FoCal: ALICE Forward Calorimeter

• FoCal is a forward calorimeter system for ALICE that extends measurements into a very forward pseudorapidity region to probe small-x QCD and gluon saturation.
• It integrates a highly granular silicon-tungsten electromagnetic section (FoCal-E) with a conventional hadronic calorimeter (FoCal-H) for precise detection of photons, jets, and mesons.
• Prototype validations demonstrate <3% EM resolution at 100 GeV and <20% hadronic resolution up to 350 GeV, confirming its capabilities for advanced forward physics studies.

Searching arXiv for the FoCal papers and closely related ALICE Forward Calorimeter references. The ALICE Forward Calorimeter, abbreviated FoCal, is a forward calorimeter system for the ALICE experiment at the Large Hadron Collider, designed to extend ALICE into the very forward pseudorapidity interval $3.2 < \eta < 5.8$ and to probe gluon densities in protons and nuclei down to $x \sim 10^{-6}$ at relatively low scales (Park, 2024). It combines a highly granular silicon–tungsten electromagnetic calorimeter with a conventional hadronic calorimeter, and its physics program centers on direct photons, neutral mesons, jets, photon–hadron correlations, and vector-meson photoproduction in ultra-peripheral collisions, all in a kinematic regime relevant to non-linear QCD evolution and gluon saturation (Park, 2024).

1. Historical setting and scientific rationale

FoCal is a new ALICE sub-detector to be installed during Long Shutdown 3, with installation foreseen for early 2028 and physics running in Run 4 during 2029–2032 (Park, 2024). It is located about $7$ m from the ALICE interaction point, close to the beam pipe, and occupies a far more forward region than the central ALICE detectors (Park, 2024).

The principal motivation is the study of small-$x$ QCD. In hadronic matter, the gluon density grows at small $x$ approximately as

$x g(x) \sim x^{-\delta/2},$

with $\delta$ determined from data (Park, 2024). The linear rise associated with DGLAP/BFKL-like evolution cannot continue indefinitely; at sufficiently small $x$, gluons overlap and recombine, and saturation sets in, in a regime commonly described as a Color Glass Condensate (Park, 2024). Saturation is characterized by a scale $Q_{\rm sat}$, with strong non-linear effects for gluons with transverse momentum $Q \lesssim Q_{\rm sat}$, and the scale increases with nuclear size as

$x \sim 10^{-6}$0

This makes nuclei especially favorable for observing saturation phenomena (Park, 2024).

FoCal is designed to access $x \sim 10^{-6}$1 at $x \sim 10^{-6}$2, beyond the reach of HERA and current LHC measurements (Park, 2024). The forward direction is critical because, for a process with transverse scale $x \sim 10^{-6}$3 at rapidity $x \sim 10^{-6}$4, the partonic momentum fractions are roughly

$x \sim 10^{-6}$5

so large positive rapidity in the FoCal direction corresponds to very small target $x \sim 10^{-6}$6 (Park, 2024). This kinematic logic underlies the detector’s emphasis on forward photons, mesons, jets, and correlations in pp, p–Pb, Pb–Pb, and ultra-peripheral collisions (Park, 2024).

A broader institutional context is that FoCal complements ALICE’s central-rapidity program and, relative to other LHC forward instruments, is distinguished by its extreme electromagnetic granularity and its explicit optimization for direct photons and saturation physics (Park, 2024). A related Letter of Interest framed FoCal as a central element of a comprehensive LHC small-$x \sim 10^{-6}$7 program extending down to $x \sim 10^{-6}$8 (Bearden et al., 2022).

2. Detector architecture and instrumentation

FoCal consists of two principal subsystems: FoCal-E, a highly granular Si+W electromagnetic calorimeter, and FoCal-H, a more conventional metal–scintillator hadronic calorimeter (Park, 2024). Together they form a compact forward calorimeter stack optimized for shower imaging, isolation measurements, and hadronic containment (Park, 2024).

FoCal-E

FoCal-E uses tungsten as absorber, exploiting its radiation length $x \sim 10^{-6}$9, small Molière radius $7$0, and compactness (Park, 2024). Its total thickness is about $7$1, with 20 active layers in total: 18 silicon pad layers and 2 silicon pixel layers (Park, 2024).

