Developing Picosecond Time-Resolved X-ray Infrastructure at the APS

在 APS 开发皮秒时间分辨 X 射线基础设施

• 批准号: 8553425
• 负责人: Philip Anfinrud
• 金额: $ 25.05万
• 依托单位: NATIONAL INSTITUTE OF DIABETES AND DIGESTIVE AND KIDNEY DISEASES
• 依托单位国家: 美国
• 资助国家: 美国
• 起止时间: 至
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/8553425
• 关键词: Achievement; Algorithms; Area; Attenuated; Capital; Chicago; Collimator; Commit; Computer software; Crystallography; Custom; Data; Diagnostic; Ensure; Equipment; European; FarGo; France; Funding; Goals; Hair; Head; Heating; Height; Hybrids; Image; Intercept; Intramural Research; Lasers; Length; Magnetism; Measurement; Microscope; Molecular Structure; Monitor; Motion; Motor; National Institute of Diabetes and Digestive and Kidney Diseases; Optics; Penetration; Performance; Photons; Physiologic pulse; Platelet Factor 4; Positioning Attribute; Procedures; Process; Programming Languages; Proteins; Protocols documentation; Pump; Pythons; Radiation; Relative (related person); Research Infrastructure; Resolution; Roentgen Rays; Sampling; Scanning; Schedule; Scheme; Science; Solutions; Source; Speed; Spottings; Staging; Synchrotrons; System; Temperature; Time; Training; United States National Institutes of Health; Universities; Vertical Dimension; Vision; Writing; X ray diffraction analysis; X-Ray Diffraction; base; beamline; conditioning; design; detector; improved; millisecond; research study; software development; synchrotron radiation

Up until FY 2008, the ID09B time-resolved X-ray beamline at the European Synchrotron and Radiation Facility (ESRF) in Grenoble, France was the only facility in the world capable of determining time-resolved macromolecular structures with 150-ps time resolution and < 2-Angstrom spatial resolution. The Anfinrud group was instrumental in helping develop that capability at the ESRF. Unfortunately, the ESRF operates in a mode that is optimized for time-resolved Laue crystallography studies only 14 days out of each year, and we had access to only a portion of this limited amount of beam time. To expand the amount of beam time available for our studies, we partnered with the Advanced Photon Source (APS) in Argonne, IL and BioCARS to develop picosecond time-resolved X-ray capabilities on Sector 14 at the APS. BioCARS is an NIH-funded beamline headed by Prof. Keith Moffat, and is operated by the University of Chicago. In FY2005, Dr. Marvin Gershengorn, then Director of Intramural Research at NIDDK, committed > $1M to procure the capital equipment needed for this effort. Our vision was to achieve picosecond time-resolved X-ray capabilities comparable to that realized at the ESRF when the APS is operated in 24-bunch mode, a common operating mode used 132 days per year. This goal required that we isolate a single bunch of X-rays from a train of pulses separated by only 153 ns, a feat that we first achieved in July 2007 using a high-speed chopper whose rotor was fabricated according to our custom design. To maximize the number of photons delivered to the sample in a single X-ray pulse, we replaced the existing U33 undulator (33-mm magnetic period) with two newly designed U23 and U27 undulators, making BioCARS the first APS beamline to operate with two inline undulators. NIDDK funded this effort, with the APS supplying the labor to design and refurbish two undulators according to our performance specifications. When the gaps of these undulators are tuned to generate 12-keV X-ray photons, the X-ray fluence is comparable to that generated at the ESRF during their 4-bunch mode. When the APS operates in their exotic hybrid mode, which is scheduled approximately 31 days per year, the X-ray fluence is a factor of 4 higher than that available with the ESRF 4-bunch mode. These achievements increase by more than an order of magnitude the amount of beamtime available worldwide to pursue 150 ps time-resolved X-ray science. The