Most of the matter in our universe is invisible. We can measure the gravitational pull of this “dark matter” on the orbits of stars and galaxies. We can see the way it bends light around itself and can detect its effect on the light left over from the primordial plasma of the hot big bang. We have measured these signals with exquisite precision. We have every reason to believe dark matter is everywhere. Yet we still don’t know what it is.
We have been trying to detect dark matter in experiments for decades now, to no avail. Maybe our first detection is just around the corner. But the long wait has prompted some dark matter hunters to wonder whether we’re looking in the wrong place or in the wrong way. Many experimental efforts have focused on a relatively small number of possible identities for dark matter—those that seem likely to simultaneously solve other problems in physics. Still, there’s no guarantee that these other puzzles and the dark matter quandary are related. Increasingly, physicists acknowledge that we may have to search for a wider range of possible explanations. The scope of the problem is both intimidating and exhilarating.
At the same time, we are starting to grapple with the sobering idea that we may never nail down the nature of dark matter at all. In the early days of dark matter hunting, this notion seemed absurd. We had lots of good theories and plenty of experimental options for testing them. But the easy roads have mostly been traveled, and dark matter has proved more mysterious than we ever imagined. It’s entirely possible that dark matter behaves in a way that current experiments aren’t well-suited to detect—or even that it ignores regular matter completely. If it doesn’t interact with standard atoms through any mechanism besides gravity, it will be almost impossible to detect it in a laboratory. In that case, we can still hope to learn about dark matter by mapping its presence throughout the universe. But there is a chance that dark matter will prove so elusive we may never understand its true nature.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
On a warm summer evening in August 2022 we huddled with a few other physicists around a table at the University of Washington. We were there to discuss the culmination of the “Snowmass Process,” a year-long study that the U.S. particle physics community undertakes every decade or so to agree on priorities for future research. We were tasked with summing up the progress and potential of dark matter searches. The job of communicating just how many possibilities there are for explaining dark matter, and the many ideas that exist to explore them, felt daunting.
We are at a special moment in the quest for dark matter. Since the 1990s thousands of investigators have searched exhaustively for particles that might constitute dark matter. By now they’ve eliminated many of the simplest, easiest possibilities. Nevertheless, most physicists are convinced dark matter is out there and represents some distinct form of matter.
A universe without dark matter would require striking modifications to the laws of gravity as we currently understand them, which are based on Einstein’s general theory of relativity. Updating the theory in a way that avoids the need for dark matter—either by adjusting the equations of general relativity while keeping the same underlying framework or by introducing some new paradigm that replaces general relativity altogether—seems exceptionally difficult.
The changes would have to mimic the effects of dark matter in astrophysical systems ranging from giant clusters of galaxies to the Milky Way’s smallest satellite galaxies. In other words, they would need to apply across an enormous range of scales in distance and time, without contradicting the host of other precise measurements we’ve gathered about how gravity works. The modifications would also need to explain why, if dark matter is just a modification to gravity—which is universally associated with all matter—not all galaxies and clusters appear to contain dark matter. Moreover, the most sophisticated attempts to formulate self-consistent theories of modified gravity to explain away dark matter end up invoking a type of dark matter anyway, to match the ripples we observe in the cosmic microwave background, leftover light from the big bang.
The scope of the dark matter problem is both intimidating and exhilarating.
In contrast, positing a new type of matter that simply doesn’t interact with light is a simple idea. In fact, we already have an example of such dark matter in the form of neutrinos—nearly massless particles that are ubiquitous but rarely interact with other matter. It’s just that we already know neutrinos can’t account for most of the dark matter in the universe. At most, they can make up about 1 percent of it.
So what about the other 99 percent? Could dark matter be the tip of an iceberg of discovery, the first revelation of one or more new particles that aren’t part of the Standard Model of particle physics? Could dark matter feel new forces that the known particles do not (in the same way that dark matter doesn’t appear to feel electromagnetic forces), or could it be linked to new fundamental principles of nature? Could dark matter solve outstanding puzzles lurking within the well-measured physics of the Standard Model, or could it reveal the earliest moments of the universe’s history? Right now the answers to all these questions are a definitive “maybe”—but the potential power of such a discovery drives us onward.
Two of the most popular proposals for the identity of dark matter are the weakly interacting massive particle (WIMP) and the axion of quantum chromodynamics (QCD). These ideas have shaped how theorists think about dark matter and inspired many of the experiments searching for it.
WIMPs are hypothetical stable particles with masses comparable to those of particles in the Standard Model. A proton’s mass is just under 1 GeV/c², and most WIMP searches have focused on the mass range between 10 and 1,000 GeV/c². (Particle physicists find it convenient to measure masses in units of energy using Einstein’s E = mc2.) The classic version of a WIMP is a new particle that interacts directly with the W and Z bosons known to carry the weak nuclear force (hence the “W” in WIMP). Such particles appear naturally in models of supersymmetry, where every known particle also has a heavier counterpart called a superpartner. A decade and a half ago our field hoped that the Large Hadron Collider near Geneva would find superpartners, but we had no such luck. If supersymmetry exists, the superpartners must be heavier than we’d initially expected. Furthermore, although many versions of supersymmetry predict WIMP dark matter, the converse isn’t true; WIMPs are viable dark matter candidates even in a universe without supersymmetry.
