A 10-3 drift velocity monitoring chamber
F. Cuna, G. Chiarello, A. Corvaglia, N. De Filippis, F. Grancagnolo, M. Manta, I. Margjeka, A. Miccoli, M. Panareo, G. F. Tassielli
PPrepared for submission to JINST
International Conference on Instrumentation for Colliding Beam Physics24 - 28 February, 2020Budker Institute of Nuclear Physics, Novosibirsk, Russia A − drift velocity monitoring chamber F. Cuna a , b , , G. Chiarello d , A. Corvaglia a , N. De Filippis e , f , F. Grancagnolo a , M. Manta b ,I. Margjeka c , e , A. Miccoli a , M. Panareo a , b , G. F. Tassielli a , a Istituto Nazionale di Fisica Nucleare, Lecce, Italy b Università del Salento, Italy c Università degli Studi di Bari, "Aldo Moro", Italy d Istituto Nazionale di Fisica Nucleare, Roma, Italy e Istituto Nazionale di Fisica Nucleare, Bari, Italy f Politecnico di Bari
E-mail: [email protected], [email protected]
Abstract: The MEG-II experiment searches for the lepton-flavor-violating decay: µ −→ e + γ . Thereconstruction of the positron trajectory uses a cylindrical drift chamber operated with a mixture ofHe and iC H gas. It is important to provide a stable performance of the detector in terms of itselectron transport parameters, avalanche multiplication, composition and purity of the gas mixture.In order to have a continuous monitoring of the quality of gas, we plan to install a small driftchamber, with a simple geometry that allows to measure very precisely the electron drift velocityin a prompt way. This monitoring chamber will be supplied with gas coming from the inlet andthe outlet of the detector to determine if gas contaminations originate inside the main chamber orin the gas supply system. The chamber is a small box with cathode walls, that determine a highlyuniform electric field inside two adjacent drift cells. Along the axis separating the two drift cells,four staggered sense wires alternated with five guard wires collect the drifting electrons. The triggeris provided by two Sr weak calibration radioactive sources placed on top of a two thin scintillatortiles telescope. The whole system is designed to give a prompt response (within a minute) aboutdrift velocity variations at the 10 − level.Keywords: Drift chambers, Particle tracking detectors, Gaseous detectors, Models and simulations,Electric fields, Charge transport and multiplication in gas Corresponding author. a r X i v : . [ phy s i c s . i n s - d e t ] J un ontents The choice of a gas mixture in a drift chamber is of utmost importance, in particular for the experi-ments, like MEG-II, in which the trajectories of low momentum particles need to be reconstructedwith high accuracy. Moreover, it is crucial to control the purity of gas injected in the drift chamberbecause uncontrolled fluctuations of the gas composition and contaminations by impurities wouldmake the drift velocity unstable and could deteriorate spatial and momentum resolution of candidatesignal tracks.Several studies about the behaviour of drift velocity as a function of the reduced electric field in amixture of He / iC H have been published. Figure 1 . Left, drift velocity as a function of the reduced electric field for different percentages of helium-isobutane mixtures [1] and right, drift velocity as a function of the applied electric field for differentconcentration of water vapours [2]. – 1 –rift velocity is the most sensitive parameter for the operation of a drift chamber with respect totiny variations of the gas mixture and so it is the "target" that we want to use in order to controlthe gas purity. As an example, the left graph in Figure 1 shows that for a mixture of He / iC H at normal pressure, variations of the electric field, around the operating value of 1 V/cm torr − , ofabout 2 V / cm induce drift velocity variations of about 1 × − .Moreover, to mitigate the ageing effect, it is useful to introduce small quantities of water vapors inthe gas mixtures, but it is mandatory to control the consequent variations of drift velocity. As anexample, the right graph of Figure 1 shows that at the operating value of the electric field of about1 V/cm torr − , variations of ≈
150 ppm lead to an increase up to 1 × − in drift velocity. The main goal of the monitoring chamber is to provide a prompt response about drift velocityvariations at 10 − level.This purpose can be obtained with a conceptually very simple structure, illustrated in Figure 2. Figure 2 . Experimental set up of the monitoring drift chamber.
We will use two Sr low-activity calibration radioactive sources placed on top of two thin scintillatortiles telescope. The sources will be collimated to select the tracks crossing the drift cells.The Sr sources produce 2 × electrons per second, with the energy distribution shown in Figure3. Only the electrons with an energy larger than approximately 0.8 MeV, amounting to about 20%of the total [3], will be able to cross the chamber and trigger the scintillator telescope.– 2 – igure 3 . Energy spectrum of electrons emitted by the Sr radioactive source. Moreover, considering the solid angle acceptance, purposedly defined by the source collimators,the total number of triggering decay electrons will be around 4 × . The mechanical structure of the monitoring drift chamber is presented in Figure 4.
Figure 4 . Cross-cut section (left) and exploded view (right) of the mechanical setup.
