A Gas Monitoring Chamber for High Pressure Applications
Philip Hamacher-Baumann, Stefan Roth, Thomas Radermacher, Nick Thamm
AA Gas Monitoring Chamber for High Pressure Applications
Philip Hamacher-Baumann ∗ III. Physikalisches Institut, RWTH Aachen University, 52056 Aachen, Germany (Dated: May 8, 2020)Recently, Time Projection Chambers (TPCs) operated at high pressure have become a topic ofinterest for future long baseline neutrino experiments [J. Mart´ın-Albo, J. Phys. Conf. Ser. ,012154 (2017)]. Pressurised gas retains the low momentum threshold of atmospheric TPCs, butoffers a larger target mass for neutrino interactions. Operation at high pressure poses several newchallenges in safety aspects regarding overpressure and high voltage safety. The presented HighPressure Gas Monitoring Chamber (HPGMC) can be used to study the suitability of various driftgas mixtures up to 10 bar and a maximum field of ∼ V / cm . The flexible construction makesit possible to exchange parts of the inner detector and test new technologies. In this work, theconstruction of a HPGMC and its commissioning using a P10 (90 % Ar + 10 % CH ) gas mixtureare presented. I. INTRODUCTION
Time Projection Chambers (TPCs) are gaseous de-tectors, that have originally been conceived for track-ing moderate numbers of particles in clean lepton inter-actions at colliders [1]. Since then, the TPC technol-ogy has evolved and is presently used from rare eventsearches [2] to ultra-high multiplicities nuclear collisionexperiments [3]. The inherent long radiation-length na-ture of gaseous detectors makes them ideal for trackingtasks, where particle shower generation or multiple scat-tering are a concern. The gas is interchangeable and canbe tailored to the given task without modification of themain detector parts or even disassembly.Since the active medium is of very low density, a lowmomentum threshold for track reconstruction can beachieved. Hence, TPCs have become a topic of inter-est for experiments at the few GeV energy scale, wherethe range of particles limits resolution, e.g. separationof tracks at an interaction vertex. Currently, the lim-iting systematic uncertainties of long baseline neutrinoexperiments are in the modelling of nuclear effects infinal state interactions [4], that can be improved bynew data in the low momentum region [5]. To achievea higher target mass, while retaining the low momen-tum threshold, High Pressure Time Projections Cham-bers (HPTPC) have been proposed for future experi-ments. The Deep Underground Neutrino Experiment’s(DUNE) near detector complex foresees to build a 10 barHPTPC to constrain interaction uncertainties in combi-nation with the liquid argon filled far detectors [6, 7].This work presents the construction and commissioningof a small-scale gaseous detector aimed at studies of driftparameters at high pressures. ∗ [email protected] II. GAS MONITORING CHAMBERS
The characteristics of the TPC’s drift gas are a neces-sary input for track reconstruction and when designingthe readout electronics. Any change in these propertiesdirectly affects performance of the detector. Since exper-iments can have run times of many years under changingenvironmental conditions and gas quality, there is a needfor continuous calibration of drift properties [8].
A. Working Principle
A Gas Monitoring Chamber (GMC) is a miniatureTPC in which ionisation tracks are created at knownpositions by radioactive sources (here Sr). Emitted β electrons can traverse the gas-filled drift volume, leavingan track of free electrons and ions, and exit into a scin-tillating fibre (Fig. 1a). The light signal from the scintil-lating fibre is used to start a fixed window measurementof signals from drifting electrons recorded at the anode.Figure 1b shows the average of 1500 such recorded wave-forms at the anode without distinguishing between thetwo (near and far) source positions. The drift velocity isthen reconstructed by measuring the time difference ∆ t between the two signal peaks and the known distance ∆ z between near and far source positions: v d = ∆ z ∆ t (1)In this scheme, the drift velocity is effectively mea-sured between the near and far position in a central driftregion of the GMC, which reduces or removes the two ma-jor systematic effects from the measurement. Firstly, thecentral field cage region has a more homogeneous driftfield than at the edges. The influence of traversing thestrong (fringe) fields close the to amplification gap alsocancel out. Secondly, constant time response character-istics (e.g. delays) of the readout and trigger electronicscancel out.GMCs have also been equipped with mono-energetic a r X i v : . [ phy s i c s . i n s - d e t ] M a y X-ray sources to monitor gain: By continuously measur-ing the spectrum of Fe, fluctuations in the gas gain canbe monitored and calibrated for [8].GMCs can be build with small inner volumes of ∼ (cid:96) ,and even be used to monitor multiple subdetectors insequence [9]. (a)(b) Figure 1: (a) Sketch of a GMC with radioactive sourcesfor drift velocity and gas gain measurement (adaptedfrom [10]). The drift fields can mirror that of a TPC orsystematically scan through different fields to create adrift velocity curve. (b) Average over 1500 wave-formsrecorded at the anode. The time difference of the twopeaks is the drift time between the near and far source’sposition.
