The AMS experiment: a magnetic spectrometer in space

PROCEEDINGS SUPPLEMENTS ELSEVIER

Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

www.elscvier.com/|~ate/n~

The AMS experiment" a magnetic spectrometer in space J. Casaus a* aDpto. Fusidn y Fisica de Particulas Elementales, CIEMAT, Avda. Complutense 22, 28040 Madrid, Spain The Alpha Magnetic Spectrometer (AMS) on the International Space Station (ISS) is the first large acceptance magnetic spectrometer to perform high statistics studies of cosmic rays in space. The experiment will address fundamental questions regarding primary antimatter and dark matter contents of the universe. In addition, the precision studies of cosmic rays in a wide energy range will result in a greatly improved understanding of the cosmic ray propagation in our galaxy. A prototype of the final detector was flown on the space shuttle Discovery in June 1998. The detector components are described and the results obtained on the recorded sample are presented. The final version of the detector will be placed on the ISS in 2005 for a 3-year exposure. The detector upgrades are presented and the sensitivity of the setup is outlined.

1. I N T R O D U C T I O N The asymmetry between matter and antimatter in the universe is one of the most important open issues in particle physics. The question whether there are cosmologically sizeable antimatter domains in our universe has deep impact on the foundations of the theories of particle physics, namely, CP-violation and baryon number non-conservation. The apparent absence of antinuclei in cosmic rays together with the lack of anomalous 7 - r a y observations originating from annihilation products are usually taken as an evidence for the standard baryogenesis approach [1]. However, baryogenesis models imply baryon number non-conservation and large levels of CP-violation which are not yet supported by particle physics experimental data. Alternative solutions which allow the present existence of primordial antimatter have been developed (see ref. [2] and references therein for a recent review). The predictions of some of these theories, which could imply antimatter domains in our galaxy [3], are achievable with sensitive searches for antinuclei in cosmic rays. Cosmological observations are consistent with the presence of dark matter (DM) as the major matter component of the universe [4]. More*On behalf of the AMS Collaboration.

over, the observation of galactic rotation curves indicates [5] the existence of a non-luminous halo which dominates the gravitational evolution of the galaxies. On the other hand, simulations of large scale structure or galactic formation [6] favor weakly interacting massive particles (WIMPs) as the main component needed to reproduce the observations. Among various proposed DM candidates, the neutralino arises as the natural one within the framework of supersymmetric extensions of the Standard Model. Neutralinos in the halo of our galaxy could be detected through their coannihilation into positrons [7], antiprotons [8], antideuterons [9] and v-rays [10] as deviations from their predicted spectra. The expected spectra of galactic cosmic rays is based on models which account for their origin and propagation. The current program in cosmic ray propagation calculations aims to describe all primary and secondary species in galactic cosmic rays (hadrons and leptons) and diffuse 7 - r a y background within a single model. The free parameters in these models (composition and injection spectra and galactic disk and halo properties) can be derived from measurements done in the heliosphere. However, only precise knowledge of the elemental and isotopic spectra of hadronic cosmic rays in a wide energy range can validate the rood-

0920-5632/03/$ - see front matter © 2003 Publishedby ElsevierScienceB.V. PII S0920-5632(02)01912-6

260

J Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

els and pin down the uncertainties in their free parameters. The faint signal searches (antimatter and dark matter) and the exclusive cosmic ray studies become difficult on earth due to the Earth's atmosphere shielding. The inaccessibility to the primary particle prevents a precise identification and measurement. A space flight provides the appropriate environment for precision measurements. AMS is a general-purpose experiment intended to perform measurements of charged cosmic rays and "~-rays with unprecedented statistics and sensitivity to faint signals. The large acceptance, long exposure and the detector identification capabilities in a wide energy range will improve by orders of magnitude the present knowledge of cosmic rays. In this paper a general description of the experiment will be presented. The layout of the prototype version of the spectrometer flown on the space shuttle will be described and the results obtained on the data sample recorded during the 10-day flight will be discussed. The detector upgrades for a long exposure (3-year) flight on the ISS and the expected performances will be outlined. 2. T H E A M S E X P E R I M E N T The AMS experiment is a fundamental physics experiment in space. The questions addressed by the experiment deal with the basics of our understanding of the universe. The AMS project is a joint effort between NASA and a collaboration of 40 High Energy Physics institutes from Europe, Asia and US and it involves more than 250 physicists and engineers. The AMS experimental program was divided into two phases: a precursor flight on the space shuttle and a long exposure flight on the ISS. The use of the space shuttle and the space station follows from an agreement between NASA and the US Department of Energy signed in 1995. The AMS collaboration has the full responsibility on the detector design, operation and scientific objectives of the experiment. NASA provides the space transportation and installation and supervises all the safety critical issues.

