NanoMAX example dataset - ptychography on siemens stars - EH1

Example (single-beam single-slice) hard X-ray ptychographic dataset recorded at the Imaging Endstation (EH1) [1] of the NanoMAX beamline [2] at the MAX IV Laboratory [3]. The datasets were recorded on a siemens star sample at different photon energies. They are kept in the folder and file structure typical to the lab and beamline. The upload contains the raw data, the script used to reconstruct the raw data, the final reconstruction results and a jupyter notebook loading the reconstructions to create a figure of the reconstructed objects. Data setes were recorded at photon energies from 5 keV to 17 keV in 1 keV steps. A pair of KB-mirrors [4] was used to focus the probing X-ray beam. The sample was positioned 1500µm downstream of the focus postion and scanned in a Fermat-spiral scanning pattern [6] over a scan region of 6µm x 6µm with an average step size of 0.3µm. At each scan position a diffraction pattern was recorded with an Eiger2 X 4M detector (DECTRIS, Switzland) [7] and an exposure time of 0.5 seconds. The detector was positioned 7.180m downstream of the sample. All measurments were taken under vacuum conditions. There were no windows between the KB mirrors, the sample and the detector. Reconstructions were performed using the ptypy-framework [8] using 1000 iterations of the difference-map (DM) algorithm [9], followed by 1000 iterations of the maximum likelihood (ML) algorithm [10] without position refinement and another 1000 iterations with positions refinement. All algorithms were implemented for GPUs using cupy [11]. We acknowledge the MAX IV Laboratory for internal beamtime on the NanoMAX beamline. Research conducted at MAX IV, a Swedish national user facility, is supported by Vetenskapsrådet (Swedish Research Council, VR) under contract 2018-07152, Vinnova (Swedish Governmental Agency for Innovation Systems) under contract 2018-04969 and Formas under contract 2019-02496. References: [1] Maik Kahnt et al., "Current capabilities of the imaging endstation at the NanoMAX beamline", AIP Conf. Proc. 27 September 2023; 2990 (1): 040018. https://doi.org/10.1063/5.0169244 [2] Ulf Johansson et al., "NanoMAX: the hard X-ray nanoprobe beamline at the MAX IV Laboratory", J. Synchrotron Rad. 28, 1935-1947 (2023). https://doi.org/10.1107/S1600577521008213 [3] Aymeric Robert et al., "MAX IV Laboratory". Eur. Phys. J. Plus 138, 495 (2023). https://doi.org/10.1140/epjp/s13360-023-04018-w [4] Maik Kahnt et al., "Complete alignment of a KB-mirror system guided by ptychography," Opt. Express 30, 42308-42322 (2022) https://doi.org/10.1364/OE.470591[6] Xiaojing Huang et al., "Optimization of overlap uniformness for ptychography", Opt. Express 22, 12634-12644 (2014). https://doi.org/10.1364/OE.22.012634 [7] Tilman Donath et al., "EIGER2 hybrid-photon-counting X-ray detectors for advanced synchrotron diffraction experiments", J. Synchrotron Rad. 30, 723-738.