A time-of-flight system for the external target facility

Chinese Physics C Vo1.37，No.5(2013)056002A time-of-flight system for the external target facilityZHANG Xue-Heng(章学恒) ； ) Yu Yu-Hong(余玉洪) SUN Zhi-Yu(孙志宇) MAO Rui-Shi(-瑞士)WANG Shi-Tao(王世陶) ZHOU Yong( j) ，。 5(AN Duo(I-3铎) LIU Long-Xiang(lJ龙翔) ，。

1 Institute of Modern Physics，Chinese Academy of Sciences，Lanzhou 730000，China。Graduate University of Chinese Academy of Sciences， Beijing 100049，ChinaAbstract： A time-of-flight system with a plastic scintillator coupled to photomultipliers is developed for the externaltarget facility(ETF).This system can satisfy the requirement of an ultrahigh vacuum( 10。mbar)，a high countingrate(-10。particles per second)and a magnetic field environment.In the beam test experiment，a total time resolutionof 580 ps FW HM was obtained for the whole system.and nuclei with a mass of up to 80 could be identified usingthis system。

K ey words： ETF，RIBLL2，TOF time resolutionPACS：29.40.Mc DOI：10.1088/1674-l137/37/5/0560021 IntroductionAn external target facility fETF 1 is constructeddownstream of the second radioactive ion beam line fRI-BLL21 in the Heavy Ion Research Facility in Lanzhou(HIRFL)1，21.It consists of a series of sub-detectorsystems，such as the bal1 f31，TOF wal f41，neutronwall51，and MWPCs61.Using beams that can accel-crate nuclei up to 23 U with a kinetic energy of severalhundreds of MeV to GeV using the main Cooling Stor-age Ring(CSRm)2I，research on the structure of exoticnuclei and the equation of state of dense nuclear mattercan be carried out at the ETF。

Like the FRS fGSI FRagment Separator)，which de-livers radioactive ion beams fRIBs1 to the ALADIN-LAND setup for decay and reaction studies at GSI781，the first half of RIBLL2 can be used to produceand separate the intcrested RIBs for ETF.The RIBscan be identified by combining the time-of-flight门 0F1the energy deposit△E and the magnetic rigidity Bp。

which is widely used by all projectile fragmentation typeseparators. Therefore，a TOF system，which providesexperimental trigger and particle identifcation，shouldbe developed for the ETF.This system should satisfVthe following stringent requirements：f 1)the start de-tector should have a large area f 100 mm×100 mm)，beused in the ultrahigh vacuum (r，410 mb)，and shouldwithstand a high counting rate r i00 particles per sec-ond1；(21 the stop detector should be operated in themagnetic field environment(<0.01 T)；and(3)the T0Fsystem should have a good mass resolution power(upto 60)。

In this paper，we would like to report the design ofthis TOF system.The measurements results will also bepresented。

2 The design of the TOF systemThe 1ayout of the first half of the RIBLL2 and ETFis shown in Fig.1.The primary beam from CSRm hitsa target at the F0 cave. The RIBs produced by pro-jectile fragmentation are separated and purified by thecombined Bp-△E-BP method I 9 l，and then delivered tothe ETF for experimental studies.Particle identificationis implemented before the second target. To obtain agood mass resolution，the flight path should be as longas possible，so we place the start and stop detectors atthe F1 cave and upstream of the second target，respec-tively.The total length of the flight path is about 26 m。

2.1 The start detectorThe start detector is designed based on a plasticscintilator from the manufacturer Eljen (EJ200)J 101。

Hamamatsu R71 1 1 photomultipliers l 11 l are used to readout the signals from both ends of the scintillator。

According to the beam optics calculations I 12 l，thebeam spots have maximum position dispersion at the F1cave.The beam size is about 100 mm×100 mm whenthe slits are fully opened.The plastic scintillator，whichhas the feature of no size restrictions，can be easily madeinto a large area detector to meet the size requirement。

