Compact Linear Collider (CLIC)

• CLIC is a staged high-luminosity e⁺e⁻ linear collider at CERN designed to deliver multi-TeV collisions for precision studies of the Higgs boson, top quark, and beyond-the-Standard-Model physics.
• It employs an innovative two-beam acceleration scheme that converts kinetic energy from a high-current drive beam into high-gradient RF bursts, achieving up to 100 MV/m for multi-TeV operations.
• Its modular, upgradeable design integrates advanced silicon-based vertex and tracking detectors with nanosecond timing to optimize particle-flow reconstruction in high-background environments.

CLIC, the Compact Linear Collider, is a proposed staged high-luminosity linear $e^+e^-$ collider developed by the international CLIC and CLICdp collaborations and studied for implementation at CERN. Its central rationale is to combine the clean initial state of lepton collisions with centre-of-mass energies ranging from a few hundred GeV to the multi-TeV scale, thereby supporting both model-independent precision measurements of the Higgs boson and top quark and broad direct and indirect searches for physics beyond the Standard Model (Adli et al., 31 Mar 2025).

1. Staged machine concept and accelerator architecture

CLIC is organized as a staged programme. In the current project description, the initial stage is optimized for $\sqrt{s}=380~\mathrm{GeV}$, including a top-threshold scan near $350~\mathrm{GeV}$, while later operation extends to the TeV domain. The updated first-stage baseline uses a site length of $11~\mathrm{km}$, operates at $100~\mathrm{Hz}$, and targets a luminosity of $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$ with a site power consumption of $166~\mathrm{MW}$. A detailed $1.5~\mathrm{TeV}$ stage has also been specified, with a site length of $29~\mathrm{km}$, $50~\mathrm{Hz}$ repetition rate, $\sqrt{s}=380~\mathrm{GeV}$0 bunches per train, $\sqrt{s}=380~\mathrm{GeV}$1 bunch spacing, and tabulated luminosity $\sqrt{s}=380~\mathrm{GeV}$2; the same report notes that updated alignment treatment would likely raise this to $\sqrt{s}=380~\mathrm{GeV}$3. Operation up to $\sqrt{s}=380~\mathrm{GeV}$4 remains part of the full programme, with performance at that energy continuing to rest on the detailed 2012 CDR and 2018 studies (Adli et al., 31 Mar 2025).

The defining accelerator feature is the two-beam acceleration scheme. A high-current drive beam is generated with low-frequency RF systems, accelerated in long pulses, compressed into short intense pulses, and passed through power extraction and transfer structures, where its kinetic energy is converted into short $\sqrt{s}=380~\mathrm{GeV}$5 RF bursts that feed normal-conducting X-band main-linac structures. The technology range discussed is $\sqrt{s}=380~\mathrm{GeV}$6–$\sqrt{s}=380~\mathrm{GeV}$7: $\sqrt{s}=380~\mathrm{GeV}$8 is chosen as the optimum for the first stage, while $\sqrt{s}=380~\mathrm{GeV}$9 is required for multi-TeV operation (Adli et al., 31 Mar 2025).

This staged design evolved from earlier baseline scenarios. In the 2018–2022 documentation, CLIC was described with stages at $350~\mathrm{GeV}$0, $350~\mathrm{GeV}$1, and $350~\mathrm{GeV}$2, with main-linac tunnel lengths of $350~\mathrm{GeV}$3, $350~\mathrm{GeV}$4, and $350~\mathrm{GeV}$5, respectively. Those studies also emphasized that the first stage was designed from the outset to be upgradeable by extending tunnels, reusing installed infrastructure, and adding new higher-gradient modules and drive-beam systems (Brunner et al., 2022).

