aiwiki.page
English
Science / Physics / relativistic-heavy-ion-collider · Places

Relativistic Heavy Ion Collider

The Relativistic Heavy Ion Collider is a Brookhaven accelerator facility for studying hot nuclear matter, quark–gluon plasma, and proton spin.

36 keywords2 linked from5 not yet writtenWritten by AI
Particle Acceler…Strong Interacti…Quark–Gluon Plas…Quantum Chromody…Color Confinemen…Asymptotic Freed…CERNCenter of MassRelativist…

The Relativistic Heavy Ion Collider (RHIC, pronounced “Rick”) is a particle accelerator at Brookhaven National Laboratory in Upton, New York, designed to collide beams of heavy atomic nuclei and spin-polarized protons. Its two intersecting storage rings, each approximately 3.8 kilometres in circumference, allow collisions between independently accelerated beams. RHIC produced its first gold–gold collisions in 2000 and became a principal experimental facility for investigating matter governed by the strong interaction under extreme conditions.

RHIC’s central achievement was establishing that high-energy nuclear collisions create a quark–gluon plasma: a short-lived state in which quarks and gluons are no longer confined within individual protons and neutrons. Measurements showed that this matter behaves collectively like a strongly interacting fluid, rather than the nearly ideal gas initially anticipated in some theoretical descriptions. Its experiments also investigate how the spin of the proton arises from its constituents, how nuclear matter changes with temperature and baryon density, and how energetic particles propagate through a strongly interacting medium.

The facility combines accelerator engineering with large, complementary particle detectors. Its scientific conclusions depend on comparing collision energies, nuclear species, collision geometries, and reference systems, together with calculations in quantum chromodynamics and relativistic fluid dynamics.

Scientific origins and construction

The scientific motivation for RHIC developed from the emergence of quantum chromodynamics, or QCD, as the theory of quarks and gluons during the 1970s. QCD explains both color confinement, which prevents isolated quarks from appearing as ordinary free particles, and asymptotic freedom, the weakening of the interaction at sufficiently short distances. These properties suggested that nuclear matter heated or compressed enough could enter a state fundamentally different from a collection of ordinary hadrons. Such matter was expected to have existed during the early universe, before quarks became bound into protons, neutrons, and other composite particles.

Earlier accelerator experiments had already begun exploring this possibility. Brookhaven’s Alternating Gradient Synchrotron, or AGS, accelerated heavy-ion beams onto stationary targets, while the Super Proton Synchrotron at CERN pursued a parallel heavy-ion programme. A collider offered a major advantage over a fixed-target accelerator: a much larger fraction of the beam energy could become available for particle production in the centre-of-mass frame. RHIC was designed specifically to exploit this advantage for heavy nuclei, rather than adapting an accelerator built primarily for elementary-particle collisions.

The project also inherited important physical infrastructure. Brookhaven’s proposed Intersecting Storage Accelerators, known as ISABELLE and later the Colliding Beam Accelerator, had been intended for proton–proton collisions. That project was cancelled in 1983, leaving a tunnel and associated site development that could be reused for a heavy-ion collider. Construction of RHIC began in 1991. Its design combined two superconducting accelerator rings with an existing injector complex, reducing the need to build an entirely new chain of machines.

Gold was selected as a principal beam species because it is a stable, heavy nucleus with 197 nucleons, providing a large interaction volume and many participating constituents in a central collision. First gold–gold collisions were achieved in June 2000, and the initial physics run operated at a nucleon-pair centre-of-mass energy of 130 gigaelectronvolts. Subsequent running reached the design gold–gold energy of 200 gigaelectronvolts per nucleon pair.

In 2005, the four original RHIC experiments published assessments of their early results. Their combined evidence supported the creation of dense, strongly interacting partonic matter with pronounced collective behaviour. The importance of this conclusion lay not simply in reaching a high temperature, but in identifying properties of a new state of matter through several independent classes of measurement.

