- Email: [email protected]
Characterization of Gas Mixtures for Ultra-Light Drift Chambers
Available online at www.sciencedirect.com
Nuclear Physics B (Proc. Suppl.) 248–250 (2014) 131–133 www.elsevier.com/locate/npbps
Characterization of Gas Mixtures for Ultra-Light Drift Chambers M. Cascellaa,b , F. Grancagnolob , P. Mazzottab , A. Miccolib , M. Panareoa,b , M. Spedicatob , G. Tassiellib,c,d a Universit` a
del Salento, Lecce, Italia sezione di Lecce, Italia c Universit` a G. Marconi, Roma, Italia d Fermilab, Batavia, USA b INFN
Abstract Low pressure helium/hydrocarbons mixtures are a key ingredient for next generation ultra-light drift chambers. Besides the obvious advantage of limiting the contribution to the momentum measurement due to multiple scattering, the operation at low pressure allows for a broad range of the drift chamber working parameters like drift velocity, diffusion, specific ionization and gas gain. Low pressure operation is of particular advantage for experiments where the tracking detector operates in vacuum. We present our campaign to characterize electron drift, primary ionization yield, gas gain, stability and the relative spatial resolution in helium based mixtures at absolute pressures down to 100 mbar. Keywords: drift chambers, gain measurement, Diethorn formula
Introduction The choice of gas mixture in a drift chamber is of utmost importance, particularly for experiments where the trajectories of low momentum particles need to be reconstructed with great accuracy. The use of helium based gas mixtures may not be sufficient to overcome the limitations due to the multiple scattering contribution. Moreover, for experiments where the tracking detector is immersed in vacuum, a gas mixture at lower than atmospheric pressure helps reducing the vessel thickness, reducing the amount of material on the path of the conversion electrons, harmful both for the multiple scattering contribution to the momentum measurement and for the fluctuations of the energy loss. To this purpose we have successfully tested some helium based mixtures at pressures below the atmosphere for proportional mode operation, confirming the good stability, in terms of both efficiency and gain, down to values of 0.1 atm. We are also pursuing a campaign of measurements to furter characterize helium based mixtures for drift http://dx.doi.org/10.1016/j.nuclphysbps.2014.02.026 0920-5632/© 2014 Published by Elsevier B.V.
Figure 1: Experimental set-up: the copper tube inside the vacuum box (top left), the sealed box with the radioactive source and the scintillators (bottom left), and the scheme of the apparatus (right).
chambers. 1. Gain Measurements 1.1. Experimental set-up A scheme of the experimental apparatus is shown in figure 1. The copper test tube is hosted in a vacuum proof aluminum box with two thin mylar widows to allow beta electrons from a 90 Sr source and X-rays in. The
M. Cascella et al. / Nuclear Physics B (Proc. Suppl.) 248–250 (2014) 131–133
1.2. Experimental results
0.08 0.07 0.06
0.04 0.03
PIb=0.1 atm
0.02 0.01 0 0
PIb=0.2 atm 0.2
0.4
0.6
0.8
1
1.2 P (atm)
Figure 2: Rate of gain variation with the high voltage as a function of P for different mixtures. 137.9 / 15
23.37 ± 3.26 V 3.365e+04 ± 1.366e+04 K 3.365e+04
80 70
26 24 22 20 18
Baaliouamer et al.
