A comprehensive study of a new versatile microchip device based liquid phase microextraction for stopped-flow and double-flow conditions

1 A comprehensive study of a new versatile microchip device based liquid phase microextraction 1 for stopped-flow and double-flow conditions. 2 María Ramos Payána*, Elia Santigosa Murillob, Jordi Coellob, Miguel Ángel Bello Lópeza 3 aDepartment of Analytical Chemistry, Faculty of Chemistry, c/Prof. García González s/n, 41012, 4 Seville, Spain 5 bDepartment of Analytical Chemistry, Universitat Autónoma de Barcelona, 08193 Bellaterra, 6 Barcelona, Spain 7 *corresponding author: [email protected] telephone number: +34954557172 8 Abstract 9 A new geometry for a versatile microfluidic-chip device based liquid phase microextraction 10 was developed in order to enhance the preconcentration in microfluidic chips and also to enable 11 double-flow and stopped-flow working modes. The microchip device was combined with a HPLC 12 procedure for the simultaneous determination of two different families as model analytes, which were 13 parabens and non-steroidal anti-inflammatories (NSAIDs): Ethyl 4-hydroxybenzoate (Et-P), Propyl 414 hydroxybenzoate (Pr-P), Butyl 4-hydroxybenzoate (Bu-P), IsoButyl 4-hydroxybenzoate (iBu-P), 15 salycilic acid (SAC), ketoprofen (KET), naproxen (NAX), diclofenac (DIC) and ibuprofen (IBU) in 16 urine samples. The new miniaturized microchip proposed in this work allows not only the possibility 17 of working in double-flow conditions, but also under stagnant conditions (stopped-flow) (SF-µLPME). 18 The sample (pH 1.5) was delivered to the SF-µLPME at 20 µL min-1 while keeping the acceptor phase 19 (pH 11.75) under stagnant conditions during 20 minutes. The highest enrichment factors (between 16 20 and 47) were obtained under stopped-flow conditions at 20 µL min-1 (sample flow rate) after 20 min 21 extraction; whereas the extraction efficiencies were within the range of 27-81% for all compounds. 22 The procedure provided very low detection limits between 0.7 and 8.5 µg L-1 with a sample volume 23 consumption of 400 µL. Parabens and NSAIDs have successfully been extracted from urine samples 24 with excellent clean up and recoveries over 90 % for all compounds. In parallel, the new device was 25 also tested under double flow conditions, obtaining good but lower enrichment factors (between 9 and 26 2 20) and higher extraction efficiencies (between 45 and 95) after 7 min extraction, consuming a volume 27 sample of 140 µL . 28 The versatile device offered very high extraction efficiencies and good enrichment factor for double 29 flow and stopped-flow conditions, respectively. In addition, this new miniaturized SF-µLPME device 30 significantly reduced costs compared to the existing analytical techniques for sample preparation since 31 this microchip require few microliters of sample and reagents and it is reusable. 32 Keywords: microextraction, miniaturization, sample preparation, microfluidic, drugs, urine sample. 33 1. Introduction 34 Liquid Phase microextraction (LPME) is a very well-known and popular technique used for the 35 extraction of acid and basic drugs based on the passive diffusion of the analytes from the sample (donor 36 solution) into an acceptor solution, through a membrane (which support an organic solvent into its 37 porous). LPME has been applied to many different fields, considering biological, pharmaceutical, 38 environmental, food, toxicology analysis, among others [1–7]. The transport phenomena based on 39 passive diffusion depends not only on the nature of the analytes, and the optimal parameters (as phase’s 40 composition, organic solvent, stirring speed, flow rate, etc), but also on the geometry of the system 41 used for LPME. Another popular technique based liquid phase microextraction, named 42 electromembrane extraction (EME), has also been frequently used since it improves the extraction of 43 compounds in many cases due to an external electrical field created to both sides of the support liquid 44 membrane [8–16]. However, EME also offers some limitations since its requirement is the use of a 45 suitable and conductor organic solvent for carrying out the extractions. Both techniques have been 46 widely used for the determination of pharmaceutical drugs either in biological samples (urine) or water 47 samples [17-23] due to the great concern that exists regarding their contribution as emergent pollutants 48 in the environment. Also, parabens have been studied due to the concern about their endocrine 49 disrupting potential [24-29]. This has required the use of powerful, fast and sensitive techniques that 50 offer better limits of quantification. 