The present paper defines the optimal extraction window (OEW) for three-phase membrane-based liquidphase microextraction (MP-LPME) in terms of analyte polarity (log P), and anchors this to existing theories for equilibrium partitioning and kinetics. Using deep eutectic solvents (DES) as supported liquid membranes (SLM), we investigated how the OEW was affected by ionic-, hydrogen bond and π-π interactions between the SLM and analyte.
Trang 1Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/chroma
Maria Schüllera, Kim Tu Thi Trana, Elisabeth Leere Øiestada, b, Stig Pedersen-Bjergaarda, c, ∗
a Department of Pharmacy, University of Oslo, P.O Box 1068 Blindern, 0316 Oslo, Norway
b Oslo University Hospital, Division of Laboratory Medicine, Department of Forensic Sciences, P.O Box 4459 Nydalen, 0424, Oslo, Norway
c Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark
a r t i c l e i n f o
Article history:
Received 5 July 2021
Revised 10 December 2021
Accepted 20 December 2021
Available online 23 December 2021
Keywords:
Sample preparation
Liquid-phase microextraction
Hollow fiber
Pharmaceuticals
Deep eutectic solvents
a b s t r a c t
Thepresentpaperdefinestheoptimalextractionwindow(OEW)forthree-phasemembrane-based liquid-phasemicroextraction(MP-LPME)intermsofanalytepolarity(logP),and anchorsthistoexisting the-ories forequilibrium partitioningand kinetics.Using deepeutectic solvents(DES) assupported liquid membranes(SLM),weinvestigatedhowtheOEWwasaffectedbyionic-,hydrogenbondandπ-π inter-actionsbetweentheSLMandanalyte.Elevenbasicmodelanalytesintherange-0.4<logP <5.0were extracted byMB-LPMEin a96-wellformat Extractionwas performedfrom 250μLstandard solution
in25 mM phosphatebuffer (pH 7.0)into50 μLof10 mMHClacceptor solution (pH 2.0)with mix-turesofcoumarin,camphor,DL-menthol,andthymol,withandwithouttheioniccarrierdi(2-ethylhexyl) phosphate(DEHP),astheSLM.TheOEWwithpureDESwasintherange2<logP <5,andlowSLM aromaticitywasfavorablefortheextractionofnon-polaranalytes.Here,extractionrecoveriesupto98% wereobtained.UponadditionofDEHPtotheSLMs,theOEWshiftedtotherange-0.5<logP <2,and
acombinationof5%DEHPandmoderatearomaticityresultedinextractionrecoveriesupto80%forthe polaranalytes.Extractionwithioniccarrierwasinefficientforthenon-polaranalytes,duetoexcessive trappingintheSLM.TheresultsfromourstudyshowthatLPMEperformsoptimallyinarelatively nar-rowlogP-windowof≈ 2–3unitsandthattheOEWisprimarilyaffectedbyioniccarrierandaromaticity
© 2021TheAuthor(s).PublishedbyElsevierB.V ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/)
1 Introduction
Microextraction techniques like liquid-phase microextraction
(LPME) are popular choices for the extraction of targeted analytes
from biological matrices Offering low consumption of organic sol-
vents and the potential for automation, they are favorable with
regards to greenness and efficiency in high-throughput applica-
tions [1] Current effort f ocus on the implementation of LPME in
routine laboratories, including the commercialization of equipment
and better understanding of the optimal area of use This work will
focus on the latter
LPME can be performed with both two and three phases where
the former is based on the partition of a substance between two
immiscible liquids, like an aqueous sample and organic acceptor
solution In a three-phase system, a second aqueous phase is in-
troduced as the acceptor, allowing for the yield of cleaner extracts
and better HPLC compatibility [2] The principle is based on liquid-
∗ Corresponding author
E-mail address: stig.pedersen-bjergaard@farmasi.uio.no (S Pedersen-Bjergaard)
liquid extraction with back extraction, with a configuration allow- ing the extraction to take place in a single step [3] Fig 1B il- lustrates the basic principle with two aqueous phases, the donor, and acceptor, separated by an immiscible organic supported liq- uid membrane (SLM) In a typical scheme for basic analytes, like many pharmaceutical drugs, the donor is alkaline, and the accep- tor is acidic A high pH in the donor promotes the extraction of the analyte as a neutral species into the SLM, while a low pH in the acceptor promotes extraction and collection of the same ana- lyte as the protonated species Extraction recovery will depend on the partition between these three phases [3] Extraction in a three- phase system is typically used for organic analytes with weak base
or acid properties
An array of configurations for LPME has been developed through the years, each with specific advantages and limitations Beginning in 1996, pioneering work for two-phase LPME was conducted by Dasgupta [4] and Cantwell [5] with the introduc- tion of single-drop microextraction (SDME), where a drop of or- ganic liquid is immersed into an aqueous sample Another im- portant milestone in the field of microextraction, pioneered by Pedersen-Bjergaard and Rasmussen, was the introduction of the
