Compound-specific isotope analysis coupled with multivariate statistics to source-apportion hydrocarbon mixtures

Area of Science:

• Environmental Chemistry
• Analytical Chemistry
• Forensic Science

Background:

• Compound Specific Isotope Analysis (CSIA) is valuable for tracking hydrocarbon degradation and weathering.
• Existing statistical methods for CSIA source apportionment are limited by error-proneness and high replicate requirements.
• Linking unknown hydrocarbons to sources is challenging due to the lack of robust statistical tools for CSIA data.

Purpose of the Study:

• To develop a rigorous statistical method for CSIA data to enable precise source apportionment.
• To overcome limitations of univariate tests and high replicate needs in multivariate CSIA analysis.
• To provide a reliable method for testing the statistical difference between unknown and potential source hydrocarbon samples.

Main Methods:

• A novel statistical approach combining pairwise Principal Component Analysis (PCA) with Multivariate Analysis of Variance (MANOVA) is presented.
• The method projects triplicate CSIA data into a simplified space for efficient multivariate analysis.
• A protocol for down-selecting putative sources using PCA and hierarchical clustering is also described.

Main Results:

• The coupled PCA-MANOVA method allows for highly precise hypothesis testing between unknown and source samples.
• This technique effectively reduces the need for numerous replicate analyses, making CSIA more practical.
• Application to Navy fuel tanks and spilled oil samples demonstrated the method's utility in source apportionment.

Conclusions:

• The developed statistical framework significantly enhances the reliability of CSIA for hydrocarbon source apportionment.
• This method, combined with traditional forensic techniques, offers a powerful toolkit for organic compound analysis.
• The approach is applicable to any organic compound amenable to Gas Chromatography-Combustion-Isotope Ratio Mass Spectrometry (GC-C-IRMS).