PSI proton facilitiesFrank Meier Aeschbacher

SINQ

TgE

TgM

The PSI experimental hall(in a slightly older floorplan)

SINQ

TgE

TgM

High Intensity Proton Accelerator (HIPA)• 590 MeV• 2 mA (capable of more)

The PSI workhorse for protons

SINQ

TgE

TgM

Target M („maigre“)• Facilitates πM1 beamline• MUSE experiment (proton radius)• Popular test beam facility (π, e, µ)

SINQ

TgE

TgM

Target E („épais“)• Facilitates πE1 and πE5 beamlines• muX: atomic parity violation in muonic atoms• MuCool and cold muonium• CREMA: hyperfine splitting in muonic hydrogen• MEG and Mu3e

SINQ

TgE

TgMπM1πE5

πE1

The nice thing about PSI:Uninterrupted beams, all beamlines operate simultaneouslyExcept for a short burst towards UCN every 300 s

SINQ

TgE

TgM

UCN• World’s most intense ultra-cold neutron

source • nEDM (finished 2017), n2EDM (under

construction)

PSI home to highest intensity DC μ+ beam: 5 x 108 μ+/s

Next generation cLFV experiments require higher muon rates (Mu3e phase-II, future MEG)

Provide new opportunities for μSR experiments

Maintain PSI leadership in high intensity muon beams and its expertise in low-energy precision physics

Muons at PSI

cLFV• Three channels in µ sector: µ → e, µ → eγ, µ → eee

• Two searches at PSI: MEG-II, Mu3eRequire DC beam

• Other facilities provide pulsed beams, suitable for µ → e:

→1010 μ-/sCOMET: Rμe = 𝒪(10-17)

→5x1010 μ-/sMu2e:Rμe = 𝒪(10-17)

Introduction to Mu3e

Mu3e is an experiment to search for

µ+ ! e+ee+

A very rare decay.

We’re in an unusual regime, hence allow for some physics background.

3 / 46

Introduction to Mu3e –– Signal in r-view

e+

e+

e–

SignalSM: < 1 1054

Ppi = 0

minv = mµ

ti = tj 8 i , jcommon vertex

e+

e+

e–ν

ν

Radiative decaySM: 3.4 105

Ppi 6= 0

minv < mµ

ti = tjcommon vertex

e+

e+

e–

Accidentalbackground

Ppi 0

minv mµ

ti tj“bad vertex”

5 / 46

Mu3e detector concepts

Phase-I configuration:

Target

Inner pixel layers

Outer pixel layers

Recurl pixel layers

Scintillator tiles

μ Beam

I High rate: 108 muon stops on target per second

I Time resolution (pixels): 20 ns

I Vertex resolution: about 200 µmI Momentum resolution: about 0.5MeV

I All inside a cryogenic 1T magnet, warm bore I.D. 1m

10 / 46

Mu3e detector concepts – Tile detector

• Separate SiPM matrix and ASIC• Place the ASICs on the mezzanine board• Will allow better assembly and replacement procedure

April 18, 2019 8

Mounting the long Mezzanine board that include the ASICs

Tile key parameters:

r 6 cm, l 360mm

7 modules with 2 4 4 tiles

Eljen EJ-228 plastic scintillatorcubes, 6.3 6.2 5.0mm3

SiPM: Hamamatsu MPPCS13360-3050PE

t 50 ps, > 99%

11 / 46

Mu3e detector concepts – Fiber detectorSciFi key parameters:

r 6 cm, l = 300mm

12 ribbons 32.5mm wide

3 layers of Kuraray SCSF-78MJfibres, 250 µm diameter

128 channel SiPM S13552-HQR(LHCb-type)

t 100 ps, > 96%, < 0.2% x/X0

12 / 46

UniGE, ETH, UZH, PSI

Mu3e detector concepts

4mm

6mm

HDI ~100µm

Mupix sensor 50µm

Mupix periphery

polyimide 15µm

SpTAB bonds

Radiation length per layer: see next slide.

