The Mu3e Experiment A new search for µ → eee https://www.psi.ch/mu3e/ A.Schöning University Heidelberg, Physics Institute
● Motivation of Searches for LFV ● Mu3e Detector Concept and Design ● Status of the Experiment ● Summary & Outlook Seminar LPNHE
3. May 2021
A. Schöning, Heidelberg 1 Seminar LPNHE, 3. May 2021 Why Mu3e? (in short)
µ µ e
γ* (→ ee )
e e e µ → e γ* (→ ee ) µ → eee
Lepton Flavor violating (LFV), and FCNC in more general, are forbidden in the SM
→ Very promising process to look for New Physics
A. Schöning, Heidelberg 2 Seminar LPNHE, 3. May 2021 SM of Particle Physics
SM
Physics after the electroweak epoch is described by the SM
A. Schöning, Heidelberg 3 Seminar LPNHE, 3. May 2021 Fermions in the Standard Model (SM)
Higgs field
neutrinos have no mass in the SM!
A. Schöning, Heidelberg 4 Seminar LPNHE, 3. May 2021 Discovery of Neutrino Oscillations
W W ν1 ν2 ν3
µ l
2 2 2 L P(να→νβ) = sin (2Θ) sin (Δ mαβ ) E ν
● Neutrino Oscillations:
solar neutrinos reactor neutrinos atmospheric neutrinos neutrino beams
A. Schöning, Heidelberg 5 Seminar LPNHE, 3. May 2021 The Fermion Masses in the SM
y ~1 (within 1%) PLOT with YUKAWA t
? SM: mν = 0
A. Schöning, Heidelberg 6 Seminar LPNHE, 3. May 2021 The Fermion Masses in the SM
TopTop QuarkQuark mm == 171171 GeV/cGeV/c2
Proton/Neutron:Proton/Neutron: mm ≈≈ 11 GeV/cGeV/c2
Elektron: Why Higgs couplings 2 m ≈ 0.5 MeV/c so different?
Neutrinos:Neutrinos: mm ≈≈ 0.010.01 -- 0.10.1 eV/ceV/c2 A. Schöning, Heidelberg 7 Seminar LPNHE, 3. May 2021 Physics Beyond the SM (BSM) l s a n t o n
i Matter-antimatter asymmetry in universe → CP-Violation t e a m v i r r Observation Dark matter e e s p b x → require new particles or interactions beyond the SM E O
Fermion generations s n
w Nature of neutrinos (Dirac or Majorana?) o n k Fermion masses (Yukawa couplings) n U → no explanation yet s
m Hierarchy “problem” (fine tuning) e l b
o Stability of Higgs field, ... r P
A. Schöning, Heidelberg 8 Seminar LPNHE, 3. May 2021 The Tree Diagrams in the SM
f f
Z0, γ H
f f Neutral Current Higgs-Yukawa coupling
→ flavor is conserved in neutral currents f
u c t ν νμ ν W± e τ (d )( s)(b) ( e )( μ )( τ )
f´ Charged Current → but flavor mixing in charged currents
A. Schöning, Heidelberg 9 Seminar LPNHE, 3. May 2021 Flavor Mixing
Quarks Leptons
Cabibbo Kobayashi Maskawa (CKM) Pontecorvo Maki Nakagawa Sakata (PMNS)
ν v v v ν d ' vud v us vub d e e 1 e 2 e 3 1 = ν = v v v ν s' vcd vcs v cb s μ μ1 μ 2 μ 3 2 ( ν ) v v v ν b' v v v b τ ( τ 1 τ 2 τ 3 ) ( 3 ) ( ) ( td ts tb ) ( ) weak mass weak mass
Q=-1/3 Q=+2/3 Q=0 Q=-1
A. Schöning, Heidelberg 10 Seminar LPNHE, 3. May 2021 Lepton Mixing & Lepton Flavor Violation γ
l l'
μ → e via ν-oscillation μ → e γ via quantum loop
W ν ν ν W γ 1 2 3 W µ e L→1/ m μ e W νμ x νe L Eν→ mW 2 2 2 L 2 2 2 2 P(να→νβ) = sin (2Θ) sin (Δ mαβ ) B(μ→e γ) ∝ sin (2Θ) (Δ m / m ) E ν αβ W
