Test bench for measurements of the NOvA scintillator properties

From AstroNuWiki
Jump to: navigation, search

NOvA scientific goals require good knowledge about its scintillator basic properties.

The new test bench was constructed at JINR. The main goal of this bench is to measure scintillator properties (for solid and liquid scintillators), namely alpha/beta discrimination and Birk's coefficients for protons and other hadrons (quenching factors). This knowledge will be crucial for recovering the energy of the hadronic part of neutrino interactions with scintillator nuclei.

alpha/beta discrimination was performed on this bench for LAB-based and NOvA scintillators. Resulting curve of recoil protons light yield was obtained. A technical description of the bench, analysis and data taking procedures, obtained results of the measurements and analysis are presented here.

What is a scintillator?

One of the methods of detecting ionizing radiation in experimental physics is to use scintillators. A charged particle passing through a substance deposits its energy within. Part of this energy goes to photon production. For some substances (scintillators) this portion is significant, so that generated light can be detected and measured by photosensors or photodetectors. The spectrum and intensity of the light signal depends on the intensity of the energy release, and the type of passing particle and attributes of the scintillator. Many scintillators, depending on radiation length, are sensitive not only to charged particles, but also to gamma-radiation and neutrons.

These are the main attributes of scintillators: light yield, the spectral composition of radiation, energy resolution, decay time, radiation resistance, radiation length and quenching factors -- Birks Law.

Birks Law

The scintillation response, S, of organic crystals depends on the nature and energy, E, of the incident ionizing particle, of residual range r.

The specific fluorescence, dS/dr, is not in general proportional to the specific energy loss dE/dr. By considering the quenching effect of the molecules damaged by the particle by the "excitons" produced by it, it is possible to show that:

dS/dr = (A*dE/dr)/(1 + kB*dE/dr)

Where A and kB are constants, which have been evaluated for scintillator from observations of S and E, and the range-energy data. The method used for evaluating the relative response is applicable to ionizing particles of any nature or energy, and also to the different organic scintillation crystals or liquids.

The quenching factor is a very important attribute of any scintillator. If we do not know the quenching factor, we cannot say what kind of particle passed through the scintillator and what primary energy it had.

Methodical aspect of the measurement and hardware setup

The bench consists of a few components. First is the Black Box with light sensitive equipment, sealing rubber, all necessary connectors and light shield. It is used to isolate the inner structure of the box from outer light. It is possible to place scintillator samples and all necessary radioactive sources (for α/β ratio measurements) inside the Black Box.

The second part of the bench is a PMT (Photomultiplier tube, 3’ Hamamatsu R12772) with divider and cuvette (it’s inner sizes are 50 × 50 × 50 mm and it is made from optical glass). The PMT and cuvette are put inside the box in a vertical orientation.

The third is all support electronics like ADC (Analog-to-digital converter, with high sampling – it gives us required precision level), High-Voltage source and laptop with required software, self-stabilized LED, neutron source (Pu-Be) with the NaI crystal, and also the scintillators and different α and β radioactive sources.

All the technical staff were located at Radio-Chemical Laboratory at JINR during this measurement.



Firstly we have to measure the alpha/beta ratio for a scintillator. This procedure is necessary for the calibration of our ADC and testing of our method. After that it is possible to measure the quenching factors for different hadrons.

In the case of NOvA we want to measure Birk's coefficients for protons with energy in the range of 1-14 MeV. But it is not so easy to deliver protons with this energy inside the scintillator. We have to use a neutron source and measure the energy of recoil protons. Recoil protons are produced in the liquid scintillator by neutron-proton scattering events using a neutron generator ING-27 with a monochromatic energy distribution or a Pu-Be source with continuous spectrum. Hence, the proton light output function can be determined from the position of the recoil proton edge in the pulse-height spectra produced by mono-energetic neutrons. In the case of a continuous spectrum it is possible to use the time-of-flight method for speed/energy determination.




Data taking

We have 3 active channels -- CH1 is a signal from NaI, CH3 is monitoring LED signal and CH4 is a signal from liquid scintillator.

Logic as follows -- we are looking for coincidence between NaI and Liquid Scintillator (CH1 and CH4) or LED and Liquid Scintillator (CH3 and CH4).


