Xenon Detector Status Liquid Xenon Group Outline Detector

Xenon Detector Status Liquid Xenon Group

Outline • Detector Setup • Operation • Performance • Problems and solutions 1

Detector Setup

PMT installation completed • All PMT successfully installed in Aug 2007 together with – – – LEDs Alpha sources (plates and wires) PT 100 temperature sensors Laser fibers Surface level meter • < 200~300 um gap btw inner slab and wall g xenon 3

Sensors, LEDs, and alpha sources Capacitor-type surface level meter Pt 100 Temp sensor 22 sensors in LXe 21 sensors on the wall Two type of a sources • Plate 20 pieces • Wire 5 x 5 wires LED 3 different attenuation x 10 4

Cabling and Al filler installation 7. 5 l 4. 5 l • Filler volume – Lateral 7. 5 lx 6 x 2 + 4. 5 lx 2 x 2 = 108 l (design) 108 – 7. 5(US) – 4. 5(DS) = 96 l (reality) – Bottom 49 l (design) 49 – 10 = 39 l (reality) 5

We were ready to close but… • The shape of the newly delivered sealing (metal gasket) was wrong ! – Does not fit the groove? ! – Radii were correct but they made a mistake in bending – Sent back to Garlok and repaired • While waiting repair work, old gaskets were recycled with indium foil around the corner – Vacuum test OK – But leak of the order of 10 -4 mbarl/sec with ~1 bar xenon in the cold vessel outgas ~1 bar xenon vacuum Warm vessel vacuum Xe leak Mass spectrometer 6

RGA of leaking xenon • Doubly charged 132 Xe (Xe++) is used to evaluate amount of xenon in the warm vessel • Production cross section with 110 ke. V electron is ~40% of Xe+ production • Used as reference data for later use 10 -4 mbarl/sec leak 7

Closing the covers • Gaskets replaced with newly delivered ones on 5/Sep • Super-insulator on the cold vessel • 240 Nmm torque 8

Pressure/leak test • • • Xenon gas (liquid) in the cold vessel Mass spectrometer on the warm vessel RGA data with recycled gasket used as a reference 10 -4 mbarl/sec leak – Doubly charged 132 Xe (Xe++) is used to evaluate amount of xenon in the warm vessel – Production cross section with 110 ke. V electron is ~40% of Xe+ production 9

Estimation of xenon leak rate Recycled gasket 10 -4 mbarl/sec leak Room temperature xenon gas New gasket Room temperature xenon gas Better at low temperature 10

Operation

Gas/liquid system Gas-phase purifier Gas line Liquid line Detector 1000 L dewar Liquidphase purifier High pressur e Storage 12

Evacuation and liquefaction • Evacuation started on 5/Sep – Thanks to cryo pump (AISIN) – 6. 9 x 10 -3 Pa in 3 days Cryo pump • Cooling started on 10/Sep • Liquefaction started on 15/Sep 0. 133 MPa Xenon gas 0. 11 MPa • Surface level was monitored with – Temperature sensors – Level meter (long and short) 13

Liquid transfer • Liquid transfer by pressure difference between two cryostats through a vacuum insulated pipe • Started on 17/Sep and completed on 20/Sep – 10 liter/hour transfer speed • Xenon filling was done in 15 days after starting evacuation Xenon liquid ~3 m 0. 25 MPa 0. 11 MPa 14

End of xenon filling temperature Level meter • Additional 10 liter was transferred for assurance 15

How many liters of liquid xenon? • Cryostat volume 1200 liter • PMT and support 142 + 43 = 185 liter • Filler 135 liter – Lateral – Bottom • Cable 96 liter 39 liter 20~30 liter • Additionally transferred amount of xenon ~10 liter • 1200 – 185 – 135 – 25 + 10 = 865 liter • Consistent with remaining amount of xenon in the dewar 16

Labview Slow Control megon 00 PC in barrack MEG Central DAQ System For shift crew use Labview (SCFE) XEC PC 2 XEC dedicated SCFE Node cooperation Alarm to experts XEC PC 1 For expert use Important controls are implemented in SCS nodes Detector, storage, dewar, purifiers 17

Liquid circulation • Circulate xenon in liquid phase – Circulation pump • 100 liter/h@3175 rpm, Dp = 0. 2 MPa – Molecular sieves Purifier Cartridge Molecular sieves, 13 X 25 g water PT Freq. Inverter OMRON • >24 g water absorption Temperature Sensor 18

