Backgrounds from Neutrons, Photons and charged particles.

Due to the location of GE1/1 (very forward region) the contribution of neutron background is something important to consider, such neutrons are produced in the pp interactions, when captured by nuclei can lead to photons emission and those photons can generate electrons (by Compton or photoelectric effect), such electrons can cause hits in the detector. The neutron background is hard to simulate, they can live up to a second which goes beyond GEANT4 simulation time limit in CMSSW, therefore special standalone simulation packages are needed to estimate the rate of those background crossing our volume of interest

Available simulation Tools.

FLUKA WebTool

FLUKA is a general purpose tool for calculations of particle transport and interactions with matter, FLUKA can simulate with high accuracy the interaction and propagation in matter of about 60 different particles, including photons and electrons from 1 keV to thousands of TeV, neutrinos, muons of any energy, hadrons of energies up to 20 TeV. FLUKA can handle very complex geometries. Inelastic cross sections for hadron-hadron interactions are represented by parameterised fits based on available experimental data. CMS collaboration has developed a web based tool in which the CMS users can specify the volume of interest, the particle id, luminosity, etc.. and the package returns the flux in Hz/cm2, such package can be found in the following link. This package is a nice one and the recommended to get fluxes values, but it does not give information about the energy of the particles, Time of Flight, position, etc..

FOCUS (FLUKA for CMS users)

Package recently developed to simplify the analysis of CMS users, it includes the geometry description (same as FLUKA webtool) and few routines which allow the user to simply specify the region of interest in the detector (in R and z positon), the beam energy, the number of primaries and as a result an output file with all particles inside the volume of interest is printed out, the first column refers to the primary number "event number", the second column is the particle id (for instance neutrons(8), photons(7)), the third column is the kinetic energy, the 4th,5th and 6th is the particle position in x,y,z. The package was released recently and is being validated by users, nevertheless it must provide same information than expected using normal FLUKA simulation.

• Output of a FOCUS job:
  

Geometry used by FLUKA and FOCUS

CMS Geometry is implemented in the FOCUS package (CMSpp 'nominal' geometry v.1.0.0.0), the materials and dimensions are specified in such a way to reproduce the real conditions in the experiment, the image below is the visualization of the CMS geometry given by FLAIR (advanced user interface for FLUKA). The second image shows a zoom to the GE1/1 area, it is important to notice that the GE1/1 detector dimensions and properties are not specified in the geometry description, instead the developers include some material between HCAL and ME1/1 to simulate the properties of an RPC detector, the assumption is that given the small width of GE1/1 and the lacking of good description of sensitivity in FLUKA the fluxes will not be greatly affected if the material in such region is "air", gems or rpcs, later we combine fluxes with properly GEM sensitivities simulated with GEANT4.

• CMS Geometry used by FOCUS:
  

• ME1/1 region in FOCUS:
  

Method to get the FLUXES+SENSITIVITY for a 8 partition GEM chamber

The same method applies for neutrons, photons and charge particles.

• Use the FLUKA Webtool to get the averaged flux in GE1/1 volume for each of the 8 eta partitions whose dimensions are taken from here, it is worth mention that the 1st partition is considered as the one with highest value in eta (as this may not be truth as the convention used inside CMSSW).
• Get the energy spectrum for using the FOCUS package, as we will use this energy spectrum as a template and normalized to the fluxes in previous point one of the important checks is to verify that the distribution shape is same as a function of eta (check was done and similar shape was observed, therefore take the one with highest statistic as the template distribution).
• Apply the GEM sensitivity (different for neutrons, photons and charged particles) as a function of energy for each of the partitions to get an "effective" rate.
• Insert this "effective" rate into the GEM digitizer to study the impact of the noise in the hits

First consideration (removing overlap between FLUKA and GEANT4 simulations)

In order to count only backgrounds coming from non-prompt processes, the time of flight of each particle has to be considered, in the usual GEANT4 simulation particles are simulated up to 500ns starting from the pp interaction, in reality only hits produced in about 250ns can be considered as prompt signals, therefore in our FLUKA studies we will only considered particles above this threshold, in the plot below it can be seen the TOF for each of the particles.

