Nanoparticles in Capillary Electrophoresis hyphenated with Mass Spectrometry -What are the benefits?

David Malmström

Department of Physical and Analytical chemistry, Analytical Chemistry Uppsala University

Licentiate Thesis 2011

Abstract Malmström, David 2011. Nanoparticles in Capillary Electrophoresis hyphenated with Mass Spectrometry – What are the benefits? Department of Physical and Analytical chemistry. 38 pp. Uppsala. Capillary electrophoresis is a technique that separate compounds based on charge and size. During the last two decades capillary electrophoresis (CE) and mass spectrometry (MS) has gained in interest, become more robust to use and able to separate neutral analytes. The sepa-ration of neutral analytes was first achieved with packed columns, but several disadvantages can be obtained with a stationary phase and the method of packing capillaries. Therefore, pseudostationary phases became a good alternative. The risk of clogging, memory effects and lower efficiency could be minimized with pseudophases. However, since mass spectrometry has become the most important analytical detector, and play a key role in the search for bio-markers in clinical applications, it is important that CE can successfully be combined with MS. To obtain this hyphenation several types of interfaces for the vital ion source exist. In paper I an atmospheric pressure photoionization interface was investigated in order to ac-complish an improved detection sensitivity. The knowledge attained with this type of inter-face could then be transferred to the one used in paper II, the electrospray ionization inter-face (ESI), where the use of a MS friendly nanoparticle based pseudostationary phase was investigated. Both studies showed that it is still possible to improve the separation technique and modify the ion source in order to improve the detection sensitivity for capillary electro-phoresis hyphenated with mass spectrometry.

David Malmström, Department of Physical and Analytical chemistry, Box 599 SE-751 24 Uppsala, Sweden

© David Malmström

”Things should be made as simple aspossible, but not any simpler"

Albert Einstein

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Axen, J., Malmström, D., Axelsson, B-O., Petersson, P.,

Sjöberg, P. J. R. (2010) Efforts to improve detection sensitivity for capillary electrophoresis coupled to atmospheric pressure photoionization mass spectrometry. Rapid Communication in Mass Spectrometry, 24(9) 1260-1264

II Malmström, D., Axen, J., Bergquist, J., Viberg, P., Spegel, P. (2011) Continuous full filling capillary electrochromatography – electrosprayingchromatographic nanoparticles. Electrophore-sis, 1(32) 1-7

The author’s contribution to the papers:

Paper I: Took part in planning the study, performed the experiments and took part in writing the paper.

Paper II: Planed the study, performed the experiments and wrote the paper.

Reprints were made with permission from the respective publishers.

Papers not included in this thesis

i Shevchenko, G., Sjödin, M. O. D.†, Malmström, D.†, Wetter-hall, M., Bergquist, J. (2010) Cloud-Point Extraction and Deli-pidation of Porcine Brain Proteins in Combination with Bot-tom-Up Mass Spectrometry Approaches for Proteome Analysis. Journal of Proteome Research, 9(8) 3903-3911

ii Eriksson, A., Bergquist, J., Edwards, K., Hagfeldt, A., Malmström, D., Agmo Hernaìndez, V., (2010) Optimized Pro-tocol for On-Target Phosphopeptide Enrichment Prior to Ma-trix-Assisted Laser Desorption-Ionization Mass Spectrometry Using Mesoporous Titanium Dioxide. Analytical Chemistry, 82(11) 4577-4583

iii Eriksson, A., Bergquist, J., Edwards, K., Hagfeldt, A., Malmström, D., Agmo Hernaìndez, V., (2011) A novel meso-porous TiO2-based experimental layout for on-target enrichment and separation of multi-and monophosphorylated peptides prior to analysis with matrix assisted laser desorption-ionization mass spectrometry. Analytical Chemistry, 83 (3) 761-766

† These authors contributed equally to this manuscript

Contents

1.� Introduction ........................................................................................... 9�

2.� Capillary Electrophoresis ..................................................................... 12�2.1.� General concept of Capillary Electrophoresis ............................ 13�2.2.� Electroosmotic Flow ................................................................... 14�2.3.� Capillary modifications .............................................................. 15�

3.� Electrokinetic chromatography ............................................................ 16�3.1.� Why Pseudostationary Phases? .................................................. 16�3.2.� Micellar Electrokinetic Chromatography ................................... 17�

4.� Mass Spectrometry .............................................................................. 19�4.1.� The use of Mass Spectrometry ................................................... 20�4.2.� Ion sources .................................................................................. 20�

5.� Modifications when hyphenating Capillary Electrophoresis to Mass Spectrometry ............................................................................................ 22�5.1.� Tapered capillaries ..................................................................... 23�5.2.� Requirements of the interface ..................................................... 23�

6.� Adaptation of Electrokinetic Chromatography and Pseudostationary Phases to Mass Spectrometry ................................................................... 26�6.1.� Traditional Micellar Electrokinetic Chromatography

adapted to Mass Spectrometric detection ................................... 27�6.2.� New Pseudostationary Phases with improved Mass

Spectrometric compatibility ....................................................... 29�6.3.� Continuous Full Filling .............................................................. 30�

7.� Concluding remarks ............................................................................. 33�

8.� Acknowledgments ............................................................................... 34�

9.� References ........................................................................................... 35�

Abbreviations

APPI Atmospheric Pressure Photoionization APCI Atmospheric Chemical Iionization BGE Background electrolyte CFF Continuous Full Filling CE Capillary Electrophoresis CEC Capillary Electrochromatography CMC Critical Micelle Concentration CZE Capillary Zone Electrophoresis EKC Electrokinetic Chromatography EOF Electroosmotic Flow ESI Electrospray Ionization HPLC High Performance Liquid Chromatography i.d. Inner dimension ITP Isotachophoresis MALDI Matrix Assisted Laser Desorption/Ionization MEKC Micellar Electrokinetic Chromatography MS Mass Spectrometry MS/MS Tandem Mass Spectrometry o.d. Outer dimension PF Partial Filling PSP Pseudostationary Phase RM Reverse Migrating SDS Sodium Dodecyl Sulfate

9

1. Introduction

Electrophoresis, which is the separation of charged molecules based on dif-ferent migration in an applied electric field, has since it was introduced by Tiselius [1] in the 1930’s and later on adapted into capillary formats by Hjertén and others [2-4] developed into a robust separation technique. John Craig Venter et al. [5] showed in 2001 the great power of capillary electro-phoresis (CE) when the human genome was sequenced well ahead of sche-dule. However, there are some limitations with CE, such as the inability to separate nonionic solutes and separate analytes with the same size-to-charge ratio. There have been ways to solve these limitations such as packed capil-laries, capillary electrochromatography (CEC), and the use of pseudostatio-nary phases (PSP). The basic principal of CEC is that a fused silica capillary is packed with a stationary phase. The packing materials are typically con-ventional high pressure liquid chromatography (HPLC) materials such as bonded silica and ion exchangers or monoliths. The most obvious advantage of CEC is that it combines the features of CE, with its flat-flow velocity profile generated by the electroosmotic flow (EOF) that reduces peak broa-dening, with the selectivity in HPLC. However, ever since the introduction of CEC it faced some major challenges. For instance, the method of packing the capillaries was problematic, since bubble formations and frit problems arises and can lead to column-to-column variation. This will affect the EOF since it is controlled mainly by the surface of the packing material. Due to the difficulties in the packing of the capillaries and thereby varying EOF CEC can be seen as irreproducible in a larger perspective. There have been many different way of improve the packing of solid particles into the capilla-ries: capillaries with pressurized ends [6], sol-gel methods with in-situ poly-merization [7] and gel based CEC [8]. But the challenges with producing good frits still exist. [9] Despite the challenges, the development in CEC is still active and developing since its breakthrough in the second half of the 1990’s as can be seen in Figure 1.

