The COMPASS++/AMBER Collaboration proposes to establish a “New QCD facility at the M2 beam line of the CERN SPS”. Such an unrivalled installation would make the experimental hall EHN2 the site for a great variety of measurements to address fundamental issues of strong interactions. The proposed measurements cover a wide range in the squared four-momentum transfer Q²: at lowest values of Q² we want to determine the proton charge radius through elastic muon-proton scattering, at intermediate Q² we want to perform spectroscopy of mesons and baryons by using dedicated meson beams, and at high Q² we plan to study the structure of mesons and baryons via the Drell-Yan process. In our Letter of Intent (LoI) [1], which was recently submitted to the SPSC, we have described physics goals, sensitivity reach and competitiveness for such a future general-purpose fixed-target facility at CERN.

In phase-1, starting in the year 2022, we plan to perform three experiments making use of the existing M2 beam line that provides muons as well as pions and protons: (1) an accurate measurement of the electric form factor G_E of the proton at small values of the squared four-momentum transfer Q² to extract the proton charge radius, using the high energy muon beam; (2) a study of the pion-induced Drell-Yan process to better determine the poorly known parton distribution functions (PDFs) of the pion and study nuclear PDFs; (3) a determination of the antiproton production cross section by scattering high energy protons on proton and helium-4 targets. A better knowledge of this cross section will improve the accuracy in interpretating the existing results from indirect Dark Matter Searches, as those obtained by the AMS-02 [2] experiment on the International Space Station. A proposal comprising all presently available details on these three measurements was recently submitted to the SPSC [3]. Brief summaries are given below in sections 2 to 4.

Beyond LS3, we propose to upgrade the M2 beam line by installing an RF-separation stage that will provide high-intensity and high-energy beams of charged kaons with a high level of purity. Such an upgrade is presently under study by CERN EN-EA in the framework of the Physics-beyond-Colliders Initiative. Once realised, it would make the CERN SPS M2 beam line unique in the world for many years to come. A brief summary is given below in sect. 5.

With RF-separated kaons, the virgin field of high-precision strange-meson spectroscopy becomes accessible, a first measurement of the kaon polarisability can be performed using the Primakoff process and measurements of the Drell-Yan process will allow access to the presently practically unknown quark-gluon structure of kaons. A variety of further measurements is proposed in the LoI, using RF-separated kaons or antiprotons in conjunction with various experiment-specific installations in the target region, including spectroscopy with low-energy antiprotons, spin-dependent Deeply Virtual Compton Scattering and Deeply Virtual Meson production, meson induced prompt-photon production, etc.

Novel instrumentation using modern detector architecture will be constructed and installed in the experimental hall EHN2, where the upgraded multi-purpose two-stage magnetic spectrometer will serve as experimental backbone of the new facility. Upgrades will be designed to serve for as many individual experiments as possible and installed along the lifetime of the facility according to actual needs and availabilities. As an example, we describe below the planned triggerless data acquisition system (see sect. 6).

The full project is expected to stretch across the next 10 to 15 years. As it continues to attract physicists world-wide, the physics scope of the facility should remain open for future exciting ideas, using either (RF-separated) hadron beams or the muon beam. Proposals for further measurements, based upon ideas already discussed in the LoI or possible new ones, will be submitted in due time.

Proton charge-radius measurement using muon-proton elastic scattering

In spite of many years of intense activity, the proton-radius puzzle remains unsolved up to now. An eventual explanation of this mismatch requires four key measurements: elastic lepton scattering and finite-size effects in atomic levels, in both cases using electrons and muons. In contrast to the atomic spectroscopy approaches, scattering experiments do not determine the proton radius directly, but by measuring the Q²-dependence of the electric form factor over an extended range and then extrapolating the form factor linearly towards Q² = 0. Results are available for three types of experiments, but not yet for muon-proton scattering. To date, a discrepancy as large as 5 standard deviations exists between the two most recent precision measurements: r^rms_CREMA = 0.841 ± 0.001 fm from line-splitting measurements in laser spectroscopy of muonic hydrogen and r^rms _MAMI = 0.879 ± 0.008 fm from elastic electron-proton scattering.

