Particle Accelerators
Discovery Machines
Why Collide Particles
In everyday experience, mass and energy feel like different things - a bowling ball versus a beam of light. But Einstein's E=mc2 revealed they are two forms of the same underlying quantity. Energy can change form, and when enough of it concentrates in a small enough space, what emerges can look like new particles of matter. When particles collide at tremendous speeds, their kinetic energy reorganizes into particles that do not normally exist in everyday life. Higher energy means heavier particles can emerge. It is the only way to produce many of these exotic forms of matter.
Nearly every particle discovered since 1930s was found either in high-energy collisions or in cosmic ray events. Cosmic rays arrive from space at random times, at random energies, hitting random targets. Accelerators let physicists choose energy, choose particles, and run experiments billions of times. They are controlled cosmic rays, engineered for precision. Instead of waiting for nature, physicists built machines to recreate conditions not seen since earliest moments after Big Bang.
You may notice the protons in the animation look like flat discs rather than spheres. This is not artistic license. At over 99.999999% of the speed of light, which is how fast protons travel inside LHC, special relativity compresses them along their direction of motion. From the laboratory's perspective, each proton is Lorentz-contracted into a pancake more than 7,000 times thinner than it is wide. Two ultra-thin discs slamming into each other at nearly light speed - that is what a collision at 13.6 TeV actually looks like.
How They Work
Accelerators use electric fields to push charged particles faster and faster. Radiofrequency cavities deliver energy in precisely timed pulses, each one giving particles a small kick at exactly the right moment. Picture a child on a swing: push at the peak of each arc and the swing climbs higher. These cavities do the same for protons circulating at nearly light speed.
Magnets shape the beam. Dipole magnets bend particle trajectories into a circular path, the principle behind synchrotrons. Quadrupole magnets squeeze the beam tighter, focusing trillions of particles into a stream thinner than a human hair. At Large Hadron Collider, superconducting magnets cooled to 1.9 Kelvin, colder than outer space, generate fields of 8.3 Tesla. Particles circulate millions of times per second, gaining energy with every lap, until they reach collision energy.
You might wonder: why protons and not electrons? LHC could in principle accelerate electrons, and some earlier machines did exactly that. But there is a catch. When charged particles travel in a circle, they radiate energy - and lighter particles radiate far more. An electron, roughly 2,000 times lighter than a proton, loses energy so rapidly in a circular machine that most of the input power goes into radiation rather than acceleration. Protons, being heavier, bend around curves with far less energy loss, making them practical for reaching the highest energies. This is also why future designs consider muons, which are 200 times heavier than electrons but still fundamental particles, offering a compelling middle ground.
Collision and Detection
Two beams travel in opposite directions and cross at designated interaction points. At LHC, protons collide at a combined energy of 13.6 TeV, roughly the kinetic energy of a flying mosquito concentrated into a space a trillion times smaller than a grain of sand. Each collision produces a shower of particles spraying outward in all directions. Detectors must catch every one.
Detectors are built in concentric layers, each designed for a different task. Innermost tracking chambers record curved paths of charged particles in a magnetic field. Surrounding calorimeters absorb particles and measure their energy. Outermost muon chambers catch muons, the only charged particles that punch through everything else. Billions of collisions happen every second. Sophisticated trigger systems select only the most interesting events, roughly one in a million, for storage and analysis.
Finding Needles in Haystacks
LHC produces roughly one billion collisions per second. Each collision generates thousands of particles, and each particle leaves traces across multiple detector layers. The raw data rate is staggering - about one petabyte per second, more than the entire internet's traffic. No computer system on Earth could record all of it. So physicists built a multi-stage trigger system that makes split-second decisions about which collisions are worth keeping. The first trigger, built into custom hardware chips, scans events in microseconds and discards 99.99% of them. A second software-based trigger examines the survivors more carefully and keeps roughly one thousand events per second for permanent storage.
Even after this brutal filtering, the stored data is enormous. CERN's computing grid spans over 170 facilities across 42 countries, collectively processing and analyzing the data. Finding a Higgs boson means identifying a handful of distinctive events among trillions of ordinary collisions. Modern machine learning algorithms have become essential for this task, trained to recognize subtle patterns in detector data that would be nearly impossible for human eyes to spot. The 2012 Higgs discovery relied on statistical analysis of billions of events, looking for a tiny excess in the data at a specific energy - a bump in a graph that changed the course of physics. Today, deep neural networks sift through collision data in real time, flagging anomalies that might hint at particles or phenomena nobody predicted.
What Accelerators Have Found
History of particle physics is a history of accelerators confirming what theorists predicted. Carl Anderson discovered positron in 1932, the first antiparticle, in cosmic ray observations. Bevatron at Berkeley produced antiproton in 1955, proving antimatter was not limited to electrons. Deep inelastic scattering experiments at SLAC in 1968 revealed quarks hiding inside protons, settling a debate that had lasted years.
CERN's Super Proton Synchrotron found W and Z bosons in 1983, carriers of weak force, confirming electroweak unification. Fermilab's Tevatron discovered top quark in 1995, heaviest known fundamental particle. And in 2012, LHC delivered its greatest prize: Higgs boson, the particle tied to the mechanism that gives mass to the W, Z, and the fundamental fermions like quarks and electrons. (Photons and gluons stay massless either way, and most of the weight you carry around is actually quark-binding energy inside protons rather than Higgs mass.) Each discovery confirmed theoretical predictions, sometimes decades after they were first written down.
