Date of Submission

5-2024

Document Type

Thesis

Degree Name

Master of Science in Forensic Science

Department

Forensic Science

Advisor

Brooke W. Kammrath, Ph.D.

Committee Member

Marisia A. Fikiet, Ph.D.

Committee Member

Peter. R. DeForest, Crim.D.

Keywords

Forensic Soil Analysis, Raman Spectroscopy, Particle-correlated Raman spectroscopy (PCRS)

MeSH

Spectrum Analysis, Raman

LCSH

Soils--Analysis, Raman spectroscopy

Abstract

Soil is a continuous material, reaching across geological bodies, and is distinctly variable and differentiable based on aspects such as geographic location, seasonal factors, and human interference. In combination with its highly transferable nature, the complexity of soil makes it valuable material for forensic trace evidence and object-to-scene association. The development of forensic soil analysis is currently focused on mineralogy. Raman spectroscopy is a chemical technique that has previously been studied for mineralogy uses. Additionally, Raman spectroscopy has been researched for use in other fields of forensic science as well. As such, it is possible to use Raman spectroscopy for the forensic analysis of soils as well.

Particle-correlated Raman spectroscopy is a novel technique that combines microscopic analysis of particles with Raman identification. PCRS provides qualitative and quantitative information. When applied to soil minerals, this information includes identification and microscopic morphological characteristics (e.g. circularity, volume, area). This study explores the use of PCRS to discriminate soil minerals transferred onto mock-evidence items from different locations. Shoes and shovels were used to collect mock-evidence from three different geographical locations an urban park, a rural woodland, and a suburban residential area. Known soil samples were collected from these locations to serve as reference samples for comparison.

The soil adhering to mock-evidence items was collected and cleaned to isolate the mineral grains. The particles in the diameter range of 90-180μm were manually dispersed onto a Ramaninactive microscope slide and analyzed using PCRS. The results were compared to the reference sample results, and source consistency could be determined using mean and distribution statistical analysis. Adding heavy mineral separation to the sample preparation process allowed for a larger number of particles to increase features usable for match criteria.

This study showed that area, diameter, perimeter, and volume are suitable metrics for PCRS analysis of soil. The mean and distribution, seen through single-factor ANOVA and Kruskal-Wallis tests, for these morphological features for multiple mineral classes were statistically indistinguishable between reference and mock-evidence samples. This study showed that PCRS is a powerful tool for mixture analysis and can demonstrate the discriminating potential of minerals for soil forensics.

Available for download on Saturday, May 09, 2026

Share

COinS