Gas Electron Multiplier (GEM) application for Time Projection Chamber (TPC) gating

A BSTRACT : A voltage-controlled Gas Electron Multiplier (GEM) can be used to block the re-injection of positive ions in large volume Time Projection Chambers (TPC). Through an accurate choice of proper geometry, gas ﬁlling and external ﬁelds it is possible to obtain a sufﬁcient level of electron transmission at very low GEM voltages (Gating GEM) despite the degradation of energy resolution due to the loss of primary electrons. The addition of a pre-ampliﬁcation GEM in front of the Gating GEM causes an improvement in energy resolution while keeping the ion feedback at the level of primary ionization. The measurements show that a small pulse of about 40 V completely closes the gate, stopping the ions produced in the ampliﬁcation stage.


Introduction
The Gas Electron Multiplier (GEM) [2] , consists of a thin Kapton insulating foil copper-clad on both sides and perforated by a high density, regular matrix of holes (50 to 100 per square millimetre). The distance between holes (pitch) is typically 140 µm and the diameter is of about 70 µm. Upon an application of a potential difference between the GEM electrodes, a high dipole field develops in the holes, focusing the field lines between the drift electrode and the readout element. Electrons drift along the channel and the charge is amplified by a factor that depends on the field intensity and on the length of the channel.
Large volume gaseous Time Projection Chambers (TPCs) suffer from the problem of ionbuilt space charge. The amplification structures (nowadays still MPWC) amplifies the primary electrons created by the interacting particle and consequently generates ions that slowly drift back to the drift volume where they modify the electric field thus changing the time-space properties of the next track. Next generation TPC endcap will maybe be equipped with Micro Pattern Gas Detector (MPGD) such as GEM and Micromegas ( [4][5][6]). Gating electrodes are needed to prevent the re-injection of positive ions into the large gas volume and during LEP (and LHC) period the wire-gating technique was extensively used. The next generation gating structures can be made by MPGDs. When a GEM foil is powered at very low potential difference (from 10V up to 40V) [1] it does not act as an electron amplifier device. Its electron transparency (the ratio between the number of electrons that are able to pass trough the GEM holes and the numbre of electrons present in front of the top GEM electrode) is reduced to few tens of percent depending on the applied potential difference, on the external fields, on the GEM geometry and on the chosen filling gas. A voltagecontrolled Gas Electron Multiplier (GEM) powered at low potential difference can be used to block the re-injection of positive ions in large volume Time Projection Chambers. A gated pulse that inverts the GEM potential difference stops all the ions produced in the amplification stages below the gating GEM. In this paper we investigated the basic physics processes under this application but real pulse gating will be matter of future work.  Figure 1 shows the small TPC scheme. Two different configurations were studied: in the first one the first GEM foil was used as Gating GEM; in the second one another GEM (Preamplification GEM) foil was added in front of the Gating GEM in order to get a preamplification factor. The final amplification stage was represented by a double GEM structure (Bottom GEMs). All the GEM foils used for the measurements are standard GEM foils (50 µm hole diameter, 140 µm pitch); the drift cathode was a metallized mylar plate and the anode was a full copper plane. The pulse height spectra were acquired using an ORTEC 142-IH preamplifier and an ORTEC 450 research amplifier; the current was read-out by means of a Keithley 6517A picoamperometer (∼ 1 pA resolution). The detector performance was studied by means of a collimated 8.9 keV X-Rays beam impinging either orthogonally or parellely to the drift cathode. When the direction was orthogonal to the drift cathode, the X-Rays could convert in three different detector gaps: in the drift gap (A region), in the transfer 1 gap (B region) or in the transfer 2 gap (C region); when the direction was parallel to the drift cathode, the conversions happened only in the drift gap (A region, see figure 8 ). The gas mixture used in all the characterization was Ar/CO 2 70%/30%.

