K 1.17121.1848 1.17541.1806 0. (...truncated)