3. A M S ON T H E S P A C E S H U T T L E A prototype of the AMS detector, AMS-01, was flown on the space shuttle Discovery on 2-12 June 1998 (STS-91 mission). The main purpose of the precursor flight was to test the spectrometer design principles and to gain experience in the operation of the detector under real conditions. In addition, the STS-91 orbit parameters, similar to those of the ISS, allowed to study the expected backgrounds for the faint signal searches. The success of the 10-day mission (about 200 hours of data taking with 100 million physical triggers) allowed to obtain relevant results concerning particle fluxes in near Earth orbit and improved limits on the presence of antimatter in space. 3.1. Detector description The main components of the AMS-01 detector [11] are sketched in Fig. 1. The charged particle bending, provided by a permanent magnet, is measured by a silicon microstrip tracking system. Scintillation counter hodoscopes provide a measurement of the particle time of flight and thus the velocity. A shell of anticoincidence counters covering the inner surface of the magnet provides a veto for non-fiducial particle trajectories and a threshold Cerenkov counter allows to perform e/p separation in a wide energy range. The detector elements went through a space qualification procedure including vibration and themo-vacuum tests to fulfill the NASA requirements for shuttle launch and landing. The AMS-01 subsystems will be briefly described. 3.1.1. Magnet The permanent magnet [12] is made of highgrade Nd-Fe-B blocks arranged in 64 sectors of a cylindrical shell with a length of 800 mm, an inner diameter of 1115 mm and an outer diameter of 1298 mm. The geometrical acceptance of the magnet is 0.82 m2sr. The arrangement provides a quite uniform dipole field perpendicular to the magnet axis, low fringe field and negligible dipole moment. The maximum bending power of the magnet is 0.15 Tm 2 and the total weight including the support structure is 2.2 tons.

J Casaus/Nuclear Physics B (Proc, Suppl.) 114 (2003) 259-273

Particletrajectory ," ," TO~Layerl I

Low EnergyParticleShield I

Sl

J

$2

e .

.

.

.

.

.

.

.

.

.

e

L

°t

.

.

.

.

.

.

.

Tracker T I Planlm "1"2

..,a i

;

0

Ta

4)

T4

~

Z T6 m I

i

I

I

, i

~

i

i

,

I

i

i

t

',

,

I,.,

i

~

i

I

, i

[ Ao~,J

i cerenk°v

Figure 1. Schematic view of AMS as flown on STS-91. The cylindrical permanent magnet, tracker planes T1-T6, time of flight layers S1-$4, Cerenkov counter and anticoincidence counters are shown.

3.1.2.

Silicon tracker

The tracker [13] consists of six planes of double sided silicon sensors which provide a position resolution of 10 #m (30 ~m) in the bending (nonbending) plane. The silicon sensors are grouped in ladders of different lengths to match the cylindrical geometry of the magnet. The ladders are mounted on thin support plates which are fixed to a carbon fiber cylindrical shell attached to the inner part of the magnet. For this precursor flight only one third of the tracker system was instrumented (57 ladders out of 180), resulting in an effective geometrical acceptance of 0.3 m2sr. The tracker measures both position and energy loss (dE/dx) allowing to reconstruct the particle rigidity (p/Z) from the trajectory bending inside the magnetic field and to derive the momentum with the measurement of the particle charge, Z, from the energy loss measurements.

3.1.3.

Time

261

of flight

The time of flight system (TOF) [14] consists of two double planes of scintillator counters located above the first and below the last tracker planes. Each plane consists of 14 scintillator modules containing a 10 mm thick, 110 mm wide long scintillator paddle connected on both ends to 3 Hamamatsu R5900 photomultipliers. The time and position of the particle when traversing a paddle are derived from the high precision measurements at both sides of the each paddle. A resolution of 115-125 ps (14.5-18.5 mm along the paddle) was obtained on the 56 flight modules. The TOF system provides the experiment with the trigger by coincidence, with a precise measurement of the velocity of the particles and with a measurement of the absolute particle charge, Z, from the scintillation light produced in the counters. 3.1.4. Anticoincidence c o u n t e r s The anticoincidence system (ACC) consists of a layer of 16 scintillator paddles covering the inner surface of the magnet. Every 800 mm long and 8 mm thick counter is read with one photomultiplier on each side. The ACC allows to reject at the trigger level the background caused by particles passing through or interacting in the magnet walls and support structure. 3.1.5. T h r e s h o l d (~erenkov c o u n t e r The threshold Cerenkov counter (ATC) [15] is located below the bottom TOF plane and consists of two layers containing 168 cells with dimensions 110× 110 × 88 mm 3. Every cell is filled with eight 11 mm thick silica aerogel blocks with a refraction index n=1,035. The Cerenkov light produced in each cell is viewed by a Hamamatsu R5900 photomultiplier. The ATC allows to reject protons from electrons up to 3,5 GeV/c. 3.2. STS-91 f l i g h t After subdetector assembly at ETHZ (Zfirich), the spectrometer was shipped to Kennedy Space Center (KSC) for integration with the space shuttle Discovery hardware in January 1998. Continuous calibrations with cosmic muons and end-to-

262

J. Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

end tests were performed before Discovery launch on 2 June 1998 for the STS-91 mission. Data taking started briefly after nominal orbit was attained. The orbital inclination was 51.7 ° and the geodetic altitude ranged from 320 to 390 kin. The data was collected in three periods: a) 25 hours before docking with MIR space station. The AMS longitudinal axis (z-axis) pointed within 45 ° of the zenith; b) 4 days during shuttle-MIR docking. The AMS z-axis pointing varied between 40 ° and 145 ° of the zenith; c) 73 hours after undocking. The attitude was subsequently constrained within 1 ° to point to 0 °, 20 °, 45 ° and 180 ° of the zenith. The AMS trigger trigger was a 4-fold coincidence of at least one counter in each T O F plane, followed by an exclusion of paddle combinations incompatible with the tracker geometry or signals in the ACC. The resulting overall acceptance was 0.42 m2sr. A total of 100 million events were recorded on tape with a trigger rate ranging from 100 up to 700 Hz. In the South Atlantic Anomaly the high rate precluded effective data taking. 3.3. D e t e c t o r c a l i b r a t i o n Although most of the detector calibrations were obtained using flight data, beam measurements were performed as a complement and cross-check. Tracker alignment was performed using data taken at the CERN PS. 3x107 14 GeV 7r- events taken at about 3000 different beam incident angles and positions were used. The achieved residual misalignment of every single tracker ladder was below 10 pro. Charge measurement calibration was crosschecked using helium and carbon beams at GSI in Darmstadt. 5x107 0.2-2 GeV/n ions taken at about 600 different angular detector positions were used. The results obtained on the beam sample confirmed the flight data calibration. 3.4. R e s u l t s The analysis of the flight data and the results obtained have been described in detail in a series of papers [16-21]. Here only a brief summary of

