Online material

Appendix A: Comments on individual molecules

We have detected seventeen molecular species, and seven carbon, sulphur, and oxygen isotopic substitutions in the 3 mm atmospheric window, between 86.2 GHz and 115.6 GHz. Here, we detail each observed transition, the Gaussian profile-fitting, cases of blended lines, and other special cases. To identify the molecules, we started from the premise that only the transitions having E_low < 100 cm^-1 and log    (^∫T_MB   dv) >  −6 nm² MHz contribute to the observed lines. After checking this initial condition, we ensured that the spectral features that do not fulfil these conditions can be considered negligible.

NGC 1068 has a heliocentric systemic velocity of v_sys = 1137 km s^-1 (value taken from NASA/IPAC Extragalactic Database – NED). Based on the values obtained from the Gaussian fits, we have chosen an average value of 1100 km s^-1 for fixing this parameter when needed. We found two different line widths for the detected molecules in NGC 1068. In general, strong lines have values in the range FWHM ~  [230−260]  km s^-1, while faint lines have line widths of about  ~[120−190]  km s^-1. The cases where it was necessary to fix this parameter are discussed below, since the value depends on each species.

• Hydrogen cyanide – HCN and H¹³CN. We detected theH¹³CN (1 − 0) transition at 86.3 GHz.It is formed by three hyperfine spectral features, although thestructure of the line is not resolved owing the broad line widthsof NGC 1068. To take the contribution ofeach hyperfine feature into account we performed a syntheticGaussian fit (details about this method can be seen in Martínet al. 2010; Aladroet al. 2011b). The resulting values,shown in Table B.1, correspond to the main com-ponent of the group. This HCN carbon isotopologue is detectedfor the first time in this galaxy. HCN (1 − 0) is identified at 88.6 GHz. In a similar way as for H¹³CN, we fitted a synthetic Gaussian profile leaving the linewidth and line position free.

• Hydrogen isocyanide – HNC and HN¹³C. One strong line of hydrogen isocyanide was detected at 90.7 GHz. The telescope beam possibly picked up some emission coming from the starburst ring that surrounds the CND of NGC 1068. We complemented this transition with the HNC (4 − 3) line observed by Pérez-Beaupuits et al. (2009). The HN¹³C(1 − 0) transition at 87.1 GHz has been tentatively detected for the first time in this galaxy. Because it is quite faint, the integrated intensity resulting from the Gaussian fit is highly affected by the baseline. This is reflected in the column density derived from the rotational diagram, which has a high associated error.

• Ethynyl – C₂H. C₂H (1 − 0), formed by six hyperfine features, is detected at 87.3 GHz. We performed a synthetic Gaussian fit to calculate the contribution of each feature. Because we only have one detected transition of this species, we calculated its column density by fixing the rotational temperature to 10 ± 5 K. This is the fourth most abundant molecule after CO and its carbon and oxygen isotopologues.

• Isocyanic acid – HNCO. Four groups of transitions were detected at 87.9 GHz, 109.5 GHz, 109.9 GHz, and 110.3 GHz. They are composed of five features each. The line at 109.9 GHz is seen as a bump at the higher frequencies of C¹⁸O, although both can be easily separated. The transition at 110.3 GHz is very close to ¹³CO and blended with CH₃CN(6_4,0 − 5_4,0). We performed a synthetic Gaussian fit to all these lines, including two other HNCO transitions that are too faint to be detected, but that fulfil the conditions mentioned at the beginning of this appendix (E_low < 100 cm^-1 and log    (^∫T_MB   dv) >  −6 nm² MHz). In the Gaussian fitting we fixed the linewidth to 230 km s^-1 and the line position to 1100 km s^-1. The integrated area of our HNCO (5_0,5 − 4_0,4) line is consistent with the values obtained by Martín et al. (2009). However, from the Boltzmann diagram we obtain a rotational temperature of T_rot ~ 30 K, which is almost five times higher than the one obtained by Martín et al. (2009). This difference is due to the transitions with the higher energy levels we included.

