Status of the Cylindical-GEM project for the KLOE-2 Inner Tracker

Abstract

The status of the R&D on the Cylindrical-GEM (CGEM) detector foreseen as Inner Tracker for KLOE-2, the upgrade of the KLOE experiment at the DAFNE phi-factory, will be presented. The R&D includes several activities: i) the construction and complete characterization of the full-size CGEM prototype, equipped with 650 microns pitch 1-D longitudinal strips; ii) the study of the 2-D readout with XV patterned strips and operation in magnetic field (up to 1.5T), performed with small planar prototypes in a dedicated test at the H4-SPS beam facility; iii) the characterization of the single-mask GEM technology for the realization of large-area GEM foils. Comment: 4 pages, 10 figures, Presented at Vienna Conference on Instrumentation (Feb 15-20, 2010, Vienna, Austria). Submitted to the Proceedings

Full text

Page 1

arXiv:1003.3770v1 [physics.ins-det] 19 Mar 2010
Status of the Cylindical-GEM project for the KLOE-2 Inner Tracker
A. Ballac, G. Bencivennic, S. Cerionic, P. Ciambronec, E. De Lucia∗,c, G. De Robertisa, D. Domenicic, G. Felicic, M. Gattac,
M. Jacewiczc, N. Lacalamitaa, S. Laucianic, R. Liuzzia, F. Loddoa, M. Mongellia, G. Morellob, A. Pelosid, M. Pistillic,
L. Quintieric, A. Ranieria, V. Valentinoa
aINFN Sezione di Bari, Bari, Italy
bINFN gruppo collegato di Cosenza, Cosenza, Italy
cLaboratori Nazionali di Frascati dell’INFN, Frascati, Italy
dDipartimento di Fisica, ”Sapienza” Universit` a di Roma and INFN Sezione di Roma, Roma, Italy
Abstract
The status of the R&D on the Cylindrical-GEM (CGEM) detector foreseen as Inner Tracker for KLOE-2, the upgrade of the
KLOE experiment at the DAΦNE φ–factory , will be presented. The R&D includes several activities: i) the construction and
complete characterization of the full-size CGEM prototype, equipped with 650 µm pitch 1-D longitudinal strips; ii) the study of
the 2-D readout with XV patterned strips and operation in magnetic field (up to 1.5T), performed with small planar prototypes in
a dedicated test at the H4-SPS beam facility; iii) the characterization of the single-mask GEM technology for the realization of
large-area GEM foils.
Key words: GEM, tracking, KLOE-2
1. Introduction
After the completion of the KLOE data taking [1], the pro-
posal of a new run with an upgraded KLOE detector, KLOE-
2 [2], at an upgraded DAΦNE machine has been accepted and
will start in spring2010[3]. The KLOE-2 physicsprogramwill
be focused on physics coming from the interaction point (IP),
where the φ-meson is produced: KS , K±, η and η
well as KSKLinterference and search for physics beyond the
Standard Model. The improvement in the reconstruction per-
formance for tracks near the interaction region is then of funda-
mental importance for the accomplishment of this physics pro-
gram.
Afterafirst phasewiththeinstallationofthelow-energye+e−
(LET) [4] and high-energy e+e−(HET) [5] tagging systems for
the identification and study of γ − γ events, the detector will
be upgraded with the insertion of an Inner Tracker (IT) be-
tween the beam pipe and the Drift Chamber (DC) inner wall.
A crucial design parameter is the resolution on the KS decay
point, occurring within few cm from the IP. An accurate study
on quantum interferometry measurement showed that an im-
provement on this resolution of about a factor of 3 with respect
to the present value (≃ 6 mm) is required [6]. The IT contri-
bution to the overall material budget has to be carefully taken
into account in order to minimize multiple scattering contribu-
tion to the track momentum resolution and photon conversions
before the DC volume. The main requirements for the IT can
be summarized as:
i. σrφ∼ 200 µm and σz∼ 500 µm spatial resolutions;
′decays as
∗Corresponding author
Email address: erika.delucia@lnf.infn.it (E. De Lucia)
ii. total material budget below 2% of the radiation length X0;
iii. 5 kHz/cm2rate capability.
The adopted solution is the Cylindrical-GEM (CGEM) [6], a
triple-GEM detector composed by concentric cylindrical elec-
trodes (fig. 1): Cathode, 3 GEM foils for the multiplication
stage (gain ∼104) and Anode, acting also as the readout circuit.
The high rate capability of the GEM (up to 1 MHz/mm2[7])
makes this detectors suitable to be placed near the interaction
point of a high-luminosity collider machine. The IT will be
Figure 1: Cross-section of the triple-GEM detector.
then composed by 5 concentric CGEM detection layers at radii
from 13 cm, to preserve the KS-KLquantum interference re-
gion, to 23 cm due to the constrained from DC inner wall at 25
cm. The total active length for all layers will be 70 cm. The
anode readout of each CGEM will be segmented with 650 µm
pitchXVpatternedstripswithastereoangleof∼ 40◦, foratotal
of about 30,000 FEE channels. The Front-End Electronics for
the IT is based on the new GASTONE ASIC [8], a 64 channels
Preprint submitted to Nuclear Instruments and Methods A March 22, 2010