10

Figure 1. Number of published articles with a topic containing (A) capillary electro-phoresis or (B) capillary electrochromatography (source ISI Web of knowledge).

Another way to circumvent the limitations of packed columns was to use a pseudophase. This interesting technique was introduced by Terabe and oth-ers in 1984 where a solution of ionic micelles was introduced in the back-ground electrolyte [10-14]. This separation principle first mentioned by Na-kagawa [15] and is based on that micelles can migrate in an aqueous solution by electrophoresis.

The use of mass spectrometry (MS) has during the past 25 years reached an outstanding position among analytical methods due to its sensitivity and low detection limits. One of the areas where mass spectrometry has become the method of choice is in the enormous field of complex protein samples – pro-teomics [16]. MS has also been recognized by the Nobel Committee by giv-ing the Nobel Prize in chemistry in 2002 for the ESI and laser desorption ionization. Olivares et al. [17] and Smith et al. [18] were the first to combine the powerful technique of mass spectrometry with capillary electrophoresis. The interface was based on the electrospray ionization (ESI) technique in-itially developed by Dole et. al. [19, 20] and Fenn et al. [21, 22]. Due to their great work CE-MS is now routinely used in areas such as metabolome anal-ysis [23]. During the last decade, nanoparticles have been a hot topic, including as PSP in CE (see Figure 10). The use of nanoparticles as PSP´s, in combination with continuous full filling (CFF) and an orthogonal ESI-interface, has shown to be a good way to separate neutral analytes without affecting the electrospray process or contaminate the ion source [24, 25]. An orthogonal method of coupling MEKC and MS utilize an alternative ionization source – atmospheric pressure chemical ionization (APCI) and atmospheric pressure photoionization (APPI) [26]. In these methods the ionization efficiency has been shown to be unaffected by the surfactant SDS using both APCI and APPI [26-28].

11

This licentiate thesis is based on fundamental studies showing how the sepa-ration as well as the coupling of CE with MS can be improved. Paper I shows how the detection sensitivity in APPI-MS was optimized by varying different ion source parameters. The results indicated that a specially de-signed APPI interface for low flow rates would be favorable. In paper II, a pseudostationary phase involving nanoparticles and the continuous full fill-ing technique was investigated. Both separation and ion source parameters were investigated and optimized leading to improvements in column-to-column variability and detection limits.

12

2. Capillary Electrophoresis

Capillary electrophoresis (CE) is an umbrella term for many different me-thods. The general benefit of CE is the highly efficient separations for all the electrophoretic driven separations in CE. However, the inability to separate neutral analytes with traditional capillary zone electrophoresis (CZE) is a major limitation. But capillary electrochromatography (CEC), which is a packed capillary with the same stationary phase as in HPLC, could solve this challenge. Although CEC and HPLC are very similar, the main difference is the flat flow velocity profile generated by the EOF in CEC, in comparison to the pressure driven parabolic flow velocity profile in HPLC. This difference in front profile lead to the highest separation efficiencies of all time. But unfortunately, the reproducibility was a major challenge in CEC. The prob-lem turned out to depend on the heterogeneous surface charges resulting in local variations in EOF. Despite several approaches to minimize the prob-lems by improving the packing methods, improved frit fabrications and in-situ polymerization the reproducibility still remain a challenge [9]. In the predecessor of capillary electrophoresis, gel electrophoresis, the gel was a good way of avoiding joule heating, but unfortunately lead to convec-tion and band broadening. When trying to exclude the gel for a liquid based system it turned out to be impossible to separate neutral analytes and ana-lytes with the same size-to-charge ratio. Gel filled capillaries has also been investigated, similar to traditional gel electrophoresis. However, producing and using gel filled capillaries was often quite difficult due to bubbles present inside the gel. The challenge was elegantly solved by using non-cross-linked gels because the gel could be injected into the capillary and after separation discarded [29, 30]. Apart from the non-cross linked acryla-mide gels, other types of non-trapped stationary phase have been used in CE and CEC (and will be discussed in chapter 3).

13

2.1. General concept of Capillary Electrophoresis The separation by electrophoresis is based on different movement of charged compounds (ions) by attraction or repulsion in an electric field [2, 3]. In CE, the electrophoretic separation is taking place in a thin glass capillary made of fused silica. The inner diameter of the capillary is typically 25-75 μm, which can be compared to a human hair shaft, typically at 100 μm in diameter. The outer surface of the capillary is coated with a polymeric substance, polyi-mide, which gives a tremendous flexibility to the otherwise very fragile ca-pillary. In comparison to traditional gel electrophoresis, CE is filled with a buffer. The choice of buffer has a major influence since it will connect the applied high electric field. In gels this applied current is in higher degree limited by Joule heating, the heating of a conducting medium as current flows through it. This will have an effect on the speed of the separation due to the limitation in using a higher potential. Electric fields between 100 and 500 V/cm is possible to use with only minimal heath generation. With higher electric fields come shorter analysis time, higher peak efficiencies and better resolution. The inner diameter (i.d.) of the capillary also has a large effect on the limitations. An increase in i.d. lead to a lowering of the surface-to-volume ratio. A capillary with 50 μm in i.d. would have around 78 times higher surface-to-volume ratio compared to a standard slab gel and 1.5 times higher than the capillary with 75 μm in i.d. The high surface-to-volume ratio will allow for very efficient dissipation of Joule heat generated from large applied electric fields. As a result, electrophoretic separations in capillary electrophoresis can easily be performed at up to 30 000 volts. The simplest form but also the most commonly utilized type of CE is capil-lary zone electrophoresis (CZE). The capillary is filled with an appropriate separation buffer with a desired pH, sample is injected in the inlet (normally the anode end) while the other end of the capillary is grounded either by placing it in another buffer vial or adding an extra liquid flow (sheath liquid) (see Figure 2). As a high voltage is applied into the two buffer reservoirs, ionic species in the sample plug will start to migrate with an electrophoretic mobility determined by their charge and mass. However, if the applied elec-tric field were the only driving force acting on the ions, just net positively-charged compounds (cations) would be separated along the capillary while neutral compounds would remain static and net negatively-charged com-pounds (anions) would be driven back into the buffer vial in the inlet end, making CZE quite limited in use. Fortunately, a force called electroosmotic flow (EOF) creates a movement of all components in the capillary towards the cathode (under normal conditions).

14

Figure 2. The principal of CE where the negatively charged silica surface in combi-nation with the applied electric current creates the EOF indicated with a blue arrow. Analytes will be migrating with an electrophoretic mobility determined by their charge and mass which then will lead to a separation of the analytes. On the capil-lary wall the static Stern Layer can be seen, as well as the moving Outer Helmholtz Plane in the middle of the capillary.

A marker is useful to add as an internal standard in order to determine rela-tive migration times for the compounds. A neutral compound, e.g. dimethyl sulfoxide (DMSO) is used as only charged species can be separated in CZE while neutral analytes will migrate with the EOF. Herein lay also one major problem with CZE, which lead to the development of electrokinetic chroma-tography, EKC, including micelles and nanoparticles.

