A Synchrotron-based Machine learning Approach for RasTer (SMART) mineral mapping was developed for the purpose of training a mineral classifier for characterization of millimeter-sized areas of rock thin sections with micron-scale resolution. An Artificial Neural Network (ANN) was used to extract relationships between coupled micro x-ray fluorescence (μXRF) data for element abundances and micro x-ray diffraction (μXRD) data for mineral identity. Once trained, the resulting classifier, i.e., the SMART mineral mapper, can identify minerals using only μXRF data. This is the real value of this machine learning approach because μXRF data are relatively fast to collect and interpret whereas μXRD data take longer to collect and interpret. Training and testing of an initial mapper were done with 192 coupled μXRF-μXRD data points sampled from a 0.25 mm² area of a shale from the Eagle Ford formation, which was scanned with 2 μm resolution. All data used in this work were obtained from the Advanced Photon Source synchrotron beamline 13-ID-E at Argonne National Laboratory. Three minerals were mapped in the Eagle Ford rock sample, for which there were 8 elements characterized. In the testing phase, the minerals were correctly classified with accuracy of 97 % and higher. The trained SMART mapper was applied for self-similar upscaling by mapping a 14 mm² scan of the Eagle Ford sample. Generated maps captured micro-scale features characteristic of the stratified texture of the rock, and the identified minerals agreed well with bulk XRD analysis of the powdered rock. The SMART mapper was also applied to a scan of a 6-mineral mixture of known composition to demonstrate ability to distinguish minerals of similar chemistry. The trained SMART mapper is transferable to scans from other x-ray microprobes because of the μXRF data normalization that accounts for sample- and beamline-specific properties like thickness, detector configuration, and incident energy.

开发了一种基于同步加速器的机器学习方法用于 RasTer (SMART) 矿物绘图，目的是训练矿物分类器，用于表征具有微米级分辨率的岩石薄片的毫米大小区域。人工神经网络 (ANN) 用于提取元素丰度的耦合微 X 射线荧光 (μXRF) 数据与矿物特性的微 X 射线衍射 (μXRD) 数据之间的关系。经过训练后，生成的分类器，即 SMART 矿物映射器，可以仅使用 μXRF 数据识别矿物。这是这种机器学习方法的真正价值，因为 μXRF 数据的收集和解释速度相对较快，而 μXRD 数据需要更长的时间来收集和解释。初始映射器的训练和测试是使用从 0.25 mm²来自 Eagle Ford 地层的页岩区域，扫描分辨率为 2 μm。这项工作中使用的所有数据均来自阿贡国家实验室的高级光子源同步加速器光束线 13-ID-E。在 Eagle Ford 岩石样本中绘制了三种矿物，其中有 8 种元素被表征。在测试阶段，矿物被正确分类，准确率达到 97% 或更高。经过训练的 SMART 映射器通过映射 14 mm ²鹰福特样品的扫描。生成的地图捕获了岩石分层结构的微观特征特征，并且识别出的矿物与粉状岩石的整体 XRD 分析非常吻合。SMART 映射器还用于扫描已知成分的 6 种矿物混合物，以证明能够区分具有相似化学性质的矿物。受过训练的 SMART 映射器可转移到来自其他 X 射线微探针的扫描，因为 μXRF 数据归一化考虑了样品和光束线特定的属性，如厚度、探测器配置和入射能量。