Chemometric classification of casework arson samples based on gasoline content.

Area of Science:

• Forensic Chemistry
• Analytical Chemistry
• Chemometrics

Background:

• Ignitable liquid (IL) detection in arson debris is crucial for investigations but complicated by complex sample matrices and pyrolysis products.
• Standard methods like ASTM E-1618 can face data interpretation bottlenecks, impacting laboratory efficiency.
• ILs, such as gasoline, are complex mixtures whose chemical signatures are altered by fire conditions.

Purpose of the Study:

• To develop and validate automated chemometric tools for the efficient detection and identification of ILs in arson debris.
• To overcome the data interpretation challenges associated with traditional analytical methods in arson investigations.
• To streamline the analysis of complex chemical mixtures found in fire debris.

Main Methods:

• Application of chemometric techniques, including Partial Least Squares-Discriminant Analysis (PLS-DA) and Similarity-based Approach (SIMCA).
• Utilized a perdeuterated n-alkane ladder for chromatographic alignment.
• Employed an automated hybrid backward elimination and forward selection approach for variable selection, guided by the cluster resolution metric.

Main Results:

• Successfully demonstrated the automated construction, optimization, and application of chemometric models to arson casework data.
• Developed PLS-DA and SIMCA classification models using 165 training samples.
• Achieved 100% specificity and sensitivity in classifying gasoline content for 55 validation set samples.

Conclusions:

• Automated chemometric tools significantly enhance the accuracy and efficiency of ignitable liquid analysis in arson investigations.
• The developed models provide a robust and reliable method for classifying gasoline presence in complex fire debris samples.
• This approach offers a cost-effective solution to data interpretation bottlenecks in forensic laboratories.