the main features will be included. The reader is referred to the published references for further details. 3.4.1. A n t i h e l i u m s e a r c h A first test of the feasibility of the antimatter search by AMS was performed on the flight sample. The main backgrounds for the antihelium search are wrongly measured protons, electrons and helium. Upward going particles are excluded by requiring positive delay between bottom and top T O F planes. IZI = 1 particles can be eliminated by the absolute charge measurement, which ensures a charge contamination probability below 10 -7 . Protons and helium can be further suppressed with quality cuts on the rigidity measurement to eliminate wrong deflection sign reconstructions. With these restrictive requirements, a sample of 2.86× 106 He events with rigidity in the range 1.6 - 140 GV was selected. As shown in Fig. 2, no antihelium candidate was found.

10 s

!

He 0

10 4

events

¢ Q) 10 3 > UJ 10 2

10

1

,

-150

1

-100

~

I

I

0 50 -50 SignxRigidity

•

100

150 GV

Figure 2. Measured rigidity times the charge sign of the selected IZI = 2 sample for the antihelium search.

J Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

70~ 16 . ~p

From this measurement an upper limit at the 95% C.L. on the relative flux of antihelium to helium of 1.1× 10 -6 was obtained 1. This measurement showed the feasibility of the AMS search on the ISS and improved the existing limits both in sensitivity and rigidity range [22]. P r i m a r y fluxes Systematic measurement of cosmic rays reaching the Earth is of paramount importance to understand the cosmic ray production, acceleration and propagation mechanisms in our galaxy. The simultaneous measurement of protons, helium, electrons and positrons performed with AMS-01 can help to reveal specific properties of these dominant species. Cosmic rays are expected to be isotropic and to show falling spectra according to a power law in energy. The shielding effect of the geomagnetic field introduces a drop off in the flux at low energy known as geomagnetic cutoff. The rigidity cutoff varies from ,,, 100 MV up to ~ 10 GV depending on the latitude relative to the Earth geomagnetic equator (geomagnetic latitude). In this context, events with a measured rigidity above the local geomagnetic cutoff will be denoted as primary cosmic rays. From the flight data sample, the primary fluxes of protons for kinetic energies of 0.1-200 GeV, helium in the range 0.1-100 GeV/nucleon, electrons from 0.2 to 30 GeV and positrons up to 3 GeV 2 were measured. The high statistics collected, ,,, 10 v H, ,~ 106 He and ,-~ 105 e +, at different geographical positions allowed to confirm the isotropy of the cosmic primary flux below the percent level. Hence, the fluxes above cutoff obtained at different detector locations were combined into a single spectrum for each particle type. The primary proton spectrum, scaled by a factor E~ 5, is shown in Fig. 3 together with recent balloon measurements [23-26]. It is customary to describe the primary spectrum by a power law in rigidity, • = ~0R -~.

'I'*

263

O E]LEAP O

Thislxpednlnt CAPRICE

+4÷ ÷

3.4.2.

1Assuming the antihelium to have the same s p e c t r u m as the helium. 2The upper limit to the positron flux measurement range is set by the proton background.

N~

uJ ~s

6

3 2

[]

IMAX

b) 20

40

60

B0 100 120 140 160 180 200

E K (GeV)

Figure 3. The AMS-01 primary proton spectrum measurement multiplied by E~ 5 in units of GeV2"5/(m 2 sec sr MeV) compared with recent balloon measurements.

The fit of the proton flux over the range 10 < R < 200 GV yields ~/ = 2.78 =t= 0.009 ( f i t ) + 0.019 (sys) and G0 = 17.1J:0.15 ( f i t ) ~ l . 3 (sys)± 1.5 (V) GV2"Ts/( m2 sea sr MV), where the systematic errors account for acceptance, resolution and selection criteria uncertainties. The primary helium spectrum, multiplied by E~ 5, is shown in Fig. 4 compared to recent balloon measurements [23,25-27]. The fit of the He flux over the range 20 < R < 200 GV yields V = 2 . 7 4 i 0.010 (fit) =t= 0.016 (sys) and G0 = 2.52=t=0.09 (fit)+0.13 (sys) =t= 0.14 (V) GVZ74/( m2 sec sr MV). The primary electron and positron spectra and

264

J. Casaus /Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

1 [a) 450

0 e

10 -1

III

8 4`

400

'l',l,+÷÷{+

350

{{{

3OO 25O

0 0 0 0 0 0 0 0

10-2~)0

ee

o

++

,I,

• O O

a)

UJ

,

I

,

•

,

P

This experiment CAPRICE RICH

Oo

0% 0% o O

o

v

x

:=

10 -5

oo

IL

0 10-6 . . . . . . . . . . .

:,

18

1

,

¢

.

0.6

4O0

•

This experiment

O

BESS

[]

IMAX

b)

÷t+÷

25O 20O

|

20

+

A

AM898 H E A T 944-9~ C A P ~ C E 94

0.4

O +

O

0,3 0.2 0.1

b) |

O

O +

3OO

@

0,5

350

15O

Oo

o

450 LL

Io.~I'~ "

eeoc

10 -4

2O0

m o~< 150

lie

,

,

,

J

,

4O

60

80

100

0 1

E K (GeWA)

Ek (GeV)

Figure 4. The AMS-01 primary helium spectrum measurement multiplied by E~ 5 in units of G e V l 5 / ( m 2 sec sr) compared with recent balloon measurements.