2. Beam environment and detector design constraints

Although CLIC is a lepton collider, its detector environment is not benign. At $350~\mathrm{GeV}$6, CLIC runs with $350~\mathrm{GeV}$7 bunches per train, bunches separated by $350~\mathrm{GeV}$8, at a repetition rate of $350~\mathrm{GeV}$9, so the train is only $11~\mathrm{km}$0 long within a $11~\mathrm{km}$1 cycle. Very small beam spot sizes, about $11~\mathrm{km}$2 and $11~\mathrm{km}$3, generate intense beamstrahlung, producing $11~\mathrm{km}$4 pairs and hadronic background concentrated along the beam axis. This background sets the minimum beam-pipe and inner-layer radii, fixes detector granularity through an occupancy limit below $11~\mathrm{km}$5 per bunch train, and drives the requirement of single-hit time stamping of roughly $11~\mathrm{km}$6 for the silicon systems (Spannagel, 2020).

The detector concept developed for this environment is CLICdet, a general-purpose detector optimized for precision measurements and beam-background rejection. The design targets include

$11~\mathrm{km}$7

transverse impact-parameter resolution near $11~\mathrm{km}$8, jet-energy resolution better than $11~\mathrm{km}$9 for $100~\mathrm{Hz}$0 jets and better than $100~\mathrm{Hz}$1 above $100~\mathrm{Hz}$2, calorimeter hit timing of $100~\mathrm{Hz}$3, vertex and tracker timing of $100~\mathrm{Hz}$4, and acceptance down to $100~\mathrm{Hz}$5 for electrons and photons. The detector uses six silicon-pixel vertex layers with $100~\mathrm{Hz}$6 single-point resolution and $100~\mathrm{Hz}$7 per layer, a silicon tracker with six barrel and seven end-cap layers at $100~\mathrm{Hz}$8 $100~\mathrm{Hz}$9-$4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$0 resolution, a $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$1-layer silicon-tungsten ECAL, a $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$2-layer scintillator-steel HCAL, forward LumiCal and BeamCal, and a $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$3 solenoid (Adli et al., 31 Mar 2025).

The detector philosophy is explicitly particle-flow based. Earlier detector studies stressed that the calorimeters must be highly granular because jet reconstruction is limited not only by intrinsic calorimetric resolution but by the confusion term, namely the incorrect association of deposits to charged or neutral particles. This requirement led to a compact design with calorimetry inside the solenoid, all-silicon precision tracking, strong forward instrumentation, and detector-wide timing capabilities in the $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$4–$4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$5 range (Pitters, 2018).

3. Silicon vertex and tracking R&D

Silicon R&D is a core technical strand of the CLIC detector programme. For the vertex detector, the physics driver is flavour tagging, which yields a target single-point resolution $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$6, material budget below $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$7 per detection layer, and power dissipation below $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$8 so that forced air-flow cooling is sufficient. The current layout uses three double layers in the barrel and three double disks on each side, with the innermost barrel at $4.5\times 10^{34}\,\mathrm{cm^{-2}s^{-1}}$9. Each active layer is envisioned as a $166~\mathrm{MW}$0 sensor plus $166~\mathrm{MW}$1 readout ASIC, with $166~\mathrm{MW}$2 pixels and total active area about $166~\mathrm{MW}$3 (Spannagel, 2020).

The most mature vertex-detector technology line is hybrid silicon. Its flagship ASIC is CLICpix2, a dedicated $166~\mathrm{MW}$4 CMOS hybrid pixel chip with active area $166~\mathrm{MW}$5, a $166~\mathrm{MW}$6 matrix at $166~\mathrm{MW}$7 pitch, and in-pixel $166~\mathrm{MW}$8-bit time-over-threshold and $166~\mathrm{MW}$9-bit time-of-arrival measurement. Laboratory and beam tests of CLICpix2 assemblies yielded spatial resolution of about $1.5~\mathrm{TeV}$0 with a $1.5~\mathrm{TeV}$1-thick sensor. The paper interprets this as evidence that the ultimate $1.5~\mathrm{TeV}$2 CLIC vertex target is not reachable with standard planar silicon sensors that are only $1.5~\mathrm{TeV}$3 thick at $1.5~\mathrm{TeV}$4 pitch, implying that sensor optimization or alternative concepts remain necessary (Spannagel, 2020).

