Alessandro Bravar

Journée de réflexion DPNC18 June ‘12

Lepton Flavor ViolationLepton Flavor Violation3e @ PSI3e @ PSI

UniGE @ PSI

Lepton Flavor Violation

The 3e experiment ( → eee)

Sci-Fi ToF Tracker

UniGE planned contributions

Lepton Flavor ConservationLepton Flavor ConservationOrigin of Lepton Flavor Number / Conservation the neutrino produced along with a + (e) in + decay, when interacting with matter will produced only s (es), → is different from e ; → Lepton Flavor (Family) Conservation

Neutrino oscillations, however, violate this ansatz:after traveling some distance the can produce a (OPERA) or an e (T2K)

Flavor Conservation in the charge lepton sector:Processes like A → e A

→ e + → e e e have not been observed yet.

The mechanism and size of LFV remain elusive.

In the quark sector the situation is quite different:quarks decay, mix, … only the baryonic number is conserved.However Flavor Changing Neutral Currents (FCNC) have not been observed.

Lepton Flavor Violation in Lepton Flavor Violation in → eee eee

neutrino oscillations SUSY exotic particles

Current experimental limitBR(→ eee) < 1012 (90% c.l., SINDRUM 1988)

This experiment (3e @ PSI)BR( → eee) < 1015 (90% c.l. exclusion) phase I (2015 – 2017)

BR( → eee) < 1016 (90% c.l. exclusion) phase II (2018 – 2020)BR( → eee) = 3 1016 (5 discovery)

Explore physics up to the PeV scaleComplementary to direct searches at LHC

LFV in the Standard ModelLFV in the Standard Model

process is heavily suppresseddue to the small mass differenceof neutrinos (m2 ~ 10-3 eV2) !

BR ( → eee) < 1050

→ measurement not affected bySM processes

Beyond the Standard ModelBeyond the Standard ModelLFV addresses issues like origin of flavor

neutrino mass generation CP violation

LFV predicted by many BSM models: Supersymmetry Higgs triplet models Little Higgs models New heavy Vector Bosons (Z’) Leptoquark (GUT models) Extra dimensions

In many models sizeable andtherefore observable LFV effectsare expected:

BR( → eee) ~ 1012 possible(just beyond SINDRUM limit)

LFV Searches : Current SituationLFV Searches : Current SituationThe best limits on LFVcome from PSImuon experiments

→ eee BR < 1012

SINDRUM 1988

+ Au → e + Au BR < 7 1013

SINDRUM II 2006

→ e + BR < 2.4 1012

MEG 2011

BR( → eee) / BR( → e) ~ o(em/)

BR(A → eA) / BR( → e) ~ o(em/)

SINDRUMSINDRUM II

MEG

Comparison Comparison → e e and and → eee eeeEffective charge LFV Lagrangian (“toy” model) (Kuno and Okada)

effective mass scale (including coupling)

– “contact” vs “loop” amplitude contribution (parameter of “toy” model)

→ e

1 PeV

+

UniGE @ PSI (MuLAN and FAST)UniGE @ PSI (MuLAN and FAST)Measurement of the lifetime and GF

“search” for W propagator effects on GF

MuLAN

FAST

Mu3e @ PSIMu3e @ PSIan experiment to search for Lepton Flavor Violation in e e e

using the most intense surface muon beam (p ~ 28 MeV/c) in the worldsensitivity ~10-16 (PeV scale !) observe ~1017 decays (over a reasonable time scale) rate ~ 2 109 decays / sec (1 y ~ 107 sec)

200 M HV-MAPS (Si pixels w/ embedded ampli.) channels10 k ToF channels

acceptance ~ 70% for m → eee decay (3 tracks!)

B ~ 1 – 2 T

surface

p ~ 28 MeV/c

How to Find How to Find → eee decays eee decays50 nsec time frames (Si “resolution”) → 100 decays @ 2 109 stops / sec

challenge : isolate → eee events

t ~ few 100 psTime of Flight ~ few 100 ps

BackgroundsBackgrounds

irreducible backgrounds accidental backgrounds (pileup)

signal

0~

0

tracks

i i

i i

t

mE

p

BR(→ eee) = 3.4 105

precise timing (ToF): t ~few 100 psprecise kinematics (p and E resolution):

p / p < 0.5% (i.e. ~ 100 keV/c) precise vertexing: x ~0.1 mm

to suppress backgrounds

Silicon Pixel Detector HV-MAPSSilicon Pixel Detector HV-MAPSHigh Voltage Monolithic Active Pixel Sensors

< 50 m thicknessactive sensorsstandard CMOS processlow noiseradiation tolerantlow power

~ 20 20 m2 pixels200 M channels

transistor logic embedded in N-well

Heidelberg

The ToF TrackerThe ToF Tracker

3000 Sci-Fi channels 250 m fibers readout with Si-PM arrays

6000 scint. tiles readout with Si-PMs

rate ~ several MHz / Sci-Fi channels

time resolution ~ few 100 ps

readout with wave-form digitizers real time analysis pileup separation background rejection

huge data rate !