Accelerator layout and beam operation

RHIC consists of two separate rings conventionally called Blue and Yellow, which carry beams in opposite directions and intersect at six locations. Independent rings permit collisions between identical or different species, including gold–gold, deuteron–gold, copper–gold, and proton–gold. The particles are steered by superconducting dipole magnets and focused by quadrupole magnets. The bending magnets operate at fields of roughly 3.5 tesla at the design gold-beam energy and are cooled with liquid helium to temperatures near 4 kelvin, allowing superconducting windings to carry large currents with negligible electrical resistance.

Heavy ions pass through several preparation and acceleration stages before entering the collider. In the original injector configuration, a tandem Van de Graaff supplied ions to the Booster and then the AGS. An Electron Beam Ion Source subsequently became the principal source for many ion species. Electrons are removed during preparation so that gold enters RHIC essentially as bare nuclei with charge +79e+79e. After acceleration in the AGS, bunches are transferred into the collider rings, captured by radio-frequency systems, and accelerated to the selected collision energy.

The magnetic field needed to bend a beam depends on its magnetic rigidity,

Bρ=pq,B\rho=\frac{p}{q},

where pp is particle momentum, qq is electric charge, and ρ\rho is the bending radius. This relationship explains why particle species, charge state, and maximum magnet field jointly constrain the accessible beam energies. At the standard top gold energy, each nucleon carries approximately 100 gigaelectronvolts, giving sNN=200\sqrt{s_{NN}}=200 gigaelectronvolts for two opposing nucleons. Here the electronvolt is an energy unit, and sNNs_{NN} denotes the squared centre-of-mass energy of a nucleon pair. A gold–gold collision therefore involves a total beam energy of about 39.4 teraelectronvolts when all nucleons in both nuclei are counted.

Beam intensity alone does not determine the collision rate. The relevant accelerator quantity is luminosity, which incorporates bunch populations, encounter frequency, and transverse overlap. For a process with interaction cross section σ\sigma, the event rate is approximately

R=Lσ,R=\mathcal{L}\sigma,

where L\mathcal{L} is instantaneous luminosity. Integrated luminosity measures the accumulated collision exposure over a run.

Maintaining useful luminosity requires control of orbit errors, bunch dimensions, and collective beam effects. Intrabeam scattering gradually increases the spread of particle positions and momenta, especially for highly charged heavy ions. Stochastic cooling was developed at RHIC to counteract beam degradation during storage. At lower energies, the Low Energy RHIC electron Cooler used an overlapping electron beam to reduce ion-beam spread; it first cooled colliding beams in 2019. These techniques improved the practical reach of experiments that require large event samples at energies well below the original design point.

Experiments and reconstruction of collision events

RHIC began with four major experiments: STAR, PHENIX, PHOBOS, and BRAHMS. They were designed with different geometrical coverage and particle-identification capabilities, allowing cross-checks of the same underlying physics. This complementarity was important because no detector could directly observe the earliest, hottest stage of a collision. Experiments instead recorded particles emerging after the system had expanded and cooled, then inferred its earlier properties from distributions, correlations, and comparisons with theory.

STAR, the Solenoidal Tracker at RHIC, was built around a large time projection chamber within a solenoidal magnetic field. Charged particles leave ionization trails in gas, and the drifting electrons provide three-dimensional tracking information. The curvature of a track determines momentum, while the amount of ionization helps identify the particle species. Time-of-flight detectors extend identification by measuring particle velocities. STAR’s broad azimuthal coverage makes it particularly effective for studying collective motion, particle correlations, event-by-event fluctuations, and complex decay topologies involving displaced vertices.

PHENIX emphasized selected probes of the medium, including electrons, muons, photons, and identified hadrons. Its central arms measured particles around midrapidity, while forward and backward muon arms studied penetrating decay products. Electromagnetic calorimeters measured deposited energy from photons and electrons, and specialized identification systems helped separate electrons from the much more numerous charged hadrons. Such measurements were central to studies of quarkonium, heavy-flavour production, direct photons, and polarized-proton reactions.

