Light readout for large scale cryogenic liquid detectors

The main physics goals of the large scale cryogenic liquid detectors are the accelerator-based long-baseline neutrino studies, dark matter searches, supernovae explosion studies, and proton decay searches. Expected or searched for signals can range in energy from a few keV to many GeV and have characteristic time duration and topological features that challenge the performance of large noble liquid Time Projection Chambers (TPCs). An essential and critical part of the Liquid Argon (LAr) TPC is the Photon Detection System (PDS), sensitive to light produced by interactions in argon. In Dual-Phase (DP) TPCs, the timing of prompt scintillation light (usually referred as S1 signal) in LAr is needed for time stamping of events and measurement of the drift coordinate of tracks in the detector. The extraction and amplification of drifted electrons in the gas phase also corresponds to the emission of light (usually referred as S2 signal) which yields information on the drift time and the amount of ionization, thus supplementing information from the charge readout on the anode plane. The interplay between the charge and light signals from an event allows to achieve precise pattern recognition as well as the measurement of the energy of interactions.

R&D activities in the framework of AIDA 2020 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 the networking activity of Task 8.4, concerns the large size scaling of cryogenic systems for the detection of the UV light produced at the same time as the charge ionization in noble liquid detectors. Task 8.4 covers the characterization and QA of large cryogenic photo-detectors (PMT, SiPM); wavelength-shifting techniques for large area readout in cryogenic detectors and digitization techniques of the scintillation signals. The WA105 3x1x1 m³ DP LAr prototype, built in the context of the neutrino platform activity at CERN, provided to WP8 in 2016 and 2017 the opportunity for benchmarking the technologies to be scaled at the level of a larger prototype,( ProtoDUNE -DP, a 6x6x6 m³ DP LAr TPC) under construction at CERN, and then to the 10 kton detector scale foreseen for DUNE. The groups involved in Task 8.4 could perform regular work by exploiting this test infrastructure and the ones set up for the preparation of ProtoDUNE -DP. Several aspects were investigated such as:

• the setting up of the QA chains
• different HV and signal schemes
• different coating techniques
• direct operation experience in cryogenic conditions over more than six months.

Table of Contents

Light readout in liquid argon

Introduction

A dual-phase photo-detection system detects light produced from two sources [1]: the scintillation process originated by ionizing particles propagating in the liquid argon (usually referred to as the S1 signal) and by the electroluminescence process due to drift electrons extracted from the liquid phase and accelerated in the argon vapor at the top of the cryostat (usually referred to as the S2 signal). Charge readout anode planes instrument the top while photodetectors are placed at the bottom of the cryostat. The interplay between the charge and light from an event enables pattern recognition and the measurement of energy of interactions.

Ionizing radiation in liquid noble elements leads to the formation of excimers in either singlet or triplet states, which decay radiatively to the dissociative ground state with characteristic S1 fast and slow lifetimes (fast is about 6 ns, slow is about 1300 ns in LAr with the so-called second continuum emission spectrum peaked at the wavelength of approximately 127 nm, 126.8 nm with a full width at half maximum of 7.8 nm [2]). This prompt and relatively high-yield (about 40000 photons per MeV at zero electric field) of 127 nm scintillation light is exploited in a LAr TPC to provide the absolute time (t₀) of the ionization signal collected at the anode, thereby providing the absolute value of the drift coordinate of fully contained events, as well as a prompt signal used for triggering purposes.

The secondary scintillation in the argon gas (i.e., the vapor phase) is unique to the dual-phase technology. It is the luminescence in gas caused by accelerated electrons in the electric field used for their extraction from the liquid and in the LEM anode holes, through Townsend amplification. The S2 signal provides information on the drift time and the amount of ionization charge, thus supplementing information from the charge readout on the anode plane. For a given argon gas density, the number of S2 photons is proportional to the number of electrons, to the electric field, and to the length of the drift path covered by the electrons in gas. In an extraction field of 3.0 kV/cm in gas, one electron generates about 100 downward-going photons that cross the liquid argon surface [3]. The time scale of S2 reflects the extraction time of original ionization in the liquid phase into the gas phase. Therefore, for about 0.5 kV/cm drift, this time scale is of the order of hundreds of µs. The time between the occurrence of the primary scintillation light and the secondary scintillation light is given by the drift time of the electrons from the ionization position to the LAr surface. This provides an accurate determination of the drift time in the active volume, and hence a correction tool for the electron attachment (and its effect on charge measurement).

Cryogenic Photomultipliers (PMTs)

Hamamatsu R5912-20Mod cryogenic PMT with the mount installed

PMT installed at the 3x1x1 m² detector,

Photomultipliers have been typically used as large area photodetectors in LAr. In particular, the 8-inch Hamamatsu R5912-20Mod cryogenic PMTs [4] have been successfully operated in the 3x1x1 m³ detector, and will be installed in ProtoDUNE -DP [5] as it can cover a large area in the detector (36 m²) and also due to its proven reliability on other cryogenic detectors. The same or similar PMTs have been successfully operated in other LAr experiments such as MicroBooNE [6,7] ,MiniCLEAN [8,9],ArDMm, [10,11] and ICARUST600 [12].

These devices have a typical gain of 10⁷ - 10⁹ at room temperature (RT), required to compensate the gain loss at cryogenic temperatures (CT). The high number of dynodes (14) of the R5912-20Mod PMT also has the advantage of requiring a lower operation voltage for a given gain compared with PMTs with lower number of dynodes, which reduces the potential risk of sparks and the heat dissipation in the PMT bases. As the PMTs are designed to operate at low cryogenic temperatures, a thin platinum layer was added between the bi-alkali photocatode and the borosilicate glass envelope to preserve the conductivity of the photocathode at these temperatures. The cathode sensitivity provides a spectral response from 300 to 650 nm. The PDS of the 3x1x1 m³ detector was formed by 5 of these PMTs and ProtoDUNE -DP will have 36 of these PMTs installed, which corresponds to 1 PMT/m².

An individual PMT support structure has been designed, manufactured and assembled at CIEMAT. This structure is mainly composed of 304L stainless steel with some small Teflon (PTFE) 6.6 pieces assembled by A4 stainless steel screws that minimize the mass while ensuring the PMT support to the cryostat membrane. The design was done taking into account the shrinking of the different materials during the cooling process to avoid breakage of the PMT glass. The PMTs will be placed at the bottom membrane of the 6x6x6 m³ detector.

