The POCAM as self-calibrating light source for the IceCube Upgrade
TThe POCAM as self-calibrating light source for theIceCube Upgrade
The IceCube Collaboration ∗ http://icecube.wisc.edu/collaboration/authors/icrc19_icecubeE-mail: [email protected], [email protected],[email protected] The planned IceCube Upgrade, consisting of seven new instrumentation strings, will be installedat the South Pole within 2022/2023. The focus of this upgrade is calibration, reduction ofsystematic uncertainties and atmospheric neutrino physics. Within this scope, the "PrecisionOptical Calibration Module" (POCAM) will be installed at a number of positions on thesenew strings, to act as a calibration light source. The POCAM is an in-situ self-calibrating,isotropic, nanosecond light source that emits flashes of adjustable intensity and pulse duration.The isotropy is achieved using a teflon integrating sphere which further allows the calibrationof the total number of emitted photons per pulse, using the integrated sensors. Prototypes havebeen deployed and operated within the GVD telescope in Lake Baikal and within the STRAWexperiment in the Pacific Ocean. We present POCAM results and experiences from the GVDand STRAW installations as well as first IceCube sensitivity studies and the following designprospects for this next-generation POCAM iteration.
Corresponding authors:
C. Fruck † , F. Henningsen , , C. Spannfellner Physik-Department, Technische Universität München, D-85748 Garching, Germany Max-Planck-Insitut für Physik, D-80805 Munich, Germany36th International Cosmic Ray Conference -ICRC2019-July 24th - August 1st, 2019Madison, WI, U.S.A. ∗ For collaboration list, see PoS(ICRC2019) 1177. † Speaker. c (cid:13) Copyright owned by the author(s) under the terms of the Creative CommonsAttribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0). http://pos.sissa.it/ a r X i v : . [ a s t r o - ph . I M ] A ug he POCAM as self-calibrating light source for the IceCube Upgrade C. Fruck
1. The Precision Optical Calibration Module
The
Precision Optical Calibration Module (POCAM) [1][2][3] is a self-calibrating light-emitter device for the in-situ calibration of large-volume neutrino detectors. Its primary goals arethe reduction and better understanding of systematic experimental uncertainties within the detector– most notably the properties of the optical detection medium as well as efficiency and angularacceptance of the photosensor instrumentation. The limited understanding of the South Pole ice inIceCube results in a systematic uncertainty of the optical detector properties of around 10% [4].It is one of the leading systematics that affects a number of analyses [2]. By providing in-situcalibrated light pulses, the POCAM aims to reduce these uncertainties down to a few percent.The baseline concept of the device is the isotropic multi-wavelength emission of nanosecondlight pulses [5] with the design shown in fig. 1. Here, the pressure housing from titanium withtwo glass hemispheres is shown together with internal components. The latter consist of a customsemi-transparent integrating sphere which makes the light pulse isotropic and the circuit boards foroperation. The LEDs are predominantly driven by Kapustinsky drivers [6] and reach a dynamicrange of 10 − photons per pulse and pulse FWHMs of 1 − −
600 nm are used, with best intensity and timingperformance achieved in the range of 400 −
470 nm. The integrated photosensors visible in fig. 1are a Silicon-Photomultiplier (SiPM) and a photodiode. These sensors monitor the pulse time onsetand intensity characteristics in-situ with an anticipated precision of a few percent.Figure 1: POCAM hemisphere assembly (left) and complete module assembly (right). The differ-ent components and their functions are explained in the text.As described in [1] and [2], the POCAM baseline design is composed of four sub-systems:pressure housing, analog and digital electronics and the light diffusion and emission components.The analog and digital boards are located in both hemispheres to allow for their independent opera-tion. In the following we will describe the baseline design operating in the STRAW experiment [1].The device goals and design improvements for the IceCube Upgrade [7] will then be discussed in2 he POCAM as self-calibrating light source for the IceCube Upgrade