Received 26 June 2012Supported by National Natural Science Foundation of China(11205210)11 E-mail：zhxh###impcas.ac.anQ 2013 Chinese Physical Society and the Institute of High Energy Physics of the Chinese Academy of Sciences and the Institute ofModern Physics of the Chinese Academy of Sciences and IOP Publishing Ltd056002-1Chinese Physics C Vo1.37，No.5(2013)056002QRIBLL2 ETFFig.1. Top view ofthe first half ofthe RIBLL2 and ETFFig.2.(color online)Schematic layout of the start detectorThe start detector will be used in ultrahigh vacuumfat a pressure of-10-0 mb1. This calls for very lowoutgassing materials in the mechanical support，opticalcoupling，a voltage divider and cables.A specific stain-less steel frame has been designed and its schematic lay-out is shown in Fig.2.The whole frame is fixed on aflange connected with the F1 chamber.The scintillatoris placed in the center of the beam line.and photomul-tipliers are fixed in the cylinders at both ends of thescintillator. Two 3 mm thick circular quartz ders to isolate the airand to transmit photons from the scintilator to the pho-tomultipliers.The couplings via optical silicon pads anda vacuum are chosen between the photomultipliers andquartz windows，and between the quartz windows andscintilator，respectively.W ith this frame al the mate-rials except the scintillator are placed in the atmosphereto ensure that the start detector is less affected by theoutgassing rate。

To ensure the air tightness of the welded interfaces，the quartz windows can only be processed into a cir-cle with a maximum diameter of 35 mm due to theconstraints of the machining technology.The l ht col-lection emciency will be influenced by the difierencein sizes between the scintillator and the quartz win-dows.The l ht collection eficiency of two prototypeshas been simulated with GEANT413].Prototype I，shown in Fig.3(a)，consists of a scintilator of dimension100 mm ×100 mm x 3 mm and two vacuum-coupled lightguides。Prototype I，shown in Fig.3(b)，is a monoblockscintilator without fonr corners.The thickness of Propo-type II is 3 mm，and the widths of the cut and uncutparts are 30 mm and 100 mm.respectively. The total length of the two prototypes，which is equal to thedistance of two quartz windows.is 500 mm.In the sim-ulation both the scintillator and 1ight guides are coveredby thin aluminum foil，the refraction of the scintilatorand light guide is 1.58.and the vacuum refraction is 1.0。

A primary beam of 500 MeV/u 0 C with a spot size of80 mm in diameter hits the center of the scintillator.Thephotons wil be produced by the scintilation process dueto the electromagnetic energy lOSS of 12 C.The light col-lection emciency is calculated by the ratio between thephotons transmitted to the end fⅣ1)and the total pho-tons(Ⅳ0).Fig.3(c)shows the light colection eficiencyas a function of the length of the vacuum layers betweenthe scintillator and light guides for Prototype I.andFig.3(d)shows the light colection eficiency change withthe length of the cut corners for Prototype II.It can beseen that the light colection efficiency slightly increaseswith a decrease in the length of the cut corners.and thevalue is similar to the emciency at the vacuum lengthof 1 mm.Considering the accuracy of machining，Pro-totype II is finally adopted with a cut corner 1ength of6 mm.The scintillator is covered b y one layer of 6 mthick aluminum foil，and the length between the scintil-lator and the quartz window is about 1 mm。

W hen the start detector is running at high rate，the056002.2Chinese Physics C Vo1.37，No.5(2013)056002O O60.040 O2O0 04O 030.020 01O 1 2length ofvacumm layer/mmO 5O 1OO l5O 200length/ramFig.3. (color online)Light collection eficiency simulated with GEANT4.(a)and(b)are a schematic diagram ofprototypes I and II，(c)is the eficiency as a function of the length of the vacuum layer of Prototype I，and(d)shows the eficiency as a function of the length of cut corners of Prototype II。

anode current of the photomultipliers increases dramat-icaly.and the non-linearity efect appears in such a wavthat the voltage drop in the last amplification stages be-comes significant and causes gain instability due to in-suficient bleeder current supply14.T0 keep the pho-tomultiplier gain stable against the incoming particleflux fup to-100 particles per second)，three types ofvoltage dividers are designed，as shown in Fig.4(a)，fb)and(c)，respectively.Type I is a traditional volt-age divider.Type II is an improved voltage divider witha large-scale regulated power supply network，and TypeIII adopts the additional power supplies at the last threedynodes based on Type II. The divider ratios of thethree types of voltage dividers have been adusted to ob-rain good gain f 15 f.In order to evaluate the rate capa-bility of our prototypes，a test was performed.A greenLED(1ight emitting diode)source[16，which was drivenby the square signa1 with an amplitude of 5 V.width of40 as，leading edge of 13 ns and trailing edge of 12.1 as，was used to simulate an increasing particle rate.Usingan optical fiber.the light output was delivered to thecenter of the R71 1 1 photomultiplier with a high voltageof-900 V.The amplitude of the photomultiplier outputsignal directly measured by an osciloscope was studiedas a function of the LED frequency. The results areshown in Fig.4(d).When the frequency is increased。