• Oxomethylium – HCO⁺ and H¹³CO⁺. The strong line of HCO ⁺    (1 − 0) at 89.2 GHz is the only detected transition in our line survey, and shows some emission coming from the starburst ring. We also used the HCO ⁺ (3 − 2) line observed at 267.6 GHz by Krips et al. (2008). The H¹³CO ⁺ (1 − 0) line is detected at 86.8 GHz, although it is blended with HCO(1_1,0 − 0_1,0) and SiO(2 − 1). To disentangle the contribution of each molecule, we performed synthetic Gaussian fits to the three lines. We first fitted H¹³CO ⁺ (1 − 0) by fixing the position to 1100 km s^-1 and the linewidth to 240 km s^-1. Then, we fitted the two other molecules (see Fig. A.1).

• Oxomethyl – HCO. Only one molecular transition of this species has been detected during the survey. It is blended with H¹³CO ⁺    (1 − 0) and SiO (2 − 1) (see Figs. 2 and  A.1). HCO (1_1,0 − 0_1,0) is located at 86.7 GHz, and contains four features. After subtracting the oxomethylium Gaussian fit, we over-fitted a Gaussian profile to the HCO features with a fixed position of 1100 km s^-1 and a linewidth of 240 km s^-1.

• Silicon monoxide – SiO. As discussed before, SiO (2 − 1) at 86.8 GHz is blended with H¹³CO ⁺ (1 − 0) and HCO(1_1,0 − 0_1,0). After subtracting the Gaussian fits of the other two molecules, we fitted another synthetic Gaussian profile to the residuals, fixing only the position to 1100 km s^-1 (see Fig. A.1). This is the strongest molecule of the three blended lines.

  	Fig. A.1 Gaussian fits to the blended lines of H13CO+ (1 − 0), HCO (11,0 − 01,0), and SiO (2 − 1). See more details in the text.
  Open with DEXTER

• Methyl cyanide – CH₃CN. This symmetric top molecule is detected for the first time in NGC 1068. The two CH₃CN (5_k − 4_k) and CH₃CN (6_k − 5_k) transitions at 92.0 GHz and 110.4 GHz lie inside our frequency coverage. Each of them are formed by 25 spectral features with the quantum number k going from 0 to J − 1. Although faint, both lines are strong enough to be detected with an S/N ≥ 3. The transition at 110.4 GHz is blended with HNCO (5_1,4 − 4_1,3). We fitted a synthetic Gaussian profile to both lines, fixing the line position of CH₃CN to 1100 km s^-1 and the linewidth to 150 km s^-1 (based on the non-blended line). Its rotational temperature is similar in NGC 1068 (T_rot ~ 29 K) and M 82 (T_rot ~ 33 K).

• Sulphur monoxide – SO. Three transitions of sulphur monoxide were detected for the first time in NGC 1068 at 99.3 GHz, 100.0 GHz, and 109.252 GHz. The first one was not blended, and its Gaussian parameters have been used to fit the other two. The line is narrow, 136 km s^-1. The other two lines are blended with HC₃N(11 − 10) and HC₃N(12 − 11), respectively. The contribution of SO was firstly subtracted, and then we fitted cyanoacetylene to the residuals. The SO (5₄ − 4₄) at 100.0 GHz does not fit well in the rotational diagram, so the values we show in Table 1 do not take this transition into account.

• Hydroxymethylidynium – HOC⁺. We detected the HOC ⁺ (1 − 0) transition at 89.5 GHz. Its peak intensity temperature is slightly above the noise, so we consider it as a tentative detection. This line shows a very narrow linewidth, of about 128 km s^-1. Fixing the rotational temperature to 10 ± 5 K, we obtained the lowest column density of the molecular survey, with N_HOC ⁺  = (1.1 ± 0.6) × 10¹³ cm^-2.

• Cyanoacetylene – HC₃N. This molecule is detected for the first time in NGC 1068. We identified HC₃N (10 − 9) at 91.0 GHz, HC₃N (11 − 10) at 100.1 GHz, and HC₃N (12 − 11) at 109.2 GHz. The transition at 91.0 GHz was not blended, so we used its Gaussian parameters to fix the other two, which were blended with SO. For those lines, we first subtracted the sulphur monoxide features (as explained before), and then fitted two synthetic Gaussian profiles to the residuals. The HC₃N rotational temperature in NGC 1068 (7.3 ± 1.4 K) is significantly lower than the ones found for M 82 and NGC 253 (~25 K and  ~ [12–73] K, respectively, Aladro et al. 2011a).

• Diazenylium – N₂H⁺. This is the first time that this species is detected in NGC 1068. Only the N₂H ⁺ (1 − 0) transition, at 93.2 GHz, lies inside our survey. It is formed by three features, so we performed a synthetic Gaussian fit in order to calculate the contribution of each one. We fixed the rotational temperature to 10 ± 5 K to calculate its column density.