Other devices (SiPMs)

On the other hand,SiPMs are also starting to be used as photon detectors in liquid argon. New SiPMs from Hamamatsu (3x3 mm – 100 um VUV3) sensitive to VUV light without wavelength shifter are being tested at CIEMAT. The aim of this first tests is to check how the devices behave at cryogenic temperatures and low wavelengths. Experts have discussed [13] the technologies (SiPMs vs. PMTs) in relation with each other. The problems that some technologies have and the research that needs to be done to solve them have been evaluated. PMTs are considered a mature technology, but with higher radioactivity than other technologies. SiPMs present the following features compared to PMTs: they can reach a quantum efficiency now near that of PMTs 35-40%; a much greater packing factor 73% vs. 40-50% coverage is feasible; infrared sensors above 900 nm could be interesting for large Ar TPCs; however, still have higher dark current rate; the readout is a challenge due to the high channel number and assembling SiPMs together; and the long-term stability and reliability need to be further studied.

Light Collectors (ARAPUCAS)

ARAPUCA array in ProtoDUNE-SP

ARAPUCA [14], is a light trap that captures wavelength-shifted photons inside boxes with highly reflective internal surfaces where they are eventually detected by SiPMs. The first large-scale implementation of an ARAPUCA module, in ProtoDUNE -SP, is composed of an array of sixteen ARAPUCA cells each one acting as an individual detector element. This configuration allows for finer spatial segmentation along the detector bar than is the case for the light guide designs. The ProtoDUNE -SP ARAPUCA design collects light from one side of the box through an optical window formed by a dichroic filter deposited with a layer of pTP (p-TerPhenyl) wavelength shifter on the external surface. This shifts the incident VUV light to a near-visible frequency that is able to pass through the filter plate to the interior of the box.

In the ProtoDUNE -SP version of the device, the inner surface of the box opposite the window houses an array of SiPMs that covers a small fraction of the area of the window (2.8-5.6%), surrounded by a foil of a highly reflective material coated with a second wavelength shifter, TPB. The TPB converts the light passing through the filter to a wavelength that is reflected by the filter. It has been shown in simulation and in prototypes that a large fraction of these trapped photons, reflecting from the filter and the lined walls of the box, will eventually fall on a SiPM and be detected. The X-ARAPUCA is a promising variant of the concept that uses a wavelength shifter-doped plate between two dichroic filter windows with SiPMs on the narrow sides of the cell; in addition to viewing scintillation light from both sides as needed for the central APA, it is expected to provide a higher light collection efficiency.

The ARAPUCA concept is relatively recent – it was first proposed in 2015 and accepted for installation in ProtoDUNE -SP in mid-2016. A series of tests in LAr have been performed with an evolving prototype design that resulted in detection efficiency measurements ranging from 0.4% to 1.8%, demonstrating the potential for substantially higher performance than the light guide designs. Monte Carlo (MC) simulations show that detection efficiencies at the level of several percent could be reasonably reached with improvements to the basic design. While the results of the experimental tests are encouraging, a deeper understanding of the optical phenomena involving emission and scattering on wavelength-shifter coated surfaces is needed to optimize the design.

[1] B. Abi at al. The DUNE Far Detector Interim Design Report, Volume 3: Dual-Phase Module, arXiv:1807.10340
[2] T. Heindl et al., “The scintillation of liquid argon,” EPL (Europhysics Letters) 91 no. 6, (2010) 62002.
[3] T. Lux, “Charge and Light Production in the Charge Readout System of a Dual Phase LAr6TPC
[4] Hamamatsu R5912-20, http://www.hamamatsu.com/resources/pdf/etd/LARGE_AREA_PMT_TPMH1286E.pdf
[5] L. Agostino et al. LBNO-DEMO: Large-scale neutrino detector demonstrators for phased performance assessment in view of a long-baseline oscillation experiment, arXiv:1409.4405
[6] T. Briese et al., Testing of Cryogenic Photomultiplier Tubes for the MicroBooNE experiment, JINST 8 (2013) T07005
[7] R. Acciarri et al., Design and construction of the MicroBooNE detector, JINST 12 (2017) P02017
[8] J. N. et al. et al., Demonstration of photomultiplier tube operation at 29 K, JINST 2 (2007) P11004
[9] T. Caldwell et al., Characterization of the R5912-02 MOD photomultiplier tube at cryogenic temperatures, JINST 8 (2013) C09004
[10] A. Marchionni et al., ArDM: a ton-scale LAr detector for direct Dark Matter searches, J. Phys.: Conf. Ser 308 (2006) 012006
[11] A. Badertscher et al., ArDM: first results from underground commissioning, JINST 8 (2013) C09005
[12] G. Raselli et al., Test and characterization of 20 pre-series hamamatsu R5916-MOD photomultiplier tubes for the ICARUS T600 detector, IEEE Nuclear Science Symposium, Medical Imaging Conference and Room-Temperature Semiconductor Detector Workshop (2016) 1–5
[13] LIDINE 2017: Light Detection In Noble Elements, https://indico.physics.lbl.gov/indico/event/545/
[14] B. Abi at al. The DUNE Far Detector Interim Design Report, Volume 2: Single-Phase Module, arXiv:1807:10327

Coating techniques for PMTs

TPB coating set-up

WA105 3x1x1 m³ PMT with TPB evaporated on it

WA105 3x1x1 m³ PMT with TPB evaporated on plate

In LAr detectors, we need to convert UV LAr scintillation at 127 nm into visible photons where PMTs (or SiPMs are sensitive by the use of suitable wavelength shifting material. The concept of coating the PMT windows with a thin film of TPB (TetraPhenyl Butadiene) has been validated with the 3x1x1 m³ detector. TPB is a wavelength shifter with high efficiency for conversion of LAr scintillation VUV photons into visible light, where PMT cathode is sensitive. The TPB is deposited on the PMT by means of a thermal evaporator which consists of a vacuum chamber with two copper crucibles (Knudsen cells) placed at the bottom of the chamber, following the sanding of the PMT window. A PMT is fixed at the top of the evaporator, with its window pointing downwards, on a rotating support in order to ensure a uniform coating. The crucibles, filled with the TPB, are heated up to 220ºC. At this temperature, the TPB evaporates through a split in the crucible lid into the vacuum chamber, eventually reaching the PMT window.

