Numerical characterization of the ARAPUCA: a new approach for LAr scintillation light detection
NNumerical characterization of the ARAPUCA: a newapproach for LAr scintillation light detection
F Marinho , L Paulucci , A. A. Machado and E Segreto Universidade Federal de S˜ao Carlos, Rodovia Anhanguera, km 174, 13604-900, Araras, SP,Brazil Universidade Federal do ABC, Av. dos Estados, 5001, 09210-170, Santo Andr´e, SP, Brazil Instituto de Fsica Gleb Wataghin, Universidade Estadual de Campinas, Rua S´ergio Buarquede Holanda, 777, 13083-859, Campinas, SP, BrazilE-mail: [email protected]
Abstract.
The ARAPUCA concept has been proposed as a simple and neat solution forincreasing the effective collection area of SiPMs through the shifting and trapping of scintillationlight in noble liquids, thus with great potential for improving timing and calorimetry resolutionin neutrino and dark matter search experiments using time projection chambers. It is expectedto achieve a single photon detection efficiency larger than 1%. The initial design consists ofa box made of highly reflective internal surface material and with an acceptance window forphotons composed of two shifters and a dichroic filter. The first shifter converts liquid argonscintillation VUV light to a photon of wavelength smaller than the dichroic cutoff, so the surfaceis highly transparent to it. When passing through the dichroic filter, it reaches the second shifterwhich allows the photon to be shifted to the visible region and be detected by the SiPM nestedinside it. When it enters the box, the photon will likely reflect a few times, including on thedichroic filter surface, before being detected. We present a full numerical description of thedevice using a Monte Carlo framework, including characterization of the acceptance window,models of reflection of different materials, and sensor quantum efficiency, that can now be usedto further improve the detection efficiency by comparing different geometries, positions of SiPMand materials. Estimates of simulated efficiencies, number of reflections and acquisition timeare presented and compared to analytical calculations. Those are very promising results, givinga total efficiency for the detection of scintillation light in liquid argon of 1.7 ±
1. Introduction
Investigations at the frontier of particle physics involve measurements of low frequency events,such as interacting neutrinos, possible proton decay, and detection of dark matter. Newtechnologies are in high demand to improve the chance of detecting such events. Scintillationlight detectors are important component systems for the next generation of LArTPCsexperiments which can provide additional information regarding calorimetry and timing thusenhancing event reconstruction performance. To explore the full potential of these systems onemust consider the best options for detector technologies under development and their impact onquality of physics measurements.One recent suggestion for photon detection at the VUV range is the use of ARAPUCA devices[1] for trapping photons and increasing the detection effective area of silicon photomultipliers a r X i v : . [ phy s i c s . i n s - d e t ] A p r igure 1. ARAPUCA device geometry (left): acceptance window (yellow), SiPM sensor (blue)and reflective cavity (gray). Acceptance window concept (right): PTP (dark magenta), dichroicfilter (gray), TPB (blue).(SiPMs). The concept consists of a box cavity with highly reflective internal walls inside whichphotons can reflect back and forth until reaching its photomultipliers. This box is equipped withan acceptance window formed by two layers of wavelength shifters and a dichroic filter whichallows photons to pass efficiently inside the cavity. It also presents high internal reflectance tothe accepted photons, therefore acting as a trap. For use in a liquid argon (LAr) environment,important for dark matter and neutrino searches, those layers were optimized as follows. Thescintillation light emitted by argon atoms, in the vacuum ultra-violet spectral region (VUV) witha wavelength around 127 nm, are converted to the detector’s range (visible). The first shifter inthe acceptance window, made of p-terphenyl (PTP) [2], converts the liquid argon scintillationlight to a photon of wavelength that peaks around 350 nm, smaller than the dichroic cutoff( ∼
400 nm). When passing through the dichroic filter, it reaches the second shifter, madeof Tetraphenyl-butadiene (TPB) which shifts to the visible region (peaking at ∼
430 nm) [3],ensuring the dichroic filter surface to now become highly reflective. The photon can reflect afew times on the internal surface before being detected by the SiPM.We present a Monte Carlo based approach to characterize the ARAPUCA light scintillationdetector device as a tool for research and development of this technology.