C. Fruck sections 3 and 4. The upgrade itself will consist of seven new and more densely instrumentedstrings, deployed within a core segment of IceCube. They will include new photosensor instrumentdesigns [8, 9] to increase sensitivity down to lower energies with more photo-active area as well asto deploy more instrumentation for better calibration, like the POCAM.The pressure housing consists of a 10 mm thick and 25 cm long titanium cylinder and twotitanium flanges to which the optical borosilicate glass is attached to using epoxy resin. The glasshas a thickness of 7 mm and a diameter of 11 . − ◦ C and are stress tested not only for temperature and pressure butalso shocks and vibrations occurring during transport and deployment.The analog boards in the POCAM electronics are responsible for the driving of light pulses andthe readout of the integrated in-situ sensors. As such, it hosts the LED drivers, the photosensors,their frontend readout chain and related peripheral components. Additionally, the temperature ofthe LEDs, their drivers and the sensors is monitored, as this is part of the reference calibrationdescribed in section 4. The main circuit used in previous POCAM versions (section 2) is theKapustinsky driver [6]. It makes use of a controlled switching of two bipolar transistors whichdischarges a capacitor through an LED with a parallel inductance. The latter will cause a bipolarswing of the pulse and thus effectively cuts it short. With this technique and selected LED-typesof 365, 405, 465 and 605 nm, we achieved FWHMs of 4 − − photons per pulse which is linear with supply voltage [1, 10]. However, the performance of thedriver is highly dependent on the used type of LED and requires a proper selection process. Thephotosensor frontend includes a charge-sensitive amplifier and an additional amplification stage forshaping of the output pulse. The output is a voltage pulse with few hundred nanosecond rising timeand microseconds of decay time. This output is equal for both the SiPM and the photodiode butdiffers only in the sensitivity of the charge amplifier. Eventually the output signals are fed to thedigital board.Communication, data aquisition (DAQ) and control of the instrument are handled on the dig-ital board which is also located in each of the hemispheres. Each digital board hosts primarily anFPGA, a micro-controller (MC) and a two-channel ADC (analog-to-digital converter). The ADCsamples the output of the sensor frontend with a rate of 10 MHz and writes the data to storage,which in this case was an on-board SD-card. Additionally it provides orientation sensors based onaccelero- and magnetometers as well as environmental monitoring with a number of temperature,pressure and humidity sensors overseeing the overall system integrity of the instrument. Commu-nication with the instrument and data transfer was realised using a serial RS-485 interface and itsoperation is controlled via ASCII commands. The digital electronics and interfaces are by designmodular and can be adapted to the respective application.Finally, the primary and central part of the POCAM light emission is the diffuser. This semi-3 he POCAM as self-calibrating light source for the IceCube Upgrade C. Fruck transparent integrating sphere from PTFE (or teflon) makes use of its Lambertian reflection to makethe light pulses isotropic. It furthermore uses a dedicated two-part geometry of a plug and spherethat removes any disturbances on the emitting surface which would cause luminosity variations.The plug is further used to couple the LED array into the integrating sphere and uses a thin layer ofteflon to diffuse the initial LED emission profile. This configuration can be seen in fig. 1. Thanks tothe constant behaviour of teflon across a broad range of visible wavelengths, this allows to integratevisible light pulses from effectively 200 nm upwards and effectively any LED opening angle. Thisemission behaviour has been extensively studied and the resulting zenithal profile is shown in fig. 5afor a diffuser made from different types of teflon. As can be seen there and also will be discussedin section 4, switching from regular to specifically-produced optical teflon results in a significantlyimproved emission profile.