Several tests were performed in order to tune some parameters like the coating thickness (TPB surface density) and the deposition rate. For the tests, a PMT mock up covered with mylar foils has been used. A TPB density of 0.2 mg/cm² was chosen for ProtoDUNE -DP as this is the value where the PMT efficiency is stable as a function of the density. Efficiency measurements were performed using a VUV monochromator by comparing the cathode current of a coated PMT with the current value of a calibrated photodiode. As a result of the efficiency tests, about 0.8 g of TPB must be placed in the crucibles at each evaporation, in order to achieve the desired PMT coating density. This value optimizes the quantity of TPB used per evaporation keeping, at the same, the coating density fluctuations below 5%. With these specifications, two to four PMTs can be coated per day at a single coating station.

The setup available at CERN described in [JINST 13 (2018) P12020] was used to coat the PMTs for the 3x1x1 m³ and 6x6x6 m³ LAr DP detectors. Also, ICARUS coated 360 PMTs with it. The position of the PMT is optimized for uniform coating. In the 3x1x1 m³ detector, two TPB coating methods were used. In 3 PMTs, TPB was coated in PMT glass, and in 2 PMTs into acrylic plates. In both cases, TPB was deposited by vacuum evaporation (0.05 – 0.2 mg/cm2). The 3x1x1 m³ a light data are compared to simulations, and preliminary results point to a light loss in the acrylic plates, as expected. For ProtoDUNE -DP, the 36 PMTs have been coated with more than 0.2 mg/cm² in the middle. The maximum difference in thickness is expected to be <20% from the extreme edge to the middle (middle-center <5%). The quantum efficiency of 4 PMTs has been measured to be 14% at 127 nm after coating.

Development of QA methods for PMTs testing

The ProtoDUNE -DP detector will consist of a 6×6×6 m³ LAr TPC to be operated at the CERN Neutrino Platform and the PDS will be formed by 8-inch cryogenic PMTs from Hamamatsu. The PMT model (R5912-20Mod) performance at CT is studied including dark current, gain, and linearity with the light intensity and pulse rate. At cold, a decrease of the PMT amplification, or fatigue effect, is measured as the PMT output current increases, either, due to high gain, light intensity or rate. Also, the characterization results of the 40 photomultipliers to be used in ProtoDUNE -DP are presented. The work presented in this section has been published in [JINST 13 (2018) T10006].

Testing set-up

A dedicated test bench was designed at CIEMAT for the PMT characterization. The measurements are performed to 10 PMTs at the same time inside a 300 L vessel at RT and filled with LN₂ at 77 K for the CT tests. A 400 L tank supplies LN₂ at 2 atm pressure through a pipe directly to the 300 L vessel where the PMTs are located at ambient pressure. The system is automatically filled, using electro-valves controlled by level probes and temperature probes through a PC. The 10 PMTs are distributed in two levels, five per level, fixed to the lid through an internal structure. The main PMT features to measure are dark current, gain, and linearity. For each feature, a different electronics set-up is arranged using VME and NIM modules from CAEN. The DAQ is remotely controlled with the aim of automating the data acquisition with LabVIEW software. The light sources used are a PicoQuant GmbH5 laser head, with a 405 nm wavelength and a pulse width less than 500 ps FWHM, and an LED pulser, with a 460 nm wavelength and a pulse width of 40 ns. The amount of light is tuned using a set of UV optical filters and the light is diffused in the detector volume.

Schematic drawing of the PMT testing experimental set-up

Picture of 10 PMTs installed in the testing dewar

Methodology

The studies carried out to fully understand the PMT behavior at CT include the dark current, the gain, and the linearity.

• Dark current: The dark current (DC) rate is the response of the PMT in absence of light. It is known that the main contribution to the DC at RT is the thermionic emission. However, a non-thermal contribution increases the DC rate at CT. The DC rate is estimated as the average rate of detected signals larger than 7 mV, which assures single photo-electron (SPE) triggering at an operating gain of 10⁷ in a completely dark state. At RT, measurements are taken after at least 15 hours of complete darkness inside the vessel.

• Gain: To characterise the gain of each PMT, a small amount of photons is sent to the PMT to obtain the SPE spectrum. The SPE spectrum is fitted to a convolution of a Poisson distribution, which models the number of photo-electrons (p.e.) generated in the photocathode, and a binomial distribution considering two possible amplification paths: through the first dynode and directly starting in the second dynode. The gain vs HV, known as gain-voltage curve, is measured from 1100 V to 1900 V in 100 V steps. Then, a fit is done following the power law G = AV^B being A and B constants dependent on the number, structure, and material of the dynodes. The PMT gain is monitored over time after the tests that produce a high PMT output current. First, the recovery time is studied at 1400 V after gain vs HV measurements. Although the average number of p.e. is lower than 1, at the maximum voltage (1900 V) the output current is high enough to reduce the effective PMT gain by 30% with respect to the initial measured gain at 1400 V and the gain recovery requires about a day at CT. Second, after the linearity measurements where PMTs are exposed to light levels of 1000 p.e. and up to a 10⁸ gain, a similar behavior is observed. Finally, a more permanent effect over the gain is caused by high rate signals of the order of MHz. In this case, the gain decreases by a factor of 2. For some PMTs, a week after illuminating them with high frequency signals, the gain is still 40% lower than the nominal gain. However, other PMTs recover faster (in 3 days)

SPE spectrum (red) of a PMT at CT at 1500 V, and fit results (black)

(a) Gain evolution at 1400 V after raising the voltage to 1900 V. (b) Gain recovery time after linearity measurements. (c) Gain recovery time after high rate measurements

• Linearity: The PMT response is studied as a function of the light intensity. The amount of scintillation light arriving to the PMTs in ProtoDUNE -DP will vary from a few p.e. to thousands of p.e. depending on the particle track energy and distance to the PMTs. Thus, the response of the PMTs should be linear in a wide dynamic range. The PMT base was designed to fulfil this goal, and the deviation of the linear response is primarily caused by the anode saturation. To avoid the fatigue effect, measurements are done from low light (SPE) to high light levels (up to 1000 p.e.). The expected amount of light is estimated relatively to the measured one in the linear region, taking three reference filters and considering their transmission factor. The 405 nm laser (<1ns pulse) and the 460 nm LED (40 ns pulse) are used to check the effect of the pulse profile on the anode saturation. For the same amount of detected light, the shorter the light pulse, the higher output peak current leading more easily the PMT to saturation due to space-charge effects on the last dynodes.