2.2. Electroosmotic Flow The silanol groups (SiO-) on the inside wall of the capillary (see Figure 2) will attract cationic species from the buffer. This attracted ionic layer has a positive charge that will decrease exponentially as the distance from the capillary wall increases. The double layer that is formed closest to the sur-face is named Stern Layer or Inner Helmholtz [31] and can be seen as a stat-ic layer. A mobile and more diffused layer is formed distal to the Stern Layer, called Outer Helmholtz Plane (OHP). When a field is applied, cations will start to move in the OHP towards the cathode but carrying water mole-cules at the same time. The cohesive nature of the hydrogen bonding of the

15

waters of hydration to the water molecules of the bulk solution will result in the entire buffer solution being pulled towards the cathode. This bulk flow is called electroosmotic flow (EOF) and can be seen as a pumping mechanism that will carry all analytes. A flat flow front profile will be created by the EOF in contrast to a parabolic flow created by a pressurized driven flow. Due to the flat flow profile all analytes will be affected in the same way leading to sharp peaks in contrast to the parabolic flow typically present in HPLC. The BGE will have a large effect on the EOF due to the pH depen-dency of the silanol groups at the capillary inner wall. Apart from pH, the ionic strength will have a large effect on the EOF. High ionic strength, or electrolyte concentration, will compress the Stern Layer, decrease the zeta potential (the potential between two layers of the opposite charge) and there-by reduce the mobility of EOF. By optimizing the EOF it is then possible to have a short analysis time while still having well separated peaks.

2.3. Capillary modifications Basic proteins and peptides have a high tendency to interact with the capil-lary wall creating a problem in protein and peptide analysis due to low effi-ciencies, low recovery, poor reproducibility, and decreased sensitivity [32, 33]. Many ways of modifying the capillary in order to improve the separa-tion and sensitivity have therefore been tried, such as extreme pH values, high salt concentrations and different additives into the BGE [34-38]. How-ever, when using ESI/MS the preferable strategy has been to permanently coat the inner capillary wall. The permanently coatings are attached to the capillary wall prior to analysis and are therefore excluded from the buffer electrolyte making it suitable with ESI/MS. However, despite the regenera-tion of the coatings by flushing the capillary in between injections with a diluted solution in order to overcome stability issues and repeatability, the time and cost required to prepare the coating could be a challenge to over-come. Coatings could be used in order to reduce, neutralize or reverse the EOF as well. Gilges et al. [39] showed for instance how polyvinyl alcohol and other polymers could be used as a dynamic coating and effectively sup-press the EOF and thereby focus the different zones, as in isoelectric focus-ing (IEF). Except modifications on the inside of the capillary wall, coatings on the tip and outside wall in order to secure the electrical connection be-tween the capillary end ESI source using metalized coatings and emitters has been developed [40-46]. This type of modification has shown an important role in sheathless ESI, which will be discussed in chapter 5.2.

16

3. Electrokinetic chromatography

“Never, never, never give up"

Winston Churchill Since CZE cannot separate neutral analytes due to the lack of electrophoretic mobility, the technique Electrokinetic chromatography (EKC) was devel-oped. In 1982 Terabe and co-workers [47] published the first successful se-paration with sodium dodecyl sulfate (SDS) as an additive in the buffer. The technique was later called micellar electrokinetic chromatography (MEKC). Apart from micelles, other types of additives have been developed and used in EKC. How the additives are introduced into the flow has been varied in order to use mass spectrometry as detection without ion suppression and ion source contamination.

3.1. Why Pseudostationary Phases? The most frequently used separation technique for analysis in liquid phases is HPLC. However, since it requires large sample and buffer volumes, mi-niaturized μLC and nanoLC systems [48, 49] have gained in popularity. To improve the separation efficiency, the particle size in the stationary phase has been decreased in parallel to the instrumentation miniaturization [50]. Due to the nature of a stationary phase the risk of sample adsorption and column clogging could get higher with the decreased column- and particle size. The problem with memory effects in the column due to “sticky sam-ples” is another problem existing with stationary phases e.g. HPLC and CEC. It is therefore not a coincidence why the use of “moving stationary phases” i.e. pseudostationary phases, as in paper II, has gained in popularity [47, 51].

17

3.2. Micellar Electrokinetic Chromatography Terabe was one of the pioneers in introducing selectivity in CE with the technique electrokinetic chromatography [10-14]. However, it was later called micellar electrokinetic chromatography (MEKC) to specify the use of micelles. These initial reports by Terabe and co-workers resulted in an ex-plosion of studies during the 1990’s with the use of different surfactants, polymers, macromolecules and particles as PSP’s [14, 47, 52, 53].The tech-nique was based on partitioning analytes between two phases, a pseudosta-tionary phase composed of charged micelles and the surrounding electrolyte phase migrating with different velocities. Terabe generated anionic micelles out of SDS and showed that the micelles had an electrophoretic mobility directed towards the positively charged anode. Despite the electrophoretic mobility out of the capillary the micelles will migrate towards the detector due to the high EOF. Analytes would be portioned between the micellar phase and the aqueous phase. A neutral analyte, which will spend no time with the micelles, will travel with a migration time of t0, while the neutral analytes that will spend all of its time with the micelles with the migration time tmc. Other neutral analytes will partition themselves in between the two phases and thereby migrate at intermediate times (see Figure 3).

Figure 3. Schematic figure of the separation mechanism in MEKC using anioic micelles. The principle is similar to the traditional liquid chromatography, i.e. the sample is distributed between two phases migrating with different speeds. But in contrast to traditional chromatography, the pseudostationary phase in EKC has an electrophoretic mobility. Despite the fact that the micelles are strongly negatively charged and therefore should migrate towards the injector end, the net mobility will be towards the detector side due to the high EOF under neutral and alkaline condi-tions.

18

By adding SDS to the BGE, anionic, cationic, and neutral analytes can be separated in the same run. The separation of neutral analytes will be based on the differences in hydrophobicity and hydrogen bonding, and can be con-trolled by varying surfactant concentration, pH and the organic additives e.g. methanol or acetonitrile. The surfactant or micelle in MEKC can be com-pared to the stationary phase in conventional liquid chromatography. The selectivity in MEKC depends on the structure of the surfactant mole-cule. The selectivity can therefore be manipulated by varying the polar head group [54] or mixing different surfactants. However, the surfactant concen-tration, buffer, pH of electrolyte solution, or the temperature are other para-meters to manipulate in order to vary the selectivity. But the most convenient way is to incorporate or modify an organic solvent to the BGE. That was also shown by S.K. Poole and C.F. Poole [54, 55] to be the most efficient way to manipulate the selectivity. The buffer composition and concentration, pH, temperature, and voltage only had a small influence on selectivity [56]. Coupling MEKC with a mass spectrometric detection has proven to be a great challenge. The main reason is due to that the most commonly used surfactant, sodium dodecyl sulphate (SDS), cause massive ionization sup-pression, contamination of the ion source and increased background signals when using ESI/MS [28, 57]. Nonetheless, there have been reports on the direct coupling of MEKC with ESI/MS utilizing SDS as a surfactant [27, 58]. Even though they could obtain the required LOD for the application, significant ionization suppression was found. Numerous attempts have been made to solve these problems, such as the partial filling (PF) technique in combination with MEKC [59] or reverse the migrating micelles to prevent the surfactants from ionizing and entering the ion source.