23 / 46

Mu3e detector concepts – Mu3e pixel cooling2

Coolingsystem

Part B (Centre)

Global

Part A (Upstream) Part C (Downstream)

Global V-fold layer 4 Gap layer 3 & 4 V-fold layer 3 Gap layer 3 & SciFi

Silicon layer 4Silicon layer 3 Silicon layer 2

Silicon layer 1

Gap layer 1 & 2Target

z

Figure 2.2: Helium cooling system of the silicon chips with detail of the centre part.

10

26 / 46

Mu3e collaboration• Switzerland:

PSI (hosting lab), ETH Zürich, Universität Zürich, Université de Genève

• Germany:Universität Heidelberg, Karlsruhe Institute of Technology, Universität Mainz

• United Kingdom:University of Oxford, University of Liverpool, University of Bristol, University College London

How we produce muons at PSI

Surface muons (~28 MeV)

Cloud muons

protons

π+

μ+

surface muons stopped pion decay

π+/-

μ+/-

cloud muonspion decay-in-flight

High Intensity Muon Beamline (HIMB)

• Future experiments need higher intensities

• 1010 µ+/s would be lovely

• How to achieve this?

Simulation for beamlines at PSIImplemented our own pion production cross sections into Geant4/G4beamline based on measured data and two available parametrizationsValid for all pion energies, proton energies < 1000 MeV, all angles and all materialsImplemented “splitting” of pion production and muon decay to speed up simulation

!22

Pion Kinetic Energy T [MeV]0 50 100 150 200 250 300 350 400 450

b/sr

/MeV

]µ

/dT

[Ω

/dσ2 d

1−10

1

10

o = 22.5θ, +πCross section on C, 585 MeV,

DataBERTBICINCL_ABLAINCLXXParametrization

R. L. Burman and E. S. Smith, Los Alamos Tech. Report LA-11502-MS (1989)R. Frosch, J. Löffler, and C. WIgger, PSI Tech. Report TM-11-92-01 (1992)F. Berg et al., Phys. Rev. Accel. Beams 19, 024701 (2016)

Simulation for beamlines at PSIImplemented our own pion production cross sections into Geant4/G4beamline based on measured data and two available parametrizationsValid for all pion energies, proton energies < 1000 MeV, all angles and all materialsImplemented “splitting” of pion production and muon decay to speed up simulation

!23

Pion Kinetic Energy T [MeV]0 50 100 150 200 250

b/sr

/MeV

]µ

/dT

[Ω

/dσ2 d

1−10

1

10

o = 90θ, +πCross section on C, 585 MeV,

DataBERTBICINCL_ABLAINCLXXParametrization

R. L. Burman and E. S. Smith, Los Alamos Tech. Report LA-11502-MS (1989)R. Frosch, J. Löffler, and C. WIgger, PSI Tech. Report TM-11-92-01 (1992)F. Berg et al., Phys. Rev. Accel. Beams 19, 024701 (2016)

Reliable results at the 10% level

Meson Production Target

Muon Rate: 4.6E8 P+/sec @ p=29.8 MeV/c

T.Prokscha et al NIM-A (2008)

Muon Transport Channel PE4 target, d=40mm

solenoids

quadrupoles

TARGET CONE Mean diameter: 450 mm Graphite density: 1.8 g/cm3

Operating Temp.: 1700 K Irrad. damage rate: 0.1 dpa/Ah Rotation Speed: 1 Turn/s Target thickness: 40 mm 7 g/cm2

Beam loss: 12 % Power deposit.: 20 kW/mA

M.Seidel, J-PARC, Oct 2015

protons

Target wheel of TgE station

40 mm polycrystalline graphite~40 kW power depositionTemperature 1700 KRadiation cooled @ 1 turn/sBeam loss 12% (+18% from scattering)

Performance of Standard TargetsStandard targets are as efficient in generating 28 MeV/c surface muons as spallation targets

1.0x 1.1x

1.5x1.4x

Surface muon rates in μ+/s for TgEgeometry of different lengths

7

Length Upstream Downstream Side

10 1.4 1010 9.0 109 1.8 1010

20 1.6 1010 1.2 1010 5.1 1010

30 1.9 1010 1.1 1010 8.5 1010

40 1.8 1010 1.1 1010 1.2 1011

60 1.8 1010 1.2 1010 2.1 1011

TABLE I. Surface muon rates in µ+/s for all muons withmomenta below 29.8 MeV/c emitted from the various sides ofTarget E for various lengths of the target in mm. The valuesfor the side rates correspond to a single side only.