https://doi.org/10.1016/j.chroma.2021.462769
0021-9673/© 2021 The Author(s) Published by Elsevier B.V This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )
Trang 2Fig 1 A) Membrane-based liquid phase microextraction in 96-well format B) The principle of three-phase liquid-phase microextraction For the extraction of basic analytes
the donor is alkalized to deprotonate the base The neutral analyte diffuses across the SLM and is protonated in the acidic acceptor solution
membrane-based technique hollow-fiber liquid-phase microextrac-
tion (HF-LPME), which initiated the development of three-phase
microextraction systems [6] In later years, several alternatives
have been developed, including dispersive liquid-liquid microex-
traction (DLLME) [7], dispersive liquid-liquid microextraction based
on solidified floating organic droplets (DLLME-SFO) [8], solvent
bar microextraction (SBME) [9], and other membrane-based tech-
niques such as membrane-based LPME (MB-LPME) in a 96-well for-
mat [ 10, 11] and electromembrane extraction (EME) [ 12, 13] In the
present study, MB-LPME in a 96-well format was used ( Fig.1A)
LPME has been applied in various fields of health and life
science, such as pharmacology [14], forensics [15], environmental
chemistry [16], clinical chemistry [17], toxicology [18], and anti-
doping control [19]with scientific publications steadily increasing
from the mid-’90s until today Although LPME has kept its rele-
vance, implementation into routine laboratories has still not been
facilitated One reason is the need for commercialized equipment,
such as seen for solid-phase microextraction (SPME) While LPME
equipment is still not commercially available, equipment for EME
is very close to market, and when used without voltage, it operates
as a MB-LPME system [20] Another reason is the need for generic
methods, where molecular descriptors can be used to predict and
select appropriate standardized extraction conditions for given an-
alytes Although a large number of validated applications have
been published [21–25], development of generic methods from this
material is difficult Extractions have been done using very differ-
ent experimental conditions, and performances have not been an-
chored sufficiently in fundamental understanding about partition-
ing and molecular interactions
Therefore, in the present work we have looked into LPME again,
now from a highly fundamental angle We have studied the extrac-
tion of selected basic pharmaceuticals as model analytes in the log
P range from −0.4 to 5.0, using selected deep eutectic solvents as
the SLM With the latter, hydrogen bonds, π-π-, and ionic interac-
tions were controlled and varied systematically For each SLM, we
investigated (1) which model analytes suffered from poor partition
into the SLM, (2) which model analytes were extracted efficiently
across the SLM, and (3) which model analytes suffered from poor
partition from SLM and into the acceptor Model analytes belong-
ing to (2) were within the optimal extraction window (OEW) of
each SLM, and OEWs were assigned for different SLMs based on
molecular interactions and analyte log P as a single molecular de-
scriptor The purpose of this was to develop a starting point for
the development of generic methods The work is therefore fun-
damental and general, and extractions were conducted only from
pure standard solutions Applications, quantifications in biological
fluids, method development, and validation, which are required to
develop the final generic methods, will follow in future papers The
intention is that the current paper will serve as fundamental refer- ence
2 Experimental
2.1 Chemicals and reagents
Coumarin, thymol, camphor, DL-menthol, di(2-ethylhexyl) phos- phate, hydrochloride acid 37%, sotalol hydrochloride, metaraminol bitartrate, atenolol, tyramine, ephedrine hydrochloride, metopro- lol tartrate, pethidine hydrochloride, haloperidol, nortriptyline hy- drochloride, loperamide hydrochloride, and methadone hydrochlo- ride were purchased from Sigma-Aldrich (St Louis, MO, USA) Sodium hydroxide, acetonitrile (HPLC grade), and formic acid 99% (LC-MS grade) were purchased from VWR (Radnor, PA, USA) Deionized water was obtained with a Milipak® (0.22 μm filter) pu- rification system from Milli-Q (Molsheim, France)
Individual stock solutions of each analyte were prepared at con- centrations of 1–4 mg mL −1and dissolved in pure deionized water