Figure 5. The AMS-01 (a) flux spectra for e + and (b) positron fraction as a function of the energy compared to previous measurements and to a model prediction.

the differential positron fraction are shown in Fig. 5 together with recent measurements [28-30]. Experimental data are compared to a model prediction [31]. There is overall agreement between the AMS01 primary spectra and balloon most precise measurements. Minor differences at low energies can obey to solar modulation seasonal variations. The accuracy achieved with the AMS-01 measurements and the wide geographic scan of the fluxes results in a qualitative improvement in the experimental knowledge of the primary cosmic rays

reaching the Earth. 3.4.3. U n d e r cutoff fluxes In addition to the primary fluxes a substantial second spectrum was observed for all species below the geomagnetic cutoff. In this second spectrum similar amount of upward and downward going particles is observed. The flux changes with the geomagnetic latitude and has its maximum at geomagnetic equatorial latitudes (~ 70 (m 2 sec st) -1 for protons). An illustration of this effect is shown in Fig. 6, where

265

J Casaus/Nuclear Physics B (Proc. Suppl,) 114 (2003) 259-273

the measured e + second fluxes are shown as a function of the geomagnetic latitude.

25 f:

2O

a)

Downward

•

0

under cutoff helium, observed in the range 0.8-3 GV with an integrated flux ~ 10 -3 (m 2 sec sr) -1, shows (see Fig. 7) that most of the events are consistent with SHe (primary 3He/4He~_0.15).

~+ e

35 3O

0

25

15

~ 2o

10

"7A

(

c

0

0

5

•

0

W

0

•

•

•

0

@

E

ee

b)

X

lO

Upward

:~ 25 IJ-

o e*

! 5 o

2O

i

.

.

.

.

.

.

.

.

1

o

Mass

(GeV/c

2)

15:-

i: i

~o:-.

0 O

5

•

•

0

o ~ 'd.;" o~ :'oi.~-~'o.~~o,s o.~ o 7 oo 0.9 IOMI

Figure 6. Integrated under cutoff fluxes (0.2-2.5 GeV) for (a) downward and (b) upward going electrons and positrons as a function of the geomagnetic latitude.

Furthermore, the second spectrum has distinctive composition. In the case of e ±, the second flux shows a predominance of positrons over electrons, whereas electrons are about one order of magnitude more abundant in the primary flux. In addition, as can be seen in Fig. 6, the e + / e ratio changes from 4 to 1 when going from the geomagnetic equator to highest latitudes. The

Figure 7. Mass distribution for the under cutoff helium flux. 3He and 4He mass values as shown as a reference.

A tracing through the geomagnetic field [32] using of all the reconstructed events was performed [33]. The AMS geographical position at the detection time and the measured track parameters were used as inputs. As a result it was shown that the particles detected below the local geomagnetic cutoff had been originated in the Earth's atmosphere. The forward tracing of the particles showed that their trajectory ended in the atmosphere as well. According to the total flight time, T, for a given trajectory, two different populations can be distinguished. Short-lived particles, with T < 0.2 S, and long-lived particles, ~- >_ 0.2 s. These two populations form separated bands in Fig. 8, where the total flight time for under cutoff e ± is shown as a function of the measured kinetic energy.

,1:. Casaus /Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

266

• ~t

a) e" ~

:~ '

80 SO

b) e +

10

4Q 20 0 -20 40 "o

{

.50

=

-8Q

m ..J

80 60 4Q 20 0 -2D .40 .50

10-1

-80 -150

10-; I

10

-1

. . . . . . .

~

1

•

r

10

. . . . . . . -1

,!

,

n

1

E~ (GeV)

Figure 8. Flight time versus energy for under cutoff e+. For short-lived particles flight times show no dependence with energy. Flight time of long-lived particles (bands A and B) decreases with increasing energy.

The trajectory tracing shows that for shortlived particles the origin point shows no longitude dependence while long-lived particle originate from well defined regions. This effect is shown in Fig. 9 for the long-lived electrons and positrons. Furthermore, positive and negative long-lived particles have complementary origin and sink position, i.e., the regions of origin for positive particles coincide with regions of sink for negative and vice versa. Particles are shown to travel in cycles across the equator where they reach maximal altitude. They are reflected at the lowest points at mid and

-i00

-50

0

50

Longitude

100

1~0

Figure 9. Geographical origin of the under cutoff long-lived (a) electrons and (b) positrons. The lines indicate the geomagnetic field contours at 380 km. The regions A and B correspond to the bands A and B marked in Fig. 8.

polar latitudes. The long-lived particles are reflected across the equator hundreds of times. The number of cycles they can make before being absorbed in the atmosphere decreases with increasing energy• The properties of the observed second spectrum are qualitatively different from the extrapolation of previous measurements. The balloon observations in the upper layer of the atmosphere [34-36] deal with splash and return albedos [37], where the particles cross the equator (at most) once. Trapped particles in the radiation belts [38], discovered by satellites above ~ 1000 km, bounce across the equator for much longer time. 3.5. Discussion The precise measurements accomplished by AMS in its precursor flight originated further research work. The primary proton and helium flux measurements were included as new inputs for the atmo-