(2023). https://doi.org/10.1107/S160057752300454X [8] Björn Enders et al., "Computational framework for ptychographic reconstructions", Proc. R. Soc. A.47220160640 (2016). http://doi.org/10.1098/rspa.2016.0640 [9] Pierre Thibault et al., “Probe retrieval in ptychographic coherent diffractive imaging,” Ultramicroscopy 109(4), 338–343 (2009). https://doi.org/10.1016/j.ultramic.2008.12.011[10] Pierre Thibault et al., "Maximum-likelihood refinement for coherent diffractive imaging," New J. Phys. 14 063004 (2012). https://doi.org/10.1088/1367-2630/14/6/063004 [11] Ryosuke Okuta et al. "Cupy: A numpy-compatible library for nvidia gpu calculations." Proceedings of workshop on machine learning systems (LearningSys) in the thirty-first annual conference on neural information processing systems (NIPS). Vol. 6. (2017).本数据集为单光束单切片型硬X射线叠层成像（ptychographic）数据集，采集于MAX IV实验室[3]的NanoMAX光束线[2]成像终端站（EH1）[1]。 该数据集采集自西门子星型标准样品，涵盖不同光子能量下的测量结果。数据按照实验室及光束线标准的文件夹与文件结构存储。本次上传内容包含原始数据、用于重建原始数据的脚本、最终重建结果，以及一份用于加载重建结果以生成重建对象图像的Jupyter笔记本（jupyter notebook）。 数据集的光子能量采集范围为5千电子伏特（keV）至17 keV，步长为1 keV。实验采用KB镜（KB-mirrors）[4]对入射X射线束进行聚焦，样品置于焦点下游1500μm处，并以费马螺旋扫描模式（Fermat-spiral scanning pattern）[6]在6μm×6μm的扫描区域内完成扫描，平均步距为0.3μm。在每个扫描位置，均使用Eiger2 X 4M探测器（DECTRIS，瑞士）[7]记录衍射图案，曝光时长为0.5秒。探测器位于样品下游7.180m处，所有测量均在真空环境下进行，且KB镜、样品与探测器之间未设置真空窗口。 重建过程使用ptpy框架（ptpy-framework）[8]完成：先通过差分映射（difference-map, DM）算法[9]迭代1000次，随后采用最大似然（maximum likelihood, ML）算法[10]迭代1000次（未进行位置校正），最后再迭代1000次并开启位置校正。所有算法均基于CuPy（cupy）[11]实现GPU加速运算。 本研究感谢MAX IV实验室提供的NanoMAX光束线内部束流时间。在MAX IV（瑞典国家用户设施）开展的研究得到了瑞典研究理事会（Vetenskapsrådet, VR）合同号2018-07152、瑞典创新系统局（Vinnova）合同号2018-04969以及Formas研究理事会合同号2019-02496的支持。 参考文献： [1] Maik Kahnt 等，《NanoMAX光束线成像终端站的当前性能》，AIP Conf. Proc. 2023年9月27日；2990(1): 040018。https://doi.org/10.1063/5.0169244 [2] Ulf Johansson 等，《NanoMAX：MAX IV实验室的硬X射线纳米探针光束线》，J. Synchrotron Rad. 28, 1935-1947 (2023)。https://doi.org/10.1107/S1600577521008213 [3] Aymeric Robert 等，《MAX IV实验室》，Eur. Phys. J. Plus 138, 495 (2023)。https://doi.org/10.1140/epjp/s13360-023-04018-w [4] Maik Kahnt 等，《基于叠层成像的KB镜系统完全对准》，Opt. Express 30, 42308-42322 (2022)。https://doi.org/10.1364/OE.470591 [6] Xiaojing Huang 等，《叠层成像的重叠均匀性优化》，Opt. Express 22, 12634-12644 (2014)。https://doi.org/10.1364/OE.22.012634 [7] Tilman Donath 等，《用于先进同步辐射衍射实验的EIGER2混合光子计数X射线探测器》，J. Synchrotron Rad. 30, 723-738 (2023)。https://doi.org/10.1107/S160057752300454X [8] Björn Enders 等，《叠层成像重建计算框架》，Proc. R. Soc. A. 472, 20160640 (2016)。http://doi.org/10.1098/rspa.2016.0640 [9] Pierre Thibault 等，《叠层相干衍射成像中的探针恢复》，Ultramicroscopy 109(4), 338–343 (2009)。https://doi.org/10.1016/j.ultramic.2008.12.011 [10] Pierre Thibault 等，《相干衍射成像的最大似然校正》，New J. Phys. 14, 063004 (2012)。https://doi.org/10.1088/1367-2630/14/6/063004 [11] Ryosuke Okuta 等，《CuPy：兼容NumPy的NVIDIA GPU计算库》，第31届神经信息处理系统大会（NIPS）机器学习系统研讨会（LearningSys）论文集，第6卷。(2017)