Some stud i es of the app l i ca t i ons of CsI photoca thodes i n gaseous de tec tors

Transcription

1 Nuc l ear Inst ruments and Me thods i n Phys i cs Research A 333 (1993) Nor th-ho l l and NUCLEAR I NSTRUMENTS & METHODS IN PHYSICS RESEARCH Sec t i on A Some stud i es of the app l i ca t i ons of CsI photoca thodes i n gaseous de tec tors G Charpak a, 1 Gaudean a, Y G i oma tar i s and D Stuar t d PPE D i v i s i on, CERN, Geneva, Sw i t zer l and "Lausanne Un i vers i ty, Sw i t zer l and Fermi Nat iona l Acce l erator Laboratory (FNAL), BT, USA d Un iversi ty of Ca l i forn ia, Davis, CA, USA b, V Peskov `, D Sc i gock i a, F Sau l i a Rece i ved 4 March 1993 Ces i um- i od i de photocathodes have been invest igated in comb i nat i on w i th gaseous detectors The quantum ef f ic iency of a semi t ransparent CsI photocathode has been measured in the UV range and found to be 10% at 140 nm We repor t exper imenta l resul ts obta ined w i th such photocathodes coup l ed to BaF, and KMgF3 scint i l lators and read out by para l le l -plate gaseous de tec tors 1 Int roduc t i on : Re f l ec t i ve Cs l photoca thodes have been shown to work w i th amp l i fy i ng gaseous de tec tors [1,2] A ma j or step i n the i r potent i a l i mpor tance comes f rom two observa t i ons [3] i ) The quantum e f f i c i ency of re f l ec t i ve CsI photoca thodes, when hand l ed under c l ean cond i - t ions, is s igni f i cant l y h i gher than the one observed in prev i ous exper i ments i i ) Adsorbed vapours of tet rak i s(d i me thy l ami ne)e thy l ene (TMAE) enhance the quantum e f f i c i ency and, i n add i t i on, protec t the CsI i n the case of contac t w i th a i r Th i s opens up a new way to read out l i ght f rom sc i nt i l l a tors, and Cherenkov l i ght i n the case of RICH counters and hadron b l i nd de tectors, and has t r i ggered ac t i ve research [4-9] We have per formed severa l measurement s connec ted ma i n l y w i th three app l i ca t i ons : - Readou t of KMgF 3 crysta l s for a pre-preshower de tec tor [10] - Readou t of BaFZ crysta l s for a preshower count er [10] and for a ca l or i me ter [11] We have l i mi ted our e f for ts to the comb i na t i on of CsI photoca thodes w i th para l l e l -p l a te gaseous de tectors because of the i r good t i me reso lut ion (< 1 ns) [12,13] We have stud i ed the mer i ts of re f l ec t i ve and semi - t ransparent photoca thodes, and have comp l e ted age i ng measurements i n v i ew of the poss i b l e app l i ca t i ons of the de tec tors to h i gh- l umi nos i ty co l l i ders 2 Preshower count er Recent l y, i t was repor ted tha t KMgF3 emi ts a t even shor ter wave l engths than BaFZ (be tween 140 and 220 nm w i th a decay t i me of about 2 ns) [14] Such a spec t rum over l aps very we l l w i th the quantum e f f i - c i ency of some sol id photoca thodes (see for examp l e re f [15] ) Our l a test tests show however tha t the best photoca thode for the readout of the KMgF3 sc int i l lat i on l i ght i s st i l l CsI A th i n KMgF3 crysta l coup l ed to a para l l e l -p l a te ava l anche chamber (PPAC) w i th a CsI photoca thode can be used as a f i rst -step detector in f ront of a ca l or i me ter i n order to d i st i ngu i sh e l ec t rons f rom gammas A th in crysta l of the KMgF3 wou l d de tec t energy depos i ted by charged par t i c l es, but be near l y t ransparent to gammas, a l l ow i ng d i f ferent i a t i on be tween showers crea ted by gammas