Capillary Zone Electrophoresis and Ion Mobility Spectrometry
Applications benefits and Shortcomings
Literature Thesis
Thesis submitted in partial fulfilment of the requirements for the degree of Master of Science at Vrije Universiteit van Amsterdam and Universiteit van Amsterdam.
Faculty of Science
Amsterdam Institute for Molecules, Medicines and Systems
Author:
Supervisors:
Maurice de Jonker
Rob Haselberg
Abstract
In this thesis the coupling of capillary zone electrophoresis (CZE), ion mobility spectrometry (IMS) and ion mobility mass spectrometry (IM-MS) will be reviewed to gain understanding about the capabilities of these coupled systems, there advantages, disadvantages and practical applications. IMS separates ions based on their mobility through an inert buffer gas induced by an electric field. The mobility of ions is based on their shape, size and charge, providing structural information. Thus, it can be used to separate ions based on their mobility. Coupled to MS it offers a powerful hybrid analytical technique with a broad array of applications. In this thesis an overview of the most prevalent forms of IMS will be given. Applications, advantages and shortcomings off an array of hyphenated CZE-IM(-M)S techniques are highlighted. Similarities and differences between separation using CZE and IMS will be highlighted to gain understanding about the orthogonality of these techniques. Understanding of orthogonality is important to assess the usefulness and specific applications of these hyphenated systems. The main difference between CZE and IMS is the phase in which the separation is occurring, CZE separates analytes dissolved in a buffer solution while IMS’ separation takes place in a gaseous buffer. Analytes need to be ionized and converted to the gas phase this is, in CZE-IMS systems, done using electrospray ionization (ESI). The workings of ESI will be discussed to gain understanding in the effect of ionization on the shape of ionized macromolecules. Main applications of CZE-IM(-M)S observed in literature today include conformational and structural characterization studies of biological macro molecules such as proteins and DNA, furthermore applications are found in quantitative studies of Lipopolysaccharides and glycans. The advantage of coupling CZE and IM(M)S as opposed to CZE-MS systems is the enhanced capability to separate chiral compounds, 1 polymeric conformers 2 and isomers, 3,4 at the cost of added complexity.
Table of Contents
1. Introduction...1
1. Fundamentals of CZE...2
1.1. Capillary Zone Electrophoresis...2
1.2. Ion Mobility in Capillary zone electrophoresis...2
1.3. Electroosmotic flow...3
2. Electron Spray Ionization...4
3. Ion mobility spectrometry...5
3.1. Ion mobility Analysers...6
3.1.1. Drift Tube Ion Mobility Spectrometry (DTIMS)...7
Resolution and resolving power of DTIMS instruments...8
3.1.2. Traveling wave ion mobility spectrometry (TWIMS)...9
3.1.3. High field asymmetric waveform ion mobility (FAIMS)...9
3.1.4. Trapped IMS (TIMS)...10
4. CZE IMS orthogonality...12
5. CZE and IMS coupled systems...13
5.1. FAIMS to reduce background noise...14
5.2. CE-TWIM-MS analysis of native and labelled glycans...15
5.3. CZE-DTIMS for native and labelled glycans...16
5.4. Characterization of polyamidoamine dendrimers...17
5.5. CE-IMS for conformational studies...19
5.6. CE-IMS Protein conformation and enzymatic activity...19
5.7. Conformation of polyproline with CE and IMS separately...21
6. Discussion...22
7. Concluding remarks...24
List of Abbreviations
ABREVIATION S
APPI Atmospheric Pressure Photoionization
BGE Background Electrolyte
CCS Collision Cross Section
CE Capillary Electrophoresis
CEM Chain Ejection Model
CRM Charged Residue Model
CZE Capillary Zone Electrophoresis
DIMS Differential Ion Mobility Spectroscopy
DMS Differential Mobility Analyzer
DTIMS Drift Tube Ion Mobility Spectroscopy
EOF Electroosmotic Flow
ESI Electrospray Ionization
FAIMS Field Asymmetric Ion Mobility Spectroscopy
HDMS High Definition Mass Spectrometer
IEM Ion Evaporation Model
IM-MS Ion Mobility Mass Spectroscopy
IMS Ion Mobility Spectroscopy
MALDI Matrix Assist Laser Desorption
OIMS Overtone Ion Mobility Spectroscopy
QQQ Triple Quadrupole mass analyser
Q-TOF Quadrupole Time Off Flight mass spectrometer
TIMS Trapped Ion Mobility Spectroscopy
TM-IMS Transversal Modulation Ion Mobility Spectroscopy
1.
Introduction
Since the introduction of capillary electrophoresis (CE) in the 1980s CE has been a continuously evolving analytical methodology. 5 By introducing different modifications in the traditional instrumentation and the introduction of new operational modes. One of the most crucial advances in the evolution of CE and more specifically CZE, has been the coupling to detection systems beyond the traditionally optical detection systems. Most noteworthy is the coupling to various mass spectrometers (MS). Hyphenated techniques brought rise to a multitude of different applications, by increasing the sensitivity, the selectivity, peak capacity and allowing for analyte identification. Which made it possible to simultaneously determine a high number of different species in complex samples, Making the technique applicable in proteomics and many other biological studies. 6, 7, 8,
The main challenge in the coupling between CZE and for example MS is to implement a capable interface between the techniques. The CZE separation is carried out in the liquid/solvated phase at atmospheric pressures and the MS separation/detection takes place in a high vacuum in the gas phase. The background electrolyte used in CZE frequently causes incompatibilities in the ionization of the analytes and can cause fouling of the detector. The incompatibilities have led to a series of obstacles that had to be overcome. Nowadays many commercially available solutions exist for that circumvent this problem and application of CE include among others the analysis of water samples, 8 inorganic compounds, 9 pharmaceuticals 10 and carbohydrates 11 as well as applications in proteomics, 12 clinical proteomics, 13 clinical metabolomics, 14 foodomics, 15 lipodomics, 16, 17 and nanoparticle characterization. 18
Another technique that has shown great potential over the last decades is ion mobility spectrometry (IMS), it is a rapid separation and detection technique developed in the late 1970s. Ion mobility spectrometry is a technique used to separate ionized analytes based on their mobility in the gas phase through a carrier buffer gas. On its own its commonly used for detecting chemical warfare agents, explosives, and volatile organic compounds and is widely used in military applications and environmental monitoring as a standalone device for its quick response, portability, ease of operation and high sensitivity. In addition to traditional applications as military chemical-agent detectors, ion mobility techniques have become popular for different purposes. Nowadays, the technique has also many laboratory analytical applications which have been expanding exponentially over the last decades. In the 1990s, several important advances allowed the possibility of the analysis of macro molecules by coupling soft ionization sources with IMS, massively broadening the potential applications of IMS and enhancing the scope of possibilities. In recent years IMS has been in increasing demand in the analysis of complex samples, biological samples in particular. Ion mobility alone will likely not be sufficient for the separation and identification of each analyte, due to its low peak capacity. 22 Many analytes in complex biological samples have the same or similar characteristics.
