Remarkable gravitational lensing by the galaxy cluster Abell 3827

Most of the matter in the universe, according to the standard cosmological model, is invisible stuff whose nature is unknown. Dark matter interacts gravitationally but as far as we know is not subject to the electromagnetic or any other interaction. Gravity is enough, however, to ensure that a spherical halo of dark matter surrounds a galaxy’s shining stars. Evidence for dark matter has accumulated from several sources; they include the anisotropy of the cosmic microwave background, galaxy-rotation data, and the gravitational lensing of light from galaxies by unseen masses.

In 2006 a pair of colliding galaxy clusters, together called the bullet cluster, provided a spectacular confirmation of dark matter. The two clusters had passed through each other 100 million years ago, and an analysis of how the bullet cluster distorted the images of background galaxies established that its intergalactic gas lagged behind its dark-matter halos. The proffered explanation was that gas–gas interactions dragged the conventional material while the halos traveled unimpeded. Indeed, observations of the bullet cluster enabled researchers to set an upper limit on how strongly dark matter can interact with itself (see Physics Today, November 2006, page 21 ).

Now a team of astrophysicists led by Durham University’s Richard Massey has spotted a galaxy in the Abell 3827 cluster that appears to be leading its halo as both are attracted toward the cluster center. ¹ Massey and colleagues suggest that the offset may result from frictional interactions between the halo and the dark matter of the cluster core.

Finding halo

The four galaxies in Abell 3827 shown in the figure lie within a few kiloparsecs of each other (a parsec is a bit more than 3 light-years). Partially surrounding them is an arc that manifests the severe distortion of a background galaxy by the gravity of Abell 3827; additional lensing can be seen in the area of the galaxy labeled N1. With the help of spectroscopic measurements obtained by the Very Large Telescope in Chile, Massey and colleagues associated each of the 30 lensed images labeled in the figure to the core of the lensed galaxy (denoted by Ao) or to one of the galaxy’s six bright, star-forming regions (labeled Aa–Af). Armed with those identifications, the research team turned to a pair of independent computer models to map the locations of the lensing dark-matter halos in the core of Abell 3827 (not shown). They found N1 to be significantly offset from its halo, by 1.6 ± 0.5 kpc in the cluster-image plane, with the luminous matter closer to the cluster center.

Massey’s group was not the first to see a halo–galaxy offset in Abell 3827, nor the first to suggest that it might be a manifestation of dark-matter self- interactions. Those honors go to Liliya Williams (University of Minnesota) and Prasenjit Saha (University of Zürich), ² who have joined Massey in the more recent work. Assuming that the offset was entirely due to dark-matter self- interactions, Williams and Saha obtained a lower bound for the interaction strength σ per unit mass m. Massey and company used the same assumptions and with their offset obtained an estimate of σ/m = (1.7 ± 0.7) × 10⁻⁴ cm²/g. The deduced interaction strength depends on the duration of the galaxies’ movement to the cluster center; the team’s value of σ/m used a ballpark estimate of 10⁹ years. The interaction strength is expressed in particle physicists’ conventional cross-section units; by way of comparison, the cross section for hydrogen gas is roughly σ/m = 10⁸ cm²/g.

An assertion that dark-matter self-interactions have been unambiguously observed would require extraordinary evidence, and Massey and company do not claim to have made an ironclad case. The challenge of determining experimental uncertainties and the modeling required to obtain the halo locations are formidable, and even given those locations, the researchers note that “interpreting an offset between mass and stars is difficult.” ¹ The combined effects of matter along the line of sight to Abell 3827 and conventional physics in the complex environment of the cluster could somehow be responsible for the inferred misalignment of dark and luminous matter. Detailed simulations in the future should help clarify whether frictional dark-matter interactions exist.

Furthermore, the dark-matter interaction model used to obtain σ/m is greatly simplified. Indeed, within a couple of weeks of the publication of the Massey work, a team led by Felix Kahlhoefer (German Electron Synchrotron) considered a more sophisticated model of dark-matter self-interaction. ³ The theorists concluded that if the displacement observed by Massey and company is totally due to unconventional dark-matter physics, then σ/m = 1.5–3 cm²/g, a value high enough to strain the upper bounds deter- mined from the bullet cluster and other observations.