CERN Accelerating science

Transforming NA62

The main goal of the NA62 experiment, located in CERNs North Area, is to precisely measure the K+→π+νν̅  branching fraction, which is highly sensitive to new physics processes. The analysis of data collected during 2016-2018 led to the most precise determination of the K+→π+νν̅  branching fraction and provides the strongest evidence so far (3.4σ) for the existence of this process. In recognition of these encouraging results, NA62 has been approved to continue running until Long Shutdown 3.

The primary physics goal for the upcoming years is to significantly improve the accuracy of the K+→π+νν̅  branching fraction measurement, with an aim to determine the branching fraction to within 10%, matching the precision of theory computations. To achieve this ambitious goal, NA62 has exploited Long Shutdown 2 to identify and prepare several key improvements to the detector. These improvements include an upgrade of the GigaTracker detector, the design and installation of a new “veto counter” detector, and an upgrade of the HASC detector. Each upgrade acts to reduce residual backgrounds identified during the data analysis.

Alongside the main physics goal, NA62 also plans to reach unprecedented sensitivity in the investigation of several Standard Model (SM) extensions involving feebly-interacting long-lived particles. To this end, NA62 has scheduled dedicated periods during which 1018 proton-on-target interactions will be recorded while operating NA62 as a beam-dump experiment. To mitigate large backgrounds identified in a small dataset collected in beam-dump mode during 2016-2018, which arise due to muons produced in the beam-dump target passing through the experiment, a new scintillator detector called the ANTI-0 has been installed at the entrance of the NA62 fiducial volume.

Following these updates, the NA62 experiment is poised to exploit protons from the SPS until Long Shutdown 3, continuing the hunt for new physics. However, the upgrades described in this article are not the only steps being taken to exploit the full potential of the NA62 detector. New electronics for the trigger and readout systems are currently being finalised, with an eye to the beginning of the data-taking in 2022. 

The NA62 GigaTracker upgrade

The GigaTracKer (GTK) is the NA62 beam spectrometer, which utilizes several stations of silicon pixel detectors to measure the time and momentum of each particle in the 800 MHz hadron beam as they enter the NA62 experiment. The GTK achieves a momentum resolution of ∆P/P ∼ 0.2%, an angular resolution of 16 µrad and a time resolution of O(70 ps). In order to reduce the inelastic interactions between the beam particles and the detector, the material budget has been kept as low as possible. This translated into a configuration of three detector planes within a magnetic field to allow the momentum measurement. The absence of geometrical redundancy is compensated by an excellent pixel time resolution of O(140 ps), allowing particle tracks to be reconstructed unambiguously. 

Although this technique is well established now, this required an extensive R&D program when the detector was designed in 2007. CERN played a major role in the development of the detector, especially concerning the readout chips and the cooling technology. Each station of the GTK has an active area of ∼ 60 × 30 mm2 made by a silicon sensor attached to two rows of five TDCPix readout-chips, for a total of 18000 pixels. The heat produced by the electronics is dissipated by a micro-channel cooling plate, representing the first use of this technology in High Energy Physics. The material budget of each station is around 0.5% interaction lengths, corresponding to a thickness of 510 µm. The detectors are installed inside the vacuum of the NA62 beam pipe, with the possibility of easily replacing them in case of damage resulting from the high-radiation environment in which they operate.

Since the 2016-2018 data-taking an additional station of the GTK has been installed. The extra station was motivated by careful study of the systematic uncertainty of the K+→π+νν̅  analysis, where the main contribution was determined to be from interactions of the beam in the GTK material and/or a mis-matching of the GTK track with the π+ candidate. The additional GTK station will help to increase the tracking efficiency and reduce the background. The new station (GTK0) has been installed close to GTK1, in a new vessel which allows installation of the two detectors from either side of the beam pipe (Fig. 1). The position of GTK0 was chosen to balance the improvement in the tracking performance with the increased backgrounds from beam interactions with the detector material. Studies based on the data collected during 2016-2018 indicate that the detector lifetime is larger than expected, therefore the increased number of stations can be sustained.

 Figure 1. Vessel containing GTK stations 0 (right) and 1 (left). The beam direction is from bottom to top.