2 2 (Δ mν) 4 −50 ∝ 4 ≈ yν ≈ 10 mt
A. Schöning, Heidelberg 11 Seminar LPNHE, 3. May 2021 Conservation of Lepton Flavor is an Accidental Symmetry!
A. Schöning, Heidelberg 12 Seminar LPNHE, 3. May 2021 Beyond the SM: μ+ →e+e+e-
4-fermion interaction heavy new particles Exotic Physics in loop
loop diagrams (similar to µ → e γ) tree diagram (Mu3e specific)
Supersymmetry Higgs Triplet Model Little Higgs Models New Heavy Vector bosons (Z') Seesaw Models Extra Dimensions (KK towers) GUT models (Leptoquarks) many other models
Many models “naturally” generate lepton flavor violation!
A. Schöning, Heidelberg 13 Seminar LPNHE, 3. May 2021 Mu3e Experiment Aiming for a sensitivity (SES) requires:
BR(μ → e e e ) < 2·10-15 (phase I) → 108 muons/s (PiE5) ~next 5 years BR(μ → e e e ) < 10-16 (phase II) → >109 muons/s (HiMB) R&D
HiMB = High intensity Muon Beamline (under study)
Mu3e phase I design BR(μ → e e e ) < 10-12 (90% CL) SINDRUM experiment (1986)
Pixel Tracker
Scintillating Fibers Scintillating Tiles
A. Schöning, Heidelberg 14 Seminar LPNHE, 3. May 2021 Challenge Number 1: Rate
Goal → 1016 muon decays
● running time: ~3 years → 3 · 107s (experimental year ~ 107s)
● detector acceptance: 30%
● required muon rate: 109 µ/s → defines technological challenge
Detector has to stand high rates! → e.g. silicon detectors for tracking
A. Schöning, Heidelberg 15 Seminar LPNHE, 3. May 2021 Challenge Number 2: Background
Number of grains of sand at all beaches in France ~ 1016
Find THE grain of sand which violates lepton flavor!
A. Schöning, Heidelberg 16 Seminar LPNHE, 3. May 2021 Irreducible Background
Radiative decay with internal conversion
- e ν e+ Mu3e signal - e e+
ν e+ μ+ → e+e+e- B(μ+ → e+e+e- νν) = 3.4 ·10- 5 e+
E = m ∑i i μ p⃗ = 0 ∑i i
A. Schöning, Heidelberg 17 Seminar LPNHE, 3. May 2021 Irreducible Background
Radiative decay with internal conversion
- e ν e+ missing energy from two neutrinos R.M.Djilkibaev, steeply falling! R.V.Konoplich PRD79 (2009) ν e+
B(μ+ → e+e+e- νν) = 3.4 ·10- 5 missing energy taken by neutrinos
very good momentum + total energy resolution required!
A. Schöning, Heidelberg 18 Seminar LPNHE, 3. May 2021 Accidental Backgrounds
Overlays of two ordinary µ+ decays with a (fake) electron (e-) Electrons from: Bhabha scattering, photon conversion, mis-reconstruction
- e e+
e+
e+ Need excellent: Vertex resolution example for Bhabha pileup Timing resolution Kinematic reconstruction
A. Schöning, Heidelberg 19 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
108 muons per second (phase I)
p=28 MeV/c
muon stopping target (hollow)
A. Schöning, Heidelberg 20 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
80 x 80 µm2 pixel
● ultra thin silicon pixel detector (HV-MAPS) with 1 per mill radiation length / layer for vertexing
A. Schöning, Heidelberg 21 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
σ(t) ~ 250 ps
● fast timing detectors (scintillating fibers) → time coincidence
A. Schöning, Heidelberg 22 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
σ(t) < 1 ns
● ultra thin silicon pixel detector (HV-MAPS) with 1 per mill radiation length / layer for outer tracking layers → momentum information
A. Schöning, Heidelberg 23 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
Long cylinder!
