Continuous-Flow Microelectroextraction for Enrichment of Low

Subscriber access provided by UNIV NEW ORLEANS

Article

Continuous-flow micro-electroextraction for enrichment of low abundant compounds Jan-Willem Schoonen, Vincent van Duinen, Amar Oedit, Paul Vulto, Thomas Hankemeier, and Petrus Wilhelmus Lindenburg Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/ac500707v • Publication Date (Web): 03 Jun 2014 Downloaded from http://pubs.acs.org on June 14, 2014

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Continuous-flow micro-electroextraction for enrichment of low abundant compounds Jan-Willem Schoonen1,2, Vincent van Duinen1,2, Amar Oedit1,2, Paul Vulto1,2, Thomas Hankemeier1,2 and Petrus W. Lindenburg1,2* 1

) Division of Analytical Biosciences, Leiden Academic Center for Drug Research, Leiden

University, the Netherlands 2

) Netherlands Metabolomics Centre, Leiden, the Netherlands

*) to whom correspondence should be sent P.W. Lindenburg, PhD Division of Analytical Biosciences Leiden Academic Centre for Drug Research Einsteinweg 55 2300 RA Leiden The Netherlands [email protected]

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 34

Abstract

We present a continuous-flow micro-electroextraction flow cell that allows for electric field enhanced extraction of analytes from a large volume (l mL) of continuously flowing donor phase into a micro volume of stagnant acceptor phase (13 µL). We demonstrate for the first time that the interface between the stagnant acceptor phase and fast-flowing donor phase can be stabilized by a phaseguide. Chip performance was assessed by visual experiments using crystal violet. Then, extraction of a mixture of acylcarnitines was assessed by off-line coupling to reversed phase liquid chromatography coupled to time-of-flight mass spectrometry, resulting in concentration factors of 80.0 ± 9.2 times for hexanoylcarnitine, 73.8 ± 9.1 for octanoylcarnitine and 34.1 ± 4.7 times for lauroylcarnitine, corresponding to recoveries of 107.8 ± 12.3%, 98.9 ± 12.3% and 45.7 ± 6.3, respectively, in a sample of 500 µL delivered at a flow of 50 µL min-1 under an extraction voltage of 300 V. Finally, the method was applied to the analysis of acylcarnitines spiked to urine, resulting in detection limits as low as 0.3-2 nM and several putative

endogenous

acylcarnitines.

The

current

flowing-to-stagnant

phase

micro-

electroextraction setup allows for the extraction of mL range volumes and is, as a consequence, very suited for analysis of low-abundant metabolites.

Keywords: electroextraction, microfluidics, flow cell, acylcarnitines, RPLC-TOFMS, phaseguide, sample pretreatment, sample enrichment

Abbreviations: EE, electroextraction


2

Page 3 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.

Introduction

Metabolomics, that is the analysis of the individual metabolites in an organism, is important for unraveling disease phenotypes.1 However, currently existing analytical methodologies are not able to address all challenges that metabolomics poses, for example the analysis of low-abundant metabolites. Often, quantitative information of these low-abundant compounds yields insight in important biological processes. A crucial step for such quantitative analysis is often the sample preparation, requiring both sample clean up as well as enrichment.2,3 Electromigration-based sample pretreatment processes are an attractive approach, since a range of focusing and stacking techniques are at disposal that may yield high enrichment factors of charged compounds, within a short time span.4,5,6 Electromigration-based techniques include amongst others free flow electrophoresis7, isoelectric focusing8, electric field gradient focusing9, isotachophoresis10, or a combination thereof11 and electromembrane extraction.12 The majority of metabolites is, or can be, ionized, making electromigration techniques widely applicable. Moreover, electromigration-based techniques are complementary to commonly used techniques such as liquid-liquid and solid phase extraction, so that other biological information can be found.13,14 Electroextraction (EE) was first described by Stichlmair et al.15 and is based on the application of an electric field over a two-phase liquid-liquid system. In EE, charged analytes migrate from an organic donor phase, typically an organic solvent, towards an aqueous acceptor phase by applying an electric field perpendicular to the interface. The electrical conductivity of the donor solvent is very low in comparison with the acceptor solvent; as a consequence, the field strength in the donor solvent is very high. This very high electric field causes charged analytes that are in the organic donor phase to migrate very fast towards the aqueous acceptor phase. Once the


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 34

analytes pass the liquid-liquid interface and enter the aqueous acceptor phase, their electrophoretic migration velocity is dramatically reduced, due to the lower field strength in this acceptor phase. As a consequence, the analytes are stacked in a narrow, highly concentrated analyte zone in the aqueous acceptor phase, just beyond the liquid-liquid interface. EE was shown to be a promising technique for small biomolecules; it is capable of achieving significant analyte signal enhancement (up to 3 orders of magnitudes) in LC-MS of peptides16,17 and acylcarnitines18, and of selective extraction of acylcarnitines from protein-rich matrices.19 Conventional EE is restricted by the sample loadability, ranging from 10 to 100 µL.16-19 A larger sample volume (i.e. donor phase) would improve the limit of detection (LOD) of the method. However, simply scaling up of the devices will result in unrealistically long analysis times due to the long migration distances. To overcome this, we designed an EE flow cell incorporated in a microfluidic chip. In this flow cell, the organic donor phase is continuously refreshed and the aqueous acceptor phase is kept stagnant. Continuous analyte extraction towards the stagnant aqueous acceptor phase takes place, while the depleted organic phase is continuously replaced with fresh, analyte-containing phase. This continuous flow microEE enables extraction of larger volumes while keeping the migration path of the analytes short. Moreover, the fact that the acceptor phase is stagnant assures that high concentration factors can be obtained. The main challenge that was encountered in developing an EE flow cell is obtaining a stable, straight interface over which a perpendicular electric field is present. When having a mismatch in flow velocity between both phases, as is required for concentrating analytes from a larger volume into a small volume, this is an even greater challenge. Several approaches to stabilize laminar flow systems exist, such as stabilization by incorporation of membranes20,21 in the