19

4. Mass Spectrometry

”Intelligence is the ability to adapt to change"

Stephen Hawking Since the 1980’s mass spectrometry has reached an unsurpassed status as analytical tool because of its universality, sensitivity, selectivity and ability to give the elemental composition, structural information and qualitative and quantitative information of the analyte [60-64]. Although several different types of MS instrumentations exist, they all include three main parts. The first part is an ionization source where the analytes are ionized and trans-ferred into the gas phase, a mass analyzer that will sort the analyte ions by their mass-to-charge (m/z) ratio, and finally a detector which will monitor the separated ions. Tandem mass spectrometry (MS/MS) involves an extra step. An extra mass analysis will occur after the ions with a specific m/z are fragmented. Fragmentation can occur through addition of energy or by colli-sion with a gas. The benefit of MS/MS is the valuable information given about the molecular structures of an unknown analyte due to the often unique fragmentation pattern for a specific analyte. Other benefits of MS/MS are the possibilities of higher selectivity and sensitivity. [65, 66]

20

4.1. The use of Mass Spectrometry Mass spectrometry has gained an enormous impact in important fields such as proteomics, peptidomics, and metabolomics [67, 68]. This is due to the introduction of soft ionization techniques, electrospray ionization (ESI) and Matrix Assisted Laser Desorption/Ionization (MALDI). These techniques made it possible to analyze intact biological molecules such as proteins, pep-tides, and polymers. The use of mass spectrometry has become an important part in the large scale study on the proteome, proteomics. Except the use of MS to discovering potential biomarkers for severe diseases like Alzheimer’s and Parkinson’s disease [69], MS has been used in other fields such as fo-rensic science. The use of gas chromatography/mass spectrometry have shown to be an important tool when detecting drug abuse in athletes, or detect accelerants in arson investigations. Modified MS instrumentations, so-called space MS is valuable in NASA’s Origins theme, and have been travel-ing in space since early 1960s and have returned information about the solar wind isotopes and other isotopic ratios. [66] Since MS can be used in so many different fields of interest, high demands are put on the instrumenta-tion. The MS need to be fast in order to keep up with hyphenated techniques and be sensitive enough to detect low amounts, but at the same time give reproducible results.

4.2. Ion sources In 1963, Richardson [20] developed a technique where he sprayed a diluted polymer solution onto a carbon coated film. The solvent was allowed to spread out in the carbon surface and thereafter be evaporated. The remaining single solid entities on the carbon surface could then be photographed in an electron microscope. This turned out to be a forerunner to the most common-ly used ionization technique today, electrospray ionization (ESI). In the late 1980’s ESI was combined with MS due to the work by Fenn et al. ESI as well as thermospray (TS) and aerospray (AS) all share the same basic prin-ciples. A sample solution enters the ion source through a stainless steel needle. The needle is maintained at a few kilovolts relative to the surround-ing chamber. As a result, the electric field at the needle tip will charge the introduced liquid. Due to Coulomb forces the liquid will disperse into a fine spray of charged droplets. The applied drying gas will evaporate the solvent, leading to smaller droplets but with higher charge density at the surface. This may then lead to a Coulomb explosion where the droplet tears apart. This progression of events can be repeated until ions exist in ambient gas. The electric field will let the droplets and ions migrate towards the inlet of the MS.

21

Figure 4. Two commonly used interfaces in MS; to the left is the APPI ion source with the photoionization lamp seen in the window underneath the heater. To the right is the ESI interface with the spray needle tip visible at the entrance to the mass spectrometer.

One of the latest among the soft ionization techniques in mass spectrometry is the APPI ion source [70]. This ionization technique has shown less depen-dency on buffer composition compare to the ESI source, and works good at low flow rates. This indicates that APPI would be suitable with CE, due to the commonly used non-volatile buffers and low flow rates [71]. In contrast to ESI the spray needle is introduced into a heater where the sample is vapo-rized (Figure 4). The ionization process will then be initiated by a discharge lamp. One of the major benefits with APPI is the possibility to discriminate between the analyte and solvent ions. By selecting a suitable UV source, so that the selected photon-emission energy is higher than the ionization ener-gies (IEs) of the target molecules, but lower than the IEs of the solvent and surrounding air it is possible to take full advantage of the APPI source. Therefore, the Krypton lamp (10.0 eV) is probably the most suitable UV source since it has a photon energy lower than the major components of air and the most commonly used solvents. For instance, hexane, isopropanol, methanol, oxygen, acetonitrile, water, and nitrogen all have higher IE than the energy of the emitted photons by the Krypton lamp and are therefore not ionized. The probability to directly ionize an analyte molecule is not very likely to occur, and therefore can a so-called dopant be introduced. This was shown by Sprangler et al. [72] in where introducing a suitable substance could significantly increase the number of ions. A dopant is effective if it is photoionizable and can function as an intermediate between the photons and the analyte, like the typical dopant acetone (9.70 eV).

22

5. Modifications when hyphenating Capillary Electrophoresis to Mass Spectrometry

When hyphenating CE to MS one has to consider several aspects, such as the ionization technique used. APPI and APCI, for instance, have a tendency to be more mass flow sensitive than ESI. This indicates that injecting a large volume of sample will give a higher response in MS. However, the low sam-ple loading abilities in CE becomes a limitation. The injection volume in CE is limited to approximate 3 % of the capillary, leading to that the only possi-bility is to increase the inner diameter or increase the capillary length in or-der to increase the injected sample volume. On the other hand, APPI has a tendency to be more suitable with nonpolar compounds and BGEs with high salt content. In protein and peptide analysis ESI/MS and MALDI/MS are the most com-monly used techniques. Since MALDI operates under high vacuum, transfer-ring solutions from a liquid based separation to the high vacuum in the ion source has been a challenge. The method of choice has been an off-line sys-tem where a fraction collector has been used to spot down the liquid on a plate before analyzing with MALDI. However, due to the high efficiency of CE, large demands are put on the fraction collector. Coupling MEKC with a mass spectrometric detection has proven to be a great challenge. The main reason is due to that the most commonly used surfactant, sodium dodecyl sulphate (SDS), cause massive ionization sup-pression, contamination of the ion source and increased background signals when using ESI/MS [28, 57]. Nonetheless, there have been reports on the direct coupling of MEKC with ESI/MS utilizing SDS as a surfactant [27, 58]. Even though they could obtain the required LOD for the application, significant ionization suppression was found. Numerous of attempts have been made to solve these problems, such as the partial filling (PF) technique in combination with MEKC [59] or reverse the migrating (RM) micelles to prevent the surfactants from ionizing and entering the ion source. These techniques therefore require specific optimization, with large restrictions in the background electrolyte (BGE) composition. Other procedures reported have involved complicated capillary connections constructed to dilute or remove surfactants from the separation capillary effluent [73-75].Yet, other

23

approaches have involved volatile surfactants or high molecular weight pseudostationary phases (PSP’s) [76-78]. Despite these technical develop-ments, coupling of MEKC with ESI-MS has so far been shown to require careful optimization and suffer from limitations in selecting interaction phase.

5.1. Tapered capillaries Tapered and narrowed restrictor capillaries have shown to be an aid in sup-pressing the bubble formation and an interesting alternative to frits in CEC [6]. Tapering the capillary tip in CE has shown to be especially good when working with low flow rates and with sheathless interfaces. The tapered ca-pillary permit a more stable electrospray and make it possible to operate at lower ESI voltages, and give better signal-to-noise ratios [79]. The tapering of the tip can be made in different ways, like pulling during heat, etching with hydrofluoric acid or by mechanical grinding. However, to reproduce a tapered capillary tip by mechanical grinding could be a challenge.

5.2. Requirements of the interface The interfaces for MS are originally designed to be used with higher liquid flows, commonly achieved with HPLC. Since CE uses significantly lower volumes and losses of analytes to the walls due to adsorption will be more crucial, it is therefore reasonable to believe that the ion sources need to be modified. This has been indicated in several studies and shown as well in paper I [71, 80, 81].

Figure 5. Modifications of the interface can be necessary when hyphenating CE with MS. Paper I showed an improved signal intensity when moving the sprayer closer to the ionization region.

