The diesel particulate matter (PM) is typically agglomerates of soot coated with condensed organic compounds. Three-dimensional (3D) measurement of PM particle structures is currently not undertaken as established techniques usually damage the structure of PM particles by dissociating or evaporating the condensed organic compounds absorbed on the soot. In this work, PM particles emitted from a diesel engine were sampled by a micro-orifice uniform deposit impactor and their soot content was measured by a thermo-gravimetric analyzer. The morphology and microstructure of individual PM particle were non-destructively studied using synchrotron soft X-ray tomography (SXT). The 3D tomographic reconstruction from the aligned tilt-series images yielded useful information about PM volumetric characters. Fractal-like agglomerate of soot and the condensed organic compounds were identified. The results showed that the agglomerates are constituted by loosely attached aggregates which are composed of spherules holding tightly to each other. Aggregate size distributions exhibited no specific trend with respect to the engine load and PM size in this work, while soot volume fractions calculated based on the 3D images of PM particles, decreased with increasing load and PM size. The fractal dimensions of agglomerates and PM particles were for the first time presented and calculated by image analysis. The essential properties of PM particles, agglomerates and aggregates are characterized. The population-averaged fractal dimension of agglomerates was smaller and more sensitive to engine load and PM size than that of aggregates, indicating that agglomerates was less compact than aggregates. This work demonstrated the significant advantages of SXT in the description of actual PM morphologies with high accuracy.

柴油机颗粒物质（PM）通常是涂有冷凝有机化合物的烟灰团聚体。当前不进行PM颗粒结构的三维（3D）测量，因为已建立的技术通常会通过分解或蒸发吸附在烟灰上的冷凝有机化合物来破坏PM颗粒的结构。在这项工作中，用微孔均匀沉积撞击器对柴油机排放的PM颗粒进行采样，并通过热重分析仪测量其烟灰含量。使用同步加速器软X射线断层扫描（SXT）无损研究了单个PM颗粒的形态和微观结构。从对齐的倾斜系列图像进行的3D层析成像重建产生了有关PM体积特征的有用信息。分形的煤烟团聚体和冷凝的有机化合物被鉴定。结果表明，团聚体由松散附着的聚集体构成，聚集体由彼此紧密保持的小球组成。在这项工作中，骨料粒度分布没有显示出与发动机负荷和PM尺寸有关的特定趋势，而基于PM颗粒的3D图像计算出的烟灰体积分数随负荷和PM尺寸的增加而降低。团聚体和PM颗粒的分形维数首次显示并通过图像分析进行计算。表征了PM颗粒，附聚物和聚集体的基本性质。与团聚体相比，团聚体的总体平均分形维数更小，并且对发动机负载和PM尺寸更敏感，表明附聚物比聚集体更不致密。这项工作证明了SXT在高精度描述实际PM形态方面的显着优势。