Gasoline discrimination using infrared spectroscopy and virtual samples based on measurement uncertainty

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Forensic Science

Background:

• Previous gasoline sample discrimination relied on physicochemical assays or distillation curves, requiring specialized equipment.
• Infrared (IR) spectroscopy presents a potentially more accessible and in situ alternative for forensic applications.
• Quantification and comparison of measurement uncertainties for IR spectral data in virtual sample generation are underexplored.

Purpose of the Study:

• To investigate the feasibility of using IR spectroscopy for a virtual sample-based methodology to compare gasoline samples in pairs.
• To evaluate the performance of handheld Near-Infrared (NIR) and benchtop Fourier Transform Infrared (FT-IR/FT-NIR) spectrometers.
• To describe a stepwise procedure for quantifying uncertainties in IR spectral acquisition for virtual sample generation.

Main Methods:

• Evaluated a handheld NIR device and a dual-range (FT-NIR and FT-IR) benchtop spectrometer.
• Developed a stepwise procedure to quantify uncertainties in IR spectral acquisition, considering repeatability, intermediate precision, and temperature variations.
• Employed virtual samples within SIMCA (Soft Independent Modeling of Class Analogy) models for sample discrimination.

Main Results:

• Complete discrimination rates were achieved for 96.3% (NIR), 93.4% (FT-NIR), and 93.7% (FT-IR) of 1431 gasoline sample pairs.
• These results are comparable to physicochemical methods (95.7%) and slightly lower than distillation curves (99.2%).
• The developed uncertainty quantification method aids in understanding IR measurement variability.

Conclusions:

• IR spectroscopy, particularly with portable devices, provides a viable, non-destructive alternative for gasoline sample comparison.
• The virtual sample methodology using IR data offers faster, simpler forensic investigations.
• This approach can be extended to other sample comparison applications in various fields.