Integral field spectroscopy of (U)LIRGs and Post-Starburst QSOs : the role of mergers in galaxy evolution

Abstract

Major gas-rich mergers are thought to be responsible of a dramatic evolution of galaxies. Attending to the changes in the morphology, Kormendy & Sanders (1992) propose that mergers of gas rich disks (U/LIRGs) lead to the formation of elliptical galaxies. Due to the large timescales needed to complete the merger, we cannot study this process following the transformation of an individual system. Instead, we need to compare the properties of systems along the different stages of the merger sequence. However, identifying all the intermediate stages of a merger and placing them in the right evolutionary order it is not an easy task. Moreover, apart from the morphological changes, Sanders et al. (1988) propose there is also a change in the ionizing mechanism along the merger sequence, from dusty starbursts in the initial stages (U/LIRGs) to unobscured QSO activity in the final ones. While it is quite clear that mergers lead to massive starburst activity, sometimes hidden by dust, and therefore, U/LIRGs (Surace et al. 2000; Veilleux et al. 2002; Kim et al. 2013), it is not so clear if the QSOs activity is merger triggered and if so, when exactly during the merger it starts. The best approach is to find systems with intermediate properties in the proposed sequence, that is, galaxies hosting bright nuclear QSOs, showing morphological disturbance typical of a past merger and with an important post-starburst population, relics of the intense starburst activity triggered in the past. The best example of this kind of systems are post-starburst QSOs (Brotherton et al., 1999; Cales et al. 2011, 2013, 2015), which have been proposed to be the missing evolutionary link. They are, therefore, ideal laboratories to study connections between the starburst and AGN phenomenon.

In this thesis we have characterized and compared the star formation histories, average stellar population properties, star formation rates (SFR), and ionized gas properties, in two small samples of galaxies in different stages across the merger sequence, three LIRGs (the two pre-mergers IC 1623 and NGC 6090 and the merger NGC 2623) and nine PSQSOs, by analysing high quality Integral Field Spectroscopy (IFS) data in the rest-frame optical range 3700 - 7000 Å, and high resolution HST imaging. Additionally, for NGC 2623 we have narrow band imaging in Halpha and [NII]6583 from OSIRIS@GTC Tunable Filters, that allow us to study the outer parts of the tidal tails. The results from the LIRGs and PSQSOs have been compared with control Sbc and Sc galaxies from CALIFA survey (González Delgado et al. 2015). The methodology applied has been the same for all, a full spectral fitting analysis was performed using the Starlight code (Cid Fernandes et al. 2005) with a combination of single stellar population (SSP) models from the literature (González Delgado et al. 2005, Vazdekis et al. 2010).

With our data we find a evolutionary sequence related to the merger progression that is in agreement with recent high resolution simulations (Teyssier et al. 2010). In the initial stages (pre-mergers) the induced star formation (SF) is extended, and enhanced, on average, by a factor ~4 with respect to the control spirals. When we resolve it spatially we find that for IC 1623 W the enhancement is the same in the central region and in the disk (by a factor 7), while for NGC 6090 NE the enhancement is higher in the central region (by a factor 9), still significant at one half light radius (by a factor 5) and less significant in the disk (only enhanced by a factor 1.5). Star formation is the dominating ionization mechanism in both pre-mergers.

Attending to the importance of stellar populations ~300 Myr, we find that the merger-induced star formation started earlier in IC 1623 W than in NGC 6090 NE. Moreover, in general, the SF is significantly enhanced in this period in the 3 LIRGs in comparison to spirals. In addition, while the specific SFR profile decline for the spirals below 1HLR, due to the presence of the bulge that inhibites the SF, for the LIRGs the profile remain flat, indicating that their central regions are still growing. In more advanced mergers, as NGC 2623, we find that most of the young SF is concentrated in the central region, enhanced by a factor 9 with respect to control spirals. However, there exists also lower level of star formation in the outer parts, enhanced by < 3 in comparison to spirals. From the global average across the whole galaxy we find that is a factor ~3 higher than in spirals. In addition, in NGC 2623 we detect fossil emission of an extended merger-induced burst occurred ~1.5 Gyr ago, probably when it was at the pre-merger stage. The mass formed during this first burst is enhanced by a factor 2 both in the center and in the outer parts with respect to the mass formed in the same period in normal spirals. In this case, apart from star formation, shocks ionization due to the central outflow is also important, as revealed by the LINER condition in the BPT diagnostic diagram. From the average of the nine PSQSOs we find that, both in terms of light and mass, the SFHs are comparable to NGC 2623 ones. Attending to the average ages we found that PSQSOs are ~400 Myr older than NGC 2623. It seems that PSQSOs are slightly more evolved, however, given the uncertainties related to the model base choice, and the heterogeneity of the sample, we can not confirm exactly to what degree. We note, however, that both present a significant contribution to mass of stellar populations younger than 1.5 Gyr (of ~6.6 x 10^9 Msun in NGC 2623 and ~10^10 Msun in the PSQSOs, Chabrier IMF), which is not present in the pre-merger LIRGs. However, we find that if pre-merger LIRGs keep forming stars at the current rate ~16 Msun/yr during ~500 Myr, then they can account for the mass in intermediate age stellar populations in NGC 2623 and PSQSOs, consistent with the maximum duration of merger triggered starbursts from numerical simulations (Di Matteo et al. 2008). The dominating ionization mechanism in PSQSOs is Seyfert, although we note that for ~1/3 (~3 out of 9) there exist also significant young star formation (from 32 - 300 Myr), as found from the spectral synthesis. Finally, given the stellar masses measured in LIRGs and PSQSOs, most of them <10^11 Msun, we find that they will form ellipticals of intermediate mass (~10^11 Msun), or S0s lenticular galaxies, in agreement with the major-merger evolutionary scenario and recent results from CALIFA and Galmer database (Hopkins et al. 2008, Querejeta et al. 2015). It is also possible that they are forming the core of future giant ellipticals. Although we do not have evidences, recent observational results and cosmological simulations suggest that dry merging is necessary to grow giant ellipticals with masses well above >>10^11 Msun (Kormendy et al., 2009; Toft et al., 2014, Cappellari et al. 2013).