24

One of the major requirements of the interface is that the current from the CE is maintained or the separation will be interrupted. There are three main ways of securing the electrical circuit; liquid junction, sheath liquid inter-face, and sheathless interface (see Figure 6). In a liquid junction interface a stainless steel union serves as a point for the high voltage contact. A nebu-lizer gas and an extra liquid could be introduced in order to create a good spray. Even though this is an easy and excellent way of define the electric field some negative aspects exist. Due to the electrode reactions occurring inside the stainless steel union gas bubble formation could interfere with the separation and spray formation. The distance between the two capillaries and the sheath liquid/nebulizing gas tubing will create a small volume, leading to band broadening in the separation. Despite the mentioned challenges and the fact that the electric field is not defined at the end of the capillary, leading to a laminar flow beyond that point, liquid junction is commonly used.

Figure 6. Three variations of ESI needles for CE-MS: liquid junction (left), sheath flow (top), and sheathless interface (right).

In both paper I and II, a sheath flow interface was used. In this type of in-terface the sheath liquid and nebulizer gas is introduced at the tip of the ca-pillary by concentrically stainless steel tubes. This will thereby minimize the risk of gas bubble formations, avoiding any post-column region with laminar flow, and give a more robust interface and reproducible results. Studies have also shown an increase in theoretical plate numbers with the coaxial sheath liquid interface compared to the liquid junction interface. One other advan-tage with the coaxial interface over the liquid junction interface is that the sheath liquid is delivered independently of the BGE. This will provide more

25

flexibility in the type of BGE used. [82, 83] However, since sheath liquid is introduced, a dilution of the sample is unavoidable.

The sheathless ESI interface on the other hand does not dilute the sample since no extra liquid is introduced. Sheathless ESI uses a thin layer of a con-ductive material on the outside of the emitter end in order to secure the elec-tric circuit. Even though several challenges arise with this method, such as instability of the conductive coating and the total reliance on the eluting so-lution to create the spray, sheathless ESI has shown a great potential. The choice of metal coating onto the fused silica has a large influence. A silver layer have shown to be unstable under red-ox conditions at the ESI tip, and sputtered gold have a finite lifetime before the electrical contact is lost [41, 46]. However, other coatings such as graphite-polyimide mixtures have shown a good performance and stability for a long time under oxidative stress [44] especially in combination with a tapered capillary. One major difference between sheathless ESI and sheath flow and liquid junction inter-faces is the added organic modifier to the separation buffer in order to facili-tate the ESI process. However, an organic modifier can be added to the BGE for other reasons than facilitate a good spray. In paper II the organic mod-ifier was added in order to vary the interaction between neutral analytes and the nanoparticles. The type and amount of organic modifier have shown to largely impact the electrospray stability with both the sheath liquid and sheathless ESI interface. A careful optimization to avoid a decrease in EOF by too high concentrations or disrupting the electrospray process is therefore needed [25, 84].

Figure 7. Paper II showed that the amount of organic modifier in the BGE and the sheath liquid had a large affect on the ionization efficiency.

26

6. Adaptation of Electrokinetic Chromatography and Pseudostationary Phases to Mass Spectrometry

On-column UV-detection is by far the most prevalent detection technique applied in CE and it is also the detection technique with which most of the commercially available CE instruments are equipped. MEKC was invented on such equipment and for this reason most of the PSPs in use today are UV-transparent. However, whereas coupling of CE with MS detection was fairly easy, a number of problems were identified. A major problem was the ana-lyte ionization suppression, high background signals and contamination of the ion-source with non-volatile PSP’s [28, 57]. Three different solutions to solve these problems will be discussed in this chapter. The first solution involved adaptation of the traditional MEKC protocols to MEKC/MS. This involved adjustments of buffer composition, surfactant concentration, partial filling (PF) of the PSP, and reverse migrating (RM) PSPs. The second solu-tion was to design new types of PSPs with improved MS compatibility. This involved high molecular weight as well as semi-volatile PSPs. The third solution involved instrumental modifications including inventions in the electrospray interface as well as application of other ionization sources such as APPI and atmospheric pressure chemical ionization (APCI). Inventions in MEKC/MS within these three categories will be reviewed in the following sections.

27

6.1. Traditional Micellar Electrokinetic Chromatography adapted to Mass Spectrometric detection

The important advantages of MEKC over HPLC are the higher efficien-cies achievable as well as selectivity originating from both electropho-retic and chromatographic mechanisms. Furthermore, as the PSP is rep-lenished after every separation, memory effects, stationary phase ageing and column clogging are minimized. However, the compatibility of MEKC with MS detection has, as previously mentioned, been poor. Still, studies were conducted in which the buffer composition and surfactant concentra-tion was altered to allow for MS detection [52, 58]. Although the LODs re-ported were high, some analytes could still be detected. Ionization suppres-sion, source contamination and high background signals were observed in all studies and the obvious solution to these problems was to exclude the surfac-tants from entering the ion source. Lamoree and co-workers [85] suggested that coupling of MEKC with MS detection could be realized by transferring the analyte zones to a capillary free of SDS by voltage switching. However, this technique requires a home built capillary coupling system. Furthermore, the efficiencies were about two-fold lower than for MEKC/UV, which was attributed to band broadening due to the application of a lower separation voltage and extra band-broadening in the coupling device. Nevertheless, the authors described a fascinating system for on-line heartcutting of sample zones in MEKC. The first technique being more general for rendering a standard MEKC pro-tocol MS compatibility was PF of the PSP (Figure 8.). With this technique only a section of the capillary at the injection side is filled with a PSP. Sub-sequently, the sample is injected and when a voltage is applied over the ca-pillary the sample migrates through the PSP and reaches the detector ahead of the PSP. The analysis is then stopped before the PSP reaches the detector, thereby avoiding introduction of the PSP into the ion source. Drawback with the PF technique includes the required introduction of additional variation with the extra injection of PSP. Furthermore, the electrolyte becomes discon-tinuous resulting in EOF variations between the PSP plug and the surround-ing electrolyte, causing pressure variations and ultimately introduction of laminar flows [86, 87]. Furthermore, the net velocity of the analytes is gen-erally higher outside the PSP plug resulting in further band broadening. Another problem is that in the border between the PSP plug and the sur-rounding electrolyte the concentration of surfactants gradually drops due to diffusion and laminar flow effects ultimately resulting in a micelle gradient.

28

Due to these problems and due to that the PSP plug generally is kept very short to avoid elution of surfactants into the ionization source, the resolution is always lower than for traditional MEKC [88]. Obviously, method devel-opment is therefore more complicated in PF-MEKC than in traditional MEKC.

Figure 8. (A) Schematic illustration of PF-EKC. Step 1: A plug of PSP is injected prior to injection of the sample. Step 2: As the voltage is applied, the analytes start to migrate through the PSP plug thereby becoming separated. Step 3: The analytes reaches the electrolyte phase prior to detection. The separation is stopped when all analytes have been detected and before the PSP reaches the detector. (B) Schematic illustration of RM-EKC. Step 1: The capillary is filled with PSP prior to injection of sample. Step 2: As the voltage is applied, the analytes start to migrate through the PSP and thereby becoming separated. The PSP migrates in the opposite direction into the inlet vial, thereby efficiently avoiding elution of surfactants into the ioniza-tion source. Step 3: The separation is stopped when all analytes have been detected.

In PF, the longest applicable plug length would be realized with a very low EOF allowing for a net reverse migration of the micelles. Thereby the theo-retical plug length would be equal to the total capillary length. Highly nega-tively charged PSPs, with low pH electrolytes or neutral capillary coatings have made it possible to realize such long plug length and the technique has been named RM-EKC (described in Figure 8B) [73, 89]. Due to the very low or absent EOF in RM-EKC, the technique is mainly applicable for positively charged analytes, although some positively charged PSPs have been applied in reverse polarity for separation of negatively charged analytes. However, due to the risk of contaminating the ionization source the whole capillary is rarely filled with PSP and therefore in practice very similar to PF. Despite the development of techniques that enable the use of traditional PSPs in EKC/MS, the counter ions of the PSP may still reach the detector causing ionization suppression. Thus, PSPs designed for use with MS detection and innovations in MS detection are required.

