墨水池
拉曼光谱
OPL公司
扩散
化学
分析化学(期刊)
偏最小二乘回归
材料科学
色谱法
计算机科学
光学
热力学
氢键
分子
物理
有机化学
机器学习
复合材料
标识
DOI:10.1016/j.microc.2023.108727
摘要
Ink type identification and ink formation time play a significant role in document examination. Current methods for dating stamp inks in questioned documents primarily rely on the analysis of volatile ink components. However, there is a dearth of literature on the investigation of ink distribution in the initial stages of stamp impressions. Such studies are scarcely reported, despite their potential significance in the field of forensic document examination. This study aims to explore the feasibility of ink diffusion based on Raman spectra monitoring and to identify the ink types using orthogonal partial least squares discrimination analysis (OPLS-DA) multivariate modeling. The kinetics of ink diffusion can be elucidated by the interplay of chromatographic effects and diffusion equilibrium. The diffusion boundary of the ink is visualized using a microscope, and 21 specific locations at varying distances from the boundary are designated for time-resolved spectrum acquisition. Typically, the diffusion process can be separated into two distinct phases an enrichment period followed by an equilibrium period. As the diffusion distance increases, the Raman peak intensity and ink concentration progressively diminish. Following the pre-processing of spectral data, any outliers were removed prior to utilizing the orthogonal partial least squares discrimination analysis method for classification purposes. The OPLS-DA model demonstrated 100% accuracy in differentiating between all types of stamp inks, affirming the significant potential of Raman spectroscopy for forensic science applications.
科研通智能强力驱动
Strongly Powered by AbleSci AI