ABSTRACT Reliable chemical analysis is essential for provenance estimation in archaeology, yet valuable iron artifacts (e.g., iron products) are often unsuitable for destructive testing. In contrast, iron slag and iron sand (a potential raw material) often allow for destructive analysis. This study investigated suitable sample preparation methods for energy‐dispersive X‐ray fluorescence (EDXRF), a nondestructive, rapid, and cost‐effective analytical technique, applied to furnace slag, tapping slag, and iron sand. Three specimen types—unprepared samples, loose powder, and pressed powder pellets—representing three different preparation methods were compared using their EDXRF‐obtained quantitative values and relative standard deviations. The results showed that unprepared samples produced inconsistent values, whereas loose powder and pressed powder pellets yielded similar. Additionally, only a minimal amount of loose powder and three to five measurements were sufficient to obtain accurate and representative data. Overall, preparing samples as loose powder emerged as the most practical, simple, and time‐efficient method for EDXRF analysis, providing a reliable approach for provenance estimation in archaeological studies where multiple samples must be analyzed.
Ichikawa et al. (Sun,) studied this question.