2. Deployments in GVD and STRAW
A first prototype of the instrument has been deployed in 2017 within the GVD neutrino tele-scope in Lake Baikal [2]. The device was mounted to one of the detection strings and lowered toits final depth of around 1 ,
000 m. During its one year of operation, several runs with POCAM havebeen carried out and allowed us to gain experience with the in-situ device operation. Furthermore,the shared data of the flasher runs – provided by the GVD collaboration – allowed an estimateof the attenuation length in Lake Baikal to ( ± ) m at 470 nm. This agrees well with previousmeasurements [11, 12] and eventually provided the first proof of concept for the POCAM.The second deployment of an improved POCAM version was the STRAW experiment in theNorthern Pacific Ocean [1]. This pathfinder experiment deployed two 150 m mooring lines to adepth of 2 .
3. IceCube Upgrade Calibration Goals
The POCAM will be a multi-purpose calibration device as it can tackle both, systematics re-lated to the South Pole ice as well as the surrounding sensor instrumentation. In total more than20 of our devices will be distributed on the seven new Upgrade strings within the IceCube vol-ume. These densely spaced and instrumented strings are in close proximity to previous IceCubestrings and will host POCAMs spread between depths of 1450 − he POCAM as self-calibrating light source for the IceCube Upgrade C. Fruck volume as well as the vicinity volumes. As the POCAM is a multi-wavelength device, all sys-tematic effects in the following, can be studied over a range of wavelengths from UV to visible.Figure 2: DOM angular acceptance curves for a strongly scatter-ing hole ice column and DOMs randomly placed inside. Figuretaken from [2].One uncertainty in the under-standing of the Antarctic ice ofIceCube is the so called holeice . It describes a stronglyscattering column of ice thatforms during refreezing of thedrill hole. This hole iceis a local ice formation thathas a predominant effect onthe angular acceptance pro-files of the sensor instrumen-tation and is so far modelledempirically [2, 13]. Sev-eral POCAMs in the denselyinstrumented upgrade volumewill be illuminating the sameDOMs with effectively planewaves of light from differentzenith angles. Thus, a sam-pling of the hole ice curve can be obtained on a per-DOM level, as is shown in fig. 2.Equally important for IceCube analyses are the bulk ice properties of the ice and the pho-todetection efficiencies of the DOMs. Both of these affect the global light propagation within theIceCube detector volume and its detection at the DOMs and thus are parameters of extensive stud-ies [4, 14]. The detector volume shows average scattering and absorption lengths of 20 m and 120 mrespectively at 400 nm. As will also be discussed in section 4, the POCAM will act as an absolutelycalibrated light source with well-known emission characteristics. As such, the previously usedLED calibration studies can be refined by using a well-defined light source in simulation studiesthat can be used to break the degeneracy of unknown light emission characteristics and ice prop-erties. As far as the POCAM development is concerned, the dynamic range is the driving factorto illuminate a large enough volume with sufficient photon statistics. As shown in fig. 3a, this isachieved with flashes ranging from 10 − photons per pulse. However, fig. 3b shows that thestrong ice scattering will diffuse out the timing of sharp light pulses, the time profile of the pulsescan thus widen in time up to a few tens of nanoseconds in order to illuminate larger volumes.A remaining open question is the ice anisotropy [14, 15]. Its effect manifests itself as an ob-served anisotropic behaviour of the in-ice photon arrival time and flux which seemingly correlateswith the direction of the ice flow. Using an isotropic light source and the surrounding DOMs, thePOCAM will be able to study its influence by comparing azimuthal intensities, however, it requiresprecise knowledge of the orientation with respect to the hole ice column and the main cable.5 he POCAM as self-calibrating light source for the IceCube Upgrade C. Fruck Distance DOM-to-POCAM [m] T o t a l ph o t o nh i t s / D O M IceCube work in progress
DOM dynamic range / 100nsDOM P O C A M d e p t h [ m ] (a) Average number of DOM hits by N = photon simulations at 400 nm andat different depths within the Upgrade volume. Time from flash [ns] − D O M ph o t o nh i t s /50 n s (b) Photon arrival time histogram for DOMs atdistances of 74 m and 149 m. Figure 3: Driving effects of ice absorption and scattering on the POCAM light pulse properties.
