Vertex and Tracking System

From ATHENA collaboration
Jump to navigation Jump to search

← Back to the Supplemental Material Home Page

Tracking Acceptance

The ATHENA tracking design leverages the impressive knowledge and developments carried out within the R&D programs (eRD3, eRD6, eRD16, eRD18, and eRD25) resulting in a Silicon and gaseous hybrid tracker. The two technologies are implemented in a complementary way which makes full use of the space available for an EIC detector, which allows for a tracking acceptance of -3.8 < η < 3.75. The locations in z of the tracking materials can be seen in Fig. 1. At the IP location there are three layers of silicon MAPS which have a length of 28 cm to provide vertexing (blue lines in Fig. 1), followed by two layers of silicon MAPS at radii of 13 and 18 cm to measure the track's sagitta, which extend about 34 and 47 cm in length, respectively (orange lines in Fig.1). Continuing to move towards larger radii, there are two closely spaced micromegas layers at radii of about 50 and 75 cm, with lengths of about 130 and 200 cm (purple lines in Fig. 1). In the backward direction there are five silicon MAPS disks (red lines in Fig.1). The two disks closest to the IP are at z positions of about -25 and -49 cm, with outer radii of about 19 and 37 cm, respectively. The remaining silicon disks are identical in size with an outer radius of about 43 cm and located at z of about -73, -109, and -145 cm. The inner radii of all the silicon disks matches the outer radii of the beam pipe. Supplementing the silicon disks are two large area annular triple-GEM detectors (green lines in Fig.1). The triple-GEM detectors are located at z of -103 and -142 cm with an inner radius of about 45 cm that extends to roughly 77 cm. The forward direction has a similar configuration with silicon MAPS disks complimented with annular triple-GEM detectors. Although since overall detector performance and physics in the proton direction is not as sensitive to material as the electron direction, there is an additional silicon MAPS disk. The inner two (closest to the IP) silicon disks are positioned at z = 25 and 49 cm, with radii of about 19 and 37 cm. The remaining four silicon disks have the same radii (~43 cm) and are located roughly in z at 73, 104, 134, 165 cm. The annular triple-GEM detectors are geometrically identical to those in the backward direction with inner radii of about 45 cm and extending to roughly 77 cm. These two triple-GEM detectors are located at z of about 106 and 162. The final tracking detector is a μRWELL layer which is located behind the forward dRICH (z ~332 cm) with an inner and outer radius of about 19 and 200 cm, respectively (magenta line in Fig.1).

The pseudorapidity coverage of the tracking system and the complementary overlaps can be seen in Fig.2, which shows the tracking material plotted against η. The colors of the various sub detectors are the same as those described in Fig.1.

Mat vs z.pdf


Figure 1: Tracking material surfaces as a function of z.

ATHENA eta coverage CanyonLandv1.2 colorKey.pdf


Figure 2: Tracking material surfaces as a function of pseduorapidity.

Simulation Validation

The tracking performance of the ATHENA detector was assessed in full simulations in which geometries were defined with DD4HEP and track reconstruction was done using ACTS. To validate these simulation results, the ATHENA tracking configuration was implemented into the Genfit based Fun4All full simulation framework, as well as a fast simulation framework. The full simulations (ACTS and Genfit) used a realistic magnet ATHENA field map, while the fast simulation used an ideal 3T solenoidal field. Figures 3 and 4 show comparisons of the relative momentum resolution as a function of momentum obtained from the three simulation frameworks divided into bins of pseduorapidity. Good overall agreement is found amongst the three frameworks, and in particular the two full simulations.

Figure 5 shows a comparison between the transverse DCA resolutions as a function of transverse momentum obtained from ACTS and fast simulation. Again, overall agreement between the two frameworks can be seen. Note, the equivalent Fun4All results were not available at the time to make a comparison.

Hybrid mat dpp backward.pdf


Figure 3: Pion momentum resolution vs. momentum comparisons between ATHENA detector full simulations using ACTS (red markers), Fun4All (blue line), and Fast Simulation (dashed line) frameworks, in backward pseudorapidity.

