Annihilation of low energy antiprotons

in silicon sensorsA. Gligorova, Member, IEEE, S. Aghion, O. Ahlen, A. S. Belov, G. Bonomi, P. Braunig, J. Bremer,

R. S. Brusa, G. Burghart, L. Cabaret, M. Caccia, C. Canali, R. Caravita, F. Castelli, G. Cerchiari,

S. Cialdi, D. Comparat, G. Consolati, J. H. Derking, C. Da Via, S. Di Domizio, L. Di Noto, M. Doser,

A. Dudarev, R. Ferragut, A. Fontana, P. Genova, M. Giammarchi, S. N. Gninenko, S. Haider, T. Huse,

E. Jordan, L. V. Jørgensen, T. Kaltenbacher, A. Kellerbauer, A. Knecht, D. Krasnicky, V. Lagomarsino,

S. Lehner, A. Magnani, C. Malbrunot, S. Mariazzi, V. A. Matveev, F. Moia, C. Nellist, G. Nebbia,

P. Nedelec, M. Oberthaler, N. Pacifico, V. Petracek, F. Prelz, M. Prevedelli, C. Regenfus, C. Riccardi,

O. Røhne, A. Rotondi, H. Sandaker, M. A. Subieta Vasquez, M. Spacek, G. Testera, E. Widmann,

P. Yzombard, S. Zavatarelli, J. Zmeskal

Manuscript received November 22, 2013.This work was supported by the Research Council of Norway and the

Bergen Research Foundation.A. Gligorova, is with the University of Bergen, Institute of Physics

and Technology, Allegaten 55, 5007 Bergen, Norway (e-mail: an-gela.gligorova@cern.ch).

S. Aghion and F. Moia are with Politecnico di Milano, Piazza Leonardoda Vinci 32, 20133 Milano, Italy and Istituto Nazionale di Fisica Nucleare,Sez. di Milano, Via Celoria 16, 20133 Milano, Italy.

O. Ahlen, J. Bremer, G. Burghart, J. H. Derking, M. Doser, A. Dudarev,S. Haider, L. V. Jørgensen, T. Kaltenbacher and A. Knecht are with theEuropean Organisation for Nuclear Research, Physics Dept., 1211 Geneva 23,Switzerland.

A. S. Belov and S. N. Gninenko are with the Institute for NuclearResearch of the Russian Academy of Sciences, Moscow 117312, Russia.

G. Bonomi and M. A. Subieta Vasquez are with the University ofBrescia, Dept. of Mechanical and Industrial Engineering, Via Branze 38,25133 Brescia, Italy and Istituto Nazionale di Fisica Nucleare, Sez. di Pavia,Via Agostino Bassi 6, 27100 Pavia, Italy.

P. Braunig and M. Oberthaler are with the University of Heidelberg,Kirchhoff Institute for Physics, Im Neuenheimer Feld 227, 69120 Heidelberg,Germany.

R. S. Brusa and L. Di Noto are with the University of Trento, Departmentof Physics and TIPFA, Via Sommarive 14, 38123 Povo, Trento, Italy.

L. Cabaret, D. Comparat and P. Yzombard are with Laboratoire AimeCotton, CNRS, Universite Paris Sud, ENS Cachan, Batiment 505, Campusd’Orsay, 91405 Orsay Cedex, France.

M. Caccia is with the Insubria University, Como-Varese, Italy.C. Canali and C. Regenfus are with the University of Zurich, Physics

Institute, Winterthurerstrasse 190, 8057 Zurich, Switzerland.R. Caravita, F. Castelli and S. Cialdi are with the University of Milano,

Dept. of Physics and Istituto Nazionale di Fisica Nucleare, Sez. di Milano,Via Celoria 16, 20133 Milano, Italy.

G. Consolati and R. Ferragut are with Politecnico di Milano, PiazzaLeonardo da Vinci 32, 20133 Milano, Italy and Istituto Nazionale di FisicaNucleare, Sez. di Milano, Via Celoria 16, 20133 Milano, Italy.

