Jacques Bouchez

CEA/DAPNIA-Saclay & APC-Paris

ISS meeting, detector WG

KEK, Jan23rd, 2006

MEMPHYS :

A megaton WČ detector in Fréjus underground laboratory

• Proton decay

• Supernovae: explosion, relics

• Solar and atmospheric neutrinos

• Oscillations: θ13 and δCP using superbeam and betabeam from CERN

PHYSICS MOTIVATION

• The Frejus site

• Results of the excavation study

• The detector

• Photodetectors

• R&D

• Physics performance

• Schedule

most transparencies borrowed from Luigi Mosca (CARE/BENE, CERN, 23/11/05)

and from Jean-Eric Campagne (GDR neutrino, Paris, 21/10/05)

LAYOUT

The MEMPHYS Project

65m

65m

Fréjus

CERN

130km130km

4800mweExcavation engineering pre-study has been done for 5

shafts

Water Cerenkov modules at FréjusCERN to FréjusNeutrino Super-beam and Beta-beam

­ ­13­km­(12 870 m)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122 232425262728293031323334

70m x 70m x 250m

­­­­­­­­­­­France ­­­­­­­­­­­­­­­­­­­­­­­­­­­Italy

Future­Lab.

Present­road­Tunnel­at­Fréjus­(grey)andfuture­Tunnel­(black)­for­safety­with­34­bypasses­(shelters)connecting­the­two­Tunnels

A Very Large Laboratory In the middle of the Fréjus tunnel at a depth of4800

m.w.e a preliminary investigation shows the feasibility to excavate up to five shafts of about 250,000 m3 each

Henderson

HK

Main results of the Preliminary Study

1) the best site (rock quality) is found in the middle of the mountain,

at a depth of 4800 mwe

2) of the two considered shapes : “tunnel” and “shaft”,

the “shaft (= well) shape” is strongly preferred

3) Cylindrical shafts are feasible up to :

a diameter = 65 m and a full height h = 80 m (≈ 250 000 m3)

4) with “egg shape” or “intermediate shape” the volume

of the shafts could be still increased

5) The estimated cost is ≈ 80 M€ X Nb of shafts

80.00

R22.07R44.53

R44.00

72.04

34.64

44.00

36.00

Exemple of “egg shape” simulation, constrained by the rock parameter measurements made during the present tunnel and laboratory excavation.

The main feasibility criterium is that the significantly perturbated region around the cavity should not exceed a thickness of about 10 m

65m

~ 4 x SK

65m

Detector basic unit

Detector:

cylinder (a la SK) 65 m diameter and 65 m height: : → 215 000 tons of water (4 times SK)

taking out 4 m from outside for veto and fiducial cut →146 000 ton fiducial target

3 modules : 440 kilotons (like UNO) BASELINE

4 modules would give 580 kilotons (HK)

→Simulations done using 440 kt

each cavity 70 m diameter and 80 m total height

PHOTODETECTION

Baseline choice: use photomultipliers

get the highest possible coverage to get the lowest possible threshold

Ideally, we want the same light/MeV as SuperK.

…but the solution with 20’’ PMT’s becomes too expensive (cf. UNO)

(40 000 20’’ PMT’s/module with 40% coverage)

R&D on HPD started in France (with Photonis)

encouraging results from ICRR/Hamamatsu with 13’’ HPD

PMT size <=> costPhotonis @ NNN05

• Diameter 20“ <=> 12“ • projected area 1660 615 cm²• QE(typical) 20 24 %• CE 60 70 %

Cost 2500 800 €

Cost of useful PE/cm2 = PM cost/(areaxQExCE)

12.6 7.7 €/PE 40% saving

• 30% coverage (12’’) gives the same # of PE/MeV as 40% coverage (20’’)

• the required # of 12’’ PMT’s is twice the # of 20’’ PMT’s

•BONUS: better timing (risetime+jitter), better pixel localization

R&D on electronics (ASICs)• Integrated readout : “digital PM (bits out)”

– Charge measurement (12bits)– Time measurement (1ns)– Single photoelectron sensitivity

• High counting rate capability (target 100 MHz)

• Large area pixellised PM : “PMm2”– 16 low cost PMs– Centralized ASIC for DAQ– Variable gain to have only one HV

• Multichannel readout– Gain adjustment to compensatenon uniformity– Subsequent versions of OPERA_ROC ASICs

ASIC requirements• High speed discriminator for autotrigger on single photoelectron

• Coincidence logics to reduce dark current counting rate (to be defined by MC studies)

• Digitisation of charge over 12 bits

• Digitisation of time of arrival over 12 bits to provide nanosecond accuracy

• Variable gain to equalize photomultipliers response and operate with a common high voltage

• Data out wireless (why not?)