zx

y

FIG. 9. Di↵erent geometries studied in our target optimiza-tion. From left to right: grooved target, trapezoidal target,fork target, rotated slab target. The red line marks the protonbeam.

cused on methods of either increasing the surface volume291

(surface area times acceptance depth) or the pion stop292

density near the surface. Each geometry was required to293

preserve as best as possible the proton beam character-294

istics downstream of the target station. The muon beam295

extraction directions considered here are sideways, back-296

wards, and forwards with respect to the proton beam.297

The accepted phase space used in our simulations roughly298

corresponds to the acceptance of the following beam-299

lines at PSI: µE4 (sideways at 90) with a maximum300

surface muon intensity of 4.8108 µ+/s [21], E5 (back-301

wards at 165) with a maximum surface muon intensity302

of 1.1 108 µ+/s [] and E1 (forwards at 8) with a303

maximum surface muon intensity of around 106 µ+/s []304

[22]. All enhancements listed below are relative to the305

standard target geometry described in Sec. IV. A model306

of each geometry investigated is shown in Fig. 9.307

The first geometry explored is a radially grooved tar-308

get whereby many grooves are place in the target surface309

parallel to the proton beam direction. The basic idea310

behind this geometry is to increases the available sur-311

face area for surface muon production. No significant312

improvement over the standard target was observed (see313

Table II). While the grooves increased the geometric sur-314

face volume by up to 45% not all of this volume is useful315

for small angular acceptance beam lines as the surface be-316

comes too steep for the surface muons with their limited317

range to still exit the surface volume. This can be seen in318

Fig. 10 which shows the initial positions for accepted sur-319

face muons. Instead of the expected half circular shape320

the distribution takes on a crescent form thereby reducing321

the surface volume gain from the grooves. It can actually322

x [mm]-6 -5 -4 -3 -2 -1 0

y [m

m]

-3

-2

-1

0

1

2

3

0

2

4

6

8

10

12

14

16

18

20

22

24

FIG. 10. Initial positions of accepted muons from the groovedtarget zoomed in to one groove. Instead of the expectedhalf circular shape the distribution takes on a crescent formthereby reducing the surface volume gain from the grooves.

be shown analytically that for an acceptance with zero323

angular opening no geometrical changes to the surface324

will lead to an enhancement in the surface muon yield.325

This is also approximately true for typical beamlines.326

The small enhancement factors still achieved for the327

grooved target stem from the fact that the pion stop328

density is not constant throughout the target. Figure 11329

shows the pion stop density through the target from one330

side to the other and integrated along its length. While331

the pion stop density is lowest at the sides where surface332

muons can actually escape the target it is approximately333

70% higher in the center. This is due to the fact that334

the lowest energy pions with only small ranges in the335

target are stopped very close to the proton path thereby336

leading to a high stopping density. Higher energy pions337

– despite being produced more copiously – are stopped338

over a larger area around the proton path and lead to a339

reduced pion stop density.340

The second geometry investigated is a trapezoidal tar-341

get with a initial transverse width of 4 mm that increases342

linearly to 6 mm at the downstream end. The basic idea343

behind this geometry is to exploit the higher pion stop344

densities close to the center of the target while still pro-345

viding the full target length for the bulk of the protons346

and a somewhat reduced length for the halo of the proton347

beam. This geometry resulted in a 15% enhancement to348

muon rates at 90 to the target, but a 2% loss to the349

backward direction (see Table II). The loss in the back-350

wards direction is due to the much reduced area of the351

backwards face of the trapezoid target that cannot be352

recovered by the gains from the side face. The geometry353

performs even worse for the forwards direction for which354

the surface muon contribution from the side face is much355

reduced.356

To resolve the ineciencies of the trapezoidal target357

and better preserve the proton beam characteristics, a358

forked target was investigated such that the full proton359

After extensive target simulations:Slanted targets are even better!

