Very high voltage (VHV)

Generation and transport of very high voltage (O(100) kV) in pure cryogenic liquid and gases has always been a rich source of R&D. Indeed cryogenic detectors have the huge advantage of generating a complete 3-D image of all particles from an interaction and reconstruct the particles’ energies well while sorting event from background. However, the collection of the charges generated by the ionization along the tracks, because of the dense medium (compared to gas) requires high drift fields of the order of 500 V/cm. This kind of experiments aiming at a maximum drift of several meters require bias voltages at their cathode in the hundreds of kilovolt range. The problem is to be able to generate this extremely high voltages and to get the voltage to the cathode over an unusually large distance without compromising the purity of the argon inside the detector by letting heat or air inside. Nonetheless there is a huge worldwide effort to tackle those issues.

A good overview of these R&D activities, before the start of the AIDA2020 project, was presented at the Workshop High Voltage in Noble Liquids which was held at Fermilab in 2013. This workshop made the point on several technical aspects related to VHV in neutrino and dark matter experiments such as MicroBooNE, ARGONTUBE, LBNE, CAPTAIN, DarkSide, ArgoNeut, LZ/Zeplin III, EXO/NEXO and NEXT. A summary of the workshop is presented in this paper.

R&D activities in the framework of AIDA2020 WP8

WP8 fosters knowledge sharing and common tools in the neutrino community as regards state-of-the-art in very large cryogenic liquid detectors. The construction of liquid argon detectors at the 10 kton scale is an essential ingredient of the future international long-baseline neutrino program unifying the European and USA efforts. WP8 activities focus on the most challenging aspects related to this detector development. One of these aspects, studied in the networking activity of Task 8.5, concerns the safe generation and transport of high voltages equal or superior to 200 kV (DC) which are necessary in order to provide a field of 500 V/cm for the drift of the ionising charge over many meters. One delicate component is the feedthrough whose role is to safely deliver the very high voltage to the cathode through the thick insulating walls of the cryostat without compromising the purity of the argon inside. This requires a feedthrough that is typically meters long and carefully designed to be vacuum tight and have small heat input. Furthermore, all materials should be carefully chosen to allow operation in cryogenic conditions. Detailed FEA simulations are required in order to optimize the design of the feedthrough but also of any other component which is biased at such large voltages especially the field cage, cathodes. In this context, a study of innovative solutions for the field cage based on light aluminum extruded profiles has been performed and a field cage based on this design is being deployed in the 6x6x6 m3 liquid argon TPC at CERN. Another aspect of this task relates to the design and test of transparent photocathodes made from an array of PMMA plates with a resistive ITO (Indium Thin Oxide) coating. This technique has advantages related to the light readout but also directly related to the photocathode design itself by limiting the propagation of bubbles which can affect the dielectric rigidity of LAr in the drift volume and in making a segmented resistive cathode with a more regular surface and lower electric fields which limit the breakdown hazards.

The main aspects of the WP8 networking activity have been then including:

• Reviewing: very high voltage methods for large noble liquids TPCs
• Development of VHV simulations, identification of critical aspects in detector design
• Development of VHV feedthroughs and generators, tests
• Construction techniques for large field cages/cathodes

WP8 used as test-bench the 3x1x1 dual-phase liquid argon TPC prototype and the related test infrastructure at CERN from which specific R&D results have been obtained.

VHV main components

The main aspects of a Very High Voltage system are:

• Generation: The power supply should be able to provide the maximal operation high voltage (HV) and the power cable should be rated for this voltage.
• Transmission: The HV feedthrough (HVFT) should be designed to sustain this maximal operation HV.
• Design: The shape of all the elements of the TPC should be carefully designed to avoid critical field regions where sparking could occur (maximal electric field above 40 kV/cm).
• Electrostatic simulations: The whole detector including the cathode, the field cage and the feedthrough has been simulated using the COMSOL multiphysics software. This work was done inside WP8 and the results have been summarised in ProtoDUNE-DP PROTOtype for the Deep Underground Neutrino Experiment - Dual Phase detector. We optimised the design of the feedthrough through simulations and we studied the influence of the different parameters of the HVFT design in the computed electric field along the whole TPC. The results are important also for future R&D on HVFT, cathode and field cage design.