The lateral cathode walls are made of a thin (190 µ m ) Cu coated PET foil, glued to a rigid frame topreserve planarity. The uniform electric field across the drift cells is obtained with a resistive25 µ m DLC foil, with an electrical resistivity around 100 M Ω m , connected at the edges to the highvoltage lateral cathode walls and with a longitudinal wire in the middle, connected to ground. Alongthe plane separating the two drift cells, four sense wires (20 µ m diameter gold plated tungsten)alternated to five guard wires (80 µ m diameter silver plated aluminum) collect the drift electrons,as Figure 5 shows. – 3 – igure 5 . The figure shows sense wires staggering used to measure drift velocity . Sense wires are staggered in the plane x = ± δ = ± . µ m to allow drift velocity measurements. The drifting electric field and the amplification field around the sense wires have been calculatedwith the Garfield++ program. For a drifting field of 1 kV / cm (corresponding to − V on thecathode planes) and a gas amplification gain of about 5 × on the sense wires, given the describedwire diameters, the voltage on the sense wires must be set at + V , whereas, on the guard wiresit needs to be > = − V in order to keep the value of the electric field on the guard wire surfaceabove − kV / cm , thus avoiding amplification of positive ions.Figure 6 shows the electron drifting lines to the sense wires in the described electrostatic configu-ration. The asymmetry introduced by the sense wire staggering is confined to a very limited regionaround the wires (i.e. for very short drift times) and, as it will become clear in the next paragraph,its effects will be systematically subtracted in the calculation of the drift velocity.– 4 – igure 6 . The figure shows the drift lines converging on the sense wires, marked with the pink color. Figure 7 . The figure shows drift cells structure and two tracks passing inside them. D i indicate the driftdistance from the crossing track to the sense wire i. – 5 –onsidering the two triplets of the wires ( ) and ( ), as shown in Figure 7, indicating with v d the constant drift velocity in the uniform electric field region, simple geometrical considerationslead to the following relations t = t + t ∓ δ v d (4.1) t = t + t ± δ v d , (4.2)where the first (second) choice of the signs refers to a track crossing the right (left) side drift cell.Defining the variable: Θ = ( t + t − t ) − ( t + t − t ) , (4.3)which, according to the track crossing side, assumes one of the two values: (cid:40) Θ + = + δ v d le f t Θ − = − δ v d ri g ht (4.4)one obtains the estimate of v d and of its variance as a function of ∆Θ = | Θ + − Θ − | : v d = δ ∆ Θσ v d = (cid:115)(cid:18) ∆ Θ (cid:19) σ δ + (cid:18) − δ ∆ Θ (cid:19) σ ∆ Θ , (4.5)where σ δ represents the error on the wire positioning and σ ∆ Θ depends statistically on the numberof events collected. Since one is interested only in the variations of the drift velocity, the firstcontribution cancels out and the precision will scale with the collected statistics. For the measurement of the drift velocity, we simulated 9000 tracks, 4500 passing through theleft side and 4500 passing through the right side of the chamber. The gas mixture chosen for thesimulation is of 90% He − iC H at pressure of 760 torr and temperature of 300 K .The tracks are generated with a uniform angular distribution within ± ◦ .For every track, four drift times are collected and the value of the variable Θ is plotted in thehistogram of Figure 8. The two peaks, corresponding to Θ + and Θ − , are highlighted by a fit to thedistribution. – 6 – igure 8 . The double peak distribution for a mixture of 90% He − iC H at pressure of 760 torr andtemperature of 300 K. The drift velocity calculated from this distribution is 2 . ± . cm / µ s , according to the relation(4.5). The number of the events simulated allows for a sensitivity of 1 × − , but an increment ofa factor 100 in the number of tracks will give the expected sensitivity of 1 × − . The monitoring drift chamber, equipped with low-activity radioactive sources and triggered by atelescope of thin scintillator tiles, allows one to monitor the drift velocity of the MEG-II centraltracker, in short time and with a high precision that allows one to evaluate variations of the operatingconditions, which would affect spatial resolution.The continuous monitoring of drift velocity variations at 1 %(cid:24) level is sensitive to variations of:• + .
4% in iC H content (from 10 .
0% to 10 . − .
2% in in iC H content (from 10 .
0% to 9 . .
4% in E/p ( ≈
6% in gas gain) at gain ≈ × • ∓ V at p ≈ bar , T ≈ ◦ C• ∓ mbar at V ≈ V , T ≈ ◦ C• − . ◦ C at p ≈ bar , V ≈ V • ≤
100 ppm variations in water vapor content around 3500 ppm– 7 – eferences [1] P. Bernardini, G. Fiore, R. Gerardi, F. Grancagnolo, U. von Hagel , F. Monittola, V. Nassisi, C. Pinto,L. Pastore, M. Primavera,
Precise measurements of drift velocities in helium gas mixtures , NuclearInstruments and Methods in Physics Research A
355 (1995) 428-433[2] V. Golovatyuk, F. Grancagnolo, R. Perrino,
Influence of oxygen and moisture content on electron lifetime in helium-isobutane gas mixtures , Nuclear Instruments and Methods in Physics Research A Beta spectra of Sr and Y , United States: N. p. (1985).(1985).