Nano-electromembrane extraction
Transcript of Nano-electromembrane extraction
Accepted Manuscript
Title: Nano-Electromembrane Extraction
Author: Marıa D. Ramos Payan Bin Li Nickolaj JacobPetersen Henrik Jensen Steen Honore Hansen StigPedersen-Bjergaard
PII: S0003-2670(13)00618-1DOI: http://dx.doi.org/doi:10.1016/j.aca.2013.04.055Reference: ACA 232558
To appear in: Analytica Chimica Acta
Received date: 10-12-2012Revised date: 25-4-2013Accepted date: 28-4-2013
Please cite this article as: M.D.R. Payan, B. Li, N.J. Petersen, H. Jensen, S.H. Hansen,S. Pedersen-Bjergaard, Nano-Electromembrane Extraction, Analytica Chimica Acta(2013), http://dx.doi.org/10.1016/j.aca.2013.04.055
This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.
Page 1 of 30
Accep
ted
Man
uscr
ipt
1
Manuscript for Analytica Chimica Acta1
Nano-Electromembrane Extraction2
3
María D. Ramos Payána, Bin Lib, Nickolaj Jacob Petersenb*, Henrik Jensenb, 4
Steen Honoré Hansenb, and Stig Pedersen-Bjergaardb,c5
a Department of Analytical Chemistry, Faculty of Chemistry, University of Seville, P.O. Box 6
41012, Seville, Spain7
b School of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, 8
2100 Copenhagen, Denmark9
c School of Pharmacy, University of Oslo, P.O Box 1068 Blindern, 0316 Oslo, Norway10
11
12
Keywords: Sample preparation13
Nano-extraction14
Membrane extraction15
Nano-electromembrane extraction16
17
* Corresponding author at: School of Pharmacy, Faculty of Health and Medical Sciences, 18
University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark, Tel.: +4519
35336184; E-mail address: [email protected]
Page 2 of 30
Accep
ted
Man
uscr
ipt
2
Abstract21
The present work has for the first time described nano-electromembrane extraction (nano-22
EME). In nano-EME, five basic drugs substances were extracted as model analytes from 200 23
µL acidified sample solution, through a supported liquid membrane (SLM) of 2-nitrophenyl 24
octyl ether (NPOE), and into approximately 8 nL phosphate buffer (pH 2.7) as acceptor 25
phase. The driving force for the extraction was an electrical potential sustained over the 26
SLM. The acceptor phase was located inside a fused silica capillary, and this capillary was 27
also used for the final analysis of the acceptor phase by capillary electrophoresis (CE). In that 28
way the sample preparation performed by nano-EME was coupled directly with a CE 29
separation. Separation performance of 42.000-193.000 theoretical plates could easily be 30
obtained by this direct sample preparation and injection technique that both provided31
enrichment as well as extraction selectivity. Compared with conventional EME, the acceptor 32
phase volume in nano-EME was down-scaled by a factor of more than 1000. This resulted in 33
a very high enrichment capacity. With loperamide as an example, an enrichment factor 34
exceeding 500 was obtained in only 5 minutes of extraction. This corresponded to 100-times 35
enrichment per minute of nano-EME. Nano-EME was found to be a very soft extraction 36
technique, and about 99.2-99.9 % of the analytes remained in the sample volume of 200 µL. 37
The SLM could be reused for more than 200 nano-EME extractions, and memory effects in 38
the membrane were avoided by effective electro-assisted cleaning, where the electrical 39
potential was actively used to clean the membrane.40
41
Page 3 of 30
Accep
ted
Man
uscr
ipt
3
1. Introduction41
Recently, electromembrane extraction (EME) was proposed as a new concept for42
analytical sample preparation [1]. In EME, charged analytes are extracted from an aqueous 43
sample, through an organic solvent (supported liquid membrane) immobilized in the pores 44
of a thin polymeric membrane, and into a µL volume of an aqueous acceptor phase. The 45
driving force for the extraction is a dc electrical potential sustained over the supported liquid 46
membrane (SLM). For extraction of cationic analytes, the cathode is located in the acceptor 47
phase, and the anode is placed in the sample. Both the sample and the acceptor phase are48
kept under pH control to ensure that the analytes are positively charged during EME. For 49
acidic analytes, the polarity of the electrical potential is reversed, with the anode located in 50
the acceptor phase. Because the analytes have to pass the SLM during EME, which is an 51
organic phase immiscible with water, high extraction selectivity can be obtained [2,3]. The 52
selectivity can easily be tuned by optimizing the chemical composition of the SLM, the 53
magnitude of the electrical potential, and by the pH conditions on both side of the SLM [2,3]. 54
In addition, because the analytes normally are transferred from a relatively large sample 55
volume (typically 0.1-10 mL) and into a small volume of acceptor phase (typically 10-25 µL), 56
EME also provides excellent pre-concentration [2,3].57
EME has been shown to give very clean extracts from complicated real samples like 58
human plasma [1,4], whole blood [5,6], urine [1,4], saliva [7], breast milk [8], waste water 59
[9,10], and drinking water [11]. Since the acceptor phase is aqueous, EME is directly 60
compatible with high-performance liquid chromatography (HPLC) [12], liquid 61
chromatography – mass spectrometry (LC-MS) [13], and capillary electrophoresis (CE) [1]. 62
Typically, less than 25 µL of organic solvent is consumed per sample to establish the SLM. 63