the amplitude remains constant up to a maximum value - 10 Hz for the three voltage dividers.and then the am-plitude increases first and then reduces.The reason forthe amplitude increase can be explained as the overlap ofthe LED electroluminescence.From Fig.4(d)we can seethat the Type III voltage divider can still keep the pho-tomultiplier gain stable at about i0 Hz.So the TypeIII voltage divider is used in our detector。

The start detector has been tested with a 60Cogamma ray source，and the coincidence signals of bothends of the detector measured with an oscilloscope areshown in Fig.5(a)，(b)and(c)when the source was putat the center，left and right positions of the scintillator，respectively.It can be seen that the time and amplitudesof two coincidence signals are similar when the source isput at the center.W hen the source is put at one side，the signal from this side is earlier and bigger than theone from the other side。

2.2 The stop detectorThe stop detector will be placed near the second tar-get.This is a focus point，and the beam size is smallerthan the one on the start detector. So a 50 mm×50 mm x 1 mm BC408 scintilator sheet[1 7coupled totwo photomultipliers at both ends with optical glue is056002-3Chinese Physics C Vo1.37，No.5(2013)056002used for the stop detector.Since the position of the stopdetector i8 near the big dipole of the ETF.the leakagemagnetic flux has a serious influence on the signa1 ampli-tude ofthe photomultiplier.Therefore the R7111 photo-multipliers have been replaced by fine mesh R7761 pho-tomultipliers I 11 I，which can be operated even in strongmagnetic fields over 1.0 T18，19].A large-scale regu-lated power supply network and additional power sup-plies for the last three dynodes were also applied for theR7761 voltage divider。

(a)(b)(c)> ( ) 善5o001O 1O 10 10 1Ofequency/ttzFig.4. (color online)Three types of voltage di-viders for the R71 1 1 photomultiplier shown in(a)，(b)and(c)，respectively.The signal amplitudesas a function of LED frequency are shown in(d)。

3 The beam testThe T0F system was instaled and used experimen-taly.A primary beam of 400 MeV/u 1 C was extractedfrom CS；Rm and implanted into a 15.1 mm thick 0Betarget at the F0 cave of RIBLL2. The fragmentationproducts were separated and identified by the first halfof RIBLL2.In the experiment.the magnetic rigidity wasset to 4.57 Tm，which was calculated with 0C nuclei.Theslits located at the F1 cave were opened at土10 mm.andthe degrader was not used。

% --c)i ～ ～I l H 20 0 ns；V 50 0mV- 麟； H 粉right ÷Fig.5. (color online)Coincidence signals of thestart detector measured with a。。Co gamma raysource.(a)，(b)and(c)show the results when thesource was put at the center，left and right of thescintilator，respectively。

The energy 1OSS was measured by a 140 LLm thick45 mm×45 mm silicon detector.Fig.6fb1 shows a TOFspectrum only for the 0C nuclei chosen in Fig.6(a1。

The root-mean-square width of this distribution is deter-mined to be 2.92.Considering the TDC channel widthof 84.7 ps I201，a time resolution of 580 ps ful widthat half maximum，FWHM1 was obtained for the wholesystem. This value includes al1 the factors that causetime dispersion，such as fragment energy dispersion andelectronics。

The mass resolution power of the TOF system isgiven by the folowing relationships[2 1]，R： ， (1)where A and T are the mass and time-oflight of thenuclei，and AA and AT are the widths of the mass peakand time-of-fiight peak.respectively.W ith the RIBLL2maximum magnetic rigidity of 10.64 Tm[2]，the highestvelocity tim is about 0.915 for the ion with the massto charge ratio of 1.5.From Eq.f11 we obtain the massresolution power R≤160，and nuclei with a mass of upto 80 could be identified using this TOF system。