Hence the use of hyphenated techniques such as IM-MS, the main advantage of IM-MS over MS is that IM-MS is capable of separating based on shape difference’s (isomers, conformers). Other hyphenated methods that have been proven to be extremely beneficial for increasing sensitivity and peak capacity are the online coupling of various front-end separation methods to IMS such as: gas chromatography19, 20, 21, liquid chromatography22–24, solid phase extraction25 and capillary electrophoresis. 22 In this literature review the coupling of IMS and CZE will be highlighted to assess
the similarities, orthogonality, the current state, various applications and future capabilities. To provide insight in their applications, benefits and shortcomings.
1.
Fundamentals of CZE
Electrophoresis is a phenomenon describing the motion of charged particles relative to a fluid induced by an uniform electric field. 26,27 The basic principle behind all electrophoretic separation methods is that different particles migrate with different velocities due to differences in size and charge between the particles. Traditionally electrophoresis has been performed in a slab gel format. The gel was used to minimize convective flow caused by Joule heating. To avoid overheating of the gels the separation voltage had to be relatively low causing lengthy analysis times. Another problem of this method is the quantitative interpretation of the results because the separated analytes had to be detected on the gel. Nowadays electrophoretic separation is not limited to gel-stations and can be performed in nanometre to micrometre capillaries. This form of electrophoresis is suitably named capillary electrophoresis (CE). CE solved many of the problems of the slab gels. The high surface to volume ratio of the capillary provides effective heat dissipation. Additionally, the high conductivity of the capillary limits the generation of Joule heating. Higher voltages can be applied, which will result in higher efficiencies and shorter analysis times. Other benefits of CE are uncomplicated and more robust automation, on-line detection and injection volumes in the order of nanolitres. Nowadays in CE there are a broad array of applications. ranging from quantificational to kinetic studies. 6, 7 Application of CE include among others the analysis of water samples, 8 inorganic compounds, 9 pharmaceuticals 10 and carbohydrates 11 as well as applications in proteomics, 12 clinical proteomics, 13 clinical metabolomics, 14 foodomics, 15 lipodomics, 16,17 and nanoparticle characterization. 18
1.1.
Capillary Zone Electrophoresis
There are several modes of CE, each with a different separation mechanism, suitable for different sample types. The most common format of CE is capillary zone electrophoresis (CZE).2627 A typical CZE setup consists of a fused silica capillary with an inner diameter of 10 – 100 µm and an outer diameter of 150 – 400 µm, with on either end of the capillary a buffer reservoir, a high voltage power supply (typically up to 30kV) with an electrode connected to both buffer reservoirs to provide a potential difference over the capillary. In CZE the separation is based on size and charge of the analytes in solution and is performed with a background electrolyte (BGE). Which is typically an aqueous acid-base buffer and its main function is to keep the separation conditions constant over the whole capillary length for the whole duration of the measurements as well as provide the right pH for the analysis. To introduce a sample into the capillary the sample vial is briefly swapped with the buffer vial, the sample can be introduced hydrodynamically or with the application of a potential difference. After introduction of the sample, the buffer vial gets replaced with the sample vial for the remainder of the separation.
1.2.
Ion Mobility in Capillary zone electrophoresis
After introducing a sample to the separation medium, an electric field is applied which causes molecules to accelerate in the direction of one of the two electrodes. The electrostatic force (F), also called the Lorentz force, exerted on a particle (I) in solution is proportional to the net charge of the particle (
qi
) and the strength of the electrical fieldE
, as shown in formula (0). 26, 28An opposing friction force ( F ) is presented by the medium through which the particles are moving. The friction force is proportional to the velocity (
vi
) at which the particles are moving. Which results in the particles reaching terminal velocity, also called the electrophoretic velocity. For a spherical particle the frictional force is expressed by the stokes equation, shown in formula (0).F=6 πη ∙ ri∙ vi (0)
Wherein
ri
is the effective radius of the particle andη
is the viscosity of the separation medium. Therefore, the electrophoretic velocity ( vi ) can be expressed as presented in formula (0).vi=
qi ∙ E
6 πη ∙ri
(0)To simplify the comparison of experimental data obtained with different field strengths. The mobility of an ion (
μi
) has been defined as shown in formula (0).μi=
vi
E
=
qi
6 πη ∙ri
(0)It is evident that ions can be separated based on differences in either qi or ri , or better yet a difference in
qi
/ri
ratios. The radius determining the mobility is the effective radius, in an aqueous solution this includes the hydration shell. The hydration shell can differ substantially from the ionic radii in crystals. The structures of larger biomolecules are strongly influenced by their hydration shells. 29 The hydration shell can be disrupted with the addition of chaotropic ions lowering the effective radius and increasing the apparent hydrophobicity. 30 The charge of amphoteric compounds is dependent on the pH of the BGE and can be manipulated by changing the pH.1.3.
Electroosmotic flow
Electroosmotic flow (EOF) is generated when a liquid near a charged surface is placed in an electrical field, this results in a movement in the liquid near the charged surface. Electroosmosis contributes to the separation in CZE, because of the small diameter capillaries that are being used. The high surface to volume ratios in CZE makes the EOF a significant factor in the migration of analytes through the capillary. The wall of a fused silica capillary is often negatively charged due to free silanol groups on the surface, that start to deprotonate above pH 1.5. The negative charge on the surface results in a distribution of positively charged species forming a layer due to electrostatic attraction. Upon application of a potential difference over the capillary. The electrical field exerts a force on the cations resulting in the movement of the fluid towards the anode. Viscous forces counteract the force acting on the thin layer near the wall resulting in a constant flow. The velocity of the EOF (
v eo
) is proportional to the electrical field strengthE
¿
)
and can be defined similarly to the ion mobility tosimplify the comparison of experimental data. The definition of the EOF mobility ( μ eo
)
is shown in formula (0).μ eo=
v eo
The EOF in CZE depends primarily on the pH of the BGE, the viscosity of the solution, and the charge of the capillary wall. The pH determines the charge density on the capillary wall. An important characteristic of the EOF is the flat flow profile along the capillary. This even migration of the BGE through the capillary minimizes band broadening due to longitudinal diffusion.
The apparent mobility ( μ app ), of an analyte is the sum of its effective mobility and the EOF.
μ app=μ eff +μ eof
(0)The magnitude of the EOF is often greater than the effective mobility of the analyte, consequently the analyte will move towards the cathode regardless of their charge. This enables the separation of positive, negative and neutral species in a single run.
The apparent mobility of an ion can be calculated from the length of the capillary up to the detector (
L det ) and the migration time( td ), as shown in formula (0).
μapp=
L det
td ∙ E
(0)The ionic mobilities can also be approached theoretically, this could be of value when comparing ion mobility of CZE with ion mobility in IMS. Ion mobilities can be calculated using the formula (0).
μ ,i=
λ 0 ,i
F
(0)Where
λ 0 , i
is the ionic conductivity and F the faraday constant.2.