infrastructure needed to pursue picosecond time-resolved X-ray studies goes far beyond delivering single X-ray pulses to the experimental hutch. We installed a picosecond laser system in a laser hutch located near the X-ray hutch, as well as an array of laser diagnostics that aid optimization of the laser performance. We have also developed a Field-Programmable-Gate-Array (FPGA) based timing system that synchronizes all time-critical components to the X-ray pulses. For example, the FPGA drives the heat-load chopper, the high-speed chopper, the picosecond laser system, a millisecond shutter, and various other motion controls that must be synchronized with the x-ray pulse arrival time. Importantly, we can set the time delay between X-ray and laser pulses from picoseconds to seconds with a precision of 10 ps. We also developed the diffractometer used to acquire time-resolved X-ray diffraction images. This effort included the design and fabrication of a millisecond shutter, a motorized support for the high-speed X-ray chopper, a support for motorized X-ray slits, detectors for non-invasively monitoring the laser and X-ray pulse energy and relative time delay, a motorized stage for the X-ray detector, supports for a collimator pipe and X-ray beam stop, beam conditioning optics that tailor the laser pulses in both space and time, beam delivery optics that focus the laser pulses onto the sample, motorized controls to center the focused laser pulse on the sample, and motorized controls to center the collimator pipe on the X-ray beam. Finally, we continue to refine the software developed to control the beamline. This software package, called LaueCollect, is written in the Python programming language, and is generalized for both time-resolved Laue crystallography and time-resolved SAXS/WAXS studies. For these experiments, it is crucial to achieve long-term stability of the laser and x-ray beams, and also achieve precise alignment of the crystal at the intersection of the laser and x-ray beams. We have made improvements this past year in all three of these areas. Because of the large distance between the laser system and the sample (> 30 m), thermal drift in the laser hutch, the hall of the synchrotron ring, and the X-ray hutch can cause the laser beam position to drift from its set point. To mitigate this problem, we have developed a protocol to monitor and maintain the laser alignment. We installed a camera to monitor the beam position within the beam conditioning optics enclosure and a second camera to image the focused spot at the sample position. If the beam position drifts from the cross hairs of either camera by more than a preset threshold, the corresponding beam-steering optics are adjusted using precision motors and the beam is re-centered. This combination of controls allows us to maintain precise spatial overlap between the laser and X-ray pulses at the sample position. We have encountered long-term drift in the laser and x-ray timing as well, which appears to track thermal drift in the experimental hall, laser hutch, and/or experimental hutch. We have no direct control over the temperature, whose drift can cause the time delay to shift by a magnitude that exceeds the X-ray pulse duration. This problem significantly compromises the accuracy of time-resolved measurements on the 100 ps time scale. We have developed a scheme to non-invasively monitor the timing drift as data are being collected, and established a protocol that periodically corrects for this drift. Due to the large mismatch between the laser and X-ray penetration depths in protein crystals, we employ an orthogonal pump-probe geometry with the laser beam directed downward and the X-ray beam horizontal. It is crucial to align the top edge of the protein crystal at the top edge of the x-ray beam: if the crystal height is set too low, we observe no diffraction; if set too high, the laser