Hybrid mat dpp forward.pdf


Figure 4: Pion momentum resolution vs. momentum comparisons between ATHENA detector full simulations using ACTS (red markers), Fun4All (blue line), and Fast Simulation (dashed line) frameworks, in forward pseudorapidity.

DCA2D Validation Canyonlandsv1.2.pdf


Figure 5: Pion transverse DCA resolution vs. transverse momentum comparisons between ATHENA detector full simulations using ACTS (red markers) and Fast Simulation (dashed line) frameworks for selected pseudorapidities.

Tracking Performance

Momentum Resolution

The relative momentum resolution (dp/p) obtained from full simulations using ACTS were parameterized by fitting the simulation results as a function of momentum with the functional form dp/p(%) = sqrt{(A*p)2 + B2}$. This is the same functional form used in the Yellow Report to produce the physics working group (PWG) requirement curves. The fits were performed in bins Δ η = 0.5. Figure 6 covers the η range -3.5 < η < 3.5, while Fig. 7 extends the η coverage to |η| < 3.7 by adding four additional bins.

P resol.pdf


Figure 6: Pion momentum resolution vs. momentum ATHENA tracking performance evaluated with ACTS for different η bins. Green line shows the fit result used to parameterize the performance.

Large eta Canyonlands v1.2.pdf


Figure 7: Pion momentum resolution vs. momentum ATHENA tracking performance evaluated with ATCS for the largest |η| bins. Green line shows the fit result used to parameterize the performance.

Transverse Momentum Resolution

The transverse momentum resolution (dpT/pT) obtained from full simulations using ACTS were parameterized by fitting the simulation results as a function of pT with the functional form dpT/pT (%) = sqrt{(A*pT)2 + B2}. This is the same functional form used in the Yellow Report to produce the PWG requirement curves. The fits were performed in bins Δ η = 0.5. Figure 8 covers the η range |η| < 3.5.

Pt resol.pdf


Figure 8: Pion transverse momentum resolution vs. transverse momentum ATHENA tracking performance evaluated with ACTS for different η bins. Green line shows the fit result used to parameterize the performance.

DCA Resolution

The transverse pointing resolution (σDCAT) was obtained from full simulations using ACTS were parameterized by fitting the simulation results as a function of σDCAT (μm) = sqrt{(A/pT)2 + B2}$, the same functional forms used to produce the Yellow Report PWG requirement curves. The fits were performed in bins Δ η = 0.5. Figure 9 shows the results for the backward region, -3.5 < η < -1, Fig. 10 the central region, -1.0 < η < 1.0, and Fig. 11 the forward region, 1.0 < η < 3.5.

CanyonLandv1.2 TransDCA Backward.pdf


Figure 9: Pion Transverse DCA resolution vs. transverse momentum ATHENA tracking performance evaluated with ACTS for different backward pseudorapidity regions. Red line shows the fit result used to parameterize the performance.

CanyonLandv1.2 TransDCA Central.pdf


Figure 10: Pion Transverse DCA resolution vs. transverse momentum ATHENA tracking performance evaluated with ACTS for different central pseudorapidity regions. Red line shows the fit result used to parameterize the performance.

CanyonLandv1.2 TransDCA Forward.pdf


Figure 11: Pion Transverse DCA resolution vs. transverse momentum ATHENA tracking performance evaluated with ACTS for different forward pseudorapidity regions. Red line shows the fit result used to parameterize the performance.

PWG Requirement Comparisons

The relative momentum and transverse DCA resolutions from the ACTS based ATHENA full simulation are binned in η bins that match those of the PWG requirements allowing for a direct comparison of the performances. Figure 12 shows the relative momentum as a function of momentum (top row) and transverse DCA resolutions as a function of transverse momentum (bottom row) for three pseudorapidity bins. The ACTS simulation results are shown by the green markers, their corresponding fit the yellow curve, and the pwg requirements are represented by the dashed blue lines. Figure 12 shows the results from the two missing intermediate pseudorapidity bins.