* Time of Flight for each neutrons, photons and charged particles:

Results

Fluxes with FLUKA webtool

The plot below shows the neutron flux for the 1st partition of the GE1/1 chamber, as can be seen the flux is averaged over the entire volume

* neutron flux in 1st GEM partition using FLUKA webtool:

• Neutron flux as a function of R:
  

• Neutron energy spectrum for each of the 8 partitions:
  

• Final table with results for all particles (including sensitivities):
  

Geant4 Simulation

The Geant4 version used to produce the following results is Geant4.9.6.p02.

The simulation source code is available online thanks to the CERN SVN Service. To check out the latest version, please use the following command in a UNIX terminal:

svn co svn+ssh://username@svn.cern.ch/reps/TrGEMG4

Goal of the simulation

To achieve this result, two steps are necessary:

• Geant4 simulates the response of a TripleGEM to various sources of photons and neutrons as a function of the energy of the incident particles;
• The Geant4 output is convoluted with the background flux of photons and neutrons.

This provides a rate dn/dE as a function of the incident particle. To obtain a single value of the flux expressed in Hz/cm, the final step consists in the integration of this rate over the energy.

Features of the simulation

Physics Lists

The Physics List used for the photon runs is QGSP_FTFP_BERT. It contains the following modules:

• G4EmStandardPhysics
• G4EmExtraPhysics (photo‐nuclear reactions)
• G4DecayPhysics
• G4HadronElasticsPhysics (elastic scattering for all long‐lived hadrons)
• HadronPhysicsQGSP_FTFP_BERT
• G4StoppingPhysics (nuclear capture at rest of negatively charged particles)
• G4IonPhysics (inelastic processes and models for the deuteron, triton and alpha)
• G4NeutronTrackingCut (neutron tracking manager)

For the neutron runs FTFP_BERT_HP was used. It contains the following modules:

• G4EmStandardPhysics
• G4EmExtraPhysics (photo-­nuclear reactions)
• G4DecayPhysics
• G4HadronElasticPhysicsHP
• HadronPhysicsFTFP_BERT_HP (important: G4NeutronHPBuilder takes care of neutrons 0 MeV < Kinetic Energy < 20 MeV)
• G4StoppingPhysics (nuclear capture at rest of negatively charged particles)
• G4IonPhysics (inelastic processes and models for the deuteron, triton and alpha)

Other information about the single modules contained in the Physics Lists can be found here.

Geometry used

The geometry used in the Geant4 simulation consists in a sequence of layers made by different materials. No external frame has been considered so far.

The left half of the table represents the currently used geometry. However it needs to be corrected in order to match the TripleGEM blueprints. These supposedly correct measures are reported on the right half of the table.

Definition of a sensitive event

Without an appropriate user-made Physics List (that would include the Photon Absorption Interaction model), Geant4 is unable to provide the correct number of electron-ion clusters, as well as its distribution, that takes place inside an ionized gas. However, a different quantity that we can trust is the deposited energy insider a material. Therefore, knowing the ionization potential leads us to the computation of the number of electron-ion pairs "by hand".

For the three gases composing the mixture, I use the following yields for the ionization potential:

Gas	Ionization Potential	Source
Ar	15.8 eV	J. Phys. B 32 L511-L516 (1999)
CO	13.78 eV	J. Electron Spectros. Relat. Phenom., 1988, 47, 167
CF	15.9 eV	Arxiv 0810.0445

Having the mixture Ar:CO2:CF4 the proportions 45:15:40, it is easy to find the average ionization potential of the mixture:

0.45*15.8 eV + 0.15*13.78 eV + 0.4*15.9 eV = 15.54 eV

Our decision has been to consider that if at least 5 times this energy has been deposited into the Drift Gap, the event is to be considered as sensitive. Otherwise, we look for the same condition inside Transfer Gap 1. The other Gaps are not considered in this algorithm.

Results

Here are the results for the sensitivity of the TripleGEM for both photons and neutrons cases.

Photons results

• Photons sensitivity results:
  

Neutrons results

• Neutrons sensitivity results:
  

The left plot shows the fraction of sensitive events, with a comparison between TripleGEM (black) and RPC (red) simulations, for the case in which the source of neutrons for TripleGEM is a parallel beam. The right plot shows the same result for a spherical isotropic source.