Transmission: The high voltage feedthrough

Giant time projection chambers filled with liquid argon is the technology of choice for future neutrino experiments like DUNE. One of the biggest challenges in this kind of detectors is to be able to bring the voltage from warm temperature (outside the detector) to the LAr which is at cryogenic temperature. To deliver the voltage inside the cryostat is used a vacuum tight feedthrough. A first prototype of such a future feedthrough – a giant insulating plug – was successfully tested at CERN under AIDA-2020 WP8 and was in operation during more than four months inside a 3m3 Dual phase TPC prototype at CERN.

The feedthrough was designed by Franco Sergiampietri at CERN and its general design comes from the experience acquired during the ICARUS experiment (see Design, construction and tests of the ICARUS T600 detector ) where lower high voltage feedthroughs were successfully tested up to -150 kV and operated for several years at - 75 kV. The 300 kV feedthrough is a scaled up and improved design. The scale up was essentially based on the following considerations:

• Increase the length of the feedthrough to penetrate the larger thicknesses of the passively insulated membrane cryostats.
• Enlarge the diameter of the inside hole to accept cable plugs transporting higher voltages.
• Keep the heat input to a minimal value to avoid formation of gas argon bubbles in the vicinity of the high voltage.
• Preserve the vacuum tightness of the cryostat.
• Reduce as much as possible the electric field at the ground termination.

The design follows a coaxial configuration with inner and outer conductors. The outer conductor consists of a 2 mm thick tubular shield with a CF250 Ultra High Vacuum flange welded on one end and an elliptical edge ring welded on the other. The inner conductor consists of a 1 mm thick stainless-steel tube filled with FR4 to provide the vacuum tightness. It has a female receptacle to accept the feedthrough cable plug on one end and a male threaded contact on the other for electrical connection to the detector. Both inner and outer conductor tube thicknesses are chosen to minimise the heat input. The insulation between them is provided by one continuous 2 meter rod of High Molecular Density Polyethylene (HMDPE). We use HMDPE RCH-1000, which is rated for operations at −269◦ C, it has a relative dielectric permittivity εr of 2.3 and a dielectric rigidity of 900 kV/cm. Detailed electrostatic simulations were performed inside WP8 to study and understand the field configuration along the feedthrough inside the LAr. The area where the outer conductor is terminated though inevitably produces high electric fields. We showed that the highest electric field in liquid argon is reached precisely in this particular area. As such great care should be taken to optimise the shape of the ground termination. For the ICARUS feedthrough the outer conductor ended with a ring of circular cross-section, whereas in the design of our feedthrough we adopted an elliptical ring based on the results obtained from electrostatic simulations. In the next figure is illustrated the electric potential and the field with the inner conductor at 300 kV and the outer at ground. The simulations were performed with COMSOL Multiphysics package. In the top plots we compare circular (right) and elliptical (left) ground terminations, where the coloured surface represents the value of the potential. The electric field along the side of the feedthrough as a function of the vertical coordinate z near the ground termination is shown as well. It can be seen that the elliptical termination allows to reduce the deviation of the equipotential surfaces, thereby lowering the critical field by 30% from roughly 120 kV/cm to 80 kV/cm.

