Bioanalytical separation using capillary

electrophoresis

Applications with microbubbles and proteins

Leila Josefsson

Licentiate thesis

KTH Royal Institute of Technology

School of Chemical Science and Engineering

Department of Chemistry

Stockholm 2017

Akademisk avhandling

Akademisk avhandling som med tillstånd av Kungliga Tekniska högskolan framlägges till

offentlig granskning för avläggande av teknologie licentiatexamen torsdag 23 november 2017,

kl. 10:00 i sal K51, Teknikringen 28, plan 5, Stockholm. Avhandlingen presenteras på engelska.

ii

Bioanalytical separation using capillary electrophoresis - applications with microbubbles

and proteins

Leila Josefsson

Licentiate Thesis

KTH Royal Institute of Technology

School of Chemical Science and Engineering

Department of Chemistry

Division of Applied Physical Chemistry, Analytical Chemistry

SE-100 44 Stockholm, Sweden

TRITA-CHE Report 2017:32

ISSN 1654-1081

ISBN 978-91-7729-560-0

Copyright©

Leila Josefsson, 2017

Printed by Universitetsservice US-AB, Stockholm, 2017

iii

Abstract

In this thesis the possibilities of using capillary electrophoresis as a separation

technique for analysis of proteins and microbubbles is presented.

A complete analytical process consists of five necessary steps of which one is the actual

analysis step. For this step a suitable analytical technique is needed. Capillary

electrophoresis (CE) is one of the common analytical separation techniques used for

analysis of a diversity of analytes, and can be both used in routine analysis and for research

purposes. The reason for using CE, compared to other liquid-based separation techniques, is

mainly short analysis time, high resolution, and negligible sample volumes and solvent

waste. Depending on the characteristics of the analytes, and the sample matrix, different

modes of CE can be used, where capillary zone electrophoresis (CZE) is the most employed

one. The basic principle of CZE is separation of the analytes due to differences in total

mobility, which is dependent on the charge and size of the analytes, and the electroosmotic

flow (EOF). The EOF can be controlled by several parameters e.g. choice of background

electrolyte (BGE), and the optimization of the parameters has been discussed throughout the

thesis.

To improve the properties of the BGE, an ethylammonium nitrate (EAN) water solution

was used as BGE for CE analysis in Paper I. The precision of the EOF with this method

was determined by adjusting the pH of the BGE, the concentration of EAN in the BGE, and

the electric field. Model proteins were thereafter analysed using the optimal parameters

yielding a precision sufficient for routine control.

One example of the applications of CE is separation of novel contrast agents, which consist

of polyvinyl alcohol microbubbles (PVA-MBs). In Paper II, a method for analysis of PVA-

MBs in biological samples using CE with UV-detection was developed. It was also

established that intact PVA-MBs could be distinguished from ultrasound degraded PVA-

MBs in the same set-up.

Keywords: Background electrolyte, capillary electrophoresis, contrast agents,

electroosmotic flow, ethylammonium nitrate, ionic liquids, microbubbles, particles,

polyvinyl alcohol microbubbles, and separation.

iv

Sammanfattning

I denna avhandling presenteras möjligheterna att använda kapillärelektrofores som

en separationsteknik för analys av proteiner och mikrobubblor.

En fullständig analytisk process består av fem nödvändiga steg varav en är själva

analyssteget. I detta steg behövs en passande analytisk teknik. Kapillärelektrofores (CE) är

en av de vanliga separationsmetoderna som används vid analys av en mångfald analyter och

används både vid rutinanalys och för forskningsändamål. Anledningen till att använda CE

jämfört med andra separationstekniker är huvudsakligen kort analystid, hög upplösning och

försumbar provvolym och lösningsmedelsavfall. Beroende på egenskaper hos analyterna

och provmatrisen kan olika typer av CE användas, där kapillärzonelektrofores (CZE) är den

mest använda. Principen för CZE är separation av analyterna på grund av skillnader i den

totala mobiliteten, vilken är beroende av laddningen och storleken på analyterna samt det

elektroosmotiska flödet (EOF). EOF kan styras genom flera parametrar, till exempel valet

av bakgrundselektrolyt (BGE), och optimeringen av dessa parametrar har diskuterats genom

hela avhandlingen.

För att förbättra egenskaperna av backgrundselektrolyten (BGE), ett etylammoniumnitrat

(EAN)-vattenlösning användes som BGE för CE analys i Artikel I. Precisionen för EOF

med denna metod bestämdes genom att justera pH-värdet för BGE, koncentrationen av

EAN i BGE och det elektriska fältet. Modellproteiner analyserades därefter med

användning av optimala parametrar vilket gav en tillräcklig precision för rutinanalys.

En av applikationerna för kapillärelektrofores är separation av nya kontrastmedel som

består av polyvinylalkoholmikrobubblor (PVA-MBs). I Artikel II utvecklades en metod för

analys av PVA-MB i biologiska prover med användning av kapillärelektrofores och UV-

detektion. Det klargjordes också huruvida intakta PVA-MBs kunde särskiljas från

ultraljudsnedbrutna MBs i samma uppställning.

v

Contents

Abstract.................................................................................................................................. iii

Sammanfattning ..................................................................................................................... iv

List of publications ................................................................................................................. 1

1 Abbreviations and symbols ................................................................................................. 2

2 Analytical chemistry ............................................................................................................ 4

3 Capillary electrophoresis ..................................................................................................... 5

4 Theory of electrophoresis .................................................................................................... 7

4.1 Electroosmotic flow .......................................................................................................... 8

4.2 Parameters effecting electroosmotic flow .................................................................... 9

5 Background electrolytes used in capillary electrophoresis ................................................ 10

5.1 Buffers ........................................................................................................................ 10

5.2 Ionic liquids ................................................................................................................ 11

5.2.1 Electroosmotic mobility dependency on the pH .................................................. 12

5.2.2 Electroosmotic mobility dependency on concentration ....................................... 13

5.2.3 Separation of proteins using ionic liquids ............................................................ 15

6 Applications in capillary electrophoresis .......................................................................... 16

6.1 Particles ...................................................................................................................... 16

6.2 Polyvinyl alcohol microbubbles ................................................................................. 16

6.2.1 Ultrasonicated polyvinyl alcohol microbubbles .................................................. 20

6.2.2 Polyvinyl alcohol microbubbles in human blood ................................................ 22

6.2.3 Modified polyvinyl alcohol microbubbles ........................................................... 24

6.2.4 Camera detection ................................................................................................. 26

7 Conclusions and future perspectives ................................................................................. 27

8 Acknowledgements ........................................................................................................... 28

9 References ......................................................................................................................... 29

1

List of publications

This thesis is based on the following publications:

Paper I Evaluation of ethylammonium nitrate as background electrolyte in

capillary electrophoresis

Leila Josefsson, Jessica Bernsteen, Åsa Emmer, Manuscript

Paper II Analysis of polyvinyl alcohol microbubbles in human blood plasma using

capillary electrophoresis

Leila Josefsson, Malin K. Larsson, Anna Bjällmark, Åsa Emmer,

Journal of Separation Science 39 (8), DOI: 10.1002/jssc.201501342.1

Reprints are published with kind permission of the journals. The contributions of the author to these papers are:

Paper I Major parts of the experiment and planning, evaluation of the data, and parts of the writing.

Paper II All experiments and major part of the writing

Parts of this work have been presented as poster and oral presentations at the following conferences: Analysis of polyvinyl alcohol microbubbles in human blood plasma using CE

Leila Josefsson, 32nd International Symposium on Microscale Separations and Bioanalysis, MSB 2016, Niagara-on-the-lake, Ontario, Canada. Analysis of polyvinyl alcohol microbubbles plasma using capillary electrophoresis

Leila Josefsson, Analysdagarna 2016, Umeå, Sweden. Analysis of polyvinyl alcohol microbubbles using different detectors

Leila Josefsson, 33rd International Symposium on Microscale Separations and Bioanalysis, MSB 2017, Nordwijkerhout, Netherlands. Analysis of polyvinyl alcohol microbubbles using different detectors

Leila Josefsson, XIX Euroanalysis 2017, Stockholm, Sweden.