Page 2

chip composed by four different stages: a charge preamplifier
with 20mV/fC sensitivity, a shaper, a leading-edge discrimina-
tor with a programmable threshold and a monostable stretcher
of the digital signal, to synchronizewith the KLOE Level1 trig-
ger.
In the following the main stages of the R&D project [6] will
be discussed: i) the construction and complete characterization
of a full-scale CGEM prototype, ii) the study of the XV strip
readout configuration and its operation in magnetic field and
iii) the construction and characterization of a large area GEM
realized with the new single-mask photolitografic technique.
2. Cylindrical-GEM prototype
A full-scale CGEM prototype has been built with 15 cm ra-
dius and 35 cm active length [9]. The cylindrical electrodes
havebeenobtainedfromverylightpolyimidefoils(50µm thick
for GEMs, 100 µm for Cathode, 100 µm for Anode) rolled onto
machined PTFE cylinders, acting as molds, and then glued ex-
ploiting the vacuum bag technique. The cylinders are then in-
sertedoneintotheotherandfiberglassannularflangesareglued
at the ends, acting as spacers for the gaps and as supportingme-
chanics [10]. The CGEM is therefore a low-mass, fully cylin-
drical and dead-zone free GEM based detector with no support
frames required inside its active area. The anode was patterned
with 650 µm pitch longitudinal strips, reconstructing the r-φ
coordinate, for a total of 1538 axial strips. The chamber, tested
with X-rays to check the uniformity over the large surface, has
been then equipped with the 16-channels prototype of GAS-
TONE[8]andtested attheCERN PS witha10GeVpionbeam.
The results show the expected spatial resolution from a digital
readout of 650 µm pitch strips [11] and a 99.6% efficiency.
3. XV readout and operation in magnetic field
A typical orthogonal XY readout can not be used for the in-
ner tracker,due to its cylindricalgeometry. The final IT readout
will bethenperformedwithanXVpatternofstripsandpadsen-
graved on a polyimide foil substrate, 100 µm thick (fig. 2). The
Figure 2: Scheme of the XV readout configuration.
X strips with 650µm pitchwill providether-φ coordinatewhile
the pads, connected through internal vias to form V strips with
650µm pitch, will provide the z coordinate. This quite innova-
tive readoutsolutionwas not implementedon the CGEM proto-
type, therefore its characteristics have been extensively studied
with dedicated planar chambers. In addition, since the IT will
operate inside KLOE’s magnetic field, the effects on the cluster
formation and electronics readout had to be studied.
To address these issues a dedicated test has been done at the
H4 permanent facility, setup at the CERN-SPS 150GeV pion
beam line within the RD51 Collaboration[12]. Five 10x10 cm2
planar triple-GEM (PGEM) detectors with 650 µm pitch read-
out have been assembled and succesfully tested: four chambers
with standard XY readout and the fifth with the XV readout un-
der investigation. The setup was 1 meter long with detectors
placed equidistantly with the XV chamber placed in the cen-
ter (fig. 3). For the operation in magnetic field, the GOLIATH
magnet was used, providing a field adjustable up to 1.5 T per-
pendicularto the horizontalbeam-plane. To fully coverthe area
Figure 3: Setup of the test beam at CERN with planar chambers with 2D-
readout.
KLOE-2 inner tracker technical design report
illuminated by the SPS beam, the planar chambers were par-
tially equipped with 22 digital readout GASTONE boards, 32
channels each, four on each XY chamber and six on the XV
chamber. The coincidence of 6 scintillators (3x3 cm2) read-
out by silicon photomultipliers,three upstream and three down-
stream, provided the trigger signal for the acquisition. The
same working point of the CGEM prototype has been used:
Ar/CO2 (70/30) gas mixture and operating voltages Vfields=
1.5/3/3/5 kV/cm and VGEM= 390/380/370 V (?VG= 1140V,
Gain=2×104). The GASTONE threshold was set at 3.5 fC.
The effect of the magnetic field (B) is twofold: a displace-
ment dx and a spread σdx of the charge over the readout
plane. The expected values obtained from simulation studies
of our chambers done with GARFIELD are dx=600 µm and
σdx=200 µm at B=0.5 T [6]. In the test beam configuration the
magnetic field effect was mainly present on the X-view. The
setup usedto measurethe displacementonthe XV chamberdue
to the magnetic field is shown in fig. 4. All four XY chambers
are likewise oriented, with the same anode-cathode configura-
tion, andprovidethe externaltrackingsystem forthe XV cham-
ber which is instead in a cathode-anode arrangement, reversed
2