Extraction recoveries in EME are normally in the range 25-75 % after 5 minutes of extraction 64
Page 4 of 30
Accep
ted
Man
uscr
ipt
4
[2,3], and enrichment factors up to 190 times have been reported [2,3]. Several applications 65
of EME have been published, among others basic drug substances [1,5,7,8,14-18], acidic 66
drug substances [10,19], lithium [20], peptides [12,13,21], lead [22], heavy metals [23], 67
alkylphosphonic acid derivatives [24], and chlorophenols [25].68
EME is essentially a micro-extraction technique, where analytes are transferred to a µL 69
volume of acceptor phase. Extractions are often close to being exhaustive, and enter steady-70
state conditions after 5-10 minutes. Both these characteristics are highly beneficial for the 71
majority of analytical routine applications. However, for more specialized studies, soft 72
extraction may be important where only small amounts (<1 %) of analyte are removed from 73
the sample. Soft extractions are important in cases where target analytes are in equilibrium 74
with components of the sample matrix, and where the free concentration of the target 75
analyte is requested. This can be the case with drugs interacting with proteins in biological 76
fluids as an example. This paper reports on experiments with down-scaling conventional 77
EME to nano-EME. The idea was to develop a nano-extraction technique providing both soft 78
extraction and high analyte enrichment, and to couple the sample preparation directly with 79
a capillary electrophoresis (CE) separation. This was realized by reducing the volume of the 80
acceptor phase, from typically 10-25 µL in conventional EME, to approximately 8 nL in the 81
present nano-EME work. A preliminary technical solution to nano-EME is presented, and 82
proof-of-principle studies are reported. The proof-of-principle studies included 83
experimentally measured enrichment factors, and were also focused on sample depletion, 84
acceptor phase saturation, and electro-assisted cleaning of the SLM which was reused for 85
more than 200 extractions. The intention of the paper is thus to present a new concept 86
which can be utilized in new analytical procedures in the future. The nano-EME system is 87
also described theoretically by a simple model which provides a good qualitative description 88
Page 5 of 30
Accep
ted
Man
uscr
ipt
5
of the system, and therefore may be helpful in optimising performance in future applications 89
of the methodology. The nano-EME system was realized in a crack in a fused silica CE 90
capillary, which was surrounded by a SLM. 91
Systems with some relation to the present work have been published previously [26-28]. 92
In one paper, analytes were transferred from the sample and through a dialysis membrane 93
directly into a CE capillary [26]. Here, the principle for mass transfer was dialysis. In another94
paper, analytes were transferred from the sample and through a SLM and into a CE capillary 95
[27]. In this case, the driving force for the extraction was a pH gradient, and the acceptor 96
phase volume was 280 nL compared to 8 nL in the present work. Very recently, a system 97
close to the current set-up was presented [28]. In this system, analytes were extracted 98
through a SLM and into a CE capillary by using the power supply of the CE-instrument, and 99
this system was used for analysis of perchlorate in biological samples.100
101
Page 6 of 30
Accep
ted
Man
uscr
ipt
6
2. Experimental101
102
2.1. Chemicals and reagents103
Analytical grades of pethidine (C15H21NO2; pKa=8.2; log P=2.2), nortriptyline (C19H21N; 104
pKa=9.7; log P=4.3), methadone (C21H27NO; pKa=9.1; log P=3.9), haloperidol (C21H23ClFNO2; 105
pKa=8.1; log P=3.8), and loperamide (C29H33ClN2O2; pKa=9.4; log P=4.1) were all obtained 106
from Sigma-Aldrich (St. Louis, MO, USA). Except for haloperidol, all these basic drugs were 107
used as hydrochlorides. 2-nitrophenyl octyl ether (NPOE) was obtained from Fluka (Buchs, 108
Switzerland). Stock solutions containing 1 mg mL-1 of each compound were prepared in 10% 109
(v/v) ethanol and stored protected from light at 277 K (4oC). Aqueous working solutions of 110
each compound were daily prepared by adequate dilutions from the 1 mg mL-1 stock 111
solutions by 10 mM HCl. Ultra-high purity water (18 MΩ cm) was used to prepare stock 112
solutions, aqueous working solutions, and separation buffer for capillary electrophoresis.113
114
2.2. Capillary electrophoresis115
Capillary electrophoresis (CE) was performed using an Agilent Technologies HP3D-CE 116
instrument (Waldbronn, Germany) equipped with an UV detector operated at 200 nm, and a117
fused-silica capillary (TSP050375, Polymicro Technologies, Phoenix, AZ, USA) of 50 μm i.d, 118
and with an effective length of 41.0 cm and a total length of 61.5 cm. The separation buffer 119
was 60 mM sodium dihydrogen phosphate adjusted to pH 2.7 with ortho-phosphoric acid. 120
The buffer was degased with helium for 10 min to avoid bubble generation during injection 121
and separation. A separation voltage of 15 kV was used, and the temperature of the capillary 122
was set at 298 K (25°C).123
124
Page 7 of 30
Accep
ted
Man
uscr
ipt
7
125
2.3. System and procedure for nano-EME 126
The nano-EME set up is illustrated in Fig. 1. The system comprised a fused silica capillary, 127
a piece of (porous) hollow fibre, a small polypropylene vial as sample reservoir, two 128