Gating GEM transparency measurements
When a low potential difference (from 5V to 50V) is applied to a GEM foil [1], it does not act as an amplification device but rather it absorbs part of the charge that approaches it and its electron transparency (ε GatingGEM , the ratio between outcoming and incoming electrons) is not 100% but depends on the ∆V GEM and on the external electric fields [3]. Figure 2 shows that the measured electron transparency has a value around 30% for 10V ≤ ∆V ≤ 40V . The Gating GEM electron transparency was obtained by the ratio between the B pulse height peak position and the C pulse height peak position (see figure 3). The energy resolution is worsened by the presence of the Gating GEM (from 20% FWHM up to 60% FWHM) and this effect is clearly visible in figure 3, that shows the pulse hight spectra acquired for different ∆V GatingGEM .   to get an improvement of energy resolution, another GEM foil (Preamplification GEM) was added before the Gating GEM. This is an useful method to preamplify primary electrons before loosing part of them in the gate electrode. The typical PH spectra obtained when the Preamplification GEM was operational is shown in figure 4.
As it is shown in figure 5, the ∆V applied on the Preamplification GEM varies between 350 V and 430 V, giving a gain (measured as G Preampli f icationGEM = A PHPeakPosition B PHPeakPosition ) between 2 and 5.5. This gain was also obtained through the optimization of external (E T 1 , E T 2 ) electric fields.
The aim of this exercise was to get a unitary first stage gain (G FirstStage = G PreampGEM * -4 - ε GatingGEM ) by adjusting ∆V PreampGEM . As it is shown in figure 6 this result was obtained with ∆V PreampGEM = 390V . It was also proven that the addition of the Preamplification GEM (figure 5) improved the energy resolution by a factor of 2, from 60% FWHM down to 30% FWHM.

Gating GEM voltage scan with preamplifier GEM
In order to prove the gating properties of the Gating GEM, a counting mode ∆V GatingGEM scan was made and the corresponding photon interaction rate was recorded. As it is shown in figure 7 the potential difference applied to the Gating GEM varied from -20 V up to +20V (where the positive : this step demonstrated that the gate is completely closed with a very small potential difference, from -10 V to -5 V. This is a very promising feature for high rate pulsed gate in TPCs. The closed gate plateau is not on the zero level because of X-Rays conversion in the C region.

Preamplification GEM Ion back flow measurement
In a real TPC operation all the ions produced in the amplification stages below the gate will be stopped by closing the gate. Being the Preamplification GEM placed above the gate, it is crucial to understand how many of the ions produced in this stage will drift back to the large gas volume. The setup used to measure the Preamplification GEM ion feedback is shown in figure 8:  Figure 9 shows that in the scanned Preamplification GEM voltage range with a fixed drift field (E d ) value of 0.1 kV/cm, a NIF around 2-3 was measured, the same order of magnitude of primary ionization current. The conclusion is that the ion feedback contribution introduced by the preamplification stage is not much higher than the primary ionization itself.

Amplification stage voltage scan
The experimental setup shown in figure 8 was used in order to understand if the Gating GEM was working independently from the amplification stage. The measurement of the NIF was performed changing the potential difference on one of the two amplification GEM (∆V BottomGEM2 ), keeping the same ∆V PreampGEM = 390 V and closing the gate (∆V GatingGEM = −20V ). Figure 10 shows that the measured NIF corresponds to a value of 2-3: this number underlines that the NIF is only due to the Preamplification GEM. If the gate is opened (∆V GatingGEM = +20V ) the point @ ∆V BottomGEM2 = 400V will correspond to a Gain FullDetector of about 3000. Figure 11 shows a ∆V GatingGEM scan that summarizes the prototype behaviour.

Full detector behaviour
When the gate is completely closed (∆V GatingGEM = −20V ) the overall gain ( I readout I ionization ) is zero (no electron can get to the amplification stage) and the NIF is 2-3 and this corresponds to the PreampGEM NIF. When the gate is open (∆V GatingGEM = +20V ) the overall gain is around 3000 but it can be increased if one of the two amplification GEM voltages is increased.

Conclusions
The standard GEM foil can be used as a gate electrode for a TPC when a low potential difference is applied to its electrodes. A small pulse (20V-40V) is sufficient to open and close the gate, giving the possibility to have a very high rate pulsed gating. The measurements prove also that the energy resolution is improved by the addition of a properly operated Preaplification GEM in front of the Gating GEM and that the Preamplification GEM does not largely contribute to the Normalized Ion Feedback, NIF (Preamp GEM NIF = 2-3). In addition, since all the ions produced in the last stage are stopped by the gate, the amplification stage gain can be as high as required by the experiment.