Electrospray Ionization
When coupling a liquid based front-end separation technique such as CZE to IMS the analytes of interest must be converted from dissolved molecules to gas phased ions. This is often done by electrospray ionization (ESI), atmospheric pressure photoionization (APPI), or by means of radiation with for example 63NI or 241Am. The most common ionization technique for biomolecules is ESI. It utilizes a strong electric field to force a liquid to become a fine aerosol. The sequence of events happening in electrospray ionization can be explained in three consecutive steps.27
Nebulization
It starts with nebulization; nebulization occurs when an electrical potential is applied at the tip of a capillary or a needle with a constant flow of liquid running through. The electrical potential causes charged particles to accumulate at the tip of the needle transferring the meniscus into a stable Taylor cone. When the electrical potential exceeds the surface tension of the solvent, nebulization occurs, and the Taylor cone emits a jet of liquid droplets. These droplets are called parent droplets or initial droplets in Figure 1.
Figure 1: A schematic illustration of an ESI source operated in positive ion mode. from 31 Solvent evaporation and droplet disintegration
The solvent that makeup parent droplets start to evaporate, causing the droplets to shrink, increasing the electrostatic repulsion between the ions. When the electrostatic repulsion exceeds the surface tension of the liquid it will result in the formation of smaller charged offspring droplets (final droplets in Figure 1). This process can benefit from the addition of a heated inert gas, called nebulizer gas. The parent droplets will experience shear forces as they are propelled through the dense gas causing deformation of the droplets. Which creates excessive electrostatic repulsion in the protrusion region of the droplet resulting in in disintegration of the parent droplet adding to the formation of offspring droplets.
Ion ejection
The emission of intact gas-phase ions from offspring solvent droplets can proceed through different mechanisms. The three different mechanisms that have been proposed are; the ion evaporation model (IEM), the charged residue model (CRM) and the chain ejection model (CEM) illustrated in Figure 2.31 These three models provide a framework that is capable of accounting for a wide range of observations.
Figure 2: illustration of ESI mechanisms. (A) IEM: small ion being ejected from a droplet. (B) CRM: globular protein being released in the gas phase. (C) CEM: ejection of an unfolded protein into the gas phase. from 31
Low molecular weight analytes are believed to follow the IEM model. This model is based on the fact that the electrical field off a nanodroplet is high enough to lead to the ejection of solvated ions from the surface of the droplet. The ejected ion with its solvation shell can be interpreted as a very small offspring droplet. Less polar ions exhibit a stronger ionization efficiency which can be explained by the fact that these ions tend to position themselves closer to the droplet surface.
It is generally accepted that large globular analytes such as folded proteins are ejected into the gas phase through the CRM model. In this model the offspring droplets containing a single analyte evaporate to dryness transferring the charges of the vanishing droplet to the analyte. Leaving only the charged analyte to remain.
Konerman et al recently hypothesized the CEM model as a third mechanism, explaining the behaviour of large nonpolar polymer chains, such as denatured proteins in ESI.31,32 The hydrophobic character of this type of analyte makes it unfavourable that they reside in the centre off the offspring droplets. They migrate to the surface instead, where they are pulled out of the droplet by repulsive electrostatic forces while acquiring charge from the droplet as they leave.
3.
Ion mobility spectrometry
IMS is widely used to separate and identify ionized molecules in the gas phase based on their mobility in a carrier buffer gas. The mobility in the gas phase depends on de balance of two forces, the pulling force of an electrical field and the frictional drag force resulted from collisions with buffer gas molecules.33 The mobility of an ion is related to its size shape and charge. The first commercially available IMS devices were known as plasma chromatographs and hit the marked in the 1970s. 34 These instruments were expensive and bulky (1.5 m × 2 m × 2 m), but further advances where provoked, due to its quick measurement times and low limits of detections for certain analytes. In the following decades the dimensions where reduced to handheld and easily transportable configurations. Until three decades ago the applicability of IMS was solely as a vapor sensor.34,33, 35 The developments of ‘novel’ ionization techniques (ESI and MALDI) made it possible to analyse liquid and solid samples, which expanded the applicability of IMS. In addition, the development of high-resolution IMS 36,37 and separations with drift-gas selectivity 38 has permitted the separation of chiral compounds, 1 polymeric conformers2 and isomers 3,4 with speeds in the order of milliseconds.39 IMS coupled with MS, occasionally referred to as ion mobility mass spectrometry (IM-MS), is a powerful analytical tool for the separation of complex samples as well as the investigation of molecular structures. After the first IM-MS instruments came to the market the amount of scientific literature on the subject has grown significantly. Nowadays the IMS instrumentation has an ever-expanding range of applications from the detection of explosives to biological and clinical analysis. In recent years IMS has utilized for the application of complex biological samples where several analytes often have similar characteristics. Nowadays the broad range of IMS applications include among others: Analysis of warfare agents, 40 food safety assessment, 41 analysis of pharmaceuticals, 42 environmental analysis, 43 water analysis, 44 illicit drug analysis, 45, 46 oil and petroleum analysis, 47 and biomolecule analysis. 48 As well as applications in: metabolomics, 49 foodomics, 50 proteomics, 51 lipodomics, 52 and structural and analytical studies of biomolecular ions. 53 IMS can also be used for the continuous detection of airborne molecular contaminants 54 and inorganic species. 55
3.1.
Ion mobility Analysers
There are many modes of IMS each with their own characteristics, most common modes include: drift tube IMS (DTIMS), traveling wave (TWIMS), high-field asymmetric waveform IMS (FAIMS), differential mobility analysers (DMA) Trapped IMS (TIMS). 35, 56, 57 for ease of understanding all these modes of IMS can be divided in two categories: ones that use a static electrical field and ones that use a dynamic electrical field. In DTIMS and DMA a static electrical field is used for the separation of ions. The main difference between the two is the use of the gas stream (illustrated in Figure 3) and that DTIMS separates ions in time and DMA separates ions in space.
FAIMS, TWIMS and TIMS all utilize dynamic electrical fields. FAIMS uses asymmetrical electrical field strengths in different directions for different amounts of time to filter the ions in space. TWIMS utilizes waves of electrical potentials to migrate the ions through a buffer gas to separate ions in time. Whereas TIMS traps the ions in a flowing buffer gas. The ions are trapped axially by an axial electrical field gradient and radially by using alternating radio frequency potentials. To detect all present ions TIMS must scan over the different mobilities.