4

Page 5 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


device, modifying the surface properties of the flow channels22, enhancing the viscosity of at least one of the liquid phases with additives23 and working at very low flow rates (low µL min-1 range).24 The group of Kitamori used so-called guiding structures (5 µm high ridges etched in glass chips) to stabilize the interface between continuously flowing liquids at low flow rates (below 1 µL min-1).25 In this paper, we demonstrate that a phaseguide, located on the bottom of the flow cell, can be used as stabilizing structure for the liquid-liquid interface in which one phase is stagnant and the other fast-flowing (up to 333 µL min-1). Phaseguides, which were first described by Vulto et al

26

are microfluidic pinning structures, present as small ridges on the

bottom of chambers in microfluidic devices. Originally they were used to control advancements of liquid-air interfaces, in order to improve filling and emptying compartments in microfluidic structures,26,27 as well as selective patterning of gels.28-30 The manufacturing process of chips with phaseguides is straightforward and uncomplicated, contrary to incorporation of membranes or surface modification of a part of the chip surface. Also, no modification of the viscosity of the phases is needed, which is beneficial for the compatibility with hyphenated techniques, such as liquid chromatography – mass spectrometry (LC-MS) or capillary electrophoresis-MS (CE-MS). Here, we will demonstrate for the first time that the presence of the phaseguide enables EE of analytes from a large volume of fast flowing organic donor solvent into a small volume of stagnant aqueous acceptor phase. The extraction volume was increased to 1 mL, but can easily be increased further. Figure 1 shows the working principle of the flow cell. In Figure 1A, the main components of the microEE flow cell are depicted. The organic donor phase, which contains the sample cations, flows through the cell, while the aqueous acceptor phase is stagnant. Thanks to the phaseguide, a stable liquid-liquid interface is present (Figure 1B). Figure 1C shows the situation when a positive electric field is switched on. As soon as cations enter the electric


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 34

field, which is perpendicular to the flow direction, they are deflected towards the aqueous acceptor phase. Since the electric field is very high in the organic donor phase and very low in the aqueous acceptor phase, the cations are concentrated just after they passed the liquid-liquid interface. First, the stabilizing effect of a phaseguide on the liquid-liquid interface was demonstrated, then, a proof of principle of the microEE flow cell is demonstrated using the cationic dye crystal violet, which allowed for visual monitoring of the stability of the liquid-liquid interface and the extraction process. The potential of the microEE flow cell for the analysis of low-abundant metabolites is demonstrated using acylcarnitines as example. Therefore, it was coupled off-line to reversed phase LC (RPLC) in combination with time-of-flight MS (TOF-MS). The analytical performance of acylcarnitine analysis is characterized and, finally, the extraction of acylcarnitines spiked to urine is demonstrated.

2.

Materials and methods

The chemicals and equipment that were used in this study, as well as the used RPLC-TOFMS and CE-TOFMS methodology and the preparation of standard solutions and urine, can be found in the Supporting Information.

2.1 Fabrication of the flow cell Based on the chemical resistance to ethyl acetate, cyclo-olefin copolymer slides were chosen as substrate. The flow cell was designed using SolidWorks Premium 2010 (Dassault Systèmes SolidWorks Corp., Vélizy-Villacoublay, France), after which Mastercam Design X3 (CNC Software, Inc., Tolland, CT, USA) was used to create G-codes and Mach3 (ArtSoft USA,


6

Page 7 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fayette, ME, USA) was used to decode the G-codes into XYZ movement of a CNC milling machine (Taig 3000 CNC-mill, Supertech EMC-xyz-GSBX driver, Phoenix, Arizona, USA). A mill of 0.50 mm diameter was used at 5.000 RPM, at a feed rate of 100 mm/min, using a plunge depth of 0.125 mm/step. Burrs were removed with a soft brush; particles were removed by rinsing the chip with demineralized water. Finally, the chip was dried with compressed air. A mask was developed to enable localized thin-film platinum sputtering, in order to incorporate electrodes and electrical contact points for connection to the power supply. Platinum deposition was performed at 0.5 kV at ~10-5 mbar for 40 min with an argon flow of 40 standard cm3/min. The sputtering process resulted in a ~100 nm thick electrode. The slides were bonded using a solvent bonding procedure. The slide surfaces were activated by placing them for 45 s in 70:30 % (v/v) toluene:2-propanol. The activation process was quenched with an excess of 2-propanol and drying the surfaces using compressed air. After aligning the substrates, a hot press, operated for 15 min at 26 kg cm-2 (±1.6 kg cm-2) and 110 °C, was used to bond the slides.

2.2 Operation of the flow cell In Figure S.1 (Supporting Information), the microEE flow cell, constructed as outlined in section 2.3, is shown. The dimensions of the analyte donor channel were 15.00 × 2.00 × 0.50 mm and the dimensions for the analyte acceptor channel were 10.00 × 2.00 × 0.50 (accuracy of the dimensions ± 0.05 mm). A ridge of substrate material (8.00 × 0.40 × 0.20 mm), i.e. the phaseguide, served to stabilize the liquid-liquid system. Two perpendicular electrodes of approximately 8.00 mm long and ~100 nm thick were sputtered alongside the channels. Holes were drilled (ID 1.2 mm) at the beginning and the end of the channels to serve as solvent inlet/outlet.


