Improving Energy Reconstruction and Background Rejection of EXO-200

中微子的双β衰变可能提供了一个独特的机会来回答粒子物理学中的一个基本问题，即中微子是狄拉克费米子还是马约纳拉费米子的本质。这种衰变尚未被观察到。EXO-200是一个寻找Xe-136的中微子双β衰变的实验。探测器是一个圆柱形的，位置分辨时间投影室充满了浓缩液体氙。该实验位于美国新墨西哥州的废物隔离中试工厂。通过测量液体氙中的闪烁光和衰变电子释放的电子数，可以探测到无中微子双β衰变。闪烁光将在两个雪崩光电二极管平面上被探测到。二次电子将通过收集（U线）和感应（V线）在其漂移过程中被检测到。在第一个测量阶段，EXO-200已经与GERDA和KamLAND Zen一起实现了对马约拉纳中微子有效质量的最佳限制之一。在2016年即将开始的第二阶段数据采集中，信号处理电子器件中的噪声被降低，从而降低触发阈值，提高能量分辨率。在本项目中，需要提高能量校准和能量重建的系统精度，从而提高对衰变半衰期的灵敏度。为了实现这一目标，将对探测器的响应进行蒙特卡罗模拟，并对模拟中探测器的建模进行改进。特别是对两个带电粒子在最终状态下的信号事件特征的电荷-光反相关的建模进行了研究。除此之外，我们还将研究如何使用v线信号来提高能量分辨率。该项目还将改进对背景事件的识别，这些背景事件主要是由光子在探测器中多次散射引起的。为此，需要模拟U线和V线中的脉冲形式，并用于训练随机决策森林。随机决策森林的应用可以提高在一个位置有能量沉积的事件（单点事件，信号签名）与在多个位置有能量沉积的事件（多点事件，背景签名）之间的分离强度。应开发一种新的分析方法，该方法基于测量已检测到的事件大小的连续度量。

Neutrinoless double beta decay may offer a unique opportunity to answer a fundamental questions in particle physics, namely the nature of neutrinos being either Dirac or Majonara fermions. This decay has not yet been observed. EXO-200 is an experiment searching for neutrinoless double beta decay of Xe-136. The detector is a cylindrical, position resolving time projection chamber filled with enriched liquid xenon. The experiment is located in the Waster Isolation Pilot Plant in New Mexico,USA. The neutrinoless double-beta decay would be detected by the measurement of the scintillation light and the number of electrons released by the decay electrons in the liquid xenon. The scintillation light would be detected in two planes of avalanche photodiodes. The secondary electrons will be detected during their drift with two wire-grids by collection (U wires) and induction (V wires). During its first measuring phase, EXO-200 already has achieved, together with GERDA and KamLAND Zen, one of the best limits on the effective Majorana neutrino mass. For the upcoming phase 2 of data taking starting in 2016the noise in the signal processing electronics was reduced so that the trigger threshold will be lower and the energy resolution will be improved. During this proposed project, the energy calibration and the systematic accuracy of the energy reconstruction shall be improved resulting in a higher sensitivity on the half life of the decay.To achieve this goal, Monte-Carlo simulations of the detector response will be performed and the modeling of the detector in the simulation will be improved. Especially the modeling of the charge-light anticorrelation for the signal event signature with two charged particles in the final state will be investigated. In addition to that we will investigate how the V-wire signals can be used to improve the energy resolution. This project also will improve the identification of background events that are mostly caused by photons scattering several times in the detector. To do this, the pulse forms in the U and V wires shall be simulated and be used to train random decision forests. The separation strength between events with energy deposition at one position (single site event, signature of the signal) against events with energy depositions at several sites (multi site event, signature of background) shall be improved by the application of random decision forests. A new analysis method shall be developed which is based on a continuous metric measuring the detected size of an event.