Figure 3: Difference between 5 forms of IMS. Above IMS analysers that use a static electrical field. Below: IMS analysers that utilize a dynamic electrical field for the separation of ions. The movement of the buffer gas is shown as a purple arrow. The higher mobility ion is shown in green and the lesser mobile ion is shown in blue. From 5652
3.1.1. Drift Tube Ion Mobility Spectrometry (DTIMS)
The simplest form of IMS is drift tube IMS. In DTIMS ions are introduced into a tube which consists of a series of stacked electrodes filled with a buffer gas. The application of a weak homogeneous electrical field causes the ions to drift through the tube. The IMS system measures how long an ion takes to pass a fixed distance through the buffer gas. 58, 59 Ions with larger rotationally averaged collision cross sections (CCS) will experience more collisions with the buffer gas causing them to migrate more slowly. The CCS averaged over all orientations represents the effective area of the gas phase ion and is a function of the structure or conformation of the analyte. The force applied by an electrical field on the analyte is balanced by the drag of the buffer gas, resulting in a steady-state velocity (
vd
). The ion mobility (K
) can be defined as a measurement of friction linked to the observed velocity (or distance / time). The ion mobility ( K ) can be determined experimental from the drift time (td
) of an ion and the length (L
) of the mobility tube, as shown in formula (0).K=
vd
E
=
L
td ∙ E
(0)Ion mobilities are commonly reported as reduced mobilities (
K 0
). Wherein the mobilities are corrected to the standard gas density ( n 0 ) (2.687 * 10E25 m-3) expressed in the standard temperature (T 0
) (273K) and the standard pressure (p 0
) (1013 hPa). The definition of the reduced mobility is shown in formula (0).K 0=K
n
n 0
=
K
T 0∙ p
T ∙ p 0
(0)K 0
Does solely correct for the difference in gas density and is still (analyte) temperature dependent. Under atmospheric or close to atmospheric pressures, Ω the momentum transfer collision integral can be assumed to be identical to the CCS. The CCS (Ω) can then be equated to K0 using the Mason-Shamp equation presented in formula = (0).CCS=Ω=
1
K 0
3 ze
16 N
[
2 π
μ KbT
]
0.5 =K 0
1
(18 π )
0.516
ze
( KbT )
0.5[
1
mi
+
1
mb
]
0.51
N
(0) Wherein Kb is the boltzman constant, mi is the mass of the ion, mb the buffer gas mass,ze
the analyte charge state, and the number density of the drift gasN
. This equation is applicable for low electrical field to buffer gas ratios ( E / N ratio <2 Townsend). At high E /N ratios there is a non-linear relationship between Ω and
1
K 0
caused by among other thingsion heating.
The Mason-Shamp equation can be simplified by regrouping the constants in the formula, giving formula (0), where only
z
,K 0
,T
, the mass of the ion (mi
) and the mass of the collision gas ( mb ) are needed to calculate the CCS. 60CCS=18500
z
K 0
[
T
mi mb
mi+mb
]
0.5 (0)And K 0 can be expressed as shown in formula (0).
K 0=
3
16
[
2 π
μ KbT
]
0.5ze
N ∙ CCS
¿
18500
z
CCS
[
T
mi mb
mi+mb
]
0.5 (0)Wherein µ is the reduced mass of the ion-gas pair.
The buffer gas choice in IMS has a great influence on the resolving power. Traditionally inert buffer gasses such as helium and nitrogen have been used for IMS, they have been extensively studied and can be used to determine the CCS of analytes.
Resolution and resolving power of DTIMS instruments
The resolving power can be determined experimentally, shown in the first part of equation (0). 61 Wherein td is the ion drift time and FWHM is the full width at half maximum for the peak of interest. Additionally the resolving power can be approached theoretically when diffusion is assumed as the
sole contributor to peak broadening, shown in the second part of equation (0) 62. The diffusion limited resolving power is dependent upon the length L of the drift tube, E the electric field strength, Q the charge of the ion of interest, k the Boltzmann constant and T the temperature of the drift gas.
Rp=
td
FWHM
=
√
LEQ
16 kT ln 2
(0)Generally ambient pressure methods allow for greater separation selectivity and higher resolution as a result of higher interaction rate of the ion and the drift gas. Experimentally acquired Rp of specific analytes can be used to compare instrument performance between laboratories due to its dimensionless nature.61 It should be noted that in equation (0) Rp is calculated based on one peak. If the separation performance between two ions is of interest the resolution can be calculated as means of direct comparison between two peaks. The resolution is defined as shown in equation (0).
¿
ta−tb∨
¿
Wb, a+w b , b
R=2 ∙
¿
(0)
Where R is the resolution, Ta and Tb are the drift times of the two ions of interest and Wb, a and Wb, b is the peak with at the base of the peak in seconds.
Regular values attained with high resolution ion mobility apparatuses vary on the nature of the analysed ion. For single charged ion kirk et al.63 reached a resolution of R=183, witch at the time was the highest reported resolution for a singly charged ion, albeit a resolution of 173 had been shown before. Better resolution values are attainable when multiple-charged ions are analysed. Puton et al reached a resolution of R= 172 when measuring a four times negatively charged ion. 40 Higher resolution are possible with continuous IMS approaches such as, TIMS or cyclone overtone ion mobility spectrometry. When measuring multiple charged larger ions resolutions of R= 250 and R=1000 are achievable using lengthy analysis times.63
3.1.2. Traveling wave ion mobility spectrometry (TWIMS)
TWIMS is different from classical DTIMS in that it does not use a homogeneous electrical field over the whole length of the tube, but instead it creates waves along a stacked ring ion guide, as shown in Figure 4. 27, 35, 64 This is achieved by raising the voltages of selected electrodes. The raised potential then propagates through the ion guide by switching to the next electrodes. The potential waves continually propagate through the tube propelling the ions forwards through a stationary buffer gas. In the Synapt (system by Waters) the voltage is raised on two electrodes spaced by 5 electrodes. In addition, adjacent electrodes carry alternating RF voltages that create a radially confined potential barrier resulting in ion focussing, suppressed diffusion and coulomb expansion. In TWIMS the experimental variables consist of the wave height and – speed, the make-up of the gas and its pressure. If the gas pressure is sufficiently low the ions experience few or no collisions and the device functions as an ion guide. If the gas pressure is higher, the ions experience more collisions and can be separated according to the transit time, based on the fact that larger ions undergo more collisions than smaller ions. Because the electrical field in TWIMS is not linear the analytical equations
describing ion movement are complicated. In practice, determination of CCS of unknowns has to be done by calibration using similar molecules with known CCS’.65
Figure 4: the difference between DTIMS (A) and TWIMS (B) illustrated. High mobility ions are in green, low mobility ions are in orange. In DTIMS a homogeneous electrical field gradient is plotted along the tube length. In TWIMS (B) the voltage wave is traveling to the exit. Higher mobility ions are picked up more easily, whereas larger ions experience more drag from the buffer gas and reach the exit of the mobility cell later. Reproduced from 65
3.1.3. High field asymmetric waveform ion mobility (FAIMS)
Field asymmetric ion mobility spectrometer (FAIMS), sometimes referred to as differential mobility spectrometer (DMS), separates ions based on the difference in an ion’s mobility through a gas at high and low electrical fields. FAIMS is a spatial based separation device rather than a time-based separation device such as DTIMS and TWIMS.56 In FAIMS, a flow of carrier gas carries the gas phased ions between two electrodes towards a detector. Asymmetric electrical fields(waves) are generated that alternate between a high field voltage in one polarity and a low field voltage in the opposite polarity.66 The duration of the high field (wave) is shorter than that of the low field. Typically, the low field portion is half the magnitude and twice the duration of the high field portion. As the polarity of the electrical field oscillates the ions will move towards one electrode or the other, while carried forwards by the carrier gas. An ion will exit the FAIMS device if there is no difference in its mobility in the high or low field as there will be no net migration towards one of the two electrodes. If an ion’s mobility differs in a high or low field, the ion will hit one of the electrodes and cannot be detected. To change the subset of ions that can migrate through the FAIMS device a small direct current compensation voltage can be added to the waveform. A schematic overview of the workings of a FAIMS analyser is shown in Figure 5.