16
60
14 12
50
Cascella et al. 0
0.1
0.2
0.3
0.4
0.5
0.1
0.2
0.3
0.4
0.5
Isobutane fraction
3
40 30
He-Ib 75-25 (0.2 atm)
20
He-Ib 75-25 (0.4 atm)
0
(Volt)
χ2 / ndf
90
10
The ADC signal is fitted with a log-normal distribution that reasonably reproduces the features of the distribution, the Most Probable Value (MPV) of the distribution is used to compute the gain of the mixture. We have studied the gain variation as a function of applied voltage for several gas mixtures and pressures ranging from 0.1 to 1.2 atm. Mixtures at pressures around 1 atm have gains ranging from 105 to 106 and remain in a proportional regime in a range of several hundred Volts, in lower pressure mixtures the gain rises quickly to significantly higher values after which a limited streamer mode sets in. For instance in the case of PIb = 0.05 atm at P = 0.1 atm The range of stability at lower pressures is narrowed to about 100 V. The slope of the exponential fits have been plotted as a function of P the mixture pressure in figure 2 showing how the gain increase with the high voltage is much slower for mixtures with a higher amount of quencher. We have tested our experimental data against Diethorn parametrization[1]. The Diethorn formula can be expressed as a function of ln G/P and of v = V/P as follows ln G v ln 2 v = ln P V0 u K0 au
PIb=0.05 atm
0.05
He-Ib 75-25 (0.8 atm) Diethorn approx. 2000 2500 3000 3500 4000 4500 5000 5500 6000
HV/P (Volt/atm)
(Volt cm-1 atm-1)
Front-end and DAQ. A preamplifier based on the MAR-6SM+ chip2 with a 20 dB gain and a bandwidth of 1.8 GHz is connected to the sense wire and hosted inside the pressure box. An additional Phillips3 775 NIM amplifier (with the same gain and bandwidth as the MAR-6) is used to increase our dynamic range. The acquisition is triggered by the coincidence of the two scintillators. The TDC hits, given by the a 3.4 mV threshold on the tube output, are only used for cross checking purposes in this study.
Rate of Gain variation with V as a function of P
∂G/∂V (Volt-1)
trigger is provided by the coincidence of two scintillators place directly below and above the mylar openings. The mixture composition is set by MKS1 mass flowmeters with an uncertainty of 1% on the iC4 H10 concentration. The Pressure is set to the level of 0.1 Torr using proportional valve controlled by an actuator. The room temperature is kept at 22 ± 1 C.
ln(G)/P (1/atm)
132
×10
100 80 60 40 20 0
0
Isobutane fraction
Figure 3: Results of the fit with the Diethorn parametrization [1] of the data points for a mixture with 25% iC4 H10 (left). Diethorn parameters for several isobutane fraction (filled markers), results by Baaliouamer et al [2] (empty markers) are shown for comparison.
ficients: V0 is the average potential between two consecutive ionization acts and K0 is the threshold value of E/P for avalanche formation. Figure 3 shows Diethorn parameters for several isobutane fraction. We are evaluating the discrepancy with Baaliouamer et al. [2] results and we plan to further extend the range of our measurements.
2. Resolution studies A three tubes system has been used to study the resolution of a cylindric cell with binary mixtures of iC4 H10 and He at P = 1 atm.
where u = ln (b/a), a is the wire radius and b is the tube internal radius. V0 and K0 are the Diethorn coef1 MKS
Instruments Andover, MA, USA. Brooklyn, NY, USA. 3 Phillips Scientific Mahwah, NJ, USA. 2 Mini-Circuits
Figure 4: Schematic view of the three tubes system.
M. Cascella et al. / Nuclear Physics B (Proc. Suppl.) 248–250 (2014) 131–133
Figure 5: the resolution σi as a function of the He content of the mixture, NHe = 7.4, NIb = 54.
133
Figure 6: An exploded view of the GasParE time expansion chamber.
3. Primary drift and diffusion measurements 2.1. Experimental apparatus The system is composed of three copper drift tubes of internal radius 8 mm in a row and a r = 10 μm gold coated tungsten wire in them. The middle tube is displaced horizontally by Δ = 500 μm with respect to the others so that for almost vertical tracks d1 + d3 − d2 ±Δ 2 the three wires are amplified with two Phillips 775 amplifier (40 dB gain) and sampled with a Tektronix oscilloscope. The trigger is given by the coincidence of three scintillators placed above and below the three tubes (see Figure 4). 2.2. Results and analysis A two thresholds algorithm4 is used to recognize the first ionization cluster, then the space-time relation is used to compute the impact parameter di of the muon. The relation is initially determined using Garfield [3] and then recursively computed from experimental data. By plotting the quantity (d1 + d3 )/2 − d2 we get a two peaks,√centered approximately in ±Δ and with a width σΔ = 2/3σd (see Figure 5 left). The resolution of the three tubes system with several mixtures of iC4 H10 and He is plotted in Figure 5 right, σd has two contributes, the first is the multiple scattering off the copper tubes σ MS 5 and the second is proportional to the number of ionization acts N = fHe NHe + NIb (1 − fHe ) parametrized as a function of the He fraction fHe . 4 For a peak expected to be n samples wide the algorithm requires at least one sample above 3σnoise and at least n/2 samples above σnoise . 5 For muons with p = 1 GeV transversing x = 700 μm of copper √ θ MS = (13.6 MeV)/(βcp) x/X0 (1 + 0.038 ln(x/X0 )) = 2.67 mrad; thus σ MS = θ MS · 2.4 cm = 64 μm .
