2.2. Absorption lines

Molecular absorption lines are also a powerful tool to study the interstellar medium of galaxies at high redshift (e.g. Combes & Wiklind 1996). A sample of the molecular lines detected is shown in Fig. 2.

Figure 2. Examples of molecular absorption lines detected in the millimeter range. Lines can be extremely narrow (below 1km/s) up to quite broad (100km/s). More than 20 different molecules or transitions have been detected in one single absorption system. Here the continuum sources are B3 1504+377 (Wiklind & Combes 1996b), PKS 1830-211 (Wiklind & Combes 1996a), PKS 1413+357 (Wiklind & Combes 1997) and B0218+357 (Wiklind & Combes 1995). The signal has been normalised to the continuum level detected.

These molecular absorption objects are the continuation at high column densities (10²¹-10²⁴ cm^-2) of the whole spectrum of absorption systems, from the Ly forest (10¹²-10¹⁹ cm^-2) to the damped Ly and HI 21cm absorptions (10¹⁹-10²¹ cm^-2). It is currently thought that the Ly forest originates from gaseous filaments in the extra-galactic medium, that the damped and HI absorptions involve mainly the outer parts of spiral galaxies. The molecular absorptions concern the central parts of galaxies. The properties of molecular absorptions detected in the millimeter domain so far, are summarised in Table 2.

Table 2. Properties of molecular absorption line systems in the mm

Source	zaa	zeb	NCO	NH2	NHIe	AV'c	NHI / NH2
			cm-2	cm-2	cm-2
Cen-A	0.00184	0.0018	1.0 x 1016	2.0 x 1020	1.0 x 1020	50	0.5
3C 293	0.0446	0.0446	3.0 x 1016	6.0 x 1020	1.2 x 1021	-	0.5
PKS1413+357	0.24671	0.247	2.3 x 1016	4.6 x 1020	1.3 x 1021	2.0	2.8
B31504+377A	0.67335	0.673	6.0 x 1016	1.2 x 1021	2.4 x 1021	5.0	2.0
B31504+377B	0.67150	0.673	2.6 x 1016	5.2 x 1020	< 7 x 1020	<2	<1.4
B0218+357	0.68466	0.94	2.0 x 1019	4.0 x 1023	4.0 x 1020	850	1 x 10-3
PKS1830-211A	0.88582	2.51	2.0 x 1018	4.0 x 1022	5.0 x 1020	100	1 x 10-2
PKS1830-211B	0.88489	2.51	1.0 x 1016d	2.0 x 1020	1.0 x 1021	1.8	5.0
PKS1830-211C	0.19267	2.51	< 6 x 1015	< 1 x 1020	2.5 x 1020	<0.2	>2.5
aRedshift of absorption line bRedshift of background source cExtinction corrected for redshift using a Galactic extinction law dEstimated from the HCO+ column density of 1.3 x 1013cm-2 e21cm HI data taken from Haschick & Baan (1985, 3C293) and Carilli et al. 1992, 1993, 1998

A systematic survey for absorption lines has also been undertaken in front of about a hundred continuum sources candidates, selected from flat-spectrum continuum sources. The continuum needs to be at least 0.2 Jy to allow detection of intervening molecular gas. The redshift of the absorbing candidate is known, either from previously detected HI absorption, or from optical lines emission. When the continuum source is strong enough, at least 1 Jy, and no redshift is known, it is possible to search for absorption lines by scanning in frequency (cf. Wiklind & Combes 1996a). This last method is the most promising with the new generation millimeter instruments, that will gain an order of magnitude in sensitivity. Indeed, the best candidates are the most obscured ones, where no redshift is available.

A lot more absorption systems could be found with the future instruments, at faint continuum flux, since the number counts of quasars are a non-linear function of flux and the local luminosity function is steep (see Peacock 1985). However, the flat-spectrum radio-loud quasars distribution is decreasing sharply at z > 3 (Shaver et al. 1996), and this is not likely due to obscuration, since they are radio-selected quasars. If quasars are associated to galaxy formation and interactions, this tends to show that the decrease of star-formation rate beyond z = 3 is real and not an extinction effect. The relative absence of strong continuum radio sources at high redshift will not help the tracing of protogalaxies through absorption techniques.