The main advantage of FAIMS is its ability to continuously analyse ions, which allows real time monitoring of samples.35 another advantage is the possibility to detect positive and negative ions simultaneously, at the cost of increased complexity. As with TWIMS measured mobilities cannot be directly related to its CCS and require calibration curves of similar ions with known CCS’ for the determination thereof.
Figure 5:schematic overview of the workings of a FAIMS analyser. An oscillating electric field E from a low level El to a high level Eh is applied perpendicular to the path of the ions. from 35
3.1.4. Trapped IMS (TIMS)
TIMS’ relies on the use of a non-uniform electric field to keep ions stationary in a gas flow.35 The force of the electrical field and the opposite frictional force of the gas on an ions cancel each other out keeping the ions in an equilibrium position in the analyser. The equilibrium of the ions in the analyser happens when the velocity of the ions through the gas ( vd ) is equal to the velocity of the buffer gas (
vg
).0=vd +vg (0)
The fundamental physical principles behind TIMS are equal to those behind DTIMS. A difference with DTIMS is that in TIMS the buffer gas flows towards the detector and the ions are held back using an electrical field. Nevertheless the same basic concepts for the determination of K, K0 and CCS still apply.60
Figure 6: Above: schematic representation of a TIMS analyser for sequential analysis with a single storage region. Below: illustrative electric field plot showing detailed steps of the separation. Analytes with low ion mobilities are blue. Analytes with high ion mobilities are in orange. From ’60
A typical TIMS analyser consists of three main regions, illustrate in Figure 6. 60, 35 The entrance funnel, TIMS tunnel, and the exit funnel. A capillary introduces ions and gas from, for example an ESI source, in between a deflection plate and the entrance funnel. Ions of the right polarity are deflected out of the gas stream into the entrance funnel. The ions are then focussed into the separation section (TIMS tunnel). In the separation section the electrical field increases along the axial section. Separating the ions based on their mobility and thus their size, shape and charge. Ions with lower mobilities will be trapped further in the trap where the magnitude of E is larger, because the electrical field required to migrate an ion through a gas scales linearly with the mobility K, as shown in Formula (0). After equilibration of the ions special position, the electrical field is reduced, and the ions elute from high to low size to charge ratios. To confine the ions radially in the analyser a pseudopotential RF voltage is applied to the electrodes. 60 This radially confining RF field has no axial component and does not interfere with the separation it merely results in ion focussing, suppressed diffusion and suppressed coulomb expansion. The main motion defining parameters in a TIMS analyser are the drift gas velocity, the ion confinement and the electrical field ramp speed.
4.
CZE IMS orthogonality
Ion mobility spectrometry, formerly referred to as plasma chromatography or gaseous electrophoresis. As the name implied shares commonalities with capillary zone electrophoresis. Both techniques utilize an electrical field to propel and separate ions through a separation fluid (liquid in CZE, gas in IMS). with both techniques the separation efficiency is proportional to the applied electrical field and inversely proportional to the diffusion coefficient of the analytes. 33
In IMS the ion mobility K depends on its charge and is inversely proportional to its reduced mass and CCS.
Ion Mobility IMS ( K )=
charge
Reduced Mass ∙CCS
(0)In CZE the ion mobility depends on the net charge of an ion and is inversely proportional to its friction coefficient in solution which is determined by the mass, shape and hydrodynamic radius of the ion in solution.
¿ ¿
Ion Mobility∈CZE=
charge
¿
(0)Separation efficiency and selectivity in CZE can further be impacted by the altering the properties of the separation buffer, such as pH, viscosity, ionic strength and temperature. As well as the properties of the capillary inner surface.
According to Mironov 27 the main difference in the techniques is the phase in which the separation occurs, solvated analytes in CZE exist in multiple co-migrating charged states, while in IMS mobility is assigned to an ion with a specific charge. Zhong et al 67 elaborates on this by stating that CZE and IMS could be orthogonal separation dimensions due to the difference in parameters that affect the “charge-to-shape ratio” in the gas phase mobility and aqueous phase mobility of an analyte. In CZE the net charge of an amphoteric compound is determined by the pKa values of the compound and the pH of the BGE and is carried by a multitude of comigrating molecules. The shape and size of the analyte in CZE refers to the apparent shape which refers to the hydrated aqueous-phase conformation of the analyte in buffer solution. Before entering an IMS analyser the analytes are converted to the gas-phase with for example ESI. In the gas-phase the charge is based on the number of charge carriers (e.g. H+) attached to the analyte. The ‘shape’ or reduced mass * CCS of an analyte in the gas phase refers to the gas phase confirmation of the analyte including adducts.
5.
CZE and IMS coupled systems
Although IMS has been used for decades its application in conjunction with CE is limited, the first documented attempt to couple CE and IMS was performed by the Hill group in 1989 who developed an ESI source for the IMS instrument. 68, 69 The Hill group focussed mainly on the coupling of CE an IMS and investigated several electrospray interfaces using standard CZE and IMS methods. These preliminary experiments were encouraging; however, the initial coupling was hindered due to spray instability and the operational robustness of CE. Problems with electrical circuit enclosure and sample dilution caused by the CE method obstructed the widespread coupling of CE and IMS. 33 in 2011 jiang et al.70 combined CE and IMS to develop a sheatless ESI interface coupling. An effective coupling was created by manually gridding the end of the capillary with emery paper to obtain a tapered tip, the tip was then coated with nickel to be electronically conductive. Jiang et al.70 demonstrated that CE and IMS could be coupled using a simple and inexpensive interface. However, in this article the coupling is only tested with injections of DMSO to evaluate performance which does not fully represent real world applications.
The application of IMS and the hyphenation of IMS with front end separation dimensions such as CZE, in literature has increased thanks in part to commercially available IMS systems. Early hyphanatable commercially available IM-MS devices include: A TWIMS-MS device by waters (Synapt) in 2006, followed by a DTIMS-MS (IM-QTOF) device brought to the market in 2014 by Agilent, and Bruker who brought a TIMS-MS (TimsTOF) to the market in 2016.
In table 1 a complete overview is given with studies using both CZE and IMS, coupled or not coupled. As shown in table 1 all coupled systems use ESI with sheath liquid except the studies looking specifically into the coupling of CZE and IMS with a simplified coupling interface. The hyphenated CZE-IMS methods all use either TWIMS or DTIMS with the exception of Li et al 71, where they utilize a FAIMS system. In this study the sole purpose of the implementation of the FAIMS was to filter out the background and enhance the detection limits. FAIMS was not used as a secondary separation dimension.
Table 1: complete overview of studies using both CZE and IMS. Front end Separation Ionization method (interface) XIMS Analyser (Trade Name)
Matrix Analyte Year Ref.