1 Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission

2

1 Abbreviations and symbols

BGE Background electrolyte

CA Contrast agents

CCD Charged coupled device

CE Capillary electrophoresis

CGE Capillary gel electrophoresis

CIEF Capillary isoelectric focusing

CITP Capillary isotachophoresis

CMEK Capillary micellar electrokinetic chromatography

CZE Capillary zone electrophoresis

E Electric field

EAN Ethylammonium nitrate

EOF Electroosmotic flow

ε Relative permittivity

η Viscosity

Fe Electric force

Ff Frictional force

Fe-PVA-MBs Polyvinyl alcohol microbubbles with superparamagnetic iron

oxide nanoparticles imbedded in the shell

HV High voltage

IL Ionic liquid

MB Microbubbles

MRI Magnetic resonance imaging

MS Mass spectrometry

μe Electrophoretic mobility

μeo Electroosmotic mobility

OPG Osteoprotegerin

3

pI Isoelectric point

PVA Polyvinyl alcohol

PVA-MBs Polyvinyl alcohol microbubbles

q Charge of an ion

r Stokes radius

RSD Relative standard deviation

RTIL Room temperature ionic liquids

SEM Scanning electron microscope

SPIONs superparamagnetic iron oxide nanoparticle

TRIS Trishydroxymethylamino methane

UV Ultraviolet light

v Velocity

vEOF Velocity of the EOF

VIS Visible light

ζ Zeta potential

4

2 Analytical chemistry

Analytical chemistry is a branch of chemistry, which seeks to find answers to questions as

“Is this compound present in the mixture?”, “In what concentration is it present?”, “Does

the concentration of this compound exceed the regulation limit?” or simply “Is this

compound pure?” Analytical chemistry is therefore common in most fields of chemistry e.g.

organic chemistry, physical chemistry, and polymer chemistry.

The analytical process includes five steps, where the first step is to formulate the general

questions stipulated above into more specific questions that can be answered through

chemical measurements. The next step in the progress is to investigate whether there is

already an existing analytical procedure that fulfils the requirements, or if the analytical

procedure needs to be developed. In both Paper I and II, the analytical procedure was

developed and optimized of for the intended purpose.

The next step is to collect the samples, and how the sampling is performed has a crucial

role, since poorly selected samples could give meaningless results. The sample should be

selected so that it is representative and so that it is stable throughout the rest of the analysis.

In many cases the sample needs to be treated with different preparation techniques so that a

suitable form of the sample is obtained e.g. dissolution of a powder, pre-concentration of

the sample, or extraction to the organic or water phase, respectively. [1] In Paper II, the

samples are collected in vacuum tubes and the analytes are extracted together with blood

plasma.

When the samples are prepared it is time to perform the analysis, and for that analytical

instruments are needed. According to the IUPAC definition, an analytical instrument is “A

device or a combination of devices used to carry out an analytical process. The analytical

process is all or part of the analytical procedure that encompasses all steps from the

introduction of the sample, or the test portion to the production of the result.” [2] There is a

diversity of analytical instruments depending on the physical and chemical characteristics of

interest. However, it is important to choose the analytical technique wisely so that high

precision is achieved, and so that the choice of analytical technique does not bias the result.

The most common analytical techniques that are used are a variety of electroanalytical

techniques, spectrophotometry, atomic- and mass spectrometry (MS), several types of

chromatography, and capillary electrophoresis (CE). In both Paper I and II, CE is used as

the analytical technique, and the methods are optimized differently to fit the applications.

The last step in the analytical workflow is the data collection, interpretation, and drawing

conclusions. It is important to understand the data so that right conclusions can be drawn.

This is also the step where the specific questions formulated at the beginning are answered,

and completing the whole analytical process.

The general aim of the work in the present thesis was to implement the complete analytical

process suitable for the specific analytical questions presented in Paper I and II. Here, the

focus is on the principles of CE as the analytical technique, and how to optimize the

methods depending on the sample.

5

3 Capillary electrophoresis

CE is a powerful separation technique, which has gained increased popularity in routine

analysis, due to the possibility of analysing a large variety of substances from ions to

complex proteins [3]. CE was first developed by Arne Tiselius in 1937 [4], and he received

the Nobel Prize 1948 for his work on electrophoresis and adsorption analysis. Tiselius’

experiments were performed in free solution, which is limited by convection and thermal

diffusion, and to evade this obstacle, gels where used instead. Gel electrophoresis has

mainly been used for separation of biological macromolecules, e.g. proteins and nucleic

acids [5].

Another method to avoid the obstacles with free solution electrophoresis was to use open

tube electrophoresis, which was first initialized in 1967 by Hjertén [6]. This was performed

in narrow capillaries, which were millimetre-sized at the time. As the dimension of the

capillaries decreased the popularity of the technique increased. Virtanen was the first to

demonstrate the advantages of using 200 µm glass capillaries and Mikkers performed the

electrophoresis in Teflon tubes [7]. It was not until the 1980s, when the fused silica

capillaries reached an internal diameter of <100 µm, when Jorgenson and Lukacs showed

that CE could be used for highly efficient separations of amino acids [8]. They also briefly

described the relationship between electro-osmosis and the mobility of ions. After this

point, introduction of commercial CE instrument moved the technique from research labs

into more practical applications.

Nowadays, there are many advantages obtained with CE, compared to other liquid-based

separation techniques such as liquid chromatography, mainly short analysis time with high

resolution, negligible waste, possibility of automatization, and minimal sample

requirement [9]. One reason for the extensive use of CE is the versatility of the technique,

where the user can choose between different modes of operation, depending on the

application. The main modes are:

∙ capillary zone electrophoresis (CZE),

∙ capillary isotachophoresis (CITP),

∙ capillary gel electrophoresis (CGE),

∙ capillary isoelectric focusing (CIEF), and

∙ capillary micellar electrokinetic chromatography (CMEK).

6

In both Paper I and II in this thesis, the CE mode CZE is utilized and the basic schematic

of a CZE system is seen in Figure 1. Components that compose the CZE system in its very

basic arrangement are a high voltage (HV) supply, buffer reservoirs, capillary, and a

detector. The electrodes from the HV supply are submerged in the buffer reservoirs,

together with the buffer filled fused silica capillary, which closes the circuit. The analytes

are introduced to the capillary by either hydrodynamic or electrokinetic injection, and the

injections are done by replacing the inlet buffer vial with a sample vial, and applying a

pressure difference or voltage over the capillary. The separation of the analytes occurs

inside the capillary, and the analytes are detected when passing the detector. In this thesis

the detection was mainly performed with a ultraviolet light (UV) -detector, where an

additional camera (not shown in figure) has been used for tracking of specific analytes.

Figure 1 - Schematic illustration of a capillary electrophoretic system with its components; high

voltage (HV) supply, buffer reservoirs, sample vial, separation capillary, and a UV-detector.

7

4 Theory of electrophoresis

CZE is the most popular CE mode of the previously mentioned modes. CZE is the

separation of analytes based on their differences in mobilities in the electrophoretic system.

There are two forces acting on an ion when an electric field is applied over the system. One

of the forces is the electric force (Fe), equation 1, and the other one is the frictional

force (Ff), equation 2, resulting from movement of the ion. [10]

Fe = Eq (1)

Ff = −6πηrv (2)

Where E is the applied external electric field, q is the charge of the ion, η is the viscosity of

the medium, r is the Stokes radius, and v is the velocity of the ion. At steady-state the sum

of the external forces must be zero, the velocity is therefore given by equation 3.

v =q

6πηr E (3)

An ion´s total velocity in CZE is dependent on the electrophoretic mobility and also the

applied electric field, described in equation 4.

vion = μe E 4)

Hence, the electrophoretic mobility of a charged analyte will be dependent on the charge of

the ion, viscosity of the buffer and the Stokes radius of the analyte, seen in equation 5.

μe =q

6πηr (5)

This leads to that those ions that are small and multiply charged will migrate faster in the

CZE system compared to ions which are larger and singly charged.