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 34

Each device was tested by filling with an aqueous 5% formic acid (v/v) solution (1.2 M), after which 10 V was applied and the resulting current was measured. When the measured current was 7 ± 0.5 µA, the chip manufacturing was considered to be successful. The acceptor phase (volume 13.4 µL) was introduced into the flow cell with a micropipette and the inlet/outlet closed off using miniaturized stopcocks. During experiments, the acceptor compartment of the flow cell was closed, while the donor compartment was connected to a syringe at the inlet and to waste at the outlet. Fluid connections were realized through pressfitting polytetrafluorethene tubing (0.71 mm inner diameter). Both electrodes were exposed at the edge of the device, in order to be able to connect them to the power source using alligator clips. The anode was connected to the donor phase and the cathode to the acceptor phase. The power source was controlled using software to program the extraction time and voltage. The inlet tubing was filled with organic donor phase until the inlet of the flow cell. Once the power source was enabled, the flow was re-enabled. The power source was switched off as soon as the desired organic donor phase volume was extracted. After extraction, the remnants of organic donor phase were removed from the flow cell. This was done by first removing the outlet tubing, then the inlet tubing and finally tilting the chip while holding a dust-free tissue against the outlet of the chip. The stagnant aqueous acceptor compartment of the flow cell was removed with a pipette and transferred to a sample vial suitable for the LC apparatus. To facilitate injection from the sample vial, 100 µL methanol:water was added, unless stated otherwise. After each experiment, the chip was rinsed with 2-propanol and dried with pressurized air.

3.

Results


8

Page 9 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.1 Feasibility experiments

3.1.1 Effect of the phaseguide

To study the effect of a phaseguide on a liquid-liquid system in which one of the phases is flowing while the other remains stagnant, two flow cells were compared at two different flow rates (50 and 333 µL min-1): one with phaseguide and one without. The stagnant compartment consisted of 1.2 M FA, the flowing compartment of ethyl acetate with saturated with a aqueous solution of crystal violet (0.3 mM). Figure 2 demonstrates that a phaseguide enables a straight, stable liquid-liquid interface, especially at 50 µL min-1. At 333 µL min-1, a minor distortion of the interface can be noticed. Without phaseguide, a non-linear interface is formed, which poses a problem for the desired EE experiments that require a straight, fixed liquid-liquid interface and aqueous acceptor phase. The purpose of these experiments was to demonstrate the stabilizing effect of a phaseguide in liquid-liquid system were one phase is flowing while the other is kept stagnant. Therefore, no electrodes were incorporated in these devices; the influence of the extraction voltage on the stability of the interface will be discussed in the next section.

3.1.2 Exploration of extraction voltage and flow Figure 3 demonstrates the feasibility of the microEE flow cell. To be able to visually monitor the process, the cationic purple dye crystal violet was used. When no voltage is applied (Figure


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 34

3A), no visible extraction of crystal violet takes place. Upon application of a voltage, the crystal violet is extracted from the flowing organic phase into the stagnant aqueous acceptor phase (Figure 3B), which was an aqueous solution of 1.2 M formic acid. No crystal violet could be observed in the organic phase once it had passed the flow cell. Since extraction starts as soon as the sample enters the electric field, the concentrated sample zone is thicker at the entrance of the flow cell. When the organic phase flow (Qorg [m3 s-1]) was increased to 333 µL min-1, a thicker concentrated crystal violet zone was obtained, but not all crystal violet could be extracted (Figure 3C). The phase separation, however, was still intact, which demonstrates the excellent capability of a phaseguide to stabilize liquid-liquid systems in which a large flow-mismatch is present. Next, the extraction voltage (Uextr 9V]) working range of the flow cell was explored. Between 30 V and 1000 V, complete extraction and a stable liquid-liquid interface were observed. Figure 3D shows the unstable liquid-liquid interface that is present when a too high voltage (2000 V) is applied. Moreover, electric discharging occasionally takes place, as can be seen in the figure, and not all crystal violet was extracted from the organic phase. It is reported in the literature that electric fields can induce interface instabilities at liquid-liquid interfaces.31,32 However, more research should be carried out to exactly determine what cause the interface instabilities at high field strength. For instance, Joule heating could also play an important role in this.

3.1.3

Considerations for device dimensions

The dimensions of the flow cell are crucial for the performance of the device. First, the flow cell should be long enough in order to provide the analytes with a sufficient residence time in the extraction field in order to be extracted. Second, the electric field in the organic donor phase


10

Page 11 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


should be sufficient to provide sufficient lateral displacement of the analyte, so that they can reach the liquid-liquid interface and be extracted. The residence time (tr [s]) of the analyte in the flow cell can be calculated from the flow of the organic donor phase (Qorg [m3 s-1]) and the volume of the extraction cell (Vcell [m3]):

= .

(Equation 1)

The lateral displacement of the analyte, (dx [m)], during tr can be calculated using the electric field strength in the organic phase (Eorg [V m-1]) and the electrophoretic mobility of the analyte (µi [m2 V-1s-1]):

= . .