and e l ec t rons I ts advant ages are the rad i at i on hardness, and the subnanosecond t i me reso lut ion prov i ded by the KMgF3 - PPAC comb i na t i on We have invest igated the response of 3 mm and 10 mm th i ck KMgF3 crysta l s us i ng a photomu l t i p l i er and a PPAC w i th a CsI photoca thode 21 Photomu l t i p l i er measurements We used e l ect rons f rom a " Sr source to compare the response of the KMgF3 crysta l and a 20 x 20 x 50 mm 3 BaFZ crysta l The source conta i ns a 9 Y compo /93/$06 00 C E l sev i er Sc i ence Pub l i shers B V Al l r i ghts reserved

2 392 G Charpak et al / Cs l photocathodes in gaseous detectors 241 Am PM SS r ô %/ / / / I / /%I l i " " I z mm Mesh a EM ADC counts F i g 1 Typ i ca l ß spec t rum f rom 90 Sr w i th a max i mum energy of 2 2 MeV 0 obta ined w i th a photomul t ipl i er and a KMgF3 crysta l nent wh i ch emi ts e l ec t rons w i th a max i mum energy of 2 2 MeV The sc i nt i l l a t i on l i ght of the crysta l s i s detec ted by a Hammama tsu R1460 photomu l t i p l i er wh i ch i s sens i t i ve f rom 110 to 300 nm A typ i ca l spec t rum pu l se he i ght i s shown i n f i g 1 By compar i ng the l i ght output at the 2 2 MeV edge f rom BaF 2 and f rom KMgF3, we deduce that the l i ght output of KMgF3 i s about 30% of that of the fast component of BaF2 Th i s cor responds to rough l y 100 photoe l ec t rons per MeV, w i th good opt i ca l contac t be tween the crysta l and the w i ndow 22 PPAC measurement 22 1 Se tup A sma l l PPAC was bu i l t as shown i n f i g 2 I t cons i sts of a KMgF3 sc i nt i l l a tor, an anode mesh, and a copper pad ca thode 4 mm away A CsI l ayer 200 or 500 nm th i ck was depos i ted e i ther on the copper pads or d i rec t l y on the KMgF3 sc i nt i l l a tor, covered by a 20 nm l ayer of W Th i s i s on l y a s l i ght mod i f i ca t i on of the se tup used in ref [11] In the present work we used a mi xture of He w i th a few percent of C2H6 or CH 4 a t a tmospher i c pressure The gas and e l ec t ron i c ga i ns were ca l i bra ted w i th a 55 Fe source Th i s a l l owed us to measure d i rec t l y the pr i mary charge (number of photoe l ec t rons) of the s i gna l s f rom the crysta l Resu l ts Al though many stud i es were made on re f l ec t i ve CsI photoca thodes, no systema t i c study was car r i ed out to compare such a photoca thode w i th a semi - t ransparent one On the other hand, for some app l i cat i ons the semi - t ransparent photoca thode may have advant ages For examp l e, in the case of crysta l readout i t can be expec ted tha t more sc i nt i l l a t i on l i ght w i l l pene t ra te a semi - t ransparent Cs I photoca thode because the di f fer u i d 02 cu I Ima tor 13 ^ Sr F ig 2 D i agram of a PPAC wi th a Cs l photocathode used to measure KMgF3 sc i nt i l l a t i on ence i n re f rac t i ve i ndex i s sma l l er be tween the crysta l and the CsI than CsI and gas To study th i s we d i d some measurement s w i th the semi - t ransparent Cs I photoca thode Some resu l ts of the quantum e f f i c i ency measurement s are present ed in f ig 3 I t can be conc l uded f rom th i s f i gure tha t the quantum e f f i c i ency of the re f l ec t i ve CsI photoca thode i s cons i derab l y h i gher than tha t of the semi - t ransparent one 0 0 semi t ransparent re f l ec t i ve 0 0, Wave l ength (nm) F i g 3 Quantum e f f i c i ency of a re f l ec t i ve and a semi - t ransparent Cst photocathode 500 mm thi ck