8

4.1 Electroosmotic flow

The previous section only focused on one ion, however in a CZE system during separation

there will be more than just one ion, and in this section the forces affecting the ions in the

buffer will be described. Above pH 2-3 of the buffer, the fused silica capillary wall will be

negatively charged because of deprotonation of silanol groups at the surface, leading to an

attraction between the wall and positively charged ions. As the pH of the buffer increases

the dissociation of the silanol groups of the wall increases until the surface is fully

dissociated. Closest to the wall there will be a tightly bound layer of positively charged

ions, called the Stern layer, and moving further away from the wall the attraction decreases

and the ions creates a more diffuse layer, see Figure 2. The differences in charge between

the capillary surface and the bulk solution will lead to a zeta-potential over the cross section

of the capillary. [11]

Figure 2 – To the left: plug shaped flow profile of the electroosmotic flow. To the right: the

negatively charged capillary wall attracts positively charged ions. The Stern layer is composed

of strongly bound ions, while the loosely bound ions create the diffuse layer.

When a voltage is applied over the capillary, the positively charged ions in the diffuse layer

will migrate towards the electrode. This movement creates a flow of buffer solution from

the anode to the cathode, called the electroosmotic flow (EOF). The EOF has a plug flow

shaped profile, see Figure 2, compared to other separation techniques, which contributes to

the increased separation efficiency of CZE.

The velocity of the EOF can be calculated using equation 6.

𝑣𝐸𝑂𝐹 = μ𝑒𝑜𝐸 (6)

The electroosmotic mobility (µeo) is defined by equation 7, where ε is the relative

permittivity of the buffer, and ζ is the zeta potential of the capillary wall.

μ𝑒𝑜 =𝜀𝜁

𝜂 (7)

9

4.2 Parameters effecting electroosmotic flow

It is essential to regulate the EOF in order to achieve good separation between analytes, and

there is several ways of controlling the EOF by either changing the operational parameters,

or the background electrolyte (BGE). Operational parameters include changing the electric

field, temperature, or by coating the capillary wall with either dynamic or permanent

coatings. The BGE can be adjusted according to the pH of the solution, ionic strength by

altering the buffer concentration, the chosen buffer-ion pair, and also the amount of added

organic phase. The parameters of the BGE either change the viscosity of the BGE, the zeta

potential of the capillary wall or both, which according to equation 7 will affect the EOF.

With dynamic and permanent coatings of the capillary wall, the EOF can either be reduced

or reversed. To reverse the EOF, positively charged surfactants or polyamines are

commonly utilized [11]. The positively charged head of the surfactant or the polyamine will

attach to the negatively charged silanol surface. For the surfactants, the hydrophobic tail

interactions will create a second layer, called a double layer, with the positively charged

surfactant head towards the bulk solution, forming a positively charged surface.

The previously mentioned parameters are primarily factors, which will affect the EOF

directly as they are changed. However, there are also secondary factors affecting the EOF,

e.g. the self-heating of the capillary arising from the applied voltage over the capillary and

movement of ions inside it [12]. This self-heating, known as Joule heating, could increase

the temperature of the capillary and change the viscosity of the buffer [10]. If the cooling of

the capillary is insufficient, a loss in resolution, efficiency and reproducibility of the

separation, and potentially loss of injected sample can be the result [13].

In Paper I, the EOF was controlled with several parameters. To begin with, the chosen

BGE consisted of the ionic liquid (IL) ethylammonium nitrate (EAN), which was diluted in

water. EAN interacts with the capillary wall and affects the EOF. Thereafter, the pH,

concentration of EAN and the electric field was varied, and EOF investigated. The effect of

ILs on EOF will be discussed in detail in sections 5.2.1 and 5.2.2.

In Paper II, the EOF was optimized by adjusting the pH of a 12.5 mM phosphate buffer

between 7 and 12, until separation of polyvinyl alcohol microbubbles (PVA-MBs) was

achieved. The results of that optimization are shown in section 6.2.

10

5 Background electrolytes used in capillary

electrophoresis

As seen in the previous section, the BGE has a profound impact on the EOF, and it is an

integral part of electrophoresis. The choice of BGE will affect the separation, since the

separation of the analytes occurs within the BGE inside the capillary. Depending on the

characteristics and additives added to the BGE the separation can occur according to one of

the five main CE modes. For CZE separation the BGE is often a buffer salt and can contain

additives in form of organic solvents. In CITP, the BGE is chosen so that a leading and

terminating electrolyte is achieved, where the leading electrolyte has anions/cations with

highest mobility, while the terminating electrolyte has the lowest anion/cation mobility. The

analytes will be separated, within a sandwich of the two buffer systems, regarding the

mobility. To separate analytes according to their isoelectric point (pI), a buffer system with

ampholytes that have pKa values within the desired pH range is used for the CIEF mode. If

a gel or a polymeric network at low concentration >10% is included in the BGE, a CGE

separation mode is obtained. To perform CMEK, a surfactant is added to the BGE with a

concentration above the critical micellar concentration.

5.1 Buffers

Commonly, BGEs are buffers used to control the pH in the CE, the generated temperature,

the charge of the analytes and capillary wall, and effects on detection. The preparation of

the buffer is extremely important, since the smallest variations in pH affects the EOF,

analyte charge, and the temperature generated by the changes in current. Those small

variations in pH could therefore have huge impacts on the CE separation. The buffer

capacity of a buffer is dependent on the concentration of the buffer salts and the pKa value,

where a higher concentration yields a higher buffer capacity. The effective buffer capacity

is considered to be within 1 pH unit of the selected buffers pKa. The most commonly used

buffers in CZE are phosphate, borate, citrate, acetate and trishydroxymethylamino methane

(TRIS). [11] A high buffer capacity is desirable, since the amount of acid or base that can

be added before a change in pH occurs is large. This is especially important due to the

electrolysis at the electrodes in the CE system. However, a high concentration of the buffer

could contribute to higher currents in the CE resulting in high joule heating. Therefore,

there is a trade-off between how high buffer capacity the selected buffer should have contra

the time required to replenish the buffer vials after a specific number of runs.

11

5.2 Ionic liquids

A more recent BGE used for CE is ionic liquids (ILs). ILs are defined as salts with a

melting temperature below 100°C, and consists of most often an organic cation and organic

or inorganic anion [14, 15]. When the compounds are liquid at or below room temperature

they belong to a subgroup called room temperature ionic liquids (RTIL). The

physicochemical properties of an IL could be tuned by the choice of anion and cation,

which makes the IL a designer solvent. The most common cations and anions used to

compose ILs are shown in Figure 3. [16]

Figure 3 - Structures of main cations and anions of IL used in CE separations. 1. Ammonium,

2. Phosphonium, 3. Pyrrolidinium, 4. Imidazolium, 5. Pyridinium, 6. Piperidinium, a. Chloride,

b. Tetrafluoroborate, c. Nitrate, d. Bromide, e. Hexafluorophosphate, f. Triflate,

g. Bis(trifluoromethylsulfonyl)imide.

Generally, ILs have low vapour-pressure, making them stable in vacuum and also

minimizing the evaporation, good long-term stability, and high electrical

conductivity [14, 17]. The first ionic liquids were reported in the late 19th and early 20th

century, but it was not until the 1980s that the ionic liquids received wider attention

amongst the scientists [15]. The reason for that was the introduction of ILs with 1-ethyl-3-

methylimidazolium as the cation by Wilkes et al. [18, 19], since those ILs were both water

and air stable. This opened the field for a number of different analytical applications like

sample preparation, chromatography, CE and MS [14, 15].

When using ILs in analytical chemistry, different physical properties are desired depending

on the application. In sample preparation, the solvating properties of the chosen IL should

be good, and if the IL also has low vapour pressure it is considered being a green solvent.

Whereas to use IL as stationary phases in gas chromatography, properties like low

flammability, polarity and negligible volatility are more appreciated. Moreover, to separate

analytes with the same behaviour as commercial polar and non-polar stationary phases

shows, the cation of the IL can be modified by changing the functional groups to either be

12

polar or non-polar. This ability makes ILs extra interesting for chromatographic

analysis [20, 21]. Except for being used as stationary phases in liquid chromatography, the

ILs can be used as additives in the mobile phase. At low concentrations of ILs in the mobile

phase, the silica surface of the carrier particles in a LC column are shielded, leading to

better peak shapes and less peak broadening [22].