~180 cm
~15 cm
not to scale!
● recurl stations with silicon pixel detector (HV-MAPS) → increased acceptance
A. Schöning, Heidelberg 24 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
Long cylinder!
~180 cm
~15 cm
σ(t) < 100 ps not to scale!
recurl stations with extra scintillating tiles → very precise timing
A. Schöning, Heidelberg 25 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
transverse view:
+ strong magnetic field (B=1T) in helium atmosphere + helium gas cooling
A. Schöning, Heidelberg 26 Seminar LPNHE, 3. May 2021 Tracking Resolution + Multiple Scattering
Muon decay (m=105.6 MeV): → electrons in low momentum range p < 53 MeV/c limited hit Multiple scattering is dominant! resolution regime Need thin, fast and high resolution tracking detectors operated at high rate (>109 particles/s @ phase II)
1 Θ ∼ X / X MS P √ 0 multiple scattering regime
A. Schöning, Heidelberg 27 Seminar LPNHE, 3. May 2021 Momentum Resolution Standard spectrometer: “Half turn” spectrometer:
multiple-scattering angle
σ p ΘMS σ ∼ (linearised) p ∼ O(Θ2 ) P Ω P MS
requires large lever arm best precision for half turn tracks
large bending angle Ω measure recurlers
A. Schöning, Heidelberg 28 Seminar LPNHE, 3. May 2021 Mu3e Design (Phase I)
transverse view:
electrons with p~33 MeV/c make roughly semi-circles!
A. Schöning, Heidelberg 29 Seminar LPNHE, 3. May 2021 Momentum Resolution (Simulation)
6 hits (recurl) 4 hits (layers) 8 hits (central)
A. Schöning, Heidelberg 30 Seminar LPNHE, 3. May 2021 Mu3e Phase I Design Long cylinder!
~110 cm σ(p) < 100-400 keV/c
σ(x) < 25 µm σ(t) < 100 ps ~15 cm
σ(t) < 500 ps
σ(x) < 25 µm not to scale! Technical Challenges: ● multiple Coulomb scattering → ultra-thin tracking layers ● high particles rates → highly granular detectors and fast online reconstruction ● compact design → high integration level (sensors, readout ASICs)
A. Schöning, Heidelberg 31 Seminar LPNHE, 3. May 2021 Mu3e Phase I Design Long cylinder!
~110 cm σ(p) < 100-400 keV/c
σ(x) < 25 µm σ(t) < 100 ps ~15 cm
σ(t) < 500 ps
σ(x) < 25 µm not to scale! Innovative Technologies: ● High Voltage Monolithic Active Pixel Sensors (HV-MAPS) for tracking
● 2 gaseous helium cooling system (<400mW/cm ) and ultra-thin pixel modules (0.1 % X0) ● MuTrig readout ASIC for timing detectors with ~30 ps time resolution
● Online filter farm based on Graphical Processing Units
A. Schöning, Heidelberg 32 Seminar LPNHE, 3. May 2021 Paul-Scherrer Institute (CH)
High intensity Proton Accelerator (HiPA ) → 2.4 mA protons at 590 MeV (1.5 MW)
Muon Beam: PiE5: Compact Muon Beamline for Mu3e
● World’s most intense continuous muon beam MEG beam ● Low momentum muons ~28 MeV/c ● PiE5 beamline shared between MEGII and Mu3e
➢ 8 + expect 1.4∙10 μ /s at Ip= 2.4 mA ➢ about half is stopped on µ-stopping target
→ Mu3e Phase I Mu3e beam
A. Schöning, Heidelberg 33 Seminar LPNHE, 3. May 2021 Mu3e Collaboration
Germany ● University Heidelberg (KIP) ● University Heidelberg (PI) ● Karlsruhe Institute of Technology ● University Mainz
Switzerland ● University of Geneva ● Paul Scherrer Institute ● ETH Zurich ● University Zurich ● [University of Applied Sciences Northwestern Switzerland] associated partner United Kingdom ● Bristol ● Liverpool ● Oxford ● UC London