C. Da Via and C. Nellist are with the University of Manchester, OxfordRoad, Manchester M13 9PL, United Kingdom.

S. Di Domizio, D. Krasnicky, G. Testera and S. Zavatarelli are withIstituto Nazionale di Fisica Nucleare, Sez. di Genova, Via Dodecaneso 33,16146 Genova, Italy.

A. Fontana and P. Genova are with Istituto Nazionale di Fisica Nucleare,Sez. di Pavia, Via Agostino Bassi 6, 27100 Pavia, Italy.

M. Giammarchi and F. Prelz are with Istituto Nazionale di Fisica Nucleare,Sez. di Milano, Via Celoria 16, 20133 Milano, Italy.

T. Huse and O. Røhne are with the University of Oslo, Dept. of Physics,Sem Sælands vei 24, 0371 Oslo, Norway.

E. Jordan, A. Kellerbauer and G. Cerchiari are with Max Planck Institutefor Nuclear Physics, Saupfercheckweg 1, 69117 Heidelberg, Germany.

V. Lagomarsino is with the University of Genoa, Dept. of Physics, ViaDodecaneso 33, 16146 Genova, Italy.

A. Magnani, C. Riccardi and A. Rotondi are with the University of Pavia,Dept. of Nuclear and Theoretical Physics, Via Bassi 6, 27100 Pavia, Italy andIstituto Nazionale di Fisica Nucleare, Sez. di Pavia, Via Agostino Bassi 6,27100 Pavia, Italy.

Abstract—The aim of the AEgIS experiment is to measurethe gravitational acceleration for anti-hydrogen in the Earth’sgravitational field, thus testing the Weak Equivalence Principle,which states that all bodies fall with the same accelerationindependent of their mass and composition. AEgIS will makeuse of a gravity module which includes a silicon detector, inorder to measure the deflection of anti-hydrogen from a straightpath due to the Earth’s gravitational field, by detecting theannihilation position on its surface. A position resolution betterthan 10 µm is required to determine the gravitational accelerationwith a precision better than 10%. The work presented here ispart of a study of different silicon sensor technologies to realisea silicon anti-hydrogen detector for the AEgIS experiment atCERN. We here focus on the study of a 3D pixel sensor withFE-I4 readout, originally designed for the ATLAS detector at theLHC, and compare it to a previous monolithic planar detectorstudied, the MIMOTERA. The direct annihilation of low energyanti-protons (∼ 100 keV) takes place in the first layers and weshow that the charged annihilation products (pions and nuclearfragments) can be detected by such a sensor. The present studyaims at understanding the signature of an annihilation eventin a 3D silicon sensor, in order to assess the accuracy thatcan be achieved by such a sensor in the reconstruction of theposition of annihilation, when the same happens directly onthe detector surface. We also present a comparison betweenexperimental data and GEANT4 simulations and previous dataobtained with a silicon imaging detector. These results are beingused to determine the geometrical and process parameters to beadopted by the silicon annihilation detector to be installed inAEgIS.

C. Malbrunot is with European Organisation for Nuclear Research, PhysicsDept., 1211 Geneva 23, Switzerland and Stefan-Meyer-Institut fr subatomarePhysik, Boltzmanngasse 3, 1090 Vienna, Austria.

S. Mariazzi, S. Lehner, E. Widmann and J. Zmeskal are with Stefan-Meyer-Institut fr subatomare Physik, Boltzmanngasse 3, 1090 Vienna, Austria.

V. A. Matveev is with the Institute for Nuclear Research of the RussianAcademy of Sciences, Moscow 117312, Russia and Joint Institute for NuclearResearch, 141980 Dubna, Russia.

G. Nebbia is with Istituto Nazionale di Fisica Nucleare, Sez. di Padova,Via Marzolo 8, 35131 Padova, Italy.

P. Nedelec is with Claude Bernard University Lyon 1, Institut de PhysiqueNucleaire de Lyon, 4 Rue Enrico Fermi, 69622 Villeurbanne, France.