In CE, the ILs has either been added to the BGE to improve its properties, or diluted with

water to a certain concentration and used as BGE without extra addition of buffer salts [22].

One reason for choosing ILs before other electrolytes is that the cations in an IL possess

two functions, where one is to modify the wall and the other is to interact with the

analytes [23]. The cations of the IL have the ability of binding dynamically to the capillary

wall, which may change the EOF and shield the negatively charged silanol groups from the

analytes, preventing them from binding to the capillary wall. This could especially be used

when analysing proteins, which otherwise are prone to adsorb to the capillary wall and get

retained or remain in the capillary. [24] One of the advantages of using ILs is that the

dynamic coating at the capillary wall can easily be removed, and the capillary can be

regenerated and used further.

Another reason why ILs has gained popularity in CE is the lower currents obtained when

using ILs-water mixtures as BGEs compared to traditional buffers. The lower the current in

the CE system, the lower Joule heating is generated, thereby less peak broadening and more

reproducible results are obtained [25, 26]. This reduction in current was also observed in

Paper I, where the IL EAN produced a current of 12 µA at a pH of 10 compared to 21 µA

when a potassium phosphate buffer was used under the same conditions.

In Paper I, the electroosmotic mobility (µeo), which is one measurement of the EOF, was

adjusted with three parameters. The parameters were pH and concentration of IL in the

BGE (also related to the ionic strength). The electric field was also altered to observe the

variations in the velocity of the EOF.

5.2.1 Electroosmotic mobility dependency on the pH

Generally, the µeo has a sigmoid relationship with pH when using fused silica capillaries

without coatings [27]. This means that at low pH, the silanol groups are protonated, which

results in the EOF being almost zero. At high pH, it is the other way around and the silanol

groups are fully protonated resulting in a high plateau value for the EOF. The largest

variation in µeo is in the pH range of 4-6 for an uncoated fused silica capillary with an ionic

strength of 0.01M. [28]

13

To investigate the µeo dependency on the pH for EAN, a 20 mM EAN-water solution was

used as the BGE in the CE-system, and the pH was adjusted to different selected values

with sodium hydroxide, seen in Figure 4 (Paper I). The unadjusted EAN solution has a pH

of roughly 5.6 at 20 °C, which also was the starting point for the analysis. The other

investigated pH values were 8, 10 and 12. The µeo was evaluated to establish the

reproducibility with at least 19 runs for each condition, performed at different occasions.

Here, the plateau value of µeo at alkaline conditions was approximately 6.2 cm2/sV, and at

pH below 8 the plateau reached a value lower than 1.5 cm2/sV. The results indicate that the

sigmoid curve has shifted towards higher pH compared to an uncoated fused silica

capillary.

Figure 4 – The electroosmotic mobility in the CE system at 20 kV, with an ethylammonium

nitrate concentration of 20 mM and varying pH. The chosen pH values were unadjusted EAN

(approximately 5.6), 8, 10 and 12. The confidence interval is at 95% (n≥19). Samples were

electrokinetically injected and the capillary had a total length of 80 cm. (Figure 1, Paper I)

5.2.2 Electroosmotic mobility dependency on concentration

From previous result, the highest precision was achieved when a higher pH was used, and to

test the µeo dependency on the ionic strength, the experiments where conducted with a pH of

10 for further studies. The ionic strength was changed by increasing the concentration of the

EAN in the BGE from 20 to 80 mM. The electric field was also altered by applying

voltages between 15 and 30 kV to the CE system to check the change in EOF.

14

Changing the ionic strength of a buffer could result in a change in both the viscosity of the

medium and the zeta potential at the capillary wall. The impact on the zeta potential is

higher than the viscosity, since the viscosity is relatively constant when working with so

low concentrations of the EAN, though. According to theory, when increasing the ionic

strength, more cations will shield the negatively charged capillary wall resulting in a lower

zeta potential between the wall and the bulk solution. A low zeta potential will lead to lower

µeo, as can be seen in equation 7. In Figure 5 the µeo is lowered when the concentration is

changed from 20 to 40 mM of EAN. However, the mobility should thus decrease further

when going from 40 to 80 mM in concentration, which is not the case. One explanation

could be that, a large portion of the silanol groups at the surface of the capillary wall are

already interacting with ethylammonium ions when using 40 mM of EAN in the BGE, and

additional EAN will not affect the EOF, since the alkyl chain is too short to create a double

layer [29].

In theory, the µeo is independent on the electric field in the CE system. Still, as shown in

Figure 5, the µeo was slightly increasing when higher voltages were applied. The increase

was roughly 2% from 15 to 20 kV, and 5% from 20 to 30 kV. This could be explained by

generation of joule heating inside the capillary, due to the higher current resulting from the

higher voltage, where the temperature increase affects both the relative permittivity and the

viscosity of the buffer as seen in equation 7. The velocity of the EOF on the other hand is

linearly dependent on the electric field, and with higher voltage the velocity increased.

Figure 5 – The electroosmotic mobility in the CE system at 15, 20 and 30 kV with an

ethylammonium nitrate concentration of 20, 40 and 80 mM at pH 10 (n=3). Samples were

hydrodynamically injected and the capillary had a total length of 98.5 cm. (Figure 2, Paper I)

15

Instead of having roughly 20 min runs for 15 kV, the analysis time got reduced to about

10 min.

5.2.3 Separation of proteins using ionic liquids

The performance of EAN as the BGE was also evaluated by separation of some protein

samples. One of the major concerns when running proteins in CE is protein absorption to

the capillary wall, which affects the recovery and separation of the proteins. The chosen

proteins in Paper I were myoglobin, albumin, and osteoprotegerin (OPG), and they were

chosen so that a distribution in size and pI was obtained. Myoglobin is a well-studied

protein [30], and is mainly found in muscle tissue. Albumin is one of the most abundant

proteins in blood [31], while OPG on the other hand is used as a potential biomarker for

several pathologic conditions that affect the skeleton and is found in very low

concentrations [32, 33].

For protein analysis, the pH of the EAN-water solution was chosen to 10 to avoid having

extremely harsh conditions. For this CE conditions the proteins had a negative charge, since

the pI values of the proteins, see Table 1, were lower than the pH of the buffer, and the

proteins migrated with a lower velocity than the EOF. The obtained mobilities can be seen

in Table 1, where both the relative standard deviation (RSD) of the peak height and the peak

area are shown.

Table 1 - The relative standard deviation (RSD) for peak height, peak area, and mobility for the

proteins myoglobin, albumin, and osteoprotegerin (n=3). (Table 1, Paper I)

Myoglobin Albumin Osteoprotegerin

pI 7.23 5.59 6.62

Molecular weight (kDa) 17 ∼70 20

RSD (%) Height Area Height Area Height Area

6.59 0.88 0.47 0.18 4.91 0.10

Averaged apparent

mobility (×10-4 cm2/sV) 4.87 3.16 6.11

RSD of mobility (%) 0.65 1.39 0.39

The low RSD values obtained for both the area, and the migration time, suggest that there

was low interaction of the proteins to the capillary wall. Having a reliable analytical method

for identification of proteins with different characteristics and abundance is very desirable.

The use of EAN as the BGE in the method for separation of proteins is within the allowed

reproducibility of routine quality control according to FDA [34].

16

6 Applications in capillary electrophoresis

As mentioned before, CE is used to analyse a large variety of compounds. The different

separation modes in CE, together with a variety of detectors e.g. UV-Vis absorption

spectrometric, fluorescence, contactless conductivity, amperometric, and

electrochemiluminescence detectors, makes CE a very universal technique, which could be

used for both quantitative and qualitative analysis. [35] CE could be used for analysing

small cations and anions such as sodium, calcium, chloride, and formate, coupled to e.g. a

contactless conductivity detector [36], or larger compounds as amino acids [37],

pharmaceuticals as residues in water samples [38], or nucleic acids [39], which can be

detected with several of the detectors mentioned before.

The sections below will cover quite uncommon application for CE.

