WE SEE its effects in how stars move within galaxies, and how galaxies move within galaxy clusters. Without it, we can’t explain how such large collections of matter came to exist, and certainly not how they hang together today. But what it is, we don’t know.
Welcome to one of the biggest mysteries in the universe: what makes up most of it. Our best measurements indicate that some 85 per cent of all matter in our universe consists of “dark matter” made of something that isn’t atoms. Huge underground experiments built to catch glimpses of dark matter particles as they pass through Earth have seen nothing. Particle-smashing experiments at the Large Hadron Collider, which we hoped would create dark matter, haven’t – at least as far as we can tell. The hunt for dark matter was never supposed to be easy. But we didn’t expect it to be this hard.
Advertisement
Find out more about dark matter at
Dark matter’s no-show means that many possible explanations for it that people like me favoured just a decade ago have now been ruled out. That is forcing us to radically revisit assumptions not only about the nature of dark matter, but also about the early history of our universe. This is the latest twist in a long-running saga: our failure to detect the particles that make up dark matter suggests that the beginning of the universe may have been very different from what we imagined.
Let’s start with what we do know about this substance – or perhaps substances. Dark matter isn’t familiar atomic matter, or any of the exotic forms of matter created at the Large Hadron Collider, buried underground near Geneva, Switzerland, or at other particle accelerators. It doesn’t appreciably interact with itself, or with ordinary matter, except via gravity. It can pass through solid objects like a ghost, and doesn’t emit, absorb or reflect any easily measurable quantities of light. It is invisible, or at least nearly so.
CERN and Mont Blanc: Explore particle physics and glaciers in Switzerland on a żěè¶ĚĘÓƵ Discovery Tour
Yet without dark matter, it is unlikely that we would be here. As galaxies and galaxy clusters were built up, dark matter played the role of scaffolding: it gathered into enormous clouds whose gravity attracted and pulled together the atomic matter that would ultimately form the luminous bit of galaxies. Without the gravity of dark matter holding stars in place, they would fly outwards, in some cases escaping into intergalactic space. Many galaxies would simply disintegrate.
We see dark matter’s imprint in many other ways, too, for example in how a galaxy cluster’s gravity deflects light that passes it. Perhaps the best evidence of all for dark matter’s existence comes from temperature patterns observed in the cosmic microwave background, the radiation left over from the big bang. Measurements of this radiation provide us with a map of how matter was distributed throughout our universe only a few hundred thousand years after its beginning. This map tells us that our universe was very uniform in its youth, with only the smallest variations in density. Without help from dark matter, there is no way that these density variations could have grown fast enough to form the galaxies and other large structures of today’s universe.
A decade or more ago, many physicists, including me, thought we knew what dark matter was likely to consist of: weakly interacting massive particles, or WIMPs. As their name suggests, these are relatively heavy particles that, besides gravity, only interact via the weak nuclear force, which also governs sub-atomic processes such as radioactive beta decay. WIMPs seemed compelling because we could understand how they would have been created in the early universe.
During the first millionth of a second or so after the big bang, all of space was filled with a hot, dense plasma in which all sorts of particles, from photons and electrons to top quarks and Higgs bosons, were constantly being created and destroyed. As space expands, however, the temperature of the plasma steadily drops. Eventually, it can’t supply the energy required to make heavier particles, and their production stops.
When this happens to a species of particle, most are destroyed – annihilated – and converted into other forms of energy. How many survive depends on how and how often the particles interact.
This leads us to a happy coincidence: for a particle species to emerge from the big bang with an abundance equal to that of dark matter today, it must have interacted through a force about as powerful as the weak nuclear force. A stronger force would have caused too many particles to be destroyed, while a feebler force would have allowed too many to survive. Rather like the temperature of Goldilocks’s porridge, the strength of the weak force seems just right to explain how dark matter came to be formed in the heat of the big bang.
But that story now seems rather a fairy tale. If dark matter does consist of WIMPs, we can estimate how much it should interact with ordinary atomic matter via the weak force, and so design experiments to detect it. These experiments, housed in deep underground laboratories to avoid the constant bombardment of cosmic radiation, started out small, deploying detectors of only a few kilograms of crystalline materials such as germanium, calcium tungstate or sodium iodide, sensitive to the light, heat and electric charge that would be produced in collisions of WIMPs with normal matter.
Over the past two decades, the size and sophistication of these experiments has hugely increased. The latest iterations are enormous, deploying anything up to tonnes of liquid xenon as their detectors. These experiments – XENON1T under the Gran Sasso mountain in Italy (pictured, right), LUX in South Dakota and PandaX-II in Sichuan, China – are each roughly 10,000 times as sensitive as the most sophisticated dark matter detectors operating in 2006.
“The longer we go without finding WIMPs, the more we must confront the possibility they aren’t there”
But they too have failed to turn up WIMPs. The only experiment that even claims to have detected anything resembling dark matter goes by the name of DAMA. Most researchers think the signal it picked up is almost certainly produced by something else: a long list of other experiments have searched for the kinds of WIMPs that could have made it, but have seen nothing.
The only other possible piece of evidence we have for WIMPs comes in the form of a strange gamma-ray signal seen emanating from the centre of the Milky Way. My collaborators and I spotted this signal in data from NASA’s Fermi space telescope more than a decade ago. It took years for us to convince most people that it was real. We continue to debate whether these gamma rays are produced by dark matter, or by something else, such as a group of thousands of rapidly spinning neutron stars. At the moment, we just can’t be sure.