• Light rate: In ProtoDUNE -DP, a continuous background of light pulses is expected due to the secondary scintillation light produced on the gaseous phase of the argon by the drifted electrons. To study how this can affect the PMT performance, dedicated tests are carried out. The light intensity is set to different levels (few p.e. to 150 p.e.) and for each light level, the light pulses emission frequency is swept from low (100 Hz) to high frequencies (10 MHz). The tests are carried out with the laser and the LED to check the effect of the pulse profile. Tthe charge spectra of the PMT is obtained to calculate the averaged amount of light observed by the PMT. Increasing the rate of the light pulses produces a proportional increase of the average output current. As the base design is based on resistors, the PMT inter-dynode voltages have a dependency with the PMT output current making the PMT response not linear when the light rate increases over a certain limit. The current through the base resistors has two components: the polarization current, constant and provided by the power supply, and the PMT output current that goes through the resistors in the opposite direction. As the total voltage applied to the base is kept constant by the power supply, a decrease on the last dynodes voltage (by the increase of the output current), makes higher the voltage on the first dynodes increasing the PMT gain. If the voltage on last dynodes continues decreasing (by the increase of the output current) at some point this voltage is not enough to maintain the current flow to the anode, and it decreases following the typical I-V curve of a vacuum diode. The only difference is that in the vacuum diode, the cloud of electrons is generated by a filament while in the PMT it is generated by the incident light and the previous dynodes. Besides that, the PMT output current does not depend any more on the light input intensity, only on the light rate that is changing the voltage on the last dynodes. Then, the expected PMT response vs. light rate can be divided in three zones: first, the linear response, from <1 Hz; second, the over-linearity region, where the PMT gain increases; and third, the saturation region, when the voltage between the last dynode and the anode is close to zero and the PMT output decreases with the light rate. The PMT output is never zero because the initial velocity of the p.e. is not zero, therefore some of them can be collected by the last dynode even if the potential difference between the cathode and the first dynode is zero. To reach the zero output this potential should be negative. At CT, the fatigue effect observed on the PMT gain, compensates the beginning of the gain increase with the light rate leading to a small reduction of the over-linearity region.

PMT validation results

• Dark current: The DC rate as a function of the HV is measured. In general, the DC at CT is higher than at RT for the same gain. For instance, for a particular PMT, the DC increases from 0.6 kHz at RT to 1.9 kHz at CT for a gain of 1.5·10⁷, being the behavior of this PMT representative of other PMTs

DC vs gain at RT and CT for one PMT.

• Gain: Gain-voltage curves at RT and CT (at 77 K) are measured and the slope of the curves follows a simple power law. For the same HV, the gain at CT is lower than at RT. Being the gain 10⁷ at RT, it decreases by 76% due to cryogenic conditions. An increase of 170 V is required on this PMT to compensate the gain loss at CT. The gain is also measured at a temperature closer to the one in ProtoDUNE -DP, which will be 94 K, taking into account the pressure at the bottom of the LAr cryostat. To achieve a higher temperature, an over-pressure of 1 bar is applied to the vessel, which is the maximum allowed by the set-up. While the temperature with LN2 at atmospheric pressure is 77 K, during these tests 83 K are reached. Gain variations are found compatible with expectations. For the same PMT, a 10⁷ gain at RT decreases 66% at 83 K. This means that an increase of 130 V in the HV is required at RT to obtain the same gain. In total, 4 PMTs were measured at 83 K, and the gain decreases on average 60%.

Gain vs HV at RT, at two CT (77 K and 83 K) for the same PMT

• Linearity: The response of several PMTs as a function of the light intensity is measured at RT and CT. For gains <10⁶ the response has been observed to be linear in the tested range (up to 1000 p.e.). For gains close to 10⁷, the PMT remains linear up to 200 p.e., while for larger gains (>10⁸) the PMT response deviates from linearity with only 75 p.e., and a saturation regime is reached. Same response is observed for all tested PMTs. The PMTs are also illuminated with the LED. The linearity range increases when the light source is wider, showing a clear dependency with the pulse shape in the PMT saturation. At CT, the PMT response is just slightly worse than at RT when the PMT is illuminated by the laser. For instance, for a 10⁸ gain and 500 p.e., the measured number of p.e. is 60% and 70% lower than expected at RT and CT, respectively. However, the linear range when the PMT is illuminated by the LED at CT is shorter than at RT, and similar to the laser at CT. Therefore, despite the very good linearity at RT, the PMT response is saturated at 300 p.e. for a gain of 10⁷ at CT.

Comparison of the measured vs expected number of p.e. illuminating the PMT with the LED at RT and CT

• Light rate: The PMT response is studied for pulsed frequencies from 100 Hz to 10 MHz using the laser and the LED as light sources and different light intensities (from 10 to 50 p.e.). At RT, the three regions explained before are observed: the PMT response is flat until a given frequency (>10 kHz) which depends on the charge; then, the overlinearity effect is observed; and the PMT saturation (>500 kHz). At CT, the PMT response is expected to be the same as at RT, as the saturation curve depends only on the base design. The expected PMT gain reduction, as the average output current increases, compensates and reduces the over-linearity region. The same result is observed for different PMTs. The frequency sweep is also done using the LED, and the results, suggest that the response is very similar, but the over-linearity starts at slightly higher frequencies when the LED is used. As the LED light pulses are wider, the PMT output peak current is smaller (for the same charge) moving the over-linearity effect to higher frequencies. For low intensity signals, as expected for the S2 light in ProtoDUNE -DP, the PMT response is linear until 1 MHz, far enough for our purpose.

PMT LED frequency response at RT and CT for different amounts of light.

ProtoDUNE -DP PMT characterization

• Dark current: The DC rate is measured for the 40 PMTs at several HVs. In particular, the DC rate at 10⁹ gain is measured to compare with the value given by the manufacturer. In general, similar results are obtained. However, two of them had to be replaced, as one had no signal and another one showed a DC rate of around 100 kHz (almost 60 times the expected rate from Hamamatsu). On average, as shown in Figure 20(b), DC rate is 0.4 kHz when the PMTs operate at 10⁷ gain at RT, and DC rate is always below 1.4 kHz. However, at CT, DC rate increases up to 1.7 kHz, and some PMTs reach up to 2.5 kHz. No correlation between DC at RT and CT is observed. The PMTs with very unstable DC rates in time were inspected by Hamamatsu and although no defects were found, they can be designated as spares.