6.1 Particles

Lately, particles of different size has increased in population in different fields [40]. The

particles are divided into subgroups depending on the size of the particle, but mainly there

are two subdivisions. Microparticles are defined according to IUPAC as entities having a

size range from 100 nm to 100 µm, and nanoparticles are defined as particles with at least

one dimension in the nanoscale (1-100 nm) [41]. It is only during the last years that

particles have found their application for CE. Particles in CE can be analysed as the sample,

used as capillary coatings, or added to the BGE as additives [42-44]. When analysing

particles with CE, information regarding particle size, the functionality at the surface and

the size distribution can be obtained. There are particles that have size-dependent

properties, which require the level of information that could be obtained with CE.

Characterization of inorganic metal oxide particles [43] and polymeric particles [45], but

also biological particles as viruses, organelles and cells [46] have all shown that crucial

information about the particles is obtained using CE.

6.2 Polyvinyl alcohol microbubbles

In the second part of this thesis (Paper II) a new type of sample was analysed using CE. In

Paper II, a method was developed for detection of microbubbles (MBs). The MBs are

hollow spheres with a hard shell of polyvinyl alcohol (PVA) polymer, which were separated

from various medium such as water and blood plasma. The polyvinyl alcohol microbubbles

(PVA-MBs) were developed to be used as contrast agents (CA) for ultrasound [47], since it

previously was known that micro-sized MBs generally increased the level of detail in the

ultrasound image [48]. However, the first initial experiment with CA performed in 1968

showed that the stability of the MBs where only a few seconds and limited their practical

use [48]. In this early experiment, the CA consisted of free gas MBs, and since then

researchers have been optimizing CA regarding to the MBs stability and their size

distribution.

17

There are already approved CA in use for clinical diagnostics. Those CA often have a soft

shell with lipids or phospholipids and a core filled with perfluoropropane or sulphur

hexafluoride [49]. Compared to the commercially available CA, the PVA-MBs have several

advantages. One is that the PVA-MBs have a shelf-life of several months, since the hard

polymer shell is more resistant to mechanical stress than the soft lipid shells are [47].

Another advantage is that the size of the fabricated PVA-MBs can be controlled to some

extent with the pH and temperature of the solution during polymerization [50]. This yields a

more homogenous MB production regarding their micro-size, with a narrower distribution

around the selected size compared with the lipid shell MBs. Keeping the size distribution to

a minimum is important from an imaging point-of-view, since MBs size distribution will

affect their efficiency as enhancers of the ultrasound signal. [51] The way PVA-MBs (and

other CA) works during a scan, is that the characteristics of the MBs enhance the ultrasound

signal by scattering the ultrasonic wave differently, due to acoustic properties of the PVA-

MBs, compared to the surrounding medium [52]. The PVA-MBs can therefore be injected

into the blood stream, and be used for facilitating the visualization of microcirculation of

tumours or the myocardium [48, 53, 54].

The characterization of the MBs can be performed with different imaging techniques

depending on the information needed. Techniques which are frequently utilized are

confocal laser scanning microscopy [47, 55, 56], light scattering [56, 57], scanning

transmission X-ray microscopy [55, 58], scanning electron microscope (SEM) [55] and

digital holographic microscopy [59]. The size of the PVA-MBs, seen in Figure 6

(Paper II), used in Paper II was determined by using ImageJ software on images obtained

from SEM measurement and a light microscope. The external diameter from five single

MBs, which were isolated from other MBs in the images and covering the visual size range,

was calculated to an average of 3.0 ± 1.1µm and 3.7 ± 1.2 µm, respectively. The difference

in the averaged diameter is most likely due to the characteristics of the two imaging

techniques [60]. When performing SEM measurements, the sample is in vacuum, compared

to a water solution during observation with the light microscope. There is a possibility that

the water evaporates from the shell or that the air escapes from the core through the shell of

the PVA-MBs during SEM-measurements. This could explain the lower diameter obtained

with the first technique. However, when utilizing the PVA-MBs in clinical diagnostics, the

PVA-MBs will be in the blood stream and the diameter obtained from the light microscope

is a more representative value of the PVA-MBs size.

18

Figure 6 - The size distribution of the PVA-MBs shown with a) SEM measurements and b)

images of PVA-MBs. The numbers indicate which MBs were used for estimations of the size

distribution. The concentration of PVA-MBs was 1.3·107 MBs/mL in water

(Part of Figure 1, Paper II).

With the imaging techniques previously mentioned, the concentration of the MBs is

obtained by taking an image of the sample with a discrete volume and counting the MBs

shown in the picture. The disadvantage with this quantification method is that it is time

consuming and prone for errors. Techniques that are more automatic like flow cytometry

requires a larger sample volume (µL-mL), and depending on the chosen detector it is

impossible to detect particles in the range of 1–30 µm due to the narrow beam light [61].

In Paper II, the goal was to develop a method where the PVA-MBs could be detected both

qualitatively and quantitatively. CE was the analytical technique of choice, since it requires

small sample volumes (nL), and as described before, CE has been already used for

separation of particles. Although the PVA-MBs are not solid particles, their size resembles

micro-sized particles.

When developing the CE-system there where several parameters that had to be investigated.

First, it is known that the PVA-MBs can be seen in visible light from the light microscope

experiments, but there is no information regarding their absorbance in the UV region. In

order to know if the detection should be performed with direct or indirect detection, an

absorbance spectrum for the PVA-MBs in the UV-region needed to be recorded. Secondly,

the choice of BGE with respect to pH and type of buffer salt would influence the separation.

19

The maximum absorbance for the PVA-MBs was determined with a spectrophotometer to

260 nm, which indicated that a direct detection could be conducted. Hence, the buffer did

not need to have a chromophore. The buffer salt was chosen so that it mimics the

physiological conditions in human blood. A phosphate buffer was selected, since that is the

major buffer salt in blood plasma, but also since phosphate has 3 pKa values covering the

whole pH-range (pKa1: 2.12, pKa2: 7.21, and pKa3: 12.67) [62]. pH 7 and 12 were selected

for further pH-analysis, and correspond well with the pKa values of phosphate. pH 7 was

chosen since it is closest to the physiological conditions and pH 12, due to deprotonation of

alcohol groups on the PVA polymer. The deprotonation affects the charge of the PVA-MBs.

Figure 7 – Electropherograms for PVA-MB with a concentration of 8.2∙107 MB/mL. Sodium

phosphate buffer was utilized with a a) pH of 7.07 at 20.1°C, and b) pH of 11.95 at 20.5°C. The

voltage was 20 kV.

The electropherograms of PVA-MBs diluted in water, seen in Figure 7, does not consist of

one single peak, rather of a high number of peaks, where each peak represents a set of MBs

with similar mobility. From the CE separation, the size distribution of the PVA-MBs can be

obtained. For the neutral pH, Figure 7 a, the migration time was approximately 11 min, with

a spread of the peaks for almost 2 min. However, when pH 12 was utilized, the migration

time was faster (roughly 10 min), and the range for the peaks was narrower (around 1 min).

For further experiments the higher pH was chosen, to minimize the risk of comigration of

PVA-MBs with blood constituents, which would have a slower migration time at that pH.

20

For the pH-analysis step, sodium phosphate was used. However, the baseline was noisy and

unstable and the baseline was more stable when potassium was used as the counter-ion

instead. Thus, potassium phosphate was used as the BGE in further experiments.

6.2.1 Ultrasonicated polyvinyl alcohol microbubbles

In this case, it is not enough to have a method that only has a response for the sample, but

rather should give a response for the degradation products as well. The PVA-MBs run the

risk of degrading inside the body when used as ultrasound enhancers, if the PVA-MBs are

exposed to a too high insonating power of the ultrasonic wave [63]. The degradation could

therefore lead to a misleading result and wrong conclusion regarding the concentration of

the MBs. To investigate whether the peaks detected in the previous CE analysis originate

from intact PVA-MBs, instead of PVA-MB degradation products, PVA-MBs were

degraded with high intensity ultrasound. The power of the ultrasound probe was 137 W, and

should have been sufficient to degrade the MBs. As a comparison, the range for the acoustic

power used in clinical ultrasound diagnostic varies between 0.3 and 440 mW depending on

the mode and the tissue investigated [64]. Before CE analysis, the degraded sample was

analysed with SEM, seen in Figure 8 (Paper II). Comparing Figure 8a with b there is few

to almost no MBs left and only degradation products are seen.

