Simultaneous derivatization and solid-based disperser liquid–liquid microextraction for extraction and preconcentration of some antidepressants and an antiarrhythmic agent in urine and plasma samples followed by GC-FID

• Displacement of disperser solvent by a safe and low-cost substance, i.e. sugar cube.
• A one-step method including derivatization and microextraction has been developed.
• Four antidepressants and an antiarrhythmic agent were determined in urine and plasma.
• Low LODs and high EFs are advantages of the proposed method.

The present work is based on a one-step method including derivatization and solid-based disperser liquid–liquid microextraction followed by gas chromatography-flame ionization detection (GC-FID) for the determination of four antidepressants (fluoxetine, fluvoxamine, tranylcypromine, and nortriptyline) and an antiarrhythmic agent (mexiletine) in human urine and plasma samples. In this method, a mixture of 1,1,2,2-tetrachloroethane (extraction solvent) and butylchloroformate (derivatizing reagent) is added on a sugar cube (solid disperser) and it is introduced into an aqueous sample containing the analytes and a catalyst, e.g. 3-methylpyridine (picoline). During dissolving the sugar cube by manual shaking, the extractant and derivatization agent are gradually released into the sample as very fine droplets. Then the resulted cloudy solution is centrifuged and the sedimented phase is analyzed by GC-FID. The influence of several variables on the efficiency of derivatization/microextraction procedure such as kind and volume of extraction solvent, type and amount of disperser, amount of derivatization agent, and catalyst volume are optimized. Under the optimum conditions the calibration curves are linear in the range of 8–100,000 μg L−1 (coefficient of determination ≥0.994). The relative standard deviations are obtained in the range of 3.0–6.0% for all compounds. Moreover, the detection limits and enrichment factors of the target analytes are obtained in the ranges 1–15 μg L−1 and 228–268, respectively, for both plasma and urine samples. The relative recoveries obtained for the spiked plasma and urine samples are between 70 and 91%. The results show that the proposed method is simple, reliable, low cost, and applicable to determine trace amounts of the studied drugs in biological samples.