1 Roger Rusack – The University of Minnesota. Outline Physics Motivation: Review the physics that drives the design of the forward region. What.

1

WHAT FORWARD DETECTOR?Roger Rusack – The University of Minnesota

2

Outline

Physics Motivation: Review the physics that drives the design

of the forward region. What do we know and what do we not

know. Radiation environment:

Reminder of the radiation levels that we can expect to see at the HL-LHC.

Discuss some detector options. One example.

3

Precision Study of Higgs

Central priority of the field will be to study in detail the Higgs-like object seen at 125 GeV.

This means we will need a high precision detector capable of working delivering the same or better quality physics as we have now in CMS.

This includes understanding and measuring the VBF scattering with good precision.

4

WW VBF

To study this process we need the forward region.

If we are to use particle flow to measure the jet energy, we need good tracking, highly segmented calorimetry and muons momentum for η ≤ 3.0 (at least).

Dan Green

5

And

As we extend searches out to the highest masses missing ET will continue to be a critical tool in our searches for SUSY particles.

Assymetries in top physics production needs angular coverage out to the highest η.

All this means that the forward region will play a central role in the Physics of

the HL-LHC.The question now is how to design the best detector for this region at a price

that we can afford in time for installation in 2022.

6

What do we have now

7

The next Ten Years of the LHC

Mike Lamont May 2012.

2019 – 2021: Physics with ‘ultimate LHC’

parameter set

Parameters ‘Ultimate’

k (# of bunches) 2808

N (bunch intensity)

1.7*1011 p

β* 0.5 m

Luminosity [cm-2s-

1]2.4*1034

E[TeV] 7

E[MJ] 541

8

9

In 2022

Long Shutdown (LS3) Upgrade LHC Replace tracker. New track trigger. New trigger. Maybe replace ECAL barrel electronics. Replace endcap electro-magnetic

calorimeter. Replace active components of endcap

hadron calorimeter (maybe absorber). Upgraded detector will need to operate

at the highest luminosity and work effectively after operation up to 3,000 fb-1 in ten years.

10

Forward Detector

High eta regions will be more important then due to physics interests.

With ~5E34 cm-2s-1 luminosities can expect to have average pile-up of ~100 – 200 events.

Radiation levels will be significantly higher than experienced before in HEP experiments.

Approaching levels of radiation typical for a reactor.

11

Design Thoughts

The best jet energy resolution in CMS has been achieved with particle flow techniques.

Particle flow requires a combination of tracking, calorimetry and muons working together.

Currently muon coverage in CMS ends at η = 2.4, the tracker at 2.5 and the endcap calorimeters at 3.0.

The number of interactions per bunch crossing will be ~ 100.

12

And

Do no harm --- CMS is a very good detector and we should not make it worse by trying to improve it.

13

Some Known Knowns

Parts of the ECAL endcap detector will have large attenuations in the crystals at the end of phase 1 and EE will in all likelihood need to be replaced.

HE will suffer significant damage in the innermost regions, with the tiles going black.

HF will receive a dose of 10 MGy at the eta of 5.

Neutron fluxes in all the calorimeters will be very large.

14

More Known Knowns

Extending the muon coverage to η = 3.0 requires removing shielding that protects against neutrons from the high eta region.

The problem to solve is how to build a high-performance detector that will survive unprecedented levels of radiation.

Time is short. And so too is money.

15

Flux of neutrons after Phase 2.

M. Huhtinen: Neutron flux for 2,500 fb-1

SLHC Workshop 2004.

16

EE Radiation Levels.

> 1015 neutrons/cm2

> 1014 neutrons/cm2

> 105

Gy

> 3.105

Gy

17

Light Yield Losses in PbWO4

18

What about HF? Current estimates are that it will

survive with light loss. But:

The neutron flux at 2,500 fb-1.

Current data indicates there will be light losses

19

and

Region LS1 EYETS 16-17 LS2 LS3 LS4

Tracker BH 12 mSv/h 35 mSv/h 50 mSv/h 65 mSv/h 125 mSv/h

EE(high h) 0.25 mSv/h 0.75 mSv/h 1 mSv/h 1.25 mSv/h 2.5mSv/h

HF(high h) 5 mSv/h 15 mSv/h 20 mSv/h 25 mSv/h 50 mSv/h

TAS region ≤15mSv/h ≤45mSv/h ≤60mSv/h ≤75mSv/h ≤150mSv/h

Expt cavern <0.5 mSv/h <1.5 mSv/h <2 mSv/h <2.5 mSv/h <5 mSv/h

Radiation Levels at contact after one month after the start of LS3

So any work on the existing detector will need to be done with shielding of the personnel.

20

Detector Options/Ideas.

Tracker. Significant work on highly rad tolerant silicon

detectors – 3D silicon. An idea whose time has come?

Current plans are to cover out to η = 2.5. Discussion of possibilities to extend to higher eta for PF calorimetry.

Micromegas have been proposed for the far forward region.

Muons: Current detectors are expected to cope with the

rate. To extend to high eta – will need to use GEM

detectors or similar.