N. Pacifico and H. Sandaker are with the University of Bergen, Instituteof Physics and Technology, Allegaten 55, 5007 Bergen, Norway.

V. Petracek and M. Spacek are with Czech Technical University in Prague,FNSPE, Brehova 7, 11519 Praha 1, Czech Republic.

M. Prevedelli is with the University of Bologna, Dept. of Physics, ViaIrnerio 46, 40126 Bologna, Italy.

978-1-4799-0534-8/13/$31.00 ©2013 IEEE

Fig. 1. Principle of the AEgIS antihydrogen beam formation [1].

I. INTRODUCTION

A. Overview of the AEgIS experiment

THE main goal of the AEgIS experiment [1] is a first direct

measurement of the Earth’s gravitational acceleration

for antimatter, with the simplest form of electrically neutral

antimatter - antihydrogen. This would test both the Weak

Equivalence Principle and the prediction of General Relativity

that matter and antimatter should behave identically in the

gravitational field of the Earth. The experiment is built at the

Antiproton Decelerator (AD) at CERN, and during its first

phase a measurement of the gravitational acceleration with

1% precision will be attempted.

B. Principle of formation and detection of antihydrogen

Antihydrogen atoms will be formed by a resonant charge

exchange reaction between Rydberg (n∼20-30) positronium

and cold (100 mK) antiprotons (see fig. 1). The antihydrogen

states will be thus be defined by the positronium states.

The positrons will be supplied from a 22Na source, and

guided towards a slab of nanoporous silica which acts as a

positronium converter [2]. The antiprotons will be first trapped

and cooled down in the catching traps in the 5 T magnet and

the resonant charge exchange will take place in the mixing

trap in the 1 T magnet (fig. 2). A beam of antihydrogen atoms

will be formed by using inhomogeneous electric field, e.g

Stark acceleration [3]. Some of the antihydrogen paths will

be selected through the two gratings of a moire deflectometer

[4], as shown in fig. 3. This effect is purely classical. After

about 1 m of free fall, the rest of the antihydrogen atoms will

be detected with a position sensitive detector, consisting of a

thin (50 µm) silicon strip detector followed by an emulsion

detector and a scintillating fiber telescope. The vertical shift

of the fringe pattern formed by the moire deflectometer is

proportional to the value of g experienced by antihydrogen

(fig. 3). If the distance between the gratings and the detector

is 0.5 m, the vertical shift is expected to be ∼ 20 µm.

Em

uls

ion

Readout ASICs

77 (or 4) K vessel

Strip detector (25 um pitch, 50 um, 300 um support ribs)

moire deflectometer

~0.5 m

Low

tem

pe

ratu

re a

ntiH

ydro

gen (

4 (

100m

) K

)

Fig. 3. Scheme of the moire deflectometer and the hybrid position sensitivedetector [6]. Antihydrogen atoms enter from the left side and get reducedin vertical components from the moire deflectometer and annihilated onthe silicon detector. The annihilation products are detected by the emulsiondetector.

II. TEST BEAM WITH ANTIPROTONS

A. First measurements with monolithic planar sensor

Understanding the process of an antiproton annihilation in

silicon is essential for the design and construction of the silicon

detector where the antihydrogen annihilations will take place.

The main motivation of this study was to test the concept

of using a silicon detector as a position sensitive annihilation

detector. This would be the first step towards the development

of an ad-hoc silicon annihilation detector with a resolution

better than 10 µm. This kind of on-sensor annihilations of

antiprotons has been observed only once, in an experiment

which has made use of antiprotons with a momentum of 608

MeV/c and a non-segmented sensor [5]. In our case, during

the two test periods (test−beams), the AD supplied the initial

antiproton beam with a momentum of 100 MeV/c (kinetic

energy of 5.3 MeV). These particles were further slowed down

with several degraders, reaching the energy value of few 100

keV according to simulations [6], just before impinging on

the detector. The first detector that we tested in June 2012

was a monolithic active pixel sensor, the MIMOTERA [7].