electrodes, and a power supply. The power supply was a model HV5448 3000V (LabSmith, 129
Livermore, CA, USA) with programmable voltage in the range up to 3000 V. The electrodes 130
for the extraction consisted of two 0.5 mm OD platinum wires. A 200 µL PCR tube (VWR 131
International) was used as the sample compartment. The piece of hollow fibre used for 132
immobilization of the supported liquid membrane was a Plasmaphan P1LX polypropylene 133
hollow fibre (330 µm internal diameter, 150 µm wall thickness, and 0.4 µm pore size) 134
purchased from Membrana (Wuppertal, Germany). A fused silica capillary (TSP050375, 135
Polymicro Technologies) of 50 µm internal diameter, 363 µm outside diameter, and a total 136
length of 61.5 cm was used for both nano-EME and capillary electrophoresis (CE).137
First, a narrow crack was made in the fused silica capillary approximately 2 cm from the 138
inlet. This crack was made with a tool for cutting optical fibres (T 108, Biccotest, Cheshunt, 139
UK). The optical fiber cleaver is originally meant for cutting optical fibers for 140
telecommunication at a perfect 90o angle, but can be used for fused silica capillaries as well. 141
The tool clamps onto the capillary thereby allowing exact positioning of the cut (crack). The 142
clamping generates a small stress in the capillary and combined with a sapphire cutting knife 143
slightly touches the capillary surface a fracture in the capillary is generated. The crack144
(fracture) was formed across the whole fused silica capillary at a perfect 90o angle. The 145
polyimide coating covering the fused silica had the same perfect cut as the fused silica but it 146
remained intact opposite to where the knife of the optical fibre cutter touched the capillary. 147
The polyimide therefore maintained the perfect alignment of the cleaved fused silica148
Page 8 of 30
Accep
ted
Man
uscr
ipt
8
capillary and assured that when the capillary was held straight the crack in the capillary was 149
closed thereby forming a sub micron channel with close to zero dead volume. When making 150
the crack on the capillary not too much force should be added on the cutter to keep capillary 151
integrity. Doing successfully, a crack was made in the fused silica capillary and the two pieces 152
were still held together by the external polyimide coating. Then, a 5 mm piece of hollow 153
fibre was placed on the outside of the fused silica capillary covering the crack. To facilitate 154
this, the piece of hollow fibre was treated with ethanol that made it slide easily onto the 155
capillary. Covering the crack, the ends of the hollow fibre were carefully melted in order to 156
fix the membrane to the fused silica capillary and to create a leak-tight connection, Fig. 1b.157
For the heat fixing, a PTFE tube was applied to control the heated length and protect the 158
other part. The heat fixing also prevented the sample from migrating into the capillary 159
without passing through the membrane. The heat fixed system was stable for long periods 160
(months) of operations. In addition, the heat fixing reduced the membrane length thereby 161
reducing the carry-over effect caused by a large volume of the SLM. 1 µL of 2-nitrophenyl 162
octyl ether (NPOE) were delivered by a micro-pipette to the piece of hollow fibre. NPOE 163
served as the supported liquid membrane (SLM) and was immediately immobilized into the 164
polypropylene membrane by capillary forces. This process was visually inspected as the 165
appearance of the membrane changed from white to transparent during immobilization, Fig. 166
1c. Excess of NPOE was removed with a medical wipe. The same NPOE was used several days 167
before a new portion of NPOE was added to the membrane. The fused silica capillary with 168
the SLM was passed through holes in the side of the sample vial and the SLM was then 169
located in the sample reservoir. From the sample reservoir, the fused silica capillary was170
directed into the CE-instrument, and a detection window for the UV detector was burned 171
20.5 cm from the outlet. The CE instrument was operated in CE-MS mode, where the electric 172
Page 9 of 30
Accep
ted
Man
uscr
ipt
9
current was connected to ground outside the instrument, with a negative potential applied 173
to the buffer reservoir inside the CE instrument during the separation. The height of outlet 174
capillary was adjusted in the same level as the reservoir inside the CE, which minimized the 175
siphoning effect and thereby assuring that no flow was present in the capillary during the 176
injection. Nano-EME and injection was being performed at the end of the capillary placed 177
outside the instrument.178
179
Prior to nano-EME, the sample reservoir was filled with 200 µL of sample consisting of 0.5 180
µg/mL pethidine, nortriptyline, methadone, haloperidol, and loperamide in 10 mM HCl. A 181
buffer vial was located at the inlet of the fused silica capillary and connected to ground with 182
a Pt electrode, and this vial contained the CE separation buffer (60 mM sodium dihydrogen 183
phosphate buffer adjusted to pH 2.7 with ortho-phosphoric acid). The Pt electrode served 184
both as the anode during the CE separation and as the cathode during nano-EME. Prior to an 185
analysis, the separation buffer was flushed into the entire capillary from the buffer vial 186