(Equation 2)

The conductivities of the organic donor phase and the aqueous acceptor phase were determined to be 0.8 and 1900 µS cm-1, respectively. Since the voltage is inversely proportional to conductivity, this means that virtually the whole extraction voltage (Uextr) is over organic donor phase, thus over the channel width of the organic phase (dorg):

=

(Equation 3)

In practice, this means that when Uextr is for example 100 V, an electric field close to 50 kV/m is present over the organic donor phase. Based on the µi-values of the analytes (see Table S.1), and Eorg it can be calculated in which regimes of Qorg and Uextr the flow cell should be operated


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 34

for successful extraction of acylcarnitines in the EE flow cell. Using these criteria, we calculated whether the device would be suited for the extraction of acylcarnitines or whether we should adapt its geometry for this purpose. In these calculations, the influence of analyte diffusion was neglected, as this is a slow process in comparison to electroextraction. Furthermore, the possible influence of changes in conductivity on Eorg was neglected; in principle, only the conductivity of the aqueous phase could increase during extraction, since the organic donor phase is continuously refreshed. As Uextr is at the starting situation almost entirely over the organic phase, increased conductivity of the aqueous phase will cause a negligible increase of Eorg. Finally, the influence of the solvent composition on µi was not taken into account. For successful extraction of an analyte molecule, dx should be at least 0.002 m, i.e. dx should be equal or larger than dorg. In Figure S.2, the results of calculations of dx at various Uextr (100, 300 and 500 V) and Qorg (50, 100 and 200 µL min-1) are shown, and it can be observed that only at the extremes, i.e. the lowest Uextr in combination with the highest Qorg, dx is smaller than channel width. Thus, the device as outlined in section 2.6 and Figure S.1 can be expected to be suited for proof-of-concept experiments with the selected acylcarnitines.

3.2 Characterization with acylcarnitines In order to determine the performance of the flow cell, a series of extractions of a test mixture containing three acylcarnitines (hexanoylcarnitine, octanoylcarnitine and lauroylcarnitine) was carried out, in combination with off-line coupled RPLC-TOFMS. Acylcarnitines represent an important metabolite class; deviations in the acylcarnitine profile have been linked to various inborn metabolic errors. 33


12

Page 13 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


First, the influence of organic donor phase flow and extraction voltage was studied. Then, under optimal Qorg and Uextr, the analytical performance of the device was studied. The data were subjected to Dixon’s Q test to check for outliers at the 95% confidence interval. 34 The samples in which the Q-value was above the threshold for n = 3 samples were omitted from the dataset. In total, two data points were removed from the series of optimization experiments.

3.2.1 Influence of Uextr and Qorg Figure 4 shows the influence of Uextr and Qorg on the extraction of acylcarnitines. A test mixture consisting of 1 mL ethyl acetate saturated with 8.33 µM of each acylcarnitine (concentration in ethyl acetate 0.21 µM) was extracted. The aqueous acceptor compartment of the cell consisted of 5% (v/v) formic acid in water (corresponding to 1.2 M), which was based on values reported in Lindenburg et al., where the concentration of formic acid in the aqueous phase in relation to EE of was optimized and it was observed that at lower formic acid concentrations in the aqueous phase extraction speed decreased.18 When no voltage was applied, low levels of acylcarnitines were extracted due to diffusion-based partitioning. During each experiment, the current was monitored. In a typical successful experiment the current was in the range of 3-25 µA, depending on the applied extraction voltage. In Figure 4B, the influence of applying an extraction voltage is clearly demonstrated; when extraction with 300 V is carried out, the peaks of hexanoylcarnitine, octanoylcarnitine and lauroylcarnitine are increased 5 times, 12.5 times and 16.5 times, respectively, in comparison with an experiment in which no voltage is applied. Two flow rates were studied: 50 and 100 µL min-1 (Figure 4). At 100 µL min-1, there is an optimal extraction voltage between 300 and 500 V for hexanoylcarnitine, >500 V for octanoylcarnitine and around 300 V for lauroylcarnitine. At this point we are not able to explain


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 34

the differences in performance of the three studied acylcarnitines. We hypothesize that the physicochemical properties play a role in this, which we consider an interesting subject for further research. At 500 V, bubble formation at the electrodes was observed, which coincided with worse EE performance, probably due to distortion of the electric field. At a flow of 100 µL min-1, the bubbles at the electrode in the organic donor phase were flushed away, while at 50 µL min-1 they were not. This could explain why at 500 V, a flow of 100 µL min-1 is more favorable for the extraction result than 50 µL min-1. When Qorg is decreased to 50 µL min-1, the optimum of all three acylcarnitines is around 300 V. Figure 4B shows the an overlay of two chromatograms, one with low peaks, obtained at 0 V and 50 µL min-1, and the other, with high peaks, obtained at the optimal conditions, i.e. 50 µL min-1 and 300 V. At a lower Qorg the analytes have a tr in the extraction cell and as a consequence, the extraction efficiency will increase. Further decreasing Qorg will possibly increase the extraction efficiency further, provided that Uextr is low enough to avoid bubble formation, and at the cost of an unfavorably long extraction time. Further measurements were, unless stated otherwise, carried out at the optimal conditions that can be derived from Figure 4A, i.e. 50 µL min-1 at 300 V.

