A new technique for luminosity measurement using 3D pixel ...

1 A new technique for luminosity measurement using 3D

2

pixel modules in the ATLAS IBL detector

3

Peilian Liu1, on behalf of the ATLAS Collaboration

4 Physics Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94704, USA

ATL-DAPR-PROC-2018-001 01 February 2018

5 Abstract

The Insertable B-Layer (IBL) is the innermost layer of the ATLAS tracking system. It consists of planar pixel modules in the central region and 3D pixel modules at the two extremities. We use the longitudinal cluster-size distributions in 3D modules of the IBL to determine the number of pixel clusters per bunch crossing produced by primary charged particles in randomly triggered collision events, and to suppress the associated backgrounds. This Pixel-Cluster-Counting algorithm can provide both bunch-integrated and bunch-by-bunch relative-luminosity measurements, and thereby contribute independent constraints to the understanding and the evaluation of the systematic uncertainties that dominate the luminosity determination at the ATLAS experiment.

6 Keywords: Luminosity, 3D Pixel Module, Pixel Cluster

7 1. Introduction

8 An accurate measurement of the delivered luminosity is a key component 9 of the ATLAS [1] physics programme. For cross-section measurements, the 10 uncertainty in the delivered luminosity is often one of the major systematic 11 uncertainties. Searches for, and eventual discoveries of, physical phenomena 12 beyond the Standard Model also rely on accurate information about the 13 delivered luminosity to evaluate background levels and determine sensitivity 14 to the signatures of new phenomena. 15 In LHC Run 2, the primary ATLAS luminometer is LUCID [2], a 16 photomultiplier-based Cherenkov detector specifically designed to measure 17 the bunch-by-bunch luminosity in every bunch crossing. It is complemented 18 by bunch-by-bunch luminosity-sampling algorithms such as track counting,

Pre1pCrionrtressupbomnidtitnedg taoutEhlosre,vieemr ail: peilianliu@

January 30, 2018

19 as well as by several bunch-integrating algorithms that do not provide 20 information on individual colliding-bunch pairs. The measurements of these 21 luminometers are compared to assess and control the systematic uncertainties 22 of the luminosity measurements at ATLAS. The level of consistency across 23 the various methods, over the full range of luminosities and beam conditions, 24 and across many months of LHC operation, provides a direct test of the 25 accuracy and the stability of the results. New algorithms such as Pixel26 Cluster-Counting (PCC) can provide additional, independent constraints to 27 the understanding of some of the instrumental biases (such as long-term 28 drifts, or the pile-up dependence of luminosity measurements) and to the 29 evaluation of the associated systematic uncertainties.

30 2. Principle of the luminosity measurement in ATLAS

The bunch luminosity Lb produced by a single pair of colliding bunches can be expressed as

Lb

=

?fr inel

(1)

31 where the pile-up parameter ? is the average number of inelastic interactions 32 per bunch-crossing, fr is the bunch revolution frequency (11245.5 Hz at LHC), 33 and inel is the pp inelastic cross-section.

ATLAS monitors the delivered luminosity by measuring ?vis, the visible interaction rate per bunch crossing (BC). The bunch luminosity can then be written as

Lb

=

?visfr vis

(2)

34 where ?vis = ?, is the efficiency of the detector and algorithm under 35 consideration, and the visible cross-section for that same detector and 36 algorithm is defined by vis = inel. ?vis is a directly measurable quantity. 37 The visible cross-section vis is calibrated by the van der Meer (vdM ) 38 method [2] under specialized beam conditions.

39 3. Insertable B-Layer 40 The ATLAS pixel detector is the innermost detector component of the 41 ATLAS tracking system. Currently, the pixel-detector layer closest to the

2

Table 1: and indices of the 3D and planar modules from the negative-z side of IBL

to the positive-z side. The IBL is constructed of 14 staves each of which consists of 20

modules.

Structure 3D Planar 3D

index -10 -7 -6 5 6 9

index

0 13

42 beam pipe is the insertable b-layer (IBL) [3], which was installed in 2014 43 between the existing pixel detector and a new smaller radius beam-pipe at 44 a radius of 3.3 cm, to maintain the performance of the pixel detector with 45 increasing luminosity. The IBL is constructed of 14 staves laid around the 46 beam pipe. Each IBL stave is instrumented along 64 cm and consists of 47 20 modules, with four 3D modules at each end and 12 planar modules in 48 the central section. Table 1 lists the and indices of the 3D and planar 49 modules1. 50 In each 3D module, one FE-I4B chip [4] is bump bonded to one 3D sensor. 51 There are 26880 pixels arranged in 80 columns on 250 ?m pitch by 336 rows 52 on 50 ?m pitch.

