8 The Ring Imaging Cherenkov detector of the AMS experiment: test beam 0 results with a prototype 0 2 Lu´ısa Arrudaa, ∗ Fernando Bar˜aoa, Patr´ıcia Gonc¸alvesa, Rui Pereiraa n a aLIP/IST J Av. Elias Garcia, 14, 1o andar 9 2 1000-149Lisboa, Portugal e-mail: [email protected] ] h p The Alpha Magnetic Spectrometer (AMS) to be installed on the International Space Station (ISS) will be - equipped with a proximity Ring Imaging Cherenkov (RICH) detector for measuring the velocity and electric o chargeof thecharged cosmic particles. This detector will contributetothehigh level of redundancyrequired for r AMS as well as to the rejection of albedo particles. Charge separation up to iron and a velocity resolution of t s theorder of 0.1% for singly charged particles are expected. A RICH protoptypeconsisting of a detection matrix a with 96 photomultiplier units,a segment of a conical mirror and samples of the radiator materials was built and [ itsperformance was evaluated. Results from thelast test beam performed with ion fragments resulting from the 1 collision of a 158GeV/c/nucleon primary beam of indium ions (CERN SPS) on a lead target are reported. The v large amount of collected data allowed to test and characterize different aerogel samples and thesodium fluoride 4 radiator. In addition, the reflectivity of themirror was evaluated. The data analysis confirms the design goals. 8 4 1. The AMS02 detector 4 . 1 AMS [1] (Alpha Magnetic Spectrometer) is a 0 precisionspectrometerdesignedtosearchforcos- 8 micantimatter,darkmatterandtostudytherel- 0 ativeabundance ofelements andisotopic compo- : v sition of the primary cosmic rays. It will be in- i X stalled in the International Space Station (ISS), where it will operate at least for three years. r a The spectrometer willbe capable ofmeasuring the rigidity (R ≡ pc/|Z|e), the charge (Z), the velocity (β) and the energy (E) of cosmic rays within a geometrical acceptance of ∼0.5m2sr. Fig. 1 shows a schematic view of the AMS Figure1. AwholeviewoftheAMSSpectrometer. spectrometer. The system is composed of sev- eral subdectors: Transition Radiation Detector (TRD), Time-of-Flight (TOF), Silicon Tracker 2. The AMS RICH detector (STD), Anticoincidence Counters (ACC), super- conducting magnet, Ring Imaging Cˇerenkov de- The RICH is a proximity focusing device with a dual radiator configuration on the top made tector and electromagnetic calorimeter (ECAL). of 92 aerogel 25mm thick tiles with a refractive index 1.050 and sodium fluoride (NaF) tiles with athicknessof5mminthecentercoveringanarea ∗IwishtothanktheFundac¸a˜oparaaCiˆenciaeaTecnolo- of 34×34cm2. The NaF placement prevents the gia for all the financial support for the journey to Siena loss of photons in the hole existing in the center andIwouldliketothanktheconferenceorganizersforthe ofthereadoutplane(64×64cm2),infrontofthe grant received to attend the meeting, as well as, for the excellentorganization. ECAL calorimeter located below. 1 2 The detection matrix is composed of 680 mul- a mechanical test. The final integration of RICH tipixelizedphotonreadoutcellseachconsistingof in AMS will take place at CERN in 2008. aphotomultipliercoupledtoalightguide,HVdi- vider plus front-end (FE) electronics, all housed and potted in a plastic shell and then enclosed in a magnetic shielding. The photon detection is made with an array of multianode Hamamatsu tubes (R7600-00-M16) coupled to a light guide. The effective pixel size is 8.5mm. A high reflectivity conical mirror surrounds the whole set. It consists of a carbon fiber re- inforced composite substrate with a multilayer coating made of aluminium and SiO2 deposited Figure2. Ontheleft: ViewoftheRICHdetector. on the inner surface. This ensures a reflectiv- On the right: Beryllium event display generated ity higher than 85% for 420nm wavelength pho- in a NaF radiator. tons. Figure 2 (left) shows a schematic view of the RICH detector. RICH was designed to measure the velocity 3. Velocity (β) and charge (Z) reconstruc- (β ≡ v/c) of singly charged particles with a res- tion olution ∆β/β of 0.1%, to extend the charge sep- aration (Z) up to iron (Z=26), to contribute to Achargedparticlecrossingadielectricmaterial