3 G Charpak et at / Cs1photocathodes in gaseous detectors 393 The spec t rum obta i ned w i th "Sr f rom the PPAC w i th the re f l ec t i ve CsI photoca thode and a 3 mm th i ck crysta l is shown in f ig 4 Not e that i t does not exhibi t the shape shown in f ig 1, but is rather exponent i a l l y peaked towards zero We est i mate that the 2 2 MeV ß- edge cor responds to rough l y 40 photoe l ec t rons of pr i mary charge In the case of a 10 mm crysta l the resu l t was 20 photoe l ec t rons The d i f ference may be par t l y due to crysta l qua l i ty When compar i ng w i th a photomu l t i p l i er (100 e - per MeV) we have to take i nto account the absorpt i on in the gr id and the interna l re f l ec t i on l osses due to the absence of opt i ca l contac t Assumi ng an energy l oss of 3 MeV/ cm for a mi n i - mum i on i z i ng par t i c l e (MIP) i n KMgF3, we expec t a t l east 20 photoe l ec t rons of pr i mary charge f rom the passage of a MIP through a 3 mm th i ck crysta l Th i s can be eas i l y de tec ted us i ng a PPAC w i th a CsI photoca thode The rad i at i on l ength of KMgF3 is 7 5 cm, wh i ch shou l d g i ve an e-y se l ec t i on of about 7% Resu l ts obta i ned w i th the semi - t ransparent CsI photoca thode were much more modest Al though we expec ted be t ter sc i nt i l l a t i on l i ght co l l ec t i on, on l y a few photoe l ec t rons were crea ted f rom the photoca thode w i th the depos i ted energy of 2 2 MeV We conc l ude f rom these measurement s tha t a re f l ec t i ve Cs I photoca thode is super i or to a semi - t ransparent one For this reason we on l y used ref l ect i ve photoca thodes in the l a test exper i ments 3 BaFZ preshower count er A preshower counter i n f ront of a ca l or i me ter i s needed for h i gh-prec i s i on pos i t i on measurements of the showers and to i ncrease the e / r r re j ec t i on capab i l - i ty Such a dev i ce may be necessary not on l y for the BaF2 ca l or i me ter, descr i bed l a ter i n th i s paper, but a l so for many of the detector cand i dates for the ca l or i met ry to be used in exper i ments at the future h i gh- l umi nos i ty co l l i ders A preshower counter made f rom 2 5 rad i a t i on l engths of BaF2 crysta l w i th a PPAC readout and a Cs I photoca thode was suggest ed in ref [11] M i n i mum i on i z i ng par t i c l es w i l l depos i t 32 MeV i n the de tec tor ; an e l ec t romagne t i c shower w i l l depos i t a f ew hundred MeV For the typi ca l photoca thode e f f i c i ency we expec t to have a few photoe l ec t rons per MeV depos i t ed in the BaF Z In this case the ga in needed i n the PPAC i s about The pos i t i on of the shower, or of the MIPs, can be measured on the pads by the cent re-of -grav i ty me thod, and the d i f fer - ence i n the energy depos i ted w i l l make i t poss i b l e to measure the e / r r separa t i on i n the preshower counter The ma i n advantages of the BaF2 preshower counter are : h i gh rad i a t i on res i stance (up to Mrad), good t i me reso l ut i on ( - 1 ns), and the poss i b i l i ty of measur i ng the depos i t ed energy In order to check these advantages and espec i a l l y measure the space reso lut ion we made and tested i n the l abora tory a sma l l prototype of the BaF2 preshower count er 31 Const ruc t i on The insta l l at ion used to test the BaFZ preshower count er is shown in f ig 5 The detector used a BaF2 crysta l of 12 cm d i ame t er and I cm th i ckness as the w i ndow and the sc i nt i l l a tor The ac t i ve s i ze of the ca thode p l ane and of the w i re p l ane i s 200 X 200 mm 2, w i th 1600 pads At h i gh ga i n and w i th a h i gh f l ux of par t i c l es the PPAC may suf fer f rom sparks [16] The prob l em has been comp l e te l y e l i mi na ted i n th i s work 600 L I É z 200 0, ~~" -vn ADC co i nts F i g 4 Charge spec t rum f rom a PPAC w i th KMgF3 i r rad i a ted by a 90 Sr source The 5 9 kev f rom 55 Fe, absorbed i n the dr i f t space, was used to ca l i bra te the ga i n i n the gap