placed inside the CE-instrument. Finally, the anode for nano-EME was placed in the sample 187
solution and extraction was accomplished by application of extraction voltage. The cationic 188
model analytes were extracted through the SLM and the crack in the fused silica capillary, 189
and into a small segment of the buffer inside the capillary. After nano-EME, the analytes 190
were separated in the fused silica capillary by applying -15 kV at the reservoir placed inside 191
the CE instrument. The sample vial (with sample) was left in the same position during the 192
analysis. The electric potential applied at the sample reservoir was removed during the 193
separation and no sample was leaking into the separation channel during the CE analysis. For 194
both the nano-EME and the separation, the electrode in the buffer reservoir outside the CE 195
instrument was grounded.196
Page 10 of 30
Accep
ted
Man
uscr
ipt
10
197
2.4. Calculations of extraction efficiency and enrichment factor198
The enrichment factor ( iEF ) for the analyte i was calculated according to the following 199
equation:200
201
i
i
s
a
i C
CEF (1)202
203
iaC is the concentration of the analyte i in the acceptor phase after extraction and isC is the 204
initial concentration of the analyte in the sample. iaC was determined by capillary 205
electrophoresis and UV-detection using external calibration.206
The extraction efficiency or recovery ( iR ) for the analyte i was defined as the fraction of 207
analyte i in the sample that was transferred to the acceptor phase by the following equation:208
209
%100
i
i
ss
aai CV
CVR (2)210
where Va is the volume of the acceptor phase and Vs is the volume of the sample.211
212
Page 11 of 30
Accep
ted
Man
uscr
ipt
11
3. Results and discussion212
3.1. Technical set-up for nano-extraction213
The technical set-up for nano-electromembrane extraction (nano-EME) is illustrated in 214
Fig. 1. This preliminary set-up was built around a fused silica capillary. A narrow crack was 215
made in the capillary. A small piece of a porous polypropylene hollow fibre was treaded onto 216
the capillary and placed over this crack. The mechanical stability of this arrangement was 217
acceptable for the current proof-of-principle studies, but a more robust design should be 218
developed in the future. The piece of hollow fibre was treated with 2-nitrophenyl octyl ether 219
(NPOE). NPOE served as the supported liquid membrane (SLM), and was immobilized in the 220
pores of the hollow fibre by capillary forces. NPOE was selected without further optimization 221
based on earlier experience from conventional EME [1-3]. The SLM was located in the 222
sample reservoir, and sample was filled into this reservoir. The anode for nano-EME was 223
placed in the sample, whereas the cathode was placed in a buffer reservoir at the “inlet” end 224
of the capillary. The inlet buffer reservoir was in liquid contact with buffer inside the 225
capillary. By application of voltage to the anode in the sample reservoir, extraction of 226
cationic model analytes was accomplished from the sample, through the SLM and the crack 227
in the capillary and into the phosphate buffer solution (pH 2.7) located inside the capillary. 228
This buffer solution also served as the running buffer in the subsequent analysis by capillary 229
electrophoresis (CE), which was accomplished inside the same capillary. Thus, after nano-230
EME, the cationic model analytes were directly separated, detected, and quantified by CE.231
A small zone of the buffer solution inside the fused silica capillary, located around the 232
crack, served as acceptor phase. The volume of this active acceptor phase was 233
approximately 8 nL when extraction was performed for 30 seconds. This volume was 234
measured by comparing the peak widths from the nano-EME / CE-system with peak widths 235
Page 12 of 30
Accep
ted
Man
uscr
ipt
12
from conventional CE with the same capillary. The model analytes were in the latter case 236
dissolved in CE buffer to avoid stacking, and injection volumes from conventional CE were 237
calculated with a computer program (CE-Expert, Beckman Coulter, Brea, CA). Since the 238
acceptor phase volume was far below 0.05 µL, we defined the extraction set-up as a nano-239
extraction system and termed it “nano-EME”.240
The experiments conducted with nano-EME in this work were all based on the basic drug 241
substances pethidine, nortriptyline, methadone, haloperidol, and loperamide as model 242
analytes. These substances were selected based on extensive experience from conventional 243
EME [1]. The sample solution contained 10 mM HCl as sample electrolyte. The acidic 244
conditions ensured full protonation of the basic model analytes, which was important for 245
their electro-kinetic migration during nano-EME. Fig. 2 shows a typical electropherogram 246
after nano-EME and CE. In this experiment the sample contained 0.5 µg mL-1 pethidine, 247
nortriptyline, methadone, haloperidol, and loperamide in 10 mM HCl, the sample volume 248
was 200 µL, the supported liquid membrane was 2-nitrophenyl octyl ether, the voltage was 249
200 V and the extraction time was 30 seconds. In spite of the small crack in the capillary, the 250
electrophoretic separation efficiency was high. Some peak tailing was observed, most 251
probably due to trace leakage of analyte from the SLM during the initial stage of the CE 252