CZE various electrospray interfaces
DTIMS Picoammeter - caffeine, tetramethyl ammonium iodide, trimethylphenyl ammonium iodide and tetra butyl ammonium iodide. 1989 68 CZE ESI sheath liquid FAIMS QQQ (Api 3000) H. influenzae strain 375 O-Deacylated Lipopolysaccharides 2004 71 CZE ESI Sheatless DTIMS - DMSO 2011 70
CZE Not coupled TWIMS Q-TOF (SYNAPT G2) Synthesized: Glycine-modified polyamidoamine dendrimers 2014 72 CZE (KCE) ESI sheath liquid TWIMS MS (Synapt G2)
Purified DNA Conformational Dynamics of DNA G-Quadruplex
2014 73 CZE ESI sheath
liquid TWIMS Q-TOF (Synapt G2) human serum protein digests aminoxyTMT-Labeled N-Glycans 2015 67 CZE (KCE) ESI sheath liquid TWIMS MS (Synapt G2) Purified human tissue transglutaminase (TG2) Conformations of human tissue transglutaminase (TG2) simultaneously with their enzymatic activities
2016 74
CZE (KZE)
Not coupled DTIMS DTIMS-TOF (home-build) Synthesized polyproline peptide Following a Folding Transition of a 13-mer polyproline peptide 2016 75 CZE ESI triple-tube sheath liquid interface DTIMS Q-TOF (Agilent 6560) Fetuin samples (fetal calf serum)
Native and APTS-labelled N-glycans 2018 76 CZE ESI Sheath liquid DTIMS A series of tetraalkylammonium bromide salts 2019 77
5.1.
FAIMS to reduce background noise
In 2004 Li et al.71 Developed a method for the analysis of liposaccharides with the use of CE-FAIMS-MS, demonstrating the potential of CE-FAIMS-MS to separate a complex mixture of liposaccharides. Although FAIMS was solely used to filter the chemical background and enhancing the sensitivity. 33,59 All three separation dimensions obtained a linear dynamic range of three orders of magnitude (0.04 to 10 µg/ml). Li et al. concluded that FAIMS is ideal for CE-MS because of its ability to act as an ion filter and continuously transmit one type of ion, independent of mass to charge ratio. FAIMS provides electronically controlled separation for gas-phased ions at atmospheric pressure, based on changes in mobility that occur at high electrical field strength. In the CE-MS and CE-FAIM-MS data are compared. A clear decrease in background noise can be seen in the CE-FAIM-MS data.
Figure 7: spectra of O-deacylated LPS (1 ng/µl) (a) CE-MS data (b) CE-FAIMS-MS data. Top panels show total ion extraction spectra, with the EIC of m/z 962.5. middle panels show the EIC between 6.5 and 7.1 minute. Bottom panels show the EIC between 6.8 to 7.1 minute. from 33,71
5.2.
CE-TWIM-MS analysis of native and labelled glycans
In 2015 Zhong et al.67 published an article describing the development of a quantitative method of glycans labelled with multiplex carbonyl-reactive tandem mass tags using CE-MS and CE-TWIM-MS. They analysed four isobaric labelled human milk oligosaccharides using both methods. As shown in spectrum A of Figure 8, LSTb and LSTc co-elute after the CE-separation. By integrating the TWIMS separation dimension they were able to separate LSTb and LSTc. They concluded that CE coupled with TWIM-CID MS/MS was demonstrated to be efficient in resolving human milk oligosaccharide isomers.
Figure 8: (A) CE-MS data, extracted ion chromatogram of [LST-aminoxyTMT + H + Na]2+ m/z 661.9 and [DLST-aminoxyTMT +
H + 2Na]3+ m/z 545.9 . (B) MS data, arrival time distribution of m/z 661.9 at different CE migration times. (C)
CE-IM-CID MS/MS data, daughter ions of m/z 661.9 (from top to bottom) LSTb, LSTc, LSTa. from 67
Although both separation dimensions are electrophoretic based separation techniques. Zhong et al. writes that the dimensions could be orthogonal, because the aqueous phase mobility and gas phased mobility of analytes are dependent on different parameters that affect the charge-to-shape ratio of the analyte in a specific environment. The average charge of an amphoteric compound in CZE is determined by the pH of the BGE and the pKa of the compound and the shape of a molecule in CE refers to the confirmation of the molecule in solution including the hydration shell and possibly co-migrating molecules. In IMS the number of charges is dependent on the charge carriers obtained during ESI and the shape refers to the gas phase confirmation including adducts. Although there are differences in CE and IMS, calling them orthogonal, or separation techniques based on fundamentally different mechanisms, would be incorrect. Nevertheless Zhong et al. has demonstrated that in certain situations coupling CE-IMS can produce complementary information.
5.2.1.
CZE-DTIMS for native and labelled glycans.
In 2018 Jooß et al.76 used CE–DTIMS-MS for the analysis of native and APTS-labelled N-glycans. The online coupling has been performed to improve the separation capability, necessary due to the high structural variability of native and APTS labelled glycans, which could not be resolved by these techniques individually. Being able to successfully separate and analyse glycans is important, due to the extensive role they play in eukaryotic systems, it is estimated that 50% of all proteins in in these systems are glycosylated.78 Many biological functions and involvement in development and progression of human disease have been associated with protein glycosylation.79 Glycoproteins originated in living cell systems often exhibit a high degree of heterogeneity. Glycosylation can occur on different sites, in general on the asparagine (N-glycans) or the serine/theanine site (O-glycans), different numbers of glycans can be attached to a protein, and there is a variety of different glycans structures occurring. Hence the challenges in glycan characterization and separation, usually obligating several orthogonal techniques. 76
The IMS used was an Agilent 6560 IM-QTOF, for the ESI coupling a commercially available triple-tube sheath liquid interface was used. Different sheath liquid compositions were investigated to optimize the ionization efficiency. Interestingly they found that the nature and type of the sheath liquid did not noticeably affect the arrival time distribution in DTIM-MS, hence the ionization efficiency could be optimized independently from the DTIM-MS dimension. Remarkedly every individual glycan peak separated in the CZE dimension presented an unexpectedly high amount of peaks wit IMS. Jooß et al. remarks that this could be explained by the presence of isomeric forms along with different linkages, or gas-phase conformers that do not interconvert on the time scale of the IMS separation. The combination of CE-DTIM-MS could somewhat resolve complexity of N-glycans that could not be resolved by DTIM-MS or CZE alone. This is shown for instance with the separation of APTS labelled G1F (1,3)/G1F(1,6) isomers shown in ′ Figure 9 as well as other glycans further discussed by Jooß et al. These examples demonstrate advantages of the hyphenation of these techniques.
The researchers in question emphasise the statement that they were first to successfully couple CZE to DTIM-MS to further enhance separation capabilities for the analysis of native and APTS labelled glycans. Jooß at al. concludes that this work demonstrates the high potential of CZE-DTIM-MS and expects it to have various applications in the future.
Figure 9: (A) CZE electropherogram 17 to 23 minutes (B) 2D heatmap m/z to drift time of double (red) and triple (orange) charged analytes (C) arrival time distribution of triply charged species and (D) doubly charged species. Analytes: APTS-labelled N-glycans: G0F (950.75, 633.50 m/z), G1F (1,3) and G1F(1,6) (1031.77, 687.51 m/z), and G2F (1112.80, 741.53 m/z).′ from 76
5.3.