The proposed measurement aims at a precision determination of the electric mean-square charge radius of the proton. This approach constitutes the “missing element” in the above mentioned fourfold variety of approaches to solve the proton radius puzzle. The Q²-dependence of the cross section in high energy muon-proton elastic scattering will be measured over the range 0.001 < Q²/(GeV²/c²) < 0.04 in order to constrain its Q²-slope near zero. The sensitivity of this measurement is illustrated in Fig. 1. The projected statistical accuracy on the proton radius is 0.01 fm or better with a considerably smaller systematic uncertainty. Using high-energy muons instead of electrons is highly advantageous, as several experimental systematic effects and also theoretical (radiative) corrections are considerably smaller.

The accuracy to be reached by the proposed muon-proton scattering experiment is expected to be comparable to that obtained in electron-proton scattering at MAMI. Comparing the results on the proton charge radius from these two complementary measurements performed with very similar techniques will allow to probe interpretations of the proton radius mismatch to be caused by lepton flavour effects.

Figure 1: The ratio between the electric form factor of the proton and its dipole formula. The sensitivity of the proposed COMPASS++/AMBER measurement is shown as a band around the line corresponding to a proton charge radius of 0.84 fm.

The measurement will employ a time-projection chamber filled with pure hydrogen up to pressures of 20 bar, which serves at the same time as a target and as detector gas. The high-pressure (up to 20 bars) hydrogen TPC (see Fig. 2) will be built for the proton-radius measurement following a design similar to that used for the electron-proton scattering experiment in Mainz. The main difference is the length of the active target region: the Mainz version has one 400 mm long drift cell, the new COMPASS++/AMBER TPC will have either two or four drift cells of 400 mm. The TPC will operate in ionisation- chamber mode, i.e. with no gas amplification. It will accurately measure the recoil-proton energy, the recoil-proton angle and the coordinate of the interaction point along the beam direction.

 Figure 2: Engineering design for the four-cell hydrogen TPC

Drell-Yan and J/ψ production using the conventional M2 hadron beam

The experimental determination of the meson structure remains a long-awaited and critical input to theoretical efforts that seek to explain the emergence of massive composite hadrons, including the large mass difference between pion and proton. A major step forward in the determination of the nearly unknown pion and kaon parton distribution functions is the main objective of the planned measurements, which will also provide benchmarks for testing recent predictions of non-perturbative QCD calculations performed on the lattice or in the framework of the Dyson-Schwinger equations. At medium and large values of Bjorken-x, a quantitative comparison between the pion and the kaon valence-quark distributions will become possible. At smaller values of Bjorken-x, improved knowledge on the onset of sea-quark and gluon distributions in the meson will help in explaining the differences between the gluon contents of pions, kaons and nucleons, and shed light on the mechanism that generates the hadron masses.

In order to determine the shape of the sea-quark distribution in the pion and better constrain the region of phase space in the Bjorken-x variable corresponding to x_π > 0.1, data will be collected with pion beams of positive and negative charge impinging on a light isoscalar target. Figure 3 shows accuracy estimates for the ratio Σ_sea/Σ_val as a function of x_π, in the dimuon mass range 4.0 < M_µµ/(GeV/c²)< 8.5, which will be accessible with a good mass resolution thanks to new vertex detectors. The curves labelled SMRS represent the predictions for three possible contributions of the sea quarks to the pion momentum, ranging from 10% to 20%. The three different assumptions for the pion sea yield increasingly different predictions for x_π values below 0.5.

Figure 3: The ratio Σ_sea/Σ_val as a function of x_π, for three different sea-quark distributions (10%,15% and 20%). The shown statistical accuracy is expected when using the foreseen COMPASS++/AMBER data-taking conditions.