about 70 members; ~15 PhD students
A. Schöning, Heidelberg 34 Seminar LPNHE, 3. May 2021 HV-MAPS Detector Technology
High Voltage-Monolithic Active Pixel Sensor MuPix10 prototype ladder (HV-MAPS) Mupix10 demo ladder transistor logic embedded in N-well (“smart diode array”)
thickness = 50 µm
I.Peric et al., NIM A 582 (2007) 876 sensor: 20 x 20 mm2 pixel: 80 x 80 μm2 ● active sensor: → hit finding + digitisation + readout MuPix prototypes characterized in lab and in several test beams ● HV-CMOS 180nm: 60-120 V ● efficiency (>99%) & noise ● low cost process (AMS, TSI) ● time resolution (<20 ns) ● ● high rates (radiation hardness) thinned to ~50 μm (~ 0.0005 X0) ● temperature-dependence ➔ specifications fulfilled
A. Schöning, Heidelberg 35 Seminar LPNHE, 3. May 2021 Preliminary Mupix10 Efficiency (PSI)
threshold 42mV (~670 e-) average efficiency ~ 99.85% (noise & rate dependent → dead time) no pixels masked! no TDAC tuning of individual pixels O(10) noisy pixel out of 64000 → lead to some deadtime losses
Preliminary
noisy pixel (not masked)
Mu3e Pixel Sensor 36 PSI, BVR 52, January 25, 2021 Pixel Tracking Detector
Ultra-thin pixel sensor modules (X/X0 = 1.15 per mille ) Gaseous He-Cooling System
simulation
MuPix (HV-MAPS) High Density Interconnect Thermo-Mechanical Mockup (vertex) d < 100 µm (LTU, Ukraine) MuPix7
50 µm 50 µm Si-heater chips
Al-Kapton Monolithic pixel sensor in 180 nm HV-CMOS A. Schöning, Heidelberg 37 Seminar LPNHE, 3. May 2021 Pixel Tracking Detector Prototype
uses PCBs instead of High Density Interconnects (HDI)
A. Schöning, Heidelberg 38 Seminar LPNHE, 3. May 2021 Scintillating Fibres
Scintillating Fibre Detector ● Scintillating fibres: Kuraray SCSF-78MJ (multi-clad) ● SiPM Hamamatsu S13552-HRQ ● MuTrig TDC ASIC (Heidelberg-KIP) for readout prototype ladder ➔ very challenging space constraints ➔ time resolution ~250 ps
➔ thickness X/X0 0.2%
SciFi detector double SciFi ladder Hamamatsu S13552-HRQ
A. Schöning, Heidelberg 39 Seminar LPNHE, 3. May 2021 Scintillating Tiles Timing Detectors
Scintillating Tile Sub-Module
DESY test beam
Scintillating Tiles ● tiles ∼ 6.5 × 6.5 × 5mm3 ● SiPM 3 x 3 mm2 ● Readout with MuTrig ASIC ( Heidelberg-KIP)
● time resolution < 100ps Time resolution < 100ps
A. Schöning, Heidelberg 40 Seminar LPNHE, 3. May 2021 Data Acquisition and Filter Farm
● Continuous readout with frontend zero-suppression ● Online track reconstruction based on new multiple scattering fit (https://arxiv.org/abs/1606.04990) ● Filter farm based on NVIDIA GPUs ● DAQ hardware is ready! ● hard working on firmware and SW
Mu3e Frontend Board with Arria V FPGA DE5aNet Receiving Board (Arria 10) “Switching Board” PCIe40 (from CPPM) (inside the magnet)
A. Schöning, Heidelberg 41 Seminar LPNHE, 3. May 2021 Mu3e Mass Plot (Simulation)
10-12
10-13
10-14 (upper limit) 10-15
A. Schöning, Heidelberg 42 Seminar LPNHE, 3. May 2021 Expecte Sensitivity versus Time
A. Schöning, Heidelberg 43 Seminar LPNHE, 3. May 2021 Experimental Status at PSI
Superconducting Solenoid PiE5
Mu3e Solenoid
B=1 Tesla
Mu3e “Skywalk”
● Superconducting solenoid produced by Cryogenic (London) was delivered in July 2020 (milestone!)