A time expansion chamber: GasParE (Gas Parameter Experiment) is being developed to characterize the transport parameters of electrons in helium based gas mixtures. A uniform electric field of up to 5 kV/cm is established in a stack of 12 copper plates placed 1 cm apart. A N2 UV laser is used to induce ionization in the gas volume trough multi-photon absorption. The transverse beam size at the focus is of the order of 10um. The longitudinal position of the beam spot can be varied with continuity with a 3 μm resolution. Drifting electrons are detected using a drift tube placed after the last plate. The setup is shown in figure 6 and is described in detail in [4, 5, 6]. We plan to measure the drift velocity vdri f t and the longitudinal diffusion σdi f f,l from the distribution arrival time of the electrons produced by the UV laser. The transverse diffusion σdi f f ⊥ can be estimated from the relative collection efficiency while varying the transverse position of the beam spot. We also intend to measure the primary ionization yield by using a low electric field (slow drift), placing the Sr90 source in front of the copper plates and counting the ionization clusters along the β path. References [1] [2] [3] [4] [5] [6]
Diethorne, W, et al. In: US. AEC Report NYO 6628 (1956) Baaliouamer, M, et al. In: NIM A A 328 (1996), 490–494. Garfield. http://garfield.web.cern.ch/garfield/ Bernardini, P, et al. In: NIM A 355 (1995), 428–433S. Golovatyuk, V, et al. In: NIM A 394 (1997), 97–102 S. Golovatyuk, V, et al. In: NIM A 461 (2001), 96–97 S.
Nuclear Physics B (Proc. Suppl.) 248–250 (2014) 131–133 www.elsevier.com/locate/npbps
Characterization of Gas Mixtures for Ultra-Light Drift Chambers M. Cascellaa,b , F. Grancagnolob , P. Mazzottab , A. Miccolib , M. Panareoa,b , M. Spedicatob , G. Tassiellib,c,d a Universit` a
del Salento, Lecce, Italia sezione di Lecce, Italia c Universit` a G. Marconi, Roma, Italia d Fermilab, Batavia, USA b INFN
Abstract Low pressure helium/hydrocarbons mixtures are a key ingredient for next generation ultra-light drift chambers. Besides the obvious advantage of limiting the contribution to the momentum measurement due to multiple scattering, the operation at low pressure allows for a broad range of the drift chamber working parameters like drift velocity, diffusion, specific ionization and gas gain. Low pressure operation is of particular advantage for experiments where the tracking detector operates in vacuum. We present our campaign to characterize electron drift, primary ionization yield, gas gain, stability and the relative spatial resolution in helium based mixtures at absolute pressures down to 100 mbar. Keywords: drift chambers, gain measurement, Diethorn formula
Introduction The choice of gas mixture in a drift chamber is of utmost importance, particularly for experiments where the trajectories of low momentum particles need to be reconstructed with great accuracy. The use of helium based gas mixtures may not be sufficient to overcome the limitations due to the multiple scattering contribution. Moreover, for experiments where the tracking detector is immersed in vacuum, a gas mixture at lower than atmospheric pressure helps reducing the vessel thickness, reducing the amount of material on the path of the conversion electrons, harmful both for the multiple scattering contribution to the momentum measurement and for the fluctuations of the energy loss. To this purpose we have successfully tested some helium based mixtures at pressures below the atmosphere for proportional mode operation, confirming the good stability, in terms of both efficiency and gain, down to values of 0.1 atm. We are also pursuing a campaign of measurements to furter characterize helium based mixtures for drift http://dx.doi.org/10.1016/j.nuclphysbps.2014.02.026 0920-5632/© 2014 Published by Elsevier B.V.