or 30% v/v methanol in deionized water From these, a polar ana- lyte mix (sotalol, metaraminol, atenolol, tyramine, ephedrine, and metoprolol) and a non-polar analyte mix (pethidine, haloperidol, nortriptyline, loperamide, and methadone) were prepared with an- alyte concentrations of 10 μg mL −1 in 25 mM phosphate buffer These analyte mixes were pipetted into the donor wells The final methanol concentration never exceeded 0.3% and was not assumed
to affect extraction recoveries To calculate the recovery, a 50 μg
mL −1 standard polar mix and 50 μg mL −1standard non-polar mix were prepared in 10 mM HCl All solutions were stored at 4 °C and protected from light
DES were prepared by weighing and mixing appropriate amounts of HBA components (coumarin or camphor) and HBD components (thymol or DL-menthol) in molar ratios of 1:1 and 1:2 The mixtures were heated in an oven (80 °C) for approximately
10 min and vortexed until homogenous In SLMs with ionic carrier, DEHP was added to the DES mixtures in volume ratios of 0.5, 2, 5,
or 10%
2.2 Liquid-phase microextraction procedure
The equipment used is previously described in [22] The ex- traction was performed with a 96-well polypropylene donor plate with 0.5 mL wells from Agilent (Santa Clara, CA, USA) The ac- ceptor plate was a 96-well MultiScreen-IP filter plate from Merck Millipore (Carrigtwohill, Ireland) The membrane material was polyvinylidene fluoride (PVDF) with a pore size of 0.45 μm A Platemax Pierceable Aluminum Sealing Film (Axygen, Union City,
Trang 3CA, USA) was used to seal the acceptor plate Agitation during ex-
traction was accomplished with a Vibramax 100 agitation system
from Heidolph (Kellheim, Germany)
The extraction was performed by pipetting 250 μL of standard
solution into the well of the donor plate The filter on the acceptor
plate was impregnated with 4 μL of DES, creating the SLM Accep-
tor solution (50 μL) was pipetted into the wells of the acceptor
plate The acceptor and donor plate were clamped together and
sealed with adhesive foil The whole set-up was placed into the
agitation device for extraction for 60 min at 900 rpm After extrac-
tion, the donor and acceptor solution were collected and analyzed
with HPLC-UV Details to the use of the extraction unit are pro-
vided in Supplementary Section 2
2.3 High performance liquid chromatography with UV-detection
HPLC-UV analysis was performed on a 30 0 0 Ultimate HPLC-
UV (Thermo Fisher Scientific, Waltham, MA, USA) with an Acquity
UPLC HSS T3 (2.1 mm I D × 150 mm, 1.8 μm particle size) pur-
chased from Waters (Wexford, Ireland) Mobile phase A consisted
of 0.1% formic acid in 95:5 deionized water/methanol (v/v) Mobile
phase B consisted of 0.1% formic acid in 95:5 methanol/deionized
water (v/v) Further details on the elution gradients, detection pa-
rameters and chromatograms are provided in Supplementary Sec-
tion 1
2.4 Calculations
The extraction recovery was calculated by Eq.(1):
R= C a, f inal
C d ,initial ×V a
Here, C a , f inal and C d ,initialare the concentrations in the acceptor
after the extraction and the concentration of analyte in the donor
before the extraction, respectively The terms V aand V ddenote the
volume of the acceptor and donor, respectively
3 Results and discussion
Eleven drug compounds were selected as model analytes with
−0.4< log P < 5.0 This range was chosen, as it represents a com-
mon range for small-molecule drugs [26] Polar analytes were de-
fined as having log P < 2.0 and non-polar analytes were defined
as having log P > 2.0 Analyte log P, log D at pH 2, and pK avalues
are shown in Table1
Table 1
Physical-chemical properties of the studied drugs
Chemicalize was used to generate structure properties (retrieved 27.02.2021,
https://chemicalize.com ) developed by ChemAxon ( https://www.chemaxon
com ) [28]
3.1 Extraction theory and definition of optimal extraction window
The partition coefficient K 1between the donor and the SLM can
be written as:
K1=C eq , SLM
Here, C eq,d and C eq ,SLMare the equilibrium concentrations in the donor and SLM, respectively Correspondingly, the partition coeffi- cient K 2between the SLM and acceptor can be written as:
K2=C C eq ,a
Here, C eq ,ais the equilibrium concentration in the acceptor The
overall partition coefficient K between the donor and acceptor can
be expressed as the product of K 1and K 2:
K= C eq ,a
Assuming that K is unaffected by the organic phase, K 2 is de- creasing when K 1is increasing, and vice versa
Based on the partition coefficients, the theoretical extraction re- covery in three-phase LPME may be calculated by:
R(%)= K × V a
K × V a+K1× V SLM+V d × 100% (5)
Here, V a, V d, and V SLM denote the volumes of the acceptor, donor, and SLM, respectively
Eq (4) is valid for estimating extraction recoveries when the system has entered equilibrium The time required to reach equi- librium is compound-dependent Previous work has shown that partition into the SLM often is the rate-limiting factor in three- phase LPME [27] Thus, kinetics are controlled by K 1, and can be modelled using the following equation:
C d(t)=C0
d· exp
−A SLM D SLM K1
V d h t
(6)
Here, C d(t) is the concentration in the donor as a function of time, C 0
d is the initial concentration in the donor ( t = 0), A SLM is the surface area of the SLM, h is the thickness of the SLM, and
D SLMis the diffusion coefficient of the analyte in the SLM Exact values of K 1and K 2 are generally not available, unless 1- octanol is used as the SLM With 1-octanol, computer-generated log P and log D values can be used to describe the efficiency of a three-phase LPME system K 1is equal to P for the neutral species, while K 2 is set to 1/D for the species at pH 2
Selecting three compounds with different log P and log D values from our set of model analytes, namely methadone, pethidine, and sotalol, the phase distribution for t = 60 min was calculated based
on Eq.(5) and Eq.(6)using computer-generated log P and log D values [28] Details of the calculations are found in Supplementary Section 3 For methadone, log P is 5.0, while log D at pH 2 is 1.5 The calculated distribution is illustrated in Fig.2 The donor is de- pleted, 70% is trapped in the SLM, and only 30% of the analyte is extracted into the acceptor Because methadone is non-polar, K 1is large, and the donor is therefore rapidly depleted Correspondingly,
K 2 is small, causing methadone to suffer from serious membrane
trapping Similar behavior is expected for other analytes with log
P higher than 3.5, including loperamide and nortriptyline from our set of model analytes, with 1-octanol as the SLM After 60 min, the extraction system has entered equilibrium, and membrane trap- ping remains unaffected upon extension of the extraction time For sotalol, log P and log D are −0.4 and −3.2, respectively So- talol is a polar substance; K 1 is low, while K 2 is high Therefore, partition into the SLM is highly unfavorable, and mass transfer is correspondingly slow After 60 min of extraction, 99% of sotalol is
Trang 4Fig 2 Calculated relative amounts of sotalol, pethidine, and methadone found
in the donor (blue), SLM (yellow), and recovered in the acceptor (green) with
t = 60 min
still left in the donor, and extraction is limited by slow kinetics
With 1-octanol as the SLM, slow kinetics is to be expected for all
substances with log P below 2.0 Recoveries can be improved by
increasing extraction time, and theoretically, exhaustive extraction
is achieved at the end However, this requires extraction far beyond
60 min, which is of little relevance for analytical applications
For pethidine, log P and log D are 2.5 and −1.0, respectively
Compared to methadone, K 1is lower and K 2is higher Due to this,
the balance between K 1and K 2is more appropriate, and the equi-
librium distribution is much more in favor of high recovery for
pethidine ( Fig.2) With K 1sufficiently high to prevent slow kinet-
ics , without causing membrane trapping, the analytes are extracted
under ideal conditions With 1-octanol as the SLM, similar behavior
is expected for other analytes with log P in the range of 2.0 to 3.5
This is the optimal extraction window (OEW) for 1-octanol, where,
in theory, exhaustive extraction is to be expected
Although the partition coefficients change if 1-octanol is re-
placed by another solvent, the principles discussed above are still
valid The here described theoretical trends have also been ob-
served in previous literature [ 29, 30] Thus, a given SLM solvent has
an OEW, where K 1 and K 2 are balanced and exhaustive extraction
can be expected More polar analytes (low K 1) are expected to suf-
fer from slow kinetics, while more non-polar analytes (low K 2) are
prone to membrane trapping
In the following, OEWs were established for different SLMs,
with reference to analyte log P
3.2 Tuning SLM properties with deep eutectic solvents
OEWs were investigated with different deep eutectic solvents
(DES), also including combinations with DEHP as the ionic car-
rier This enabled individual assessment of hydrogen bond-, π-π
-, and ionic interactions-, and their impact on the OEW Four eu-
tectic components, two HBA components, and two HBD compo-
nents were mixed in molar ratios of 1:1 and 1:2 Coumarin and
camphor were selected as HBA components, while thymol and DL-
menthol were selected as HBD components Coumarin and thymol,
provided aromatic character to the deep eutectic solvents Table2
gives an overview of the tested membrane compositions, includ-
ing the number of HBA and HBD sites, and aromatic properties
(aromatic ring count) Mixtures of coumarin and DL-menthol did not form stable deep eutectic mixtures, seen as precipitation after heating, and were therefore not tested DES compositions with or without ionic carrier were selected based on literature and previ- ous experience [31–33]