Succeeded! Circulation period 50. 63 Hz ! Circulation speed evaluation • change of the surface level after closing the inlet valve • 3. 6% / 30 sec ~ 432% / h • 1% corresponds to 0. 165 liter 0. 165 x 432%/h = 71 liter /hour 19

Liquid-phase purification • • Light yield for 17. 6 Me. V g saturated around 23/Nov (180 h purification time) Necessary to continue longer than we expected – 5 hours purification was enough in the LP test (100 L LXe) – Probably due to • slower circulation speed (100 L/h 70 L/h) • Worse initial condition compared to the LP • • Needed longer time to prepare monitoring tools due to PMT HV feedthru problem (reported later) Noise from the pump (freq. inverter? ) affected other detectors C-W run 17. 6 Me. V gamma 180 h 205 h 23/Nov 2/Dec 70 h 14/Nov 20

Absorption Length l > 3 m @95 % C. L. • Ratio Data/MC vs distance fitted with an exponential curve. • Inner and Outer face PMTs • Cosine of incidence angle < -0. 2 • Slope compatible with zero (no absorption). After 50 h purification, 4/Nov Alpha source PMT 21

2 D display, charge/time 2 D hist, charge: PMT# 3 D reconstructed position Performance Waveform 1 D histograms Charge: event#

PMT Calibration • PMT calibration – LED • PMT gain – Flushing LEDs at different intensities – Npe~1/s 2 • Time offset calibration – Viewing one LED flushing with many PMTs simultaneously – Alpha • QE and absorption length evaluation • Liquid and cold gas 23

Time offset determination • • Possible method only in non-segmented detector like ours c’ is obtained by using all data Different LEDs viewed by one PMT Measured Flashing Time Of Speed of LED light Offset of j-th PMT i-th LED Tj RD run (ultra low) ti –ti-1 1/c’ li, j-li-1, j txe - t. TC 24

C-W run • Li at 14. 6, 17. 6 Me. V • B at 4. 4, 11. 7 and 16. 1 Me. V • Details in Giovanni’s presentation 25

CEX run - Pi 0 calibration • Eg CEX process – 170 o • p 0(28 Me. V/c) g g • 54. 9 Me. V < E(g) < 82. 9 Me. V q Eg • 54. 9 Me. V p-p p 0 n LH 2 target p 0 1. 3 Me. V for q>170 o 175 o 0. 3 Me. V for q>175 o Eg q • 82. 9 Me. V Eg Na. I tagging counter – 3 x 3 crystals, APD readout g • Pb collimator panel in front of the Xe detector 26

LH 2 Target • Pressure test of cell 4. 5 bar (abs) • Time to liquefy – 80 min from start of LHe flow • Liquid stability – 1. 2 bar operating pressure (96% cell full) – 1. 3% RMS, 6% max excursion • Liquid Helium consumption – 2. 4%/h – 42 h auto 27

(+/ bea +/- m ax is 21 - 0. 07) deg. Na. I Detector Up/down +/- 60 (+/- 0. 005) deg. 28

CEX run data analysis, preliminary • • • Position cut – Cut shallow events (< 2 cm) – Select only center events ( |u|, |v| < 5 cm) Position correction Pile-up rejection by light distribution Select center event on Na. I detector Not applied QE correction – If applied worsen resolution. Pedestal has 2% spread – Needs better baseline evaluation – Check hardware for 2008 run 55 Me. V gamma pedestal sup = 2. 4% FWHM = 6. 5% 29

Time Measurement • Using only 12 PMTs around the center Intrinsic resolution by T-B analsysis 115 ps Practical resolution by t. Xe – tsci 280 ps T : weighted mean of inner PMT timing after subtracting photon propagation time Effective velocity 10 cm/ns 30

Position Measurement • Using collimator run data, 31

Physics run • RD run gamma energy – Data : Xe self trigger threshold=3. 5 V – MC : RD event generation + event overlaps + trigger simulation • Vertical scale is scaled assuming, – Mu stop = 5 e 6 – Calorimeter acceptance = 0. 1 – Calorimeter detection efficiency = 0. 6 32

Problems found in 2007 and solutions for 2008 run

Feedthru • We could not apply required voltage on all PMTs at the beginning • We found that this had been caused by spark in the feedthru • Needed to prepare “new ones” for 2007 run – Commercial products or hand made? 34