4 39 4 C Charpak et al / CsI photocathodes in gaseous detectors by l oad i ng each anode w i re w i th a 10 Mf1 res i stor The res i stor l i mi ts the cur rent and makes the t rans i t i on of a st reamer i nto a spark a l most i mposs i b l e A h i gh vo l tage i s app l i ed on the anode w i res, and the s igna l is read out on each pad w i th a charge amp l i f i er as descr i bed above The PPAC worked w i th a gas mi xture composed of He + 1% CH, Z Tor r of e thy l fer rocene (EF) a t a tmospher i c pressure The ref l ect i ve photoca thode is depos i t ed on the pads by vacuum evapora t i on Immed i a te l y a f ter the 500 r im thi ck Cs I photoca thode was depos i ted, the ca thode p l ane was t ransfer red to a box where i t was pumped be fore be i ng put i nto contac t w i th the EF vapour The photoca thode was in contact w i th a i r for l ess than 2 mi n, and a f ter a few hours of gas f l ush i ng w i th EF vapour + argon, an adsorbed l ayer of EF had formed and protec ted the photoca thode f rom the react ion w i th a i r Then the photoca thode was mounted i n the de tector i n a i r The sc int i l l at ion of the BaF2 crysta l was produced by a pu l sed ( - 10 ns) X- ray source w i th the max i mum energy of photons - 60 kev Th i s source cou l d depos i t an equ i va l ent energy of up to 100 GeV per pu l se I f necessary, thi s energy can be reduced by a set of l ead a t tenua tors In most measurements a l ead co l l i ma tor (a ho l e of 500 R i m) was used to s i mu l a te a s i ng l e-par t i c l e cross i ng In some tests a pu l sed (15 ns) vacuum ul t rav iol et (VUV) source was a l so used to ext rac t e l ec t rons f rom the CsI photoca thode r i r r r r~r r i r r i r F ig 6 PPAC signal f rom the preshower detector at 26 kv (a) The hor i zonta l sca le i s 5 Ws per division and the ver t ical scale is 50 mv per division (b) The hor i zonta l sca le i s 50 ris per division and the ver t ical sca le i s 100 mv per division 4 Resu l ts F i g 6a shows the s i gna l obta i ned f rom the co l l i - ma ted pu l sed X- ray genera tor a t an app l i ed vo l tage of Pads Lead co l l i ma tor Pu l sed X-Ray genera tor BaF 2 Amp l i f l ers F ig 5 Exper imenta l setup for the study of the BaFZ preshower prototype 2 6 kv The e l ec t ron par t of the s i gna l i s shown be t ter i n f i g 6b Our measurements were made star t i ng f rom the i on i za t i on chamber mode (ga i n 1) and up to ga i n 10` ' The typi ca l dependence of the ga in on the vo l tage is shown in f ig 7 The de tec tor can work in a stab l e fash i on a t a h i gh ga i n of 10 5 F i g 8a shows the prof i l e of the co l l i ma ted pu l sed X- ray genera tor s i gna l on a l ine of pads Each b i n represents one pad (10 mm) The ver t i ca l ax i s i s the co l l ec ted charge on the cor respond i ng ADC F i g 8b i s a s i mi l ar measurement i n the same geome t ry, but w i th a pu l sed UV source wh i ch produces a spot of l i ght of a