ATHENA achieves the required momentum resolution set by PWG requirements in all η bins except the most backward region. As discussed in the proposal achieving the required momentum resolution is not possible. In the backward and forward regions, the transverse DCA resolution under performs relative to the PWG requirements at pT below about 2 GeV/c and exceeds the requirements above 2 GeV/c. In the central region, the transverse DCA resolution is comparable to pwg requirement, except in the range of about 1 GeV/c < pT < 5 GeV/c, where it slightly misses requirement. Figure 14 and 15 shows the ATHENA momentum and transverse DCA performance in narrower η bins than the PWG requirements were binned in.

Athena tracking performance 11-24-21.pdf


Figure 12: Relative momentum (top) and transverse DCA (bottom) resolutions, obtained from full simulations using ACTS, as a function of momentum and transverse momentum, respectively (in proposal).

Athena tracking performance inter eta 11-23-21.pdf


Figure 13: Relative momentum (top) and transverse DCA (bottom) resolutions, obtained from full simulations using ACTS, as a function of momentum and transverse momentum, respectively.

Athena tracking performance fine eta 11-29-21.pdf


Figure 14: Relative momentum resolutions, obtained from full simulation using ACTS, as a function of momentum.

Athena tracking dca performance fine eta 11-30-21.pdf


Figure 15: Transverse DCA resolutions obtained from full simulation using ACTS, as a function of transverse momentum.

Magnetic Field Study

One way to improve the momentum resolution performance of ATHENA in the backward pseudorapidity range -3.5 < η < -2.5 is to increase the magnetic field strength. To determine the field strength required to meet the backward requirements listed in the Yellow Report, the ATHENA tracker was implemented into the Fun4All full simulation framework, along with the ATHENA field map. The momentum resolution as a function of momentum was calculated in the range -3.5 < η < -2.5 using several different field strengths (1T, 2T, 3T, 5T, and 10T). The 3T field is the official ATHENA field map, with field strength of 3T. The field strength was then scaled by a constant factor to effectively produce a desired field strength. The results of this study are shown in Fig.16, which shows that the Yellow Report requirements (black curve) can only be met with a magnetic field strength of order 10T.

FieldStudy Zoom.pdf


Figure 16: Pion momentum resolution vs. momentum for various magnetic field strengths. Results are from ATHENA tracking implementation in the Fun4All full simulation.

Silicon technology

As described in the ATHENA proposal the technology chosen for the silicon vertex and tracking detector is a MAPS sensor in 65 nm co-developed with the ALICE ITS3 collaboration.

This choice was motivated by the work carried out by the eRD16/18/25 collaborations over more than five years that led to definition of the requirements for an EIC silicon vertex and tracking detector designed to achieve the required physics performance and the identification of the ITS3 technology as the best candidate to satisfy these requirements. This technology is now adopted by all three proto-collaborations for their detector implementation. Work to develop the ITS3 technology for use at the EIC is done by members of the three proto-collaborations within the EIC Silicon Consortium (EIC SC), that grew out of the eRD25 collaboration.

The choice of this technology is based on the following reasons:

  • The specified ALICE ITS3 65 nm MAPS sensor has performance that meets or even exceeds the EIC tracking and vertexing requirements as demonstrated in simulations presented in the Yellow Report and in the eRD25 proposal and report.

  • The development of the EIC tracking and vertexing system can leverage on a well-funded, large effort at CERN, open to non-ALICE members to contribute to the R&D to develop and use the technology for other applications. 

  • The ALICE ITS3 approach to develop an ultra-low mass vertex detector is particularly attractive for the EIC vertex layers that could be easily designed with the same technologies developed for the ITS3 detector. 

  • The schedule of the ITS3 and EIC projects are well aligned (until TDR, construction and installation are on a different time scale).