Our ADC is DRS4 manual.

Main parameters of the data taking

  • Flight path -- 1.6 m in our case.
  • 4.43 MeV gamma like the START point.
  • We are interested in time difference between START and STOP (signal from liquid scintillator) points higher than 46 ns and lower than 120 ns (ToF for neutrons). It corresponds to neutrons energy between 1-6 MeV. These energy limits and START point are connected with reaction channel (alpha + 9Be -> 13C* -> 12C* + n -> 4.43 MeV gamma -- first excitation state of 12C). E_{excitation}(13C*) = 10.6 MeV.
  • Connection between time of flight and energy is: T(ns) = (72.3[ns] * L[m])/sqrt(E[MeV])
  • Result of the energy calibration -- we are using several gamma sources (60Co, 137Cs, 228Th). We are looking for Compton edge (half-height) by fitting the spectra with Erf function. Compton edge energies as follow -- 478 keV for 137Cs, 1041 keV for 60Co, 405 keV and 2381 keV for 228Th.

Energy calibration

NOvA Th fit.jpeg


Cerenkov component

Gamma lines with energy higher than 180 keV - Compton edge (it depends on refractive index of liquid scintillator) have Cerenkov component in light output. For better measurement precision one has to take it into account. Main idea as follows: measure the spectra of gamma sources with lines higher and lower than 180 keV. Higher points should have their own incline like lower ones but lower ones do not have Cerenkov component so it is possible to "paste together" both lines and find this unknown part.

228Th is well suitable for this measurement. It has 3 main lines — 239 keV, 583 keV, 2614 keV. It's spectra with error function fit like an example (PMT voltage is 1600 V for first and second lines and 1100 V for the third line):

228Th first line 1600V.jpeg 228Th second line 1600V.jpeg 228Th third line 1100V.jpeg

57Co, 137Cs, 60Co, 22Na will be used too.

Analysis procedure

Our main parameter is time difference between CH1(NaI) and CH4(LS). It is possible to select neutrons with defined energy using this time distribution.

After that one can extract light output on liquid scintillator (charge) on condition that NaI signal corresponds to 4.43 MeV gamma and build Recoil proton energy vs Scintillator response curve. We are looking for half-height of the edge in light yield of liquid scintillator. This point corresponds to single interaction between incident neutron and proton inside the scintillator when neutron transmits all of his energy.

For more details about analysis procedure please see section Additional information.

Birks coefficient evaluation with average dE/dx

General Birks formula is dL/dx = (S×dE/dx)/(1 + kB*dE/dx) , where L is a light yield, E is a proton energy, S is a scintillator efficiency (in our case S=1 because we make our calculations in terms of electron–equivalent energy), dE/dx is a energy loss, kB is a Birks coefficient.

We made several assumptions:

• dE/dx ≈ E/∆x; where ∆x is s a stopping range – we got it from PSTAR tables.

• dL/dx × ∆x ≈ L.

With this assumptions general Birks formula turns into kB = (E - L)/(L×E/∆x).


We performed α/β ratio measurements on the bench for LAB-based and NOvA liquid scintillators using the special method. Also we performed a measurement of the light yield (for Birks coefficient extraction) for recoil protons using a Pu-Be neutron source and compared it with the German article results. We assumed special way of Birks coefficient evaluation using average dE dx and obtained values for different LAB-based liquid scintillators are in a good agreement. Results are as follows:

PMT gain.jpeg

NOvA alpha beta.jpeg


For the Gd+Cm source quenching factors for LAB-based scintillator (\alpha) are equal to 22,2 for Gd (energy 3,183 MeV) and to 16,5 for Cm (energy 5,795 MeV). For the NOvA scintillator they are equal to 23,58 and 17,6.


NOvA LAB all data.jpeg


kB evaluation results

We tested out evaluation method on German LAB points – very well matching between results.

Method gives:

0.0092±0.0001 and 0.0088±0.0001 cm/MeV for our LAB scintillator

0.0132±0.0002 and 0.0126±0.0002 cm/MeV for NOvA scintillator

KB gLAB.jpeg

KB Dubna LAB NOvA.jpeg

Additional information

!It is possible to find current status and all technical details here!.

AlexanderAntoshkin (talk) 02:53, 12 October 2018 (MSK)