4. Device Improvements for the IceCube Upgrade
While previous POCAM iterations have been succesfully deployed in environmental condi-tions similar to the South Pole, the instrument leaves room for precision optimizations necessaryfor an IceCube application. Components related to the light emission and in-situ calibration arehence subject to further improvements, this includes isotropy, dynamic range as well as in-situsensor characterization and readout.In order to illuminate large-enough fractions of the IceCube detector, the high-end of thedynamic range needs to be increased to 10 − photons per pulse. Therefore, new dedicatedLED drivers are under investigation. They will operate alongside the fast Kapustinsky driversand thus, are allowed to have pulse profiles up to a few tens of nanoseconds to provide necessaryintensities. Promising systems – based on the Gen1 driver of IceCube [16] – using different gatedrivers and GaN-FETs (Gallium-Nitride field effect transistors) are being developed. In addition,the default Kapustinsky driver will be accompanied by an additional one, tuned to sharp timeprofiles of the order of 1 − . . . . . . . P h o t o n s / pu l s e Kapustinsky - fastKapustinsky - defaultGen1 gate driver ∼
10% systematics (a) Intensity of POCAM LED drivers using thedefault 405 nm LED at different supply voltages.
Time [ns] . . . . . . N o r m a li ze d c o un t s Kapustinsky - fastKapustinsky - defaultGen1 gate driver (b) POCAM light pulse time profiles for differentLED drivers and using the default 405 nm LED.
Figure 4: Measurements of light intensity and pulse shape for the default POCAM LED drivers.6 he POCAM as self-calibrating light source for the IceCube Upgrade
C. Fruck In addition, the LED wavelengths will be adjusted for the IceCube Upgrade, with a defaultLED of 405 nm installed in every device. A secondary wavelength colour is selected from a spectralrange of 320 nm to 550 nm and will depend on the final location of the POCAM in the ice, as thespectral interest in optical properties is depth-dependent. As mentioned in section 2, the rightchoice of the LEDs greatly affects the performance of the drivers. In the scope of finding themost suitable LED-type, automated characterization measurements on a large batch of LEDs willbe performed to investigate their characteristics in connection with the driver. Also the readoutelectronics of the in-situ sensors are going to be revised with a new analog frontend. This willinclude new charge- and pre-amplifiers with significantly reduced bias current and thus highersignal-to-noise ratio for both SiPM and photodiode. In addition, the fast SiPM signal is split, asmall fraction will trigger a giga-hertz comparator for optical onset pulse timing and the remainderis charge-integrated. Other components of the DAQ will stay as described in section 1 with anadditional board that will handle timing and communication with the IceCube detector and itsprotocols. As mentioned already in section 1, diffusers used in GVD and STRAW, were madeout of regular teflon. As can be seen, in fig. 5a this achieves promising isotropic behaviour inzenith angle. In the scope of optimizations, a specifically produced teflon for optical applicationswas proposed. This optical teflon further increases the diffusion of the initial LED flash with onlyminor loss of intensity with respect to the regular one. However, fig. 5a shows that it notablyincreases the isotropy and as such is the material of choice for an IceCube application.In order for the POCAM to have the capability to calibrate itself, it is necessary to conductprecise characterization measurements of the light pulsers. These act as a reference during lateroperation. The main goal of this procedure is an automated relative characterization of the pulserswith respect to driver bias voltage and temperature. The latter poses more complicated, as it has tobe carried out over a range of −
55 to + ◦ C. Apart from an intensity and time profile calibration,it also becomes necessary to quantify the LED emission spectrum. As shown in fig. 5b, it hasrecently been observed that pulsed LED drivers vary in mean emission wavelength with appliedbias voltage. This wavelength shift is important to account for different wavelength-dependentquantities within the IceCube detector, in order to achieve better calibration. . . . . I n t e n s i t y [ a . u .] SimulationSimulationIdeal emissionOptical PTFE + flangeRegular PTFE + flange (a) Zenith emission profile of a POCAM hemi-sphere using regular and optical teflon diffusers. . . . . . . . . . . . . λ o b s e r v e d − λ n o m i n a l [ n m ] − . . Kapustinsky - default (405nm)
10 15 20 25 30 − − Kapustinsky - default (470nm) (b) Mean emission wavelength versus voltage fora Kapustinsky driver.