Furthermore, an analysis that simultaneously accounts for the differential cross section and for the degree of polarization of the produced charmonia resonances is expected to provide stringent experimental constraints on their production mechanisms. Hence J/ψ production provides an alternative access to both quark and gluon distributions in the incoming meson. In parallel to meson-structure measurements, the availability of heavier nuclear targets in the setup will allow the study of cold nuclear effects such as nuclear PDFs and parton energy loss.

Measurement of proton-induced antiproton production cross sections

The purpose of this experiment is the measurement of the antiproton production cross sections in proton-proton and proton-⁴He scattering for projectile energies from several ten to a few hundred GeV. In combination with similar measurements by LHCb in the TeV range, the COMPASS++/AMBER measurements will provide a fundamental data set that is expected to allow for a significantly higher accuracy of the predicted natural flux of antiprotons in the galactic cosmic rays. This is of great importance as the indirect detection of dark matter (DM) is based on the search for products of DM annihilation or decay, which are expected to appear as distortions in the spectra of rare cosmic ray components like positrons, antiprotons, or even antideuterons. The new data set will thus substantially improve the sensitivity of existing (and future) very accurate AMS antiproton flux measurements to DM signals, which is presently limited by the poor knowledge of the antiproton production cross sections.

In order to be able to profit from the AMS-02 high-precision data, a similar accuracy has to be achieved in the computation of the antiproton source term for all the production channels. Figure 4 reports the extrapolated AMS relative uncertainty on the antiproton/proton ratio. The collection of new data using a proton beam with energies ranging between 60 and 280 GeV in conjunction with a ⁴He (or H) target would allow to extensively characterise the antiproton production spectrum. This is a key point to derive and/or constrain antiproton production models, which in turn may lead to a further decrease in the overall uncertainty on the antiproton production cross section.

Figure 4: Relative uncertainty afflicting the prediction for the p̄/p ratio, shown in dependence on the rigidity p/Ze (expressed in GigaVolt).  In light blue the up-to-date astrophysical uncertainty (based on AMS-02 data analysis), in dark yellow the mean of the nuclear physics uncertainties. In black for comparison the AMS-02 measurement uncertainties as reported in [2]

The existing M2 hadron beam line with its momentum range between 20 and 280 GeV/c is an ideal place to perform this measurement. The double-differential antiproton production cross section will be measured using the spectrometer in EHN2 equipped with liquid-hydrogen or liquid-helium targets and using the antiproton-identification capabilities of the RICH detector. Measuring for several beam momenta the cross section in 20 bins each for antiproton momentum and pseudorapidity, a 1% statistical uncertainty will be reached for the cross section with an anticipated point-to-point systematic uncertainty of less than 5%. 

RF-separated beams

A study of a possible enrichment of desired particle species in the M2 beam has been launched by EN-EA in the context of the Physics Beyond Colliders Initiative.

While alternative techniques can be employed for particles at lower energies, the method of RF-separation is the only effective way to provide the desired high-purity kaon and antiproton beams in the M2 beam line. The method of RF-separation was first employed at CERN in the 1960s based on ideas of Panofsky and Schnell and it is based on exploiting the different velocities of particle species in a beam with defined momentum. As displayed in Fig. 5, two dipole RF cavities (RF₁ + RF₂) with frequency f are implemented at a given distance L. The transverse kick of RF₁ is either amplified or compensated by RF₂ depending on the phase difference that is given by the difference of velocities of the various particle species. For two species i (i = 1, 2) with masses m_i and velocities β_i, this quantity reads ΔΦ = 2π(Lf/c)(β₁⁻¹ − β₂⁻¹).  In the limit of large momenta, ΔΦ can also be expressed as a mass difference between the two species at the beam momentum p.

For kaons as wanted particles, the phase difference could be chosen at ΔΦ_πp = 2π, which results in ΔΦ_πK = 94^◦. This means that the kick for both protons and pions would be compensated by RF2 and they would be absorbed in the beam stopper. The kaons would receive a close-to-maximum transverse kick and mostly go around the stopper. For antiproton beams, the phase difference could be chosen at ΔΦ_πp̄ = π, which results in ΔΦ_p̄K = 133^◦ and ΔΦ_p̄e = 184^◦. In this case, the antiprotons would receive an acceptable deflection while electrons and pions are dumped effectively.