* Electrostatic simulation of the HVFT:

The first feedthrough was manufactured according to these design requirements by the company CINEL Strumenti Scientifici. The first and most delicate step is to drill with high precision a continuous hole along the 2 meters rod of HMDPE. The diameter of the hole reduces from 43 to 32 mm to accept the feedthrough cable plug on one end and to allow the insertion of the inner conductor on the other. This step is quite critical mechanically, as the HMDPE is a rather soft material and it is not trivial to drill a precise and aligned hole in a 2 meter long rod. The inner conductor is then introduced in the rod and the last step consists in ”cryo-fitting” them in the outer conductor. Pictures of the feedthrough are provided in the figure below:

* The 300 kV high voltage feedthrough:

300kV High voltage test in LAr

The schematic representation of the test setup used in the test performed in September 2016 at CERN inside the WP8-AIDA work package is illustrated in the next figure. The setup consists of a 1 meter diameter vacuum insulated dewar with a 4 cm thick stainless steel cover that hosts the high voltage. feedthrough through its center. Inside the dewar the feedthrough is terminated by a circular 10 cm2 Rogowski shaped electrode. Some LEDs are installed in the upper part of the dewar. They can be switched on to help visual inspections of the LAr level and conditions inside the dewar. O2 impurities are also monitored through a gas analyzer model AMI 2001RS, with a sensitivity of 100 ppb. Before the filling, the vessel is evacuable in order to remove air traces, favour the outgassing of the materials and check the absence of leaks to atmosphere. The dewar is filled with liquid Argon purified through a molecular sieve (ZEOCHEM Z3-06), which filters water molecules, and a custom-made copper cartridge, which removes oxygen and other electronegative molecules. The liquid Argon level is visually checked through a vacuum sealed viewport and estimated also using a temperature sensor placed in the vicinity of the requested LAr level. The nominal value of liquid Argon level used in the test is about 500 mm from the bottom of the dewar. In this configuration the liquid is 100 mm above the top of the termination of the ground outer conductor of the feedthrough (see Figure 6). There is no active cooling of the liquid Argon. Instead the pressure inside the dewar is controlled by exhausting the boil off Argon through an external liquid argon bubbler. The dominant heat input to the liquid argon comes from the non-insulated top cover situated at about 1 meter above the nominal level. The voltage is provided by a 300 kV power supply through a HV coaxial cable. A 1:1 transformer coupled with the inner conductor of the cable and insulated by the polyethylene shield of the cable has been added to the cable to inductively couple the high voltage wire to an oscilloscope. When a pulsed current flows through the cable, it is detected on a scope. This sensor is used to monitor the frequency of possible micro discharges as well as the current delivered by the power supply during charging up. The residual heat input through conduction from the feedthrough in this particular setup was calculated to be ≃ 1.5 W.

* Experimental setup of the 300 kV test performed at CERN on September 2016:

Two series of tests were performed in September 2016. The procedure consisted in ramping up the high voltage while constantly monitoring the values of the current and the high voltage to verify the absence of discharges. Large discharges would result in a significant increase of current and a voltage drop directly visible on the power supply display. The transformer installed in the high voltage cable allows for further monitoring of micro-discharges that would potentially be out of the resolution range of the power supply (< 1mA).

For the first series of tests we slowly raised the high voltage on the power supply at a rate of about -10 kV/min. Thanks to the bubbler, we were able to maintain the pressure inside the dewar 50 mbar above the atmospheric pressure. O2 impurities were repeatedly measured in the Argon vapour and were found to be below 100 ppb. The surface of the liquid argon could be observed from the viewport and the absence of large waves or boiling could be checked visually. This exercise which was repeated twice is illustrated in the following figure. In both occasions we were able to reach the maximum voltage delivered by the power supply.

• Applied High voltage as a function of time during the tests:
  

A second series of tests were aimed at understanding the stability of the feedthrough during longer term operations. During these runs, we were able to ramp up until the end of the scale of the power supply in seconds. The results are summarised in the table below, where are indicated the voltages applied and the conditions during the three runs. In all cases we could reach and keep the voltage stable.

