From 'something is there' to 'what is it?'
In the previous guide we collected the case that mass is missing — flat galaxy rotation curves, galaxies swarming too fast inside clusters, and foreground mass bending background light far more than the visible stuff can account for. Every one of those is a gravity measurement, and together they are about as solid as anything in astrophysics. But notice the limit of what they prove: they tell us *how much* unseen mass is out there and *where* it sits, and nothing at all about *what it is made of*. This guide takes the next, harder step — from the fact of the missing mass to the question of its identity.
The honest starting point is humility. Dark matter is, at this moment, a name for our ignorance: a placeholder for whatever supplies the extra gravity. Nobody has ever held a gram of it, seen it glow, or caught it in a detector. What we *can* do is interrogate the gravity it produces, ask what kind of stuff is consistent with all of it at once, and then go hunting for that stuff. Remarkably, the clues already rule out almost every easy answer and point, with surprising sharpness, at something genuinely new.
Two words that narrow the field: non-baryonic and cold
Start with the most tempting boring answer: maybe the missing mass is just ordinary matter that happens not to shine — faint stars, cold gas clouds, free-floating planets, dead stellar remnants. Call all of that baryonic matter, meaning it is built from protons and neutrons, the same stuff as you and the Sun. The problem is that we have two independent ledgers of exactly how many baryons the universe contains, and both come up far short. The first is Big Bang nucleosynthesis: in the universe's first few minutes, the amount of helium, deuterium and lithium that got forged depended sensitively on the density of protons and neutrons. The measured abundances pin that density tightly — and it is only a small fraction of the total mass we need.
The second ledger comes from the relic glow of the early universe and the pattern of its faint temperature ripples — the acoustic peaks. Their relative heights act like a fingerprint that reads out, separately, how much *ordinary* matter and how much *total* matter the universe holds. The two numbers disagree by roughly a factor of five. Both methods reach the same verdict from completely different physics: ordinary baryonic matter simply cannot be the bulk of the missing mass. Whatever dark matter is, it must be non-baryonic — outside the familiar roster of protons and neutrons entirely.
The second word is cold, and it comes from the way galaxies were built. In the early universe, dark matter's gravity gathered the first clumps that ordinary gas later fell into to make galaxies. If the dark particles had been born moving near light-speed — 'hot' — they would have streamed freely out of small clumps and smeared them away, so the smallest structures would have formed first only on huge scales, then fragmented. Instead the universe built itself bottom-up: small galaxies first, merging into big ones. That demands particles that were already sluggish early on, so their gravity could pull the smallest seeds together. We call such slow matter cold dark matter, and it is the 'CDM' baked into our standard model of the cosmos.
The leading suspects
The recipe — non-baryonic, cold, neutral, barely interacting — is so specific that no particle in the known Standard Model fits it. So theorists look to particles that other ideas predict for entirely separate reasons, and ask whether one of them could double as dark matter. Three suspects lead the field. The longtime favourite is the WIMP, a Weakly Interacting Massive Particle: a heavy particle, perhaps tens to hundreds of times a proton's mass, that feels only gravity and the weak nuclear force. Its appeal is an elegant accident called the 'WIMP miracle' — a particle with roughly the weak force's strength, left over from the hot early universe by freeze-out, would naturally survive in just about the right abundance to be the dark matter. That coincidence made WIMPs the hunt's prime target for decades.
The second suspect is the axion — a feather-light particle, perhaps a trillion times lighter than an electron, first dreamt up not for cosmology at all but to fix an unrelated wrinkle in the theory of the strong nuclear force. It would have been produced in vast, slow-moving numbers in the early universe, which makes it count as cold despite its tiny mass. The third is the sterile neutrino, a heavier, shy cousin of the ordinary neutrinos you met in the Sun guide; 'sterile' because it would ignore even the weak force, touching our world through gravity and the faintest of mixings. Each suspect is well motivated, each is testable, and — this matters — they are radically different from one another, which is exactly why we need experiments to choose between them rather than taste.