~12 cm diameter24 ribbons 16 mm wide

Hamamatsu MPPC array 5883

250 effective pitch

UniGE + ?

Sci-Fi ArraysSci-Fi Arrays5 staggered layersof 250 m fibers double cladding Kurarayscintillating fibers

SCSF-81M peak ~ 437 nm decay ~ 2.4 ns att > 3.5 m

minimize thickness to reduce multiple scatteringminimal thickness for good time resolution (light output)effective thickness ~ 1 mm (+ glue and / or TiO2 paint and / or support structure)

track “topologies”

light propagation in multicladding fibers

Si-PMs DemystifiedSi-PMs Demystified

ADC ch.

distance between peaks is constant -> gain from ADC spectra

gain vs. HV linearin +-0.5 V window

effect of cross talk

part of Labo III program(A. Bravar and S. Orsi)

peak #15

Si-PM Photo Detection EfficiencySi-PM Photo Detection Efficiency

Hamamatsu KETEK

max (SciFi) ~ 440 nm

P.D.E. ~ 30 % P.D.E. ~ 60%

detect 2 more photons → gain ~ 2 in time resolution

SciFi SciFi

55-60%

PDE = geometrical QE Geiger

(50–70 %)(60–90%)

Read-OutRead-Outamplifier ~10x flash ADC

or SCA FPGA

CFD algorithm(real time)

optical link

Si-PM

> 1 GHz> 10 -12 bit

several MHz rate → very fast amplifier rise ~ 1 nsec decay ~ 10 nsec(noise not critical because read-out with WFD)

digitizer: start with DRS4 switched capacitor array (phase I) and later DRS5 (phase II)SCA time stretcher: GHz sampling → MHz readout

best timing can be obtained using waveform digitizing(e.g. real-time algorithms simulating the functioning of a constant fraction discriminator)

huge data rate → processing of DRS information in real time (on board)

data reduction (hit processing and matching) also in real time

UniGE + PSI

DRS4 @ PSI DRS4 @ PSI http://drs.web.psi.ch

DRS4 Evaluation Board4 channels 1 – 5 GSPS

12 bitUSB power

S. Ritt

again part of Labo III equipment

Next Generation SCA (DRS5)Next Generation SCA (DRS5)

Short sampling depth Deep sampling depth

only short segments of waveform need fast sampling and readout

PSI (S. Ritt)

UniGE plansUniGE plansMu3e collaboration: Geneva, Heidelberg, PSI, Zurich, ETHZ, + …

Develop ToF system (SciFi and scintillating tiles) hardware (SciFi ribbons, Si-PMs, …) in coll. with UniZH electronics (amplifiers, DRS, firmware, …) in coll. with PSI digitizing electronics: possible synergy with NA61 and AIDA

Simulations and optimization of Mu3e detector, in particular ToF system in coll. with ALL

More concrete (next 6 months) - build SciFi array prototype and test / optimize for time resolution rate capabilities - develop fast amplifiers, readout based on commercial DRS electronics - develop (offline → real time) algorithms for DRS electronics - R&D on Si-PM PDE

FNS request: 1 PostDoc + 1 CanDoc

ADDITIONAL TECHNICAL STUFF

Late ’80s – Early ’90s (the beginning)Late ’80s – Early ’90s (the beginning)

first Position-Sensitive PMswith “crossed wires” anodebefore the advent of multi-anode PMs

16 x 16 wires (channels)delay line readout

RD-17 / FAROS

Sci-Fi arrays and Si-PMTsSci-Fi arrays and Si-PMTsHamamatsu MPPC 5883

alternative solutions single fiber readout

Zecotek linear array of 18 1mm2 MAPDs(CMS HCAL upgrade) green light !could use Hamamatsu Si-PMs

can couple (glue) Sci-Fi ribbon directlyto the photosensor+ direct mapping of the Sci-Fi array+ best optical transmission- limited sensor size

monolithic photosensor (no dead regions)blue light !

total surface ~ 10 mm2

note: this is max surface for Si-PM50 x 50 m2 pixels5 columns of pixels grouped in a single readout ch. instead of a single ch.effective readout pitch 250 m

How To Measure Best TimingHow To Measure Best Timing

J.-F. Genat et al., arXiv:0810.5590 (2008) D. Breton et al., NIM A629, 123 (2011)

Simulation with realistic noiseand best discriminators

beam measurements@ SLAC and FNAL

17 ps () can be achieved with waveform digitizing and 40 photoelectrons(no jitter from scintillator decay)

Switched Capacitor Array (DRS Chip)Switched Capacitor Array (DRS Chip)

Shift RegisterClock

IN

Out

“Time stretcher” GHz MHz“Time stretcher” GHz MHz

Waveform stored

Inverter “Domino” ring chain0.2 - 2 ns

FADC 33 MHz

The DRS5 DigitizerThe DRS5 Digitizer

coun

ter

latc

hla

tch

latc

h

writepointer

readpointer

digital readout

analog readout

trigger

FPGA

100 ps sample time. 3.1 ns hold time 2 times better timing resolutiondata driven readout(almost) dead-time-less waveform digitizing 2 MHz sustained event rate planned for 2013

S. Ritt