  • The ITS3 project has a 180 nm MAPS sensor fallback solution in case the main 65 nm 
development would not prove feasible on the required me scale. The decision will be taken 
later this year after results from the first submission in 65 nm (more comments on this below). 

  • A large group of EIC institutes have expressed interest in this proposed development and joined 
to form the EIC Silicon Consortium providing the required broad base of expertise for the full detector implementation.

The EIC SC has defined a strategy for the sensor development, including fall back options and associated detector infrastructure, that is described in the document at this link: https://www.eicug.org/web/sites/default/files/2021_04_26_silicon_tracking_strategy.pdf

The EIC SC has joined the sensor development from the beginning contributing to the first submission of test structures organised by CERN through a shared Multiple Layers per Reticle (MLR) run in Q1 2021. Preliminary results from all test structures presented recently show good performance. Whilst more in-depth testing is needed to characterise the technology, preliminary results are encouraging and no show stoppers have been identified. A summary of recent results can be found here: https://indico.cern.ch/event/1091910/

The EIC SC continues work on sensor development with more of its institutes joining the first engineering run where a prototype stitched wafer-scale sensor will be implemented as well as further IP blocks for data and power distribution. Whilst the initial ITS3 specifications aimed at a 10 um pixel pitch, the first sensor prototype will have a slightly larger pixel pitch around 18-22 um. Earlier eRD25 simulations carried out during the yellow report phase of the project showed that for the same material budget, a pixel pitch of 10 or 20 um has very little influence on pointing and momentum resolution, indicating that the key performance benefit comes mainly from the reduction in material thickness and not the reduction in pixel size. These results can be found in chapter 6.2 in the PhD thesis at this link https://etheses.bham.ac.uk/id/eprint/11699/. Performance simulations of the ATHENA vertex and tracking detector with pixel pitch up to 20 um are ongoing. It is also worth stressing that this first prototype submission is an exploration to prove stitching and that further optimisations will be carried out in the next submissions, so the pixel size is not yet final and could be closer to 10 um.

Gaseous technology

GEM

As part of the end cap tracking system, ATHENA makes use of annular triple-GEM detectors. Each of the annular rings consists of 6 triple-GEM modules, whose inner to outer radii span from about 45 cm to 76 cm, respectively. The triple-GEM modules are designed to minimize material and are assembled from only foil layers, consisting of the windows, drift cathode, amplification (3 GEMs), and readout elements, totaling 0.4% X/X0 of material in the active area. Such all foil designs were successfully demonstrated by the eRD3 and eRD6 projects as part of the generic EIC R&D program EIC R&D program. Two large area (1 m long) all-foil triple-GEM detectors were built and tested, one making use of a CMS type of module assembly. While the other used a gluing approach and reached an active area material budget of 0.4% X/X0 (2018 test beam results). Both detectors have participated in a second beam test at FNAL this past summer and are under analysis.

The triple-GEM detectors will make use of a 2D-strip readout structure to provide precise 2D hit points. The final readout structure is still being optimized and is one of the main goals of the eRD108 R&D effort.

μRWELL

Micro-resistive well (μRWELL) technology is not as advanced as GEM and Micromegas technology, but has been in active developed for several years with R&D showing promising results. The μRWELL is simpler in design than the triple-GEM detector, as its amplification elements consists of one GEM-like foil on top of a thin resistive layer, which eliminates the need to stretch amplification elements when assembling the detector. The ATHENA tracking system envisions having a μRWELL detector located behind the forward dRICH. This tracker would be located about 332 cm from the IP and have a radial coverage of about 2 m. With the μRWELL tracker being outside of the other tracking layers and located behind the dRICH, the material budget requirement of the detector does not need to be as stringent (we expect around ~1-2% X/X0) as that other tracking layers, allowing one to more easily gain better detector rigidity.

Similar to the GEM, the μRWELL will use a 2D-strip readout structure, which is still being optimized as part of the R&D for eRD108.