3.3 Pure eutectic solvents
In the first set of experiments, the effect of hydrogen bond in- teractions was investigated using SLMs with different HBA/HBD ra- tios and an aromatic ring count equal to zero Two SLMs were compared, namely camphor:DL-menthol in molar ratios 1:1 and 1:2 (HBA/HBD ratio 2.0 and 1.5, respectively) To establish mass balance data, both the acceptor and the donor were analyzed Experimental data obtained with camphor:DL-menthol (1:1) are summarized in Fig.3A The results were in accordance with the theoretical discussion in Section3.1 The polar model analytes so- talol, metaraminol, atenolol, tyramine, ephedrine, metoprolol, and pethidine largely remained in the donor These compounds, repre- senting the log P range from −0.4 to 2.5, suffered from slow ki- netics For haloperidol, nortriptyline, loperamide, and methadone, with log P values between 3.7 and 5.0, recoveries were generally high, and this indicated that the extraction was under ideal or near-ideal conditions in the OEW When the molar ratio of cam- phor and DL-menthol was changed from 1:1 to 1:2 (results shown
in Fig.3B), the HBA/HBD ratio decreased from 2.0 to 1.5 The OEW was seemingly unaffected, but the overall membrane trapping in- creased for the model analytes within this region Since the aro- matic ring count was zero with both SLMs, the slight increase in membrane trapping was caused by increased HBD activity
In a next set of experiments, the effect of aromatic ring count was investigated Three different SLMs were compared, namely coumarin:thymol, camphor:thymol, and camphor:DL-menthol, all
in molar ratios of 1:1 These SLMs represented three, one, and zero aromatic ring counts, respectively, while the HBA/HBD ratio was 2.0 in all cases
Experimental data obtained with coumarin:thymol (1:1) are summarized in Fig.3C This SLM was highly aromatic, with a ring count of three Sotalol, metaraminol, atenolol, tyramine, ephedrine, and metoprolol ( −0.4< log P < 1.8) were prone to slow kinetics Pethidine (log P = 2.5), however, was extracted with 65% recovery and was within the OEW for this SLM For haloperidol, nortripty- line, loperamide, and methadone, with log P between 3.7 and 5.0, mass balance data verified significant membrane trapping in the range between 64 and 100% With camphor:thymol (1:1) ( Fig.3D), the aromatic ring count was one Again, the polar analytes in the range −0.4 < log P < 1.8 suffered from slow kinetics and were not extracted from the donor Here, the OEW was shifted towards lower log P values compared to previously discussed SLMs The an- alytes in the log P range from 2.5 to 5.0 were extracted to some extent, but membrane trapping still dominated This was espe- cially evident for loperamide, possibly due to strong π-π inter- actions between the SLM (ring count one) and loperamide (ring count three)
With camphor:DL-menthol (1:1), the SLM was non-aromatic Pethidine was no longer within the OEW, while haloperidol, nor- triptyline, loperamide, and methadone were all extracted with high recoveries ( Fig.3A) Since all the non-polar model analytes are aro- matic, the change to a non-aromatic SLM increased their extrac- tion into the acceptor Therefore, the optimal extraction window was shifted slightly towards higher log P Loperamide was now ex- tracted with high recovery, most probably because the π-π inter- actions were absent No significant membrane trapping was ob- served for this SLM, but is assumed to be present above the tested log P range
Trang 5Fig 3 Relative amounts of model analyte found in the donor (blue), SLM (yellow), and recovered in the acceptor (green) after 60 min of extraction for A) camphor:DL-
menthol (1:1) and B) camphor:DL-menthol (1:2), C) coumarin:thymol (1:1), D) camphor:thymol (1:1) The error bars represent the standard deviation (SD) of the acceptor with n = 4
3.4 Eutectic solvents with DEHP
In the next set of experiments, the effect of ionic interactions
was investigated by adding 0.5, 2, 5, and 10% DEHP to the non-
aromatic SLM camphor:DL-menthol (1:1)
Experimental data obtained with 0.5% DEHP are summarized
in Fig.4A The polar analytes sotalol, metaraminol, atenolol, tyra-
mine, and ephedrine and the non-polar analytes haloperidol, nor- triptyline, and loperamide were all extracted with low recoveries ( < 30%) As observed in the mass balance data the majority of an- alyte was trapped in the SLM Satisfactory extraction recoveries ( > 60%), were measured for the non-polar analytes pethidine and methadone The data show no apparent correlation with analyte log P values When increasing the DEHP percentage to 2% (See