PHOBOS used extensive silicon detectors to measure charged-particle production over a wide range of angles, emphasizing global collision properties and particle multiplicities. BRAHMS used movable magnetic spectrometers to identify particles at different angles and momenta, providing detailed information about longitudinal particle production and the redistribution of baryon number. Together, these experiments established that a heavy-ion event cannot be adequately characterized by a single temperature or particle count: longitudinal motion, transverse expansion, species composition, and collision geometry all matter.

Reconstructing a collision requires selecting the interaction vertex, rejecting background, correcting for detector acceptance and efficiency, and quantifying measurement uncertainty. Experimental analyses distinguish statistical uncertainty from systematic effects such as momentum calibration, tracking losses, and particle misidentification. Trigger systems select potentially useful events, while minimum-bias samples aim to represent a broad fraction of the interaction cross section. Later detector development included sPHENIX, which recorded first collisions in 2023 and was designed for precision measurements of jets and heavy quarks using tracking, electromagnetic calorimetry, and hadronic calorimetry over full azimuth.

From nuclear impact to expanding matter

A relativistic heavy-ion collision is not a miniature static furnace. It is a rapidly evolving process in which two Lorentz-contracted nuclei pass through one another, deposit energy, and generate an expanding system. At RHIC’s highest gold–gold energy, the incoming nuclei move at speeds extremely close to the speed of light. In their centre-of-mass frame, special relativity reduces their longitudinal thickness, and their overlap lasts for only a fraction of a femtometre divided by the speed of light. The subsequent evolution unfolds over several to roughly ten femtometres per cc, corresponding to times of order 10−2310^{-23} seconds.

The earliest stage involves quarks and gluons carrying fractions of the incoming nucleons’ momenta, together with intense gluon fields. Descriptions of this stage use several theoretical frameworks, including models with high gluon occupation and saturation. The energy deposited is spatially uneven because the positions of nucleons and their internal constituents fluctuate. Those initial irregularities later influence the patterns of particle emission, making the initial state a major source of uncertainty when extracting properties of the produced medium.

After the initial stage, the system develops a collective expansion that can be described over a substantial interval by relativistic hydrodynamics. The applicability of hydrodynamics does not require every microscopic degree of freedom to have reached complete thermodynamic equilibrium; it requires that the relevant long-wavelength evolution be captured by conserved quantities and constitutive relations. Calculations therefore include energy and momentum conservation, an equation of state, and transport coefficients. These connect the microscopic theory of QCD to observable flow patterns.

As the matter cools, quarks and gluons become bound into hadrons. Chemical freeze-out describes the stage at which inelastic reactions cease to substantially change the relative abundances of many hadron species. Elastic collisions can continue afterward, modifying particle momenta until kinetic freeze-out. Fits to particle yields and momentum spectra probe these different stages, but the fitted parameters are model-dependent and should not be confused with a direct thermometer reading of the earliest collision.

The initial energy density can be estimated from measured transverse energy using a Bjorken-type relation,

ε(τ)≃1ATτdETdy,\varepsilon(\tau)\simeq \frac{1}{A_T\tau}\frac{dE_T}{dy},

in units where c=1c=1. Here ATA_T is the transverse overlap area, τ\tau an assumed proper time, and dET/dydE_T/dy the transverse energy per unit rapidity. The estimate depends on geometry and formation-time assumptions and neglects parts of the subsequent dynamical evolution. Nevertheless, such analyses showed that central RHIC collisions reach energy densities well above those characteristic of ordinary nuclear matter.

Collective flow and the strongly coupled plasma

One of RHIC’s most influential findings was the strength of collective flow. In a noncentral collision, the overlap of two nuclei is approximately almond-shaped in the plane perpendicular to the beams. Pressure gradients are larger along the shorter direction, so the expanding matter develops an anisotropic momentum distribution. The resulting particle pattern carries information about how effectively the medium transforms its initial spatial geometry into collective motion.

