In catalysis, nanoparticles enable chemical transformations and their structural and chemical fingerprints control activity. To develop understanding of such fingerprints, methods studying catalysts at realistic conditions have proven instrumental. Normally, these methods either probe the catalyst bed with low spatial resolution, thereby averaging out single particle characteristics, or probe an extremely small fraction only, thereby effectively ignoring most of the catalyst. Here, we bridge the gap between these two extremes by introducing highly multiplexed single particle plasmonic nanoimaging of model catalyst beds comprising 1000 nanoparticles, which are integrated in a nanoreactor platform that enables online mass spectroscopy activity measurements. Using the example of CO oxidation over Cu, we reveal how highly local spatial variations in catalyst state dynamics are responsible for contradicting information about catalyst active phase found in the literature, and identify that both surface and bulk oxidation state of a Cu nanoparticle catalyst dynamically mediate its activity.

在催化作用中，纳米颗粒能够实现化学转化，其结构和化学指纹控制活性。为了加深对此类指纹的理解，在现实条件下研究催化剂的方法已被证明是有用的。通常，这些方法要么以低空间分辨率探测催化剂床，从而平均化单颗粒特征，要么仅探测极小部分，从而有效地忽略大部分催化剂。在这里，我们通过引入包含 1000 个纳米粒子的模型催化剂床的高度多重单粒子等离子体纳米成像来弥补这两个极端之间的差距，这些模型集成在纳米反应器平台中，可实现在线质谱活性测量。以铜上的 CO 氧化为例，我们揭示了催化剂状态动力学的高度局部空间变化如何导致文献中发现的催化剂活性相的信息相互矛盾，并确定铜纳米颗粒催化剂的表面和本体氧化态动态介导它的活动。