21

Calorimetry

Two main lines of thought: Build a new ECAL and fix the hadron

calorimeter. Replace the electromagnetic calorimeter

and re-furbish the hadron calorimeter with rad hard detector material.

Replace both with one homogeneous calorimeter.

22

Detector Material.

What detector materials do we know that will work at these ultra-high radiation levels. Amorphous silicon Quartz. 3D silicon – may be needed for the pixels at

high eta. Glasses. Liquids. LYSO ….

High precision timing.

23

Possible EM Calorimeter

Shashlik design with LYSO Main idea is that theWLS acts a light source, but does nottransport the light to the photosensor

The quartz performs that function

quartz

wls

Idea proposed by Randy Ruchti (Notre Dame)

24

New Photodetectors

First results from a GaAs SiPM photodetector

Lightspin & UVA

25

Silicon Carbide

Bandgap of 4H-c SiC is 2.32 eV. High bandgap material used in making

LEDs, now showing up more in semiconductor and nan0-technology industry. You can now buy JFETs and MOSFETs in SiC.

MOSFETs have been tested to 7.5 × 1014 neutrons/cm2 and were operational.

Significant degradation at 1016 n/cm2 seen by RD50.

Interesting detector development possibilities.

26

4H-SiC PIN480 Avalanche Photodiode: Recessed Window

PECVD + Thermal SiO2

n :+ substrate

n: 2000 nm, 4.5x10 18cm -3

p: 200 nm, 2x1018cm-3

p : 480 nm,- 1x1016cm- -3

p+: 200 nm, 1x10 19cm-3p-contact

(Ni/Ti/Al/Au)

n-contact(Ni/Ti/Al/Au)

Thickness of p+ : ~ 35 nmAR Coating (2300 Å)

Joe Campbell – Electrical Engineering – University of Virginia

Quantum Efficiency

27

28

Amorphous Silicon

Amorphous Silicon (a:Si-H) has been studied as a possible rad hard detector materila for several years. Idea is that it is like a severely damaged material to

start with, so changes after irradiation are relatively small.

Problem good materials had an effective collection depth of ~5μ and charge collection times of order 100 ns. Making an efficient a:Si-H tracker was not feasible.

Never studied for calorimetry where there is more charge deposition.

Now a:Si-H is a major industry . Making it a very cheap material.

Also new meta-materials based on a:Si-H have much larger mobility than standard material.

29

One Example Idea

Use Čerenkov signal and the ionization signal in hadron calorimetry. Not as in DREAM calorimeter for optimum

resolution, but for a restricted volume calorimeter.

Main idea in the CMS – HF: sample hadronic shower in EM core only.

Tag only EM core with Čerenkov and measure the ionization.

Benefit: EM core of a hadron shower is in a cone ~10 cm diameter, all ionization is in a cone 1 m diameter.

Significantly reduces overlap between showers.

30

Straw-man Idea:

Cerenkov Light Detection: Quartz plates with a layer of amorphous

silicon Use Silicon Carbide or GaAs APDs to detect

the Čerenkov light in the quartz. Couple signal to PCB – readout at outer

radius. Ionization detection.

Use a:Si-H as a readout material. Lots of technical problems with this idea,

but let’s look at how it would work.

31

Simulation:

Simulate ideal detector to understand detector performance.

Current on-going work lots of questions still to be understood.

Geant simulations done using CATS system developed here by Hans Wenzel.

All results presented here are from work at Minnesota by Peter Hansen - 0th year grad student.

32

CATS Simulation

CATS – http://home.fnal.gov/~wenzel/CaTS.html

1.01m x 1.01m, 1 cm square tiles 1.5 m total absorber width (brass) Separate runs with different absorber

25 mm, 50 mm and 75 mm per layer. Total absorber depth kept constant at 1.5

m. Each layer detectors:

1 cm2 tiles of 7 mm quartz & 5 mm scintillator.

Count only Cerenkov photons and ionization

33

100 GeV hadron shower

34

A 500 GeV Shower

35

Another 100 GeV Shower

36

Shower Size

Study spatial extent of the showers by defining square annuli and adding up the ionization and Čerenkov light inside the annuli.

Study resolution we get using energy from tiles where there is Čerenkov light.

37

Shower Size

50 mm absorber plates.

38

Resolution with 75 mm plates

1 10 100 10000.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1.000

Sigma/E for 75 mm absorber with pions

All ScintTagged ScintTagged Scint in WindowAll Scint in Window

Energy Gev

Sigma./E

39

Resolution with 50 mm plates

1 10 100 10000.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1.000

Sigma/E for 50 mm absorber with pions

All Scint

Tagged Scint

Tagged Scint in Window

All Scint in Window

Energy Gev

Sigma./E

40

Questions

Simulation has so far concentrated on the resolution that can be obtained. Need to study the tails in the distributions.

Explore further the benefits of tracking the hadron shower development.

Two particle separation. Optimize jet reconstruction with a

tracker. Evaluate performance inside the CMS

detector. Test structure in a test beam.

Can we reuse existing ones?

41

Conclusion

2022 is not that far away to design test and qualify a completely new calorimeter for CMS endcaps.

Technical challenges are large. Collaboration with the Fermilab calorimeter

R&D group would be very welcome.