1
Capillary Electrophoresis-Mass Spectrometry Interfacing: Principles and Recent Developments

Lukas Naumann, Jasmin Schairer, Alisa Höchsmann, Elahe Naghdi, and Christian Neusüß

Aalen University, Faculty of Chemistry, Beethovenstraße 1, 73430 Aalen, Germany

1.1 Introduction

Capillary electrophoresis (CE) is an attractive separation technique in many fields of analytical chemistry. This is primarily due to its selectivity based on charge and size. Mass spectrometry (MS) is an excellent characterization tool. Thus, the coupling of CE with MS is highly desirable, though by far not so widespread as liquid chromatography (LC) with MS. Separation techniques can be coupled with MS either offline or online. CE is a microfluidic technique with an inner diameter (i.d.) of the separation capillary in the 20-100?µm range. The overall volume of CE capillaries (length is typically 20-100?cm) is in the order of 1?µl, with injection volumes in the low nl range. Thus, fraction collection for subsequent offline MS characterization is difficult and rarely performed for subsequent MS characterization. Only when CE is combined with matrix-assisted laser desorption/ionization (MALDI) fractions can these be collected for offline MALDI-MS characterization (see Section 1.7). Mostly electrospray ionization (ESI) is used in online coupling approaches due to its ability to ionize efficiently most of the molecules being important in bioanalysis. The coupling of CE with ESI-MS is obvious since ionic compounds are separated in most electromigrative techniques and the low flow rates in CE perfectly fit to ESI. This is especially the case when nanoESI is considered: the ionization in flow rates below 1?µl/min without additional nebulizing gas is known for sensitivity and reduced matrix effects. However, CE-ESI-MS interfacing is complicated by two major aspects: (i) the variable flow rates in CE (electroosmotic flow [) and (ii) the required voltages both for CE and for ESI. Both electrical fields need to be contacted at the end of the separation capillary without hampering the separation efficiency. Thus, special designs are needed for CE-MS, which are more complicated than those applied for LC-ESI-MS.

In this chapter, we summarize the actual status of CE-MS interfaces and discuss the most relevant techniques. Homebuilt systems are also included, although commercialization of efficient ionization techniques has made significant progress in the last decade. Due to the importance of ESI in bioanalysis, we focus on CE-ESI-MS and start with general considerations of CE-ESI-MS interfaces. This is followed by a section about the widespread sheath liquid (SL) interface setups, including both the traditional coaxial design and modern nanoflow sheath liquid (nanoSL) approaches. Afterward, sheathless approaches are presented and discussed with a focus on the commercially available porous-tip interface. Then, miscellaneous CE-ESI-MS interfaces are briefly summarized. Next, coupling microchip electrophoresis (MCE) with MS is discussed in a short section. Thereafter, alternative ionization techniques are discussed including MALDI and inductively coupled plasma (ICP). The chapter ends with a short concluding discussion and outlook on future trends.

1.2 General Considerations of CE-ESI-MS

ESI is by far the most widely applied ionization technique for CE-MS. Many interface designs have been developed addressing the special needs regarding the handling of variable flow rate in CE and the provision of electrical contact for both CE and ESI. Most relevant CE-ESI-MS interface designs are presented in Figure 1.1.

All indicated designs will be discussed in the following sections. However, we start with a short overview about ESI properties (Section 1.2.1) and general aspects of CE-MS (Section 1.2.2 and 1.2.3).

Figure 1.1 Most important interface designs for CE-ESI-MS and MCE-ESI-MS.

1.2.1 Electrospray Ionization

In principle, electrospray (ES) ionization creates ions in gas phase from a solution. To perform ESI, a conductive liquid along an electrical field is required to apply electrospray voltage. ESI takes place in three steps: (i) creation of the fine droplets in strong electrical field, (ii) droplet desolvation and shrinkage (Coulomb repulsions), and (iii) formation of the gas-phase ions from the charged offspring droplets.