3.2.2 Quantitative aspects A calibration curve of octanoylcarnitine and lauroylcarnitine was constructed to determine the linearity of the method. For this, six concentration levels (20, 40, 60, 100, 150 and 200 nM in ethyl acetate) were extracted and measured in triplicate. The sample mixture also contained 100 nM hexanoylcarnitine, which served as an internal standard. The organic donor phase (500 µL) was delivered with a flow rate of 50 µL min-1 and extracted at 300 V. Both calibration curves (see Figure S.3 and table S.2) were through origin (based on 95% confidence interval). Both calibration curves showed good linearity and repeatability over the tested range of


14

Page 15 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


concentrations. The repeatability of the method was studied by comparing the relative standard deviations of the slope of three calibration curve repetitions. The repeatability was 7.9% for the octanoylcarnitine curve and 8.1% for the lauroylcarnitine curve. The concentration LOD, i.e. the concentration that can be detected at a signal-to-noise ratio of three, as defined by the International Union for Pure and Applied Chemistry 35, were based on the signal-to-noise ratio of the peaks belonging to the lowest calibration value. The LOD was determined for the starting sample, i.e. prior to mixing with ethyl acetate.

The recovery was determined in triplicate by comparing the organic donor phase prior to EE with the aqueous acceptor phase after EE of 1 mL organic phase. The reference sample was 1mL 0.21 µM acylcarnitines in ethyl acetate. Since this solvent is not suitable for injection into the LC system, it was evaporated to dryness and reconstituted in 100 µL milliQ; during this process it was concentrated 10 times in order to obtain interpretable peaks in RPLC-TOFMS. In the calculation of the recovery, this concentration factor was corrected for. At complete extraction, the sample should be 74.6 times concentrated, i.e. from 1 mL donor phase into 13.4 µL acceptor phase. Based on the RPLC-TOFMS data, it was calculated that hexanoylcarnitine was concentrated 80.0 ± 9.2 times, octanoylcarnitine 73.8 ± 9.1 and lauroylcarnitine 34.1 ± 4.7 times, which corresponds to recoveries of 107.8 ± 12.3% for hexanoylcarnitine, 98.9 ± 12.3% for octanoylcarnitine and 45.7 ± 6.3% for lauroylcarnitine. Possibly, the lower recovery found for lauroylcarnitine can be explained by its lower solubility in the aqueous acceptor solution (it is the most hydrophobic acylcarnitine in this study). A similar behavior of acylcarnitines in an EE study was reported in Raterink et al.. 19


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 34

3.2.3 Demonstration of urine analysis The urine sample was spiked with 1.25 µM hexanoylcarnitine, octanoylcarnitine and lauroylcarnitine. Then, the spiked urine was diluted in ethyl acetate 2.5% (v/v). A total of 500 µL of this organic donor phase (corresponding to 12.5 µL urine) was extracted; 0.1 µL of the undiluted acceptor phase was injected and analyzed by RPLC-TOFMS. EE was performed using a voltage of 300 V and a flow rate of 50 µL min-1. Figure 5 shows the results of urine analysis.

Based on the signal-to-noise ratios, the LOD values of the flow cell in combination with RPLC-TOFMS are estimated to be in the low nM range, similar to the LOD values obtained with test solutions (section 3.2.2 and Table S.2). Several endogenous acylcarnitines, namely 2octenoylcarnitine, nonaoylcarnitine, decanoylcarnitine and myristoylcarnine, could be putatively identified, based on mass data and retention time order. These acylcarnitines are known to occur in urine in low concentrations.36 For absolute certainty of these identifications, experiments with reference standards or tandem MS should be carried out. These results support the promising potential of the microEE flow cell for the analysis of lowabundant metabolites. In Lindenburg et al.

18

, where capillary EE was coupled to LC-MS, the

LOD of hexanoylcarnitine spiked to urine was 25 nM. EE offers an alternative to techniques such as protein precipitation, solid phase extraction and liquid-liquid extraction, which are often used in LC-MS based acylcarnitine analysis. 37 The advantage of EE over these techniques is that it does not involve evaporation to dryness and reconstitution, thus preventing introduction of experimental errors and affection of recovery.16 The LOD values obtained with our system are comparable to LOD values obtained with superior LC-MS equipment, e.g. ultra performance RPLC-triplequad-MS where sample pretreatment involved derivatization to butyl esters and two evaporation/reconstitution steps (LOD values of a wide variety of acylcarnitines in 6 µL plasma


16

Page 17 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


4 – 357 nM)38, and ultra-performance hydrophilic interaction chromatography - triple quad-MS where serum sample pretreatment was protein precipitation (LOD values of acylcarnitines in 15 µL serum ~1.7 nM). 39

5.

Discussion and outlook

parameter

scaling with EE efficiency

expected complication at high end

expected complication at low end

unstable liquid-liquid system Uextr

bubble formation at electrodes

incomplete extraction

Joule heating Qorg

ionic strength of aqueous phase


unfavorable long analysis times

compatibility issues with hyphenated technique

analyte degradation at electrode

chemical resistance of device material analyte solubility in organic phase

precipitation due to saturation

no analyte in solution

analyte solubility in aqueous phase

precipitation due to saturation

no analyte in solution

length of flow cell

decreased concentration factors


gap between phaseguide and top substrate

unstable interface


dorg


unstable liquid-liquid system

daq

decreased concentration factors

unstable liquid-liquid system analyte degradation at electrode

Table 1 In this research paper we have introduced a microEE flow cell that is capable of extracting charged analytes from a flowing donor phase into a stagnant acceptor phase. A crucial factor in this context was to obtain a stable liquid-liquid interface between the two liquid phases of which one was motile. Here we showed that a simple pinning barrier, a phaseguide, was sufficient to


