A GEM Detector System for an Upgrade of the CMS Muon Endcaps

where we have a higher percentage of CF4 than CO2. A more detailed study is underway for the effect of CF4
gas in GEMs and other detectors.
4.4 Avalanche simulation
The studies done here use “single-electron” avalanches as also described in the paper[19]. We generate a stack
of single electrons randomly at a distance of 250 µm in the z-direction, from the GEM. The plane lies in the x-y
direction. The electron is given an initial energy of 0.1 eV and a random initial direction of motion. Then one
of the processes listed above is selected at random, weighed by their relative probability considering the electron
energy at the time of the collision. An additional electron is added to the stack in case of Penning transfer or
ionization; the electron is removed from the stack in case it is attached or attempts to leave the drift medium. The
process is repeated until the stack is exhausted.
The total gain, Gtot which is the total number of electrons produced in the avalanche, differs from the measured
or effective gain Geff because the latter is derived solely from the current in the anode plane. Avalanche electrons
which terminate on the GEM electrodes or insulator do not contribute to the anode current and henceGeff is always
smaller than Gtot. The “single-electron” avalanche simulation has been repeated at least 1000 times for varying
GEM voltages and Penning parameters. All simulations were performed at standard temperature and pressure, i.e.
T0 = 293.15 K, P0 = 1atm = 760 Torr.
Some of the primary electrons are lost due to attachment to the gas molecules (attachment loss) or to the metal
(geometric loss). The losses due to attachment and due to geometric losses can be seen in Fig. 39. The attachment
loss as expected is independent of the GEM potential difference, while the geometric losses decrease with increasing
GEM potential (or higher electric field), as electrons are accelerated and sucked into the GEM hole faster with
increasing electric fields. This leads to lower geometric losses. The attachment loss is∼ 12% in this gas mixture.
Figure 39: Loss rate for primary electrons for rp = 0.6. The losses occur due to attachment to the quencher gas
molecules (left) and due to the electrons hitting the surface of copper metal or polyimide walls (right)
The electrons which make it into the GEM hole give rise to the avalanche. All of these secondary electrons do
not contribute to the signal as some are lost on hitting the walls of the polyimide or hitting the copper metal and
some are lost due to attachment to the gas molecules. The losses due to attachment and due to geometric losses can
be seen in Fig. 40. The geometric loss increases with increasing GEM potential because the electrons exiting the
hole are attracted back by the lower copper metal which is positive with respect to the upper copper metal plate.
The secondary attachment loss reduces to ∼ 8% as the electrons gain energy inside the GEM hole due to the high
electric field inside the hole.
39