Figure 5: Panofsky-Schnell method for RF-separated beams. The unwanted particles (red) are stopped by a beam stopper while the wanted particles (green) receive a net deflection by the combination of the RF₁ and RF₂ dipole RF cavities out of the central axis, which is sufficient to go around the stopper.

In the current optics, the two FODO sections of the M2 beam have been kept unchanged, which has the benefit of keeping the possibility to change back to the muon beam configuration at a rather moderate cost and within a shorter time compared to a complete change of all M2 beam line elements. Depending on evolving requirements and further optimisation, the option to go back to muon beams could be checked in more detail.

The ”triggerless” DAQ

The development of a data acquisition system for the COMPASS++/AMBER experiment is challenged by diverse requirements imposed by the wide physics program and difference in detector compositions, and the requirement of high precision measurements (i.e. high statistics and high beam rates). These requirements can be met as the rapid development of technology allows for a transition  from the classical trigger based data acquisition to a continuous read-out scheme, in which detector subsystems deliver continuous time-stamped data streams for real-time processing in later stages of the DAQ (e.g. High Level Triggering / Feature extraction) and dead time free storage of the whole data stream.

The development of the Compass DAQ in the past led to a new design, where most of the traditional computers were substituted with FPGAs. This new iFDAQ was introduced for the COMPASS run in 2014 and successive further development has led to a very stable and modular DAQ system, which was successfully used in the last years of the Compass II data taking. For the design of the new data acquisition system it is proposed to adopt a rather far-looking approach to allow the use in a wide range of physics cases and needs in the Compass++/AMBER framework. The logical step is to go to a continuous data acquisition with a digital trigger system, which is tightly integrated in the iFDAQ framework. (Fig. 6).

Figure 6: Evolution of iFDAQ.

For this approach all detectors have to be equipped with front-end electronics, which is able to digitize the analog signals of the detectors in real time, perform zero suppression, and stream the data continuously to the next DAQ level. After the front-ends, the data is split into two data streams. Each detector sends a data-stream to the DAQ, where it is buffered until the trigger decision is made. The front-ends of the detectors, which participate in the trigger decision, have a second data stream that is sent to the digital trigger processor and is used to make the selection of data that is later stored to disk.

The trigger processor is a multi stage FPGA unit where the first stage corrects the data with calibrations and builds time correlated event candidates. These so-called "Events of Interest" are then further processed in several stages to work out the trigger decision, according to the different physics programs. The output of the hardware trigger is broadcasted via the Time and Control System to all DAQ modules where the corresponding data are extracted and sent to the hardware event builder, then distributed between online computers and stored on local disks. The DAQ can run in two modes, i.e. the complete not-triggered mode where everything is directly written to disk and the triggered mode where sophisticated trigger algorithms are used to reduce the outgoing bandwidth to 20 Gbit/s, which is the maximum sustained bandwidth to the central data storage at the moment. The mode of operation and the needed reduction factor depend on the physics program and covers a wide range. The modularity of the DAQ enables us to scale the system according to the different requirements. 

Future Reading

[1] B. Adams et al., Letter of Intent: a “New QCD facility at the M2 beam line of the CERN SPS” (COMPASS++/AMBER), ArXiv:1808.00848, Tech. rep., CERN- SPSC-2019-003 (SPSC-I-250).

[2] M. Aguilar et al., ”Antiproton Flux, Antiproton-to-Proton Flux Ratio, and Properties of Elementary Particle Fluxes in Primary Cosmic Rays Measured with the Alpha Magnetic Spectrometer on the International Space Station” Phys. Rev. Lett. 117 (2016) 091103

[3] B. Adams et al., The COMPASS++/AMBER Collaboration, ”Proposal for Measurements at the M2 beam line of the CERN SPS, Phase-1: 2022-2024”, CERN-SPSC-2019022, SPSC-P-360, June 5, 2019.