53 4. Pixel-Cluster Counting Algorithm

54 The principle of the PCC-based luminosity measurement is that the 55 number of primary clusters produced by primary particles is assumed to 56 be proportional to the luminosity. The absolute PCC luminosity scale is 57 fixed by cross-calibrating it to LUCID in a reference run. 58 Only those primary clusters in the 3D modules of the IBL, that are 59 produced by primary particles from pp collisions, are counted. The number 60 of background clusters depends not only on the luminosity, but also on 61 beam conditions and on material effects. The 3D modules are located at 62 high ||. Particles from the interaction point (IP) traverse 3D modules at

1ATLAS uses a right-handed coordinate system with its origin at the nominal

interaction point in the centre of the detector, and the z-axis along the beam line. The

x-axis points from the interaction point to the centre of the LHC ring, and the y-axis

points upwards. Cylindrical coordinates (r, ) are used in the transverse plane, being

the azimuthal angle around the beam line. The pseudorapidity is defined in terms of the

polar

angle

as

=

-

ln

tan(

2

).

3

63 shallow incidence, producing long clusters, which is key to signal-background 64 separation. Figure 1 compares the experimental longitudinal cluster size 65 distributions for the four 3D modules at positive-z ( indexed 6 through 66 9). The module farthest from the interaction point ( indexed 9) sees the 67 longest primary clusters. Data-volume constraints prevent saving all the pixel 68 clusters in the entire IBL detector with reconstructed data. 3D modules 69 were selected for saving all pixel clusters (as required for PCC) because 70 they have the best signal-to-background ratio, due to their high || location. 71 Additionally, the 3D modules extend beyond the || < 2.5 acceptance of the 72 rest of the ATLAS tracker, which makes PCC independent from the track73 based luminosity measurement.

Number of Clusters

5000 4000 3000 2000

ATLAS Preliminary s= 13 TeV

IBL Module 6 IBL Module 7 IBL Module 8 IBL Module 9

1000

0 1 3 5 7 9 11 13 15 17 Longitudinal Cluster Size [pixels]

Figure 1: Comparison of longitudinal cluster size distributions for the four most forward 3D modules on the positive-z side of IBL.

74 4.1. Gaussian distribution of the longitudinal cluster size for the primary

75

clusters

The longitudinal size of primary clusters is sensitive to the particle

incidence angle, and could be calculated by

Longitudinal size = sensor thickness ? zIP - z3D

(3)

pixel pitch

rIBL

76 Here z3D is the z position of the 3D module, zIP is where the interaction occurs 77 in the z direction, and rIBL is the radius of the IBL. Since the zIP distribution 78 is approximately Gaussian, the longitudinal cluster size distribution of 79 individual 3D modules is also expected to exhibit an approximately Gaussian 80 shape.

4

81 The primary clusters produced on the module edge are shorter than 82 expected due to missing pixels. Similarly, clusters can be interrupted 83 (broken) by dead or inefficient pixels. Figure 2 illustrates the longitudinal size 84 distributions of broken and on-edge primary clusters as well as of complete 85 primary clusters. The shapes were obtained from simulated single-interaction 86 minimum-bias events. In collision data, on-edge clusters are avoided by 87 requiring cluster centers to be a minimum distance away from an edge. This 88 defines a fiducial area for each module. Broken clusters with a gap of a single 89 pixel constitute about 5.6% of all primary clusters, and they are removed

from the analysis.

Figure 2: Longitudinal size distributions of primary clusters in the most forward IBL 3D module on the negative-z side, obtained from simulated single-interaction minimum-bias events. Only clusters originating from primary particles are used.

90

91 4.2. Background clusters 92 In the products of pp collisions, aside from the primary particles produced 93 in the primary collisions, there are some secondary particles from the 94 interaction of the primary particles and photons with the detector material, 95 as well as beam backgrounds. These lead to background clusters. There 96 are also some 1-hit background clusters from hot pixels. In addition, when 97 charged particles travel through the detector, they can excite the detector 98 material. This excitation decays away radioactively, which cause small hits 99 to be seen in the detector for a brief period of time after an event. This 100 "afterglow" effect is studied in special runs in which several empty BCs

5

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download