<div>Mass spectrometry (MS) is a powerful and versatile analytical tool that is extensively used for the identification and analysis of complex mixtures. The ability to couple MS to atmospheric pressure ionization techniques and high-performance liquid chromatography (HPLC) or gas chromatography (GC) provides a high degree of experimental flexibility. MS is based on the analysis of gas-phase ions. Gas-phase ions are manipulated within the mass spectrometer and separated for detection based on their mass-to-charge (m/z) ratio.</div><div>One of the most commonly used techniques for complex mixture analysis is tandem mass spectrometry (MSⁿ). MSⁿ involves the isolation of the desired ion and allowing it to undergo reactions, such as collision-activated dissociation (CAD) or ion-molecule reactions. Based on the generated product ions, structural information can be obtained for unknown analytes in complex mixtures. In addition, MSⁿ methods based on diagnostic gas-phase ion-molecule reactions have been demonstrated to provide a general and predictable tool to identify specific functional groups in unknown ionized analytes and to classify unknown analytes into different compound classes depending on their functionalities.</div><div>The research described in this dissertation mainly focuses on the development of tandem mass spectrometric methods based on gas-phase ion-molecule reactions and/or CAD for the identification of the <i>N</i>-nitroso functionality, which is present in some potentially mutagenic drug impurities. Furthermore, the dissertation discusses combining machine learning and MSⁿexperiments based on diagnostic ion-molecule reactions of 2-methoxypropene to predict reaction outcomes in a semiautomated fashion for protonated analytes containing specific functional groups. Lastly, chemical characterization and gas-phase reactivity of the boron-centered radical dianion [B₁₂I₁₁]^2-• toward some organic molecules are discussed.</div>