Trang 6Fig 4 Relative amounts of model analyte found in the donor (blue), SLM (yellow), and recovered in the acceptor (green) after 60 min of extraction for A) camphor:DL-
menthol (1:1) + 0.5% DEHP, B) camphor:DL-menthol (1:1) + 2% DEHP, C) camphor:DL-menthol (1:1) + 5% DEHP, and D) camphor:DL-menthol (1:1) + 10% DEHP, E) camphor:DL-menthol (1:1) + 5% DEHP, F) camphor:thymol (1:1) + 5% DEHP, and G) coumarin:thymol (1:1) + 5% DEHP Sample size n = 4 The error bars represent the standard deviation (SD) of the acceptor with n = 4
Trang 7Table 2
Overview of tested membranes, including computer-generated HBA sites, HBD sites, aromatic ring count, and molecular structures of the eutectic components
Aromatic ring count a
Ionic carrier (Y/N)
Coumarin Thymol Camphor DL-menthol DEHP
a Chemicalize was used to generate structure properties (assessed by 01.2020, https://chemicalize.com ) developed by ChemAxon ( https://www.chemaxon.com ) [28]
Fig 4B), the extraction recoveries of the polar analytes generally
increased The increase was both due to increased donor deple-
tion and decreased membrane trapping Extraction recoveries all
decreased for the non-polar analytes From the mass balance data,
it is evident that this was due to an increase in membrane trap-
ping
For higher DEPH concentrations, namely 5 and 10%, similar ob-
servations were made ( Fig.4C and 4D) The extraction recoveries of
the polar analytes were around 50%, except for atenolol, which suf-
fered from membrane trapping For the majority of the non-polar
analytes, the recoveries were below 10%, and mass balance data re-
vealed serious membrane trapping Pethidine was extracted close
to ideal conditions when DEPH concentrations were high, while
the extraction recovery of methadone fluctuated with the higher
DEPH concentrations
From the obtained mass balance data, addition of DEHP gener-
ally increased the extraction of polar analytes from 5- to a 50%-
level This shows that ionic interactions are essential for the ex-
traction of these analytes, even with deep eutectic solvents DEHP
increases the partition into the SLM, and the donor is depleted
For the non-polar analytes, except pethidine and methadone, the
addition of an ionic carrier significantly increased the amount of
membrane trapping For nortriptyline, as an example, membrane
trapping increased from 20 to 93% upon addition of 10% DEHP
It is suspected that the ion exchange at the membrane-acceptor
interface is too weak to accommodate for the strong ionic and
hydrophobic interactions between the non-polar analytes and the
SLM Pethidine and methadone are behaving similarly to the polar
analytes, where higher DEHP concentrations increased donor de-
pletion
In the last set of experiments, the combined effect of aromatic-
ity and ionic interactions were investigated Hydrogen bonding was
not investigated further, as previous experiments revealed an in-
significant effect on the OEW For this last set of experiments,
5% DEHP was added to coumarin:thymol, camphor:thymol, and
camphor:DL-menthol, all in molar ratios of 1:1 These SLMs rep-
resented zero, one, and three aromatic ring counts, respectively;
while the HBA/HBD ratio was equal to 2.0 in all SLMs Mass bal-
ance data are shown in Fig 4 For the polar analytes, the high-
est extraction recoveries were obtained for camphor:thymol + 5% DEHP ( Fig.4F) This SLM has an aromatic ring count equal to one Further increasing or decreasing the aromatic ring count increased the amount of membrane trapping This SLM was the overall best for the polar analytes For the non-polar analytes, the increase
in aromatic ring count decreased the extraction recoveries due to membrane trapping This shows that aromatic SLMs in combina- tion with DEHP are highly unfavorable for the extraction of non- polar analytes
4 Conclusion
In the present study, we proposed the terms optimal extrac- tion window (OEW), slow kinetics, and membrane trapping, to express the log P range where a given three-phase membrane- based liquid-phase microextraction (MB-LPME) system can be ex- pected to be optimal We investigated a selection of supported liq- uid membranes (SLM) based on deep eutectic solvents (DES) with varying HBA/HBD and aromatic ring count, and established their OEWs using a set of pharmaceutical drugs ( −0.4 < log P < 5.0)
as model analytes With pure DES, extraction was primarily fa- cilitated by hydrogen bond and π-π interactions, and the OEWs were typically within the range 2 < log P < 5 Model analytes with log P <∼2 suffered from slow kinetics, while model analytes with log P > ∼3.5 were prone to membrane trapping OEWs were slightly affected by the HBA/HBD ratio and shifted towards higher log P values with increasing aromatic ring count Within the OEWs, the model analytes were extracted with high recoveries, except for highly aromatic ones, which were prone to strong π-πinteractions and membrane trapping Although the pure DES were strong sol- vents regarding hydrogen bond and π-π interactions, they were insufficient for the extraction of polar analytes (log P < 2) With the addition of ionic carrier (DEHP) to the SLMs, polar analytes were efficiently extracted and the OEWs shifted to the range −0.5