DC histograms for 10⁷ gain at RT and CT measured at CIEMAT for the 40 PMTs. At RT, DC rate is on average 0.4 kHz, and at CT, 1.7 kHz

• Gain: The gain results from the characterization of the 40 PMTs are presented here. A good correlation between the HV determined at CIEMAT for a 10⁹ gain, and the HV required according to manufacturer specifications (both values at RT) is observed, with only 2.3% deviation possibly due to differences in the setup or in the gain determination method. To achieve a 10⁷ gain, higher HV, 170±72 V on average, needs to be applied at CT to reach the same gain, which is equivalent to 70% gain drop at CT.

Histogram of HV required to obtain a 10⁷ gain at RT and CT. The average HV is 1154 V at RT, and 1324 V at CT

Development of HV/signal unified cabling

In order to choose the optimal PMT base design for the 6x6x6 m³, two possible voltage configurations (positive and negative bias) were considered. More details about this work have been also published in [JINST 13 (2018) T10006].

In the so called positive base (PB), a positive HV is applied at the anode and the photocatode is grounded, which reduces the noise; whereas in the negative base (NB), a negative HV is applied at the cathode and the anode is grounded. In the negative bias configuration, the photocathode is connected to HV, and special care must be taken to prevent spurious pulses due to HV leakage through the glass tube envelope to nearby grounded structures. The NB configuration requires two cables, one for HV and the other for the signal readout. Nevertheless, in this configuration it is easier to read the PMT signal as it is referenced to ground. On the other hand, one advantage of the PB is that only one coaxial cable is required to carry the positive HV and to receive the signal from the PMT, but a decoupling circuit is needed to split the HV and the PMT signal. A dedicated splitter circuit was designed to perform this function out of the cryostat. The splitter affects the real voltage value that is sent to the PB which is expected to be 7% lower than the one read on the power supply. After some validation measurements, the PB design was selected, as the total number of cables and feedthroughs in the detector is reduced and its behavior in terms of linearity and dark current is slightly better than the NB.

Schematic drawing of the two HV divider options for the base

Diagrams of the two bases considered: (a) positive base and (b) negative base

Once the PB was selected, all the bases were assembled, cleaned and tested in air, Ar gas and liquid nitrogen. Two tests were performed to the bases before being soldered to the PMTs and tested at CT: a resistance value of 13.4 MOhm was confirmed, and 2000 V were applied to the bases in Ar gas to verify the absence of sparks.

In order to validate the design of the PMT base, the PB and NB are compared at RT and CT in terms of gain, dark rate, and linearity. As expected, the PMT gain with both configurations is similar. In order to compare DC rate vs voltage dependence, both bases are tested on the same PMT and on the same darkness conditions. The results show that the behavior of both bases is similar at voltages up to 1600 V, and at higher HV the DC rate increases more on the NB reaching a rate around 50% higher than the PB at 1900 V. The base with negative power supply shows higher dark rate than the positive one because the photocathode is at high voltage and spurious pulses could appear due to current leakage through the glass or due to electro-luminescence in the glass. For the linearity response, the positive base shows slightly better results. On the positive bias base, the power supply filter capacitor is closer to the anode which increases the charge reservoir for the PMT output increasing slightly the linearity range. For the study of the response with the light rate. The positive circuit shows also better results. The difference between the two bases is also due to the different position of the filtering capacitor on the base. To verify it, a PB without this capacitor was also tested and its behavior was slightly worse than with the NB, making clear that the presence of this capacitor close to the anode improves the linearity of the PMT. Increasing the capacity of this capacitor did not improve the response any further at the tested light levels. These tests confirm the better performance of the PB.

Response vs light rate for NB, PB and PB without filtering capacitor at RT with the laser

Development of transparent cathode

Another point developed in Task 8.4 activity consists in the possibility of integrating the wavelength shifting technique in a transparent photocathode made of PMMA plates with a resistive ITO (Indium Thin Oxyde) coating and at PTB coating on the upper surface closing the drift volume (development performed in common with the Very High Voltage task 8.2). This technique allows for a clean definition of the light production volume exactly corresponding to the charge readout volume and for the separation of the wavelength shifting function from the surface of the photodetectors. An intensive R&D program was been carried out in this direction, fully demonstrating the validity of the wavelength shifting technique integrated in the transparent cathode. The design aspects for the implementation of this transparent cathode in the 6x6x6 m³ prototype were completely developed. Its implementation was discarded for ProtoDUNE -DP as the scientific committees found the level of experience with this technique still premature to mitigate the risks related to a full implementation in ProtoDUNE -DP. Then ProtoDUNE -DP and the 10 kton dual-phase detectors are based on the traditional cathode design. This R&D line will be further tested in the future.

Digitization of light signals

In order to design the system to digitize the light signals, one has to take into account that the following information needs to be extracted from the PMT signals:

• S1 fast component shape, charge and timing;

• S1 slow component shape;

• S2 shape, charge and timing (distance from S1 and duration);

• Single photoelectron (SPE) charge spectrum for gain calculation during PMT calibration;

• Trigger signal generation by the coincidence of several PMT signals.

In general, the PMT signal dynamic range goes from the mV level to several volts (over 50 W load). During the operation of the 3x1x1 m³ DP demonstrator, PMT signals larger than 2 V were observed with PMT gains around 10⁶. The light levels in the future DP detectors will have a larger dynamic range due to its large volume, therefore higher gains will be required to see the far light signals. However, higher gains increase the output from closer light signals, requiring that the FE electronics cover a large range of input voltages. To cover a dynamic range of 10 V with a resolution below the mV level, 14 bits are necessary (least significant bit (LSV) 0.6 mV). For 2 V of dynamic range 12 bits would be sufficient (LSB 0.5 mV). Results from ProtoDUNE -DP and relevant simulations are needed to determine the required dynamic range for future detectors.

The sampling frequency also affects the time tagging precision. The time uncertainty due to the PMT alone is around 3 ns (transit time spread). Other factors, e.g., Rayleigh scattering, increase this uncertainty, as does the sampling period; therefore, the lower sampling frequency, the better. In the 3x1x1 m³ DP demonstrator 4 ns sampling was used to digitize waveforms. The rate of the events observed in the demonstrator was around 300 kHz with the threshold at the SPE level. The light signal digitization has to be synchronized with the DAQ. All the ProtoDUNE -DP DAQ electronics use the White Rabbit (WR) protocol for synchronization. A dedicated White Rabbit Micro Telecommunications Computing Architecture (µTCA) slave node on the light readout FE electronics distributes clocks and timing information to the different FE digitization cards.

