Exploring band-free Raman microspectrometry combined with PCA and MCR-ALS for size-resolved forensic analysis of uranium in aerosols in a model nuclear atmosphere.

Journal of environmental radioactivity(2023)

引用 0|浏览3
暂无评分
摘要
Achieving non-destructive micrometer-scale molecular and structural analysis of uranic materials in atmospheric aerosols with traditional methodologies is a challenge. Spatially resolved analysis of uranium in actinide-bearing aerosols is critical for nuclear forensics. Although laser Raman microspectrometry enables this, for the normally low uranium concentrations in the aerosols the spectra are indiscernible (band-free) against pronounced background: trace analysis requires a push in analytical strategy. We combined laser Raman microspectrometry (utilizing two lasers (λ = 532 nm, λ = 785 nm)) with principal component analysis (PCA) and multivariate curve resolution-alternate least squares (MCR-ALS) to perform size-resolved analysis of uranium in aerosols. Uranium-specific Raman scatter bands corresponding to uranyl nitrate (860 cm), uranium sulphate (868 cm), uranyl chloride (816 cm) and uranium trioxide (839 cm) were detected. The 816 cm, 854 cm, 868 cm bands were resolved by MCR-ALS and used to identify and map uranium in PM size aerosols. Based on spectral feature selection of the signature bands, PCA identified two sources of aerosol particles in model nuclear atmosphere - Sea spray for PM and re-suspension of 'nuclear' dust from a rare earth element (REE) mine for PM. The MCR-ALS-resolved uranium bands showed the potential for attributive nuclear forensic analysis.
更多
查看译文
关键词
Raman microspectrometry, Machine learning, Nuclear forensics, Aerosols
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要