Tim M. P. Tait and Jen Christiansen
One of the reasons many physicists love the WIMP idea is that these particles naturally would have generated the same amount of dark matter in the universe that we observe. As the thinking goes, when the cosmos was much smaller, denser and hotter than it is now, even weak interactions were enough to produce WIMPs when known particles collided. And a similar reaction happened in reverse—when WIMPs collided, they created regular particles. If the big bang hadn’t produced WIMPs originally, the known particles would have made them. And collisions of WIMPs that transmuted their energy into known particles would have destroyed most WIMPs, leaving only a residual abundance. A WIMP with a mass around that of the Higgs boson would produce the correct amount of dark matter, for instance. This mechanism is simple and appealing.
WIMPs appeal to many experimentalists because they must interact significantly with the known particles—that’s how they arrive at the right amount of dark matter. There are three classic ways to search for WIMPs: collider experiments, where we hope to reproduce the conditions of the early universe by colliding Standard Model particles together to generate dark matter; direct-detection experiments, which use extremely sensitive detectors to look for visible particles “jumping” when they are struck by a dark matter particle; and indirect detection, where we look out into space to search for familiar particles being produced when dark matter particles collide and annihilate one another. The third approach in particular tests exactly the same destructive processes that would have set the abundance of WIMPs in the universe. Therefore, if these reactions behave in the same way today as they did in the early universe, we have a definitive prediction for how often they occur. For the first two approaches, the predictions are not so clear-cut. In collider searches, our ability to detect WIMPs depends on how heavy they are: more massive WIMPs may require more energy to produce than the collider has available. And in direct detection, we don’t know how often WIMPs will bump into regular particles.
Astrophysical observations—indirect detection—have revealed several signals that might be hints of dark matter annihilation, but there are also more mundane explanations for what we see. For example, the Galactic Center GeV Excess is a glow of gamma-ray light from the heart of the Milky Way; it has the right rate and the right energy to be a WIMP-annihilation signal. It was discovered in 2009, so why haven’t we declared victory? Unfortunately, we know that certain spinning neutron stars can produce gamma rays at similar energies, and it’s quite possible that the excess is the first sign of a new population of such stars. We hope this question will be resolved in the coming years: finding a counterpart signal in a direct-detection or collider experiment would support the dark matter interpretation, whereas finding radiation from the neutron stars at other wavelengths would rule it out.
In the next decade or so future large gamma-ray telescopes (such as the Cherenkov Telescope Array being built in Chile and Spain and the Southern Wide-field Gamma-ray Observatory planned for somewhere in South America) could test the WIMP mechanism for producing dark matter up to the highest masses where it is viable. Yet even if we don’t observe dark matter annihilation, there are loopholes to save WIMP theory. In some models, the annihilation process that created WIMPs in the early universe switches off at later times. In those cases, however, WIMPs should generally still show up in collider experiments and direct detection.
If we ask "What could dark matter be?" the possibilities are nearly endless.
Direct-detection experiments have made amazing progress in improving their sensitivity to rare events. Within 10 years the next generation of experiments could be so sensitive that they will start detecting neutrinos from the sun streaming through the detector. Until we reach that point, there are no other processes that could masquerade as dark matter, and no seemingly insurmountable technical challenges stand in the way. There are still many simple WIMP models that could show up in this range.
The QCD axion is a very different type of dark matter candidate, and until recently we haven’t had nearly the same ability to test it. Like the WIMP, it would be a new fundamental particle, though much tinier: axions are far lighter than any known particle, even neutrinos. If these particles exist—whether they make up all the dark matter or not—they could resolve long-standing puzzles in our understanding of the strong force, which holds atomic nuclei together. Plus, axion theories make distinct predictions: if you know the mass of the axion, you can estimate how strongly it interacts with the known particles. Unfortunately, those interactions depend on the axion mass and can be exceedingly weak for the lighter axions.
Still, axion interactions could have striking effects because to account for dark matter they would have to be so plentiful that they would manifest as a wave rather than as individual particles. According to quantum mechanics, every fundamental particle is also a wave and has an associated wavelength inversely proportional to its mass. At scales smaller than this wavelength, the classical picture of a particle breaks down. Axions are so light that we could expect to see such quantum effects over distances comparable to the typical size of an experiment on Earth.
Because of how weakly QCD axions are expected to interact with regular matter, fewer experiments have looked for them, and they have searched in only a tiny fraction of the possible mass range. New detection strategies and quantum sensor technologies, however, have opened up prospects for hunting the QCD axion over many orders of magnitude in mass. The latest version of a long-running experiment called ADMX-G2 is extremely sensitive, and upcoming projects such as DMRadio promise to greatly extend the search.
Over the next decade dramatic experimental advances will test both the WIMP and the QCD axion over the bulk of their natural mass range for the first time. The theoretical groundwork has been laid, and the plans for experiments are in place. We could leave it at that—there’s a good chance that these strategies will give us the solution.