Power pulsing is treated as a central enabling technique. In CLICpix2, only the preamplifier and discriminator are power pulsed by switching their supply current between a full-power DAC and a reduced-current DAC. At the default low-power setting, the front-end settles in about $1.5~\mathrm{TeV}$5; if the low-power current is reduced to zero, the wake-up time extends to about $1.5~\mathrm{TeV}$6. Even so, the average analog power consumption drops by a factor five, from $1.5~\mathrm{TeV}$7 to $1.5~\mathrm{TeV}$8. The same study states that if additional DACs are included in future pulsing schemes, a factor-$1.5~\mathrm{TeV}$9 reduction is possible, bringing total analog power below $29~\mathrm{km}$0 (Spannagel, 2020).

For the tracking detector, the baseline direction is monolithic silicon, motivated by the very large active area and the need to reduce material, complexity, and cost. The principal prototype is CLICTD, fabricated in a $29~\mathrm{km}$1 CMOS imaging process with a small N-well collection electrode in a $29~\mathrm{km}$2-thick p-type high-resistivity epitaxial layer. It contains $29~\mathrm{km}$3 pixels over $29~\mathrm{km}$4, with elongated $29~\mathrm{km}$5 pixels, $29~\mathrm{km}$6-bit ToA, $29~\mathrm{km}$7-bit ToT, and an eight-sub-pixel analog segmentation along the long dimension. After threshold equalization, the measured threshold dispersion is $29~\mathrm{km}$8 electrons and the pixel noise RMS is $29~\mathrm{km}$9 electrons. The same R&D programme also investigates anisotropic conductive film bonding for fine-pitch hybridization and notes that silicon-on-insulator designs are under investigation, though without detailed performance data in that conference paper (Spannagel, 2020).

Simulation is treated as part of the detector technology stack. The CLIC silicon programme explicitly identifies Allpix$50~\mathrm{Hz}$0 as the framework used to simulate the full detection chain, combining Geant4 particle transport with detailed electric-field descriptions imported from device simulations such as TCAD. This supports end-to-end optimization of geometry, biasing, signal formation, timing, and digitization under realistic stochastic sensor physics (Spannagel, 2020).

4. Precision Higgs and top programme

The staged physics logic of CLIC begins with the $50~\mathrm{Hz}$1 stage, where Higgsstrahlung and top-pair production coexist. In the updated 2025 summary, the recoil-mass measurement of Higgsstrahlung remains the anchor for model-independent Higgs coupling extraction without assumptions on exotic Higgs decays. That report quotes a projected precision of $50~\mathrm{Hz}$2 on $50~\mathrm{Hz}$3, with $50~\mathrm{Hz}$4 and $50~\mathrm{Hz}$5 at similar precision in the model-independent fit, percent-level sensitivity to $50~\mathrm{Hz}$6, and $50~\mathrm{Hz}$7 precision on the total Higgs width using the first two stages (Adli et al., 31 Mar 2025).

Higher energies change both event rates and observable content. The same summary emphasizes that vector-boson-fusion Higgs production rises strongly at the TeV stages, making them especially powerful for precision Higgs measurements and for new-physics sensitivity. The Higgs self-coupling is a principal justification for multi-TeV running: CLIC expects a $50~\mathrm{Hz}$8 observation of double Higgsstrahlung $50~\mathrm{Hz}$9 at the second stage and evidence for $\sqrt{s}=380~\mathrm{GeV}$00, while the combination of second-stage $\sqrt{s}=380~\mathrm{GeV}$01 and $\sqrt{s}=380~\mathrm{GeV}$02 information with differential distributions at $\sqrt{s}=380~\mathrm{GeV}$03 yields an expected precision

$\sqrt{s}=380~\mathrm{GeV}$04

The role of the $\sqrt{s}=380~\mathrm{GeV}$05 channel and of differential information is explicitly to remove the ambiguity that would arise from the inclusive $\sqrt{s}=380~\mathrm{GeV}$06 cross section alone (Adli et al., 31 Mar 2025).