056002-4Chinese Physics C Vo1.37，No.5(2013)0560020200 -100 0 100TOF，channel0- 180 - 160 -140TOF，ch nelFig.6. (color online)The particle identifcation spectrum for c。Be at 400 MeV/u(a).The TOF spectrum of。C is shown in(b)，and the red line denotes the fitting results with Gaussian function。

4 SummaryIn this paper，a TOF system with a plastic scintil-lator coupled to photomultipliers was developed for theETF.The start and stop detectors are installed at theF 1 cave of RIBLL2 and upstream of the second target。

respectively.The distance of this TOF system is about26 m.The special design of the frame makes the startdetector work in ultrahigh vacuum.To obtain high lightcollection e币ciency.a monoblock scintilator withoutfour corners was adopted by the simulations.Photomul-tipliers for a high magnetic environment were used inthe stop detector to resist the leakage magnetic flux ofthe big dipole.The voltage dividers for both detectorswere designed a8 a large-scale regulated power supplynetwork with additional power supplies at the last threedynodes，which makes the photomultiplier gain stable atincoming particle flux up to 10。particles per second。

This TOF system was used experimentaly and a totaltime resolution of 580 Ds FW HM was obtained.Nucleiwith a mass of up to 80 could be identified using thissystem 。

The authors gratefuly acknowledge Meng Jun-HouShen-Jun and Zhao Yu-Gang for their support on thevacuuyt technique。

Referencesl XIA J W .ⅥrANG Y F.RA0 Y N et a1.HIRFL Status andHIRFL-CSR Proect in Lanzhou.Proceeding of the First AManPaticle Accelerators Conference.Japan：KEK，1998.3422 xIA JW .ZHAN stum .M ethods。

A，2002，488：113 YEI K.Design and Simulation Studies of CsI(TI1 ScintilatorDetector Array for External Target Facility at HIRFL-CSR(Ph.D.Thesis).Lanzhou：Instumitute of Modern Physics，CAS，2010(in Chinese)4 YU Y H.Development and Construction of Fast Plastic Scintil-lator Detector Array for External Target Experiment at CSRmfPh.D.Thesis).Lanzhou：Instumitute of Modern Physics，CAS，2009(in Chinese15 XU H G.Design and Construction of Neutron W all for ExternalTarget Experiment at CSRm of HIRFL-CSR (Ph.D.Thesis1。

Lanzhou：Instumitute of Modern Physics，CAS，2006 fin Chi-nese)6 LU C G.Study of RIBLL2 External nⅡget Gaseous TrackDetector (Ph.D.Thesis).Lanzhou： Instumitute of ModernPhysics，CAS，2011(in Chinese)7 Blaich T H，Elze T W Emling H et a1.Nuc1.Instrum .M ethodsA，1992，314：1368 Geisse1 H，Armbruster P，Behr K H et a1.Nuc1.Instrum.Meth-ods B，1992，70：2869 Schmidt K H，Hanelt E et a1.Nuc1.Instrum.Methods A，1987，260：28710 http：//ories/Data.Sheets/Plastic-Scintmat0rs/EJ2O0%20data%20sheet.pdf11 http：//M T，YANG X D，XIA J W et a1.HEP & NP，2001，25f5)：443 fin Chinese)13 http：//geant4.cern.ch14 LE0 W R.Techniques for Nuclear and Particle Physics Exper-iments.2 Revised Edition.Berlin：Springer.199415 DING X L，WANG J C，QI H R et a1.Nuclear Electronics&Detection Technology.2005，25(4)：39216 http：///Upload/PDF/20110804135342887552.pdf17 http：//www.phys.uf1.edu/courses/phy4803L/group-I/muon/bicron-bc400-416.pdfl8 Liima T，Amam i M ，Adachi I et a1.Nuc1.Instrum.MethodsA.1997.387：6419 Bonesini M .Strati F.Baccaglion et a1.Nuc1.Instrum.M ethodsA.2006.567：20020 http：//Prod.jspparentll&idmod3521 SUN Z Y.ZHAN W L.GU0 Z Y et a1.Nuc1.Instrum.MethodsA，2003，503：496056002-5如鲁0uI3昌高舌