Characterization of polyamidoamine dendrimers
Leriche and colleges 72 synthesized and characterized glycine-modified polyamidoamine dendrimers (PAMAM dendrimers) with among other things CE and IM-MS. The goal of the characterization was to understand and interpreted their physicochemical and biochemical properties as well as to control the chemical design. The objective with the CE and IM-MS analysis was to semi-quantitate the different species and estimate the average number of grafted glycine units. CE was utilized to separate ammonium cored PAMAM dendrimer GlynG1(N) (n = 0-6) oligomer ions in condensed phase
and IM-MS to further separate them in the gas phase. In addition, IM-MS was used to estimate the CCS to compare them to theoretical values.
Figure 10: comparison of (A) and (B) CE electropherograms and (C) and (D) ion mobility spectra (arrival time distributions). Of ammonium cored PAMAM dendrimer GlynG1(N) (n = 0-6) oligomer. Peak identification: GlynG1(N) (n = 0-6), Peak a: n = 0,
peak b: n= 1, etc. insert in (C) shows doubly charged ion distribution of said analytes. (B) and (C) show an electropherogram and arrival time distribution of Gly6G1(N).
As shown in Figure 10, A and B, the CE separation did not resolve individual species making peak integration not feasible and impeding the estimation of the number of grafted glycine units. The alternative method for the investigation of the GlynG1(N) distribution was IM-MS, in these experiments the Synapt G2 (Waters) traveling wave IMS was used. Two ion-mobility spectra are shown in Figure 10, C and D, it shows that doubly charged GlynG1(N) are well separated from other multiply charged ions. Two dimensional ion-mobility (drift time) vs mass spectrometry (m/z) spectra have been constructed which enabled differentiation of the seven GlynG1(N) (n = 0-6) oligomer ions (inset in Figure 10 C). Comparing the CE electropherograms to the IMS spectra, what stands out is that the obtained migration orders were the same. The separation of different dendrimers was only
achieved by using a combination of IM-MS involving the combination of IMS and MS data. The CE method could not differentiate al the dendrimers for it was not coupled to an MS. However, the CE analysis was performed in the minute time scale while the IM analysis was in the millisecond time scale. The CCS of the dendrimers were measured using polyalanine for the calibration of the IMS cell. The CCS was calculated as demonstrated in the protocol for IM-MS analysis of large protein complexes described by Ruotolo et al.80
5.4.
CE-IMS for conformational studies
In 2014 Mironov et al.73 Used CE coupled to IMS-MS to study the conformational dynamics of DNA G-Quadruplex in solution. The challenge in monitoring the conformation of DNA is the fact that they have the same nucleotide sequence and constitution, and therefore the same molecular weight. The differential affinity of different DNA conformers to potassium ions can aid CE separation and has a stabilizing effect. In Figure 11 results are shown with on the x axis the CE migration time and on the y-axis the IMS drift time. The different DNA conformations (GM and GQ) are after the addition of KCl well separated in the CE dimension and poorly in the IMS dimension. The addition of potassium ions had a large effect on the CE separation wile it did not aid the IM separation significantly.
Figure 11: Kinetic CE-IM-MS data of the separation of GM and GQ DNA sequences. CE migration time on the x-axis and IM drift time on the y-axis. GM and GQ conformations are well resolved in the CE dimension and poorly by the IMS dimension. from 73
5.5.
CE-IMS Protein conformation and enzymatic activity
In 2016 Mironov et al.74 developed a method using capillary electrophoresis coupled with UV detection and IMS-MS for the detection of protein conformational isomers, their enzymatic activity and their inhibition in a single experiment. The protein conformers were separated, and their interconversion dynamics were monitored using kinetic CE-UV. IM-MS was used to measure the conformers sizes, exact molecular weights and the structures of the enzyme and its substrate. The coupled CE-UV-IM-MS method made it possible to observe the effect of small-molecule inhibitors on the enzymatic activity and the conformational distribution in real time. The enzyme studied was human tissue transglutaminase TG2. The method was based on an earlier study where the researchers used kinetic CE-UV for the monitoring of the confirmation of TG2 enzyme. 81 Mironov et al.74 writes that CE-UV-IM-MS opens a new pathway for the regulation and modulation of cellular functions as well as drug development and protein engineering.
Kinetic CE is a form of CE whereby molecules that interact in the capillary during electrophoresis are separated. An advantage of kinetic CE-UV-IM-MS over the more traditional Kinetic CE with fluorescence (laser induced) or UV-detection is the increased selectivity which allows for less problematic simultaneous analysis of enzymes, substrates and reaction products. An advantage of UV detection is that the absorption is directly proportional to the concentration of the analyte protein. This makes it possible to determine the concentration of each protein conformer, as long as they are spatially separated. The disadvantage of only using UV-detection is that it is difficult to detect and identify small molecules at low concentrations or in complex samples. The high selectivity of MS makes it a more suitable detection technique for complex samples. Ion mobility was used to compare the relative sizes of different conformers carrying the same charge. A schematically representation of the method is shown in Figure 12.
Figure 12: (A) representation of the CE-UV-IM-MS method. A, P and D represent the acceptor substrate, the donor substrate and the product. TG2-O is the open confirmation of the TG2 enzyme (protein Y) and TG2-C is the closed confirmation of the TG2 enzyme (protein X). from 74
Ion mobility results are shown in Figure 13. Mironov et al. writes about the results shown in Figure 13 A that baseline separation of the two conformers was prevented on extracted ion electropherograms, because of poor ionization efficiency and ion loss in the ion mobility chamber. What can be seen from Figure 13 A is that the last chunk of the peak had a lower drift time which should be the result of a lower CCS. The lower resolution of the IM separation compared to the CE separation is a result of
multiple factors: for one the IM buffer gas is less dense than the BGE in the CE experiments, diffusion of molecules in the gas-phase is much higher than diffusion in the liquid/solvated state. Another factor is the stronger effect of coulomb repulsion of ions in the gas-phase caused by the absence of solvating molecules. Water molecules play a key role on the confirmation stability of aqueous solvated proteins. Ejection of these water molecules during ESI could lead to conformational changes inducing non-native unfolded structures, hindering the IM separation of an unfolded and folded protein.
Figure 13: CE migration time (x-axis) and IMS- arrival time distribution (y-axis) of TG2-O (open confirmation of human tissue transglutaminase) and TG2-C (closed confirmation of human tissue transglutaminase). (A) the dashed lines represent peak apexes. from 74
There are multiple parameters that limit this CE-UV-IM-MS method. To detect conformational changes of the enzyme caused by, for instance ligand binding interaction, the conformational changes must be substantial enough for CE or IM separation. the conversion must happen within the timeframe of the analysis and for the IM separation the conformations of the proteins ought to be stable in the gas phase.
5.6.