The pad layers use silicon pad sensors with transverse cell area $7$2 and HGCROC readout ASICs from CMS HGCAL, providing ADC, TOA, and TOT functionality (Park, 2024). Sampling is performed at 40 MHz with configurable phase relative to the 25 ns LHC bunch crossing, and timing information is used to improve pile-up rejection and dynamic range (Park, 2024). These layers provide the primary electromagnetic energy measurement and support cluster-shape, isolation, and invariant-mass analyses (Park, 2024).

The two pixel layers are placed at the 5th and 10th sampling layers and use ALPIDE monolithic active pixel sensors with pixel size approximately $7$3 (Park, 2024). Their readout is ITS2-like, with ALPIDE strings assembled via SpTAP rather than wire bonding, and data rates of roughly 1.2 Gbps for inner-barrel chains and 400 Mbps for outer-barrel chains (Park, 2024). These layers provide extremely fine sampling of shower cores, enabling separation of nearby electromagnetic showers at sub-mm to mm scales and supporting detailed shower-shape analysis for $7$4 discrimination (Park, 2024).

A prototype Si–W FoCal calorimeter based on MAPS established the technical viability of this concept earlier, using $7$5 pixels and digital calorimetry, and showed electromagnetic shower imaging in unprecedented detail (Haas et al., 2017). This suggests continuity between the R&D lineage and the Run-4 detector implementation.

FoCal-H

FoCal-H is a hadronic calorimeter based on Cu capillary tubes of outer diameter 2.5 mm filled with plastic scintillating fibers, with effective thickness about $7$6 (Park, 2024). It has no longitudinal segmentation and is laterally segmented into towers (Park, 2024). Readout uses SiPMs and an H2GCROC front-end ASIC, an adaptation of HGCROC for SiPM signals, with ADC/TOA/TOT functionality, programmable analog gain via a current conveyor, and per-channel programmable bias voltage (Park, 2024).

Its detector functions are hadronic energy measurement for jets and hadrons, measurement of the hadronic halo around EM clusters for photon isolation, and provision of hadronic depth and containment in the multi-hundred-GeV to TeV regime (Park, 2024).

Readout and integration

FoCal’s readout follows ALICE standards: ITS2-like ALPIDE chains for the pixel layers, HGCROC/H2GCROC at 40 MHz for the pad and hadronic systems, and integration into the Run-4 O$7$7 framework (Park, 2024). The combination of fine spatial granularity and timing information provides pile-up mitigation, isolation support, and inputs for high-level triggers on forward photons, jets, and vector-meson decays (Park, 2024).

3. Detector performance and prototype validation

FoCal’s design has been validated through detailed simulation and through a full-length prototype tested in CERN PS and SPS beams from 2021 to 2023 (Park, 2024). A separate prototype study using CMOS pixel sensors had already demonstrated that a Si–W FoCal concept could sustain extremely fine-grained digital calorimetry with about 39 million pixels in a modest tower width (Haas et al., 2017).

Prototype performance summary

Subsystem or observable	Reported result	Source
EM response linearity	deviations $7$8 from $7$9	(Park, 2024)
EM resolution at 100 GeV	$x$0	(Park, 2024)
Pixel shower-separation agreement data vs MC	within $x$1 mm	(Park, 2024)
Hadronic resolution up to 350 GeV	$x$2	(Park, 2024)

For FoCal-E pad layers, electron-beam studies up to 300 GeV showed that the mean charge signal is well described by

$x$3

with deviations from linearity below $x$4 (Park, 2024). The relative energy resolution,

$x$5

agrees with GEANT-based simulation within uncertainties for $x$6 GeV, and the measured resolution is below $x$7 at 100 GeV, better than the design goal of about $x$8 (Park, 2024).

For the pixel layers, two-electron 300 GeV beam tests resolved two distinct shower cores separated by about 1 cm in the hit map (Park, 2024). The projected lateral profiles yielded FWHM values of about 1.2 mm and 2.4 mm at 20 GeV in layers 5 and 10, respectively, shrinking to about 0.8 mm and 1.2 mm at 300 GeV (Park, 2024). Data and simulation agree within 0.5 mm (Park, 2024). This is directly relevant to rejecting merged $x$9 backgrounds in direct-photon analyses.