Figure 8 - Difference between a) the intact PVA-MBs solution with a magnification of 500, and

b) the ultrasonicated PVA-MBs solution with a magnification of 600 illustrated with scanning

electron microscope images. (Figure S2, Paper II)

21

Running the ultrasonicated MBs in CE, and comparing it to the electropherograms of PVA-

MBs and water, seen in Figure 9 (Paper II), shows several differences between the three

samples. The striking dissimilarity between the intact MBs and the degraded MBs

(Fig. 9a and b) are the group of individual peaks that has been replaced by a single wide

peak after the negative water peak. One explanation for having a wide peak is that the

degraded pieces are smaller in size and have a smaller size distribution than intact MBs.

However, the migration time of the wide peak is similar to the intact MBs, which could be

explained by the identical functional group on both analytes [43]. With the CE method, it is

possible to distinguish intact MBs from degraded MBs.

Figure 9 – Electropherograms from (A) a sample of intact PVAMBs, (B) a sample of degraded

PVA-MBs, and (C) water as a reference. In all cases a potassium phosphate buffer was utilized

as BGE with a pH of 11.93 at 21.3C, and the voltage was 20 kV. The concentration of PVA-

MBs was 3.2 × 108 MBs/mL. (Figure 2, Paper II)

22

6.2.2 Polyvinyl alcohol microbubbles in human blood

In clinical use, the PVA-MBs will be circulating in the blood stream, and the easiest way of

determining the MB concentration in the blood would be through a blood test. To simulate

that procedure, the samples were collected in vacuum tubes, and the human blood was

spiked with PVA-MBs. The vacuum tubes were treated according to clinical procedures to

achieve either blood plasma or blood serum. The size of the MBs would suggest that the

PVA-MBs would be found in the lower layers of the centrifuged blood. However, the PVA-

MBs have high buoyancy due to the air-filled core, which make them arise to the surface,

where they could simply be separated from other blood components with the supernatant,

seen in Figure 10a and d (Paper II). To confirm that the white line above the supernatant,

see Figure 10a and d, are MBs and not foam, the supernatant where observed under light

microscope and compared to plasma and serum without PVA-MBs, Figure 10 b-c, and e-f.

Figure 10 – Centrifuged samples of human blood spiked with PVA-MBs in vacuum tubes; (A)–

(B) with K2EDTA (generating a plasma) and (D)–(F) with a gel containing coagulation

activators (generating a serum). Light microscope images of plasma (B) with MBs and (C)

without, and serum (E) with MBs and (F) without. (Figure 3, Paper II)

23

After confirming the presence of MBs in plasma, the plasma samples were analysed in CE,

see Figure 11 (Paper II). In Figure 11a, the plasma sample without PVA-MBs is shown,

where the negative peak with a migration time of roughly 10 min belongs to water. The

other peaks seen in the electropherogram are related to unidentified blood components,

which are highly abundant e.g. albumin. The PVA-MBs in plasma showed a difference

compared to previous electropherogram, which is enlarged in Figure 11c. The number of

peaks with migration times between 10 and 13 min corresponds to the PVA-MBs, and that

time range has increased compared to when the PVA-MBs where only diluted in water. The

longer migration times suggest that the mobility of the PVA-MBs has altered. From

equation 5, it is evident that either the size of the MBs are increased, the charge of them is

decreased, or both. This could be explained by protein adsorption to the surface of the MBs,

and the absorbed proteins could create a corona around the PVA-MBs [65] resulting in an

altered radius, as well as a change in surface charge [50]. However, the majority of PVA-

MBs remains within the same migration time range, which suggests that most of the

proteins are desorbed from the MBs surfaces when entering the CE system. This protein

desorption could occur due to changes in pH, diffusion into the buffer or caused by the

applied voltage [66].

Figure 11 – Electropherograms of (A) a sample of human blood plasma, and (B) human blood

plasma mixed with PVA-MBs at a concentration of 4.4 × 107 MBs/mL. (C) Is an enlargement of

(B) at the migration time 9.5–11.5 min. A potassium phosphate buffer with a pH of 11.92 at

20.2°C was used. The applied voltage was 20 kV. (Figure 4, Paper II)

24

6.2.3 Modified polyvinyl alcohol microbubbles

In the previous sections, the PVA-MBs used had neither modification to the surface, nor to

the shell of the MBs. However, over the recent years researchers have experimented with

the PVA-MBs as potential drug delivery systems by modifying the surface of the MBs with

target specific groups [56]. The MBs shell structure could also be modified by incorporating

nanoparticles, resulting in a multimodal imaging CA. One of the nanoparticles that are

added to the PVA mixture during MB fabrication is superparamagnetic nanoparticles of

iron oxide (SPIONs) [54]. The PVA-MBs with SPIONs imbedded in the shell (Fe-PVA-

MBs) can be used in diagnostics, both for ultrasound imaging and magnetic resonance

imaging (MRI). The basic principal of MRI is the relaxation time of the protons after

excitation of the nuclear spin due to a pulse of radio waves in a magnetic field. The contrast

within the image can be varied by changing the pulse sequence, or the relaxation properties

of the protons in the medium. Fe-PVA-MBs enhance the MRI due to the stronger spin–spin

relaxation time effect of the protons due to the magnetic properties of the SPIONs. The

clinical market is dominated by gadolinium chelated complexes as CA for MRI [67, 68],

but compared to those complexes, SPIONs are more biocompatible and can easily be

degraded using the cells’ normal iron metabolism [69]. Generally, CA in MRI facilitates the

detection of cardiovascular diseases, organ/lesion perfusion, and renal disease among

others. [67]

Running the Fe-PVA-MBs both diluted in water and in human blood plasma, with the same

CE conditions as for the plain PVA-MBs, resulted in electropherograms similar to those

obtained for PVA-MBs, seen in Figure 12. The only distinguishable difference is the higher

intensity of the MB peaks for the Fe-PVA-MBs. This increased intensity is likely due to the

higher optical density of the Fe-PVA-MBs compared to the PVA-MBs, since the

concentration of MBs where in the same order of magnitude.

25

Figure 12 - Electropherogram for Fe-PVA-MBs a) diluted in water, and b) diluted in human

blood plasma. The peak marked with a star have higher absorbance then shown here. Potassium

phosphate buffer was utilized as BGE with a pH of 11.9 at 20°C, and the voltage was 20 kV.

The current was 15-18 µA and the concentration of Fe-PVA-MBs was 3.2 · 108 MBs/ml.

(Figure S1 and S3, Paper II)

This confirms that the CE method can be used for different types of PVA-MBs.

a) b)

26

6.2.4 Camera detection

While the UV-detection used in the electropherograms above worked well, modification of

the MBs resulted in altered signals. To obtain more universal results, a new type of

detection is under development. Since the PVA-MBs are visible under a regular

microscope, an initial study has been performed where an analogue charged coupled device

(CCD) camera is used as a complement to the existing UV-detector.

The CCD-camera was started simultaneously with the CE run and recorded the progress.

The flow of the PVA-MBs can be seen in Figure 13, where three MBs are tracked. Initial

experiments have been conducted where the footage has been processed through a

MATLAB program, where this program converted the film into an electropherogram.

However, the resolution of the CCD-camera was rather low, which makes the smallest and

dizziest features in the film not accountable by the program. This leads to a lower count of

MBs than injected. By an upgrade of the camera to a digital camera, the frequency of the

streamed video can be increased, which could allow the operator to determine the speed of

the microbubbles in the capillary. The increased resolution of a digital camera compared to

an analogue one, enables determination of the size of the microbubbles, and microparticles

in general, with higher precision compared to the setup used hitherto.

Figure 13 – Camera images of the capillary during analysis of PVA-MBs at a) t=0, b) t=t1, and

c) t=t2. The same three PVA-MBs are marked in each image.

27

7 Conclusions and future perspectives

The aims of the work in the thesis have been fulfilled by the development of the complete

analytical processes used for separation of proteins and MBs using CE as an analytical

technique. The advantages and possibilities that could be achieved with CZE are stated, but

also how CZE methods could be optimized by adjusting the EOF using several parameters

as buffer concentration, pH of the BGE, and electric field.