few mm d i ame ter on one pad The spread due to the sc i nt i l l a t i on l i ght i n the crysta l i s c l ear l y shown by compar i ng the two f i gures as we l l as i s the neg l i g i b l e cont r i but i on due to the ava l anche s i ze in the PPAC I t can be seen that the pos i t i on of the ava l anches can be de f i ned by a cent reof -grav i ty me thod The un i formi ty of the s i gna l between the pads was measured to be bet ter than 15% The energy reso l ut i on of the de tec tor a t a gas ga i n of 10 5 i s shown in f ig 9 In this test about 10 photoe l ect rons were in j ected by the at tenuated l ight beam f rom the VUV pu l sed l amp An energy reso lut ion of 70% FWHM was ach i eved, wh i ch conf i rms the resu l ts

5 G Charpak et al / CsI photocathodes to gaseous detectors 395 obta i ned i n re f [11] Al though these pre l i mi nary resul ts are encourag i ng, the prototype has to be tested in a beam to measure prec i se l y the pos i t i on reso l ut i on for e l ec t rons and gamma photons, as we l l as the e l ec t ronp i on re j ec t i on capab i l i ty 1000 '~ a) 5 Test of a prototype of a BaF2 ca l or i me ter modu l e A BaF2 ca l or i me ter i s be i ng d i scussed as a poss i b l e cand i da t e for the LHC and SSC exper i ments (see for examp l e ref [171) A poss ib l e l ayout of a BaF2 ca l or i me ter for the SSC and LHC exper i ments i s descr i bed i n re fs [17,18] I t cons i sts of an ar ray of BaF2 crysta l s cover i ng the acceptance I q I < 3 8 Photod i - odes were proposed for the readout of the fast component ( - 1 ns) of the sc i nt i l l a t i on l i ght w i th A < 220 nm We deve l oped another approach, based on the PPAC w i th a so l i d photoca thode The ma i n advantage of thi s type of readout compared to that of photod i - odes is that the PPAC i tsel f is very cheap and a l l ows gas ga i n A f i rst prototype of th i s type of readout was tested by the authors of re f [18] The energy reso l ut i on was measured as 3%/ v~e a t ga i n 1 [191 and 7% a t ga i n 10 5 [11] In this sect i on we present some resul ts of tests of one ca l or i me ter modu l e w i th PPAC readout of the fast BaF2 sc i nt i l l a t i on l i ght Exper i menta l se tup The prototype of the BaF2 ca l or i me ter modu l e i s present ed in f ig 10 I t cons i sts of a BaF2 crysta l F i g 8 (a) Prof i l e on the pads for the co l l i ma ted pu l sed X- ray generator Each l ine on the hor i zonta l axis represents a pad number and the ver t ical axis represents the charge col l ected on the ADC (b) A simi l ar pi cture for a pul sed UV l amp signal w i th the same geomet ry The si ze of the l ight spot on the pad 1s a few mm m d i ame ter C 10 He + 0 8% CH 4 I i i i H V (kv) F ig 7 Ga i n measurements of the preshower detector w i th a gas mi xture composed of He+0 8% CH Tor r of EF vapour mm l ong and 55 mm i n d i ame ter d i rec t l y coup l ed to the PPAC The PPAC is bui l t f rom the anode mesh and the sta inl ess-stee l (ss) ca thode p l ane 70 mm in d i ame ter and 7 mm away The anode mesh was put a t 1 mm d i stance f rom the BaF2 crysta l i n order to avo i d poss i b l e charg i ng of the crysta l by the co l l i d i ng e l ect rons The sta i n l ess-stee l ca thode p l ane was covered by a 55 nm th i ck CsI l ayer w i th an EF adsorbed f i l m by a procedure s imi l ar to the one descr i bed above The pure Cs I was f i rst depos i t ed on the ca thode by the vacuum evapora t i on techn i que and mounted i ns i de the de tec tor i n a g l ove box f l ushed w i th argon to l i mi t the contact w i th a i r Then the EF adsorbed l ayer was formed by cont i nuous l y f l ush i ng the gas mi xture conta i n i ng EF vapour through the gaseous detector In th i s test the gaseous de tec tor was ma i n l y used i n i on i za t i on chamber mode w i th a gas mi xture composed of He + d i f ferent concent ra t i ons of CH, + _ 10-2