Micromegas

The outer barrel tracker of ATHENA is based on the Micromegas technology developed for the CLAS12 Micromegas Vertex Tracker. The current design foresees Micromegas modules of the same size, about 50x70cm2, that will be curved at different radii. This choice will simplify the design and the production line, therefore lowering the costs and time of the final detector. The main difference with respect to CLAS12 is that the Micromegas will be equipped with a 2D-strip readout structure. The design consists of four layers cylindrical layers, two inner ones at a radial distance of about 48 cm and two outer ones at about 76 cm. The length of the layers is such that it covers the pseudo-rapidity interval |η| < 1.05. Each layer will be composed of several Micromegas modules (tiles): the inner layers will be composed by 6 (φ) x 2 (z) modules, while the outer layers will consists of 10 (φ) x 3 (z) modules. In order to provide full coverage, i.e. no acceptance gaps, both in polar angle and in z, the tiles will overlap. The front-end electronics will be placed on the support structures at η = +- 1.1 in order to minimize the material budget in the active region. For the outer layers, the inner modules will be connected to the FEE with (max) 1m long flex cables.

eRD108 Effort

The eRD108 program looks to continue the MPGD related R&D that is needed for the EIC. Presented below are R&D objectives from eRD108 that pertain to the MPGDs implemented in ATHENA.

GEM

  • Reduce support material: With the active area of the triple-GEM modules reduced to 0.4% X/X0. When the 6 triple-GEM modules are placed together to form the tracking ring, the support material of the GEM module sides occupy part of the tracking acceptance. One objective of eRD108 is try and minimize this support material, which is not trivial since it is supporting the triple-GEM stack.

μRWELL

  • Large-area μRWELL module: Design, build, and fully characterize on a test bench (cosmics, x-rays) and in a large-area μRWELL detector.
  • The triple-GEM modules could potentially be replaced by μRWELL that uses the same foil windows, drift cathode, and readout structure as the triple-GEM. This would not only reduce the complexity of the detector assembly, but also has the potential to reduce further the material budget compared to the triple-GEM detector. This is planned to be tested within eRD108, by designing and building the large-area μRWELL module with a foil based drift-cathode and readout.

Micromegas

The curved Micromegas technology is already mature. The main focus of the R&D will be the optimization of the 2D readout structures.

Although the current Micromegas technology meets already the requirements in term of material budget (0.4% X/X0 in the active region), an ongoing R&D program is focusing on lowering the mass of each module below 0.1% X/X0.

Readout (Common)

A common goal of the readout structure used with the different MPGD technologies (GEM, μRWELL, Micromegas) is to have a readout that has a low channel count and still provides excellent spacial resolution.

  • 2D Zig-Zag readout structure: Testing a 2D zig-zag strip readout within eRD108 follows naturally, building on the work done within eRD6 related to 1D zig-zag strip readout structure, which was implemented on both small and large GEM prototypes. The interwoven zig-zag structure of the electrodes enhances the charge sharing among neighboring strips compared to traditional straight strips, resulting in excellent position resolution with less channel counts.
  • Capacitive sharing: Another area of R&D covered by eRD6 that is showing interesting results and can reduce the number of channel counts is capacitive-sharing readout structures. These can be either pad or strip readout and are applicable to the three MPGD ATHENA technologies and any 2D (or 1D) readout structure (U-V, zig-zag,...). Replacing the standard readout with one that consists of a vertical stack of strip (or pad) layers, the charge produced by the MPGD amplification element is transferred from layer to layer via capacitive coupling. The pitch of the strips in each layer becomes larger as you move away from the MPGD layer. The capacitive-sharing scheme preserves the spacial resolution, which is set by the pitch of the strips on the layer closest to the MPGD, allowing one to readout the larger pitched strips but with the resolution defined by the finer pitched strips. A capacitive sharing readout can also be used to embed a 3 coordinate structure (X-Y-U, U-V-X, etc.). This idea is being investigated to address possible issues that could arise with trying to associate the proper detector hit point to the track if there are multiple hits in the detector.

← Back to the Supplemental Material Home Page