analysis, but overall peak shapes were found satisfactory for the proof-of-principle studies 253
(Fig. 2).254
255
3.2. Extraction performance256
In a first series of practical experiment, enrichment factor (EF) was measured as function 257
of applied voltage. The results are summarized in Fig. 3. In these experiments, with a nL-258
volume of acceptor phase, the sample volume was 200 µL, NPOE was used as SLM, and the 259
Page 13 of 30
Accep
ted
Man
uscr
ipt
13
extraction time was 30 seconds. As seen from Fig. 3, enrichment factors for all of the model 260
analytes increased with increasing voltage from 0 to 500 V. With increasing voltage, both the 261
magnitude of the sample-to-SLM distribution coefficient and the electrophoretic migration 262
across the SLM increased. A similar behaviour has been reported for conventional EME [29]. 263
The extraction current was in the range of 14-16 µA for a 200 V extraction. Above 200 V, our 264
preliminary system for nano-EME became less stable. Relative standard deviations from 265
replicate extractions increased, and the current during the subsequent CE-analysis became 266
less stable. This indicated that gas bubbles were formed due to electrolysis at high voltages 267
with nano-EME, and this was the reason for the instability. Therefore, it was decided to 268
operate the system at 200 V for the rest of this study. At 200 V, the voltage drop over the 269
SLM was estimated to about 20-40 V. The rest of the voltage drop was in the buffer in the 270
fused silica capillary between the acceptor phase zone and the inlet buffer reservoir with the 271
cathode.272
In another experiment, the enrichment factor (EF) was measured as function of extraction 273
time. The results are summarized in Fig. 4. In these experiments, the sample volume was 200 274
µL, NPOE was used as SLM, and the extractions were carried out at 200 V. As seen from Fig.275
4, enrichment factors for all of the model analytes increased linearly with increasing 276
extraction time from 0 to 300 seconds. The enrichment factors were compound dependent, 277
due to differences in the magnitude of the distribution coefficient and due to differences in278
electrokinetic mobility in the SLM.279
Fig. 4 illustrates a very important point with the nano-EME system. Enrichment per unit 280
time was extremely high for the nano-EME system. With loperamide as an example, an 281
enrichment factor exceeding 500 was obtained in only 5 minutes of extraction. This 282
corresponded to 100-times enrichment per minute of nano-EME. For comparison, a recent 283
Page 14 of 30
Accep
ted
Man
uscr
ipt
14
work on EME in a micro-chip format, demonstrated a 107 times enrichment of loperamide 284
after 30 minutes of extraction [30]. In the latter case, the acceptor phase volume was 0.6 µL, 285
and the enrichment per minute was 3.6. In addition to very high enrichment per unit time, 286
the nano-EME system also provided high extraction selectivity. This is illustrated in Fig. 5a+b, 287
where human urine was spiked with pethidine, nortriptyline, methadone, haloperidol, and 288
loperamide, and subsequently processed by nano-EME. Although the urine sample 289
contained a broad range of matrix components, as illustrated by direct CE analysis in Fig. 5a, 290
most matrix components were removed when nano-EME was combined with CE analysis 291
(Fig. 5b). This high selectivity has been discussed previously for conventional EME [1-3].292
The influence of the ionic strength of the sample was investigated with the nano-EME 293
system by comparing extractions performed from pure water samples, 2 M sodium chloride, 294
and 1 M sodium sulphate, and all acidified to pH 2.0 using HCl. As illustrated in Table 1, 295
enrichment factors for the given model analytes were independent of the ionic strength. 296
Also, as seen from Table 1, extractions were repeatable within 1.4 % RSD. Detection limits 297
were 15, 3, 2, 4, and 0.2 ng mL-1 for pethidine, nortriptyline, methadone, haloperidol, and 298
loperamide, respectively.299
300
3.3. Sample depletion301
In a subsequent experiment, the level of sample depletion was investigated. The results 302
are summarized in Table 2. In these experiments, the sample volume was 200 µL, NPOE was 303
used as SLM, the extractions were carried out at 200 V, and the extraction time was 120 304
seconds. As seen from Table 2, recoveries were in the range 0.10 %-0.79 %. Although small 305
amounts of analyte also were trapped inside the SLM (discussed below), nano-EME served as 306
a very soft extraction technique, extremely small amounts of analyte were removed from 307
Page 15 of 30
Accep
ted
Man
uscr
ipt
15
the sample, and about 99.2-99.9 % of the analytes remained in the sample. Although 308
recoveries were low, the nano-EME system still provided high enrichment due to the very 309
small volume of acceptor phase (Table 2). This is a unique feature of nano-EME.310
In another experiment, nano-EME was conducted with stirring of the sample. This 311
experiment provided results (recoveries) comparable with data reported above without 312
stirring. The observation supported that extraction in the current nano-EME system occurred 313
only in the boundary layer of the SLM, while the rest of the sample remained relatively 314
undisturbed. This conclusion was further supported by an experiment where the sample 315
volume was varied between 50 and 300 µL. From this experiment, the absolute amount of 316