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 34

stabilize the liquid-liquid interface. This is a simplification with respect to other stabilization techniques such as membrane-based techniques. To our knowledge, we have shown for the first time that it is, by implementing a phaseguide, possible to stabilize a liquid-liquid system over longer periods of time, in which one phase remains stagnant while the other is flowing. In Table 1, the main parameters that affect the efficiency of EE, i.e. the amount of analyte that is successfully collected in the aqueous phase per time unit, are listed in order to provide the reader with an overview of what should be taken into account when developing a microEE flow cell. As shown in this research paper, a higher Uextr leads to enhanced extraction, but there is an upper limit where no stable liquid-liquid system can be maintained anymore, and where the influence of Joule heating and bubble formation impair the EE process. At the lower end, incomplete EE can be expected to take place. To cope with this, Qorg could be decreased as well, but this results in unfavorably long analysis times. On the other hand, at the higher end of Qorg, incomplete extraction takes place. The ionic strength of the aqueous phase influences EE via conductivity; the lower the ionic strength, the lower the conductivity, the higher the field strength in the aqueous phase and the faster the analytes will migrate towards the electrode, where they risk to be degraded. As a consequence, it is beneficial to work with an aqueous phase with high ionic strength but care must be taken that compatibility with hyphenated techniques is maintained. Furthermore, for successful EE, the analyte should be soluble in both phases. Studying the relation between analyte properties and extraction performance is of major importance for fully understanding the EE process and has our current attention. Device geometry is crucial for obtaining the optimal extraction performance. In this research paper we presented a prototype to demonstrate the proof of principle of microEE in a flow cell. The key element of the device is the phase guide, which was chosen to be 200 µm in a channel height of


18

Page 19 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


500 µm. Given the obtained recoveries for hexanoylcarnitine and octanoylcarnitine, which were around 100%, this phaseguide was suited and its height does not have to be decreased. Increasing the gap between the phaseguide and the top substrate increases the risk of interface instabilities. Decreasing this gap might fortify interface stability, but will also cause incomplete extraction due to the fact that Eorg will not be over the whole donor compartment. In fact it is not only the height of the phaseguide that is crucial for its performance as a pinning barrier, but also its shape, angles with the sidewall and contact angles of the liquids with the wall material. That said, the pressure that is required for a meniscus to overflow the phaseguide reduces with an increased channel height. Extensive calculations on phaseguide stability as a dependency of geometry have been made and are currently in press.40 For future work it would be important to determine aqueous liquid contact angles in relation to the chip material in order to have a quantitative estimate of the phaseguide barrier function. The length of the flow cell increases the area covered by Eorg and could expand the upper limit of Qorg, but simultaneously, the acceptor volume is increased, which lowers the concentration factors that can be achieved. To maximize concentration factors, the length of the flow cell could be shortened, but care must be taken to avoid incomplete extraction. The channel widths of the organic donor phase and aqueous acceptor phase (dorg and daq) affect EE performance. The higher Qorg, the more analyte can be extracted per time unit, but increasing dorg will lower Eorg, at the risk of incomplete extraction. To maximize Eorg, dorg should be decreased, but this could result in an unstable liquid-liquid system. When daq is too small, analytes will reach the electrode during EE and are at risk of degradation; increasing daq means increasing the acceptor volume and impairing the enriching effect of EE. When constructing a future flow cell in which maximal concentration factors can be achieved,


19


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 34

while maintaining complete extraction, we would, in comparison with the prototype presented in this paper, decrease the length of the flow cell and daq and increase dorg. The results presented here may have significant implications for metabolomics research and sample preparation techniques. The focus of this manuscript was mainly on demonstrating the principle and improvement of limits of detection as a result of high concentration factors. This may have significant benefits for when hyphenated with e.g. capillary electrophoresis techniques, that require small, but preferably concentrated analyte plugs. The EE procedure also has an inherent sample pretreatment component. By dissolving the sample in an organic donor phase, proteins precipitate. Furthermore, a major part of the lipids is not participating in the EE process, as they are neutral, and are left behind in the donor phase. This means that metabolites can be extracted from complex sample matrices at relatively high purity. Our future work will elaborate on these advantages.

6.

Conclusion We have successfully developed and characterized an EE flow cell that uses a phaseguide to

stabilize a liquid-liquid interface between a flowing organic donor phase and a stagnant aqueous acceptor phase. The working principle was demonstrated using the cationic dye crystal violet and the device was further characterized for extraction acylcarnitines. For this, the EE flow cell was coupled off-line to RPLC-TOFMS, a combination that is demonstrated to be promising; good calibration curves, repeatability and recovery were obtained. The LOD values for the three studied acylcarnitines were in the low nM range in case of test mixtures. Urine analysis resulted in an estimated LOD value that was also in the low nM range, and several low-abundant


20

Page 21 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


endogenous acylcarnitines could be detected. This is similar to or even better than LOD values that were obtained with more advanced LC-MS equipment.

7.

Acknowledgements

We would like to express our gratitude to Daan Boltje and Marcel Hesselberth from the Leiden Institute of Physics for aiding us in the thin film deposition process that we used to manufacture the flow cell electrodes. Authors would like to thank the Netherlands Metabolomics Centre and the Netherlands Genomics Initiative for funding this research.


21


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

8.