2. Numerical simulation
The computational description of the ARAPUCA was developed in order to include its maincomponents and their correspondent optical functions correctly. For that, the Geant4 simulationframework [4, 5] provides a suitable set of functionalities that allows the propagation of opticalphotons through various materials and interfaces considering different models for light reflection,refraction, absorption, emission, etc. Figure 1 shows a complete 3D model of a detectorindicating its acceptance window, reflective cavity and a silicon photomultiplier (SiPM). Thegreen line indicates a photon optical path and the yellow dots its reflection points on the cavityand filter internal reflective walls.The acceptance window consists of a dichroic filter with one layer of ∼ . / cm PTPdeposited on top and one layer of ∼ × − mg / cm TPB deposited on the bottom of thefilter as illustrated in figure 1. The widths of the PTP and TPB are chosen to ensure optimalwavelength shifting [2]. The incoming VUV photons are absorbed by the PTP material whichin turn emits photons isotropically in the 330-400 nm range. As a consequence, half of theseemitted photons pass through the dichroic material, reach the TPB and are absorbed. TheTPB emits photons in the 400-560 nm range, which therefore, are trapped within the cavity. igure 2.
Dichroic filter (left): Transmission (solid black) and reflection (dashed red) [6].Window emitted photon spectra (right).
Figure 3.
SiPM photon detection efficiency as function of incoming photon energy.Figure 2 shows the average transmittance and reflectance of the dichroic filter at 45 o incidenceangle (left) and the spectra for the accepted and reflected light (right) both as a function of thephotons’ energy.For the internal walls there are two main possibilities for modelling the reflection typesoccurring due to the materials envisioned for the device. The box can be made of Teflon TM PTFE which offers high reflection (
R > TM
3M which is a specular reflective film ( R ∼ . Results A detailed description of the photons trajectories within the different materials of the devicegeometry allows to evaluate all the relevant parameters for its characterization and overalloperation. The following sections describe results obtained with the complete simulation ofthe ARAPUCA or its components. Comparison with analytic calculations or measurements isprovided when those are accessible.
The total efficiency of the detector can be defined as: (cid:15) total = N detected N total , (1)where N detected is the number of observed photons from the SiPM signal and N total is the numberof incoming photons which arrive at the acceptance window from outside.However, it is important to characterize each component in terms of its own efficiency. Hence,one can factorize the total efficiency as: (cid:15) total ≈ (cid:15) acceptance × (cid:15) collection × (cid:15) SiP M , (2)where (cid:15) acceptance is the window efficiency on converting incoming VUV photons to visible lightinside the cavity and (cid:15) collection gives the fraction of accepted photons which reach the SiPM (i.e.collected and not absorbed in the cavity walls). The SiPM efficiency is given by: (cid:15) SiP M = 1 N collected (cid:90) E E (cid:15) SiP M ( E ) dN collected dE dE, (3)where N collected is the number of photons which arrive at the surface of the SiPM and (cid:15) SiP M ( E )is the sensor photon detection efficiency as a function of the photon energy.Note the factorization in equation 2 is possible because (cid:15) acceptance is constant around the VUVincoming photons energy. The (cid:15) collection is also assumed constant as it should only depend on R ,the active coverage f and geometry of the reflective cavity. In both aforementioned materials R is above 95% and approximately constant across the whole wavelength range inside the cavity.The other two parameters are independent.Table 1 shows the value obtained for these efficiencies calculated for a device with 36 mm ×
25 mm × R = 95%. An experimental test realized with a prototypefollowing the same configuration has provided a value (cid:15) exptotal ∼ .