The MIMOTERA was installed in a six way cross vacuum

chamber (∼ 10−6 mbar, at room temperature) which was

attached at the end of the main AEgIS apparatus and detecting

a fraction of the antiprotons not trapped by the apparatus. A

photo of this detector mounted on a PCB is shown in fig. 4.

Its total size is 2x2 cm2 and the thickness of the sensitive

volume is 14 µm, glued on a silicon mechanical support with

a total thickness of ∼ 600 µm. The MIMOTERA is back-

illuminated through an entrance window of ∼ 100 nm (SiO2

passivation layer). The maximum readout rate is configurable

up to 20 MHz. The readout system of the detector is divided

into four sub-arrays of 28 x 112 pixels, which are readout in

parallel. Each of the pixels consists of two independent readout

matrices, each of them built by 2 x 81 interconnected diodes,

5 x 5 µm2 each. This double-readout architecture allows the

MIMOTERA to operate without dead-time. The detector was

mounted ∼ 4 cm off axis (the same position adopted for

Fig. 2. Schematic view of the AEgIS apparatus. It consists of two magnets, 5 T and 1 T, which enclose the antiproton trap and the mixing trap respectively.The positrons are guided towards the main apparatus by their own transfer line. The gravity module is attached at the end of the 1 T magnet.

the 3D sensor, described later on) in order to have a lower

luminosity detecting only the antiprotons slightly deviated by

a fringe field from the 1 T magnet used for the antihydrogen

production. The simulated kinetic energy distribution of the

antiprotons implies that most of the antiprotons annihilate

within the first few microns of the detector. A schematic

view of the cross−section of this detector is given in fig.

5 showing an antiproton annihilation and the annihilation

products penetrating the active sensor volume.

Fig. 4. The MIMOTERA detector mounted on its PCB. The dimensions ofthe sensor are 2x2 cm2.

SiO2 layer (100 nm)

EPI silicon active

bulk (14 um)

A B A B A BA B readout matrices

Carrier substrate (600 um)

...

Antiprotons

protons

nucl.

fragments

pions

Fig. 5. Schematic view of the MIMOTERA detector, showing an annihilationevent occurring in the first layers of the detector and annihilation prongstravelling inside the sensitive volume.

B. Second measurements with 3D pixel sensor

While the small thickness and the relatively low granularity

of the MIMOTERA detector (153 µm pixel pitch) was im-

portant to determine the typical energy range of the clusters

and thus the needed dynamic range of the final detector, we

also wanted to study in more detail the tracks to understand

better the possible achievable resolution of the annihilation

point. For this reason we chose to test a thicker pixelated

detector because of a better tracking quality, and a 3D pixel

sensor (CNM 55)[8] was installed during the second test-beam

(fig. 6). The installation was done in the same way as for the

Fig. 6. The CNM 55 sensor bump bonded the FE-I4 R/O chip mountedon single chip card. The whole system is mounted on a flange before beinginstalled in the six way cross vacuum chamber.

MIMOTERA detector, providing as similar conditions with the

first test beam as possible.

This 3D pixel sensor and the readout ASIC [9] were origi-

nally designed for the Insertable B Layer [10] currently being

installed at the ATLAS [11] experiment at CERN, optimised

for looking at high energy particles from events occurring

every 25 ns. In spite of this the technology offered interesting

features relevant for our application described below.