• Carbon monosulfide – CS and C³⁴S. We detected the CS (2 − 1) line, which is blended with ¹³CO (1 − 0) coming from the image band. Since these two are the only transitions of these molecules detected in our survey, it is not possible to disentangle the contribution of each one. Thus, no Gaussian fit was done. Nevertheless, we used the CS (3 − 2), CS (5 − 4), and CS (7 − 6) lines observed by Bayet et al. (2009b) to do the rotation diagram. Two gas components with different temperatures and column densities are seen. C³⁴S (2 − 1) was detected at 96.4 GHz. This transition was complemented with the C³⁴S (3 − 2) line observed by Martín et al. (2009). The rotational temperature is one of the lowest obtained in the survey, with only  ~ 4 K.

• Carbon monoxide – CO, C¹⁸O, and ¹³CO. CO (1 − 0), at 115.3 GHz, was blended with NS (3₁ − 2₁), but since NS was, comparatively speaking, very faint, we assumed that its contribution is negligible. Since the emission of CO (1 − 0) is quite spread in the nucleus of NGC 1068, it clearly shows a deviation from a Gaussian profile, which points to the complex gas dynamics in that region, as well as a possible contribution by the starburst ring emission. We fitted a three Gaussian profile, which is shown in Fig. A.2. We also used the CO (2 − 1), CO (3 − 2), and CO (4 − 3) lines from Israel (2009). Like CS, CO shows two components in the rotation diagram. Even though CO (1 − 0) is the strongest feature in the survey, it is not seen in the image band. This is due to the variable rejection along the 3 mm band. ¹³CO (1 − 0) was detected at 110.2 GHz. Similarly to ¹²CO (1 − 0), this transition showed a deviation from a Gaussian profile. It was complemented with the ¹³CO (2 − 1) and ¹³CO (3 − 2) lines observed by Israel (2009). Two different components are also seen in the Boltzmann diagram. As discussed before, ¹³CO  (1 − 0) does appear in the image band, blended with CS (2 − 1). C¹⁸O (1 − 0) appears at 109.8 GHz. It was not blended, but it showed a bump at the higher frequencies due to the proximity of an HNCO transition. To avoid the isocyanic acid contamination, we fixed the linewidth to 230 km s^-1 and then fitted three Gaussians, which accounts for the line profile better. We also used the C¹⁸O (2 − 1) line observed by Martín et al. (2009) for complementing our observations. The rotational temperature is the lowest one among all the molecules detected (T_rot = 3.3 ± 0.1 K).

• Cyanogen – CN. Two separated lines of cyanogen were identified at 113.2 GHz (CN (1_0,1 − 0_0,1)) and 113.5 GHz (CN (1_0,2 − 0_0,1)). The first one is formed by four different features, while the second one has five. We fitted a synthetic Gaussian profile to both lines. We included other CN transitions observed by Pérez-Beaupuits et al. (2009). We also detected two multi-feature lines of ¹³CN at 108.4 GHz and 108.6 GHz.

  	Fig. A.2 Line profile of the CO (1 − 0) line. Y-axis is on a Tmb scale.
  Open with DEXTER

• Methanol – CH₃OH. Only the CH₃OH (2_k − 1_k) line is seen at 96.7 GHz, which is composed of four features with the quantum number k varying between  − 1,0,1. However, we calculated the integrated intensities of other transitions that lie inside our frequency range, which fulfil the conditions imposed on the energy and logarithm of the integrated intensities, as explained at the beginning of this appendix. Considering all these features (thirteen in total), we performed a synthetic Gaussian fit.

• Nitrogen monosulphide – NS. We detected the NS (3₁ − 2₁) transition in our survey. It is split into two separated multi-lines with six features each. The first one is blended with CO (1 − 0), while the second one is located at higher frequencies, so that it can be easily separated from CO. We first subtracted the CO contribution, and then fitted a synthetic Gaussian profile to all the NS features. We assumed T_rot = 10 ± 5 K to calculate the column density.

	Fig. A.3 Boltzmann diagrams of the molecules observed in the survey for which we have detected more than one transition, or for which we have taken other transitions from the literature (marked with *). The resulting rotational temperatures are indicated.
Open with DEXTER

Appendix B: Gaussian fits parameter results

© ESO, 2012