Ionic Liquid-Mediated Multi-Walled Carbon Nanotube-Polydimethylsiloxane Fiber for Headspace Solid-Phase Microextraction of Phenolic Compounds in Aqueous Samples by Gas Chromatography Coupled to Flame Ionization Detector

• Published in: Journal of chromatographic science 2017-02, Vol.55 (2), p.174-181
• Main Authors: Sadri, Minoo, Vatani, Hossein
• Format: Article
• Language: English
• Subjects: Dimethylpolysiloxanes - chemistry; Drug Stability; Gas Chromatography-Mass Spectrometry - methods; Ionic Liquids - chemistry; Limit of Detection; Linear Models; Nanotubes, Carbon - chemistry; Phenols - analysis; Phenols - chemistry; Phenols - isolation & purification; Reproducibility of Results; Solid Phase Microextraction - methods; Water Pollutants, Chemical - analysis; Water Pollutants, Chemical - chemistry; Water Pollutants, Chemical - isolation & purification
• ISSN: 0021-9665; 1945-239X
• DOI: 10.1093/chromsci/bmw164

An ionic liquid-mediated multi-walled carbon nanotube (MWCNT)-polydimethylsiloxane (PDMS) sorbent was developed for headspace solid-phase microextraction (HS-SPME) of phenolic compounds from human urine samples. Sol-gel method was used to prepare this sorbent. For this purpose, MWCNTs were functionalized covalently and were attached chemically to the hydroxyl-terminated PDMS. Prepared fiber showed high thermal stability (over than 320°C) and good lifespan (>210 times). These good performances can be attributed to the performance of carbon nanotubes and sol-gel method. Affecting parameters on the efficiency of HS-SPME were investigated and optimized. Under the optimal conditions, linear dynamic ranges were observed over a range of 0.002-200 ng mL with limits of detection from 0.0005 to 0.005 ng mL and limits of quantitation between 0.002 and 0.02 ng mL The relative standard deviations for one fiber (repeatability) (n = 5) at three different concentrations (0.05, 2 and 100 ng mL ) were obtained from 4.6 up to 6.7% and between fibers or batch to batch (n = 3) (reproducibility) in the range of 5.7-7.8%. Urine samples were used as real samples. All real samples were spiked at 0.5 ng mL of understudy analytes and the relative recovery percentages obtained from 90.7 to 102.1%.