Performance of the ALICE Inner Tracking System 2 Journal: Nuclear Instruments and Methods in Physics Research A

Nicolò Valle, on behalf of the ALICE Collaboration · 2025

Read Performance of the ALICE Inner Tracking System 2 Journal: Nuclear Instruments and Methods in Physics Research A by Nicolò Valle, on behalf of the ALICE Collaboration (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-17233.txt.

Introduction

Nicolo Valle reports on the inner tracking system of the ALICE experiment for the full collaboration in two thousand twenty-five. ALICE studies matter made in high-energy crashes at the Large Hadron Collider near Geneva. Its upgraded tracker uses pixel sensors built into the chip itself, with no separate readout layer.

The device spans about ten square meters and holds twelve and a half billion pixels in seven rings. Fine detail, low mass, and a first ring close to the crash point suit it for tracing short-lived particles and slow tracks. Valle sums up how the tracker ran in the third run of the collider, how it was tuned, and what the run taught the team.

Catalog record

CREATORNicolò Valle, on behalf of the ALICE Collaboration
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE18 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-17233

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Performance of the ALICE Inner Tracking System 2 Journal: Nuclear Instruments and Methods in Physics Research A Nicolò Valle, on behalf of the ALICE Collaboration Abstract The upgraded Inner Tracking System (ITS2) of the ALICE experiment at the CERN Large Hadron Collider is based on Monolithic Active Pixel Sensors (MAPS). With a sensitive area of about 10 ​ m 2 10\,\mathrm{m^{2}} and 12.5 billion pixels, ITS2 represents the largest pixel detector in high-energy physics. The detector consists of seven concentric layers equipped with ALPIDE pixel sensors manufactured in the TowerJazz 180 ​ nm 180\,\mathrm{nm} CMOS Imaging Sensor process. The high spatial resolution and low material budget, in combination with small radial distance of the innermost layer from the interaction point, make the detector well suited for secondary vertex reconstruction as well as for tracking at low transverse momentum. This paper will present the detector performance during the LHC Run 3 and give an overview on the calibration methods and running experience. 1 The ITS2 detector concept The ALICE experiment at the LHC underwent a major upgrade during the 2019-2021 Long Shutdown, targeting high-precision studies of the quark-gluon plasma with low transverse momentum ( p_{\mathrm{T}} ) probes. The upgrade includes a beam pipe with reduced outer radius, moved from 28 mm to 18 mm, a new Inner Tracking System (ITS2), a new Muon Forward Tracker (MFT), a new GEM-based TPC readout, and a new Fast Interaction Trigger (FIT) detector. Furthermore, the readout and trigger system has been extensively upgraded to record Pb–Pb collisions at up to 50 kHz in continuous mode, ensuring sensitivity to signals that do not have a triggerable signature. The new ITS2 significantly enhances ALICE’s tracking capabilities, especially at low p_{\mathrm{T}} [ 1 , 2 ] . It consists of seven cylindrical silicon pixel layers, made of ALPIDE chips arranged in longitudinal staves as shown in Fig. 1 , for a total of 12.5 billion pixels covering a sensitive area of about 10\,\mathrm{m^{2}} . The innermost three layers form the Inner Barrel (IB) and are installed at radii between 22 and 40\,\mathrm{mm} . The outermost layer of the Outer Barrel (OB) is located at a radius of 40\,\mathrm{cm} . The ALPIDE (ALICE Pixel Detector) chip is a 3\times 1.5\,\mathrm{cm^{2}} matrix of 1024\times 512 monolithic pixel sensors with O(30\,\mathrm{\muup m}) pixel pitch, reaching an intrinsic spatial resolution of about 5\,\mathrm{\muup m} . It is manufactured in the TowerJazz 180\,\mathrm{nm} CMOS imaging process with a 25\,\mathrm{\muup m} thick active layer allowing to reduce the silicon thickness to just 50\,\mathrm{\muup m} in the IB. Together with an ultra-light mechanical support, this allowed to achieve a material budget of 0.36\%\,X_{0}/\mathrm{layer} for the IB and 1.1\%\,X_{0}/\mathrm{layer} for the OB. Full CMOS circuitry is implemented directly into the pixel cell, which contains the sensing diode, amplification and shaping stage, discriminator, and a digital section with multi-event buffers and mask registers. A