Beyond Discovery
Accelerators do far more than discover particles. Hospitals use antimatter every day in positron emission tomography, where positrons annihilate with electrons inside a patient's body to map metabolic activity. Synchrotron light sources produce intense X-ray beams used for protein crystallography, materials science, and even archaeology, revealing hidden layers in ancient manuscripts without touching them.
Proton therapy targets cancer tumors with a precision that conventional radiation cannot match, depositing most energy right at tumor depth while sparing healthy tissue. Ion implantation, a form of low-energy acceleration, is essential for manufacturing semiconductors in every phone and computer. Technology developed for fundamental research has reached into medicine, industry, and daily life in ways its inventors never imagined.
Next Generation
The Large Hadron Collider is roughly at the energy ceiling of its tunnel. Pushing further requires either a much larger machine or a fundamentally different design. CERN's proposed Future Circular Collider would span 91 kilometers and reach 100 TeV, nearly eight times LHC energy, but at a cost of tens of billions of euros and a construction timeline measured in decades. The International Linear Collider, the Compact Linear Collider, and the Circular Electron Positron Collider in China are competing electron-positron proposals, each with different physics goals.
Plasma wakefield acceleration takes a completely different approach. A laser or particle beam punches through a plasma, creating a wake of electric fields a thousand times stronger than any conventional cavity. Particles surf this wake, gaining in centimeters what traditional machines need kilometers to achieve. It is still experimental, and beam quality remains a challenge, but it could reshape accelerator design in coming decades.
The Muon Shot
The most ambitious idea on the table is the muon collider, and it carries the loudest divide between optimism and skepticism in particle physics. The promise is straightforward. A muon is a heavier cousin of the electron, about 207 times heavier. Unlike a proton, it is a point particle, so when two muons collide the full energy of the collision goes into producing new particles instead of being diluted across the partons that make up a proton. Unlike an electron, the muon is heavy enough that bending it around a ring does not bleed away most of its energy as synchrotron radiation. A relatively modest 10-kilometer ring could in principle reach 10 TeV in the centre of mass, comparable to a proton-proton collider many times its size.
The catch is the lifetime. A muon at rest survives for 2.2 microseconds before it decays into an electron, an electron antineutrino, and a muon neutrino. Every step of the machine has to happen before then, in the muon's own clock. Time dilation helps at relativistic energies, but only by factors of hundreds, not infinitely. So in the same fraction of a second, the muon collider has to create muons from a proton-on-target reaction, capture them, cool the resulting cloud into a tight beam, accelerate that beam to TeV energies, deliver it into a storage ring, and let it collide. The decay products are a constant background in every detector. The whole machine is racing the clock.
The technical bottleneck is the cooling step. Muons emerge from the production target spread over a wide range of energies and angles, an effectively hot gas in beam terms. To bring them to the brightness a collider needs, the beam has to be cooled, and standard cooling techniques used for protons or electrons all take far longer than a muon lives. The solution, called ionisation cooling, is to pass the muon beam through a low-density absorber that drains transverse momentum, then re-accelerate it back up to its original longitudinal energy. Repeated many times, the procedure narrows the beam without losing too many particles. The principle was demonstrated by the Muon Ionization Cooling Experiment at Rutherford Appleton Laboratory in the UK, which published its results in Nature in 2020. MICE proved cooling works. A full six-dimensional cooling channel at the brightness an actual collider needs has not yet been built.
The decisive political event in the muon-collider story was the December 2023 P5 report, the recommendation document that sets US particle-physics priorities for the next decade. The report formally placed a muon collider at the centre of US ambitions, calling the project a "muon shot" and recommending that the US join the International Muon Collider Collaboration. The IMCC, headquartered at CERN with a globally distributed engineering effort, had been formalised in 2020 and produced an interim design report in July 2024 outlining a path to a 10 TeV machine. A Department of Energy and CERN agreement to formalise national-lab participation is, as of 2026, in active drafting. The current published target is a demonstrator facility in the 2030s and a full machine in the late 2040s, with substantial caveats on funding.
The unsolved engineering challenges are real and large. The full six-dimensional cooling channel has to be designed, built, and demonstrated at brightness. The muon production rate has to be increased by orders of magnitude beyond what current proton drivers deliver. The detectors have to be redesigned to live inside a constant rain of decay products from muons circling around them at a rate of one decay per muon per microsecond. And there is a stranger problem the public is not used to thinking about: at TeV-scale muon energies, the neutrinos produced when muons decay form a tightly collimated beam that, after travelling kilometres through the rock around the accelerator, surfaces with enough flux to constitute a measurable radiation dose. The muon collider would essentially shine a faint neutrino beam at the population downstream of it. This is not science fiction; it is the subject of multiple working papers in the IMCC and a constraint on where such a machine could legally be built.
Whether the muon collider ever runs is genuinely uncertain. The engineering is harder than anything CERN has ever attempted, and several plausible alternatives, including the FCC and various wakefield proposals, are competing for the same constrained budget. But the physics reach is hard to argue with. If the next major discovery in fundamental physics lives in the 1-10 TeV regime, a muon collider is one of very few realistic ways to look there. The next decade of demonstrator work will decide whether the dream becomes the machine.
The Bigger Picture
Accelerators are how we ask universe its deepest questions. Every fundamental particle in the Standard Model was found by colliding things together at higher and higher energies. Yet the Standard Model accounts for only about 5% of what exists. Dark matter, dark energy, the mechanism behind neutrino mass, the origin of matter-antimatter asymmetry, none of these have answers yet. The next generation of colliders may find them. Or they may find something nobody predicted, which historically is how the biggest breakthroughs have happened.