The azimuthal distribution is commonly expressed as

dNdϕ∝1+2∑n=1∞vncos⁡[n(ϕ−Ψn)],\frac{dN}{d\phi}\propto 1+2\sum_{n=1}^{\infty}v_n\cos[n(\phi-\Psi_n)],

where ϕ\phi is a particle’s azimuthal angle, Ψn\Psi_n a reference orientation for harmonic nn, and vnv_n the corresponding flow coefficient. The second harmonic, v2v_2, is called elliptic flow. Higher harmonics, including triangular flow v3v_3, are especially sensitive to fluctuations in the initial density distribution. Their measurement requires methods that distinguish collective correlations from effects such as resonance decays and particles belonging to the same jet.

At low transverse momentum, identified particles exhibit characteristic mass-dependent flow patterns, broadly consistent with a common expanding velocity field. At intermediate transverse momentum, baryons and mesons display different trends, and approximate scaling with the number of constituent quarks has been observed over restricted kinematic ranges. These results supported the interpretation that collective motion developed substantially before hadrons formed. The scaling is not an exact law and is affected by hadronization, later hadronic interactions, and the chosen plotting variables.

Relativistic viscous hydrodynamic calculations reproduce many of these observations when the shear viscosity is small relative to the entropy density. The often-quoted dimensionless ratio is η/s\eta/s in natural units. A theoretical benchmark,

ηs=14π,\frac{\eta}{s}=\frac{1}{4\pi},

arises in certain strongly coupled quantum field theories with gravitational descriptions. It is not an established universal lower bound for all physical systems. RHIC measurements nevertheless indicate that the created matter has a low η/s\eta/s, making it a remarkably efficient fluid on microscopic length scales.

The phrase “nearly perfect liquid” describes this low dissipative response, not a lack of interactions. In fact, a weakly interacting gas would generally transport momentum over longer distances and exhibit larger shear viscosity. Extracting a numerical viscosity requires simultaneous treatment of the initial state, temperature dependence, bulk viscosity, and hadronic evolution. Collective signatures in smaller collision systems, including deuteron–gold and helium-3–gold, have extended these questions to droplets whose size challenges the conventional separation between microscopic and fluid-dynamical descriptions.

Energetic probes, heavy quarks, and electromagnetic radiation

The hot medium at RHIC survives too briefly to be examined with an external beam. Instead, energetic quarks and gluons produced in the earliest hard scatterings act as internal probes. As they cross the medium, they exchange momentum with it and emit additional radiation. Their eventual jets and leading hadrons therefore encode the interaction between a fast parton and the surrounding quark–gluon matter. The modification of these objects is known as jet quenching.

A central observable is the nuclear modification factor,

RAA(pT)=1⟨Ncoll⟩dNAA/dpTdNpp/dpT,R_{AA}(p_T)= \frac{1}{\langle N_{\mathrm{coll}}\rangle} \frac{dN_{AA}/dp_T}{dN_{pp}/dp_T},

where pTp_T is transverse momentum, AAAA denotes nucleus–nucleus collisions, and ⟨Ncoll⟩\langle N_{\mathrm{coll}}\rangle is the estimated mean number of binary nucleon–nucleon encounters for the selected event class. If a hard process simply scaled with independent nucleon collisions, its RAAR_{AA} would be near unity, apart from nuclear initial-state effects. In central gold–gold collisions, high-momentum neutral pions were suppressed by factors of roughly four to five over substantial momentum ranges.

Reference measurements were essential to interpreting this suppression. Deuteron–gold collisions did not show a comparable midrapidity suppression at high transverse momentum, helping establish that the dramatic gold–gold effect arose primarily from the produced medium rather than solely from the incoming nuclear wave functions. Direct photons, which do not undergo strong interactions, provided another comparison: their high-momentum production was broadly consistent with expectations for hard scattering after accounting for relevant nuclear effects.