How did we make new ones? Wataru’s Design Air Metal body Glass insulator Xe Production procedure • Fix pins in the holes and fill with silica • Bake in argon atmosphere • Cool down No need to change connector Replacement can be done quickly Body made of insulator (not metal) 35

Installation • Oct. 10 -12: LXe recovery to 1000 L Dewar • Oct. 13 -14: GXe recovery • Oct. 15 -16: Mounting new flanges & testing – With flushing dry nitrogen gas in the cryostat • Oct 17 -19: evacuation • Gas filling ~0. 13 MPa – Successfully took gas alpha data at 800, 900, 1000 V – LED data • Liquid transfer started at almost same time – Until the detector is cooled we continued LED and alpha DAQ – 21, 22, 23/Oct transfer speed 15~20 liter/hour – Completed early morning of 24/Oct • 2 weeks interruption of DAQ 36

PMT status after replacing feedthru • Stable operation after replacing feedthru • LED intensity optimization – Better gain evaluation than before 37

As a bonus… • Xenon recovered through purifier – Practice to transfer the liquid to the dewar – Water contamination suppressed • Cryostat was warmed up to ~220 K – 1 st experience of temperature cycle – Test of the gasket – RGA: I = 3. 8 x 10 -13 A 5. 7 x 10 -7 mbarl/sec 132 Xe++ 38

Future plan (currently ongoing) • Replacement during winter shutdown – 48 pin x 6 x 4 – Need to change connectors – Used in the small prototype and PMT test chamber ceramic Kyocera Ultra High Vacuum Feedthrough welding 39

Heat load • • • Pressure is slowly increasing under normal operating condition. Refrigerator cooling power: 200 W Expected redundant cooling power: ~100 W – – – Calculation based on LP modeling PMT: 37 W (Vmean = 775 V) Conduction: 64 W • • Cable (50), Chimney (4. 8), SI(3. 1), Support(6. 3) Cables Heat income through Cu cooling pipes was not taken into account – Cu heat conduction: 390 W/m/K • • • ? PMTs 10 mm diam 1 mmt pipe, 20 cm 390 x 135 (K) x 2. 83 x 10 -5/0. 2 (m) = 7. 4 W 6 pipes 44. 4 W Cf. Steel heat conduction ~20 W/m/K Pipes There seems to be other heat leak… – – Larger than 50 W Super-insulator? 40

LN 2 cooling pipe • Cu pipe feedthru SUS pipe feedthru 41

Cooling Power 42

Blow up of low temperature valve • Low temperature valves blew up – A few liters of liquid xenon was lost • Purifier cryostat was opened – Misleading valve design – No documentation on the manual • Piping was modified and no valve is in use now Wilson Seal Air Wilson Seal xenon Plug or shaft PT Cup nut collar bellows O-ring Liquid xenon 43

PMT Signal Splitter • Spark in feedthru’s destroyed protection zener diodes on PMT splitter boards – base-line shift at splitter output – Signal was out of range of WFD • Fixed by replacing all zener diodes splitter DRS PMT Trigger 44

Light Yield Discrepancy between a and g data • Energy scale discrepancy btw alpha and gamma – Too small light yield from g events (~1/2) – Not due to magnetic field • Confirmed by taking C-W data w/o COBRA field • Purity seems good Improvement and plateau of light yield of both gammas and alphas • Have a look on WF a 45

Waveforms • triplet= 22 ns • recomb= 45 ns g = 34 ns a = 21 ns g Xe Xe a e ! Careful treatment of electronics time constant is necessary e Xe Before purification Q/A was 1. 93+/-0. 02 in LP test A Q Electronegative impurity? Oxygen? ? 46

Electronegative Impurity Removal • O 2 getter cartridge – Developed for LAr use at CERN – be mounted at the exit MEG liquid-phase purifier with by-pass valves – Preparing an oxygen purity monitor also – will be ready at an early stage of 2008 run 47

Current Status and Schedule • Xenon recovered to the 1000 liter dewar – Gas analysis will be done on site • Cryostat is opened now – All PMTs and cables are checked – Replacing feedthru is in progress – LN 2 cooling pipe modification • Cryostat will be moved back to Pi. E 5 at the end of March – Evacuation, liquid transfer, purification – Ready on 19/April – Schedule at http: //meg. web. psi. ch/subprojects/install/xenon. html 48