J Casaus/Nuclear Physics B (Proc. SuppL) 114 (2003) 259-273

spheric neutrino calculations [39] and the possible implications of the detected under cutoff fluxes were discussed [40,41]. An intensive program to reproduce the under cutoff fluxes started [42-46]. The aim was to understand the observed fluxes as a result of the interaction of the primary cosmic rays with the Earth's atmosphere and subsequent drift in the geomagnetic field. Monte Carlo simulations were performed using the measured primary fluxes, modulated by the geomagnetic cutoff, as particle beams and a model for the Earth's atmosphere [47] as a target. The particle interactions were included using either FLUKA package [48] or parametrized cross sections. The interaction products were traced along the geomagnetic field [32] and the STS-91 covered surface at flight altitude was introduced as a detection boundary. The results of the simulations are in agreement with the measured data. The measured absolute fluxes and spectra are well reproduced by the Monte Carlo models. The main features of the geographical origin and flight time of the under cutoff particles are present in the simulations and they follow naturally within the formalism of adiabatic invariants [49]. The distinctive composition of the under cutoff e + is understood as being originated by geomagnetic effects, in particular the east-west asymmetry in the primary proton rigidity cutoff. The anomalous isotopic composition and the spectrum of the under cutoff helium is reproduced when a coalescence model [50] for the light element yield in the interaction of the primary particles with the atmosphere nuclei is included. 4. A M S O N T H E ISS

The final version of the AMS detector will be placed at the $3 truss segment of the ISS in July 2005 for an exposure of 3 years. The long exposure together with the notably improved capabilities of the experiment will result in unprecedented precision studies of charged cosmic and "),-rays up to the TeV region. Currently, all subdetectors are in the construction phase. The subsystem construction and vet-

267

ification will take place until the first half of 2004. In summer 2004 all the subdetectors will be shipped to ETHZ (Ziirich) for final assembly and tests. The spectrometer will be transported to KSC at the beginning of 2005. There it will be integrated with the space shuttle and space station interface hardware. Systematic end-to-end tests will be performed before the shuttle launch. 4.1. D e t e c t o r d e s c r i p t i o n Several detector upgrades have been included in the final version of the spectrometer. The main components are displayed in Fig. 10.

J_ l

Figure 10. Schematic view of AMS spectrometer for the ISS. The superconducting magnet, the transition radiation detector (TRD), the time of flight (TOF), the silicon tracker, the ring imaging Cerenkov counter (RICH) and the electromagnetic calorimeter are shown.

The most substantial difference with respect to the prototype version is the improved bending power provided by a superconducting magnet. The tracker completion will double the geometrical acceptance and the inclusion of a Transi-

268

J Casazts/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

tion Radiation Detector (TRD), a Ring Imaging Cerenkov Counter (RICH) and an Electromagnetic Calorimeter (ECAL) will increase the particle identification capabilities of the spectrometer. The detector design had to face strict weight and power budgets coming from shuttle transportation and ISS service limitations. The total weight of the detector is limited to 6200 kg and the power consumption must be below 2 kW. The AMS subsystems will be briefly described. 4.1.1. S u p e r c o n d u c t i n g M a g n e t The cryogenic superconducting magnet [51] consists in an arrangement of 2 main dipole coils and 12 racetrack coils as shown in Fig. 11.

+Z AMS R a ~ r ~ C~ AINmml~ t~

I

~C~'~

+X

•I,Y J

" He

7 ~ialn

Vm

C r y t x ~ (4)

Figure 11. The AMS superconducting magnet layout. The magnet coils are surrounded by the helium vessel and located inside a vacuum case.

The inner dimensions of the AMS prototype permanent magnet are kept in the new magnet in order to minimize the changes in the inner subdetector design. The cryomagnet design ensures an intense and quite uniform field perpendicular to the vertical axis, a negligible dipole moment and the fulfillment of NASA safety regulations concerning

fringe magnetic fields on the ISS. The magnet operation temperature (1.8 K) is achieved with the use of a cryogenic system which includes 2500 liters of superfluid helium in a large dewar. The cryomagnet and the helium vessel are located inside an aluminum vacuum case which has additional purpose as primary structural support of the experiment. The total weight of the system including the vacuum case is 2.3 tons for a nominal cryomagnet bending power of 0.86 T m 2. 4.1.2. Silicon tracker The tracker will be composed of 8 layers of double sided silicon sensors with a total instrumented surface of 6.45 m 2. The design of the silicon sensors remains unchanged from the first flight. The 8 layers will be distributed in 5 support planes. The external ones will mount a single layer whereas the 3 central ones will support one silicon layer on each side. The improved bending power and the additional particle trajectory measurements in the silicon tracker will result in a resolution in the rigidity measurement of 1.5% at 10 GV and a maximum detectable rigidity above 1 TV. 4.1.3. T i m e o f flight Although the T O F basic design and functionality remains unchanged from the precursor flight, the increase in the AMS magnetic field forced a change in the photomultiplier model and the design of complicated light guides connecting the scintillator end with the PMT. In addition, the total number of scintillator modules has been reduced to 34. The expected resolution in the velocity measurement is 3.5% for ~=1 particles. 4.1.4. Transition R a d i a t i o n D e t e c t o r The TRD [52] is composed of 20 layers of straw tube modules and polypropylene fiber radiators mounted in an octogonal structure as shown in Fig. 12. The upper 4 layers and the lower 4 layers are oriented in the perpendicular direction of the central 12 layers. The TRD is located on top of the experiment above the upper T O F plane. The 5248 straw tubes, operated at high voltage

J Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

269

Figure 12. Artistic view of AMS TRD support octagon. The magnetic field direction is indicated in the figure. The 20 layers of straw tubes are mounted along the octagon axis.