B. Continuous Calibrations
Drifting electrons are accelerated and deflected by elec-tric and magnetic fields between collisions with the gasatoms or molecules. The rate at which these collisionshappen depends on the gas density and follows pressureand temperature fluctuations. Under the assumption,that the drift velocity directly depends on the field alongthe electron’s path between two subsequent collisions,a density correction can be obtained by evaluating thedensity-corrected drift velocity [2]: v d ( E ∗ ) = v d (cid:18) ETp (cid:19) . (2)The quantity E ∗ is often referred to as the reduced elec-tric field.TPCs are generally not well controlled in pressure ortemperature, but rather follow environmental fluctua-tions [8]. As a result, the drift properties of the usedgas mixture vary with them, which has to be accountedfor when reconstructing particle tracks or combining datasets over long running periods.One strategy used is to operate a GMC in parallel toa main detector TPC with gas supplied by the detec-tor’s return line [8]. The drift velocity is continuouslymeasured at the TPC’s drift field. In addition, temper-ature and pressure of the gas in the TPC and GMC arerecorded. Both data sets can be calibrated towards acommon T and p to make the data comparable over theruntime by correcting changes in gas density. Changesthrough varying impurity concentrations or supply gasquality are picked up and calibrated by this strategy. ForND280, the T2K near detector, the gas gain fluctuationsare reduced to 1 % compared to ∼
10 % without calibra-tion [8].
III. HIGH PRESSURE OPERATION
The operation principle of an atmospheric GMC canbe transferred almost identically to a High Pressure GasMonitoring Chamber (HPGMC). The highest pressureconsidered in this work is 10 bar, following the design ofthe DUNE near detector [6].Many commonly used quenching gases are flammable,adding another source of danger to the high overpres-sure. To ensure safe operation, the critical parts aresourced from industry with a certified and verified qual-ity. Pressure-holding parts are manufactured accordingto DIN standard, which restricts the vessel’s capacity toa maximum of 50 bar (cid:96) , or 5 (cid:96) at 10 bar [11]. A small innervolume is also desirable to provide a frequent exchange ofthe gas inside and thus low latency detection of changesin the gas mixture or quality.
A. Range of Electrons in High Pressure Gas
The electrons from the Sr sources are strongly atten-uated and scattered due to the high pressure and hencemore often prevented from reaching the start-trigger fi-bres than at atmospheric pressure. A considerable frac-tion of the β decay electrons are produced with an energyaround 500 keV (Fig. 2) [12], that have an approximaterange of 10 cm in 10 bar argon [13]. It is necessary, thatcrossing electrons have enough energy to reach and thenproduce scintillation in the start-trigger fibres after cross-ing the pressurised gas. Therefore, the diameter of thefield cage, across which the β electrons are to traverse,was chosen to be smaller than 10 cm. The electrons enterthe fieldcage in a collimated cone that is further widenedby multiple scattering. The opposing 8 × start-trigger fibres are aligned so, that only tracks are selected,that have not been deflected more than the fibre’s heightalong the drift direction.A more fundamental difference to an atmosphericGMC comes from the fact that the sources have to beplaced inside the pressurised volume, as the beta elec-trons can not penetrate the pressure vessel; a 2.5 MeVelectron has a range of only ∼ I n t e n s i t y / a . u . r a n g e / c m Figure 2: Range (blue) and spectrum (red) of electronsemitted by the Sr decay chain. The two peakscorrespond to the Sr (left) and Y (right) decays inthe source. To retain reasonable rates, the traverseddistance in the gas should not be larger than 10 cm.Values computed using CSDA range listed in NIST’sESTAR database [13].
