PM2.5 is a complex pollutant composed of chemical species such as sulfates, nitrates, and ammonium compounds, posing threats to human health and global climate change. This study investigates the global spatiotemporal variations of PM2.5 chemical compositions between 2005 and 2016 and identifies their key meteorological and socioeconomic drivers. Through results from a combination of heatmap visualizations and the random forest machine learning model, the study reveals persistent pollution hotspots in East and South Asia, with notable co-occurrence of ammonium and nitrate components. The analysis highlights that PM2.5 concentrations are influenced primarily by meteorological factors such as precipitation and planetary boundary layer height, the chemical components are associated more with anthropogenic factors such as population density and energy use. The findings underscore the differential roles of natural and human drivers across pollutant types, thus providing a foundation for targeted mitigation strategies for environmental policymaking.