Figure 13. The AMS RICH detector layout. The components shown are, from top to bottom, the Cerenkov radiator, the external conical reflector and the detection plane.

with a mixture of Xe and CO2, are able to detect the transition radiation light produced by ultra relativistic particles. The TRD is able to perform electron proton separation over the energy range 1.5-300 GeV with a rejection factor 102-103.

4.1.6. E l e c t r o m a g n e t i c C a l o r i m e t e r The ECAL [54] consists of layers of lead foils with glued scintillating fibers. Every 11 layers are arranged into a superlayer with the fibers oriented along the same direction. The complete setup includes 9 superlayers with alternating fiber directions which results into a total radiation depth of 15 X0 for shower development. The scintillation light is transported by the fibers to a set of photomultiplier tubes located at the 4 sides of the central square active part. A total of 1296 readout channels (324 Hamamatsu R7600-00-M4 multianode PMTs) provide a 3D image of the electromagnetic shower with 18 samples in depth. An artistic view of the complete ECAL assembly is displayed in Fig 14. The AMS electromagnetic calorimeter will provide the experiment with a precise measurement of the energy of electromagnetic particles and an electron proton separation in a wide energy range. The expected energy resolution is -~ 3% for 100 GeV electrons with a hadron rejection factor of 104 for E <~ 1 TeV. In addition, thanks to the 3D imaging capabilities of the detector, the shower shape reconstruction allows to determine the particle direction with an resolution better than 1° for E ~> 100 GeV.

4.1.5. R i n g I m a g i n g (~erenkov C o u n t e r The RICH [53] has three main components as seen in Fig. 13. The Cerenkov radiator, attached to the lower TO F honeycomb, consisting of a set of silica aerogel tiles with low refraction index and a small fraction of NaF tiles 9, the detection plane, instrumented with 680 R7600-M16 multianode Hamamatsu PMTs and the external conical reflector, which covers the expansion length between the radiator and the detection plane. The RICH will measure the Cerenkov cone opening angle of relativistic particles and thus provide AMS with a precise particle velocity measurement (a(13)/~ ~ 0.1%). In addition, with the use of a photon counting algorithm on the associated signals, it will provide the experiment with an independent absolute charge determination for Z ~<26 with a ~-10 -2 charge confusion probability. 3The final decision on the RICH aerogel refraction index and on the size of the NaF area will be taken based on the results of ongoing tests.

270

J Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

Figure 14. Artistic view of the AMS ECAL. The photomultiplier tubes are mounted on the sides of the central lead and scintillating fiber active part.

4.2. Expected performances The upgrades in the AMS detector design, the large geometrical acceptance (0.5 m2sr) and the long exposure of the experiment will allow to extend the present range of precise measurements up to energies of O(1 TeV). The detector strategy is now based on redundant measurements which ensure a proper background rejection for rare signal searches. The electron hadron rejection makes use of both TRD and ECAL measurements to reach a rejection factor of 10 -~ up to 300 GeV. The particle absolute charge measurement is based on tracker, T O F and RICH for light nuclei with an expected proton helium confusion probability below 10 -9 . In addition, the new subdetectors provide novel capabilities for light isotope identification and Vray measurement. The precise velocity measurements coming from the RICH allow to perform light isotope separation for E ~< 10 GeV/n. The combined TRD and tracker measurements allow to identify and to measure the photon conversions into electron-positron pairs in the upper part of the detector. The v - r a y energy resolution achieved with this conversion mode is ~ 2% at 10 GeV with an angular resolution better than 0.03 ° for E ~> 10 GeV for an effective geometrical acceptance of ,,, 0.06 m2sr. Finally, the implementation of an ECAL stand-alone trigger will provide direct measurements of v-rays by requiring the absence of any charged track in the spectrometer.

The use of the ECAL stand-alone measurement profits from the calorimeter energy and angular resolution up to the TeV region with an effective geometrial acceptance of 0.06 m2sr. The overwhelming statistics available after the 3-year mission can be illustrated with the following estimations. The experiment will collect a total of ~109 helium nuclei which will allow to discover (set a limit to) a relative flux of antihelium to helium at the 10 -9 level up to rigidities O(1 TV). Antiprotons will be identified up to 400 GeV with a total expected statistics of about 106 events. Concerning the primary hadron flux measurements, AMS will collect l0 s protons and l0 T He with energies above 100 G e V / n and 105(104) B(C) with E ~> 100 G e V / n and it will precisely measure the individual element fluxes up to energies O(1 TeV/n). As regards the light isotope separation, AMS will identify l0 s deuterium (3He) nuclei from protons (4He) up to 10 GeV/n. In addition, the ratio of the radioactive 1°Be to the stable 9Be will be measured up to energies around 10 GeV/n. With respect to the lepton component, AMS will measure the electron flux up to energies O(1 TeV) and it will identify and measure positrons up to 400 GeV. Finally, the new v - r a y detector capabilities will allow to precisely measure the galactic and extragalactic diffuse Vray background in the energy range from 1.5 GeV to 1 TeV and to identify ")'-ray sources with sensitivities comparable to dedicated experiments. A few examples of the fundamental physics issues which can be faced with the expected data sample will be outlined. According to a possible scenario with m a t t e r antimatter domains within our galaxy in a matter dominated universe [55], which is one of the experimentally non excluded m a t t e r - a n t i m a t t e r asymmetric models [56], antimatter coming from a globular-like star cluster could reach the heliosphere. The observed v - r a y spectrum sets an upper limit the mass of the cluster of ~105 solar masses. On the other hand, the globular cluster mass should exceed ,,~103 solar masses to survive annihilation. Within this limits, an antihetium to helium flux ratio in cosmic rays ranging from 10 -6 to 10 -s is expected. The AMS sensitivity is thus beyond the predicted limits for such a model.

J Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

Regarding the dark matter detection capabilities, AMS provides an unique oportunity to perform a multichannel search for neutralino coannihilation in the galactic halo. The most promising signatures arise in the antimatter and in the "/-ray channels. Although antiproton detection at low energies was traditionally considered the smoking gun for dark matter detection due to the clean kinematic cutoff for secondary production, new precise measurements [57-59] and the reevaluation of the theoretical predictions for secondary production [60] have converged into a quantitative agreement. However, it has been pointed out [61] that a number of supersymmetric configurations could be excluded with a precise antiproton flux measurement at higher energies as expected from AMS. Recently, it has been suggested that a cleaner signal could be observed in the low energy antideuteron measurement [9]. For energies below 1 GeV/n, the secondary production is expected to be several orders of magnitude below the neutralino signal. The antideuteron production cross sections are so small that no single antideuteron has been detected in cosmic rays to date. The AMS large acceptance and identification capabilities will result in a few secondary antideuterons detected after the 3-year exposure. The positron channel has become the most popular one after the HEAT ballon experiment reported [62] an excess in the positron to electron ratio for energies in the 10-50 GeV range. Although the possibility of a neutralino annihilation origin of this measurement has already been discussed [63], the low statistical significance of the measurement and the fact that the excess appears close to the high energy limit for positron identification of the experiment require new precise meaurements. The "y-ray channel exhibits the advantage of providing a monochromatic beam for the annihilation into 7"Y and Z'~ with an energy equal (or close) to the neutralino mass [64]. Moreover, the signal over background ratio can be increased by pointing towards the galactic center were the coannihilation rate is expected to be maximum. The AMS v-ray detection capabilities both in sensitivity and energy range will allow to perform

271

the search for a significant range of neutralino masses [65]. In general, the neutralino coannihilation probability depends quadratically on the galactic halo radial density, which is poorly known. This strong dependence can be reduced in the case of v-rays, as proposed in [66], with the search for spectral features in the extragalactic "/-ray background. In addition to the antimatter and dark matter search, the AMS experiment will constrain the models for production, acceleration and propagation of cosmic rays with the simultaneous measurement of the hadronic and the e ± components. For instance, carbon is expected to be of primary origin, i.e., present at the cosmic ray sources, whereas boron is believed to be a spallation product of the primary carbon in its path through the interstellar medium. The ratio B/C is thus used to estimate the amount of matter traversed by cosmic rays since their acceleration. In specific models, this ratio defines the cosmic ray escape path length distribution (leaky-box model) or the spatial diffusion coefficient (diffusion models). Furthermore, of the ratio of the /3-radioactive 1°Be to the stable 9Be can be used to detemine cosmic ray confinement time in the galaxy and, in diffusion models, the effective thickness of the galactic halo. The B/C and l°Be/gBe ratios are only two examples of the basic measurements needed to validate the current models of galactic cosmic ray propagation and to constrain their free parameters. The precise measurements provided by AMS experiment will define the benchmark data for those models. Finally, the AMS "r-ray detection capabilities will be used for v-ray point source detection in an unexplored energy range [67]. The inclusion of an astro mapper system [68] will provide the experiment with the required pointing accuracy. The estimated expected sensitivity for 5a detection of point sources is 10-s (cm 2 sec) -1 for E > 1 GeV and 10 -9 (cm 2 sec) -1 for E > 10 GeV. Although these are modest values when compared with a dedicated experiment as GLAST [69], AMS has the advantage of becoming operative at least one year in advance.

272

J Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

5. C O N C L U S I O N S The AMS spectrometer is a fundamental physics experiment performed in space. The results obtained with a prototype version of the detector on a short precursor flight in 1998 showed the feasibility of a long duration mission on the ISS. The inclusion of a superconducting magnet and new subdetectors in the final detector layout will result into greatly improved capabilities of the experiment. Currently, all subdetectors are in the construction phase. In July 2005 the spectrometer will be transported to the ISS where the experiment will take data for a 3-year period. The large data sample collected after the experiment long exposure will allow a sensitive search for antimatter and dark matter in cosmic rays and astrophysics studies regarding cosmic ray production, acceleration and propagation. As a ~,-ray observatory AMS will open an unexplored energy window in the study of "),-ray sources.

REFERENCES

1. S. Weinberg, Phys. Rev. Lett. 42 (1979) 850; E.W. Kolb, M.S. Turner, Ann. Rev. Nucl. Part. Sci. 33 (1983) 645. 2. Yu. Galaktionov, Rep. Prog. Phys. 65 (2002) 1243. 3. M.Yu. Khlopov, Gravitation & Cosmology 4 (1998) 69. 4. B. Sadoulet, Rev. Mod. Phys. 71 (1999) $197; M.S. Thrner, J.A. Tyson, Rev. Mod. Phys. 71 (1999) S145; N.A. Bahcall et al., Science 284 (1999) 1481. . Y. Sofue, V. Rubin, Ann. Rev. Astron. Astrophys. 39 (2001) 1375. , A. Jenkins et al., Astrophys. J. 499 (1998) 20. E. Diehl, Phys. Rev. D 52 (1995) 4223. 7. 8. G. Jungman, M. Kamionkowski, Phys. Rev. D 49 (1994) 2316. . F. Donato et al., Phys. Rev. D 62 (2000) 043003. 10. M. Srednicki et al., Phys. Rev. Lett. 56 (1986) 263; Erratum-ibid 56 (1986) 1883.