29

6.2. New Pseudostationary Phases with improved Mass Spectrometric compatibility

In the development of new PSPs with improved MS compatibility two main approaches can be identified. On one hand, researchers have strived to re-duce ionization suppression by developing high molecular weight PSPs with a CMC close to or equal to zero [77, 78, 90]. With these PSPs, accumulation of surfactants at the water-air-interface is reduced causing less ionization suppression. Furthermore, electropherograms with less background signals in the low m/z-range are generated and the high molecular weight PSPs are in less extent ionized. Also, method development is simplified as the electro-lyte composition no longer affects the integrity of the PSP. On the other hand, reduced contamination of the ion source has been assessed by the use of semi-volatile surfactants [58, 91]. These two approaches are thus very similar to traditional MEKC. However, several new types of phases, not falling into these categories, have also been developed. Magnetic nanopar-ticles have been applied in a variety of analytical applications, such as in vivo examination via magnetic resonance imaging (MRI) [92] and in chip based systems (microfluidics) [92, 93]. The main characteristics of these particles are that they can be functionalized on the bead surface while still having the embedded magnetic properties, allowing for manipulation of the particle with permanents magnets or electromagnets. Such magnetic beads have been applied in sample preparation and pre-concentration [94, 95]. In 1997 Rashkovetsky et al. [96] performed an enzymatic assay, affinity ad-sorption and isotachophoretic (ITP) focusing using magnetic beads in a CE system. In that application a small amount of magnetic beads, containing immobilized antibodies, was injected and trapped by a magnet in a neutral hydrophilic-coated fused-silica capillary. Subsequently, the sample, contain-ing the antigen, was injected and the antigen was trapped onto the beads. The beads could then be washed and the antigen eluted by a change in the buffer pH. Finally, the separation voltage was switched on and ITP was performed. An interesting application of nanoparticles was presented by Kan and Barron [97] and later by others [98][97][96][96][96]. In their application, poly(N-isopropylacrylamide) (PNIPAAm) modified temperature-sensitive latex beads were trapped in the capillary by an increase in the capillary tempera-ture. Thereby, these particles could be used to trap analytes in the capillary. Lowering the temperature brought the beads back into suspension and the analytes could in that way be eluted from the capillary. These particles have also shown to be useful as coating in CE capillaries [97, 99]. The magnetic beads and the temperature sensitive latexes are not PSPs per se, but, they do

30

fulfill one of the requirements of a PSP, being a non-immobilized interaction phase. Even though these particles have not yet been applied in CEC/MS, they do present an interesting means of combining CEC with a non-trapped stationary phase with MS detection.

6.3. Continuous Full Filling Today, ESI is the dominating ionization technique used in liquid phase sepa-rations coupled with MS detection. As ESI originally was developed for HPLC, the performance of many of the earlier separations published on MEKC/MS were hampered due to the use of ESI interfaces not suited for the capillary format. As the ESI interfaces were miniaturized and adapted to the capillary format significant improvements in MEKC/MS were achieved. Further improvement was obtained with the orthogonal ESI-interfaces. These interfaces, spraying at an angle to the mass spectrometer inlet, had a greatly improved tolerance to non-volatile electrolyte constituents [100]. However, further developments are required to realize the full potential of CE/MS [71, 80, 101]. Viberg [24] elegantly showed that nanoparticles could be used as a PSP in a continuous full filling (CFF) technique using an ortho-gonal ESI interface (Figure 9) in the same manner as in paper II. In CFF the PSP is suspended in the electrolyte similar to traditional MEKC. Thus, the electrolyte is continuous, avoiding the problems resulting from disconti-nuous electrolytes in e.g. PF-EKC. Due to an orthogonal ESI interface the nanoparticles was exclude from entering the MS. The particles could not, due to their larger momentum, deviate from the off-axis spray direction. If these highly concentrated nanoparticles in electrolyte suspensions would be sprayed on-axis several microgram per minute of non-volatile nanoparticles would risk entering the mass spectrometer. Several studies, mainly involving improvements in the nanoparticle based PSP, followed this proof of principle report [25, 102-108]. The later studies, including paper II, showed that CFF with nanoparticles could be performed without any negative effects on the analyte signal, without any contamination of the mass spectrometer and with efficiencies as high as one million plates per meter [25, 106, 107].

31

Figure 9. A schematic picture showing the principal of CFF-CEC/ESI-MS, where the added nanoparticles to the BGE are transported through the open capillary. This will allow for a separation of neutral analytes as well due to the analytes different partitioning to the nanoparticles. The electroosmotic flow is sufficiently high to generate a net transport towards the cathode de-spite any anodic migration of the nanoparticles.

32

The number of publications involving nanoparticles in EKC and related techniques has increased exponentially during the past two decades (Figure 10). Besides their ESI-MS compatibility, these phases also have beneficial mass transfer and a zero CMC. Despite developments of new PSP’s and ESI-interfaces, the full potential of EKC/MS may not yet have been realized. Recent developments in ionization techniques indicate that other ionization sources may be better suited for coupling of EKC with MS detection. Recently, two conceptually different ionization techniques were applied in MEKC/MS. Atmospheric pressure chemical ionization (APCI) and atmospheric pressure photo ionization (APPI) applies electrical dis-charges and photons, respectively, to ionize the analytes in the gas phase. Interestingly, these two techniques were found to be virtually insensitive to the presence of surfactants in the BGE. Analogous to the improvements in ESI with orthogonal spraying, orthogonal APCI was also found superior to on-axis APCI [80]. However, although proven to possess an extensive poten-tial in EKC/MS, ESI-MS is generally still more sensitive [80, 109-111].

Figure 10. The number of published articles with a topic concerning nanoparticles and CE and the sum of PSP-CE related techniques (source ISI Web of knowledge).

33

7. Concluding remarks

”If you can’t explain it simply, you don’t understand it well enough"

Albert Einstein

Since CZE is unable to separate neutral analytes, the use of pseudostationary phases is an excellent way of combining the high efficiency achieved in CE with the selectivity in HPLC. As paper II indicates, the new ESI/MS friend-ly pseudostationary phases could be a good way of solving limitations with complex matrixes using CE-ESI/MS without tedious modification steps or loosing sensitivity due to ion suppression. Despite the different ionization techniques in APPI and ESI the same spray needle is used. Therefore, sever-al of the drawn conclusions could be used in paper II as well. Paper I indi-cated that optimizing the protruding capillary end from the stainless steel sprayer and that the nebulizing gas flow had a large influence on the detec-tion sensitivity. It could also be concluded that the available APPI interface is not fully optimized for CE-MS. However, small modifications and minia-turization of the interface, as well as changing the capillary inner diameter, are some ways of optimize the conditions for CE-APPI/MS. Based on paper I and II, and all the number of publications during the last decade, there are definitely a high interest in CE and CE-related techniques. But more research is needed to fully use their potentials. However, the use of MS friendly pseudostationary phases will lead to higher interest in the tech-nique and hopefully become even more appreciated and used in important field such as clinical applications. The use of nanoparticles in sheathless ESI would be an excellent way of combining ESI friendly pseudostationary phases with the good properties of the sheathless interface, since paper II showed the importance of organic modifier in the BGE when working with nanoparticles in CE. When working with samples with high salt content, nanoparticles in CE-APPI/MS could however be the method of choice.

