Preprint typeset in JINST style - HYPER VERSION
CERN-PH-EP-2011-147
Submitted to JINST
A study of the material in the ATLAS inner detector using
secondary hadronic interactions
The ATLAS Collaboration
ABSTRACT: The ATLAS inner detector is used to reconstruct secondary vertices due to hadronic
interactions of primary collision products, so probing the location and amount of material in the
inner region of ATLAS. Data collected in 7 TeV pp collisions at the LHC, with a minimum bias
trigger, are used for comparisons with simulated events. The reconstructed secondary vertices
have spatial resolutions ranging from ∼ 200 µm to 1 mm. The overall material description in the
simulation is validated to within an experimental uncertainty of about 7%. This will lead to a better
understanding of the reconstruction of various objects such as tracks, leptons, jets, and missing
transverse momentum.
KEYWORDS: Material measurement, Hadronic Interactions, Secondary Vertexing.
arXiv:1110.6191v2 [hep-ex] 15 Dec 2011
Contents
1. Introduction 2
2. Inner Detector 2
3. Data samples, track selection and reconstruction 4
3.1 Data samples 4
3.2 Track selection 4
3.3 Track reconstruction in data and MC 5
4. Description of vertex reconstruction and resolution 5
4.1 Vertex reconstruction 5
4.2 Vertex resolutions 6
5. Reconstructed vertices in 7 TeV data 7
5.1 Qualitative comparison of data and MC 9
5.2 Position of beam pipe and pixel detector 11
5.3 Details of modules in the pixel detector 12
6. Systematic uncertainties 13
6.1 Tracking efficiency 13
6.2 Selection criteria during vertex finding 15
6.3 Other sources 15
6.4 Total systematic uncertainty 16
7. Numerical comparison of data and MC 16
7.1 Vertex yields 16
7.2 Details of interactions in beryllium part of beam pipe 17
7.3 Comparison of vertex yields in data and MC 18
7.4 Vertex yields within the modules of the pixel detector 18
8. Conclusions 19
9. Acknowledgements 20
– 1 –
1. Introduction
An accurate description of material in the ATLAS inner detector is crucial to the understanding
of tracking performance, as well as other reconstructed objects such as electrons, jets and miss-
ing transverse momentum. Traditionally, photon conversions, which are sensitive to the radiation
length of material, are used to map the detector. To quantify the amount of material in terms of
interaction lengths, the measurement must be converted from radiation lengths, which requires a
very precise knowledge of the actual composition of the material, or a direct measurement of quan-
tities sensitive to the interaction length must be made. This paper describes a direct measurement
using the reconstruction of secondary vertices due to hadronic interactions of primary particles,
and is based on a careful comparison of the secondary vertex yield in data with a simulation of the
ATLAS inner detector. The simulation implements precise information about the inner detector
components, and this study aims to validate its correctness.
In addition to directly probing the number of hadronic interaction lengths of material, another
advantage of studying such interactions is the excellent spatial resolution of the resulting recon-
structed secondary vertices. This property is exploited to study the precise location of the material.
Since hadronic interactions vertices usually result from low to medium energy primary hadrons
(with average momentum, <p> around 4 GeV, and with about 96% having p < 10 GeV) the out-
going particles have low energy and large opening angles between them. This contrasts with photon
conversions, where the opening angle between the outgoing electron-positron pair is close to zero.
Consequently, the technique presented here has a much improved spatial resolution. Hadronic in-
teractions will often produce more than two outgoing particles with momenta high enough to be
reconstructed by the tracking system. An inclusive vertex finding and fitting package is used to
reconstruct these vertices.
This paper is structured as follows: Section
2 gives a brief description of the inner detector,
Section 3 gives details of the data sample and track selection criteria used in this analysis, and
Section 4 contains a description of the vertex-finding algorithm. Section 5 contains qualitative
results from data and comparisons with Monte Carlo simulations (MC). Section 6 describes the
various systematic uncertainties, which are used in Section 7 to make quantitative comparisons
with MC.
2. Inner Detector
The inner detector consists of a semi-conductor pixel detector, a semi-conductor microstrip detector
(SCT), and a transition radiation tracker (TRT), all of which are surrounded by a solenoid magnet
providing a 2 T field [
1, 2]. It extends from a radius
1
of about 45 mm to 1100 mm and out to |z| of
about 3100 mm. A quarter section of the inner detector is shown in Figure 1. It provides excellent
track impact parameter and momentum resolution over a large pseudorapidity range (|η| < 2.5),
and determines the positions of primary and secondary vertices.
1
ATLAS uses a right-handed coordinate system with its origin at the nominal interaction point (IP) in the center of
the detector and the z-axis along the beam pipe. The x-axis points from the IP to the center of the LHC ring, and the y
axis points upward. Cylindrical coordinates (R,φ ) are used in the transverse plane, φ being the azimuthal angle around
the beam pipe. The pseudorapidity is defined in terms of the polar angle θ as η = −lntan(θ /2).
