Effect of sample matrices on supported liquid membrane: Efficient electromembrane extraction of cathinones from biological samples

In this work, the effect of sample matrix on electromembrane extraction (EME) was investigated for the first time using cathinones (log P < 1.0) as polar basic model analytes. Ten supported liquid membranes (SLMs) were tested for EME from spiked buffer solutions, urine, and whole blood samples, respectively. For buffer solutions, SLMs containing aromatic solvents provided higher EME recovery than non-aromatic solvents, which confirmed the significance of cation-π interactions for EME of basic substances. Interestingly, when applied to urine and whole blood samples, aromatic SLMs were less efficient, while non-aromatic SLMs containing abundant hydrogen-bond acidity/basicity were efficient. These observations were explained by SLM fouling, and the antifouling property of the SLM was clearly dependent on the nature of the SLM solvent. Accordingly, a binary SLM containing aromatic 1-ethyl-2-nitrobenzene (ENB) and non-aromatic 1-undecanol (1:1 v/v) was developed. This binary SLM was not prone to fouling, and provided high recoveries of cathinones from urine and whole blood. EME based on this SLM was optimized and evaluated in combination with liquid chromatography tandem mass spectrometry (LC-MS/MS), and the linear ranges with R² ≥ 0.9903 for cathinones in whole blood and urine were 5–200 ng/mL and 1–200 ng/mL, respectively. The LOD and LOQ of cathinones were ranged from 0.12 to 0.54 ng/mL and 0.38–1.78 ng/mL, respectively. The repeatability and accuracy bias at three levels were ≤11% and within 10%, respectively. In addition, the matrix effect ranged from 88% to 118% was also in compliance with guidelines for bioanalytical method validation provided by the European Medicines Agency.