Whatever the resolution of that argument, the longer we go without directly detecting WIMPs, the more we are forced to confront the uncomfortable possibility that they might not be there. And yet dark matter must exist – alternative explanations, such as modifying gravity to produce the same sort of effects, don’t seem to work (see “A disturbance in the force“). If not WIMPs, then what?
A whole new world
One possibility is that dark matter could interact with other forms of matter and energy even less than we had imagined – perhaps only through gravity or some force so feeble that we haven’t even discovered it yet. Such a particle would be even more difficult to detect in underground experiments or to produce with particle accelerators.
The problem is that such non-interacting particles would probably survive the big bang in vast numbers, and wildly exceed the abundance of dark matter in our universe today. But if they interact rarely enough, perhaps these particles were never produced in great quantities in the first place, instead building up an appreciable abundance only gradually over the first fraction of a second of cosmic history.
It could be, too, that dark matter is just one of several kinds of particles that almost never interact with any known forms of matter and energy. This “hidden sector” of particles would involve forces and interactions that we have never observed, and that allow dark matter to evolve in a rich variety of ways. These interactions may have depleted the amount of dark matter, without leading to any appreciable interactions with ordinary matter.
The hidden-sector particles might become bound to each other, forming dark nuclei or dark atoms. One day, we could even discover something like a periodic table of the hidden sector elements. For that reason, of all the plausible ideas about dark matter that have grown in popularity in recent years, this is perhaps my favourite.
A disturbance in the force
Despite dark matter’s long-standing refusal to reveal itself, most physicists remain confident that it exists – the evidence in its favour is just too great. A few, however, champion a very different possibility. Rather than explaining the motions of stars around galaxies with new forms of matter, they speculate that a different conception of gravity may be the answer.
These ideas fall under the general umbrella of modified Newtonian dynamics, or MOND. This postulates that gravity works ordinarily here on Earth and in our solar system, but differently in the low-acceleration environments experienced by stars throughout the Milky Way and other galaxies.
In these circumstances, the force of gravity is effectively stronger than Newton or Einstein thought. This strengthening of gravity creates the illusion that unseen dark matter must be present.
Many versions of MOND have been proposed over the past few decades, but they have suffered from a range of problems, both observational and theoretical. Perhaps the single biggest failure is MOND’s inability to explain the temperature patterns observed in the cosmic microwave background, the relic radiation of the big bang. Whereas dark matter enables us to explain and understand the observed features of this light in incredible detail, no version of MOND has ever remotely done the same.
Compounding these problems is the fact that no version of MOND has been able to explain the observed dynamics of galaxy clusters.
Many of these alternative dark matter candidates call for experiments very unlike those designed to hunt for WIMPs. One example is the , based at the University of Washington in Seattle and managed by scientists at my institute, Fermilab. It uses powerful magnetic fields to try to convert one hypothesised type of ultra-light dark matter particle, axions, into photons.
Some physicists are trying to produce dark matter using particle beams originally designed to study neutrinos. Others are designing tunable electronic circuits that could pick up signals of dark matter waves, much like a radio picks up electromagnetic waves consisting of photons. There are even ideas involving gravitational wave detectors. These ideas may not seem to have much in common, but they are all motivated by testing previously overlooked possibilities for dark matter.
There is an even more dramatic possibility that many cosmologists are considering. Our surprise at dark matter’s no-show is based on our current understanding of the early universe. Maybe we haven’t seen the particles because dark matter is different from what we had expected – or perhaps because the universe’s first moments were.
The amount of dark matter that was created in and survived the big bang depends on how our universe evolved during its hot and volatile youth. We know a great deal about most of our universe’s 13.8-billion-year history, but we have no direct observations that enable us to study the first fraction of a second, the window in which dark matter is thought to have formed.
One possibility my colleagues Hooman Davoudiasl and Sam McDermott and I have investigated is that our universe experienced a brief period of hyper-fast expansion during this era. We already think it did something similar right at the beginning, in an event known as cosmic inflation.
Another – somewhat less explosive – burst of expansion may have occurred somewhat later as well, still within the first fraction of a second of cosmic history. It would have diluted the amount of dark matter in the early universe, and thereby changed our expectations for how strongly this substance should interact – and how difficult it should be for us to detect.
Alternatively, there may have been a population of particles that decayed at some point in the early universe, disappearing and creating dark matter. Dark matter particles created in this way could be extremely feebly interacting, explaining why they have gone undetected for so long.
A third possibility is that our universe went through an abrupt change during its first moments – not merely a steady cooling, but a total phase transition. We already know of two such transitions in which the nature of particles and their interactions changed within the universe’s first second, what are known as the QCD and electroweak phase transitions. But there may have been others. A phase transition in dark matter interactions could have influenced how dark matter formed in the early universe, again altering our expectations for the kinds of experiments that might detect it today.
It is too early to say whether the right answer is one, some or none of the above possibilities. Perhaps an experimental breakthrough will change the game yet again. But the stubborn elusiveness of dark matter has left many physicists and cosmologists surprised and confused. In droves, we are returning to our chalkboards, revisiting and revising our assumptions – and with bruised egos and a bit more humility, desperately attempting to find new ways to make sense of a very dark and hidden universe.