8% which is in reasonableagreement with the Monte Carlo estimate.
Table 1.
Monte Carlo estimated ARAPUCA efficiencies. (cid:15) acceptance (cid:15) collection (cid:15)
SiP M (cid:15) total . ± .
0% 19 . ± .
4% 25 . ± .
3% 1 . ± . (cid:15) collection estimates above analytic calculations given by [1]: (cid:15) analyticcollection = f − R (1 − f ) . (4)Note that it is not possible to obtain efficiencies higher than the analytic estimates with theproposed non-focusing box design from figure 1. The cylindrical cavity with a disk shaped igure 4. Focusing geometries.acceptance window provides (cid:15) collection = 14 . ± .
4% while the analytic estimate for an equivalentsetup (same R and f ) gives 13 . (cid:15) collection = 16 . ± .
4% and the analytic estimate is 13 . The evaluation of the time characteristics of the device is another important aspect that needsto be understood if considering it to be used for timing purposes in a LArTPC experiment.There are three major delay sources that determine the acquisition time of the device: t acquisition = t P T P + t T P B + t collection , (5)where t P T P and t T P B are the emission decay time of the wavelength shifters and t collection isthe time it takes for the photon accepted in the cavity to arrive on the SiPM after successivereflections.All these times are random quantities in a event-by-event basis. Figure 5 on the left showsan example of the t collection distribution obtained from the Monte Carlo simulation. As expectedthe distribution has a sharp decay as the overall distance traveled is very small inside the cavity.One can evaluate a reasonable distribution for t aquisition assuming a mixture of exponentialdistributions for each of the emission times and their respective measured values [9]. Figure 5on the right shows the obtained distribution with a most probable acquisition time of ∼ t forthe interacting primary particle in a physics event or to help as trigger device to distinguishbeam events from other sources in a surface installed experiment.
4. Conclusion
An ARAPUCA full simulation was implemented and is currently employed as a R&D designtool for this detector technology. The total efficiency obtained from a prototype simulation andits experimental value were in good agreement. Comparison between the analytic estimates andsimulation for the collection efficiency, number of internal reflections and timing as function ofthe f and R parameters follow the same behavior. It is a flexible software application whichallows not only parameter optimization but geometrical and physical modifications with ease. igure 5. Expected distribution for t collection (left) and t acquisition (right).In particular one can evaluate the performance of each component according to its features suchas wavelength shifters types, width, spectra emission dependence on temperature, dichroic filtertransmission and reflection as function of the angle, etc. One can also choose the main surfacemodels according to the type of cavity walls and adjust the response of the SiPM used accordingto the operating parameters. In addition a set of analysis algorithms were also developed toevaluate all the relevant figures of merit of interest. Acknowledgments
The authors would like to thank Funda¸c˜ao de Amparo `a Pesquisa do Estado de S˜ao Paulo(FAPESP) for financial support under grants n o o References [1] Machado A A and Segreto E 2016
JINST
ARAPUCA a new device for liquid argon scintillation light detection C02004[2] DeVol T A, Wehe D K, Knol G F 1993
Nucl. Instr. Meth. Phys. Res.
A Evaluation of p-terphenyl and 2,2”dimethyl-p-terphenyl as wavelength shifters for barium fluoride et al
JINST
VUV-Vis optical characterization of Tetraphenyl-butadiene films on glass andspecular reflector substrates from room to liquid Argon temperature P09006[4] Agostinelli S et al
Nucl. Instr. Meth. Phys. Res.
A Geant4 - a simulation toolkit et al
Nucl. Instr. Meth. Phys. Res.
A Recent developments in GEANT4 TM Enhanced Specular Reflector Film (ESR), 3M.com/displayfilms[8] Sensl C-Series SiPM, sensl.com/products/c-series[9] Segreto E 2015
Phys. Rev.
C Evidence of delayed light emission of tetraphenyl-butadiene excited by liquid-argon scintillation light91