34

8. Acknowledgments

”As we express our gratitude, we must never forget that the highest appreciation is not to utter words, but to live by them"

John F. Kennedy

There are many I would like to show my appreciation, especially my family and friends for all their love and support! I would like to show my biggest gratitude to my supervisors, Jonas Berg-quist and Charlotta Turner for giving me this opportunity to pursue as a Ph.D. student at department; guiding me into the world of science. Marcus Sjödin, you have given me a great time with many laughs both at and outside work - one can’t ask for a better friend and colleague! This work couldn’t have been possible without Peter Spégel, Jakob Axén, Peter Viberg and all my other excellent collaborators! Thank you for all the scientific discussions and practical help you have given me. Peter Spégel, I would in particular show my appreciations to you and I hope we will have many more projects together in the future. I would also like to give many thanks to Nanosep AB (PerÅke Oldentoft) for giving me this opportunity to work with your nanoparticles and thereby getting into the field of pseu-dostationary phases in CE. Magnus Wetterhall, thank you for the all the help you have given me. I’m looking forward to continue our projects together, and I’m very happy to start having you as a supervisor. Last but not least, a gratitude to all at the department, in particular Jörg Ha-nrieder and Roland Pettersson.

35

9. References

[1] Tiselius, A., Transactions of the Faraday Society 1937, 33, 0524-0530. [2] Hjertén, S., Chromatographic Reviews 1967, 9, 122-219. [3] Jorgenson, J. W., Lukacs, K. D., Anal. Chem. 1981, 53, 1298-1302. [4] Mikkers, F. E. P., Everaerts, F. M., Verheggen, T., Journal of Chromatography 1979, 169, 11-20. [5] Venter, J. C., et al, Science 2001, 291, 1304-+. [6] Lord, G. A., Gordon, D. B., Myers, P., King, B. W., J. Chromatogr. A 1997, 768, 9-16. [7] Allen, D., El Rassi, Z., Electrophoresis 2003, 24, 3962-3976. [8] Hjertén, S., Elenbring, K., Kilár, F., Liao, J.-L., et al., J. Chromatogr. A 1987, 403, 47-61. [9] Svec, F., Electrophoresis 2009, 30, S68-S82. [10] Terabe, S., Otsuka, K., Ichikawa, K., Tsuchiya, A., Ando, T., Anal. Chem. 1984, 56, 111-113. [11] Terabe, S., Trac-Trends Anal. Chem. 1989, 8, 129-134. [12] Terabe, S., Matsubara, N., Ishihama, Y., Okada, Y., Journal of Chromatogra-phy 1992, 608, 23-29. [13] Terabe, S., Otsuka, K., Ando, T., Anal. Chem. 1985, 57, 834-841. [14] Terabe, S., Ozaki, H., Otsuka, K., Ando, T., J. Chromatogr. A 1985, 332, 211-217. [15] Nakagawa, T., Newsl., Div. Colloid Surf., Chem. Soc. Jpn. 1981, 6, 1. [16] Aebersold, R., Mann, M., Nature 2003, 422, 198-207. [17] Olivares, J. A., Nguyen, N. T., Yonker, C. R., Smith, R. D., Anal. Chem. 1987, 59, 1230-1232. [18] Smith, R. D., Olivares, J. A., Nguyen, N. T., Udseth, H. R., Anal. Chem. 1988, 60, 436-441. [19] Mack, L. L., Kralik, P., Rheude, A., Dole, M., The Journal of Chemical Physics 1970, 52, 4977-4986. [20] Dole, M., Mack, L. L., Hines, R. L., Mobley, R. C., et al., The Journal of Chemical Physics 1968, 49, 2240-2249. [21] Whitehouse, C. M., Dreyer, R. N., Yamashita, M., Fenn, J. B., Anal. Chem. 1985, 57, 675-679. [22] Yamashita, M., Fenn, J. B., J. Phys. Chem. 1984, 88, 4451-4459. [23] Takada, Y., Sakairi, M., Koizumi, H., Anal. Chem. 1995, 67, 1474-1476. [24] Viberg, P., Jornten-Karlsson, M., Petersson, P., Spegel, P., Nilsson, S., Anal. Chem. 2002, 74, 4595-4601. [25] Malmström, D., Axen, J., Bergquist, J., Viberg, P., Spegel, P., Electrophoresis 2011, 32, 261-267.

36

[26] Himmelsbach, M., Haunschmidt, M., Buchberger, W., Klampfl, C. W., Anal. Chem. 2007, 79, 1564-1568. [27] Hommerson, P., Khan, A. M., de Jong, G. J., Somsen, G. W., J. Chromatogr. A 2008, 1204, 197-203. [28] Schappler, J., Guillarme, D., Rudaz, S., Veuthey, J. L., Electrophoresis 2008, 29, 11-19. [29] Wu, D., Regnier, F. E., Journal of Chromatography 1992, 608, 349-356. [30] Guttman, A., Cooke, N., Anal. Chem. 1991, 63, 2038-2042. [31] Helmholtz, H., Annalen der Physik 1879, 243, 337-382. [32] Stutz, H., Bordin, G., Rodriguez, A. R., Analytica Chimica Acta 2003, 477, 1-19. [33] Simó, C., Elvira, C., González, N., San Román, J., et al., Electrophoresis 2004, 25, 2056-2064. [34] Puerta, A., Axén, J., Söderberg, L., Bergquist, J., Journal of Chromatography B 2006, 838, 113-121. [35] Rodriguez, I., Li, S. F. Y., Analytica Chimica Acta 1999, 383, 1-26. [36] Horvath, J., Dolník, V., Electrophoresis 2001, 22, 644-655. [37] Kaši�ka, V., Electrophoresis 2003, 24, 4013-4046. [38] Liu, C.-Y., Electrophoresis 2001, 22, 612-628. [39] Gilges, M., Kleemiss, M. H., Schomburg, G., Anal. Chem. 1994, 66, 2038-2046. [40] Barnidge, D. R., Nilsson, S., Markides, K. E., Anal. Chem. 1999, 71, 4115-4118. [41] Barnidge, D. R., Nilsson, S., Markides, K. E., Rapp, H., Hjort, K., Rapid Com-mun. Mass Spectrom. 1999, 13, 994-1002. [42] Selditz, U., Nilsson, S., Barnidge, D., Markides, K. E., Chimia 1999, 53, 506-510. [43] Wetterhall, M., Nilsson, S., Markides, K. E., Bergquist, J., Anal. Chem. 2002, 74, 239-245. [44] Nilsson, S., Wetterhall, M., Bergquist, J., Nyholm, L., Markides, K. E., Rapid Commun. Mass Spectrom. 2001, 15, 1997-2000. [45] Amirkhani, A., Wetterhall, M., Nilsson, S., Danielsson, R., Bergquist, J., J. Chromatogr. A 2004, 1033, 257-266. [46] Nilsson, S., Svedberg, M., Pettersson, J., Bjorefors, F., et al., Anal. Chem. 2001, 73, 4607-4616. [47] Terabe, S., Procedia Chemistry 2010, 2, 2-8. [48] Chervet, J. P., Ursem, M., Salzmann, J. B., Anal. Chem. 1996, 68, 1507-1512. [49] Ishii, D., Asai, K., Hibi, K., Jonokuchi, T., Nagaya, M., J. Chromatogr. A 1977, 144, 157-168. [50] Swartz, M. E., J. Liq. Chromatogr. Relat. Technol. 2005, 28, 1253-1263. [51] Somsen, G. W., Mol, R., de Jong, G. J., J. Chromatogr. A 2010, 1217, 3978-3991. [52] Ozaki, H., Itou, N., Terabe, S., Takada, Y., et al., J. Chromatogr. A 1995, 716, 69-79. [53] Ozaki, H., Terabe, S., Ichihara, A., J. Chromatogr. A 1994, 680, 117-123. [54] Poole, S. K., Poole, C. F., J. Chromatogr. A 2008, 1182, 1-24. [55] Poole, K. S., Poole, F. C., Analyst 1997, 122, 267-274.