extracted analyte was unaffected by the sample volume. Thus, whereas conventional EME 317
can extract exhaustively [2,3], nano-EME may provide a very interesting alternative for very 318
soft extractions in a strongly limited sample volume segment.319
320
3.4. Acceptor phase saturation321
With a nL-volume of acceptor phase, the extraction capacity was expected to be limited. 322
This was confirmed by linearity measurements as illustrated in Fig. 6. In these experiments, 323
the sample volume was 200 µL, NPOE was used as SLM, the extractions were carried out at 324
200 V, and the extraction time was 120 seconds. As illustrated in Fig. 6, the nano-EME 325
system provided linear data up to sample concentrations of about 2.5 µg mL-1 for 326
haloperidol, and up to 1 µg mL-1 for pethidine, nortriptyline, and methadone. For the most 327
hydrophobic substance among the model analytes, namely loperamide, which also provided 328
the highest enrichment factor, deviation from linearity was observed at even lower 329
concentration (between 0.25 and 0.5 µg mL-1). For loperamide, the enrichment factor was 330
approximately 200 after 120 seconds. With a concentration of 0.25 µg mL-1 in the sample, 331
Page 16 of 30
Accep
ted
Man
uscr
ipt
16
this corresponded to a final concentration of about 50 µg mL-1 in the acceptor phase 332
segment. At higher analyte concentrations (or with longer extraction times) the amount of 333
analyte extracted into the acceptor phase approached the capacity of the system, and 334
significant deviation from linearity was observed. 335
336
3.5. Membrane cleaning337
With the current preliminary set-up for nano-EME, re-use of the SLM for many extractions 338
was mandatory to avoid time-consuming changing of the small SLM in-between each 339
extraction. Although nano-EME may be developed into other technical configurations in the 340
future, re-use of the very small SLM will probably remain as an important issue because very 341
small membranes are difficult to handle. In the current nano-EME system, traces of analyte 342
were trapped inside the SLM after each extraction. This is illustrated in Fig. 7, where a 343
sample free of analytes (pure 10 mM HCl) was extracted right after a sample containing the 344
analytes. As seen from Fig. 7a), carry-over was observed in the system due to trapping in the 345
SLM. However, this carry-over was eliminated by immersing the SLM in pure 10 mM HCl, 346
applying the potential (200 V) for 2 minutes, while flushing the CE capillary with separation 347
buffer. In this electro-assisted cleaning procedure, the SLM was efficiently rinsed using the 348
electrical potential. With the electro-assisted cleaning, the SLM was re-used for more than 349
200 extractions over a period of 2 months before replacement. During this time period, both 350
water samples (low complexity) and urine samples (moderate complexity) were extracted. 351
Highly complex samples like human plasma were not extracted in these experiments. Also, 352
with electro-assisted cleaning, all nano-EME experiments were highly repeatable with RSD 353
values less than 15 %.354
355
Page 17 of 30
Accep
ted
Man
uscr
ipt
17
3.6. Theoretical considerations356
The nano-EME system has several unique features that do not allow previous EME theory 357
to be applied directly. In the present setup, the volume of the acceptor phase was much 358
smaller than the volumes of sample and SLM. As discussed in section 3.3, the experimental 359
results showed that stirring (convective mass transport in the donor solution) had no effect 360
on the extractions efficiency. In order to develop a theoretical model for nano-EME, we 361
assume the following:362
363
1) No significant analyte depletion occurs in the sample (as reported in section 3.3).364
2) After a lag time, tlag, the limiting step to mass transport is the overall electro-365
migration (related to analyte concentration and mobility) in the SLM. Although there 366
may be a small diffusive contribution to mass transport, we shall omit it in the 367
present description. The validity of this assumption has been addressed in the 368
literature [1]. 369
3) The steady state concentration in the SLM is governed by the sample-to-SLM 370
distribution coefficient, Kd,i*, which in turn is related to the Galvani potential 371
difference across the sample-SLM interface. 372
4) The performance is not limited by the solubility in the acceptor phase. This 373
assumption is obviously not valid if very high enrichment factors are achieved (as 374
discussed in section 3.4). In such a situation deviations from the theoretical 375
predictions are likely to occur. 376
377
Under these assumptions, the analyte (i) concentration in the SLM, imc , is governed by the 378
distribution coefficient: 379
Page 18 of 30
Accep
ted
Man
uscr
ipt
18
380
t> tlag: *,,, idiDim Kcc (3)381
382
where cD,I is the analyte concentration in the sample. Kd,i* can be estimated as previously 383
described [29] and is dependent on the standard distribution potential of the ion and the 384
Galvani potential difference across the sample-SLM interface, which in turn is related to the 385
overall applied extraction potential.386
387
The analyte flux across the SLM (NPOE membrane), NPOEiJ , can be expressed as: 388
389
dEKcudEcuJ memidiDNPOEimemim
NPOEi
NPOEi // *
,,, (4)390
391