51 3 Up to date, parabens and non-steroidal antiinflamatories haven been determined by traditional LPME 52 and EME procedures resulting in very good enrichment factors [1-4,17,19, 20, 30-33]. Those 53 procedures allowed good enrichment factors but low extraction efficiencies. In the last years, liquid54 liquid extraction has been miniaturized into microfluidic devices in order to address the limitations 55 from traditional procedures and these chip devices are becoming an attractive alternative due to the 56 many advantages that it presents [34-42]. The microchip devices for sample treatment have two 57 channels that allow working in two different ways based on the flow rate of each phase: double-flow 58 or stopped-flow conditions. In double-flow conditions, both phases (sample and acceptor) are moving 59 at some flow rate. However, in stopped-flow conditions, the acceptor phase keep stagnant while the 60 sample solution is used at some flow rate. Recent microchip devices based LPME, have been 61 demonstrated to work only under double-flow conditions but not under stopped-flow conditions in a 62 single step since the latter required to collect several extracts for its direct injection into HPLC. This 63 was due to the low sample volume available in the acceptor channel ( 2 µL) [41,42] and consequently, 64 the analysis time increased and the reproducibility decreased when an enrichment factor was necessary 65 prior to the sample analysis. On the other hand, the devices did not allowed high preconcentration 66 factors although the sample flow rate was significantly increased under double-flow conditions. 67 Based on the current limitations of microfluidic devices for microextraction procedures, the aim of this 68 work was to develop a new versatile and effective microfluidic device in order to overcome the 69 limitations from previous microfluidic devices, increasing the preconcentration and allowing working 70 under stopped-flow conditions compatible with direct analysis. 71 Based on geometry aspects, an increase of the depth channel would increase the volume capacity 72 contained in the channel but it could decrease the transport phenomena by passive diffusion since the 73 analytes are farther away from the membrane. Microfluidic systems that follow a laminar regimen do 74 not carry agitation, so diffusion can be slow if the distance between the analytes and extraction solvent 75 is increased. Moreover, an increase of the channel´s width would increase the contact surface between 76 4 the sample and the analytes, however, very wide channels could destabilize the laminar flow and affect 77 the membrane stability. 78 In this work, we present for the first time a new versatile and effective microfluidic chip based LPME 79 which allow the possibility of working under two different working modes (double-flow or stopped80 flow conditions). The microchip was applied to the simultaneous determination of two different 81 families in urine samples. This way, a comprehensive study between both different working conditions 82 was carried out. The microchip decreased the sample volume and time of analysis since no collecting 83 samples were needed for direct injection. The proposed stopped-flow device (SFµLPME) is the 84 easiest microfluidic chip for the simultaneous extractions of different drugs resulting in higher 85 enrichment factors with lower cost instrumentation, simple handling, reusability and is still considered 86 a “green method” by keeping low organic solvent (< 5µL) consumption. The proposed device has been 87 successfully applied to urine samples. 