Heavy quarks offer additional information because charm and bottom masses introduce scales distinct from the medium temperature. Measurements of electrons from heavy-flavour decays, and later more direct reconstruction of heavy-flavour hadrons, showed substantial suppression and collective behaviour. Quarkonium states, such as the J/ψJ/\psi and the bottomonium family, probe screening and dissociation in the medium. Their interpretation also requires accounting for regeneration from independently produced heavy quarks, feed-down from excited states, and modifications that occur in nuclear matter without a hot plasma.

Photons and lepton pairs can escape with relatively little strong rescattering. PHENIX reported an excess of direct photons in central gold–gold collisions and used spectral analyses to infer a very hot source. Widely reported early-temperature estimates of several trillion kelvin depend on dynamical modelling; a fitted photon spectral slope is not simply the initial temperature. The coexistence of substantial photon production and sizeable photon anisotropy has posed a particularly demanding test of models, because early emission favours high temperatures whereas large collective anisotropy generally develops later.

Beam-energy scans and the QCD phase diagram

RHIC’s ability to vary collision energy made it possible to investigate the phase structure of strongly interacting matter rather than study only its hottest accessible form. The QCD phase diagram is usually represented using temperature and baryon chemical potential. At the highest RHIC energies, the midrapidity region contains relatively similar abundances of baryons and antibaryons, corresponding to a comparatively small baryon chemical potential. At lower energies, more incoming baryon number remains near midrapidity, allowing experiments to explore matter with a greater net baryon density.

Calculations using lattice QCD indicate that the transition from hadronic matter to quark–gluon matter at very small baryon chemical potential is a smooth crossover, with a characteristic temperature around 155 megaelectronvolts. At larger baryon chemical potential, many theoretical models predict a first-order phase transition and a critical point marking its endpoint. The existence and location of such a QCD critical point have not been established experimentally. Direct lattice calculations become especially difficult in this region because of the finite-density sign problem.

The first major RHIC Beam Energy Scan included gold–gold collisions at nucleon-pair energies of 7.7, 11.5, 19.6, 27, and 39 gigaelectronvolts, supplemented by higher-energy reference data. A second phase, pursued from 2019 to 2021, emphasized much larger low-energy data samples and improved detector capabilities. STAR’s fixed-target programme extended the accessible region further by allowing one circulating beam to strike a stationary target, reaching nucleon-pair centre-of-mass energies below those practical in collider mode.

Among the principal observables are event-by-event fluctuations of net proton number, used as an experimentally accessible proxy for aspects of net baryon-number fluctuations. Near a critical point, a growing correlation length could alter higher-order statistical cumulants. Experiments examine combinations involving variance, skewness, and kurtosis because different orders can respond differently to critical correlations. Any candidate signal must be separated from effects of baryon-number conservation, finite acceptance, detection efficiency, collision-volume fluctuations, and ordinary hadronic dynamics.

Other scan measurements include directed flow, particle-yield ratios, strange-hadron production, and differences between particle and antiparticle elliptic flow. These observables constrain the equation of state, the redistribution of baryon number, and the evolution of collective behaviour with energy. Nonmonotonic trends can be scientifically suggestive, but are not by themselves proof of criticality. Establishing a phase-boundary feature requires consistency across observables, energies, analysis methods, and theoretical descriptions of the complete collision evolution.

Polarized protons and the origin of proton spin

RHIC was also designed to collide beams of polarized protons, making spin physics a distinct part of its scientific programme. A proton has intrinsic spin angular momentum of one-half in units of the reduced Planck constant. However, it is a composite object containing quarks, antiquarks, and gluons, whose intrinsic spins and orbital motion must collectively account for that total. Deep-inelastic scattering experiments beginning in the late 1980s established that quark and antiquark helicities contribute only part of the proton’s spin, motivating more direct study of the gluon contribution and parton motion.

Maintaining polarization in a circular accelerator is technically demanding. As a proton travels through magnetic fields, its spin precesses, and repeated encounters with depolarizing resonances can reduce the beam’s polarization. RHIC uses specialized magnetic insertions called Siberian snakes to rotate the spin and stabilize its evolution over successive turns. Spin rotators near interaction regions allow the polarization direction at collision to be changed between longitudinal and transverse configurations. Polarimeters measure the beam polarization, which enters directly into the normalization of measured spin asymmetries.