SPE waveforms and amplitudes from the 3x1x1 m³ detector at different voltages.

Event rates for different trigger thresholds observed in the 3x1x1 m³ DP demonstrator.

Light calibration system

One of the main goals of the PDS is to provide trigger for non-beam physics. The trigger is based on the amplitude of PMT-signals. The amplitudes of the PMT-signals are summed for groups of certain PMTs and/or for all PMTs and then, these input signals are discriminated according to the trigger logic. An equalized PMT response allows to use the same threshold definition for all PMT groups, simplifying the determination of the trigger efficiency. Beside measuring the PMT gain, it is also designed to monitor the stability of the PMT response, and so its quantum efficiency, during the performance of the 6×6×6 m³ detector. As concluded from the operation of the 3×1×1 m³ LAr TPC detector and the PMT characterization at cryogenic temperature, a LCS is strongly recommended during the experiment data taking period. The details about ProtoDUNE -DP LCS have been published in [arXiv:1902.07127].

Design

An LED-driven fiber calibration system is designed so that a configurable amount of light reaches each PMT. The calibration light is provided by a blue LED of 465 nm using a Kapustinsky circuit as LED driver and transmitted by a fiber system ending with an optical fiber installed at each PMT. There are 6 LED placed in a hexagonal geometry and a reference sensor to check the LED performance in the center of each group. The direct light goes to the fiber, and the stray light to the SiPM used as reference sensor. Each LED is connected to an external fiber going to one feedthrough. Then, fibers are connected inside the cryostat and each one of these fibers is attached to a 1-to-7 fiber bundle, so that one fiber is finally installed pointing at each PMT. The PMTs are oriented with the first dynode perpendicular to the Earth magnetic field and the fiber parallel to the first dynode to have a similar gain to the one obtained with diffuse light. The components placed outside the cryostat at room temperature form the external system and the ones installed inside it at cryogenic temperature the inner system.

Diagram of the PMT calibration system designed for ProtoDUNE-DP

The design of the external components is driven by the aim to have a cost-effective system. An additional requirement is that a reference light sensor monitors the amount of the injected light. The light is injected in form of several ns long pulses provided by a Kapustinsky circuit. The setup consists of a commercial black box in which a light guide structure is mounted to guide the light. On each of the six arms, an electronics board containing a Kapustinsky circuit is mounted. The LED, NSPB300B from Nichia Corp, with a peak wavelength of 465 nm is placed on the PCB in front of an optical SMA to SMA feedthrough. On the other side of each feedthrough an optical fiber, FG105LCA -CUSTOM-MUC from Thorlabs is connected. It transports the light to one of the feedthroughs in the instrumentation flange on top of the cryostat. While a large fraction of the LED light is emitted in the forward direction, a small fraction, the stray light, is emitted under a large angle and reaches by reflection to the central region of the light guide structure where it is detected by a SiPM, MicroFJ -30035-TSV-TA from SensL .

(a) Photo of the light guide structure during the development phase. The 6 arms are visible, and on one of them a prototype LED driver is mounted. During these tests, the SiPM was mounted on a commercial development board. (b) Schematics of the light way from the LED to the reference sensor

The inner system is designed to minimize the light losses at cryogenic temperature. The external fibers are connected to 6 female optical feedthroughs from Allectra installed at 2 flanges. Inside the cryostat, a single long fiber, FT800UMT from Thorlabs, goes down from each optical feedthrough routed along the walls of the cryostat to its bottom where a 1-to-7 fiber bundle, composed by FT200UMT fibers from Thorlabs, is connected to each long fiber. A total of 36 of these fibers are guided to the PMTs at the bottom of the detector. The end of the fiber is fixed at the PMT support structure pointing the photocathode. The fibers and bundles are 0.39 NA TECS^TM hard-clad, multimode, step-index fibers with high OH to increase the light transmission at low wavelengths. In order to optimize the light transmission of the fiber-bundle connection, the inner fibers have a diameter of 800 µm, big enough to distribute uniformly the light at the bundle entrance, total diameter 700 µm. From the mechanical point of view, the described approach of bundles attached to fibers is safer than connecting directly the LAr to freeze inside the connector which would reduce the light transmission.

1-to-7 fiber bundle end where the 7 fibers can be seen

Characterization

A systematic characterization of the external and internal components is carried out in the laboratory to evaluate the feasibility of the system for ProtoDUNE -DP.

The external components, the LED driver PCBs and the central PCB with the Sensl SiPM are characterized using a calibrated photodiode and a 1 inch PMT. The photodiode provides the average power over many light pulses while the PMT signal is used to study the pulse shape of the different LED driver PCBs. It is studied if the repetition rate of the light pulses leads to any saturation effect. The optical fiber is directly connected to the powermeter using a suitable SMA adapter. The output curves indicate that the drivers show no frequency depending effect within the tested range (100 - 10100 Hz). The time stability of the LED driver is studied measuring the output power during almost 45 hours and variation of the output power is found to be stable within the precision of the powermeter of 0.01 nW. The response of the SiPM acting as reference sensor is also studied. The stray light is measured with the SiPM and the built-in ADC of the BeagleBone, while the output power is measured at the end of an optical fiber. The measured signal describes well the output power until 18 V LED bias voltage, where saturation effects set in.

All the inner components are characterized at RT and CT, using liquid nitrogen for the cryogenic tests. The expected light transmission from the flange to the PMTs is calculated and compared with the amount of light detected by the PMTs in a dedicated set-up. In order to characterize the fibers and bundles individually, they are directly connected to a 460 nm LED and the light output is measured with a powermeter at the end of the fiber. A homogeneous light output is measured among the fibers and bundles, with alight uniformity of 90%. The light attenuation at CT with respect to RT is 0.8 dB for the fibers and 1.9 dB for de bundles. Finally, the fiber - SMA to SMA vacuum MS - bundle system is measured at RT and CT. In total, 3.1 dB of attenuation is measured for CT operation. The total attenuation expected at 465 nm is <16 dB at RT and <19.1 dB at CT, dominated by the geometrical efficiency, which accounts for dividing from one to seven fibers (12 dB). The estimated light attenuation of the total system is compared with the attenuation from dedicated PMT measurements at CT. The set-up designed for the ProtoDUNE -DP PMT characterization is used. The attenuation of the inner system in LN₂ measured by the PMTs is 20.4 dB consistent with the expected one.