6 396 G Charpak et a l / Cs l photocathodes to gaseous detectors 225 F u 200 F ,25 i 0o F r 75 r so 25 i 1 1, L M~MMM b) CGLLECIEO CHARGE F i g 9 Pu l se d i st r i but i on obta i ned f rom the PPAC i r rad i a ted by the VUV source ext ract ing 10 photoe l ect rons per pul se f rom the CsI photoca thode Tor r of EF vapour ( - 50% of EF vapour pressure at room tempera ture) The ion i zat ion source was the pu l sed ( - 10 ns) X- ray genera tor ment i oned ear l i er (see f i g 5) Us i ng a set of l ead absorbers we cou l d decrease this energy The HV was app l i ed on the ca thode, and the s i gna l was read out on the gr i d w i th a fast charge amp l i f i er ( r i se t i me < 10 ns, 360 mv per pc) A photomu l t i p l i er was mounted d i rec t l y (see f i g 10) on the crysta l to a~- / / /0 " / F ig 11 Ioni zat ion chamber signal at I kv (a) The hor i zonta l sca le i s 1 ws per division and the ver t ical sca le 7s 20 mv per division (b) The hor i zonta l sca le is 50 ns per division and the ver t ical sca le is 20 mv per division mon i tor the depos i ted energy by measur i ng the i ntens i ty of sc i nt i l l a t i on l i ght w i th A < 220 nm E PPAC F ig 10 Exper i menta l setup for the study of the BaF2 prototype ca l or i me ter ce l l PM 52 Resu l ts A typ i ca l i on i za t i on chamber s i gna l obta i ned i n the mi xture He + I% CH 4 + EF (10-2 Tor r ) i s shown i n f i g 11 The r i se- t i me of the s i gna l i s about 200 ns The dr i f t ve l oc i ty i s about 50 ns /mm for an e l ec t r i c f i e l d of 2 5 kv/ cm For pure CH 4 and h i gher f i e l ds, the dr i f t ve l oc i ty reached about 10 ns /mm The s i gna l cor re l a t i on be tween the i on i za t i on chamber and the photomu l t i p l i er i s shown i n f i g 12 The average max i mum s igna l in the gaseous detector was about photoe l ec t rons wh i ch cor responds to a depos i t ed energy of about 100 GeV The energy resolut ion was 0 73% rms Ana l ys i s of the da ta, obta i ned w i th the l ead a t tenua tor, when the depos i t ed energy was decreased by a factor of 2, shows that the energy reso lut ion rema i ns prac t i ca l l y unchanged a t - 07 rms, i nd i ca t i ng tha t the constant term i s domi na t i ng Th i s qua l i tat i ve l y con-

7 G Charpak et al / Cs I photocathodes in gaseous detectors 397 :700 ra 12 Ent r i es 9B7 i 2oo ' F i g 12 Cor re l a t i on be tween the i on i za t i on chamber and the photomu l t i p l i er for the pu l sed X- ray genera tor s i gna l f i rms resu l ts obta i ned i n re f [19] : the energy reso l ut i on i s <3% E -1/2 [GeV] + 0 5% I t shou l d be not ed that the detector worked for more than three months w i th h i gh stab i l i ty us i ng the same photoca thode Cs I photoca thodes covered w i th TMAE or EF adsorbed l ayers were observed ear l i er to be more res i stant to a i r than pure Cs I l ayers [10, 18] We have ver i f i ed th i s phenomenon for EF w i th our detector The photoca thode was exposed to a i r for four hours The s i gna l decreased by on l y 9% Af ter f l ush i ng w