The sensor consists of 80 (columns) x 336 (rows) = 26880

cells. The pixel size is 250x50 µm2 and the electrodes have a

diameter of 10 µm. The thickness of the active volume is 230

µm and its passivation layer is ∼ 3 µm thick and composed of

three layers of material: 1.5 µm Al, 0.8 µm doped polysilicon

and 1.150 µm thermal oxide, one on top of the other. A

schematic view of the architecture of the pixels is given in

fig. 7. The internal box drawn with a dotted line represents

a typical cell that consists of two readout n-columns around

six ohmic p-columns. The 3D design of the electrodes results

in a higher average electric field between the electrodes and a

shorter collection path, which is the reason for a larger signal

to noise ratio than the one for the planar pixel technology.

The USBpix hardware, used for the readout, is based on a

multipurpose IO-board (S3MultiIO) with a USB2.0 interface

to a PC and an adapter card which connects the S3MultiIO

to the Single Chip Adapter Card (SCC), where the FE and

the sensor are mounted. The S3MultiIO system contains a

programmable (Xilinx XC3S1000 FG320 4C) FPGA, which

provides and handles all signals going to the FE.

III. RESULTS

A. Monolithic planar sensor

The results of the MIMOTERA detector are fully reported

in [6]. What follows is a summary of the features of antiproton

Fig. 7. Schematic view of the 3D electrodes in CNM 55 pixel sensor. Thevolume within the dashed lines represents one pixel consisting of two readout(red) and six ohmic (blue) electrodes.

MIMOTERA WIDTH [pixel #]20 40 60 80 100

MIM

OT

ER

A H

EIG

HT

[p

ixe

l #]

20

40

60

80

100

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

22000

24000

Pix

el E

nerg

y [k

eV

]

no Al9 μm

6 μm

3 μm

Fig. 8. Sample of a triggered frame from the monolithic planar pixel detectorafter applying the noise cut. The dashed lines mark the areas that were coveredwith 3, 6 and 9 µm Al foil for studying the number of annihilations in eachof them [6]

tot

clu

s)/

Npix

(nclu

sN

-410

-310

-210

-110

1 FTFPChips

Data

Cluster Size [pixels]

2 4 6 8 10 12 14 16 18 20

DA

TA

/NS

IMN 0

2

4

Fig. 9. Cluster size (in number of pixels) distribution for clusters observedin the monolithic planar pixel detector and comparison with two GEANT4models. FTFP matches the data points better than CHIPS.

annihilations in a planar pixel detector. In these first tests we

successfully detected on-detector annihilations for the very

first time at such low antiproton energies. Some areas of the

detector were covered with 3, 6 and 9 µm Al foil for studying

the number of annihilations in each of them. The analogue

readout of the MIMOTERA detector required an off-line noise

cut. This cut was fixed to 150 keV, the value being extracted

as 5 RMS (30.3 keV) from the noise distribution of single

pixels. A sample frame (after the noise cut) is shown in fig.

tot

clu

s(E

)/1

Me

V/N

clu

s N

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

FTFPChips

Data

[MeV]totE

0 5 10 15 20 25

D2σ

+S2

σ)/

D-p

S(p

-15

0

15

Fig. 10. Cluster charge distribution for clusters observed in the monolithicplanar pixel detector and comparison with two GEANT4 models. Again, FTFPis closer to the data points, especially for energies > 5 MeV.

8.

For the cluster analysis, clusters were defined as conglom-

erates of pixels neighbouring in the horizontal, vertical and

diagonal direction, without using a seed-driven algorithm.

This would require assumptions on the geometrical profile

of the energy deposited, which is not possible due to the

heterogeneous nature of the clusters. In spite of the small

thickness, the relatively high statistics allowed observation

of a number of tracks from long-range annihilation products

travelling inside the thin detector plane. We found prongs

(most likely being protons) with lengths up to 2.9. mm.

The small ratio depth/pixel width resulted in small clusters

of mostly one pixel (∼ 70%), about 20 % of two pixels,

and the remaining part included three or more pixels. Part

of the one pixel cluster was due to secondary particles from

annihilations that happen elsewhere in apparatus. This noise

was simulated separately and also studied in the data [6]. The

cluster size (in number of pixels) is shown in fig. 9. The cluster

sizes ranged between 1 and 20 pixels. The total deposited

energy for each cluster was also measured and the results are

shown in fig. 10. This energy was measured to be up to 40

MeV. Charge saturation of pixels was observed in very few (<

10) annihilation events, due to the generation of slow charged

fragments characterized by a high dE/dx.