Figure 1: Experimental set-up: the copper tube inside the vacuum box (top left), the sealed box with the radioactive source and the scintillators (bottom left), and the scheme of the apparatus (right).
chambers. 1. Gain Measurements 1.1. Experimental set-up A scheme of the experimental apparatus is shown in figure 1. The copper test tube is hosted in a vacuum proof aluminum box with two thin mylar widows to allow beta electrons from a 90 Sr source and X-rays in. The
M. Cascella et al. / Nuclear Physics B (Proc. Suppl.) 248–250 (2014) 131–133
1.2. Experimental results
0.08 0.07 0.06
0.04 0.03
PIb=0.1 atm
0.02 0.01 0 0
PIb=0.2 atm 0.2
0.4
0.6
0.8
1
1.2 P (atm)
Figure 2: Rate of gain variation with the high voltage as a function of P for different mixtures. 137.9 / 15
23.37 ± 3.26 V 3.365e+04 ± 1.366e+04 K 3.365e+04
80 70
26 24 22 20 18
Baaliouamer et al.
16
60
14 12
50
Cascella et al. 0
0.1
0.2
0.3
0.4
0.5
0.1
0.2
0.3
0.4
0.5
Isobutane fraction
3
40 30
He-Ib 75-25 (0.2 atm)
20
He-Ib 75-25 (0.4 atm)
0
χ2 / ndf
90
10
The ADC signal is fitted with a log-normal distribution that reasonably reproduces the features of the distribution, the Most Probable Value (MPV) of the distribution is used to compute the gain of the mixture. We have studied the gain variation as a function of applied voltage for several gas mixtures and pressures ranging from 0.1 to 1.2 atm. Mixtures at pressures around 1 atm have gains ranging from 105 to 106 and remain in a proportional regime in a range of several hundred Volts, in lower pressure mixtures the gain rises quickly to significantly higher values after which a limited streamer mode sets in. For instance in the case of PIb = 0.05 atm at P = 0.1 atm The range of stability at lower pressures is narrowed to about 100 V. The slope of the exponential fits have been plotted as a function of P the mixture pressure in figure 2 showing how the gain increase with the high voltage is much slower for mixtures with a higher amount of quencher. We have tested our experimental data against Diethorn parametrization[1]. The Diethorn formula can be expressed as a function of ln G/P and of v = V/P as follows ln G v ln 2 v = ln P V0 u K0 au
PIb=0.05 atm
0.05
He-Ib 75-25 (0.8 atm) Diethorn approx. 2000 2500 3000 3500 4000 4500 5000 5500 6000
HV/P (Volt/atm)
Front-end and DAQ. A preamplifier based on the MAR-6SM+ chip2 with a 20 dB gain and a bandwidth of 1.8 GHz is connected to the sense wire and hosted inside the pressure box. An additional Phillips3 775 NIM amplifier (with the same gain and bandwidth as the MAR-6) is used to increase our dynamic range. The acquisition is triggered by the coincidence of the two scintillators. The TDC hits, given by the a 3.4 mV threshold on the tube output, are only used for cross checking purposes in this study.
Rate of Gain variation with V as a function of P
∂G/∂V (Volt-1)
trigger is provided by the coincidence of two scintillators place directly below and above the mylar openings. The mixture composition is set by MKS1 mass flowmeters with an uncertainty of 1% on the iC4 H10 concentration. The Pressure is set to the level of 0.1 Torr using proportional valve controlled by an actuator. The room temperature is kept at 22 ± 1 C.
ln(G)/P (1/atm)
132
×10
100 80 60 40 20 0
0
Isobutane fraction
Figure 3: Results of the fit with the Diethorn parametrization [1] of the data points for a mixture with 25% iC4 H10 (left). Diethorn parameters for several isobutane fraction (filled markers), results by Baaliouamer et al [2] (empty markers) are shown for comparison.
ficients: V0 is the average potential between two consecutive ionization acts and K0 is the threshold value of E/P for avalanche formation. Figure 3 shows Diethorn parameters for several isobutane fraction. We are evaluating the discrepancy with Baaliouamer et al. [2] results and we plan to further extend the range of our measurements.