Average number of p.e. measured by the PMTs versus the number of photons at the flange

Before installation at CERN, the performance of the complete LCS is validated with the final components, and with the inner system at CT. The goal of the LCS is to measure the PMT gain and to study the PMT response stability. Three different measurements will be performed during the ProtoDUNE -DP operation: gain stability at the operating voltage, gain calibration curve, and PMT response at several p.e. level to monitor the quantum efficiency, once the gain is known. The same measurements are carried out during the full system validation. In each case, measurements are taken with one LED providing simultaneously light pulses to 6 PMTs at a time. The gain of each PMT is measured taking SPE spectra at the operating HV and compared to the gain of these PMT measured using a diffuser inside the vessel, and similar gains are obtained. Also, the gain curve is mapped successfully with the gain as a function of the high voltage applied to the PMTs.The response of the PMTs for different light levels can be studied to confirm the optical path integrity, look for relative changes in the PMT quantum efficiency, or as alternative method to determine the PMT gain. On average, the maximum light expected from the LCS at CT is 60 p.e. which is enough for the planned measurements. The light output is limited by the maximum LED voltage (19.5 V) and by the SiPM saturation level. It could be increased if the reference sensor is not needed, i.e. once the correct functioning of the LEDs is ensured

Performance of the PDS of the 3x1x1 m³ DP detector

To validate the concept of a non-evacuated industrial cryostat, test several key sub-systems for ProtoDUNE -DP, and demonstrate for the first time the capabilities of the DP LAr TPC technology on a tonne scale, a demonstrator with an active volume of 3x1x1 m³ was built in 2016 and operated in 2017 at CERN. This is the largest DP LAr TPC operated to date, and the details have been published in [JINST 13 (2018) P11003].

The demonstrator consists of a 3x1x1 m³ (4.2 t) active volume DP LAr TPC inside a passively insulated cryostat with internal volume of 23 m³. The detector is suspended under a 1.2 m thick insulating lid called top-cap where the necessary feedthroughs are hosted. Ionisation charges drift vertically towards the liquid-vapour boundary where they are extracted into the gas phase, amplified by LEMs, and collected on finely segmented anodes. The electron extraction, amplification, and collection are performed inside a 3×1 m³ charge readout planes (CRP). The CRP is electrically and mechanically independent from the drift cage and can be remotely adjusted to the liquid level.

Schematic drawing of the 3×1×1 m³ DP LAr TPC in the cryostat.

Five PMTs are mounted underneath the TPC drift cage. They detect the scintillation light from argon excimer states formed by charged particles crossing the liquid volume (primary scintillation, S1), as well as the secondary scintillation light (S2) from the electroluminescence of the electrons extracted in the argon vapour. TPB is used to convert the deep ultra violet photons of the argon scintillation peaked at 127 nm to the visible spectrum within the sensitivity of the PMT photo-cathode. The PMT signals give the absolute time on an event, T₀, and provide a trigger for the data acquisition system. The amount of detected light can also be useful for calorimetric measurements of the deposited energy.

The light readout signals are acquired using commercial electronics. The analogue signals from each PMT are digitised over a 1 ms window, which corresponds approximately to the maximum drift time of the electrons. The digitisation is performed with a resolution of 12 bit sampled at 250 MHz using a CAEN v1720 board. To limit the data volume in some runs, a 4 µs event time window was acquired to record only the S1 signal. The ADC has a total dynamic range of 2 V limiting the PMT gain. The board is read out via an optical link to a PC equipped with a CAEN A2818 PCI CARD. The software for the display and acquisition is based on the MIDAS framework and runs on the same PC. The board also allows to program a simple majority coincidence trigger.

One of the technological milestones of the 3x1x1 m³ is the detection of prompt and secondary scintillation. The 3x1x1 m³ detector would allow to study the properties of LAr scintillation and the light propagation in 3 m³ volume. The detection of the electroluminescence (secondary scintillation), produced by the ionisation charge in the argon vapour and never before measured in a LAr TPC with a 1 m drift, could give additional handle on the amount of charge reaching the CRP.

Prior to installation, the gains of the five PMTs are measured in air at RT by studying the response to a SPE produced at the photo-cathode. Since at cryogenic temperatures a decrease of the gain is expected, the latter must be remeasured once the detector is filled in order to equalise the PMT responses in liquid argon. Dedicated runs were taken using a pulse generator as a trigger running at a frequency of 100 Hz. Data were digitised with the sampling of 250 MHz and the gain was measured in an offline analysis of the collected waveforms. Every fluctuation from the baseline along the waveform is integrated to get the corresponding charge. Since a random trigger is used, the most common signal should be dark current at the level of a SPE. The PMT’s equipped with positive base provides larger gain for equal applied voltage.

Left: distribution of the integrated charge for one PMT in the single PE analysis. Right: PMT gain in liquid argon for different operating voltages

We first study the prompt liquid argon scintillation by not applying any drift electric field on the TPC. The average of multiple digitised waveforms from one PMT is fitted with a superposition of three exponential functions convoluted with a Gaussian in order to include both the scintillation time structure and any experimental effects such as the jitter of the trigger device (5 ns), the time of flight of the primary ionising particle and the light propagation time (10 ns). The scintillation time profile of liquid argon consists of a fast de-excitation of both a singlet state (fast component) and a triplet state (slow component), whose value depends on the purity of the liquid. A third, intermediate, component is needed in order to improve the agreement with the data. The intermediate component has been observed at different setups ranging from tenths of ns to values above 100 ns in good agreement with present results.

Average digitised signals from one PMT fitted with the function described in the text. The lifetime of the fast component is fixed to 7 ns in the fit to the gas data

By operating the TPC with a drift and extraction field, we are able to observe the proportional scintillation of the ionising charges in the argon vapour. The proportional scintillation, also referred to as electroluminescence or S2, is expected to take place in the gas in regions where the electric field is above 2 kV/cm, which is the case in the extraction gap but also inside the LEMs and in the induction gap with electric fields of the order of 30 kV/cm. The proportional scintillation signal lasts until the farthest electrons reach the extraction region. The maximum time extension of the secondary light signal is therefore comparable to the maximum electron drift time. The secondary light contribution extends for about 600 µs, which is the expected time for an electron drifting over 1 m in a field of 0.5 kV/cm. The almost flat S2 continuum qualitatively indicates that the liquid argon purity is sufficiently high so that there is no substantial attenuation in the extracted ionising charge over the entire drift distance. The effect of the larger electric field inside the LEM is clearly visible and contributes as an enhancement of the S2 signal, as well as to an increase of the lifetime. Nevertheless, at those electric field settings, the amplitude of the S2 continuum remains several orders of magnitude lower than that of the primary scintillation peak. Since oxygen contamination would reduce the light signal because of quenching, the liquid argon purity can be verified by studying the lifetime of the triplet state in liquid. In the liquid phase, the measured value of 1.6 µs is compatible with oxygen equivalent impurities better than 1 ppm according to estimations based on previous experimental data.