The cluster parameters (size and energy) from data analysis

were compared with two GEANT4 models: CHIPS (QGSP

BERT) [12] and FTFP (FTFP BERT TRV) [13], both suitable

for anti-proton at-rest-annihilation with nuclei, though none

of the models was to-date validated for annihilations at rest in

silicon. Pertinent validation was performed at rest for CHIPS

with Uranium and Carbon data, while the newer FTFP still

lacks validation for antiproton energies below 120 MeV [14].

The cluster energy distribution in fig. 10 shows an excess

of clusters with energy < 1 MeV in the data compared with

both simulations, which can be ascribed to a lower energy

deposition from the noise present in the data. The FTFP model

shows a better agreement, especially for energies > 5 MeV.

The cluster size distribution (fig. 9) shows that there is a

small systematic overestimation of the cluster size by CHIPS,

but the overall agreement is good with the two simulation

models adopted, with FTFP being the closer one to the data

points.

B. 3D sensor

The tests of the 3D sensor provided material for further

analysis of the antiproton annihilation in silicon, in particular

of tracks of the annihilation prongs. Given the same position-

ing of the two detectors in the vacuum chamber, the energy

distributions for the incoming antiprotons were similar. This

distribution is given in fig. 11. A sample acquisition frame is

shown in fig. 12. When compared to the typical frame of the

thin monolithic detector tested, more and longer tracks can

be observed, as a result of the thicker active volume. Tracks

are more likely due to the higher geometrical acceptance for

annihilation products travelling at angles different from the

direction of the beam.

Kinetic energy [MeV]0 0.1 0.2 0.3 0.4 0.5 0.6

tot

p)/

26

.7 k

eV

/Nkin

(Ep

N

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

Fig. 11. Kinetic energy distribution for the incoming antiprotons impingingon the 3D sensor.

Cluster charge distribution is shown in fig. 13, the size

distribution in fig. 14. Large clusters are observed, even in

excess of 80 pixels, due to long tracks. Cluster energy is lower

than what was observed with the MIMOTERA detector. The

reason for this discrepancy is to be found in two parameters:

the thicker passivation region is more likely to stop heavy frag-

ments, that would produce high energy deposits. In addition

to this, saturation of the single channel amplifiers was often

observed: the measured charge is thus expected to be lower

than the charge effectively deposited in the bulk in some cases.

When travelling through the detector, annihilation prongs

produce tracks from a few mm to 1.5 cm long. The annihilation

point can be reconstructed by fitting the tracks and calculate

the errors on their interception point. The position resolution

we were able to achieve applying this procedure with the CNM

55 sensor is 56.5 µm on X (pixel size of 250 µm) and 24.3 µm

on Y (pixel size of 50 µm). This resolution could very likely

be improved by using weighted fitting algorithms accounting

the charge of the pixels composing the tracks. However a

sense-full application of such an algorithm would require no

saturated pixels composing the tracks. The results shown in

this paper are the first we obtained using this procedure and

Fig. 12. Frame with antiproton annihilations observed in 3D sensor. Longtracks with different energy deposits, according to dE/dx: pions blue tracks(0.1-0.3 keV/µm); fast protons, green to turquoise tracks (> 0.5 keV/µm).

Fig. 13. Cluster charge distribution for clusters observed in 3D pixel sensor.

our further work will be focused on a better tracking algorithm

for antiproton annihilations, improved clustering etc.