i th gas for one day fu l l recovery was observed As was ment i oned in the i nt roduc t i on, one of the advant ages of th i s type of readout is the poss ibi l i ty to work w i th gas ga i n Th i s cou l d be use fu l for the ca l or i me ter des i gn where the crysta l i s segmented a l ong the shower axis, as was suggest ed in ref [10] Gas ga i n cou l d be useful in the par ts cor respond i ng to the beg i nn i ng and the end of the shower, where the depos i ted energy i s re l a t i ve l y l ow Our pre l i mi nary study shows that a l though w i th a ga in of about 10, energy reso l ut i on i s a l ready degraded, th i s w i l l not not i ceab l y a f fec t the energy reso l ut i on of the ca l or i me ter 6 Age i ng proper t i es of the CsI photoca thodes Af ter the observa t i on that the Cs I photoca thode cou l d be used successfu l l y for the readout of the BaF2 ca l or i me tey and the preshower counters, the cruc i a l po i nt i s the stab i l i ty and age i ng proper t i es of the photoca thode Concern i ng these quest i ons there is some cont roversy in the l i terature In some works [3,12,20] i t was found that the Cs I photoca thode i r rad i ated cont i nuous l y by the VUV l i ght s l ow one In the case of the fast component the quantum e f f i c i ency drops by a fac tor of e a t a co l l ec ted charge of about Qf nc/ cm2 ; i n the case of the s l ow component a t a charge Qs i < 5 mc/ cm2 In one work [211 i t was repor ted that the co l l ected charge for wh i ch the s l ow componen t degrades by a fac tor e i s approx i ma te l y Q,, < 100 mc/ cm2 The cont rad i ct i ons in the measured va l ues of Q, i cou l d be exp l a i ned by the resu l ts obta i ned i n re f [22] I t was f i rst shown i n th i s work tha t the degrada t i on i n the CsI quantum e f f i c i ency i s revers i b l e i n most cases, by keep i ng the photoca thode for some t i me w i thout VUV rad i a t i on or by warmi ng i t up to 90 C Th i s exp l a i ns why i n re f [21] Q,, was ach i eved 200 t i mes bet ter than in ref [3] : in ref [21] the quan tum ef f i - c i ency of the photoca thode was measured one hour l a ter, a f ter the VUV beam was shut down Tak i ng into account the observa t i on made in ref [22] we car r i ed out a study of the stab i l i ty of the CsI quantum e f f i c i ency The resu l ts, descr i bed i n re f [221, were conf i rmed both in vacuum and in a gas a tmosphere At l ow VUV f l ux (cor respond i ng to a cur rent <_ 1 na) and c l ean gas cond i t i ons (sta i n l ess-stee l tubes, ox i sorb f i l ter ) no i nd i ca t i on of the degrada t i on of quantum e f f i c i ency was observed up to the co l l ec ted charge 100 mc/ cm2 The same resu l ts were obta i ned a t a gas ga i n of These age i ng proper t i es are enough for record i ng the sc i nt i l l a t i on l i ght f rom the BaF2 cor - respond i ng to - 10 Mrad depos i ted energy Never thel ess, the rad i a t i on proper t i es of the CsI photoca thode i tse l f shou l d be stud i ed 7 Conc l us i on We have i nvest i ga ted the sc i nt i l l a t i on of a KMgF3 crysta l us i ng both a photomu l t i p l i er and a photosens i - t i ve PPAC w i th a CsI photoca thode We found tha t the photomu l t i p l i er s i gna l f rom KMgF3 i s 30% of the s i g- na l for the BaF2 fast component The PPAC sens i t i v i ty w i th BaF2 was about 20 photoe l ec t rons per MeV depos i ted energy, wh i ch w i l l a l l ow de tec t i on of MIPs us i ng a th i n crysta l Th i s shows the i nterest of put t i ng such a counter i n f ront of ca l or i me ters The advantage over other types of detectors is the t i me reso lut ion of the order of 1 ns toge ther w i th the rad i a t i on hardness In study i ng the coup l i ng of BaF2 to the PPAC we have made some progress in the const ruct i on of BaF2 h i gh-granu l ar i ty preshower de tec tors and e l ec t romagne t i c ca l or i me ters w i th proper t i es ma tch i ng the re-