A. Schöning, Heidelberg 44 Seminar LPNHE, 3. May 2021 June 2021 Integration Run
Pixel Detector Prototype Scintillator Tiles (2x)
Scintillating Fibers
Goal: final test of all detector systems including data acquisition (DAQ) experimental cage before starting mass production A. Schöning, Heidelberg 45 Seminar LPNHE, 3. May 2021 Summary
Mu3e has an unique discovery potential for New Physics Technical Design Report published in 2020 (https://arxiv.org/abs/2009.11690) First Integration Run with all detector systems planned for May/June 2021 ➢ production readiness ➢ construction phase of about two years
Start of data taking in 2023 → goal for Phase I B(μ+ → e+e+e-) ≤ 5 · 10-15 (90% CL)
A. Schöning, Heidelberg 46 Seminar LPNHE, 3. May 2021 Mu3e Phase II and High Intensity Muon Beamline (HiMB) Goal: deliver ~1010 muons/s to two experiments (Mu3e, muSR)
Mu3e
large capture solenoid
solenoidal beam transport
Possible design: new target M
A. Schöning, Heidelberg 47 Seminar LPNHE, 3. May 2021 Mu3e Phase II and High Intensity Muon Beamline (HiMB) Goal: deliver 1010 muons/s to two experiments (Mu3e, muSR)
● Mu3e Phase II: B(μ+ → e+e+e-) ≤ 10-16 (90% CL)
● HiMB Physics Case Workshop 6.-9. April 2021 (https://indico.psi.ch/event/10547/)
A. Schöning, Heidelberg 48 Seminar LPNHE, 3. May 2021 Backup
A. Schöning, Heidelberg 49 Seminar LPNHE, 3. May 2021 History of LFV Decay experiments
4 orders of Mu3e I magnitude Mu3e II
A. Schöning, Heidelberg 50 Seminar LPNHE, 3. May 2021 Model Independent Comparison
dipole μeee contact IA e e
Effective cLFV Lagrangian: e e m μ e μ dipole κ eμ ν ,ee L = H + J ν J Λ2 (1+ κ) Λ2 (1+ κ) μ e
μ e κ = parameter κ → 0 Λ = common effective mass scale κ → ∞
B(μ+ → e+e+e-) B(μ+ → e+e+e-) ~ 0.006 → ∞ B(μ+ → e+γ) B(μ+ → e+γ)
A. Schöning, Heidelberg 51 Seminar LPNHE, 3. May 2021 Model Independent Comparison
dipole μeee contact IA e e
e e
μ e
μ e
μ e B( μ →eee ) < 10-12 κ → 0 κ → ∞
B(μ+ → e+e+e-) B(μ+ → e+e+e-) ~ 0.006 → ∞ B(μ+ → e+γ) B(μ+ → e+γ)
A. Schöning, Heidelberg 52 Seminar LPNHE, 3. May 2021 μ+ →e+e+e- Diagrams
μ+ → e+e+e- e e e e
γ Z
μ e μ e photon penguin Z - penguin
e e e e
H
μ e μ e Higgs-penguin box diagram
A. Schöning, Heidelberg 53 Seminar LPNHE, 3. May 2021 LFV-Effective Field Theory
A.Crivellin et al., PSI-PR-16-15
Representation by Wilson coefficients and higher-dimensional operators:
A. Schöning, Heidelberg 54 Seminar LPNHE, 3. May 2021 Other Possible Searches with Mu3e Cheng et al. (2013) ● Search for μ → eγ (LFV) with converted photons ➢ better reduction of accidental BG than MEG
● Search for familons ➢ pseudo Goldstone bosons of spontaneously broken flavor symmetry ➢ dark matter candidate μ+ e+ μ+ e+ X X invisible e+ prompt or long-lived e-
● Search for dark photons A’:
➢ process μ → e νν A’ Essig et al. (2014)
A. Schöning, Heidelberg 55 Seminar LPNHE, 3. May 2021 Search for Familons μ+→ e+X
●Familons are the Goldstone Bosons of a spontaneously broken flavor symmetry
m =60 MeV ● + + X μ → e X is 2-body decay ●Search for a peak on the e+ momentum distr. thesis A.-K.Perrevoort
Michel BG
Projected Mu3e Sensitivity (Phase I)
A. Schöning, Heidelberg 56 Seminar LPNHE, 3. May 2021 Side Remark: Exotic LFV Decays