!26

Test of HiMB Slanted Target

current TgE slanted TgE

Test of slanted target at TgE station

To be installed in week 48 (Nov. 25th) for the remainder of the accelerator period

Mechanical and thermal simulations completed and no show-stopper found

A-A ( 1:2 )

A

A

1

1

2

2

3

3

4

4

5

5

6

6A

A

B

B

C

C

D

D

8°

Protons

Test of HiMB Slanted Target

TgE

+30%

+30%

+60%

+35%

Test of HiMB Slanted Target30-60% gain in surface muons expected

Pions and cloud muons unaffected

Verification of expected beam intensities and profiles

Split Capture Solenoids

!29

Two normal-conducting, radiation-hard solenoids close to target to capture surface muons

Central field of solenoids ~0.35 T, field at target ~0.1 T

500 mm 250 mm

solenoid500 mm aperture

500 mm250 mm

solenoid500 mm aperturep

Solenoid Beamline: HiMBSource1.2 x 1011 μ+/s 1.3 x 1011 μ+/sTgE TgM*

Capture7.2 x 109 μ+/s C ~ 6%

3.4 x 1010 μ+/s C ~ 26%

Transmission5 x 108 μ+/s T ~ 7%Total ~ 0.4%

1.3 x 1010 μ+/sT ~ 40%Total ~ 10%

Existing μE4 beamline

Proposed solenoid beamlineGain due to high capture

and transmission efficiency

Solenoid Beamline: HiMB@EH

Schematic of the layout in the experimental hall

Still a lot to do!Beam:

Final beam opticsBeam spot at final focusPerformance of separator…

Target:Slanted target within small gapChange of target stationNew shielding and beam channelsDisposal of highly radioactive material…

But exciting prospect and certainly worth the effort!

MuCool: cold muon beam

standard muon beam

our muon beam

IN OUT

phase space compression: 1010

efficiency: 10-3

MuCool principle

1st stage

2nd stage

→ Transforms a standard beam into a high-brightness low-energy beam

D. Taqqu, PRL97, 194801 (2006)Y. Bao et al., PRL 112, 224801 (2014)

IN

OUT

MuCool prototype (1st stage)

Target

Prototype exists and operated successfully!

HIMB + MuCool• HIMB: 1010 µ/s

• MuCool ε ≈ 10-3

• Together: 107 µ/s of < 1 eV muons

• Those muons can be re-accelerated

• Think of this

SINQ

TgE

TgM

Not mentioned but not forgotten:Proton Irradiation Facility (PIF)

Conclusions• PSI is home of unique beamlines:

• intense p beam ≥ 2 mA

• intense π and µ beams → MEG, Mu3e, MUSE, …

• ultra cold neutrons → n2EDM

• Plans for an upgrade to 1010 µ+/s: HiMB

• Feasibility study finished by end 2020

• MuCool: < 1 eV muons at high rate

IMPORTANT NOTICEThe following slides are considered as additional material beyond the scope of this presentation. No guarantee is made that

you may understand this without the context of a proper presentation. So use with caution or contact the speaker. ;)

Backup slides

Andreas Knecht NUFACT2017, 25. - 30. 9. 2017

Alternate MaterialsSearch for high pion yield materials → higher muon yield

relative µ+yield ∝ π +stop density ⋅µ+Range ⋅ length

∝n ⋅σπ + ⋅SPπ + ⋅1

SPµ+

⋅ ρC (6 /12)Cρx (Z / A)x

∝Z 1/3 ⋅Z ⋅ 1Z⋅ 1Z

∝ 1Z 2/3

39

p

π+μ+

Andreas Knecht NUFACT2017, 25. - 30. 9. 2017

Alternate Materials

Several materials have pion yields > 2x Carbon

Relative muon yield favours low-Z materials, but difficult to construct as a target

B4C and Be2C show 10-15% gain

40

Atomic Number Z0 10 20 30 40 50 60 70 80 90

Rel

ativ

e Su

rface

Muo

n Yi

eld

0.5

1

1.5

2

2.5 Pion YieldRelative Muon Yield /p

/cm

]+ π

Pion

Yie

ld [

0.001

0.002

0.003

0.004

0.005

0.006

0.007Osmium

Carbon

Nickel