< log P < 2 The SLM with 5% DEHP and moderate aromaticity re-
sulted in extraction recoveries of up to 80% for the polar analytes For the non-polar analytes, however, the same SLM suffered from membrane trapping The results from our study show that a given three-phase liquid-phase microextraction system is efficient only
Trang 8in a relatively narrow log P range within the optimal extraction
window (OEW) Minor shifts of the OEW can be expected from al-
terations of HBA, HBD, and aromatic ring count, while major shifts
can be expected when introducing ionic carriers
Identifying OEWs for different SLMs is a step towards a bet-
ter understanding of LPME In the future, this can help users in
routine laboratories to define the possibilities and limitations of a
given LPME system More research on SLMs with an added car-
rier is planned to increase knowledge on carrier-mediated LPME
This will help developing better theoretical models to predict op-
erational conditions and performance
Supplementary information
HPLC-UV method – Additional details; MB-LPME in 96-well for-
mat – Equipment and handling; Calculations for theoretical extrac-
tion model
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests that could have appeared to influence the work re-
ported in this paper
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi: 10.1016/j.chroma.2021.462769
CRediT authorship contribution statement
Maria Schüller: Conceptualization, Methodology, Formal analy-
sis, Investigation, Writing – original draft, Writing – review & edit-
ing, Visualization Kim Tu Thi Tran: Methodology, Formal analysis,
Investigation, Writing – review & editing Elisabeth Leere Øiestad:
Supervision, Writing – review & editing Stig Pedersen-Bjergaard:
Conceptualization, Methodology, Formal analysis, Writing – origi-
nal draft, Writing – review & editing, Supervision
References
[1] S.H Hansen , S Pedersen-Bjergaard , Bioanalysis of pharmaceuticals: sample
preparation, separation techniques, and mass spectrometry, 1st Ed., John Wi-
ley & Sons, Ltd, Chichester, UK, 2015
[2] M Ghambarian , Y Yamini , A Esrafili , Developments in hollow fiber based liq-
uid-phase microextraction: principles and applications, Microchimica Acta 177
(2012) 271–294
[3] S Pedersen-Bjergaard , B Gammelgaard , T.G Halvorsen , Introduction to Phar-
maceutical Analytical Chemistry, 2nd ed., Wiley, Hoboken, NJ, 2019
[4] H Liu , P.K Dasgupta , Analytical chemistry in a drop Solvent extraction in a
microdrop, Anal Chem 68 (1996) 1817–1821
[5] L Araujo , M.E Troconis , D Cubillán , J Mercado , N Villa , A Prieto ,Single drop
microextraction and gas chromatography–mass spectrometry for the determi-
nation of diflufenican, mepanipyrim, fipronil, and pretilachlor in water sam-
ples, Environ Monit Assess 185 (2013) 10225–10233
[6] S Pedersen-Bjergaard , K.E Rasmussen , Liquid-liquid-liquid microextraction for
sample preparation of biological fluids prior to capillary electrophoresis, Anal
Chem 71 (1999) 2650–2656
[7] M Rezaee , Y Assadi , M.-R.Milani Hosseini , E Aghaee , F Ahmadi , S Berijani ,
Determination of organic compounds in water using dispersive liquid–liquid
microextraction, J Chromatogr A 1116 (2006) 1–9
[8] M.-.I Leong , S.-.D Huang , Dispersive liquid–liquid microextraction method
based on solidification of floating organic drop combined with gas chromatog-
raphy with electron-capture or mass spectrometry detection, J Chromatogr A
1211 (2008) 8–12
[9] X Jiang , H.K Lee , Solvent bar microextraction, Anal Chem 76 (2004)
5591–5596
[10] A Gjelstad , A T Andresen , A Dahlgren , T.E Gundersen , S Pedersen-Bjergaard ,
High-throughput liquid-liquid extraction in 96-well format: parallel artificial
liquid membrane extraction, LC GC Eur 30 (2017) 10–17
[11] L Vårdal , G Wong , ˚A.M.L Øiestad , S Pedersen-Bjergaard , A Gjelstad , E.L Øies- tad , Rapid determination of designer benzodiazepines, benzodiazepines, and Z-hypnotics in whole blood using parallel artificial liquid membrane extrac- tion and UHPLC-MS/MS, Anal Bioanal Chem 410 (2018) 4 967–4 978 [12] L Vårdal , E.L Øiestad , A Gjelstad , S Pedersen-Bjergaard , Electromembrane ex- traction of substances with weakly basic properties: a fundamental study with benzodiazepines, Bioanalysis 10 (2018) 769–781
[13] S Pedersen-Bjergaard , K.E Rasmussen , Electrokinetic migration across artifi- cial liquid membranes New concept for rapid sample preparation of biological fluids, J Chromatography A 1109 (2006) 183
