Observations of two faint galaxies with a low abundance of elements heavier than helium show that the galaxies have an efficiency of star formation less than one-tenth of that of the Milky Way and similar galaxies. See Letter p.335
Star formation is well studied in bright galaxies such as the Milky Way, where it occurs by localized gravitational collapse in dense cold clouds. The clouds are mostly composed of gaseous atomic and molecular hydrogen and helium, with a small fraction of the clouds' mass being in the form of dust particles made from condensed 'metals' (elements heavier than helium). Trace amounts of carbon, oxygen and other elements also make heavy molecules, such as carbon monoxide. Dust is important for star formation because it prevents most starlight from getting inside the clouds, allowing heavy molecules to radiate away their heat and cool to the point at which gravity overcomes gas pressure and stars are made. Astronomers know much less about star formation in small faint galaxies, which is the subject of the paper by Shi and colleagues1 on page 335 of this issue.
Small galaxies have a lower abundance of heavy elements (metallicity) than large galaxies, and their average gas densities are also low, making gravity inside the clouds relatively weak and dust absorption of starlight relatively small. As a result, star formation seems to be slow, as Shi et al. report. The basic difference between small and large galaxies is their rotation speed, which is related to mass. The Milky Way rotates rapidly around its centre, with a speed of more than 200 kilometres per second2. Rapid rotation means that gravity binds our galaxy tightly, that the speed at which material can escape the galaxy is high, and that metal-rich debris from stellar winds and supernova explosions gets trapped in the interstellar medium. This debris is the material that has been processed by nuclear reactions inside stars and is the origin of all heavy elements. After a cosmic age of stars forming and dying, the trapped metals add up to a few per cent of the total mass in gas and stars for galaxies the size of the Milky Way. This is enough for the resulting dust to block starlight from cloud interiors and allow molecules to form, cool and collapse into stars3.
By contrast, the two galaxies studied by Shi and co-workers, Sextans A (Fig. 1) and ESO 146-G14, have low rotation speeds (23 and 70 km s−1, respectively) and masses that are only 0.2% and 13% of the Milky Way's mass4,5. These galaxies are too tiny to have trapped most of their heavy elements from a lifetime of supernovae, and indeed the abundance of heavy elements in these galaxies relative to hydrogen is less than 10% of that in the Milky Way6,7. Weak gravity also means that they have low gas pressures, on average. As a result, we do not expect molecules to form in dense clouds, and so the presence of young stars in these galaxies is a puzzle.
Shi and colleagues' study bypasses the molecules and looks for the associated dust instead. The problem with low-metallicity galaxies is that their molecular gas is very difficult to observe. Molecular hydrogen at the low temperatures required for star formation does not emit radiation efficiently, and carbon monoxide, the next most abundant molecule, is rare when both carbon and oxygen are rare8. The dust mixed with the gas can be detected, however, because it radiates in the infrared regime of the electromagnetic spectrum, at wavelengths between 10 and 1,000 μm. Detection requires a large telescope on a satellite because Earth's atmosphere absorbs most infrared light and makes cosmic sources nearly invisible from the ground. Shi et al. combined observations of dust from the Herschel and Spitzer infrared space telescopes with observations of regions that contain hot young stars from an ultraviolet space observatory, the GALEX Space Telescope, to determine the dust masses and star-formation rates in Sextans A and ESO 146-G14. The authors also derived the mass of atomic hydrogen from archival ground-based radio observations9,10.
The main result of Shi and colleagues' study is that there is much more infrared light than would be expected for the atomic hydrogen and star-formation rates that are present in these two galaxies. More infrared means that there is more dust than anticipated, and much more gas considering the low abundance of heavy elements there. Moreover, this gas has to be molecular because not enough atomic hydrogen is observed. Shi et al. conclude that a large mass of unseen molecules has to be present near the observed regions of star formation. However, then there is a problem with the rates at which star formation occurs, which should be ten times larger than they are if the efficiency of star formation, the rate per molecule, is the same as in the Milky Way and similar galaxies. The reasons for these peculiarities are unknown. Previous models11,12 for low star-formation rates in such galaxies were based on molecules being prevented from forming in the first place, but that is apparently not happening in Sextans A and ESO 146-G14.
Astronomers should understand much more about molecular clouds and star formation at low levels of metallicity in the next few years. A new interferometric telescope, the Atacama Large Millimeter/submillimeter Array (ALMA) in Chile, was designed to detect faint emission from molecules and dust13. Now in its third year of observations, ALMA is powerful enough to map even rare molecules such as carbon monoxide at the sparse elemental abundances of faint galaxies. This would allow the temperatures, densities and motions of the star-forming gas to be determined from spectral signatures of the molecules.
Sextans A and ESO 146-G14 are examples of what galaxies might have looked like in the first billion years after the Big Bang. Even future Milky Way-like galaxies were small then, and had relatively few heavy elements14. Understanding star formation in the smallest galaxies of our own backyard may give us considerable insight into the earliest star formation in the Universe.
Shi, Y. et al. Nature 514, 335–338 (2014).
van der Kruit, P. C. & Freeman, K. C. Annu. Rev. Astron. Astrophys. 49, 301–371 (2011).
McKee, C. F. & Ostriker, E. C. Annu. Rev. Astron. Astrophys. 45, 565–687 (2007).
Filho, M. E. et al. Astron. Astrophys. 558, 18 (2013).
Karachentsev, I. D., Karachentseva, V. E., Huchtmeier, W. K. & Makarov, D. I. Astron. J. 127, 2031–2068 (2004).
Kniazev, A. Y. et al. Astron. J. 130, 1558–1573 (2005).
Bergvall, N. & Rönnback, J. Mon. Not. R. Astron. Soc. 273, 603–614 (1995).
Elmegreen, B. G. et al. Nature 495, 487–489 (2013).
Ott, J. et al. Astron. J. 144, 123 (2012).
Peters, S. P. C. et al. Preprint at http://arxiv.org/abs/1303.2463 (2013).
Ostriker, E. C., McKee, C. F. & Leroy, A. K. Astrophys. J. 721, 975–994 (2010).
Krumholz, M. R. Mon. Not. R. Astron. Soc. 436, 2747–2762 (2013).
Hand, E. Nature 495, 156–159 (2013).
Mannucci, F. et al. Mon. Not. R. Astron. Soc. 398, 1915–1931 (2009).