Using EAN as the BGE, the precision of the µeo and the precision for protein mobility could

be increased to values acceptable in routine control analysis, since EAN-water solution

contributes to lower currents and has an adequate buffer capacity.

CZE can be used for size characterization and concentration of particles, here exemplified

by PVA-MBs. The intact PVA-MBs could easily be separated from ultrasound degraded

MBs and proteins existing in human blood plasma.

With an analogue camera, the PVA-MBs can be monitored optically inside the capillary. A

new digital camera detector would increase the versatility and sensitivity of the capillary

electrophoresis instrumentation when analysing microparticles, since more accurate

information regarding the concentration, size and speed would be achieved.

28

8 Acknowledgements

I would like to express my gratitude for the School of Chemistry for financing my

Excellence PhD student position. However, this would not have been possible without my

supervisors Åsa Emmer, Mark Rutland and Johan Jacksén, who accepted me with open

arms and guided me during those years.

I would like to thank all my co-workers, both past and present for all the fine memories we

share. Johan Jacksén, thank you for all the advices from you regarding my work, my

rebuilding of my home, and for teaching me to be more practical and not so theoretical. It

would not be the three musketeers for so long without Saara Mikkonen and hers “Hur svårt

ska det vara?” I’m glad that Joakim Romson finally joined the group, where he always

belonged. Now I will not be the only one concerned about the brightness of the computer

screens. What would our office time be without the smile you put on our faces Linus

Svenberg, and thank you for accepting all the stuff me and Alex tried to get rid of. Thank

you Yuye for translating Chinese words and website for me, and for all the discussions we

have had in the office.

This work would not have been possible without the help of my collaborators Malin

Larsson, Anna Bjällmark, Jessica Bernsteen and the help with all of my students.

Most of all, I would like to acknowledge my family for being so supportive throughout my

time at KTH. Dziekuje Ci mamo że pomagałaś mi ile mogłaś w szkole, i że zawsze tu jesteś

jak potrzebuje Ciebie. Nawet jak tylko nazekasz na moją wagę! Tack Booo :* för att du

finns där för mig, stöttar mig och läser mina arbeten innan jag vågar skicka in dem till Åsa.

Sharing my work at conferences (and enjoying some good times abroad) would not be

possible without the financial support from Stiftelsen Tornspiran, Knut & Alice

Wallenbergs stiftelse, ÅForsk, Gålöstiftelsen, Linders stiftelse and Roos stiftelse.

29

9 References

1. Harris, D.C., Quantitative Chemical Analysis. 2010: W. H. Freeman.

2. Guilbault, G.G. and M. Hjelm, Nomenclature for automated and mechanised analysis

(Recommendations 1989). Pure and Applied Chemistry, 1989. 61(9).

3. Harstad, R.K., et al., Capillary Electrophoresis. Analytical Chemistry, 2016. 88(1): p.

299-319.

4. Tiselius, A., A new apparatus for electrophoretic analysis of colloidal mixtures.

Transactions of the Faraday Society, 1937. 33: p. 524.

5. Brody, J.R. and S.E. Kern, History and principles of conductive media for standard

DNA electrophoresis. Anal Biochem, 2004. 333(1): p. 1-13.

6. Hjertén, S., Free zone electrophoresis. Chromatographic Reviews, 1967. 9(2): p. 122-

219.

7. Camilleri, P., Capillary electrophoresis: Theory and Practice. Second edition ed. 1997,

Salem, MA, United states of America: CRC Press. 576.

8. Lukacs, K.D. and J.W. Jorgenson, Capillary zone electrophoresis: Effect of physical

parameters on separation efficiency and quantitation. Journal of High Resolution

Chromatography, 1985. 8(8): p. 407-411.

9. Mayer, B.X., How to increase precision in capillary electrophoresis. Journal of

Chromatography A, 2001. 907(1): p. 21-37.

10. Kok, W., Capillary electrophoresis: Instrumentation and operation. Chromatographia,

2000. 51(3331): p. XI.

11. Whatley, H., Basic Principles and Modes of Capillary Electrophoresis, in Clinical and

Forensic Applications of Capillary Electrophoresis, J.R. Petersen, Mohammad, A.A.,

Editor. 2001, Humana Press: Totowa, NJ.

12. Knox, J.H. and K.A. McCormack, Temperature effects in capillary electrophoresis. 1:

Internal capillary temperature and effect upon performance. Chromatographia, 1994.

38(3): p. 207-214.

13. Rathore, A.S., Joule heating and determination of temperature in capillary

electrophoresis and capillary electrochromatography columns. Journal of

Chromatography A, 2004. 1037(1): p. 431-443.

14. Sun, P. and D.W. Armstrong, Ionic liquids in analytical chemistry. Analytica Chimica

Acta, 2010. 661(1): p. 1-16.

15. Shamsi, S.A. and N.D. Danielson, Utility of ionic liquids in analytical separations. J

Sep Sci, 2007. 30(11): p. 1729-50.

16. Lopez-Pastor, M., et al., Ionic liquids and CE combination. Electrophoresis, 2008.

29(1): p. 94-107.

17. Berthod, A., M.J. Ruiz-Angel, and S. Carda-Broch, Ionic liquids in separation

techniques. J Chromatogr A, 2008. 1184(1-2): p. 6-18.

18. Wilkes, J.S., et al., Dialkylimidazolium chloroaluminate melts: a new class of room-

temperature ionic liquids for electrochemistry, spectroscopy and synthesis. Inorganic

Chemistry, 1982. 21(3): p. 1263-1264.

19. Wilkes, J.S. and M.J. Zaworotko, Air and water stable 1-ethyl-3-methylimidazolium

based ionic liquids. Journal of the Chemical Society, Chemical Communications,

1992(13): p. 965.

20. Liu, J.F., J.A. Jonsson, and G.B. Jiang, Application of ionic liquids in analytical

chemistry. Trac-Trends in Analytical Chemistry, 2005. 24(1): p. 20-27.

21. Anderson, J.L., D.W. Armstrong, and G.-T. Wei, Ionic Liquids in Analytical Chemistry.

Analytical Chemistry, 2006. 78(9): p. 2892-2902.

22. Sun, P. and D.W. Armstrong, Ionic liquids in analytical chemistry, in Analytica

Chimica Acta. 2010. p. 1-16.

23. Soares, B., et al., Ionic liquids in chromatographic and electrophoretic techniques:

toward additional improvements in the separation of natural compounds. Green

30

chemistry : an international journal and green chemistry resource : GC, 2016. 18(17): p.

4582-4604.

24. Huang, Y., S. Yao, and H. Song, Application of ionic liquids in liquid chromatography

and electrodriven separation, in J. Chromatogr. Sci. 2013. p. 739-752.

25. Qi, S., et al., Rapid separation and determination of aconitine alkaloids in traditional

Chinese herbs by capillary electrophoresis using 1-butyl-3-methylimidazoium-based

ionic liquid as running electrolyte, in Biomed. Chromatogr. 2006. p. 294-300.

26. Qi, S., et al., Rapid and sensitive determination of anthraquinones in Chinese herb

using 1-butyl-3-methylimidazolium-based ionic liquid with β-cyclodextrin as modifier in

capillary zone electrophoresis. Journal of Chromatography A, 2004. 1059(1): p. 191-

198.

27. Huang, T.-L., On the pH hysteresis of electroosmotic mobility with capillary zone

electrophoresis in silica capillary. Chromatographia, 1993. 35(7): p. 395-398.

28. Lambert, W.J. and D.L. Middleton, Ph Hysteresis Effect with Silica Capillaries in

Capillary Zone Electrophoresis. Analytical Chemistry, 1990. 62(15): p. 1585-1587.

29. Horn, R.G., D.F. Evans, and B.W. Ninham, Double-layer and solvation forces

measured in a molten salt and its mixtures with water. The Journal of Physical

Chemistry, 1988. 92(12): p. 3531-3537.

30. Totzeck, M., et al., Assessment of the functional diversity of human myoglobin. Nitric

Oxide, 2012. 26(4): p. 211-216.