The works related to the installation of GTK0 ended in June 2021, with the detector installed and ready to take data since the beginning of the 2021 data-taking. The hitmap of the four GTK stations with the beam profile measured by them is shown in Fig.2. The processing and analysis of the data is ongoing.

Figure 2. Beam profile as seen by each of the GigaTracker stations. 

The NA62 Veto Counter detector

During the analysis of the 2016-2018 data, one of the main sources of background in the K+→π+νν̅  analysis was identified to be K+ in the beam decaying while passing through the GTK. The new Veto Counter detector was designed to detect the decay products of background K+ decays, typically charged pions and photons, before they are absorbed by the final collimator. This important task is complicated by the presence of an intense muon halo around the NA62 hadron beam. A major challenge when designing the Veto Counter was to achieve excellent detection efficiency for the kaon decay products while also being able to distinguish between those and the muon halo. The Veto Counter design goals are a time resolution close to 200 ps and a detection efficiency above 99%.

The final design of the Veto Counter has three planes of rectangular plastic scintillator tiles placed only a few mm from the beam. Scintillation light is collected on both sides of each tile by fast photomultipliers coupled through light-guides. The first and third detector planes are dedicated to minimum ionizing particles (MIPs), the latter being located just after the collimator. The second plane is preceded by a layer of lead, which acts as a photon converter, and is dedicated to the detection of the ensuing electromagnetic shower. With this configuration, a rough particle identification can be performed: a pion will leave a signal on the first plane only, a photon on the second plane only, and a muon on the first and last planes, as it is the only particle that can pass through the collimator without being absorbed. With this information, the muon halo can be distinguished from the pions and photons produced in K+ decays.   

Figure 3. Gluing of the light-guides on the scintillator tiles. They have been processed one side at a time in several batches. A curing time of 24h was necessary for the glue.

To develop the Veto Counter, a test setup was established at a collaborating institute in Louvain-La-Neuve, Belgium. Several test modules were assembled to measure the efficiency and time resolution with different plastic scintillators in order to optimise the choice of detector materials. The mechanical design of the detector and its vessel started in July 2020 in Louvain-La-Neuve in close collaboration with CERN: a challenging aspect of the design was to fit the detector in the limited space, of about 15 cm, taking into account obstructions due to other elements of the NA62 beam line. The construction of the detector itself took place in Louvain-la-Neuve, CERN, and at an external company. Fig. 3 shows the gluing of light guides to the scintillator tiles, while Fig. 4 shows the assembly of one part of Veto Counter station.

The full Veto Counter detector was successfully installed in the experiment and commissioned during the 2021 data-taking. Data from the Veto Counter were acquired successfully through both paths of a dual-readout system: one based on the common readout electronics of NA62; the other based on the FELIX system currently being developed at CERN. New electronic boards were also developed and tested during 2021 that will allow the full potential of the Veto Counter to be exploited in 2022 and beyond.

Figure 4. The assembly of one part of Veto Counter station 1 in its aluminium frame. In the center, the scintillator tiles held by the light guides are ready to be light-tight wrapped in the black paper. On either side, the PMTs and their voltage dividers are firmly held in contact with the light guides. The back station has already been cabled, the front one will be cabled on the connectors after the front cover is fitted.

Successful upgrade of HASC

Another important piece of the upgraded NA62 detector is the hadron sampling calorimeter (HASC). The HASC was originally designed to veto rare topologies of the K+ → π+π+π- decay that are background to the K+→π+νν̅  analysis. In 2018, the K+→π+νν̅  analysis revealed that the HASC is also effective at mitigating backgrounds from certain configurations of the K+ → π+π0 decay, providing an additional 30% background reduction. After extensive studies, it was understood that the HASC is able to detect positrons created when the γ produced from the π0 decay interacts with the beam pipe. An upgrade of the HASC was proposed whereby a second calorimeter station would be placed on the opposite side of the beam axis in order to double the π0 rejection. 

Each HASC station is made up of 9 identical modules recovered from a prototype developed by NA61 Collaboration. Each module is a sandwich of lead plates interleaved with scintillator tiles, organized in 10 longitudinal read-out sections, with each scintillator tile optically coupled via round wavelength shifting fibers. In the rear side of each module there are 10 optical connectors coupled to silicon photomultiplier (SiPM) sensors.

