Secondary Emission Calorimetry R&D

CHEF2013, Calorimetry for High Energy FrontierApril 22-25, 2013 Paris, France

Burak Bilki (for David Winn)

University of IowaArgonne National Laboratory

2

Why Secondary Emission Ionization Calorimeters?

• Secondary Emission (SE) signal: SE surfaces inside em/had showers:

– SE yield Scales with particle momentum

– e-: 3 < <100, per 0.05 <e-<100 keV (material dependent)

– ~ 0.05 - 0.1 SEe- per MIP

• SE is: Rad-Hard + Fast

– a) Metal-Oxide SE PMT Dynodes survive > 100 GigaRad

– b) SE Beam Monitors survive 1020 MIPs/cm2

Ex: ~60-240 SEe- per 100 GeV pion shower w/ MIPs aloneBUT SEe- Must be Amplified! Exactly like photoelectrons (p.e.)!

3

An SEe is statistically exactly like a p.e.

g~105–106 e per SEe (SEe as generated by shower particles) (just like p.e. gain from collected photons w/ g=105-6)

A) Muon MIP:~ 0.1 SEe- per sample module x 25 ModulesMIP Signal: We predict ~2.5 SEe/Muon (in 25 X0 EM module)

but we measure ~8x this in Test Beam!

B) EM Showers:~900 shower electrons/GeV yields 45-90 SEe/GeVNB: Lower Estimates! MIP SE, not true SE from particles

In CMS Hadron Forward Calorimeter Phase II Upgrade Framework

25 X0 Forward EM Calorimeter1 X0 W/Fe + SEe Sensor ~ 1.5 – 2.5 cm thick

4

SEe Dynodes: a) Etched Metal Sheets

Hamamatsu Dynodes15 cm now ~50 cmAlready diced from largesheets

5

Dn-Dn+1: 0.9 mmC-C mesh: 13 µmWire diameter: 5 µm

MESH DYNODE VARIANTS

SEe Dynodes: b) Metal Screen Dynodes: 15D - g~105

6

Beam Tests of SE Sensor

We Expect ~500 Shower electrons to Cross Mesh ~25-50 SEe assuming all shower e = MIPs

Mesh PMT and BaseFacing DownstreamPhotocathode Reverse Bias

100 GeV e-

3 cm Pb

19 Stage Mesh

CERN SPS, Oct. 2011

7

BEAM TEST: 100 GeV electrons 3 cm Pb ~ 5 X0 Radiator ~ Shower Maxdownstream the mesh PMT w/ photocathode turned off

Peak corresponds to ~40 SE electrons (mesh stack gain ~105)

PRELIMINARY!

Fluctuations High!- PMT Dia ~ Shower

Dia- Beam not centered

NOTE µ MIP:Detection Eff ~10%Response ~1-2 “SEe”

CERN SPS, Oct. 2011

8

100 GeV Electrons – SE Mode

Charge > 160 fC cut applied. (1 pe ~ 160 fC) Scales with X0. Note: Shower not laterally contained!

~4 X0

CERN SPS, Oct. 2011

9

80 GeV Electrons – SE Mode

Charge > 160 fC cut applied. (1 pe ~ 160 fC) Scales with X0 Note: Shower not laterally contained!

CERN SPS, Oct. 2011

10

Muon Efficiency ~ 80%(1 pe ~ 160 fC)

Muons pass through the mesh dynodes (selected by the wire chambers)

Below Thresh.

SEe Efficiency with Muons CERN SPS, Oct. 2011

11

SE Module Beam Test

51, … , 59: PMT IDs

Beam position(into the page)

2 datasets

Selected with Wire Chamber

515253

545556

575859

Using mesh dynodes from PMTsCERN SPS, Nov. 2012

12

SE Module Beam Test

80 GeV e- Beam Variable absorbers

0 – 9 X0

SE SE SE SE

Shower not contained laterally or longitudinally Results require estimates and approximations

2-cm iron absorbers:

X0 = 1.75 cmMolière Radius: 1.72 cm

13

SE Module Beam Test

No upstream absorber80 GeV e- Beam hitting 1-51

Different measures of efficiency possible, here Charge > 45 fC (pedestal RMS) is chosen.

Efficiency ~ 41%

14

SE Module Tests – Preliminary Results

Normalizing responses of different layers

Example: Normalization of Layer 3 response to Layer 2 response using 7X0 sampling

(Also works in the reverse order next slide)

Charge > 20 fC

15

SE Module Simulation

Geant4 simulation of the SE module test beam setup:80 GeV e- beam19 stage mesh dynodes generate SE electrons (dynodes ~ sheets)Gain is simulated offline (106)Landau fluctuations are implemented offlineSingle parameter to tune: Efficiency of SEe production(mesh dynodes are simulated as solid sheets) 0 - 0.35% flat random

16

25 X0 sampling calorimeter1.75 cm Fe absorbers19-stage SE sensor ~ 2 cmLateral size 1 m x 1 m

SE Calorimeter SimulationUsing SE module MC tune

Landau fit MPVs and MPV errors

17

Conclusions

Secondary Emission calorimeter is radiation-hard and fast.

Progressive beam test results are encouraging for better prototypes.

More beam test needed Next generation calorimeter prototype in preparation.

SE calorimetry is feasible for large-scale applications.

Large implementation options once the proof of concept is established (forward calorimetry for hadron/lepton colliders, beam monitors, Compton polarimetry for lepton colliders, etc. ).

18

Backup Slides

SE: ionization e- random walk to surface oxide (10-50 nm thick), from ~1-2 µm in the underlying metal, escape into t vacuum under applied E-field = “secondary e-” = SEe.

SEYield gb -Semiempirical SternglassL = Escape Depth, Ep, Ap = particle energy, mass

Proton SE yield vs Ep

X200 peak-valley

Secondary Emission

20Shower +/- electron Energies

GEANT4: Cu Block, 1cm “plates”,100 GeV e incident. Shower e+/- that cross the 1 cm “gaps” arebinned in both energy and depth in Cu

21

10 GeV e

100 GeV e

~900 e+-/GeVCross 1cm Cu gaps ~Indep of Einc

Yields 45-90 SEe/GeV generated.(assume all mips)

GEANT4: e’s in Cu1 cm absorber PlatesCount e+/- crossing gapsdNe/dE vs Shower e+- KE

Shower e+- energies crossing gaps

SEe Detector Module Concept -HCal

3D Stackable x,y,z

22

Example SE: CERN LHC Beam Loss Monitor

24

Al2O3,TiO2

~0.05 e- per MIP. Yet used as beam monitors:Rad Hard!

HadronShowers:Sub-mipCharged particles

Signal Generation from protons – DESY BPM