Weakly Interaction Slim Particles (WISP)
light axions are theoretically well motivated axions X could be LFV familons ➢ pseudo-Nambu-Goldstein boson of spontaneously broken family asymmetry ➢ addressing dark matter X could weakly couple to SM particles: μ+ → e+ X ( μ+ → e- γ X ) X would be long-living or decay in detector: X → e+e- , νν μ+ → e+ ν ν μ+ → e+ X Search topologies: peak in Michel spectrum displaced e+e- vertex μ+ → e+e+e- signature with M(ee) peak Michel spectrum
Ee A. Schöning, Heidelberg 57 Seminar LPNHE, 3. May 2021 Search for Dark Photons in A’→ e+e-
signal ● Search for Dark Photons with zero or short lifetime: + + µ → e νe νµ A’ with A’ → ee ● peak in m(e+e-) invariant mass
thesis A.-K.Perrevoort
backgrounds
A. Schöning, Heidelberg 58 Seminar LPNHE, 3. May 2021
A. Schöning, Heidelberg 59 Seminar LPNHE, 3. May 2021 PiE5 Beamline + Experimental Region
Compact Muon Beamline was successfully commissioned providing up to 108 muons/s
mockup for Mu3e solenoid
A. Schöning, Heidelberg 60 Seminar LPNHE, 3. May 2021 Experimental Cage & Services
experimental cage (Heidelberg)
Intro, Mu3e 61 PSI, BVR 52, January 25, 2021 Mupix10 Design & Specifications
Specification from TDR
Mu3e Pixel Sensor 62 PSI, BVR 52, January 25, 2021 Beam Test Results and Mupix10 Telescope
Telescope: 3+1 (DUT) layers Note, all following results are very DESY & PSI testbeams (despite Corona) first results and preliminary! MuPix works fine in general!
Hit map (electrons @DESY) Time correlation between layers (PSI)
Mu3e Pixel Sensor 63 PSI, BVR 52, January 25, 2021 Mupix10: Noise and Efficiency Scan
wide efficiency plateau! efficiency > 99% for thresholds< 80 mV noise is rather flat and it includes here many scattered beam particles (~90%) noise < 0.1 Hz/pixel after beam particle and hot pixel removal (~10 out of 64000) (→ not shown)
Preliminary
Mu3e Pixel Sensor 64 PSI, BVR 52, January 25, 2021 2-Comparators
(from TDR) Motivation of 2-comparator design use lower threshold for reducing time walk (ToA) use higher threshold for hit validation use higher threshold for measuring falling edge more precisely → better ToT
Two methods to measure ToT: ➔ rising and falling edge from single or high threshold (“high”) ➔ rising lower edge and falling higher threshold (“mix”) → not yet tested
Mu3e Pixel Sensor 65 PSI, BVR 52, January 25, 2021 MuPix10 Delay Circuit
Issue: ● Hits should be read out after completion of ToT measurement ● ToT measurement depends on pulse height → disturbs chronological order of hits ● Solution: read hits after adjustable fixed delay Challenges: ➔ Handling of overflows (~huge pulses) is required → counter stops ➔ Delay dispersion of pixels should be small
time-over-threshold Mupix10
Mu3e Pixel Sensor 66 PSI, BVR 52, January 25, 2021 Mupix10: Pixel Tuning
3 bit tune dac (TDAC) per pixel tune with charge injection significant dispersion reduction measured
Mu3e Pixel Sensor 67 PSI, BVR 52, January 25, 2021 Mu3e Timeline
Schedule
R&D R&D
HiMB = High Intensity Muon Beamline → delivers more than 109 muons per second
A. Schöning, Heidelberg 68 Seminar LPNHE, 3. May 2021