* Long-term tests performed during one hour and with the liquid level 10 cm above the nominal value. The simulated electric field at the critical point (i.e. ground termination) it also shown for each settings.:

Long-term operation of the VHV feedthrough tested inside the 3x1x1 m3 dual phase LAr TPC prototype

• HVFT inside the 3x1x1 m3 detector:
  

The 300 kV high voltage feedthrough has been operating in a stable way for several months in the 3x1x1 prototype at the lower voltage (-56.6 kV) required in order to obtain a field of 500V/cm over the the short drift space (1 m) of that detector and for shorter periods up to -75 kV. A detailed description of the construction and the running experience of the 3x1x1 detector is available on arXiv:1806.03317 [physics.ins-det].

Design and simulations of very large drift cages

Once the VHV has been generated and safely delivered inside the pure noble liquid, it is fundamental to ensure that all the detector components which are biased under such high voltage are accurately designed to minimize local electric fields. As example for the drift cage in liquid argon TPCs a solution has been found by using commercially available extruded Aluminum profiles. Those profiles are manufactured in large quantities at a low cost; they are polished in order to avoid sharp edges. In the context of task 8.5 detailed FEA simulations have been undertaken to study the electric fields and address the suitability of the design up to –300 kV. Simulations have been performed, the design optimized and an entire drift cage with external dimensions of 6x6x6 m3 is in the process of being installed. A detailed description of its design is for instance available at this presentation given at the ProtoDUNE dual-phase Technical Design review of April 2017.

• Simulation activities for the design of the ProtoDUNE dual-phase field cage. Left: picture of an extruded Al profile used for the 6x6x6 m3 prototype drift cage inside liquid argon. Center and right: images showing the design and simulations of large field cages made out of those profiles.:
  

• The ProtoDUNE dual-phase field cage after completion of its assembly:
  

The design and simulation studies of the 3x1x1 detector have also been at the basis of the conception of the drift cage system for the DUNE 10 kton dual-phase detector module. The Interim Design Report of the DUNE Far Detector 10 kton dual-phase module is extensively described in arXiv:1807.10340 [physics.ins-det]. Chapter 4 (page 75) describes the design of the VHV system which is based on similar structural elements as those developed for the ProtoDUNE dual-phase 6x6x6 m3 prototype detector at CERN. These elements (field cage sub-modules and cathode sub-modules) stay all within an envelope of 3x3 m2 which is required for their trasportation and underground installation.

Transparent ITO coated cathodes

This activity is in common with Task 8.4. An overview is shown at this presentation given at the first AIDA-2020 annual meeting. It consists of the possibility of building a cathode made of PMMA plates, integrated in a metallic structure, with a resistive ITO (Indium Thin Oxide) coating and a PTB coating on the upper surface closing the drift volume. This technique has advantages related to the light readout but also directly related to the photocathode design itself by limiting the propagation of bubbles, which may affect the dielectric rigidity of LAr in the drift volume and in making a segmented resistive cathode with a more regular surface and lower electric fields which limits the breakdown hazards. An intensive R&D program was carried out in this direction in order to develop the technology of ITO coating on large surface plates, to understand the behaviour of these plates at cryogenic temperatures and to define and simulate the support structure in which these plates are integrated. The design aspects for a possible implementation of this transparent cathode in the 6x6x6 m3 prototype were completely developed (see SPSC-SR-206 2017 annual report). An application of this concept for the dark matter experiment ArDM has also been developed (see this presentation).

Publications

First test of a high voltage feedthrough for liquid Argon TPCs connected to a 300 kV power supply C. Cantini et al 2017 JINST 12 P03021

A 4 tonne demonstrator for large-scale dual-phase liquid argon time projection chambers B. Aimard et al 2018, arXiv:1806.03317 [physics.ins-det]

Master thesis

ProtoDUNE-DP PROTOtype for the Deep Underground Neutrino Experiment - Dual Phase detector Master thesis ETHZ 2017 P. Chiu

Presentations

Task 8.5 Status Report, S. Murphy, AIDA-2020 Third Annual Meeting Bologna 24-27 April 2018

Outreach

Feeding the voltage through to the cage On track issue #5, WP8

A dual-phase DUNE, Symmetry magazine