• Low cost (aim at 200 euros/channel )

R&D started at LAL Orsay, in connection with Photonis

Other interested french labs

Mechanics & PMT tests

Basic unit that we want to build and test under water

Electronic box water tight

IPNO

Taken in charge by IPNO: well experienced in photodetectors (last operation: Auger). With PHOTONIS tests of PMT 8”, 9” 12” and Hybrid-PMT and HPD

MEMPHYS physics reachA) non accelerator-based physics

• Nucleon decay (for 5 Megaton.years)

- 1035 yrs (p→e+0) - 2 1034 yrs (p→K+)- complementarity with liquid argon- some chance of discovery

– Neutrino bursts from Super-Novae explosion– 200,000 events from SN at 10kpc– 30 events from Andromeda– 2 events at 3 Mpc– collapse studies, explosion alerts (grav.antennas, telescopes)– mass hierarchy(θ13>10-3), θ13 sensitivity in [3 10-6 – 3 10-4]

– Relic Neutrinos from past Super-Novae explosions– 100 events in 5 Mt.y (with pure water)– 2000/4000 events in 5 Mt.y (with Gd loaded water)

MEMPHYS physics reachB) with CERN super and beta beams

ongoing studies within the ISS physics working group

A possible schedule for MEMPHYS at Frejus

Year 2005 2010 2015 2020

Safety tunnel Excavation

Lab cavity ExcavationP.S Study

detector PM R&D PMT production

Det.preparation InstallationOutside lab.

Non-acc.physics P-decay, SN

Superbeam Construction Superbeam

betabeam Beta beamConstruction

decision for cavity digging decision for SPL construction decision for EURISOL site

CONCLUSIONS

• The Frejus site can house a large scale (megatonne) detector

• The preferred geometry is made of cylindrical shafts

• 3 detectors give an overall fiducial mass of 440 kT

• A solution based on 12’’ PMT’s saves costs with same light as SK

• ongoing R&D for electronics (ASIC’s) and mechanics

• physicswise, MEMPHYS compares favourably to UNO and SK

• A full study on cavern + detector should be launched for EU FP7 in collaboration with liquid argon (GLACIER) and scintillator (LENA)

• Important milestone will be 2010

• Physics would start before 2020

• MEMPHYS welcomes all interested collaborators

• A document submitted to CERN council strategy group is available (see http:// council-strategygroup.web.cern.ch/council-strategygroup/SGcontrib.html)

BACKUP

SUPERBEAM BETABEAM

→ e e →

→ e e →

Superbeam + beta beam together

2 ways of testing CP, T and CPT : redundancy and check of systematics

2 beams

1 detector

2yrs

8yrs

5yrs

5yrs

pure4 flavours + K

ASICs submissions• MAROC : 64 ch multianode readout

– 64 fast digital outputs (2ns risetime)– Charge measurement with variable shaping– Gain adjustment (6bits)– 3 Digital thresholds (10bits)– Submitted June 05 (SiGe 0.35µm )

• MECANO2– Large dynamic range variable gain preamps– Fast unipolar shaper for 100 MHz counting rate– Submitted June 05 (SiGe 0.35µm)

MAROC1: BLOCK FUNCTIONALITY DIAGRAM • Complete front-end chip

with 64 channelsSubmitted in June 2005

Expected in October 2005

• Gain and Bandwidth flexibility:

–Gain adjustment per channel (6 bits: 0 to 4)

–Bipolar Fast Shaper:• Gain=5mV/fC• BW=10MHz

–Unipolar Fast Shaper:• Gain:5mV/fC• BW:100MHz• 3 thresholds: LSB=3mV, range=1V

to 3.5V

–Multiplexed charge measurement Peaking time with variable feedback network

• Tp=25ns to 200ns

Hold­signal­

Photomultiplicator­

Photons

Variable­Gain

Preamp.­

VariableSlow­Shaper

­ S&H­

BipolarFast­Shaper­

­

Unipolar­Fast­Shaper ­

64­Trigger­

outputs­

Gain­correction­(6­bits)­

4­discriminator­thresholds­(4*11bits­DACs)­

Multiplexed­

charge­output­

cmd_LUCID­

FS

_choice

­­­­­­­­

LUCID

LUCID

Vth(Bip­FS)­=­2.3VVth1(Unip­FS)=­1.07V­(1/3­pe-)Vth2(Unip­FS)=­1.3­V­(1.5­pe-)Vth3(Unip­FS)=­1.7­V­(3.5pe-)

• Possible use of IPs (expensive)

• Huge effort started in in2p3/CEA– Several designs in institutes

– 10 bit pipeline ADC (LPCC) 10MHz

– 10 Bit C/2C SAR (LAL) 1 mW 1 MHz

– 10 bit FADC (LAL) 100 MHz

– 12 bit Wilkinson (CEA,LAL,LPCC)

© J. Lecoq

Integrating the ADCs :

Pipeline ADC ©J. Lecoq100 MHz FADC ©V. Tocut