New approaches for capillary electrophoresis–mass spectrometry in drug analysis : Evaluation of photo-, chemical and thermospray ionization

Abstract

During the past two decades, capillary electrophoresis–mass spectrometry (CE–MS) has emerged as a powerful analytical tool that is well suited to the analysis of pharmaceutical samples. CE–MS combines efficient and fast separation with mass-selective detection and can be considered as orthogonal to LC–MS. Until now, the coupling of CE with MS has predominantly been performed through electrospay ionization (ESI). However, conventional CE background electrolytes (BGEs) like phosphate and borate buffers may cause ion suppression in ESI, and thus lead to reduced or even complete loss of analyte signals. This especially hinders the application of special CE-modes such as micellar electrokinetic chromatography (MEKC) in which (nonvolatile) surfactants are added to enhance selectivity and/or to achieve separation of neutral compounds. In addition, ESI is less suited to analytes that are not readily ionized in solution. In this thesis, the design and performance of CE–MS systems based on alternative ionization techniques, viz. atmospheric pressure photoionization (APPI), atmospheric pressure chemical ionization (APCI) and theromospray ionization (TS), has been investigated. In both APCI and APPI, the sample is first vaporized after which ionization is initiated by a corona discharge or VUV irradiation, respectively. Under specific conditions, polar compounds could be detected in the absence of corona discharge or VUV-photons. In this case, ion formation largely appeared to occur through TS. APPI and APCI were both found suitable for ionization of neutral compounds that were poorly detected by electrospray ionization. APPI enabled the formation of odd-electron ions, and therefore provided a means to detect apolar compounds upon separation by MEKC. The gas-phase-ion formation of ionic species, such as quaternary ammonium compounds, does not rely on charge-exchange or proton-transfer and can therefore not be analysed by APCI-MS or APPI-MS. Instead, this type of compounds may be analysed by TS-MS, which can be carried out using either an APPI or APCI source. Furthermore, under specific conditions, APPI and TS can be employed simultaneously, which ensures the applicability of CE–APPI-MS for a wide variety of compounds ranging from ionic to apolar. With ESI and TS analyte signals were significantly suppressed in the presence of nonvolatile buffers or surfactants. By contrast, APCI and APPI showed a strong compatibility with these constituents. With all CE–MS systems, impurities in drugs down to 0.1% (w/w) could be detected and identified. Overall, ESI and TS appeared the most efficient ionization techniques, especially for impurities that are charged in solution, resulting in detection limits down to 100 ng/mL in the full-scan mode. The signal response for APCI and APPI was often lower than for ESI, but still enabled detection at the 0.1% level by injection of 1 mg/mL of the parent compound. These results show that APPI, APCI and TS are suitable ionization techniques for drug impurity profiling by CE–MS. The availability of alternative ionization methods for CE–MS enhances the compatibility of this technique with nonvolatile buffers and surfactants, and allows a larger number of impurities in drugs to be detected.