C. Conclusions

Two types of silicon pixel detectors with different electrodes

geometry were used for studying the antiproton annihilation in

silicon. The data taken with the monolithic planar sensor, due

to the small thickness and low granularity, were more suitable

for analyses on the deposited energy. On the other hand, the

3D pixel sensor provided data that allowed a more precise

resolution on the annihilation point and study of the various

tracks produced by the annihilation prongs. The main results

so far, as presented in [6] and in this paper for the 3D sensor,

are:

• First on−sensor detection of antiproton annihilations at

rest in segmented silicon detectors.

• Total energy deposition up to 10 MeV per antiproton

annihilation, value highly dependant on the saturation of

pixels (40 MeV in the case of MIMTOERA).

• Identification of tracks from annihilation prongs up to 1.5

cm long (2.9 mm in the case of MIMOTERA).

Fig. 14. Cluster size distribution for clusters observed in 3D pixel sensor.

Fig. 15. Sample hitmap of the 3D sensor with two fitted proton tracks comingfrom an antiproton annihilation. Such long tracks are used to reconstruct theannihilation point.

• Comparison and agreement to some extent with GEANT4

(CHIPS and FTFP) models for the size and the charge of

the clusters produced from the antiproton annihilation.

• Identification of the annhilation prongs, such as protons,

pions, apha particles, heavy ions etc.

• Position resolution of ∼ 20-30 µm on the annihilation

point.

ACKNOWLEDGMENT

We thank the Bergen Research Foundation and the Research

Council of Norway for their support to this project. We also

thank Ole Dorholt for the technical support and the ATLAS 3D

community asfor the putting their detector at our disposal and

their help with the testing. Finally we thank Andrea Micelli

for his support throughout the test-beam campaign with the

3D CNM-55 sensor.

REFERENCES

[1] A. Kellerbauer et al., Proposed antimatter gravity measurement with an

antihydrogen beam Nuclear Instruments and Methods in Physics ResearchB 266 (2008) 351356

[2] Giovanni Consolati et al., Mesoporous materials for antihydrogen pro-

duction Chem.Soc.Rev., 2013, 42, 3821[3] G. Testera et al. Formation of a cold antihydrogen beam in AEGIS for

gravity measurements AIP Conference Proceedings 1037, 5, 2008[4] Markus K. Oberthaler et al. Inertial sensing with classical atomic beams

Physical Review A, vol 54, 1996 (3165-3176)[5] McGaughey et al. Low energy antiproton-nucleus annihilation radius

selection using an active silicon detector / target Nuclear Instrumentsand Methods, vol 249, 1986 (361-365)

[6] S.Aghion et al.,Annihilation of low energy antiprotons in silicon

arXiv:1311.4982 [physics.ins-det][7] R. Boll et al. Using Monolithic Active Pixel Sensors for fast monitoring

of therapeutic hadron beams - Radiation Measurements vol. 46, Issue 12,2011 (1971-1973)

[8] S. Grinstein et al. Beam Test Studies of 3D Pixel Sensors Irradiated

Non-Uniformly for the ATLAS Forward Physics Detector arXiv:1302.5292[physics.ins-det]

[9] V. Zivkovic et al. The FE-I4 pixel readout system-on-chip resubmission

for the insertable B-Layer project 2012 JINST 7 C02050[10] C. Da Via et al. 3D silicon sensors: Design, large area production and

quality assurance for the ATLAS IBL pixel detector upgrade, NuclearInstruments and Methods in Physics Research A 694 (2012) 321330

[11] The ATLAS Collaboration The Performance of the ATLAS Detector

Reprinted from Eur. Phys. J. C, Volume 70 (2010) and 71 (2011) atCERN for the benefit of the ATLAS collaboration 2010 and 2011 2011

[12] P. V. Degtyarenko, M. V. Kossov, and H.P. Wellisch - Chiral invariant

phase space event generator, I. Nucleon-antinucleon annihilation at rest

- Eur. Phys. J. A 8, 217-222 (2000)[13] A. Galoyan, V. Uzhinsky Simulation of Light Antinucleus-Nucleus

Interactions arXiv:1208.3614[14] Geant4 Physics Reference Manual