The applied electrical gradient penetrates the liquid surface at the tip of the sprayer and forms a meniscus. Further, meniscus deforms and a cone, is called a Taylor cone, is created. A fine jet of charged droplets emerges from the cone apex with sufficient electrical field. The charge at the droplet surface corresponds to the electrical field's polarity and voltage. The subsequent solvent evaporation leads to a cascade of Coulomb repulsions, where the surface charge at the droplet overcomes the surface tension. It leads to smaller charged droplets. Two different models describe the subsequent formation of the gas-phase ions: the charge residue model (CRM) [1] and the ion evaporation model (IEM) [2]. The CRM suggests Coulomb repulsions continue until an ultimate droplet is formed which contains only a single analyte molecule. This molecule retains a part of the droplet's excess charge and becomes a free gas-phase ion. Proteins are expected to follow the CRM due to their relatively large size and the presence of their polar side chains, which are located at the protein surface to stabilize their solvation within a single droplet [3, 4]. Small molecules are expected to leave the shrinking droplet according to the IEM.

In general, the nanoESI provides better ionization efficiency than classical ESI because of the smaller initial droplet size, which is related to the initially formed Taylor cones. As a result, the sensitivity is higher in nanoESI. In Figure 1.2 the Taylor cone size is illustrated for the three mostly used CE-MS interface designs.

Figure 1.2 Comparison of the Taylor cone size of different spraying approaches in CE-ESI-MS at the same scale. (a) TTS, (b) nanoflow sheath liquid interface, and (c) porous-tip interface.

It obviously shows the benefit of the nanoSL (Figure 1.2b) and sheathless interface (Figure 1.2c) compared with the coaxial sheath liquid interface (Figure 1.2a) regarding initial droplet size.

However, the kind of solvent also strongly influences the ESI efficiency. A mixture of mostly water-based background electrolyte (BGE) with a high content of a medium polar solvent (methanol, acetonitrile, isopropanol) is aspired, potentially (partly) compensating or even outperforming the effect of dilution in SL interfaces.

ESI is prone to analyte signal suppression by high buffer concentration, nonvolatile components, and surfactants inherent to its liquid-phase mechanism. Besides, nonvolatile components may cause source contamination in ESI, producing high background signals. Thus, similar to LC-MS, volatile BGEs and solvents are generally used in CE-MS to increase ionization efficiency of the analytes and to avoid contamination of the MS. Nevertheless, the possibility of MS source contamination is slightly lower because of the lower CE capillary effluent than the LC column effluent [5].

1.2.2 Electrical Circuit in CE-ESI-MS

After the voltage-driven separation in the CE capillary, ESI requires a second electric circuit between the capillary exit and the MS endplate. In order to avoid any peak broadening, both electrical circuits share the same electrode, placed close to the end of the separation capillary. Ideally the potentials of both the CE and ESI can be chosen independently, in order to be flexible in the CE separation and achieve a stable and reproducible electrospray. Depending on the mass spectrometer, two approaches are possible: the shared electrode of CE and ES is grounded (mass spectrometers from, e.g. Agilent Technologies and Bruker Daltonics) or carries a potential (mass spectrometers from, e.g. Thermo Fisher Scientific, Waters, Sciex). The grounded electrode at the CE outlet/ES simplifies the independent choice of CE parameter and ES conditions. Also, the maximum CE voltage (typically 30?kV) can be used for separation, when the same polarity is applied to CE and ES. Nevertheless, it is also possible to perform efficient separation and stable ES when the shared CE outlet/ES electrode is put on (ES) voltage. In this case the remaining voltage from CE needs to be considered, as well as the involved currents. Typically, the currents of CE and ESI are very different (CE: 5-50?µA, ESI: <1?µA). Therefore, an additional electrode at the end of the CE capillary is needed! The difference between these currents needs to be handled by the ESI source of the mass spectrometer. Additional resistors guiding the high currents out of the system might be required, in...