Novel ultra-fast photodetectors for near reconstruction-less time-of-flight positron emission tomography

摘要 飞行时间正电子发射断层扫描(TOF-PET)扫描仪提供更好的信噪比(SNR) 与传统的PET系统相比，可以减少伪影。TOF-PET扫描仪的性能得到改进 凭借其探测器的定时精度：伽马光子的时间探测精度越高越好 这场演出。TOF-PET的最终目标是在半高时达到10ps的全宽(FWHM) 符合时间分辨率(CTR)，以精确地分辨三维正电子湮没点。 最先进的PET探测器由耦合到硅光电倍增管(SiPM)和SHOW的闪烁晶体组成 定时分辨率约为100-200 ps半高宽。 在这个项目中，我们致力于显著改善SiPM的时序特性，作为这样的改进 将对TOF-PET产生很大影响，因为它将改善大多数使用 闪烁晶体和/或切伦科夫光发射器增加了几倍。最先进的SiPM针对 窄波长范围(λ)，因为不同波长的穿透深度不同。为 λ=45 0 nm和λ=5 90 nm的光子，衰减深度分别为0.4μm和2μm。权衡之处在于 或者a)具有更厚的耗尽层以吸收更宽范围的波长但增加时间抖动， 或者b)具有更薄的耗尽层以减少时间抖动，但仅吸收窄范围的波长。 因此，不存在同时提供非常快的时间响应和高性能的最先进的SiPM 广泛波长范围内的光子探测效率(PDE)。 我们建议开发一种集成在耗尽层中的光子捕获微结构的SiPM原型 这使光横向分散，并使人们能够在广泛的波长范围内获得高探测效率 在1μm的耗尽层内。利用这样的薄层，电子漂移时间中的抖动减少到10ps和 暗电流预计也会减少。这种新的光传感器可能会给TOF-PET带来革命性的变化。 利用周期性微结构使光在垂直方向上弯曲并捕获光子 最近显示出增强了与材料的相互作用、高检测效率和快速响应 用于光通信的波长在800-900 nm之间。在这份提案中，我们将开发一种基于 在这项技术上。首先，我们将模拟集成空穴捕获的最佳层结构 微结构和雪崩区，以提供105的增益。第二，我们会做一个电子的 不同类型微电池的表征，包括增益校准和测量量子效率 对于每个单元格的不同λ。最后，我们将用全尺寸SiPM(3x3mm2)制造晶片，并测试 具有闪烁晶体和切伦科夫发射体的SiPM。

SUMMARY Time-of-Flight Positron Emission Tomography (TOF-PET) scanners provide better signal-to-noise ratio (SNR) and artifact reduction compared to conventional PET systems. The performance of TOF-PET scanners improves with the timing precision of its detectors: the more accuracy in the time detection of gamma photons the better the performance. The ultimate aim of TOF-PET is to reach a 10 ps full width at half maximum (FWHM) coincidence time resolution (CTR) to resolve precisely the positron-electron annihilation point in 3 dimensions. State-of-the-art PET detectors consist of scintillation crystals coupled to silicon photomultipliers (SiPM) and show timing resolutions in the order of 100-200 ps FWHM. In this project, we focus on improving dramatically the timing properties of the SiPMs, as such improvement would have a strong impact on TOF-PET as it would improve the timing performance of most detectors that use scintillation crystals and/or Cerenkov light emitters by several-fold. State-of-the-art SiPMs are optimized for a narrow range of wavelengths (λ) because of the difference in penetration depth at different wavelengths. For photons of λ=450 nm and λ=590 nm, the attenuation depth is 0.4 μm and 2 μm, respectively. The trade-off is to either a) to have a thicker depletion layer to absorb a wider range of wavelengths but to increase the time jitter, or b) to have a thinner depletion layer to reduce the time jitter but absorb only a narrow range of wavelengths. Therefore, there is not a state-of-the-art SiPM that provides, simultaneously, very fast time response, and high photon detection efficiency (PDE) across a wide range of wavelengths. We propose to develop an SiPM prototype with photon-trapping microstructures integrated in the depletion layer that disperses the light laterally and allows one to obtain high-detection efficiency for a wide range of wavelengths within a depletion layer of 1 μm. With such a thin layer, the jitter in the electron drift time decreases to 10 ps and the dark current is expected to decrease as well. This new photosensor could revolutionize TOF-PET. The utilization of periodic microstructures to bend light in a perpendicular orientation and trapping photons for enhanced interaction with materials, high detection efficiency and fast response have been recently shown for wavelengths between 800-900 nm for optical communication. In this proposal, we will develop a new SiPM based on this technology. First, we will simulate the optimum layer structure to integrate the hole-trapping microstructures and an avalanche region to provide a gain of >105. Second, we will do an electronic characterization for the different type of microcells, including a gain calibration and measure quantum efficiency for different λ for each cell. Finally, we will manufacture a wafer with full-size SiPMs (3x3 mm2) and test the SiPMs with scintillation crystals and Cerenkov emitters.