e/p separation and to albedo rejection. of refractive index n, with a velocity β, greater In order to validate the RICH design, a pro- than the speed of light in that medium emits totype with an array of 9×11 cells filled with 96 photons. The aperture angle of the emitted pho- photomultiplier readout units similar to part of tonswithrespecttotheradiatingparticletrackis the matrix of the final model was constructed. known as the Cˇerenkov angle, θ , and it is given c The performance of this prototype has been by cosθ = 1 (see [2]). c β n tested with cosmic muons and with a beam of It followsthatthe velocityofthe particle,β, is secondary ions at the CERN SPS produced by straightforward derived from the Cˇerenkov angle fragmentation of a primary beam in 2002 and reconstruction,whichisbasedonafittothepat- 2003. Thelightguidesusedwereprototypeswith tern of the detected photons. Complex photon a slightly smaller collecting area (31×31mm2). patterns can occur at the detector plane due to Different samples of the radiator materials were mirror reflected photons, as can be seen on right tested and placed at an adjustable supporting display of Figure 2. The event displayed is gen- structure. Different expansion heights were set erated by a simulated beryllium nuclei in a NaF in order to have fully contained photon rings on radiator. the detection matrix like in the flight design. A The Cˇerenkov angle reconstruction procedure segmentofa conicalmirrorwith1/12ofthe final relies on the information of the particle direction azimuthalcoverage,whichisshowninleftpicture providedbythetracker. Thebestvalueofθ will c of Figure 3, was also tested. resultfromthemaximizationofalikelihoodfunc- The RICH assembly has already started at tion, built as the product of the probabilities, p , i CIEMATinSpainandisforeseentobefinishedin that the detected hits belong to a given(hypoth- July 2007. A rectangular grid has already been esis) Cˇerenkov photon pattern ring, assembled and has been subject to a mechani- cal fit test, functional tests, vibration tests and nhits L(θ )= pni[r (θ )]. (1) vacuum tests. The other grids will follow. The c i i c refractive index of the aerogel tiles is being mea- iY=1 suredandtheradiatorcontainerwassubjectedto Here r is the closest distance of the hit to the i 3 E31 • DATA *1.0.9 102 MC Db/b0.8 0.7 DATA 0.6 MC 0.5 10 0.4 0.3 0.2 1 0.1 -0.2-0.15-0.1-0.05 0 0.05 0.1 0.15 0.2 0 0 5 10 15 20 25 1-BETA LIP x 10-2 Z(scint&std) Figure4. Comparisonofthe(β−1)∗103distribu- tion for helium data (black dots) and simulation Figure 3. Protype with reflector (left). Top view (shaded)(left). Evolutionoftheβresolutionwith of the test beam 2003 experimental setup using thechargeobtainedforthesameaerogelradiator. CERN SPS facility (right). Simulated points for Z=2, 6, 16 are marked with full squares (right). Cˇerenkov pattern and n is the hit signal. For a i more complete description of the method see [3]. correspond to particles impinging vertically and The Cˇerenkov photons produced in the radia- generating fully contained rings. The beta re- torare uniformly emitted alongthe particle path constructed from simulated helium data is also insidethedielectricmedium,L,andtheirnumber shown superimposed with a good agreement be- per unit of energy (N) depends on the particle’s tween data and Monte Carlo (MC). charge, Z, and velocity, β, and on the refractive The charge dependence of the velocity relative index, n. Therefore electric charge (Z) is deter- resolution for the same radiator is shown in the mined from the signalevaluation and taking into right plot of Figure 4. The observed resolution account the different detection efficiencies. varies according to a law ∝1/Z, as it is expected 1 from the charge dependence of the photon yield 2 N ∝Z ∆L 1− β2n2 (2) intheCˇerenkovemission,uptoasaturationlimit (cid:18) (cid:19) setbythepixelsizeofthedetectionunitcell. The 3.1. Results with the RICH prototype function used to perform the fit is the following: The large amount of collected data in the last test beam at CERN, in October 2003,performed 2 A with ion fragments resulting from the collision of σ(β)= +B2 (3) s Z a 