[14] P.K Jagtap , K Tapadia , Pharmacokinetic determination and analysis of nor- triptyline based on GC–MS coupled with hollow-fiber drop-to-drop solvent mi- croextraction technique, Bioanalysis 10 (2018) 143–152
[15] T Kuuranne , T Kotiaho , S Pedersen-Bjergaard , K.Einar Rasmussen , A Leinonen ,
S Westwood , R Kostiainen , Feasibility of a liquid-phase microextraction sample clean-up and liquid chromatographic/mass spectrometric screening method for selected anabolic steroid glucuronides in biological samples, J Mass Spectrometry 38 (2003) 16–26
[16] E Zeini Jahromi , A Bidari , Y Assadi , M.R Milani Hosseini , M.R Jamali , Dis- persive liquid–liquid microextraction combined with graphite furnace atomic absorption spectrometry: ultra trace determination of cadmium in water sam- ples, Anal Chim Acta 585 (2007) 305–311
[17] L Dong , X Shen , C Deng , Development of gas chromatography–mass spec- trometry following headspace single-drop microextraction and simultaneous derivatization for fast determination of the diabetes biomarker, acetone in hu- man blood samples, Anal Chim Acta 569 (2006) 91–96
[18] I Kohler , J Schappler , T Sierro , S Rudaz , Dispersive liquid–liquid microex- traction combined with capillary electrophoresis and time-of-flight mass spec- trometry for urine analysis, J Pharm Biomed Anal 73 (2013) 82–89 [19] Z Zhang , D Wang , L Zhang , M Du , G Chen , Determination of diuretics in hu- man urine by hollow fiber-based liquid-liquid-liquid microextraction coupled
to high performance liquid chromatography, Analyst 133 (2008) 1187–1194 [20] M Drobnjak , F.A Hansen , E Øiestad , T Løvli , R Trones , Ø Martinsen , S Peder- sen-Bjergaard , Electromembrane extraction with vials of conducting polymer,
LC GC N Am 38 (2020) 435 [21] M.F dos Santos , C.C Ferri , S.C Seulin , V Leyton , C.A.G Pasqualucci , D.R Muñoz , M Yonamine , Determination of antidepressants in whole blood using hollow-fiber liquid-phase microextraction and gas chromatography–mass spectrometry, Forensic Toxicol 32 (2014) 214–224
[22] K.S Ask , M Lid , E.L Øiestad , S Pedersen-Bjergaard , A Gjelstad , Liquid-phase microextraction in 96-well plates - calibration and accurate quantification of pharmaceuticals in human plasma samples, J Chromatography A 1602 (2019) 117–123
[23] R.A Menck , C.D.R de Oliveira , D.S de Lima , L.E Goes , V Leyton , C.A Pasqualucci , D.R Muñoz , M Yonamine , Hollow fiber–liquid phase microex- traction of barbiturates in liver samples, Forensic Toxicol 31 (2013) 31–36 [24] J Zhang , M Zhang , S Fu , T Li , S Wang , M Zhao , W Ding , C Wang , Q Wang , Simultaneous determination of imperatorin and its metabolite xanthotoxol in rat plasma by using HPLC–ESI-MS coupled with hollow fiber liquid phase mi- croextraction, J Chromatogr B 945-946 (2014) 185–192
[25] S Yudthavorasit , C Chiaochan , N Leepipatpiboon , Simultaneous determination
of multi-class antibiotic residues in water using carrier-mediated hollow-fiber liquid-phase microextraction coupled with ultra-high performance liquid chro- matography tandem mass spectrometry, Microchimica Acta 172 (2011) 39–49 [26] S Knoll , T Rösch , C Huhn , Trends in sample preparation and separation meth- ods for the analysis of very polar and ionic compounds in environmental water and biota samples, Anal Bioanal Chem 412 (2020) 6149–6165
[27] A Gjelstad , H Jensen , K.E Rasmussen , S Pedersen-Bjergaard , Kinetic aspects
of hollow fiber liquid-phase microextraction and electromembrane extraction, Anal Chim Acta 742 (2012) 10–16
[28] "Chemicalize." Access date: 27.02.2021, from http://www.chemicalize.com [29] S Pedersen-Bjergaard , K.E Rasmussen , A Brekke , T.S Ho , T.Grønhaug Halvorsen , Liquid-phase microextraction of basic drugs – se- lection of extraction mode based on computer calculated solubility data, J Sep Sci 28 (2005) 1195–1203
[30] S Pedersen-Bjergaard , T Ho , K Rasmussen , Fundamental studies on selectivity
in 3-phase liquid-phase microextraction (LPME) of basic drugs, J Sep Sci 25 (2002) 141–146
[31] F.A Hansen , E Santigosa-Murillo , M Ramos-Payán , M Muñoz , E Leere Øies- tad , S Pedersen-Bjergaard , Electromembrane extraction using deep eutectic solvents as the liquid membrane, Anal Chim Acta 1143 (2021) 109–116 [32] P Mako ´s , A Przyjazny , G Boczkaj , Hydrophobic deep eutectic solvents as
“green” extraction media for polycyclic aromatic hydrocarbons in aqueous samples, J Chromatogr A 1570 (2018) 28–37
[33] D.J.G.P Van Osch , C.H.J.T Dietz , J Van Spronsen , M.C Kroon , F Gallucci , M Van Sint Annaland , R Tuinier , A search for natural hydrophobic deep eutectic solvents based on natural components, ACS Sustain Chem Eng 7 (2019) 2933–2942