Modified graphene-silica as a sorbent for in-tube solid-phase microextraction coupled to liquid chromatography-tandem mass spectrometry. Determination of xanthines in coffee beverages.

Given the increasing need for analyzing natural or contaminating compounds in complex food matrices in a simple and automated way, coupling miniaturized sample preparation techniques with chromatographic systems have become a growing field of research. In this regard, given the low extraction efficiency of conventional sorbent phases, the development of materials with enhanced extraction capabilities is of particular interest. Here we present several synthesized graphene-based materials supported on aminopropyl silica as sorbents for the extraction of xanthines. The synthesized materials were characterized by infrared spectroscopy and scanning electron microscopy. Aminopropyl silica coated with graphene oxide and functionalized with octadecylsilane/end-capped (SiGOC18ecap) showed the best performance for xanthines extraction. Hence, this material was employed as an in-tube solid phase microextraction (in-tube SPME) device coupled online with ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) and applied for the analysis of xanthines in roasted coffee samples. Extraction parameters and detection conditions were optimized. The method showed low limits of quantification (0.3-1.0 µg L-1), precision as relative standard deviation (RSD) values lower than 10%, recoveries between 73 and 109%, and pre-concentration factors from 5.6 to 7.2. Caffeine was determined in all ground roasted and instant coffee samples, in a wide range (0.9 to 36.8 mg g-1), and small amounts of theobromine and theophylline were also detected in some samples. This work demonstrated that functionalized graphene-based materials represent a promising new sorbent class for in-tube SPME, showing improved extraction capacity. The method was efficient, simple, and fast for the analysis of xanthines, demonstrating an excellent potential to be applied in other matrices.