2. Resolution studies A three tubes system has been used to study the resolution of a cylindric cell with binary mixtures of iC4 H10 and He at P = 1 atm.
where u = ln (b/a), a is the wire radius and b is the tube internal radius. V0 and K0 are the Diethorn coef1 MKS
Instruments Andover, MA, USA. Brooklyn, NY, USA. 3 Phillips Scientific Mahwah, NJ, USA. 2 Mini-Circuits
Figure 4: Schematic view of the three tubes system.
M. Cascella et al. / Nuclear Physics B (Proc. Suppl.) 248–250 (2014) 131–133
Figure 5: the resolution σi as a function of the He content of the mixture, NHe = 7.4, NIb = 54.
133
Figure 6: An exploded view of the GasParE time expansion chamber.
3. Primary drift and diffusion measurements 2.1. Experimental apparatus The system is composed of three copper drift tubes of internal radius 8 mm in a row and a r = 10 μm gold coated tungsten wire in them. The middle tube is displaced horizontally by Δ = 500 μm with respect to the others so that for almost vertical tracks d1 + d3 − d2 ±Δ 2 the three wires are amplified with two Phillips 775 amplifier (40 dB gain) and sampled with a Tektronix oscilloscope. The trigger is given by the coincidence of three scintillators placed above and below the three tubes (see Figure 4). 2.2. Results and analysis A two thresholds algorithm4 is used to recognize the first ionization cluster, then the space-time relation is used to compute the impact parameter di of the muon. The relation is initially determined using Garfield [3] and then recursively computed from experimental data. By plotting the quantity (d1 + d3 )/2 − d2 we get a two peaks,√centered approximately in ±Δ and with a width σΔ = 2/3σd (see Figure 5 left). The resolution of the three tubes system with several mixtures of iC4 H10 and He is plotted in Figure 5 right, σd has two contributes, the first is the multiple scattering off the copper tubes σ MS 5 and the second is proportional to the number of ionization acts N = fHe NHe + NIb (1 − fHe ) parametrized as a function of the He fraction fHe . 4 For a peak expected to be n samples wide the algorithm requires at least one sample above 3σnoise and at least n/2 samples above σnoise . 5 For muons with p = 1 GeV transversing x = 700 μm of copper √ θ MS = (13.6 MeV)/(βcp) x/X0 (1 + 0.038 ln(x/X0 )) = 2.67 mrad; thus σ MS = θ MS · 2.4 cm = 64 μm .
A time expansion chamber: GasParE (Gas Parameter Experiment) is being developed to characterize the transport parameters of electrons in helium based gas mixtures. A uniform electric field of up to 5 kV/cm is established in a stack of 12 copper plates placed 1 cm apart. A N2 UV laser is used to induce ionization in the gas volume trough multi-photon absorption. The transverse beam size at the focus is of the order of 10um. The longitudinal position of the beam spot can be varied with continuity with a 3 μm resolution. Drifting electrons are detected using a drift tube placed after the last plate. The setup is shown in figure 6 and is described in detail in [4, 5, 6]. We plan to measure the drift velocity vdri f t and the longitudinal diffusion σdi f f,l from the distribution arrival time of the electrons produced by the UV laser. The transverse diffusion σdi f f ⊥ can be estimated from the relative collection efficiency while varying the transverse position of the beam spot. We also intend to measure the primary ionization yield by using a low electric field (slow drift), placing the Sr90 source in front of the copper plates and counting the ionization clusters along the β path. References [1] [2] [3] [4] [5] [6]
Diethorne, W, et al. In: US. AEC Report NYO 6628 (1956) Baaliouamer, M, et al. In: NIM A A 328 (1996), 490–494. Garfield. http://garfield.web.cern.ch/garfield/ Bernardini, P, et al. In: NIM A 355 (1995), 428–433S. Golovatyuk, V, et al. In: NIM A 394 (1997), 97–102 S. Golovatyuk, V, et al. In: NIM A 461 (2001), 96–97 S.