where NPOEiu is the electrokinetic mobility of analyte i in NPOE, memE is the potential drop 392
across the SLM, and d is the SLM thickness. We now infer that at t> tlag we have equality of 393
fluxes from analyte leaving the SLM and of analyte entering the acceptor phase, AiJ :394
395
NPOEiJ = A
iJ (5)396
397
From the definition of flux, the amount of analyte transport into the acceptor phase as a 398
function of time (t> tlag) , niA(t), can be expressed as:399
400
AtJAtJtn NPOEi
Ai
Ai )( (6)401
402
Page 19 of 30
Accep
ted
Man
uscr
ipt
19
where t is time and A is the "area of extraction" between the SLM and the acceptor phase. 403
Combining equation (4) and (6) above gives us: 404
405
AtdEKcutn memidiDNPOEi
Ai /)( *
,, (7)406
407
From equation (7) the enrichment factor can be obtained (t> tlag) : 408
tdEKuV
AcVtntEF memid
NPOEi
aiDa
Aii ///)()( *
,,
(8)409
The enrichment factor is independent on the analyte concentration in the sample solution410
provided that the assumptions are valid. Equation (8) predicts a linear relation between 411
enrichment factor and time, which also was observed during the practical experiments 412
discussed in section 3.2. Equation (8) also shows that the most lipophilic compounds (high 413
Kd,i*) are extracted with the highest efficiency, and this was in agreement with the 414
experimental results in section 3.2. It should be noted that the model is not valid for very 415
high analyte concentrations, where the extraction efficiency becomes limited by the 416
solubility of the analyte in the SLM or in the acceptor phase. In the latter case the 417
enrichment factor will level off to a constant value after a long extraction time as discussed 418
in section 3.4.419
420
Page 20 of 30
Accep
ted
Man
uscr
ipt
20
4. Conclusions420
The present work has for the first time described nano-electromembrane extraction 421
(nano-EME) using an electrical field as the driving force for mass transfer. Attention was 422
focused on both practical experiments and theoretical considerations. Essentially, nano-EME423
was accomplished by down-scaling conventional EME, reducing the volume of the acceptor 424
phase approximately by a factor of 1000 to only 8 nL. Although the technical set-up used for 425
the nano-EME experiments was preliminary in nature and requires further development, the 426
results clearly demonstrated several very interesting features of nano-EME. First, nano-EME 427
provided extremely high analyte enrichment factors per unit time. Secondly, nano-EME was 428
found to be a very soft extraction technique. Thus, in nano-EME, extraction was performed 429
only in the boundary layer around the supported liquid membrane (SLM), and less than 0.8 430
% of the analyte molecules were extracted from the 200 µL sample. The combination of very 431
soft extraction, high enrichment from very small samples volumes, and transfer of the 432
analytes into a nL aqueous solution, is unique for nano-EME. This should be interesting for 433
the future, among others for non-disturbing investigation of biochemical reactions taking 434
place in very small compartments. 435
436
5. Acknowledgement437
The Danish Council for Independent Research, Technology and Production Sciences (11-438
106647) is acknowledged for financial support.439
440
441
Page 21 of 30
Accep
ted
Man
uscr
ipt
21
References441
[1] S. Pedersen-Bjergaard, K.E. Rasmussen, J. Chromatogr. A 1109 (2006) 183-190.442
[2] A. Gjelstad, S. Pedersen-Bjergaard, Bioanalysis 3 (2011) 787-797.443
[3] N.J. Petersen, K.E. Rasmussen, S. Pedersen-Bjergaard, Anal. Sci. 27 (2011) 965-972444
[4] S.S.H. Davarani, A.M. Najarian, S. Nojavan, M.-A. Tabatabaei, Anal. Chim. Acta 725 445
(2012) 51-56.446
[5] A. Gjelstad, K.E. Rasmussen, S. Pedersen-Bjergaard, Anal. Bioanal. Chem. 393 (2009) 447
921-928.448
[6] L. Strieglerová, P. Kubáň, P. Boček, J. Chromatogr. A 1218 (2011) 6248-6255.449
[7] S. Seidi, Y. Yamini, A. Saleh, M. Moradi, J. Sep. Sci. 34 (2011) 585-593.450
[8] I.J.O. Kjelsen, A. Gjelstad, K.E. Rasmussen, S. Pedersen-Bjergaard, J. Chromatogr. A451
1180 (2008) 1-9.452
[9] K. Alhooshani, C. Basheer, J. Kaur, A. Gjelstad, K.E. Rasmussen, S. Pedersen-Bjergaard, 453
H.K. Lee, Talanta 86 (2011) 109-113.454
[10] M.R. Payán, M.Á.B. López, R.F. Torres, M.V. Navarro, M.C. Mochón, Talanta 85 (2011) 455
394-399.456
[11] I.K. Kiplagat, T.K.O. Doan, P. Kubáň, P. Boček, Electrophoresis 32 (2011) 3008-3015.457
[12] M. Balchen, L. Reubsaet, S. Pedersen-Bjergaard, J. Chromatogr. A 1194 (2008) 143-458
149.459
[13] M. Balchen, T.G. Halvorsen, L. Reubsaet, S. Pedersen-Bjergaard, J. Chromatogr. A460
1216 (2009) 6900-6905.461
[14] M. Eskandari, Y. Yamini, L. Fotouhi, S. Seidi, J. Pharmaceut. Biomed. Anal. 54 (2011) 462
1173-1179.463
Page 22 of 30
Accep
ted
Man
uscr
ipt
22
[15] L.E.E. Eibak, A. Gjelstad, K.E. Rasmussen, S. Pedersen-Bjergaard, J. Chromatogr. A464
1217 (2010) 5050-5056.465
[16] S. Nojavan, A.R. Fakhari, J. Sep. Sci. 33 (2010) 3231-3238.466
[17] M. Rezazadeh, Y. Yamini, S. Seidi, J. Chromatogr. B 879 (2011) 1143-1148.467
[18] S. Seidi, Y. Yamini, T. Baheri, R. Feizbakhsh, J. Chromatogr. A 1218 (2011) 3958-3965.468
[19] M. Balchen, A. Gjelstad, K.E. Rasmussen, S. Pedersen-Bjergaard, J. Chromatogr. A469
1152 (2007) 220-225.470
[20] L. Strieglerová, P. Kubáň, P. Boček, Electrophoresis 32 (2011) 1182-1189.471
[21] M. Balchen, H. Jensen, L. Reubsaet, S. Pedersen-Bjergaard, J. Sep. Sci. 33 (2010) 1665-472