FoCal-H beam tests with hadrons up to 350 GeV showed energy resolution below $x$0 across the scanned range, with discrepancies between data and simulation below $x$1 (Park, 2024). That level is regarded as sufficient for forward jet measurements and photon isolation (Park, 2024).

The combination of $x$2 pads, $x$3 pixels, 20 longitudinal samplings, and TOA/TOT timing is specifically exploited for $x$4 discrimination, isolation, background rejection, and forward jet reconstruction (Park, 2024).

4. Physics program: direct photons, mesons, jets, and correlations

FoCal’s physics program is centered on small-$x$5 QCD and gluon saturation, but it spans multiple observable classes (Park, 2024).

Direct photons

Direct photons are primarily produced through the Compton process

$x$6

making them sensitive to the gluon density while remaining unaffected by final-state strong interactions (Park, 2024). The main challenge is the overwhelming background from decay photons, especially $x$7 (Park, 2024).

FoCal addresses this background through three complementary techniques: isolation cuts using both FoCal-E and FoCal-H, invariant-mass cuts on cluster pairs consistent with $x$8 decays, and shower-shape analysis using the pixel layers, including the long-axis ellipticity of the shower (Park, 2024). Simulations indicate that these methods can enhance the direct-photon signal fraction to $x$9 at $x g(x) \sim x^{-\delta/2},$0, corresponding to about a factor 11 gain in signal purity (Park, 2024).

The expected impact on nuclear PDFs is framed via the direct-photon nuclear modification factor $x g(x) \sim x^{-\delta/2},$1 at $x g(x) \sim x^{-\delta/2},$2 TeV. NLO QCD with current nPDFs yields uncertainties of order $x g(x) \sim x^{-\delta/2},$3 at $x g(x) \sim x^{-\delta/2},$4, and incorporating FoCal pseudo-data through reweighting reduces this uncertainty by about $x g(x) \sim x^{-\delta/2},$5 (Park, 2024). Combining FoCal with existing forward $x g(x) \sim x^{-\delta/2},$6 data from LHCb further constrains the small-$x g(x) \sim x^{-\delta/2},$7 gluon nPDF (Park, 2024). This makes direct photons a central FoCal observable in global nPDF fits.

Neutral mesons

FoCal-E reconstructs neutral mesons such as $x g(x) \sim x^{-\delta/2},$8, $x g(x) \sim x^{-\delta/2},$9, and $\delta$0 through $\delta$1 decays and invariant-mass analysis of cluster pairs (Park, 2024). Backgrounds include combinatorial pairs, cluster splitting, and correlated contributions from other meson decays (Park, 2024). Polynomial fits, event mixing, and a rotational method were studied for combinatorial-background estimation, with the rotational method providing the best description, especially below the $\delta$2 mass peak where correlated backgrounds are important (Park, 2024). After subtraction, the signal is well described by a Crystal Ball function, and realistic pp simulations at $\delta$3 TeV show a clear $\delta$4 peak (Park, 2024).

These forward meson yields in pp and p–Pb constrain low-$\delta$5 gluon PDFs and nuclear modification effects, complementing the cleaner but rarer direct-photon channel (Park, 2024).

Forward jets

FoCal reconstructs jets with the anti-$\delta$6 algorithm and radius parameter $\delta$7 at both particle and detector level (Park, 2024). Performance is quantified using

$\delta$8

with the Jet Energy Scale defined as the mean and the Jet Energy Resolution as the RMS or Gaussian width (Park, 2024).

For jets with centroid $\delta$9, the JES becomes more negative up to about 600 GeV and then stabilizes, while the JER remains below $x$0 up to 3 TeV under a Gaussian-fit definition (Park, 2024). Numerical integration yields slightly worse JER below 400 GeV, and further refinements are possible by accounting for biases in the neutral energy fraction (Park, 2024).

These forward jet measurements in pp and p–Pb are intended to constrain small-$x$1 dynamics via jet spectra, dijet balances, and correlations with rapidity and energy (Park, 2024).

