The collision-induced dissociation of forty-one 3-keto anabolic steroids and related compounds has been studied using both triple quadrupole (QqQ) and hybrid quadrupole-time of flight (QTOF) instruments. Due to the complexity of the product ion spectra of these analytes, which generate a large number of ions, only two specific regions were studied in depth: the product ions near the precursor ion (m/z > or =M-100) and the most abundant product ions at a collision energy of 30 eV. Accurate mass measurements were used in order to obtain an unequivocal assignment of the empirical formula and the origin of each selected product ion. Analytes have been divided into eight groups according to the number and position of double bonds and the presence of functional groups such as hydroxyl- or nitrogen-containing rings. A correlation between the steroid structure and the product ions obtained has been postulated. The application of these correlations can be useful in the elucidation of feasible structures for unknown steroids and/or their metabolites.

The ionization of 46 anabolic steroids has been studied. The absence of basic or acidic moieties in most of these analytes makes their direct ionization as [M + H]+ by atmospheric pressure interfaces difficult. The formation of adducts with different components of the mobile phase has been found to be an efficient way to ionize anabolic steroids by electrospray. Different mobile phases using methanol (MeOH) or acetonitrile as organic solvent and HCOOH, Na+ or NH4+ as additives have been tested to favor the adduct formation. A direct correlation between the chemical structure of the anabolic steroid and the possibility to ionize it in a particular chromatographic condition has been found. According to their ionization, anabolic steroids can be divided into seven different groups depending on both the nature and the relative position of their functional groups. The formation of different adducts such as [M + Na + MeOH]+ or [M + H + CH3 CN - H2O]+ is required in order to ionize some of these groups and the optimal mobile phase composition for each group of anabolic steroids is proposed. Despite the ionization limitations due to their chemical structure, most of tested anabolic steroids could be ionized using the adduct formation approach.

The detection of new anabolic steroid metabolites and new designer steroids in urine is a challenge in doping analysis. An approach based on precursor ion scanning for the detection of unknown anabolic steroids and metabolites is proposed. The study of the MS/MS spectra of selected anabolic steroids revealed different fragmentation pathways at low and medium collision energy depending on the steroid structure. However, after analysis at high collision energy three common ions at m/z 105, m/z 91, and m/z 77 were found for all studied anabolic steroids. These ions can be explained by the fragmentation of the steroid structure and corresponded to the methyl tropylium, tropylium, and phenyl ions, respectively. Because of the theoretical low specificity of these ions, the simultaneous presence of all of them was used as a starting point to consider a substance as a possible anabolic steroid. Hence, the developed approach is based on the simultaneous acquisition of the precursor ion scan of m/z 105, 91, and 77. The specificity of this approach has been checked by the injection of several doping agents including beta-agonists, corticosteroids, beta-blockers, and diuretics. In general, only compounds with a steroidal structure showed a signal at all three selected m/z values although some exceptions have been found. The applicability of the method was tested for three different scenarios: the detection of steroid metabolites, the detection of unknown steroids, and the analysis of prohormones. In metabolic studies, several recently reported fluoxymesterone metabolites were also found using this method. For detection of unknown steroids, some negative urine samples were spiked with the designer steroid THG and 33 other anabolic steroids and treated as blind samples. Finally, the applicability of the developed approach for the analysis of dietary supplements was checked by the analysis of a prohormone where several impurities and/or degradation products were found.