The recipe and the leading suspects:
REQUIRED: non-baryonic + cold + neutral + barely interacting
candidate rough mass scale how it's cold
--------- ---------------- ------------
WIMP ~10-1000 x proton heavy + freeze-out
axion ~1e-12 x electron born slow, in bulk
sterile neutrino keV-ish (light, heavy
for a neutrino) mass keeps it slow
None is in the known Standard Model -> dark matter = new physics.Three ways to catch a ghost
If dark matter is a particle streaming through us right now — and if the halo is real, billions pass through your body each second — then in principle we can catch one. The search splits into three complementary strategies, and it helps to picture each. The first is direct detection: build an exquisitely quiet detector, often a vat of liquid xenon, bury it a kilometre underground to hide from cosmic-ray noise, and wait for the once-in-a-blue-moon nudge of a passing particle bouncing off an atomic nucleus, depositing a whisper of energy. This is the front line of dark-matter detection, and the experiments have grown to tonnes of target while seeing — so far — nothing.
The second strategy is indirect detection. Even if dark particles never touch our detectors, two of them might occasionally meet in space and annihilate, or one might slowly decay, producing a faint spray of ordinary particles — gamma rays, positrons, neutrinos — that we *can* see. So telescopes stare at places where dark matter should pile up densely: the galactic centre, dwarf galaxies, the cores of clusters, looking for an excess glow that ordinary sources cannot explain. The third strategy is collider production: instead of waiting for dark matter to come to us, try to *make* it. Slam protons together at enormous energy and watch for events where momentum and energy go missing, carried off invisibly by a freshly made dark particle that sails out of the detector unseen.
Axions need their own cleverness, since they are far too light to bump a nucleus. The trick is that a strong magnetic field can coax an axion to convert into a faint photon, so an 'axion haloscope' is essentially a tuned radio looking for a whisper of microwave light at exactly the right frequency. The crucial, honest fact across all of this: after decades of ever-more-sensitive searches on every front, no experiment has produced a confirmed dark-matter signal. That is not failure — it is how the field narrows the possibilities, ruling out swathes of WIMP territory and pushing the hunt toward axions and lighter, stranger candidates.
Or is gravity the thing we have wrong?
There is a serious, respectable alternative we owe a fair hearing. What if no mass is missing at all, and instead our law of gravity is slightly wrong in the regime of the very feeble accelerations found in a galaxy's outskirts? That is the proposal of modified Newtonian dynamics, or MOND: below some tiny threshold of acceleration, gravity is tweaked to fall off more gently than Newton's inverse-square. With one new constant, MOND reproduces the flat rotation curves of individual galaxies remarkably well — in places more crisply, and with fewer adjustable knobs, than dark matter manages. If galaxies were the only evidence, this would be a genuine contest.
But galaxies are not the only evidence, and this is where most cosmologists are won over to real particles. Three things MOND struggles with, dark matter handles naturally. It cannot fully account for the speeds of galaxies inside large clusters without adding *some* unseen mass anyway. It has no clean explanation for the acoustic peaks in the early-universe glow, whose heights demand a non-baryonic component that does not push back against light. And most vividly, there is the Bullet Cluster: two galaxy clusters caught mid-collision, where the hot gas — most of the ordinary matter — was slowed and left lagging in the middle, while gravitational lensing shows the bulk of the mass sailed straight on through with the nearly collisionless galaxies, cleanly separated from the gas.
So where does that leave us? With a sharp recipe and no confirmed dish. The missing mass is almost certainly non-baryonic and cold; the front-running candidates are WIMPs, axions and sterile neutrinos; the searches are ingenious and, so far, empty-handed; and a minority of careful people still bet on modified gravity. None of that is settled. What *is* settled is that the visible universe — every star, planet and person we have spent this whole subject learning — is a thin frosting on a much larger, darker cake. The next guide turns from the matter we cannot see to something stranger still: a dark *energy* that is not pulling galaxies together but pushing the whole universe apart, faster and faster.