Photon–hadron correlations

FoCal studies isolated photon–hadron correlations as a proxy for direct photon–hadron correlations, with an isolated photon candidate in FoCal correlated against $x$2 candidates (Park, 2024). The azimuthal correlation $x$3 at detector level exhibits no near-side peak, a small dip at $x$4, and a clear away-side peak corresponding to the recoil jet (Park, 2024). Gaussian fits to the away-side peak provide a width that narrows with increasing photon and associated $x$5 transverse momentum, indicating a more collimated recoil jet (Park, 2024). For Run-4 pp luminosity of $x$6, the relative statistical uncertainty on that width is below $x$7 (Park, 2024).

A plausible implication is that such precision makes the correlation width a sensitive discriminator of broadening or suppression effects in p–Pb relative to pp, which the FoCal program interprets in terms of transverse-momentum broadening and saturation-driven modification (Park, 2024).

5. Ultra-peripheral collisions and vector-meson photoproduction

FoCal provides access to heavy vector-meson photoproduction, specifically $x$8 and $x$9, in p–Pb and Pb–Pb ultra-peripheral collisions (Park, 2024). In this channel, one beam particle acts as a photon emitter and the other as target, and at LO pQCD the cross sections are proportional to the square of the gluon density, making them especially sensitive to gluon distributions (Park, 2024).

The accessible photon–proton energy extends up to

$Q_{\rm sat}$0

well beyond current ALICE data and overlapping with future EIC coverage (Park, 2024). Projections show sensitivity to deviations from simple power-law growth in $Q_{\rm sat}$1, with such deviations expected when saturation becomes relevant (Park, 2024). Coherent $Q_{\rm sat}$2 and $Q_{\rm sat}$3 photoproduction can be reconstructed through $Q_{\rm sat}$4 decays in FoCal-E, and the invariant-mass distribution of cluster pairs is described by a sum of double-sided Crystal Ball functions that clearly separates the two resonances (Park, 2024).

A related performance note emphasized that these UPC measurements probe nuclear shadowing and saturation at unprecedented energies in ALICE (Huhta, 2024).

6. Operation, timeline, and role within ALICE and the LHC

FoCal is an approved ALICE upgrade project, with LHCC approval in March 2024 (Park, 2024). The documented timeline is 2024 to mid-2027 for module mass production and assembly, early 2028 for installation, 2028 for commissioning, and data taking from 2029 onward in Run 4 (Park, 2024).

Operational challenges include radiation hardness, occupancy and data rate, and cooling (Park, 2024). The design relies on LHC-proven technologies such as ALPIDE and HGCROC, studied radiation levels in the FoCal region, fast MAPS readout at 1.2 Gbps and 400 Mbps, online data compression, O$Q_{\rm sat}$5 processing, and integrated cooling compatible with the compact support and services (Park, 2024).

Within ALICE, FoCal complements ITS2 and the TPC at central rapidity by extending sensitivity to extremely small $Q_{\rm sat}$6 in the forward region (Park, 2024). In LHC context, it is commonly contrasted with LHCf, CMS HF, ATLAS forward calorimetry, CASTOR, and LHCb, but its combination of $Q_{\rm sat}$7 coverage and extreme EM granularity is distinctive (Park, 2024).

A common misconception is that FoCal is merely a generic forward calorimeter addition. The documented design instead indicates a detector whose electromagnetic granularity, longitudinal sampling, and mixed pad/pixel architecture are explicitly driven by the physics requirement of separating isolated photons from dense decay backgrounds in a forward environment (Park, 2024). Another possible misconception is that FoCal is only a photon instrument; in fact, the detector concept and program include jets, hadronic isolation, meson reconstruction, photon–hadron correlations, and UPC vector mesons (Park, 2024).

Taken together, the detector concept, prototype performance, and projected physics reach indicate that FoCal is intended to turn ALICE into a forward small-$Q_{\rm sat}$8 QCD laboratory, with direct sensitivity to gluon PDFs and non-linear QCD evolution in a region of phase space that is currently only weakly constrained (Park, 2024).