pulse cannot penetrate to the depth of the X-ray pulse, and we observe no pump-induced change in the crystal diffraction. To that end, we developed microscope imaging software that allows us to define visually the top edge of the crystal by pointing and clicking at several points along the length of the crystal. Repeating this process at various phi angles generates a three-dimensional wire-grid definition of the crystal edge, and is used as a starting point for determining the edge of the crystal using X-rays, which is much more precise. To minimize radiation damage when executing the edge finding algorithm, we acquire diffraction images from 10-fold attenuated X-ray pulses. We scan the crystal vertically in 20 &#956;m steps (half the vertical dimension of the X-ray beam) and determine the integrated spot intensity from a single X-ray pulse at each position. We developed an efficient algorithm for quantifying the integrated intensity of spots on the detector, with each step of the vertical scan taking less than 0.5 s. The integrated spot intensity is proportional to the crystal volume intercepted by the X-ray pulse, and is used to define the edge of the crystal. Once found, the crystal is raised 40 &#956;m to position the top edge of the crystal at the top edge of the X-ray beam. This procedure ensures precise positioning of the crystal, and has demonstrably improved the quality of the time-resolved diffraction data.

截至 2008 财年，位于法国格勒诺布尔的欧洲同步加速器和辐射设施 (ESRF) 的 ID09B 时间分辨 X 射线束线是世界上唯一能够以 150 ps 时间分辨率和< 2埃空间分辨率。 Anfinrud 小组在帮助 ESRF 开发这种能力方面发挥了重要作用。不幸的是，ESRF 的运行模式针对时间分辨劳厄晶体学研究进行了优化，每年只有 14 天，而我们只能使用有限的射束时间中的一部分。为了延长我们研究可用的光束时间，我们与伊利诺伊州阿贡的高级光子源 (APS) 和 BioCARS 合作，在 APS 的第 14 区开发皮秒时间分辨 X 射线功能。 BioCARS 是一个由 NIH 资助的光束线，由 Keith Moffat 教授领导，由芝加哥大学运营。 2005 财年，时任 NIDDK 校内研究主任的 Marvin Gershengorn 博士承诺投入超过 100 万美元来采购这项工作所需的资本设备。 我们的愿景是实现皮秒时间分辨 X 射线能力，与 APS 在 24 束模式（每年使用 132 天的常见操作模式）下运行时在 ESRF 中实现的能力相当。这一目标要求我们从间隔仅 153 ns 的脉冲串中分离出单束 X 射线，这是我们于 2007 年 7 月首次使用高速斩波器实现的壮举，该斩波器的转子是根据我们的定制设计制造的。为了最大限度地增加在单个 X 射线脉冲中传递到样品的光子数量，我们用两个新设计的 U23 和 U27 波荡器取代了现有的 U33 波荡器（33 毫米磁周期），使 BioCARS 成为第一个使用两个内联波荡器。 NIDDK 资助了这项工作，APS 提供劳动力根据我们的性能规格设计和翻新两个波荡器。当这些波荡器的间隙调整为产生 12 keV X 射线光子时，X 射线注量与 ESRF 在 4 束模式下产生的 X 射线注量相当。当 APS 在奇特的混合模式下运行时（每年大约 31 天），X 射线注量比 ESRF 4 束模式高 4 倍。这些成就使全球可用于 150 ps 时间分辨 X 射线科学的波束时间增加了一个数量级以上。 进行皮秒时间分辨 X 射线研究所需的基础设施远远超出了向实验室提供单个 X 射线脉冲的范围。我们在 X 射线室附近的激光室中安装了皮秒激光系统，以及一系列有助于优化激光性能的激光诊断系统。我们还开发了基于现场可编程门阵列 (FPGA) 的计时系统，可将所有时间关键组件与 X 射线脉冲同步。例如，FPGA 驱动热负载斩波器、高速斩波器、皮秒激光系统、毫秒快门以及必须与 X 射线脉冲到达时间同步的各种其他运动控制。重要的是，我们可以将 X 射线和激光脉冲之间的时间延迟设置为皮秒到秒，精度为 10 ps。我们还开发了用于获取时间分辨 X 射线衍射图像的衍射仪。这项工作包括设计和制造毫秒快门、高速 X 射线斩波器的电动支架、电动 X 射线狭缝的支架、用于非侵入性监测激光和 X 射线脉冲能量及相关能量的探测器。时间延迟是 X 射线探测器的电动平台，支持准直管和 X 射线光束阻挡器、在空间和时间上定制激光脉冲的光束调节光学器件、将激光脉冲聚焦到目标上的光束传输光学器件。样品，电动控制装置将聚焦激光脉冲集中在样品上，电动控制装置将准直管置于 X 射线束的中心。最后，我们继续完善为控制光束线而开发的软件。该软件包名为 LaueCollect，采用 Python 编程语言编写，适用于时间分辨 Laue 晶体学和时间分辨 SAXS/WAXS 研究。 对于这些实验，实现激光束和 X 射线束的长期稳定性以及在激光束和 X 射线束交叉处实现晶体的精确对准至关重要。去年我们在这三个领域都取得了进步。 由于激光系统与样品之间的距离较大 (> 30 m)，激光室、同步加速器环的大厅和 X 射线室中的热漂移会导致激光束位置偏离其设定点。为了缓解这个问题，我们开发了一种协议来监控和维护激光对准。我们安装了一台摄像机来监测光束调节光学器件外壳内的光束位置，并安装了第二台摄像机来对样品位置处的聚焦点进行成像。如果光束位置偏离任一摄像机的十字线超过预设阈值，则使用精密电机调整相应的光束转向光学器件，并使光束重新居中。这种控制组合使我们能够在样品位置保持激光和 X 射线脉冲之间的精确空间重叠。 我们也遇到了激光和 X 射线定时的长期漂移，这似乎可以跟踪实验大厅、激光室和/或实验室中的热漂移。我们无法直接控制温度，其漂移会导致时间延迟偏移的幅度超过 X 射线脉冲持续时间。这个问题严重影响了 100 ps 时间尺度上时间分辨测量的精度。我们开发了一种方案，可以在收集数据时非侵入性地监测时间漂移，并建立了一个定期纠正这种漂移的协议。 由于激光和 X 射线在蛋白质晶体中的穿透深度之间存在很大的不匹配，我们采用正交泵浦探针几何结构，其中激光束向下，X 射线束水平。将蛋白质晶体的顶部边缘与 X 射线束的顶部边缘对齐至关重要：如果晶体高度设置得太低，我们将观察不到衍射；如果设置得太高，激光脉冲无法穿透到 X 射线脉冲的深度，并且我们观察到晶体衍射中没有泵浦引起的变化。为此，我们开发了显微镜成像软件，使我们能够通过指向并单击沿晶体长度的几个点来直观地定义晶体的顶部边缘。以不同的 phi 角度重复此过程会生成晶体边缘的三维线栅定义，并用作使用 X 射线确定晶体边缘的起点，这要精确得多。为了在执行寻边算法时最大限度地减少辐射损伤，我们从 10 倍衰减的 X 射线脉冲中获取衍射图像。我们以 20 μm 的步长（X 射线束垂直尺寸的一半）垂直扫描晶体，并确定每个位置的单个 X 射线脉冲的积分光斑强度。我们开发了一种有效的算法来量化检测器上斑点的积分强度，垂直扫描的每一步花费的时间不到 0.5 秒。积分光斑强度与 X 射线脉冲截获的晶体体积成正比，用于定义晶体的边缘。找到后，将晶体升高 40 μm，将晶体的顶部边缘定位在 X 射线束的顶部边缘。该过程确保了晶体的精确定位，并明显提高了时间分辨衍射数据的质量。