B. High Voltage Safety
The drift velocity is a function of the reduced field E ∗ [Eq. 2]. An increase in pressure will expand driftvelocity curves along that axis. The effect can be can-celled by a likewise increase of the drift field E to keep E ∗ constant. Typical drift fields are of the order of afew 100 V / cm / bar ; to reach corresponding reduced fieldsat 10 bar, the HPGMC is designed for a maximum cath-ode voltage of 30 kV.One concern with high voltages are electric break-downs, that could damage parts of the detector or read-out electronics. The breakdown in a gas can be calculatedby Paschen’s Law V b ( pd ) = B pd ln (A pd ) − ln (cid:2) ln (cid:0) γ − (cid:1)(cid:3) , (3)that predicts the breakdown voltage V b between two ide-alised coplanar surfaces separated by a gas at pressure p ,the separation distance d , gas-specific parameters A , Band a material constant γ se [15, 16]. The parametersA , B = A , B( V applied ) are experimentally found to be con-stant over a limited range of E / p [16]. For scans overlarger ranges, A , B can not be assumed to be constant,but can be calculated from Townsend simulations of puregasses with
MagBoltz [17].The secondary electron emission coefficient γ se de-scribes how many electrons are released from ions im-pacting on the cathode and later reach the anode. Itdepends on the involved electrode materials and is high-est for alkali metals [16]. Its influence on the resulting V b is negligible for high voltages and pressures, where V b ∝ pd .The fraction V b /V applied is used to determine a safeoperation limit by choosing insulation distances such that V b ( pd ) ! > · V applied , with p = 1 bar . (4)Figure 3 presents contours of constant fractions betweenapplied and breakdown voltage by evaluating [Eq. 4]. Atthe maximum design voltage of 30 kV for the HPGMC,the clearance between the cathode on high voltage andthe pressure vessel on ground potential would have tobe larger than 12 cm. This would introduce too muchdead space to comply with a limit of 5 (cid:96) for the inner vol-ume. Therefore, the vessel’s wall of the HPGMC is linedwith 4 mm PFA, a PTFE derivative with high dielectricstrength of 80 kV/mm [18] and low outgassing [19]. Ar-gon at high pressure requires significantly less insulationdistance.Special care has to be taken for the location of the ra-dioactive sources. To retain a reasonable measurementrate, the sources have to be inside the pressure vesseland as close as possible to the drift volume. This ex-poses them to voltages close to 30 kV for the far source,which sits close to the cathode plane. Electrostatic sim-ulations with Agros2D [21] have been performed for the V a pp l . / ( k V ) . . . . . . Figure 3: Contours of constant ratio of breakdownvoltage to voltage applied over a gap d in 1 bar argon(Fig. from [20]). Distances are chosen such that V b /V applied stays above 2 (green region). The orangeregion (1 ≤ V b /V applied ≤
2) is avoided as a safetybuffer. Corresponding distances for high pressure argonare generally smaller.far source position to determine the minimum requireddistance from [Eq. 4]. The 2 mm ×
10 mm sources areinstalled in brass capsules, which then are placed closeto the field cage inside a solid holder made from POM(Fig. 4 and Section IV). The near position is generallyexposed to less strong fields, but is placed at the samedistance from the field cage as the far source.Figure 4: Electrostatic simulation for the far sourceposition, done with
Agros2D (Fig. from [20]). Thecathode has been fixed at the maximum voltage andfield rings at their corresponding degraded voltage.Source and capsule are made from brass metal, but leftfloating. The calculated breakdown field inside the boreis always more then twice of the actual field. The insetshows the actual sources holders with a 1 Yen coin forscale.
IV. THE HIGH PRESSURE GAS MONITORINGCHAMBER: HPGMC
The HPGMC uses a stainless steel industry standardflanged cross pipe with modified blind flanges for elec-trical and gas connections to the inside (Fig. 6) with aninner volume of ∼ . (cid:96) and a wall thickness of ∼ Sr source holders on opposite sides,which are connected by the field cage in the centre. Thisyields a good precision of the alignment, especially in∆ z , while still avoiding permanent installation. Replace-ment of most inner parts can be accomplished in shorttime and different technologies can be tested without theneed for construction of a new field cage or pressure ves-sel. The maximum operation range of the HPGMC is ∼ kV / cm / bar at 1 bar, or ∼ V / cm / bar at the maximumpressure 10 bar at room temperature. Gas is suppliedfrom premixed bottles and regulated with a bottle pres-sure regulator in high pressure operation. A pressure re-lieve valve limits the over pressure of the vessel to 11 bar.The crate is equipped with pressure and temperaturesensors to monitor the operation and environmental con-ditions. It has been designed to be mobile for use with ahost detector, e.g. a TPC at a test beam facility. V. FIRST MEASUREMENTS