Figure 5. Time difference between the muon-veto detectors MUV3 signal tagging a muon and the closest in time signal from the respective HASC station. The blue (orange) curve corresponds to muons incident on the old (new) HASC station.

In 2019 nine new HASC-like calorimeter modules were purchased from the NA61 Collaboration. The team from National Institute for Physics and Nuclear Engineering (IFIN-HH) in Bucharest, Romania, developed a new mechanical structure to support the SiPM sensors and the associated front-end electronics (FEE), as well as a new FEE board to read-out the 10 SiPM sensors. A hybrid cooling system based on a water-air heat exchanger and Peltier coolers was also designed and produced to cool the FEE and SiPM’s installed on the new HASC station to around 21ºC. Due to mechanical incompatibilities, the old HASC station is equipped with air cooling only, which lowers the temperature of the SiPM’s from 36 ºC to about 24 ºC.

The modules of the new HASC were tested with cosmic rays in the IFIN-HH lab during 2019-2020. In June 2021 the installation and commissioning of the new HASC was finalized, along with the cooling system of the old HASC. Fig. 5 shows the time resolution of the two HASC stations, which are both less than 200 ps. The difference in time resolution between the new and old stations is caused by the difference in operating temperature.

Installing the ANTI-0 detector

The final addition to the NA62 detector made during Long Shutdown 2 is the ANTI-0 detector. The ANTI-0 detector is a plastic-scintillator hodoscope that covers the entrance of the NA62 fiducial volume, which is designed to detect muon halo entering the detector in the region 118-1080 mm transverse to the beam axis. This setup allows halo muons to be distinguished from muons produced in kaon decays inside the fiducial volume. However, the ANTI-0 is even more important when the experiment is operating as a beam-dump experiment, where the 400 GeV/c proton beam from the SPS is dumped on a target absorber close to the NA62 detector. In this case, the ANTI-0 is responsible for identifying muons produced in the target absorber, which constitute the main background to searches for feebly-interacting long-lived particles that decay inside the fiducial volume.  

Figure 6. The top half of the ANTI-0, with the Scintillator tiles are seen wrapped in white paper

Figure 7. The ANTI-0 detector installed in the NA62 experiment. The detector is seen as the square silver box at the centre of the image.

The ANTI-0 detector was successfully prepared by close collaboration between scientists at CERN and at the Institute for High Energy Physics (IHEP) in Protivino, Russia. The detector consists of 280 square plastic scintillator tiles, see Fig. 6. Each tile covers an area of 124 × 124 mm, with the Scintillation light collected via four SiPMs acting as a single electronic channel. The SiPMs are placed in slots cut in a Plexiglas plate at a distance of 40 mm from the horizontal edges of the scintillator to help with uniformity of the light collection. The tiles are placed in a chessboard style, with the tiles placed alternately on either side of an aluminum central support plate. Each scintillator overlaps by 4 mm with the four neighboring tiles on the other side of the central plane, ensuring better than 99.9% geometrical coverage.

The ANTI-0 detector was constructed and installed during Long Shutdown 2, and collected its first data during 2021. The installed detector is shown in Fig. 7. The overlap of the tiles can be seen by plotting the coordinates of reconstructed tracks extrapolated more than 80 meters from the NA62 spectrometer to the position of the ANTI-0, having selected events with exactly two coincident hits in the ANTI-0 (Fig. 8). The distribution illustrates both the precision of the NA62 spectrometer, as well as the lack of gaps in the geometrical coverage of the ANTI-0

Figure 8. Histogram with XY distribution of tracks reconstructed and extrapolated to the ANTI-0 surface for events with 2 hits in ANTI-0.


The NA62 experiment is a collaborative effort of laboratories and institutes around the globe. The successful completion of the upgrades described in this article testify the ability of the diverse team behind the experiment to push through challenges and accomplish the goals of the upgrade programme, even when access to the usual facilities was restricted and remote working became the norm. Building, testing and operating the different subdetectors of NA62 paves the way to an exciting physics programme during the next runs. The successful installation of the upgraded systems and the plethora of ideas for the future shows that Kaon physics at CERN is far from over!