Nucleation site distribution is ubiquitous in many natural and industrial processes, such as liquid-to-vapor phase change, gas-evolving reactions, and solid-state material growth. However, a comprehensive understanding of nucleation site distribution remains elusive. These limitations are due to the challenge of probing micro/nanoscopic nucleation sites and inadequate statistical interpretation of the nucleation process. Here we report direct experimental observation of nucleation site distribution in droplet condensation using phase-enhanced environmental scanning electron microscopy. We also use statistical theory to demonstrate that the population of nucleation sites is governed by the Poisson distribution, whereas the nearest-neighbor distance follows the Rayleigh distribution instead of the commonly used Poisson distribution. We further show the broad applicability of these insights into nucleation site distribution to hydrogen-evolving reactions and chemical vapor deposition. Our platform, combining precise characterization and theory, advances the fundamental understanding of nucleation phenomena and guides designs from materials to devices.

成核位点分布在许多自然和工业过程中无处不在，例如液相到气相的相变，气体反应和固态材料的生长。但是，对成核位点分布的全面了解仍然难以捉摸。这些限制是由于探测微观/纳米镜形核位点的挑战以及对形核过程的统计解释不足所致。在这里我们报告使用相增强的环境扫描电子显微镜对液滴凝结中成核位点分布的直接实验观察。我们还使用统计理论来证明成核位点的数量由泊松分布控制，而最近邻居距离遵循瑞利分布而不是常用的泊松分布。我们进一步展示了这些见解在成核位点分布到析氢反应和化学气相沉积中的广泛适用性。我们的平台结合了精确的特性描述和理论，可增进对成核现象的基本理解，并指导从材料到器件的设计。