Sum of averaged waveforms from the 3 negative based PMTs showing the prompt and proportional scintillation components. The prominent peaks centred at around 200 µs correspond to the prompt scintillation in liquid (S1) and the continua extending for 600 µs after are from the proportional scintillation in gas (S2). The higher S2 yield is clearly visible when the LEMs are polarised (red curve)

It is studied the correlation between the amount of collected charge versus the amount of detected S2 light with the PMT array after the contribution of the prompt S1 signal is subtracted in the analysis. The evident correlation between two signals demonstrates the sensitivity of the photon detection system to the electroluminescence in the vapour phase and the integrated amount of the S2 light could serve to provide additional information on the total amount of deposited charge.

The S2 light collected by the PMT array as function of the integrated charge collected on the anode.

Prospects: Large scale LAr detectors (DUNE and ProtoDUNE -DP)

The construction of LAr detectors at the 10 kton scale is an essential ingredient of the future international long-baseline neutrino program unifying the European and USA efforts. The DUNE experiment aims to address key questions in neutrino physics and astroparticle physics. It includes precision measurements of the parameters that govern neutrino oscillations with the goal of measuring the CP violating phase and the neutrino mass hierarchy with a muon neutrino beam produced at Fermilab. The physics program also addresses non-beam physics as nucleon decay searches and the detection and measurement of the electron neutrino flux from a core-collapse supernova within our galaxy. DUNE will consist of a near detector placed at Fermilab close to the production point of the muon neutrino beam of the Long-Baseline Neutrino Facility (LBNF), and four 10 kt fiducial mass LAr TPCs as far detector in the Sanford Underground Research Facility (SURF) at 4300 m.w.e. depth at 1300 km from Fermilab. At least one of the four modules will use the DP technology (arXiv:1807:10327).

DUNE DP module with cathode, PMTs, field cage and anode plane.

As mentioned before, in order to gain experience in building and operating such large-scale LAr detectors, an R&D program is currently underway at the CERN Neutrino Platform. It consists of two prototypes with the specific aim of validating the design, assembly, and installation procedures, the detectors operations, as well as data acquisition, storage, processing, and analysis. The two prototypes employ LAr TPCs as detection technology. One prototype only uses LAr, called ProtoDUNE Single-Phase (SP) and is taking data since 2018, and the other uses argon in both its gaseous and liquid state, thus the name ProtoDUNE -DP. Both detectors have similar sizes. ProtoDUNE -DP, currently under construction, has an active volume of 6x6x6 m³ corresponding to a fiducial mass of 300 t. In ProtoDUNE -DP the charge is extracted, amplified, and detected in gaseous argon above the liquid surface allowing a finer readout pitch, a lower energy threshold, and better pattern reconstruction of the events. In addition, the scintillation light signal is used as trigger for non-beam events, to determine precisely the event time, needed for a full 3D reconstruction of non-beam events, and for cosmic background rejection.

Installation of the PMTs in ProtoDUNE-DP

Resources

Publications

C. Cuesta et al. Photon detection system for ProtoDUNE dual phase, JINST 12 (2017) C12048

B. Aimard et al. A 4 tonne demonstrator for large-scale dual-phase liquid argon time projection chambers, JINST 13 (2018) P11003

D. Belver et al. Cryogenic R5912-20Mod photomultiplier tube characterization for the ProtoDUNE dual phase detector, JINST 13 (2018) T10006

D. Belver et al. A Light Calibration System for the ProtoDUNE-DP Detector, submitted to JINST (2019), arXiv:1902.07127

Presentations

A. Gallego et al. ProtoDUNE Dual Phase, X CPAN DAYS, October 2018, Salamanca, Spain

C. Lastoria et al. The light detection system in ProtoDUNE-DP, XXVIII International Conference on Neutrino Physics and Astrophysics, June 2018, Heidelberg, Germany.

C. Cuesta et al. Light readout, AIDA-2020 Third Annual Meeting, April 2018, Bologna, Italy.

C. Cuesta et al. ProtoDUNE experiments at CERN, IX CPAN DAYS, October 2017, Santander, Spain

C. Cuesta et al. Photon detection system for ProtoDUNE dual phase, LIDINE 2017: Light Detection in Noble Elements, September 2017, SLAC, CA, USA

C. Cuesta et al. Light readout, AIDA-2020 Second Annual Meeting, April 2017, Paris, France

S. Jimenez et al. Light detection in large scale cryogenic liquid detectors, AIDA-2020 First Annual Meeting, June 2016, Hamburg, Germany

Master thesis

Carlos Alonso, Análisis de la detección de luz con el prototipo de argón líquido WA105 del CERN, Universidad Complutense de Madrid, July 2018.

Marcos Allende, Estudio de la colección de luz en el experimento WA105 del CERN, Universidad Complutense de Madrid, July 2017.

Dissemination activities

Primeras partículas en el prototipo de un gran experimento internacional de física de neutrinos, Ciemat web 19/09/2018

A dual-phase DUNE, Symmetry magazine, 07/08/2018

Reunión del proyecto AIDA-2020, infraestructuras europeas avanzadas para detectores y aceleradores, Ciemat web 22/05/2018

Los grandes desafíos del siglo XXI están en manos de estas 12 científicas, Quo magazine, 22/03/2018

Illuminating the hunt for neutrinos, OnTrack AIDA2020 Newsletter, 12/09/2017

Shedding light on ProtoDUNE-DP light detection, Nus to SURF, 16/11/2017

Neutrinos y el Proyecto DUNE, Radio show ‘España vuelta y vuelta’, 27/07/2017

El CIEMAT participa en la conferencia LIDINE 2017, Ciemat web, 10/10/2017

Reunión del proyecto DUNE, experimento internacional sobre neutrinos, Ciemat web, 28/04/2017