Department of Chemistry, Queen's University, Kingston Ontario, K7L 3N6, Canada.
School of Computing, Queen's University, Kingston Ontario, K7L 3N6, Canada.
Talanta. 2024 Jan 1;266(Pt 1):124959. doi: 10.1016/j.talanta.2023.124959. Epub 2023 Jul 21.
DNA evidence in sexual assault cases have proven increasingly difficult to obtain and analyse due to increased condom use. With more interest in alternatives to DNA evidence, prophylactic lubricants, spermicides and residues may be interesting prospects. Current interest in the analysis of prophylactic residues focuses on the evaluation and identification of lubricants and constituents, primarily through gas chromatography or Fourier transform infrared spectroscopy. Though cost-effective methods, extensive sample preparation and destructive modes of analysis remain an area for improvement. As a result, the focus has since shifted to ambient ionization methods that offer adequate sensitivity and reduced sample preparation. The Liquid Microjunction Surface Sampling Probe (LMJSSP) is a versatile ambient ionization source that employs a probe that supports a continuously flushing droplet that extracts analytes when placed in contact with a surface. The analytes are aspirated into the mass spectrometer with a Venturi pressure. In this work we use the LMJSSP to analyse the trace transfer of condom lubricant to different types of fabric (cotton, cotton-spandex, and denim). Furthermore, we examine the sensitivity and storage conditions for the direct analysis method on different swab types (cotton, silicone, and foam). Additionally, Principal Component Analysis (PCA) and Maximally Collapsing Metric Learning (MCML) are utilized for visualization of differentiability of commercially available condom brands including Durex™ and Trojan™, and product subtypes. The results present an interesting multi-disciplinary approach of using a direct liquid extraction ambient ionization technique and machine learning to improve the overall workflow for the analysis of lubricants, swabs and fabrics. Machine learning algorithms were able to differentiate between inherent differences of Durex™ and Trojan™ condoms.
由于避孕套使用的增加，性侵犯案件中的 DNA 证据越来越难以获取和分析。随着人们对 DNA 证据替代品的兴趣增加，预防性润滑剂、杀精剂和残留物可能是有趣的研究方向。目前，人们对预防性残留物的分析主要集中在评估和识别润滑剂及其成分上，主要通过气相色谱或傅里叶变换红外光谱法进行。尽管这些方法具有成本效益，但广泛的样品制备和破坏性分析模式仍然是一个需要改进的领域。因此，研究重点已经转移到具有足够灵敏度和减少样品制备的环境电离方法上。液相微萃取表面采样探头（LMJSSP）是一种多功能的环境电离源，采用探头支撑一个连续冲洗的液滴，当与表面接触时，该液滴提取分析物。分析物通过文丘里压力被吸入质谱仪。在这项工作中，我们使用 LMJSSP 分析了避孕套润滑剂转移到不同类型织物（棉、棉氨纶和牛仔布）的痕量转移。此外，我们还研究了不同拭子类型（棉、硅酮和泡沫）上直接分析方法的灵敏度和储存条件。此外，主成分分析（PCA）和最大压缩度量学习（MCML）用于可视化可商购避孕套品牌（包括 Durex™和 Trojan™）和产品亚型的可区分性。结果展示了一种有趣的多学科方法，即使用直接液体提取环境电离技术和机器学习来改进润滑剂、拭子和织物分析的整体工作流程。机器学习算法能够区分 Durex™和 Trojan™避孕套的固有差异。