11. S.C.C. Ting in: M.Chaian, K.Huitu, R.Orava (Eds.), The First Arctic Workshop on Future Physics and Accelerators, World Scientific, Singapore, 1995, p. 1. 12. S. Ahlen et al., Nucl. Instr. and Meth. A 350 (1994) 351. 13. J. Alcaraz et al., Nuovo Cimento II A 112 (1999) 1325. 14. D. Alvisi et al., Nucl. Instr. and Meth. A 437 (1999) 212. 15. D. Barancourt et al., Nucl. Instr. and Meth. A 454 (2000) 174. 16. J. Alcaraz et al., Phys. Lett. B 461 (1999) 387. 17. J. Alcaraz et al., Phys. Lett. B 472 (2000) 215. 18. J. Alcaraz et al., Phys. Lett. B 484 (2000) 10. 19. J. Alcaraz et al., Phys. Lett. B 490 (2000) 27. 20. J. Alcaraz et al., Phys. Lett. B 494 (2000) 193. 21. M. Aguilar et al., Phys. Rep. 366 (2002) 331. 22. T. Saeki et al., Phys. Lett. B 422 (1998) 319. 23. M. Boesio et al., Astrophys. J. 518 (1999) 457. 24. E.S. Seo et al., Astrophys. J. 378 (1991) 763. 25. T. Sanuki et al., Astrophys. J. 545 (2000) 1135. 26. W. Menn et al., Astrophys. J. 533 (2000) 281. 27. J. Buckley et al., Astrophys. J. 429 (1994) 736. 28. R.L. Golden et al., Astrophys. J. 436 (1994) 769. 29. S.W. Barwick et al., Astrophys. J. 498 (1998) 779. 30. M.A. DuVernois et al., Proceedings of 26th ICRC, Vol. 3, 1999, p. 49. 31. I.V. Moskalenko, A.W. Strong, Astrophys. J. 493 (1998) 694. 32. N.A. Tsyganenko, D.P. Stern, J. Geophys. Res. 101 (1996) 27187. 33. M.A. Huang et al., Chinese J. Phys. 39 (2001) 1. 34. S.D. Verma, J. Geophys. Res. 72 (1967) 915. 35. M.H. Israel, J. Geophys. Res. 74 (1969) 4701. 36. S.W. Barwick et al., J. Geophys. Res. 103 (1998) 4817. 37. S.B. Treiman, Phys. Rev. 91 (1953) 957; E.C. Ray, J. Geophys. Res. 67 (1962) 3289. 38. S.B. Treiman, Phys. Rev. 93 (1954) 544.

J. Casaus/Nuclear Physics B (Proc. Suppl.) 114 (2003) 259-273

39. T.K. Gaisser et al., Proc. 27th ICRC, 2001, p. 1643. 40. P. Lipari, Astropart. Phys. 16 (2002) 295. 41. V. Plyaskin, Phys. Lett. B 516 (2001) 213. 42. L. Derome et al., Phys. Lett. B 489 (2001) 1. 43. L. Derome et al., Phys. Lett. B 515 (2001) 1. 44. L. Derome et al., Phys. Lett. B 521 (2001) 139. 45. E. Fiandrini et al., astro-ph/0106241. 46. P. Zuccon et al., astro-ph/0111111. 47. A.E. Hedin, J. Geophys. Res. 96 (1991) 1159. 48. A. Ferrari et al., Phys. Medica 17 (2000) 1. 49. C.E. Mc Ilwain, J. Geophys. Res. 66 (1961) 3681. 50. W.J. Lope et al., Phys. Rev. C 52 (1995) 2004. 51. B. Blau, 1st Int. Conf. on Particle and Fundamental Physics in Space, Elba 2002, to appear in Nucl. Phys. B, Proc. Suppl. 52. Th. Siedenburg, 1st Int. Conf. on Particle and Fundamental Physics in Space, Elba 2002, to appear in Nucl. Phys. B, Proc. Suppl. 53. J. Casaus, 1st Int. Conf. on Particle and Fundamental Physics in Space, Elba 2002, to appear in Nucl. Phys. B, Proc. Suppl. 54. F. Pilo, 1st Int. Conf. on Particle and Fundamental Physics in Space, Elba 2002, to appear in Nucl. Phys. B, Proc. Suppl. 55. M.Yu. Khlopov et al., Phys. Rev. D 62 (2000) 083505. 56. A.D. Dolgov, Nucl. Phys. Proc. Suppl. 95 (2001) 42. 57. T. Maneo et al., Astropart. Phys. 16 (2001) 121. 58. Y. Asaoka et al., Phys. Rev. Lett. 88 (2002) 051101. 59. M. Boezio et al., Astrophys. J. 561 (2001) 787. 60. I.V. Moskalenko et al., Astrophys. J. 565 (2002) 280. 61. A. Bottino et al., Phys. Rev. D 58 (1998) 123503. 62. S.W. Barwick et al., Astrophys. J. 482 (1997) L191; S. Coutu et al., Proc. 27th ICRC, 2001. 63. E.A. Baltz et al. Phys. Rev. D 65 (2002) 063511. 64. L. Bergstr6m, H. Snellman, Phys. Rev. D 37 (1988) 3737. 65. J.L. Feng et al., Phys. Rev. D 63 (2001) 045024.

273

66. L. Bergstr6m et al., Phys. Rev. Lett. 87 (2001) 251301. 67. G. Lamanna, 1st Int. Conf. on Particle and Fundamental Physics in Space, Elba 2002, to appear in Nucl. Phys. B, Proc. Suppl. 68. http://carso.area.trieste.in/amical.html 69. R.P. Johnson, 1st Int. Conf. on Particle and Fundamental Physics in Space, Elba 2002, to appear in Nucl. Phys. B, Proc. Suppl.