Top physics is the other pillar of the first stage. Detailed studies summarized in the 2019 physics-potential review state that more than one million top quarks and antiquarks are produced already in the initial phase. For a threshold scan with $\sqrt{s}=380~\mathrm{GeV}$07 collected in $\sqrt{s}=380~\mathrm{GeV}$08 energy steps, the expected statistical uncertainty on the top mass is $\sqrt{s}=380~\mathrm{GeV}$09–$\sqrt{s}=380~\mathrm{GeV}$10, with total systematic uncertainty of order $\sqrt{s}=380~\mathrm{GeV}$11. At higher energy, direct measurement of the top Yukawa coupling through $\sqrt{s}=380~\mathrm{GeV}$12 at $\sqrt{s}=380~\mathrm{GeV}$13 with $\sqrt{s}=380~\mathrm{GeV}$14 reaches $\sqrt{s}=380~\mathrm{GeV}$15 precision (Zarnecki, 2019).

These observables are not treated as isolated measurements. CLIC studies repeatedly frame Higgs, top, diboson, and fermion-pair data within SMEFT analyses, with the high-energy stages providing special leverage because many operator effects grow with energy. A plausible implication is that the staged programme is designed not only to accumulate statistics but to change the effective theory directions that can be separated experimentally (Zarnecki, 2019).

5. Beyond-the-Standard-Model reach and hidden-sector sensitivity

CLIC’s BSM case combines precision observables, direct production near the kinematic limit, and sensitivity to nonstandard signatures. A representative summary of the full staged programme quotes model-independent sensitivity to invisible Higgs decays

$\sqrt{s}=380~\mathrm{GeV}$16

heavy-singlet mixing constrained to $\sqrt{s}=380~\mathrm{GeV}$17, Higgsino reach beyond $\sqrt{s}=380~\mathrm{GeV}$18 in disappearing-track searches, $\sqrt{s}=380~\mathrm{GeV}$19 discovery reach up to $\sqrt{s}=380~\mathrm{GeV}$20 for SM-like couplings, and sensitivity to lepton-flavour-violating scales above $\sqrt{s}=380~\mathrm{GeV}$21. The same report emphasizes that the clean environment, precise initial-state knowledge, and multi-TeV energy together make CLIC particularly strong for electroweak states, hidden sectors, long-lived particles, and energy-growing EFT effects (Roloff et al., 2018).

A concrete case study is the search for invisible scalar decays using hadronic recoil against a reconstructed $\sqrt{s}=380~\mathrm{GeV}$22 boson. At $\sqrt{s}=380~\mathrm{GeV}$23, a hadronic $\sqrt{s}=380~\mathrm{GeV}$24 analysis with WHIZARD event generation and Delphes fast simulation of CLICdet yields expected $\sqrt{s}=380~\mathrm{GeV}$25 C.L. upper limits

$\sqrt{s}=380~\mathrm{GeV}$26

for integrated luminosities of $\sqrt{s}=380~\mathrm{GeV}$27 and $\sqrt{s}=380~\mathrm{GeV}$28, respectively. The same study quotes $\sqrt{s}=380~\mathrm{GeV}$29 discovery thresholds of $\sqrt{s}=380~\mathrm{GeV}$30 and $\sqrt{s}=380~\mathrm{GeV}$31, and extends the search to an additional neutral scalar $\sqrt{s}=380~\mathrm{GeV}$32 with $\sqrt{s}=380~\mathrm{GeV}$33 over $\sqrt{s}=380~\mathrm{GeV}$34–$\sqrt{s}=380~\mathrm{GeV}$35 at $\sqrt{s}=380~\mathrm{GeV}$36 and $\sqrt{s}=380~\mathrm{GeV}$37–$\sqrt{s}=380~\mathrm{GeV}$38 at $\sqrt{s}=380~\mathrm{GeV}$39 (Mekala et al., 2021).