37

[56] Terabe, S., J. Pharm. Biomed. Anal. 1992, 10, 705-715. [57] Rundlett, K. L., Armstrong, D. W., Anal. Chem. 1996, 68, 3493-3497. [58] Ishihama, Y., Katayama, H., Asakawa, N., Analytical Biochemistry 2000, 287, 45-54. [59] Nelson, W. M., Tang, Q., Harrata, A. K., Lee, C. S., J. Chromatogr. A 1996, 749, 219-226. [60] Domon, B., Aebersold, R., Science 2006, 312, 212-217. [61] Kasicka, V., Electrophoresis 2010, 31, 122-146. [62] Nesbitt, C. A., Zhang, H. X., Yeung, K. K. C., Analytica Chimica Acta 2008, 627, 3-24. [63] Hoffmann, E. d., Stroobant, V., Mass Spectrometry: Principles and Applica-tions, John Wiley & Sons, Ltd, West Sussex, England 2002. [64] Skoog, Holler, Crouch, Principles of Instrumental Analysis, David Harris, Bel-mont, USA 2007. [65] Smith, R. D., Loo, J. A., Edmonds, C. G., Barinaga, C. J., Udseth, H. R., Anal. Chem. 1990, 62, 882-899. [66] Ekman, R., Silberring, J., Westman-Brinkmalm, A., Kraj, A., Mass Spectrome-try: Instrumentation, Interpretation, and Applications, John Wiley & Sons, Inc., New Jersey, USA 2008. [67] Desiderio, C., Rossetti, D. V., Iavarone, F., Messana, I., Castagnola, M., Jour-nal of Pharmaceutical and Biomedical Analysis 2010, 53, 1161-1169. [68] Ramautar, R., Mayboroda, O. A., Somsen, G. W., de Jong, G. J., Electrophore-sis, 32, 52-65. [69] Lopez, M. F., Mikulskis, A., Kuzdzal, S., Bennett, D. A., et al., Clin. Chem. 2005, 51, 1946-1954. [70] Robb, D. B., Covey, T. R., Bruins, A. P., Anal. Chem. 2000, 72, 3653-3659. [71] Axen, J., Malmstrom, D., Axelsson, B. O., Petersson, P., Sjoberg, P. J. R., Rap-id Commun. Mass Spectrom. 2010, 24, 1260-1264. [72] Raffaelli, A., Saba, A., Mass Spectrometry Reviews 2003, 22, 318-331. [73] Liu, Z., Sam, P., Sirimanne, S. R., McClure, P. C., et al., J. Chromatogr. A 1994, 673, 125-132. [74] Quirino, J. P., Terabe, S., Science 1998, 282, 465-468. [75] Quirino, J. P., Terabe, S., Anal. Chem. 2000, 72, 1023-1030. [76] Hong, C.-Y., Chen, Y.-C., J. Chromatogr. A 2007, 1159, 250-255. [77] Shamsi, S. A., Anal. Chem. 2001, 73, 5103-5108. [78] Wang, B., He, J., Shamsi, S. A., Journal of Chromatographic Science 2010, 48, 572-583. [79] Kirby, D. P., Thorne, J. M., G��tzinger, W. K., Karger, B. L., Analytical Che-mistry 1996, 68, 4451-4457. [80] Hommerson, P., Khan, A. M., de Jong, G. J., Somsen, G. W., J. Am. Soc. Mass Spectrom. 2009, 20, 1311-1318. [81] Hommerson, P., Khan, A. M., de Jong, G. J., Somsen, G. W., Electrophoresis 2007, 28, 1444-1453. [82] Pleasance, S., Thibault, P., Kelly, J., Journal of Chromatography 1992, 591, 325-339. [83] Wachs, T., Sheppard, R. L., Henion, J., J. Chromatogr. B-Biomed. Appl. 1996, 685, 335-342.

38

[84] Samskog, J., Wetterhall, M., Jacobsson, S., Markides, K., J. Mass Spectrom. 2000, 35, 919-924. [85] Lamoree, M. H., Tjaden, U. R., van der Greef, J., J. Chromatogr. A 1995, 712, 219-225. [86] Nelson, W. M., Lee, C. S., Anal. Chem. 1996, 68, 3265-3269. [87] Amini, A., Paulsen-Sorman, U., Westerlund, D., Chromatographia 1999, 50, 497-506. [88] Muijselaar, P. G., Otsuka, K., Terabe, S., J. Chromatogr. A 1998, 802, 3-15. [89] Quirino, J. P., Terabe, S., Anal. Chem. 1998, 70, 149-157. [90] Hou, J., Zheng, J., Shamsi, S. A., J. Chromatogr. A 2007, 1159, 208-216. [91] Petersson, P., Jörntén-Karlsson, M., Stålebro, M., Electrophoresis 2003, 24, 999-1007. [92] Gijs, M. A. M., Microfluidics and Nanofluidics 2004, 1, 22-40. [93] Verpoorte, E., Lab Chip 2003, 3, 60N-68N. [94] Sui, W. G., Huang, L. L., Dai, Y., Chen, J. J., et al., Clin. Exper. Med. 2010, 10, 259-268. [95] Tomlinson, A. J., Guzman, N. A., Naylor, S., J. Capillary Electrophor. 1995, 2, 247-266. [96] Rashkovetsky, L. G., Lyubarskaya, Y. V., Foret, F., Hughes, D. E., Karger, B. L., J. Chromatogr. A 1997, 781, 197-204. [97] Kan, C. W., Barron, A. E., Electrophoresis 2003, 24, 55-62. [98] Malmstadt, N., Yager, P., Hoffman, A. S., Stayton, P. S., Anal. Chem. 2003, 75, 2943-2949. [99] Buchholz, B. A., Doherty, E. A. S., Albarghouthi, M. N., Bogdan, F. M., et al., Anal. Chem. 2000, 73, 157-164. [100] Serwe, M., Ross, G., Chromatographia 1999, 49, S73-S77. [101] Axen, J., Axelsson, B. O., Jornten-Karlsson, M., Petersson, P., Sjoberg, P. J. R., Electrophoresis 2007, 28, 3207-3213. [102] Nilsson, C., Birnbaum, S., Nilsson, S., J. Chromatogr. A 2007, 1168, 212-224. [103] Nilsson, C., Nilsson, S., Electrophoresis 2006, 27, 76-83. [104] Nilsson, C., Viberg, P., Spegel, P., Jornten-Karlsson, M., et al., Anal. Chem. 2006, 78, 6088-6095. [105] Spegel, P., Schweitz, L., Nilsson, S., Anal. Chem. 2003, 75, 6608-6613. [106] Spegel, P., Viberg, P., Carlstedt, J., Petersson, P., Jornten-Karlsson, M., J. Chromatogr. A 2007, 1154, 379-385. [107] Viberg, P., Spegel, P., Carlstedt, J., Jornten-Karlsson, M., Petersson, P., J. Chromatogr. A 2007, 1154, 386-389. [108] Viberg, P., Spegel, P., Nilsson, J., Petersson, P., et al., Chromatographia 2007, 65, 291-297. [109] Keski-Hynnila, H., Kurkela, M., Elovaara, E., Antonio, L., et al., Anal. Chem. 2002, 74, 3449-3457. [110] Thurman, E. M., Ferrer, I., Barcelo, D., Anal. Chem. 2001, 73, 5441-5449. [111] Hommerson, P., Khan, A. M., Bristow, T., Harrison, M. W., et al., Rapid Commun. Mass Spectrom. 2009, 23, 2878-2884.