Page 22 of 34

References 1. Griffin, JL. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 2004, 359, 857-871. 2. Hyotylainen T. Anal. Bioanal. Chem. 2009, 394, 743-758. 3. Vuckovic, D. Anal. Bioanal. Chem. 2012, 403, 1523-1548. 4. Lindenburg, P.W.; Ramautar, R.; Hankemeier, T. Bioanalysis 2013, 5, 2785-2801. 5. Breadmore, M.C.; Shallan, A.I.; Rabanes, H.R.; Gstoettenmayr, D.; Abdul Keyon, A.S.; Gaspar, A.; Dawod, M.; Quirino J.P. Electrophoresis 2013, 34, 29-54. 6. Britz-Kibbin P.; Terabe S. J. Chromatogr. A 2003, 1000, 917-934. 7. Turgeon, R.T.; Bowser, M.T. Anal Bioanal Chem 2009 394, 187-198. 8. Sommer, G.J.; Hatch, A.V. Electrophoresis 2009, 30, 742-757 9. Meighan M.M.; Staton, S.J.; Hayes, M.A. Electrophoresis 2009 30, 852-865. 10. Smejkal P.; Bottenus, D.; Breadmore, M.C.; Guijt, R.M.; Ivory, C.F.; Foret, F.; Macka, M. Electrophoresis 2013, 34, 1493-1509. 11. Quist J.; Vulto P.; Hankemeier, T. Anal Chem 2014, Epub ahead of publication. dx.doi.org/10.1021/ac403764e. 12. Gjelstad, A.; Pedersen-Bjergaard, S. Bioanalysis 2011, 7, 787-797. 13. Kamphorst, J.J.; Tjaden, U.R.; Van der Heijden, R.; De Groot, J.; Van der Greef, J.; Hankemeier, T. Electrophoresis 2009, 30, 2284-2292. 14. Ramautar, R.; Nevedomskaya, E.; Mayboroda, O.A.; Deelder, A.M.; Wilson, I.D.; Gika, H.G.; Theodoridis, G.A.; Somsen, G.W.; De Jong, G.J. Mol. Biosyst. 2011, 7, 194-199. 15. Stichlmair, J.; Schmidt, J.; Proplesch, R. Chem. Eng. Sci. 1992, 47, 3015-3022. 16. Lindenburg, P.W.; Seitzinger, R.; Tempels, F.W.A; Tjaden, U.R.; Van der Greef, J.; Hankemeier, T. Electrophoresis 2010, 31, 3903-3912.


22

Page 23 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


17. Lindenburg, P.W.; Tempels, F.W.A; Tjaden, U.R.; Van der Greef, J.; Hankemeier, T. J. Chromatogr. A 2012, 1249, 17-24. 18. Lindenburg, P.W.; Tjaden, U.R.; Van der Greef, J.; Hankemeier, T. Electrophoresis 2012, 33, 2987-2995. 19. Raterink, R.J.; Lindenburg, P.W.; Vreeken, R.J.; Hankemeier, T. Anal. Chem. 2013, 85, 7762-7768 20. Hereijgers, J.; Callewaert, M.; Lin, X.; Verelst, H.; Breugelmans, T.; Ottevaer, H.; Desmet, G.; DeMalsche, W. J. Membr. Sci. 2013, 436, 154-162. 21. Cervera-Padrell, A.E.; Morthensen, S.T.; Lewandowski, D.J.; Skovby, T.; Kiil, S.; Gernaey, K.V. Org. Process Res. Dev. 2012, 16, 888-900. 22. Hibara, A.; Nonaka, M.; Hisamoto, H.; Uchiyama, K.; Kikutani, Y.; Tokeshi, M.; Kitamori, T. Anal. Chem. 2002, 74, 1724-1728. 23. Collins, C.J.; Berduque, A.; Arrigan, D.W.M. Anal. Chem. 2008, 80, 8102-8108. 24. Novak, U.; Pohar, A.; Plazl, I.; Žnidaršič-Plazl, P. Sep. Purif. Technol. 2012, 97, 172178. 25. Tokeshi, M.; Minagawa, T.; Uchiyama, K.; Hibara, A.; Sato, K.; Hisamoto H.; Kitamori, T. Anal Chem 2002, 74, 1565-1571. 26. Vulto, P.; Podszun, S.; Meyer, P.; Hermann, C.; Manz, A.; Urban, G.A. Lab Chip 2011, 11, 1596-1602. 27. Vulto, P.; Medoro, G.; Altomare, L.; Urban, G.A.; Tartagni, M.; Guerrieri, R.; Manaresi, N. J. Micromech. Microeng. 2006, 16, 1847-1853. 28. Vulto, P.; Dame, G.; Makohliso, S.; Zahn, P.; Maier, U.; Urban, G.A. Lab Chip 2010, 10, 610-616.


23


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 34

29. Vulto, P.; Kuhn, P.; Urban, G.A. Lab chip 2013, 13, 2931-2936. 30. Trietsch, S.J.; Israels, G.D.; Joore, J.; Hankemeier, T.; Vulto, P. Lab Chip 2013, 13, 3548-3554. 31. Samin, S.; Tsori, Y. J Chem Phys 2009, 131, 194102 32. Lin, Z.; Kerle, T.; Baker, S. M.; Hoagland, D.A.; Schäffer, E.; Steiner, U.; Russell, T. P. J Chem Phys 2001, 114, 2377-2381. 33. Longo, N.; Di San Filippo, C.A.; Pasquali, M.; Am. J. Med. Genet. Part. C Med. Genet. 2006, 142c, 77-85. 34. Rorabacher, D.B. Anal. Chem. 1991, 63, 139-146. 35. McNaught, A.D.; Wilkinson, A.; Nic, M.; Jirat, J.; Kosata, B.; Jenkins, A. IUPAC. Compendium of Chemical Terminology. 2nd ed.; Blackwell Scientific Publications: Oxford, 2006. 36. Bouatra, S.; Aziat, F.; Mandal, R.; Guo, AC.; Wilson, M.R.; Knox, C.; Bjorndahl, T.C.; Krishnamurthy, R.; Saleem, F.; Liu, P.; Dame, Z.T.; Poelzer, J.; Huynh, J.; Yallou, F.S.; Psychogios, N.; Dong, E.; Bogumil, R.; Roehring, C.; Wishart, D.S. PLoS One 2013, 8, :e73076. 37. Mansour, F. R.; Wei, W.; Danielson, N. D. Biomed Chromatogr 2013, 27, 1339-1353 38. Gucciardi, A.; Pirillo, P.; Di Gangi, I.M.; Naturale, M.; Giordano, G. Anal. Bioanal. Chem. 2012, 404, 741-751. 39. Kivilompolo, M.; Öhrnberg, L.; Orĕsič, M.; Hyötyläinen, T. J Chromatogr A 2013, 1292 189-194.