31. Anderson, N.L. and N.G. Anderson, The human plasma proteome: history, character,

and diagnostic prospects. Molecular & cellular proteomics : MCP, 2002. 1(11): p. 845-

867.

32. Boyce, B.F. and L. Xing, Biology of RANK, RANKL, and osteoprotegerin. Arthritis

Research & Therapy, 2007. 9(Suppl 1): p. S1.

33. Krämer, H.H., et al., Osteoprotegerin: A new biomarker for impaired bone metabolism

in complex regional pain syndrome? Pain, 2014. 155(5): p. 889-895.

34. Administration, F.a.D., Methods, Method Verification and Validation. ORA Laboratory

Procedure. p. 1-19.

35. Tůma, P. and F. Opekar, Detectors in Capillary Electrophoresis, in Analytical

Separation Science. 2015, Wiley-VCH Verlag GmbH & Co. KGaA.

36. Elbashir, A.A., O.J. Schmitz, and H.Y. Aboul-Enein, Application of capillary

electrophoresis with capacitively coupled contactless conductivity detection (CE-C4D):

An update. Biomedical Chromatography, 2017. 31(9): p. e3945-n/a.

37. Denoroy, L. and S. Parrot, Analysis of Amino Acids and Related Compounds by

Capillary Electrophoresis. Separation & Purification Reviews, 2017. 46(2): p. 108-151.

38. Hamdan, I.I., Capillary electrophoresis in the analysis of pharmaceuticals in

environmental water: A critical review. Journal of Liquid Chromatography & Related

Technologies, 2017. 40(3): p. 111-125.

39. Lian, D.S. and S.J. Zhao, Capillary electrophoresis based on nucleic acid detection for

diagnosing human infectious disease. Clinical Chemistry and Laboratory Medicine,

2016. 54(5): p. 707-738.

40. Guihen, E., Nanoparticles in modern separation science. TrAC Trends in Analytical

Chemistry, 2013. 46(Supplement C): p. 1-14.

41. Vert, M., et al., Terminology for biorelated polymers and applications (IUPAC

Recommendations 2012), in Pure and Applied Chemistry. 2012. p. 377.

42. Zarei, M., M. Zarei, and M. Ghasemabadi, Nanoparticle improved separations: From

capillary to slab gel electrophoresis. TrAC Trends in Analytical Chemistry, 2017.

86(Supplement C): p. 56-74.

43. Petersen, S.L. and N.E. Ballou, Separation of micrometer-size oxide particles by

capillary zone electrophoresis. Journal of Chromatography A, 1999. 834(1): p. 445-

452.

44. Pyell, U., Characterization of nanoparticles by capillary electromigration separation

techniques. ELECTROPHORESIS, 2010. 31(5): p. 814-831.

31

45. Riley, K.R., et al., Using capillary electrophoresis to characterize polymeric particles.

Journal of Chromatography A, 2016. 1463(Supplement C): p. 169-175.

46. Subirats, X., D. Blaas, and E. Kenndler, Recent developments in capillary and chip

electrophoresis of bioparticles: Viruses, organelles, and cells. ELECTROPHORESIS,

2011. 32(13): p. 1579-1590.

47. Cavalieri, F., et al., Stable polymeric microballoons as multifunctional device for

biomedical uses: Synthesis and characterization. Langmuir, 2005. 21(19): p. 8758-

8764.

48. Gramiak, R. and P.M. Shah, Echocardiography of the aortic root. Investigative

Radiology, 1968. 3(5): p. 356-366.

49. Mulvagh, S.L., et al., American Society of Echocardiography Consensus Statement on

the Clinical Applications of Ultrasonic Contrast Agents in Echocardiography. Journal

of the American Society of Echocardiography, 2008. 21(11): p. 1179-1201.

50. Stigler, J., et al., Protein Interactions with Microballoons: Consequences for

Biocompatibility and Application as Contrast Agents, in Ultrasound Contrast Agents,

G. Paradossi, P. Pellegretti, and A. Trucco, Editors. 2010, Springer Milan. p. 53-66.

51. Grishenkov, D., et al., In vitro contrast-enhanced ultrasound measurements of capillary

microcirculation: Comparison between polymer- and phospholipid-shelled

microbubbles. Ultrasonics, 2011. 51(1): p. 40-48.

52. Streeter, J.E., et al., Improving Sensitivity in Ultrasound Molecular Imaging by

Tailoring Contrast Agent Size Distribution: In Vivo Studies. Molecular imaging, 2010.

9(2): p. 87-95.

53. Harvey, C.J., et al., Developments in ultrasound contrast media. European Radiology,

2001. 11(4): p. 675-689.

54. Brismar, T.B., et al., Magnetite Nanoparticles Can Be Coupled to Microbubbles to

Support Multimodal Imaging. Biomacromolecules, 2012. 13(5): p. 1390-1399.

55. Tzvetkov, G., et al., Water-dispersible PVA-based dry microballoons with potential for

biomedical applications. Materials Science and Engineering: C, 2010. 30(3): p. 412-

416.

56. Cavalieri, F., et al., Tethering functional ligands onto shell of ultrasound active

polymeric microbubbles. Biomacromolecules, 2006. 7(2): p. 604-611.

57. Paradossi, G., et al., Tailoring of physical and chemical properties of macro- and

microhydrogels based on telechelic PVA. Biomacromolecules, 2002. 3(6): p. 1255-

1262.

58. Fernandes, P.A.L., et al., Quantitative analysis of scanning transmission x-ray

microscopy images of gas-filled PVA-based microballoons. Langmuir, 2008. 24(23): p.

13677-13682.

59. Saglimbeni, F., et al., Optical characterization of an individual polymer-shelled

microbubble structure via digital holography. Soft Matter, 2012. 8(34): p. 8822-8825.

60. Tzvetkov, G., et al., Soft X-ray induced modifications of PVA-based microbubbles in

aqueous environment: a microspectroscopy study. Physical Chemistry Chemical

Physics, 2009. 11(7): p. 1098.

61. Givan, A.L., Shapiro, H.M., Stewart, C.C., Perez, O.D., Maecker, H.T., Liu, L.,

Telford, W.G., Eaker, S.S., Bertoncello, I., Dooley, D.C., Hawley, T.S., Pruitt, S.C.,

Galbraith, D.W., Chan, F.K-M., Lievens, S., Feldhaus, M., Cavrois, M., Ormerod,

M.G., El-Naggar, A.K., Kapoor, V., Perfetto, S.P., Flow Cytometry Protocols. 2 ed.

Methods in Molecular Biology. 2014: Humana Press. XIV, 434.

62. Vidaud, L., et al., Mat- pKa calculation tool development for evaluation of acidity

constants from solubility profiles - Large study of 41 compounds. International Journal

of Pharmaceutics, 2012. 437(1-2): p. 137-155.

63. Lentacker, I., S.C. De Smedt, and N.N. Sanders, Drug loaded microbubble design for

ultrasound triggered delivery. Soft Matter, 2009. 5(11): p. 2161-2170.

32

64. Henderson, J., et al., A survey of the acoustic outputs of diagnostic ultrasound

equipment in current clinical use. Ultrasound in Medicine and Biology, 1995. 21(5): p.

699-705.

65. Sakulkhu, U., et al., Significance of surface charge and shell material of

superparamagnetic iron oxide nanoparticle (SPION) based core/shell nanoparticles on

the composition of the protein corona. Biomaterials Science, 2015. 3(2): p. 265-278.

66. Belder, D., et al., Cross-linked poly(vinyl alcohol) as permanent hydrophilic column

coating for capillary electrophoresis. Electrophoresis, 2001. 22(17): p. 3813-3818.

67. Hasebroock, K.M. and N.J. Serkova, Toxicity of MRI and CT contrast agents. Expert

Opinion on Drug Metabolism and Toxicology, 2009. 5(4): p. 403-416.

68. Kato, H., et al., Lanthanoid endohedral metallofullerenols for MRI contrast agents.

Journal of the American Chemical Society, 2003. 125(14): p. 4391-4397.

69. Bulte, J.W.M. and D.L. Kraitchman, Iron oxide MR contrast agents for molecular and

cellular imaging. NMR in Biomedicine, 2004. 17(7): p. 484-499.