Page 3

Figure 4: Test beam setup and definition of the measured quantity.
with respect to the other chambers. Since the XY chambers
are subject to the same Lorentz force, the reconstructed track
will be shifted by the same offset dx with respect to the true
track trajectory. The displacement in the XV chamber instead
will be of the same magnitude dx but with opposite direction,
dueto the reversedcathode-anodearrangement. First, with zero
magnetic field (B=0 T), the setup was aligned to a few microm-
eter precision and then, with the magnetic field turned on, the
total displacement (D) between the track reconstructed by the
XY telescope and the point in the XV chamber was measured:
D = 2×dx (fig. 4). The displacement dx was measured for 5
values of the magnetic field and found in good agreement with
the value obtained from the GARFIELD simulation at B=0.5 T
(fig. 5). To study the XV chamber performance, we measured
Figure 5: Displacement dx as a function of the magnetic field (green points)
with the result from GARFIELD simulation at B=0.5 T (blue square).
the resolution in both X and Y coordinates: the X coordinate
is measured directly from the X strips while the Y coordinate
is obtained from the crossing of both X and V strip readout.
Fig. 6 shows the resolution on the X coordinate as a function of
the magnetic field, the values ranging from 200 µm at B=0 T
up to 380 µm at B=1.35 T. The resolution on the Y coordinate
measurement is ∼370 µm at B=0 T, in agreement with what
expected from the digital readout of the two X and V views
(fig. 7).The performance of the front-end chip GASTONE
have been studied measuringthe cluster size and reconstruction
efficiency, defined as the presence of a cluster in the XV cham-
ber when a candidate track was reconstructed using four XY
chambers. Fig. 8 shows the reconstruction efficiency measured
as a function of the magnetic field and operating the XV cham-
ber with four different gain values. The efficiency for the nom-
Magnetic field (T)
0 0.20.40.60.81 1.2 1.4
Resolution X (mm)
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Resolution X vs magnetic field
Figure 6: Resolution on the X coordinate as a function of the magnetic field.
Entries Entries 19609 19609
Constant Constant 19.8 19.8
±±
2041 2041
Mean Mean 0.00269 0.00269
±±
-0.03705 -0.03705
Sigma Sigma
0.0024 0.0024
±±
0.3688 0.3688
mm
-4-2024
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
Y residuals for X-V plane B = 0T
m
µ
= 370
Y
σ
Figure 7: Resolution on the Y coordinate at B=0 T.
inal KLOE magnetic field B=0.52 T and voltage settings was
measured to exceed 99%, slightly decreasing at higher B fields.
In fact the increase of the magnetic field leads to a larger spread
of the charge over the readout strips and causes a reduction of
the charge seen by each single pre-amplifier channel, with a
consequent efficiency drop. Thus the increase of the magnetic
field requires for higher gains to efficiently operate the cham-
ber. The variation of the magnetic fields within the KLOE-2
planned values (0.3-0.5 T), to improve the acceptance for low
momentum tracks, has a negligible effect on the reconstruction
efficiency in the voltage range around our working point. The
Figure 8: Efficiency as function of the magnetic field obtained operating the
XV chamber with four different gain values.
charge sharing, grounding and cross-talk between strips could
have been in principle different for X and V views, due to the
3

