Black Holes Grow Better With A Friend

Title: Low-mass Galaxy Interactions Trigger Black Hole Activity

Authors: Mićić, M., Irwin, J. A., Nair, P., Wells, B. N., Holmes, O. J., and Eames, J. T.

First Author’s Institution: Homer L. Dodge Department of Physics and Astronomy, The University of Oklahoma, Norman, OK 73019, USA

Status: Accepted to ApJ Letters [open access]

Most galaxies have supermassive black holes at their centres. Over time, these black holes grow by eating up lots of gas and stars, and occasionally merging with another black hole. As telescopes have improved, and especially with the recent launch of the James Webb Space Telescope, astronomers have discovered supermassive black holes that are more massive than we originally thought black holes should be in the early universe (you can read more about black hole growth here and here). This observation implies that black holes can grow very quickly during the first 700 million years of the universe’s existence.

But understanding this early phase of black hole growth is no easy task. In the early universe, the majority of galaxies were low-mass, dim dwarf galaxies that are difficult to observe, even with our strongest telescopes. One way that astronomers can overcome this challenge is by studying nearby dwarf galaxies that we think resemble early-universe dwarf galaxies.

When a black hole is growing and eating a lot of gas, we call it an active galactic nucleus, or AGN. The process also releases a huge amount of energy, making an AGN very bright, especially in X-rays. Astronomers think that black holes build up a lot of their mass during these AGN phases, but it’s unclear what triggers the beginning of an AGN phase and whether AGN triggers are more common in the early universe. 

One possibility is that galaxy mergers trigger AGN phases. Studies have found evidence that a higher percentage of galaxies that have recently undergone mergers also host an AGN, supporting this idea. However, these studies focus on galaxies that are much more massive than early universe galaxies, making it difficult to say whether we can extrapolate these results and apply them to low-mass dwarf galaxies. 

Four panels showing images of dwarf galaxy systems. Each panel is labelled with an ID number (clockwise from top left: id36, id49, id60, and id59).
Figure 1: Four dwarf galaxy systems found by today’s authors. The red boxes show the galaxies that they’ve determined to be close to each other. The original images are from the 3D-HST survey. Adapted from Figure 2 of today’s survey.

The authors of today’s paper set out to answer the question of whether galaxy mergers could trigger enough AGN activity to explain the existence of supermassive black holes in the early universe. To do so, they studied pairs of dwarf galaxies that are much closer to us, but might be representative of early universe galaxies. 

To find these elusive pairs of dwarf galaxies, the authors used data from the 3D-HST survey, which was conducted using the Hubble Space Telescope and observed over 200,000 galaxies. From this huge catalogue, they selected galaxies with at least ten times fewer stars than the Milky Way. They then inspected the images taken of these galaxies to see whether there was at least one other dwarf galaxy within 100 kiloparsecs (that’s around 3×1018 km!). Using this method, they identified 93 systems that contained at least two nearby dwarf galaxies. As you can see in Figure 1, these dwarf systems vary in terms of the number of galaxies in the system and the distance between each galaxy.

For each dwarf galaxy in a pair or group, the authors also identified a galaxy at a similar distance from us with a similar mass but without any nearby neighbours. This sample of isolated galaxies acted as the control sample against which the authors could compare their results. 

To determine whether each galaxy was hosting an AGN, the authors crossmatched their galaxies with X-ray data taken by the Chandra X-ray Observatory. Since the AGN in these galaxies are expected to be fairly faint, very high-quality X-ray data is needed in order to definitively say whether a galaxy has an AGN or not. As a result, the authors were only able to check for AGN in 29 of their dwarf galaxy pair systems and found 7 AGN within this subsample. In the control sample, they only found 3 AGN in a sample of 183 galaxies. 

A scatter plot with AGN frequency on the y-axis and median log stellar mass on the x-axis. Red, green, yellow, purple, and black points span the entire mass range and are at low AGN frequency. Two blue points have a much higher AGN frequency and represents the results of the paper.
Figure 2: The AGN frequency (the two blue points) for interacting dwarf galaxies is much higher than the AGN frequency for the control sample (black point) and the AGN frequency reported in previous papers (red, yellow, green, and purple points). Figure 5 in today’s paper.

Since the dwarf galaxy pairs are fairly close to each other, Chandra is unable to separate out the light from each galaxy and the authors are unable to say whether both dwarf galaxies have AGN, or just one of the two. If every single galaxy in the sample hosts an AGN, then the AGN fraction might be as high as 16.4%, but if only one of the pair hosts an AGN, then the fraction may be as low as 9.8%. Regardless, both of these fractions are significantly higher than the AGN fraction in the control sample, which was just 1.6%, and is generally higher than the AGN fractions reported in other papers (see Figure 2). 

The results of today’s paper support the idea that dwarf galaxy mergers might be a trigger of AGN activity. Since the majority of galaxies in the early universe are dwarf galaxies, this might result in a higher AGN fraction, which could help to explain how black holes are able to grow so quickly in the early universe. Further X-ray observations of the dwarf galaxy pairs identified by the authors will help to confirm this result and better constrain the AGN fraction in dwarf galaxy mergers, so stay tuned! 

Astrobite edited by Archana Aravindan

Featured image credit: Mićić et al., 2024.

Author

Submit a Comment

Your email address will not be published. Required fields are marked *