1672.473
[22] C. Basheer, S.H. Tan, H.K. Lee, J. Chromatogr. A 1213 (2008) 14-18.474
[23] P. Kuban, L. Strieglerova, P. Gebauer, P. Bocek, Electrophoresis 32 (2011) 1025-1032.475
[24] L. Xu, P.C. Hauser, H.K. Lee, J. Chromatogr. A 1214 (2008) 17-22.476
[25] J. Lee, F. Khalilian, H. Bagheri, H.K. Lee, J. Chromatogr. A 1216 (2009) 7687-7693.477
[26] L. Nozal, L. Arce, B.M. Simonet, A. Rios, M. Valcarcel, Electrophoresis 28 (2007) 3284-478
3289.479
[27] B.A.P. Buscher, U.R. Tjaden, J. van der Greef, J. Chromatogr. A 764 (1997) 135-142.480
[28] P. Kubáň, I.K. Kiplagat, P. Boček, Electrophoresis 33 (2012) 2695-2702.481
[29] K.F. Seip, H. Jensen, A. Gjelstad, S. Pedersen-Bjergaard, submitted for 482
Electrophoresis.483
[30] N.J. Petersen, S. Taule Foss, H. Jensen, S. Honoré Hansen, C. Skonberg, D. 484
Snakenborg, J.P. Kutter, S. Pedersen-Bjergaard, Anal. Chem. 83 (2011) 44-51.485
Page 23 of 30
Accep
ted
Man
uscr
ipt
23
486
Fig. 1. a) Schematic illustration of the nano-EME system in nano-extraction mode and photo of a 375 487
µm OD fused silica capillary with crack and SLM b) before and c) after filling the pores with NPOE.488
489
490
Fig. 2. Electropherogram after nano-EME and capillary electrophoresis. Sample: 0.5 µg mL-1 of 491
pethidine, nortriptyline, methadone, haloperidol, and loperamide in 10 mM HCl; Sample volume: 200 492
µL; Supported liquid membrane: 2-nitrophenyl octyl ether; Extraction voltage: 200 V; Extraction time: 493
30 seconds.494
Page 24 of 30
Accep
ted
Man
uscr
ipt
24
495
Fig. 3. Analyte enrichment versus extraction voltage. Sample: 0.5 µg mL-1 of pethidine, nortriptyline, 496
methadone, haloperidol, and loperamide in 10 mM HCl; Sample volume: 200 µL; Supported liquid 497
membrane: 2-nitrophenyl octyl ether; Extraction time: 30 seconds.498
499
500
Fig. 4. Analyte enrichment versus extraction time. Sample: 0.5 µg mL-1 of pethidine, nortriptyline, 501
methadone, haloperidol, and loperamide in 10 mM HCl; Sample volume: 200 µL; Supported liquid 502
membrane: 2-nitrophenyl octyl ether; Voltage: 200 V.503
504
Page 25 of 30
Accep
ted
Man
uscr
ipt
25
505
Fig. 5. Electropherograms showing a) blank urine injected without nano-EME, b) urine containing 5 506
µg mL-1 of model analytes without nano-EME, and c) urine containing 5 µg mL-1 of model analytes 507
after nano-EME. 1) pethidine, 2) nortriptyline, 3) methadone, 4 ) haloperidol and 5 ) loperamide. 508
Sample volume: 200 µL; Supported liquid membrane: 2-nitrophenyl octyl ether; Voltage: 200 V; 509
Extraction time: 15 seconds.510
511
512
Fig. 6 Peak area divided by concentration as function of concentration (µg mL-1). Sample: pethidine, 513
nortriptyline, methadone, haloperidol, and loperamide in 10 mM HCl; Sample volume: 200 µL; 514
Supported liquid membrane: 2-nitrophenyl octyl ether; Voltage: 200 V; Extraction time: 120 seconds. 515
Left hand axis is for pethidine, nortriptyline and methadone. Right hand axis is for haloperidol and 516
loperamide.517
518
Page 26 of 30
Accep
ted
Man
uscr
ipt
26
519
Fig. 7. Electropherogram after nano-EME and capillary electrophoresis of pure 10 mM HCl (no 520
analytes), following a sample containing 0.5 µg mL-1 of pethidine 1), nortriptyline 2), methadone 3), 521
haloperidol 4), and loperamide 5) in 10 mM HCl. a) No cleaning of the SLM, b) after cleaning of the 522
SLM by 200 V for 2 minutes. Sample volume: 200 µL; Supported liquid membrane: 2-nitrophenyl 523
octyl ether; Voltage: 200 V; Extraction time: 120 seconds.524
525
526
527
Page 27 of 30
Accep
ted
Man
uscr
ipt
27
Table 1 Enrichment factor obtained with different sample matrix 527
Sample matrix 528
Water 2 M NaCl 1 M Na2SO4529
Analyte (n=3) (n=3) (n=3)530
531
Pethidine 4.9 (1.3) 4.7 (1.2) 4.6 (0.9)532
Nortriptyline 12.6 (1.1) 12.3 (0.9) 12.4 (1.3)533
Methadone 32.2 (1.4) 30.1 (1.1) 30.1 (1.0)534
Haloperidol 6.0 (1.3) 5.4 (1.2) 5.7 (1.3)535
Loperamide 58.1 (0.9) 56.4 (1.0) 57.9 (1.3)536
537
Sample: 0.5 µg mL-1 of pethidine, nortriptyline, methadone, haloperidol, and loperamide in water, 2 538
M NaCl and 1 M sodium sulphate, all adjusted to pH 2 with HCl; Sample volume: 200 µL; Supported 539
liquid membrane: 2-nitrophenyl octyl ether; Extraction voltage: 200 V; Extraction time: 60 seconds. 540
RSD (in %) are given in parentheses.541
542
543
544
Page 28 of 30
Accep
ted
Man
uscr
ipt
28
Table 2 Amount of analyte extracted into the acceptor phase544
Analyte Recovery in Amount left Enrichment545
acceptor phase in sample factor546
(n=3) (n=3) (n=3)547
548
Pethidine 0.10 % ~99.9 25549
Nortriptyline 0.27 % ~99.7 69550
Methadone 0.46 % ~99.5 114551
Haloperidol 0.28 % ~99.7 63552
Loperamide 0.79 % ~99.2 196553
554
Sample: 0.5 µg mL-1 of pethidine, nortriptyline, methadone, haloperidol, and loperamide in 10 mM 555
HCl; Sample volume: 200 µL; Supported liquid membrane: 2-nitrophenyl octyl ether; Extraction 556
voltage: 200 V; Extraction time: 120 seconds.557
558
559
560
561
Page 29 of 30
Accep
ted
Man
uscr
ipt
29
EME (electromembrane extraction) was coupled directly with a CE separation561capillary.562
The acceptor volume was only in the order of 8 nL 563 High efficient CE separations could be obtained without the need of sample stacking. 564 The system presented high enrichment factors exceeding 500.565 The supported liquid membrane (SLM) was stable for more than 200 extractions.566
567
Page 30 of 30
Accep
ted
Man
uscr
ipt
*Graphical Abstract