The commissioning of the HPGMC was carried outusing premixed standard P10 (Ar:CH – 90:10) with amaximum pressure of ∼ T = 296 . / p of the measurements.It is found, that [Eq. 2] holds for the full range of morethan 5 bar, from almost atmospheric to above 6 bar pres-sure. VI. OUTLOOK AND CONCLUSION
The HPGMC is a miniature TPC, capable of operationwith drift gases up to 10 bar pressure, given sufficient gasamplification. It can be used for continuous calibrationof a host detector or as a stand-alone experiment to studygas properties. The anode plane can be exchanged to in-vestigate the response and stability of different amplifica- m / n s ) m / ( d v Figure 7: Drift velocity measurements performed byHPGMC at three different pressure levels. The widerange in pressure of the three measurements is causedby a drift of the gas bottle pressure reducer used to setthe chambers pressure. Scaling against ET / p holds overthe full tested span of more than 5 bar.tion geometries in novel gas mixtures at atmospheric orhigh pressure. The very high maximal field of ∼ kV / cm can also be reached at 1 bar. The small inner volume of ∼ . (cid:96) enables a quick change of the gas mixture underinvestigation. Scaling of drift velocity curves at chang-ing pressures was found to be consistent with E ∗ = ET / p over a wide span of 5 bar. The system is mobile and hasalready accompanied a test beam experiment at CERN.Attempts are underway to test the operation with argonbased drift gases at both very low and very high quencherfractions. ACKNOWLEDGMENTS
I thank Stefan Roth for many discussions on con-struction and operation of drift chambers. I thank theT2K/TPC working group at the 3rd Physics Instituteof RWTH Aachen University for fruitful discussions dur-ing all phases of the project. Special thanks go to theteams in the workshops and Jochen Steinmann who madeeven the most complicated parts of the detector a reality.PH-B is supported by DFG (Germany) Grant No. RO3625/2-1. [1] H. Hilke, Reports on Progress in Physics , 116201(2010).[2] D. Gonz´alez-D´ıaz, F. Monrabal, and S. Murphy, NuclearInstruments and Methods in Physics Research Section A:Accelerators, Spectrometers, Detectors and AssociatedEquipment , 200 (2018).[3] Dellacasa et al. (ALICE Collaboration), ALICE time pro-jection chamber: Technical Design Report , Technical De-sign Report ALICE (CERN, Geneva, 2000).[4] L. Alvarez-Ruso et al. , Progress in Particle and NuclearPhysics , 1 (2018), arXiv:1706.03621v2.[5] K. Abe et al. (T2K), (2019), arXiv:1901.03750[physics.ins-det].[6] J. Martin-Albo, Journal of Physics: Conference Series , 012154 (2017), arXiv:1610.07803v1.[7] B. Abi et al. , (2020), arXiv:2002.03010.[8] N. Abgrall et al. (T2K ND280 TPC), Nuclear In-struments and Methods in Physics Research SectionA: Accelerators, Spectrometers, Detectors and Asso-ciated Equipment
A637 , 25 (2011), arXiv:1012.0865[physics.ins-det].[9] G. Altenh¨ofer, T. Hebbeker, C. Heidemann, H. Reithler,L. Sonnenschein, and D. Teyssier, Nuclear Instrumentsand Methods in Physics Research Section A: Acceler-ators, Spectrometers, Detectors and Associated Equip-ment , 1 (2017). [10] D. Terhorst, “Entwicklung einer Monitorkammer zur¨Uberwachung des Driftkammergases der T2K-TPC,”(2008).[11] Deutsches Institut f¨ur Normung e. V., “Unfired pressurevessels - Part 1: General; German version EN 13445-1:2014,” (2011).[12] H. Hansen, The International Journal of Applied Radia-tion and Isotopes , 1241 (1983).[13] S. Seltzer, “Stopping-Powers and Range Tables for Elec-trons, Protons, and Helium Ions, NIST Standard Refer-ence Database 124,” (1993).[14] Deutsches Institut f¨ur Normung e. V., “Radiological pro-tection - Sealed radioactive sources - General require-ments and classification (ISO 2919:2012),” (2012).[15] F. Paschen, Annalen der Physik , 69 (1889).[16] M. A. Lieberman and A. J. Lichtenberg, Principles ofPlasma Discharges and Materials Processing , second edi-tion ed. (John Wiley & Sons, Inc., 2005).[17] S. Biagi, Nuclear Instruments and Methods in PhysicsResearch Section A: Accelerators, Spectrometers, Detec-tors and Associated Equipment
A421 , 234 (1999).[18] DuPont, “DuPontTM Teflon R (cid:13) PFA 345 Molding and Ex-trusion Resin,” (2017).[19] NASA, “Outgassing Materials Database,” (2017).[20] P. Hamacher-Baumann, (2019), arXiv:1911.04846v1.[21] P. Karban, F. Mach, P. K˚us, D. P´anek, and I. Doleˇzel,Computing95