30 March 2004 1

Global Mice Particle Identification

Steve Kahn30 March 2004

Mice Collaboration Meeting

Steve Kahn30 March 2004 Particle Identification Page 2

Particle ID Elements Upstream Detectors:

Time of Flight system Rely on the time difference between stations

TOF0and TOF1 to provide velocity measurement. Upstream Cherenkov System

Verifies that the track is a muon.

Downstream Detectors: Time of Flight system

Verifies the existence and its time for a track exiting the detector solenoid.

Steve Kahn30 March 2004 Particle Identification Page 3

Particle ID Elements Downstream Detectors (cont.)

Downstream Cherenkov System Verifies that the track seen in TOF2 is an electron. Not sensitive to muons.

EM Calorimeter Shows whether a track that is seen in TOF2 has

an EM shower or not.

Tracking Systems Needed to know the particle momentum.

Steve Kahn30 March 2004 Particle Identification Page 4

Steve Kahn30 March 2004 Particle Identification Page 5

Upstream TOF System Plots on the right show:

Upper plot shows time distributions at TOF0 and TOF1:

T=30 ns for these stations. Lower plot shows the transit

time of individual tracks. <T>=37.6 ns. T=176 ps The T corresponds well to

what we expect for with P=200 MeV

37.6 ns.

Obsolete

Steve Kahn30 March 2004 Particle Identification Page 6

Separating from in the Real World Tom Roberts has shown

an analysis using the Tof timing along with the momentum from the tracker to separate from .

Tof information is not sufficient by itself since there is some overlap in the and velocity distributions.

There are 17 in the lower plot; 5 of the overlap the distribution.

Steve Kahn30 March 2004 Particle Identification Page 7

Requirements for ToF Particle ID

The TofParticleID class will need: Access to TOF0 and TOF1 digitizations for the

same event (actually for the same track). Our ROOT structure keeps digitizations separate.

Knowledge of the Tracker reconstructed momentum.

All of the TOF0 hits in the pile-up time interval We need to estimate the likelihood that a time

coincidence is real or coincidental.

Steve Kahn30 March 2004 Particle Identification Page 8

The Upstream Cherenkov : Ckov1

C6F13 Radiator

CL

MIRROR

C6F14

PMT

UV

WINDOW

beam

mirror

PMT

0 20 40 60 80 100

Npe

e Candidates 186MeV/c 1cm C6F14

The figure shows clean separation of 186 MeV/c e, ,. However at 250 MeV/c we should expect significant overlap between and .

Steve Kahn30 March 2004 Particle Identification Page 9

Ckov1 Parameters and Status Status: we now see hits and digits for Ckov1. Radiator is C6F14 with nrefr=1.25

Thresholds are 0.7 (e), 140 () and 190() MeV/c respectively.

The Ckov1 alone will not be able to distinguish from since the velocities are close as we have seen.

The discrimination comes from the pulse height analysis, where the (dE/dx)Č is largely a function of alone.

With the knowledge of the momentum from the tracker we should be able to separate from .

The Ckov1 may be less sensitive to the background than TOF since TOF I is located in a position with lots of background.

Steve Kahn30 March 2004 Particle Identification Page 10

Ckov1 Particle ID Software Needs

The Ckov1 acts independent of the other particle ID detectors. Since it has a single PMT it cannot

measure the radius of the Cherenkov cone.

It will only measure a pulse height. It will need to know the momentum

from the tracker reconstruction.

Steve Kahn30 March 2004 Particle Identification Page 11

Electrons from Muon Decay are Present in the Downstream Track Sample

~1% of downstream tracks may be electrons, not muons.

80% of these electrons can be removed by kinematics, but this could bias the emittance measurement

Momentum Distribution

Angular Distribution

Spacial Distribution

Steve Kahn30 March 2004 Particle Identification Page 12

Downstream Detectors The figure on the

right shows the placement of the downstream detectors.

Both the Ckov2 and EMcal need a coincidence with the TOF III.

Steve Kahn30 March 2004 Particle Identification Page 13

Downstream Cherenkov Ckov2 is a threshold Cherenkov

The refraction index is nrefr=1.02 This corresponds to p>525 MeV/c for

muons But only pe>2.5 MeV/c for electrons If we see a signal in Ckov2 and TOF III it

is EM energy Single ionizing electron Twice minimum ionizing photon

Steve Kahn30 March 2004 Particle Identification Page 14

The Muon Beam Expands as the Field Falls Off

The calorimeter subtends 6060 cm

Steve Kahn30 March 2004 Particle Identification Page 15

Muon vs electron identification in EMcal

We have carried out simulation studies in G4MICE to optimize the mu/electron separation capabilities by varying: sampling fraction, i.e. lead layer thickness: 0.5-0.2 mm readout segmentation, i.e. cell size: 3.75x3.75 cm2, 3.25x3.25 cm2, 2.5x2.5 cm2, 2.5x4.0 cm2

We only consider a “pattern-based” identification algorithm, i.e. detection efficiency in layers >1

Stolen from A. Tonazzo

Steve Kahn30 March 2004 Particle Identification Page 16

Calorimeter signal, efficiency definition

Detection efficiency is defined by a cut on signal above noise threshold: 3-4 p.e.

PMT signalEnergy deposit“digitization”:•Light attenuation along fibers•Winston cone collection efficiency•Photocatode quantum efficiency

Stolen from A. Tonazzo

Steve Kahn30 March 2004 Particle Identification Page 17

Steve Kahn30 March 2004 Particle Identification Page 18

Steve Kahn30 March 2004 Particle Identification Page 19

Software Requirements of Particle ID Common Requirements:

Need to access information from more than one detector unit:

Upstream TOF requires two stations Downstream Ckov2 or Emcal also require TOF2 Need results of Tracker reconstruction.

Primarily the track momentum and error. Need to create an Event class that contains:

All detector results Digitizations for particle ID detectors Reconstruction for tracker detectors.

We are currently not organized that way.

Steve Kahn30 March 2004 Particle Identification Page 20

Particle ID Classes There should be a ParticleID class

for each ParticleID algorithm. There may be one or more than one

way to handle the information from each particleID detector system.

Each of these ParticleID classes should inherit from a ParticleIDBase class

Steve Kahn30 March 2004 Particle Identification Page 21

ParticleIDBase class Contain common quantities. Should contain:

Detector/Algorithm identification Reconstructed P, P from tracker for

track. Link to reconstructed track.

Probabilities of being ,,,e. Quality factor from algorithm

determination.