24

Page 25 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


40. Yildirim, E., Trietsch, S. J., Joore, J., Berg, A. van den, Hankemeier, T., Vulto, P. Phaseguides as Tunable Passive Microvalves for Liquid Routing in Complex Microfluidic Networks, currently under review at Lab on a Chip.


25


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 34

FIGURE AND TABLE CAPTIONS

Figure 1

The working principle of the continuous-flow microEE flow cell. A)

scheme representing the main components of the cell B) the phaseguide enables a stable liquid-liquid interface C) upon application of the electric field, extraction into the acceptor phase takes place. Upper drawings are top view; lower drawings are cross sectional view. 1) aqueous acceptor phase; 2) organic donor phase; 3) phaseguide; 4) electrodes; white arrow represents flow direction of organic phase; black arrows represent extraction direction; purple dots represent sample molecules; dotted line represents liquid-liquid interface

Figure 2

Comparison of microEE flow cells with and without phaseguide, at two

different flow rates (50 µL min-1 and 333 µL min-1). Dashed line indicate location of interface; i) channel for flowing organic donor phase ii) channel for stagnant aqueous acceptor phase iii) location of phase guide

Figure 3

EE of crystal violet in the microEE flow cell. A) start situation (Uextr = 0 V, Qorg

30 µL min-1 B) stable EE situation (Uext = 1000 V, Qorg 30 µL min-1 C) unfavorable situation due to too high Qorg (Uext = 1000 V, Qorg = 333 µL min-1) D) unfavorable situation due to too high Uext (Uext = 2000 V, Qorg = 166 µL min-1).

Figure 4

A) Influence of organic donor phase flow and extraction voltage on LC-MS peak

areas obtained for the electroextraction of three acylcarnitines (n=3); error bars represent


26

Page 27 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


standard deviation. B) overlay of RPLC-TOFMS base peak acylcarnitine chromatograms resulting from two experiments, showing the influence of the application of 300 V (high peaks) in comparison when no voltage was applied (low, black peaks). AU stands for arbitrary units.

Table 1 regression data of calibration curves obtained with the microEE flow coupled offline to RPLC-TOFMS, and recovery data. The calibration curves were through origin (based on 95% confidence interval). Hexanoylcarnitine served as internal standard.

Figure 5 Urine analysis, showing extracted ion chromatograms of three spiked acylcarnitines, marked with *, and 4 putatively identified endogenous acylcarnitines. The peak marked with # represents a contamination not related to urine

Table 1 Important EE parameters, how they scale with extraction efficiency and expected limitations at lower and upper end.


27


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 34

GRAPHICAL ABSTRACT


28

Page 29 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1 The working principle of the continuous-flow microEE flow cell. A) scheme representing the main components of the cell B) the phaseguide enables a stable liquid-liquid interface C) upon application of the electric field, extraction into the acceptor phase takes place. Upper drawings are top view; lower drawings are cross sectional view. 1) aqueous acceptor phase; 2) organic donor phase; 3) phaseguide; 4) electrodes; white arrow represents flow direction of organic phase; black arrows represent extraction direction; purple dots represent sample molecules; dotted line represents liquid-liquid interface 232x183mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2 Comparison of microEE flow cells with and without phaseguide, at two different flow rates (50 µL and 333 µL -1). Dashed line indicate location of interface; i) channel for flowing organic donor phase ii) channel for stagnant aqueous acceptor phase iii) location of phase guide 424x205mm (300 x 300 DPI)


Page 30 of 34

-1

Page 31 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3 EE of crystal violet in the microEE flow cell. A) start situation (Uextr = 0 V, Qorg 30 µL min-1 B) stable EE situation (Uextr = 1000 V, Qorg 30 µL min-1 C) unfavorable situation due to too high Qorg (Uextr = 1000 V, Qorg = 333 µL min-1) D) unfavorable situation due to too high Uext (Uextr = 2000 V, Qorg = 166 µL min-1).org 867x270mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4 A) Influence of organic donor phase flow and extraction voltage on LC-MS peak areas obtained for the electroextraction of three acylcarnitines (n=3); error bars represent standard deviation. B) overlay of RPLC-TOFMS base peak acylcarnitine chromatograms resulting from two experiments, showing the influence of the application of 300 V (high peaks) in comparison when no voltage was applied (low, black peaks). AU stands for arbitrary units. 242x392mm (300 x 300 DPI)


Page 32 of 34

Page 33 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5 Urine analysis, showing extracted ion chromatograms of three spiked acylcarnitines, marked with *, and 4 putatively identified endogenous acylcarnitines. The peak marked with # represents a contamination not related to urine 235x182mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 1 Important EE parameters, how they scale with extraction efficiency and expected limitations at lower and upper end. 352x122mm (300 x 300 DPI)


Page 34 of 34

Continuous-Flow Microelectroextraction for Enrichment of Low

Recommend Documents