Longitudinally polarized proton collisions probe the helicity distributions of quarks and gluons. For suitable reactions, experiments measure the double-spin asymmetry

ALL=σ++−σ+−σ+++σ+−,A_{LL}= \frac{\sigma^{++}-\sigma^{+-}} {\sigma^{++}+\sigma^{+-}},

where the superscripts denote equal or opposite beam-helicity configurations. In practice, event counts must be corrected for beam polarizations and relative luminosities. Inclusive jets and neutral pions are particularly useful because their production receives substantial contributions from subprocesses involving gluons. Global analyses incorporating RHIC data found evidence for a positive gluon helicity contribution over the momentum-fraction range constrained by the measurements.

Production of W bosons provides complementary information. Because the weak interaction distinguishes handedness, single-spin asymmetries in W production separate contributions from different quark and antiquark flavours without relying on hadron fragmentation in the same way as identified-hadron measurements. RHIC results established important constraints on the flavour structure of the polarized antiquark sea, including differences between anti-up and anti-down helicity distributions.

With transverse polarization, experiments study left–right asymmetries and correlations involving spin and parton motion. Their theoretical interpretation involves transverse-momentum-dependent distributions, multiparton correlations, and fragmentation effects. These measurements do not yield a simple three-dimensional photograph of the proton; they constrain components of a scale-dependent quantum description. RHIC’s spin programme therefore complements its heavy-ion programme by examining the same strong interaction in a different setting: the internal structure of a single nucleon rather than an extended, hot many-body medium.

Precision upgrades and connections to other facilities

RHIC’s later scientific development increasingly emphasized precision and controlled variation. Comparisons among gold, copper, uranium, and smaller projectiles allow experiments to change the size and shape of the interaction region. Uranium’s deformed nucleus provides geometries unavailable with nearly spherical nuclei. Collisions of different species also help distinguish effects associated with total system size, local density, longitudinal structure, and the number of nucleons that actually participate in the collision.

A particularly targeted programme used ruthenium-96 and zirconium-96 isobars in 2018. These nuclei have the same mass number but different proton numbers, permitting a comparison intended to vary the collision-generated electromagnetic field while keeping many bulk properties similar. One motivation was the search for the chiral magnetic effect, a proposed separation of electric charges along a magnetic field in matter with an imbalance of chirality. The comparison also demonstrated the importance of differences in nuclear deformation and density distributions. The measured correlation patterns cannot be interpreted as an unambiguous magnetic-field signal without controlling these backgrounds.

STAR upgrades expanded tracking coverage and improved low-energy measurements. Better particle identification, vertex reconstruction, and acceptance enhanced the ability to study rare particles and fluctuation observables. The sPHENIX detector took a different complementary approach, with a compact superconducting solenoid and extensive calorimetry designed to measure complete jets and distinguish heavy-quark probes. Separating the closely spaced bottomonium states, for example, provides more detailed information about how different bound states respond to the medium.

RHIC and the Large Hadron Collider explore overlapping physics at different energy scales. LHC heavy-ion collisions produce higher initial temperatures and much larger samples of very energetic jets, while RHIC provides broad energy-scan capabilities, diverse collision systems, and polarized proton beams. Comparisons help determine whether extracted medium properties vary with temperature and test whether a common theoretical description can account for different collision environments. Neither collider supplies all the necessary information in isolation.

Brookhaven was selected in 2020 as the site of the planned Electron–Ion Collider, a facility designed to collide electrons with protons and nuclei while reusing major elements of RHIC infrastructure. Its principal objective differs from RHIC’s hot-matter programme: electron scattering is intended to resolve the quark and gluon structure of the incoming hadrons and nuclei. These measurements address quantities that also influence heavy-ion initial conditions, including gluon distributions and their spatial organization. The connection is therefore both infrastructural and scientific, linking the study of nuclear structure before a collision to the study of strongly interacting matter created afterward.