The broader CLIC physics programme also includes extended Higgs sectors, compositeness, neutrino-mass mediators, flavour-violating signatures, and disappearing or stub tracks from nearly degenerate electroweak multiplets. In that literature, the multi-TeV stages are repeatedly presented not as incremental luminosity upgrades but as the part of the machine where high-energy electroweak processes become a direct BSM probe in their own right (Roloff et al., 2018).

6. Implementation, maturity, and project outlook

CLIC is presented in recent project reports as a mature proposal rather than a purely conceptual study. System-level demonstrations at CTF3 validated the drive-beam scheme, RF transfer efficiency, power extraction, and two-beam acceleration up to $\sqrt{s}=380~\mathrm{GeV}$40, with the required stability achieved. Major subsystems are assessed mostly at TRL $\sqrt{s}=380~\mathrm{GeV}$41–$\sqrt{s}=380~\mathrm{GeV}$42, with X-band technology and HOM damping at TRL $\sqrt{s}=380~\mathrm{GeV}$43, positron source and two-beam acceleration at TRL $\sqrt{s}=380~\mathrm{GeV}$44, damping rings at TRL $\sqrt{s}=380~\mathrm{GeV}$45, emittance preservation at TRL $\sqrt{s}=380~\mathrm{GeV}$46, and interaction-point spot size and stability at TRL $\sqrt{s}=380~\mathrm{GeV}$47 (Adli et al., 31 Mar 2025).

Implementation studies near CERN are correspondingly detailed. The 2025 baseline includes two simultaneous interaction regions fed by a dual beam-delivery system, so that two detectors can operate in parallel and each sees on average $\sqrt{s}=380~\mathrm{GeV}$48 at $\sqrt{s}=380~\mathrm{GeV}$49. The $\sqrt{s}=380~\mathrm{GeV}$50 baseline cost is quoted as

$\sqrt{s}=380~\mathrm{GeV}$51

with major cost drivers including main-linac modules, civil engineering, drive-beam injectors and beam transport, cooling, and ventilation. The upgrade to $\sqrt{s}=380~\mathrm{GeV}$52 is estimated at about $\sqrt{s}=380~\mathrm{GeV}$53 billion CHF, and one CLICdet detector at about $\sqrt{s}=380~\mathrm{GeV}$54 (Adli et al., 31 Mar 2025).

The current planning scenario assumes a six-year preparation phase ahead of a Technical Design Report by 2031, earliest approval in 2031, earliest construction start in 2033, and first beams and collisions in 2041. Construction and commissioning of the first stage are estimated to take seven years, with a little over five years for excavation, lining, infrastructures, and installation, eight months for system commissioning, two months for final alignment, and one year for beam commissioning (Adli et al., 31 Mar 2025).

The same report adds an explicit sustainability layer. At $\sqrt{s}=380~\mathrm{GeV}$55, annual electrical energy consumption is estimated as $\sqrt{s}=380~\mathrm{GeV}$56, and operation at an assumed future electricity carbon intensity of $\sqrt{s}=380~\mathrm{GeV}$57 corresponds to about $\sqrt{s}=380~\mathrm{GeV}$58. A notable conclusion is that embodied carbon from construction is estimated to be $\sqrt{s}=380~\mathrm{GeV}$59–$\sqrt{s}=380~\mathrm{GeV}$60 times larger than emissions from ten years of operation, which shifts optimization pressure toward civil engineering, site layout, and injector and surface installations as much as toward wall-plug efficiency (Adli et al., 31 Mar 2025).

Taken together, these studies present CLIC as a staged linear-collider programme whose scientific identity depends on a specific combination: absolute Higgs measurements at the first stage, top-threshold and continuum precision, multi-TeV electroweak reach, and a detector system explicitly engineered for nanosecond timing, ultra-low-mass silicon tracking, and particle-flow reconstruction in the presence of beam-induced background.