And yet ... even though the WIMP and the axion are beautiful ideas, there is no guarantee that the universe conforms to our aesthetic preferences. And if we ask, “What could dark matter be?” the possibilities are nearly endless.
An entire landscape of theories manages to describe everything that dark matter needs to do to explain the universe, but each invokes different particles and forces to make it happen. Theorists have thoroughly mapped out which ideas have a hope of working and which ones are inconsistent with observations. Many of the viable hypotheses are surprisingly different from WIMPs or axions. Some, for instance, include massive aggregate objects composed of many tinier constituents—akin to dark matter atoms composed of different dark particles.
There is a limit to how small dark matter particles can be. If they were much lighter than axions—about 25 orders of magnitude lighter than the mass of the electron—their wavelengths could be close to the size of star clusters or small galaxies. If this were the case, the distribution of dark matter and its gravitational footprint would be observably different.
What about the other end of the mass scale? The smallest clumps of dark matter we can directly observe are tens of millions of times the mass of the sun. Individual dark matter particles should be smaller than that, but how much smaller? If dark matter were made of dense, dark objects—often called massive compact halo objects (MACHOs), as a tongue-in-cheek contrast to WIMPs—then their gravity could deflect light and disrupt orbits as they barreled through the galaxy in ways we could see. MACHOs could take the form of tiny black holes, born in the first moments after the big bang. These black holes would not form from stars—because dark matter predates stars—and could be much lighter than the sun. The only way these black holes could account for all of the dark matter would be if they had about the same mass as the asteroids in our solar system, between around 100 billion and 100,000 trillion metric tons. That would give them individual masses one one-thousandth of the mass of the moon, making them 75 orders of magnitude heavier than the smallest possible dark matter particles. (For comparison, the ratio between the radius of our observable universe and the radius of a proton is only about 41 orders of magnitude.) That’s quite a lot of ground to cover.
And in the vast region between these two extremes, we have a plethora of options. The process that could produce WIMPs in the early universe would also work for many other particles. If dark matter were lighter than a proton and born through this mechanism, it could be just one of many new particles inhabiting a “dark sector” of physics. These other particles would generally be unstable, so there would be very few of them out in space. Yet they could show up in particle accelerators, especially if they were also relatively light. Light dark matter and dark sectors could also exist without relying on the WIMP mechanism to produce the right amount of dark matter—there are myriad other possibilities for how to generate the observed abundance of dark matter.
If the dark sector is out there, we need new experimental methods to find it. Classic WIMP detectors, for instance, lose sensitivity once the dark matter is much lighter than atomic nuclei because they look for a strong “kick” on nuclei by incoming dark matter. New technology can seek signs of electrons (which are 2,000 times lighter than protons) being kicked instead or use even more creative strategies to detect tiny energy transfers from dark matter to standard particles. The recent advent of ultrasensitive quantum sensors could help.
The only way we know to search over such a wide range of possibilities is to build many small experiments, each sensitive to different types of dark matter, rather than focusing our resources on a few huge projects. We can also use these small experiments to develop new technologies and try out novel ideas; if one of those strategies proves powerful or detects something that could be an initial hint of dark matter, we could then scale it up.
Indirect-detection searches in space already span a vast range of energy scales. If the dark matter were slowly decaying into visible particles, with a typical lifetime as long as a billion times the current age of the universe, we would know it by now for many possible dark matter masses. We can test primordial black holes, for instance, with this kind of search; this is how we know that if black holes make up all the dark matter, they can’t be lighter than about 100 billion metric tons (lighter black holes decay faster).
And even if we don’t see a signal, we’ll continue to learn more about dark matter by mapping its gravity in space. Current and upcoming instruments will measure the distribution of stars and distant galaxies with fantastic precision and depth. Developments in precision cosmology and artificial intelligence are driving techniques to help us glean as much as we can from these data. Such observations could provide new clues to the fundamental nature of dark matter that will complement what we can learn in the lab.
After all the Snowmass discussions, the physics community opted to embrace a balanced strategy. We plan to delve deeply into our favorite theories of dark matter while also searching widely (at a shallower level) to explore as many possibilities as we can.
If we are lucky, one of these experiments will make a clear detection. Once that happens, it will trigger a paradigm shift. The broad and varied search will collapse to focus on that signal, and we’ll plan future experiments to better understand it. A discovery would also prompt theorists to study the bigger picture of how to connect dark matter with the rest of the particle zoo we’re familiar with.
But what if none of these experiments finds a signal? Perhaps physicists at the next Snowmass Process, about a decade hence, will have to use null results to chart the direction for future searches. We can’t deny this outcome would be disappointing, but it would still count as a major achievement. Science moves forward one step at a time, and the results that teach us where not to look for the next insights are just as important as those that confirm a particular idea is correct. If we could predict with certainty what dark matter will turn out to be, it would mean that we already know the answer, making our jobs much less exciting. And although we can’t say exactly when or even whether we’ll find dark matter, we know that the universe is filled with it. We’re optimistic that the next years of our quest will lead us to a deeper understanding of what it is.