– 2 –
Envelopes
Pixel
SCT barrel
SCT end-cap
TRT barrel
TRT end-cap
255<R<549mm
|Z|<805mm
251<R<610mm
810<|Z|<2797mm
554<R<1082mm
|Z|<780mm
617<R<1106mm
827<|Z|<2744mm
45.5<R<242mm
|Z|<3092mm
Cryostat
PPF1
Cryostat
Solenoid coil
z(mm)
Beam-pipe
Pixel
support tube
SCT (end-cap)
TRT(end-cap)
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8
Pixel
400.5
495
580
650
749
853.8
934
1091.5
1299.9
1399.7
1771.4
2115.2
2505
2720.2
0
0
R50.5
R88.5
R122.5
R299
R371
R443
R514
R563
R1066
R1150
R229
R560
R438.8
R408
R337.6
R275
R644
R1004
2710
848
712
PPB1
Radius(mm)
TRT(barrel)
SCT(barrel)
Pixel PP1
3512
ID end-plate
Pixel
400.5
495
580
650
0
0
R50.5
R88.5
R122.5
R88.8
R149.6
R34.3
Figure 1. Plan view of a quadrant of the inner detector showing each of the major detector elements with
their active dimensions and envelopes. The lower part shows a zoom of the pixel region.
In the barrel region, the precision detectors (pixel and SCT) are arranged in cylindrical layers
around the beam pipe, and in the endcaps they are assembled as disks and placed perpendicular
to the beam axis. The TRT is made of drift tubes, which are parallel to the beam axis in the
barrel region, and extend radially outward in the endcap region. The envelope of the barrel pixel
detector covers the radial region from 45 mm to 242 mm, which includes the active layers, as well
as supports, and extends to about ±400 mm in z. The barrel SCT envelope ranges from 255 mm
to 549 mm, and the barrel TRT sub-system covers the radial range from 554 to 1082 mm. The
latter two sub-systems extend to about ±800 mm in z. Outside the beam pipe, the regions without
material are filled with different gases, N
2
and CO
2
in the silicon and TRT volumes, respectively,
and for simplicity they are referred to as air gaps.
All pixel sensors in the pixel detector, in both barrel and end-cap regions, are identical and
have a nominal size of 50 ×400 µm
2
, and there are approximately 80.4 million readout channels.
The pixel detector in the barrel region has three layers, containing 22, 38, and 52 staves in azimuth,
respectively. The layers are concentric with the beam pipe. Each stave contains 13 modules along
z, and each module contains about 47000 individual pixels. A ‘zoomed-in’ view of a module can
be seen in Figure 4.4 of Ref. [
2]. In the SCT barrel region there are small angle stereo strips in each
layer, with one set parallel to the beam direction to measure R −φ and the other set at an angle of
40 mrad, which allows for a measurement of the z-coordinate. The endcap SCT detector has a set
– 3 –
of strips running radially outward and a set of stereo strips at an angle of 40 mrad to the former.
The total number of readout channels in the SCT is approximately 6.3 million. The TRT consists
of 298,000 drift tubes with diameter 4 mm, and provides coverage over |η| < 2.0. The material
measurements in this paper are focused mainly on the beam pipe and the pixel detector in the barrel
region.
3. Data samples, track selection and reconstruction
3.1 Data samples
The data used in this analysis were collected during March-June 2010 in proton-proton collisions
at a center-of-mass energy of 7 TeV. During this initial period the instantaneous luminosity was
approximately 10
27
−10
29
cm
−2
s
−1
. Data were collected using minimum bias triggers [3] and
correspond to approximately 19 nb
−1
of integrated luminosity. Later runs with higher instantaneous
luminosity were not used. The minimum bias triggers collect single-, double- and non-diffractive
events, with the majority belonging to the last category. In order to facilitate comparisons with
MC, single- and double-diffractive contributions are effectively removed from the data and the
remaining events are compared with a simulated sample of non-diffractive events [
3]. This is
achieved by requiring a large track multiplicity at the primary vertex. This approach works best
when the number of additional pp interactions per event (pile-up) is small, and so only the low
luminosity runs are used. It is required that there be exactly one reconstructed primary vertex in
the event, and that it should have at least 11 associated tracks; this requirement is expected to keep
less than 1% of single- and double-diffractive events, while retaining ∼ 68% of non-diffractive
events. At this stage there are ∼ 40.9 (13.5) million events in data (MC), respectively. MC events
are weighted such that the mean and width of the z-coordinate distribution of the primary vertex
position match the data. MC events were generated using PYTHIA6 [
4] with the AMBT1 tune [5],
simulated with GEANT4 [6], and processed with the same reconstruction software as data. The
ATLAS simulation infrastructure is described elsewhere [7].
3.2 Track selection
Since the main goal of the track reconstruction software is to find particles originating from the
primary vertex, it puts stringent limits on the allowed values of transverse and longitudinal im-
pact parameters. As a result, the reconstruction efficiency for secondary track candidates strongly
depends on both R- and z-coordinates of the vertex they originate from.
In order to reconstruct secondary interactions, well-measured secondary track candidates should
be selected, and tracks coming from the primary vertex rejected in order to reduce combinatorial
background. Tracks are required to have: (a) transverse momentum above 0.3 GeV, (b) transverse
impact parameter relative to the primary vertex of at least 5 mm, and (c) fit χ
2
/dof < 5. The im-
pact parameter requirement removes more than 99% of the primary tracks, as well as many tracks
produced in K
0
S
decays and γ conversions. In general, particles produced in secondary hadronic
interactions have much larger impact parameters, especially in comparison to γ conversions, which
tend to point back to the primary vertex. There is no requirement on the number of hits in the pixel
detector, since that would limit the scope of this analysis to a radius less than that of the third layer.
– 4 –