8 398 G Charpak et al / Cs I photocathodes in gaseous detectors qu i rements of future col l iders, i e fast t iming, radi a- t i on hardness, and good granu l ar i ty and energy reso l u- t i on In add i t i on, the const ruc t i on techn i ques for the de tec tor are s i mp l e, ow i ng to the res i stance to a i r of the photoca thode wh i ch works at room t empera ture and a tmospher i c pressure, and no w i ndow i s needed be tween the sc i nt i l l a tor and the gaseous de tec tor There fore the pr i ce of such a readout is neg l ig ib l e compared to tha t us i ng other c l ass i ca l techn i ques, such as photod i odes For app l i ca t i on to h i gh- l umi nos i ty mach i nes, the age i ng proper t i es of the photoca thodes are of pr i me i mpor tance Our l a test resu l ts i nd i ca te tha t CsI i s not mod i f i ed by i r rad i a t i ons i n the range 100 mc/ em2 As a f i na l demonst ra t i on, a prototype of a l arge enough si ze has to be bui l t and tested in a beam env i ronment Acknow l edgement s Th i s work was per formed i n the f rame of the LAA pro j ec t and we thank Prof A Z i ch i ch i for h i s suppor t and d i scuss i ons, and R Bouc l i er, I Croty and G M i l l i on for the i r techn i ca l suppor t Re ferences [1] G Charpak et al, prepr int CERN EP/ (1989) [2] V Daugendor f e t a l, Nuc l Inst r and Me th A289 (1990) 322 [3] J Segumot et al, Nuc l Inst r and Me th A297 (1990) 133 [4] B Hoene i sen e t a l, Nuc l Inst r and Me th A302 (1991) 447 [5] B Guerard et al, Nuc l Inst r and Me th A310 (1991) 116 [6] D Imr i e e t a l, Nuc l Inst r and Me th A310 (1991) 122 [7] Y G iomatar i s and G Charpak, Nuc l Inst r and Me th A310 (1991) 589 [8] M Star i c e t a l, Nuc l Inst r and Me th A307 (1991) 145 [9] R Arno l d et al, Uni vers i ty Lou i s Pasteur prepr int CRN/HE (1991) [10] G Charpak e t a l, Proc Large Hadron Col l i der Workshop, Aachen, 1990, eds G Jar l skog and D Re i n, CERN 90-10, EFCA (1990), vol 3, p 384 [1 l ] G Charpak et a l, Proc 9th I NFN Workshop on Perspect ives for New De tec tors in Future Superco l l iders, I ta ly, 1989, eds L Ci fare l l i, R W i gmans and T Ypsi l ant is (Wor l d Sc i ent i f i c, S i ngapore, 1991), p 43 ; G Charpak et al, prepr int CERN-EP/ (1990) [12] G Charpak e t a l, Nuc l Inst r and Me th A307 (1991) 63 [13] V Daugendor f et al, Nuc l Inst r and Me th A308 (1991) 519 [14] J L Janson et al, Sol ide State Commun 67 (1988) 183 [15] D Imr i e et al (paper in preparat ion) [16] P Fonte e t a l, Nuc l Inst r and Me th A305 (1991) 91 [17] S C T i ng (spokesman), Express ion of interest to the SSC Labora tory by the L Co l l abora t i on, May 1990 [18] R Zhu, Proc Large Hadron Col l i der Workshop, Aachen, 1990, eds G Jar l skog and D Rem, CERN 90-10, EFCA (1990), vo l 3, p 411 [19] Y G i oma tar i s e t a l (unpub l i shed da ta) [20] S Kwan et al, prepr int Fermi l ab-pub 91/ 98 (1991) [21] G Charpak et al, Nuc l Inst r and Me th A310 (1991) 128 [22] D Anderson et a l, Fermi l ab Tech Memo TM-1753 (1991)