pulse injection capacitor allows to inject test charges into the input of the analog front-end, a crucial feature for calibration as explained in section 2 . Additionally, the memory cells of the pixels can be directly set, to emulate hits with any desired pattern. The ITS2 Detector Control System has been programmed to simulate physics events using this feature, enabling the commissioning of the entire data acquisition chain under realistic particle load conditions even without beam collisions. Figure 1: ITS2 layout and photo of an ALPIDE chip. The periphery of each chip houses control and readout functionalities, along with 14 DACs for biasing the pixel front-end circuitry. These include registers for tuning the discrimination charge threshold, which can be adjusted individually for each chip. The chip matrix is read out with zero suppression using high-speed links supporting 1.2 Gb/s. In the OB, a slower parallel data port is used for inter-chip data transfer, allowing for the integration of compact multi-chip modules where a single high-speed link handles the readout of seven chips. Each of the 192 staves is connected to a FPGA-based Readout Unit (RU), working in the 0.5\,\mathrm{T} magnetic field of ALICE and in radiation environment ( 10\,\mathrm{krad} of TID [ 3 ] ). The ITS2 has been operated so far in continuous mode. The RU firmware generates periodic triggers with a programmable rate, steered by a signal received with the frequency of the LHC orbit of approximately 11\,\mathrm{kHz} . Normally, a framing rate of 202\,\mathrm{kHz} is used for pp data acquisition, with the ALICE interaction rate usually leveled at 500\,\mathrm{kHz} . During Pb-Pb data acquisition, where the hadronic interaction rate in ALICE was pushed to 50\,\mathrm{kHz} for the first time, a framing rate of 67\,\mathrm{kHz} is used. More details on the detector system architecture and the ALICE software system can be found in [ 1 , 2 ] . The remainder of this work focuses on the ITS2 calibration methods and presents a selection of operational highlights and physics results. 2 Calibration: charge threshold, noisy pixels and acceptance maps The optimization of the in-pixel threshold is achieved by injecting test charges into the input of the analog front-end. The process is fully managed by the ITS2 Detector Control System, which triggers the chips at a custom frequency of 1 kHz using the internal sequencer equipped in the RUs, independent of the ALICE trigger. During a threshold scan, fifty charges (from 0 to 500 electrons) are injected fifty times each. The response of the detector is processed synchronously in the ALICE computing farm, where the threshold and noise levels are determined from the s-curve measured for each pixel. If tuning of the threshold is required, this procedure is repeated for a representative subset of pixels at each step during a scan of the values of the relevant on-chip biasing DACs. Threshold values ranging from approximately 60 to 170 electrons guarantee a detection efficiency of over 99% and a sufficiently low rate of fake hits, as shown in [ 2 ] . The chosen working point is 100 \mathrm{e}^{-} . The stability of the threshold is monitored daily. After tuning, the average chip value is uniform across the entire detector at the percent level. The stability over time is illustrated in Fig. 2 . Small fluctuations are primarily induced by variations in the chip supply voltages. A modest shift effect is also observed in the IB and attributed to the accumulated radiation dose. Those effects are consistently recoverable with a new tuning. Figure 2: Evolution of the average in-pixel discriminating threshold per layer, from March 2023 to October 2024. Periods without beams in the LHC are marked in gray. After tuning the thresholds, a noise calibration is performed to identify the noisy pixels and mask them. Pixels firing more than 10^{-6} times per event in the OB or 10^{-2} times in the IB during readout without beams are classified as noisy. The higher threshold in the inner layers is set to prioritize the detection efficiency and remains affordable in terms of both bandwidth and event reconstruction. As a result of this strategy, the fraction of masked pixels is lower than 10^{-5} in the IB and 10^{-4} in the OB. The rate of fake hits during data taking is kept at the level of 10^{-7} hits/pix/evt in the IB and 10^{-8} hits/pix/evt in the

Read the complete text