Phenomenological And Experimental Studies On Problems In Neutrino Detections

After the discovery of neutrino oscillation,most of the oscillation parameters have been measured by solar,atmosphere,accelerator and reactor neutrino experiments.However,the neutrino mass hierarchy（MH）is still unknown to us.In 2012,Daya Bay experiment obtained a nonzero value for the neutrino mixing angleθ₁₃with a significance of 5.2σ,which enables the possibility to determine neutrino mass hierarchy.As one of the future large neutrino experiments,Jiangmen Underground Neutrino Observatory（JUNO）aims to determine neutrino MH at 3σsignificance within six years of running.This requires JUNO to reach an unprecedented energy resolution of 3%/（?）.To achieve this goal,20,000 large area PMTs with high quantum efficiency will be used for photon detection.Thus the simulation and performance evaluation of these PMTs become a key problem in the experiment.In this thesis,we first explore on the possibility to determine neutrino MH with supernova neutrino.Two cases of neutronization burst and black hole formation during supernova explosion are considered.Then,we construct a new optical model for PMTs based on optical principles.This model gives an accurate description of PMT response in JUNO detector simulation.Moreover,a feasible method to measure the absolute detection efficiency of PMTs were still unknown.In this thesis,we propose a beam-splitting method,with which a Hamamatsu 20inch PMT is successfully calibrated.Supernova neutrino is one of goals of JUNO while how to discriminate neutrino MH with supernova neutrino is a hot topic.With the short-time characteristics of supernova neutronization burst,we can try to discriminate MH with the flight time difference between different neutrino mass eigenstates.For different MH,the measured spectra in the detector will have different time structures.Moreover,since there will be a sudden signal termination of the signal if a black hole is formed during supernova,we can also use this signal termination to discriminate the MH of neutrinos.This method require a long distance between earth and supernova or even larger detectors,so it will not be applicable in JUNO.The difficulty of PMT simulation lies in the construction of an accurate and reliable response model.In fact,photons undergo reflection,refraction and absorption in an air（water）-glass-photocathode-vacuum system.The coefficients between them can be obtained using Snell’s law and Fresnel’s equations.All the PMTs in JUNO work under water.Therefore when the incident angle of photon is large,total internal reflection happens,which will significantly enhance PMT detection efficiency.Moreover,the bottom surface and inner structure of PMT can also affect its detection efficiency.After including the interference effect on photocathode in the simulation,the new PMT optical model gives an accurate description of PMT angle response This model is used in batch testing system of the PMTs.Efficiency of strong-light detection device（like PD）can be measured by its respon-sivity on a beam with known power.However,for PMTs which work in weak light,efficiency can not be measured using the same method.Using an integrating sphere as beam splitter to split the incident light into two beams with a proportion 1663:1,we can enable the PMT to work in single photon model.Using the beam intensity measured by the PD for the stronger beam,the intensity of the weaker beam going into PMT can be deduced.Moreover,On the other hand,absolute quantum efficiency of PD can be measured easily.Combining with the detection efficiency value,we can obtain collection efficiency of PMTs.The value of these efficiencies will be stable for different PMTs due to their similar structures.We review the work done in this thesis and discussion possible future extension in the last chapter.