158 GeV/c/nucleon primary beam of indium (cid:18) (cid:19) ions(CERNSPS)onaleadtarget,allowedtotest where, A means the β resolution for a singly the beta and charge reconstruction algorithms, chargedparticlewhileB meanstheresolutionfor as well as to characterize the used radiators [4]. a very high charge generating a large number of Figure 3 (right) showsa generalview of the 2003 hits. The fitted values are A = 0.872 ± 0.003 testbeamsetupintheexperimentalareaH8-SPS and B = 0.047 ± 0.001 for the run conditions at CERN. stated above. Simulated data points for Z=2, Theresolutionoftheβ measurement,obtained 6, 16 are marked upon the same plot with full as explained in Section 3, was estimated using squares. Once more the agreement between data a Gaussian fit to the reconstructed β spectrum, and MC measurements for charges different of shown in left plot of Figure 4 for helium nu- Z=2 is good. clei. Data were collected with the aerogel radi- The distribution of the reconstructed charges ator n=1.05, 2.5cm thick together with an ex- in an aerogel radiator of n=1.05, 2.5cm thick is pansion height of 35.31cm. The events shown shown in left plot of Figure 5. The spectrum en- 4 hchaanrcgeespaeasktrsuocvteurrethoefwwheolllesreapnagreatuepdtiondZi=vi2d8u.al HeLiBeBC s zrec000.4..455 Thechargeresolutionforeachnuclei,shownin H NOFNeNa 0.35 rdpiiegvahikdtsupasaelnlGeeclatouefdsFsibiagynutrfihetes5it,nowdtaehpseeernvedacleounnatsttmerdueatchtseurdoruecmghhaernigntes- MgAlISiPSClArKCaSTciVCrMFneCNoi 0000..12..2355 performed by two scintillatorsand silicontracker 0.1 0.05 detectors. A charge resolution for proton events 00 5 10 15 20 25 slightly better than0.17chargeunits is achieved. Zrec RICH Z(scint&std) Thechargeresolutionasfunctionofthe charge Z of the particle followsa curvethat corresponds to the error propagation on Z which can be ex- Figure 5. Charge peaks distribution measured pressed as: with the RICH prototype using a n=1.05 aero- gel radiator, 2.5cm thick. Individual peaks are σ(Z)= 1 1+σp2e +Z2 ∆N 2 . (4) identified up to Z∼28 (left). Charge resolution 2s N0 (cid:18) N (cid:19)syst versus particle Z for the same aerogel radiator. The curve gives the expected value estimated as Thisexpressiondescribesthetwodistincttypesof explained in the text (right). uncertaintiesthataffectZ measurement: the sta- tistical and the systematic. The statistical term rithms on real data taken with in-beam tests at is independent of the nuclei charge and depends CERN, in October 2003 was done. The detec- essentially on the amount of Cˇerenkov signal de- tor design was validated and a refractive index tected for singly charged particles (N0 ∼ 14.7) 1.05 aerogel was chosen for the radiator, fulfill- and on the resolution of the single photoelec- ing both the demand for a large light yield and a tron peak (σ ). The systematic uncertainty pe goodvelocityresolution. The RICHdetetector is scaleswithZ,dominatesforhigherchargesandis being constructedanditsassemblingtothe AMS around1%. Itappearsduetonon-uniformitiesat complete setup is foreseen for 2008. the radiator level or at the photon detection. In order to keep the systematic uncertainties below REFERENCES 1%,theaerogeltilethickness,therefractiveindex and the clarity should not have a spread greater 1. S. P. Ahlen et al., Nucl. Instrum. Methods A than 0.25mm, 10−4 and 5%, respectively; at the 350, 34 (1994). detection level a precise knowledge (< 5% level) 2. T.Ypsilantis and J.Seguinot, Nucl. Instrum. of the single unit cell photo-detection efficiency Methods A 343, 30 (1994). and gains is required. 3. F.Bara˜o,Nucl.Instrum.MethodsA502,510 Runs with a mirror prototype were also per- (2003). formed and its reflectivity was derived from data 4. P. Aguayo, L.Arruda et al., NIM A560, 291 analysis. Theobtainedvalueisingoodagreement (2006). with the manufacturer value. 4. Conclusions AMS-02 will be equipped with a proximity focusing RICH enabling velocity measurements with a resolution of about 0.1% and extending the charge measurements up to the iron element. Velocity reconstructionis made with a likelihood method. Charge reconstruction is made in an event-by-event basis. Evaluation of both algo-