Figure 5: Measurements of integrating sphere emission profile and LED spectrum.Together with the angular emission profile, four characterization measurements need to be con-ducted for every module. In order to streamline this process, two fully automated characterization7 he POCAM as self-calibrating light source for the IceCube Upgrade
C. Fruck stations are being developed. The first setup consists of a freezer, capable of reaching temperaturesdown to − ◦ C, the sensor instrumentation for the different characterization measurements anda reference light source for stability checks. The instrument electronics are mounted inside thefreezer and the LED drivers are connected to the characterization sensors inside the dark box viaoptical fibers. The sensors include a spectrometer for the wavelength measurement, an avalanchephotodiode to sample the time profile of the pulse via time-correlated single photon counting, aswell as a photomultiplier for low and a photodiode for high intensities. Secondly, the angular emis-sion scan is carried out with two rotational stages to which the POCAM hemisphere is mounted to.A photodiode measures the emitted intensity for different azimuth and zenith angles. All measure-ment points shown in figs. 4 and 5 are carried out with prototypes of these setups.To conclude, the POCAM as an isotropic in-situ calibrated device will help to improve theunderstanding of the IceCube detector by providing means to reduce systematic uncertainties ofthe optical medium and the sensor instrumentation. The streamlined process of production andcalibration will further ensure a precise knowledge of individual instrument characteristics.
Acknowledgements
We acknowledge the support by SFB1258 and the Cluster of Excellence. Additionally wethank the Baikal and STRAW collaborations for the deployments of the POCAM and sharing ofrelated detector data.
References [1]
STRAW
Collaboration, M. Boehmer et al.,
JINST (2019) P02013.[2] IceCube Gen2
Collaboration, E. Resconi et al.,
PoS(ICRC2017)934 (2018).[3]
IceCube-Gen2
Collaboration, K. Krings,
PoS(ICRC2015)1133 (2016).[4]
IceCube
Collaboration, M. G. Aartsen et al.,
Nucl. Instrum. Meth.
A711 (2013) 73–89.[5]
IceCube Gen2
Collaboration, M. Jurkovic et al.,
EPJ Web of Conferences (2016) 06001.[6] J. Kapustinsky et al.,
Nucl. Instr. Meth. Phys. Res. Sec.A (1985) 612 – 613.[7]
IceCube
Collaboration,
PoS(ICRC2019)1031 (these proceedings).[8]
IceCube
Collaboration,
PoS(ICRC2019)855 (these proceedings).[9]
IceCube
Collaboration,
PoS(ICRC2019)966 (these proceedings).[10] B. K. Lubsandorzhiev and Y. E. Vyatchin,
JINST (2006) T06001.[11] BAIKAL
Collaboration, I. A. Belolaptikov et al.,
Astropart. Phys. (1997) 263–282.[12] NEMO
Collaboration, V. Balkanov et al.,
Nucl. Instrum. Meth.
A298 (2003) 231–239.[13]
IceCube
Collaboration, M. Rongen,
EPJ Web Conf. (2016) 06011.[14]
IceCube
Collaboration, M. G. Aartsen et al., arXiv:1309.7010 .[15]
IceCube
Collaboration,
PoS(ICRC2019)854 (these proceedings).[16]
IceCube
Collaboration, M. G. Aartsen et al.,
JINST (2017) P03012.(2017) P03012.