Conformation of polyproline with CE and IMS separately
In 2016 Barr et al.75 used IM-MS and CE-UV separately to follow a folding transition. In this experiment the folding transitions of a 13-mer polyproline were followed from the all-cis polyproline (PPI) to the all-trans polyproline (PPII) form upon immersion in aqueous solvent. IMS was used to follow the transition in the gas phase and CE was used to follow the transition in solution. Pro-13 was synthesized and dissolved in excess 1-propanol and incubated for 72 hours to generate the PPI conformation. After which the solution was dried and dissolved in a water containing solution to initiate the transition to the PPII confirmation. The dried pro-13 was dissolved in two different concentrations, to facilitate the lower sensitivity CE-UV method. At defined time intervals during the transition, aliquots were analysed with IM-MS and CE-UV. The results are shown in Figure 14.
Figure 14: (A) IMS data: Collisional cross section distribution for Pro-13, obtained at different transition times from a PPI(s) to PPII(aq) confirmation (B) CE-UV data: converted to electrophoretic mobility of Pro-13, obtained at different transition times from a PPI(s) to PPII(aq) confirmation. (C) CE separation of said conformational transformations with underlying peaks shown in colour. (D) comparison of relative abundance of conformers monitored with IMS (filled symbols) and CE (hollow symbols). from 75
A homemade multifield drift tube/ion funnel design IM-MS instrument has been used for the IM-MS experiment, earlier described by Koeniger et al. 82 A difference in peak order can be seen from Figure 14 (A) and (B), the shift can be caused by the solvation conditions or the movement of bonds beyond the peptide bond. An additional variable causing a divergent peak pattern can be a difference in net charge in the solvated state as opposed to the gas phase. Fewer conformers are resolved with CE than with IMS. Five peaks were partially resolved using the CE method and eight with IMS. The CE and IMS, solvated and gas phased, analysis showed similar conformer patterns as a function of time. Barr et al. have demonstrated that IMS gas phase measurements reflect solution phase populations for the polyproline system.
6.
Discussion
The application of coupled CE-IMS methods is quite recent while the earliest preliminary study dates from 1989. The first scientific literature where CE-IMS was applied was published in 2004 wherein IMS was being used as an ion filter. 68, 71 Two years later (2006) Waters first shipped their Synapt HDMS, a device combining TWIMS and MS.62 Since then new commercial IM-MS instruments have developed, waters launched their second and third generation of TWIMS-MS equipment (Synapt G2 and Vion IMS QTOF). Meanwhile other manufacturers have introduced IM-MS devices based on other IMS modes. Agilent Technologies introduced an IM-MS device based on DTIMS, Thermo Fisher Scientific introduced a FAIMS device and Bruker developed a device based on TIMS. As a consequence of these advances it can be expected that there are many new applications being developed of which CE-IM-MS applications that could establish a foundation for further implementation of CE-IM-MS or CE-IMS in laboratories if it proves to be successful. The majority of published applications that use a coupled CE-IM-MS system focus on macromolecules of which most studies are conformational, characterization or semi-quantitative studies. When looking at the conformation of macromolecules it seems that coupling CE to IM-MS enhances the separation capabilities allowing for the separation of conformers that could not be separated with CE alone as shown by [67, 74, 76]. Which brings rise to the question if those conformers could be separated using IM-MS alone, Leriche and colleges 72 and Barr et al.75 have shown that for the characterization of ammonium cored PAMAM dendrimers and following a folding transition of a 13-met polyproline respectively. IM-MS performed as a better separation technique for these specific samples than CE-UV and performed in the order of milliseconds while the CE separation took minutes. However, the samples used in these studies were synthesized and purified analytes which do not represent the complex samples dealt with in other fields of research such as the -omics and biomedical analysis. In those samples a front-end separation will be necessary to decrease ion suppression and increase peak capacity. When choosing a first and a second separation dimension for a certain application it is wise to opt for the most orthogonal option that works for the specific application. In practice it has been shown that the separation order of analytes analysed with CE and IMS show similar patterns. 72, 75 This is not an unexpected result considering both separation mechanisms are based on size, shape and charge. Size and charge in the liquid phase can differ from size and charge in the gas phase after ESI which makes it possible to obtain additional information after CE-IM-MS as opposed to CE-MS. Zhong et al. 67 showed that labelled glycans that could not be resolved with CE-MS could be resolved with CE-TWIMS-MS (figure 8). Additionally Jooß et al.76 showed that additional information could be obtained on labelled glycans using CE-DTIMS-MS opposed to CE-MS (figure 9). The additional selectivity often comes at the cost of sensitivity resulting from ion losses at the IMS-MS junction.83 These sensitivity issues have been minimized over the last decade by the development of more efficient ion funnel designs, and might be minimized even further in the future. Li et al.71 showed that IMS can be used to enhance sensitivity by using FAIMS as a background filter increasing the S/N ratio, instead of an additional separation dimension. The CE-FAIMS-MS method showed great improvement over the CE-MS method in terms of sensitivity.
Overall there are few scientific studies published wherein CZE and IM-MS are coupled which in part is due to the novelty of commercially available IM-MS systems. An additional explanation for the lack of studies is that more prevalent front-end separation techniques such as LC or SPE shown a higher degree of orthogonality with IMS and are therefore a more obvious candidates for the coupling with IM(-M)S for most applications.
It has been shown that the determination of CCS can be very advantageous for qualitative determination studies, but this parameter also has potential in quantitative studies as peak determination tool as CCS is far more precise than retention time and ion ratios can vary due to ESI instability. 49 CCS is a parameter based on fundamental characteristics of the analytes and should be reproducible between laboratories. CCS in combination with m/z values offer a great benefit in the identification of compounds. The main drawback today with CCS determination is the absence of automation in calibration for the determination thereof, the lack of standardized methods to obtain CCS’ and the lack of databases with known CCS’ based on those standardized methods.
Another application of IMS is as a standalone detector for CE, not coupled to a mass spectrometer. Guo et al. showed the feasibility of this application with a series of tetraalkylammonium bromide salts.77 IMS has a rapid analytical response and a high sensitivity. Furthermore IMS devices can be portable and affordable which are perfect features for laboratories with lower budgets. 41 CE-IMS could prove useful as a replacement for UV detectors (CE-UV). The main advantage being that compounds without UV absorption can be ionized and detected. Moreover, IMS provides an additional qualitative information besides the retention time and a second separation dimension which could result in faster method development and analysis times respectively.
7.
Concluding remarks
IMS offer an amplitude of configurations, applications and possibilities in the field of analytical chemistry. IMS was once restricted to small molecules. However, nowadays with the advances in ionization methods IMS can be seen as a general analytical method. It can be utilized as a standalone detector; it can be hyphenated with MS and it can be coupled to a broad array of front-end separation techniques using modern ionization methods such as ESI. The hyphenation of IMS with MS is often vital for the analysis of complex samples IMS can either be used as a filter to increase MS sensitivity as done by Li et al. 71 or to enhance the resolution of compounds with the same m/z and slight differences in there conformation. It is shown in the literature discussed in section 5 of this review that CE-IM-MS can be more effective than CE-MS due to IMS’ ability to effectively separate conformers. The ability to separate conformers seems to be the best advantages of CE-IM-MS today. The lack of reproduced studies provides little evidence for hard conclusions, but it seems that CE-IM-MS can provide additional resolution with minimal loss of sensitivity as opposed to CE-CE-IM-MS.