Page 4

different readout geometry. Our measurements demonstrated
good behavior of both X and V readout views, which appear to
have equalized response.
4. Large Area GEM
To build the IT outermost layer, a GEM foil as large as
1440x700 mm2is needed. This foil can be obtained splicing
3 separate 480x700 mm2foils with a technique that we have
developed, using an epoxy adhesive and a vacuum bag. The
urge for larger GEM foils has driven a change of the produc-
tion procedure by CERN TS-DEM-PMT laboratory, switch-
ing to a single-mask etching, more suitable for a large sur-
face [13]. The new GEMs have quasi-cylindrical holes and a
new characterization is necessary. In order to check the uni-
formity of the new single-mask GEMs over a large area, we
will build a 700x300 mm2planar triple-GEM. Dedicated tools
for the stretching, handling and assembling of such large foils
have been designed and realized. The chamber, that will be the
largest GEM detector ever operated, will be equipped with the
GASTONE 64-channels final release and readout with the Off
Gastone Electronic (OGE) Board [6]. A dedicated test beam
of the Large Area GEM is foreseen this year at the CERN. The
externaltrackingsystem will be providedbythe fourXY cham-
bersusedfortheXVreadoutstudies, replacingtheXV chamber
with the Large Area GEM.
5. Finalizing the project: 3-D Finite Element simulation
ANSYS [14] finite element 3-D simulation of the CGEM has
been developedto estimate the structural response undertensile
loads: induced strain, stress and displacements [15]. A detailed
characterizationof the materials has been doneto implementan
accuratedescriptionofthe mechanicalbehavior. Themodelhas
been then validated by comparison with measurements done
with CGEM and PGEM prototypes. Fig. 9 shows the results
Figure 9: ANSYSsimulation of the 700 mm length CGEM: radial displacement
of the generatrix profile of a GEM foil obtained applying several tensile loads.
obtained with the simulation of the 700 mm length CGEM for
the generatrix profile of a GEM foil, with gravitational effect
only and once an axial load is applied to improve the stability
of the foils. The gravitational sag is O(µm) therefore negligible
and the maximum value of the expected radial displacement
is 20 µm for 500 N axial load and becomes 40 µm with 1000
N. The peculiar effect observed on the edges of the distribu-
tions is due to the presence on the GEM foil edges of 5 mm of
kapton connecting two more rigid structures radially displaced:
the GEM active zone (two-side copper-clad kapton foil) and
the Permaglass ring. This originates an S-shaped deformation
which translates into the observed egde effect. Besides this,
the radial displacement is constant in the GEM active zone and
similar distributionshavebeenobtainedforCathodeandAnode
showing that the structure of the CGEM gaps is preserved.
6. Conclusions
The construction, safe operation and extensive test of an al-
mostfull-sizeCylindrical-GEMprototypehasdemonstratedthe
feasibility of such a novel low-mass and dead-zone-free ver-
tex detector. The final readout configuration has been validated
with the successful test of the small planar prototypes operated
in magnetic field: very good results in terms of spatial reso-
lution, efficiency and cluster size have been obtained. A large
planar GEM prototype will be built using foils realized with
the new single-mask technique to test their quality and homo-
geneity together with the GASTONE 64-channels readout fi-
nal release. The R&D phase on Cylindrical-GEM is practically
concluded. The project of the KLOE-2 Inner Tracker has been
recently approved and its construction will start this year in or-
der to be ready for insertion in the KLOE detector by summer
2011.
References
[1] F. Bossi,E. De Lucia,J. Lee-Franzini, S. Miscetti, M. Palutan and KLOE
Collaboration, Rivista del Nuovo Cimento Vol.31, N.10 (2008).
[2] R.Beck etal, KLOE-2
terestfor thecontinuation
gram atDAΦNEupgraded
http://www.lnf.infn.it/lnfadmin/direzione/roadmap/LoIKLOE.pdf
[3] R. Beck et al, KLOE-2 collaboration, A proposal for the roll-in of the
KLOE-2 detector, LNF-07/19(IR),INFN-LNF, Frascati, 2007.
[4] D. Babusci et al., doi:10.1016/j.nima.2009.09.110
[5] F. Archilli et al., doi:10.1016/j.nima.2009.06.082
[6] F. Archilli et al., KLOE-2 Collaboration, Technical Design Report of the
Inner Tracker for the KLOE-2 experiment , arXiv:1002.2572v1 and LNF-
10/3(P) INFN-LNF, Frascati,2010.
[7] PhD thesis of Marco Poli Lener: Triple-GEM detectors for the in-
nermost region of the muon apparatus at the LHCb experiment,
http://doc.cern.ch/archive/electronic/cern/preprints/thesis/thesis-2006-013.pdf.
[8] A. Balla et al., Nucl. Inst. & Meth. A 604 (2009) 23.
[9] G. Bencivenni et al., Nucl. Inst. & Meth. A 572 (2007) 168 and Nucl. Inst.
& Meth. A 581 (2007) 221.
[10] G. Bencivenni et al., NSS Conference Record, 2007 IEEE, Volume 6 pp.
4666-4670.
[11] G. Bencivenni et al., NSS Conference Record, 2008 IEEE, 19-25 Oct.
2008 pp. 1345-1346 .
[12] RD51 Proposal, CERN-LHCC-2008-011 (2008)
[13] M. Villa et al., paper presented at this Conference.
[14] www.ANSYS.com
[15] L. Quintieri et al., Finite Element Model of the Cylindrical GEM Detector
as New Inner Tracker of Kloe2 and Mechanical Characterization of the
Employed Materials, LNF-09/12(IR), INFN-LNF, Frascati, 2009.
collaboration,
of
inluminosity
Expression
KLOE
and
of in-
the physics
in
pro-
energy,
4