88 2. Experimental 89 2.1. Chemicals and solutions 90 Ethyl 4-hydroxybenzoate (Et-P), Propyl 4-hydroxybenzoate (Pr-P), Butyl 4-hydroxybenzoate (Bu-P), 91 IsoButyl 4-hydroxybenzoate (iBu-P), salicylic acid (SAC), ketoprofen (KTP), naproxen (NPX), 92 diclofenac (DIC), ibuprofen (IBU),1-octanol, dihexyl ether, 2-nitrophenyl octhyl ether (NPOE), 93 formic acid, sodium hydroxide, chloride acid,sodium chloride and methanol were purchased from 94 Fluka–Sigma–Aldrich (Madrid, Spain). 100 mg L-1 stocks solutions were prepared in methanol except 95 SAC, DIC and IBU that were prepared in Milli-Q Plus water (Elga, purelab option S-R 7-15 (Madrid, 96 Spain). All working dilutions were prepared using ultrapure water from a Milli-Q Plus by adequate 97 dilutions from stored at 4ºC. A membrane (Celgard 2500) of 25 µm thickness, 55% porosity, and 0.21 98 µm x 0.05 µm pores was obtained from Celgard (Charlotte, NC, USA). 99 2.2 Fabrication of the microfluidic-chip device 100 5 Figure 1 shows a scheme of the microfluidic device based liquid phase microextraction. This 101 microfluidic device has been re-designed and modified in order to overcome the limitations and 102 disadvantages from previous microfluidic devices. The optimal poly(methyl methacrylate(PMMA) 103 device consisted of two symmetrical patterned plates with one channel of 23 mm length, 120 µm depth 104 and 3 mm width each. Four holes of 3 mm and 1.35 mm diameter were drilled for assembling and 105 fixing in/outlets Teflon tubes, respectively. A flat polypropylene membrane piece of 27 mm length x 106 5 mm width separated the acceptor phase (channel 1) and the donor phase (channel 2). Firstly, the 107 membrane was placed over one channel and impregnated with 4 µL of dihexyl ether. Once the 108 extracting solvent was immobilized along the membrane by capillary forces, the channels were aligned 109 and the device was closed using four small crews. The final size of a microfluidic device for one single 110 extraction was 47×29×6 mm, however by increasing the size of both PMMA plates, an arbitrarily large 111 number of extraction channels can be implemented and independently addressed. Also, the microchip112 device can be opened any time when exchange membrane is needed. 113 A laser cutter (Epilog Mini 24-30 W) was used to fabricate this chip. The best quality was obtained 114 using a writing speed of 40%, power of 33%, a resolution of 1500 and a frequency of 5000. 115 Inlets Teflon tubes (acceptor and donor inlets) were connected to two separate micro-syringe pumps 116 (Cetoni GmbH, Korbussen, Germany). The sample (pH 1.5) was pumped into the microfluidic device 117 at 20 µLmin-1 while keeping the acceptor phase (pH 11.75) constant. The microfluidic device was also 118 tested under double-flow conditions as described below, in order to compare different working modes. 119 The acceptor phase was collected using a micropipette and was directly injected into a HPLC for 120 analysis. 121 2.3. Chromatographic conditions 122 An Agilent 1100 series (Barcelona, Spain) liquid chromatography equipped with a G1312A Bipump 123 and an autosamplerG1313A for 5 µL of sample injection was used as HPLC system. The column used 124 for the separation of the nine compounds was a LiChroCART® 75-4 Purospher® STAR RP-18e 3 µm 125 6 (75 mm x 4.0 mm i.d.) (VWR, Barcelona, Spain) proceeded by a guard column Kromasil1 100 Å, C18, 126 5 µm (20 mm x 4.6 mm i.d.) (Scharlab S.L., Barcelona, Spain). 127 The mobile phase consisted of 0.1% formic acid (pH 2.6) (component A) and methanol (component 128 B) at a flow rate of 0.5 mL min-1. Separation was performed at 25ºC. An initial 60% component B 129 was used in isocratic mode for 2 min, and then a linear elution gradient was programmed from 60% to 130 80% (B) for 3.4 min and from 80 % to 86 % B for another 2.4 minutes. Three minutes were waited 131 between injections which allowed re-equilibration of the column to the initial conditions. 132 The wavelengths used for DAD were 235, 255, 230, 280 and 225 nm for SAC, KTP, NAX, DIC and 133 IBU, respectively and 255 nm for all parabens. The chromatogram was completed in less than 10 134 minutes and the retention time was 3.1, 3.3, 4.7, 5.3, 6.3, 6.6, 6.8, 8.9 and 9.1, for SAC, Et-P, Pr-P, 135 KTP, NPX, iBu-P, Bu-P, DIC and IBU, respectively. 136 2.4. Preparation of biological samples analysis using µLPME extraction 137 Spiked urine samples were adjusted to pH 1.5 with HCl and filtered through Pall NylafloTM nylon 138 membrane filter 0.45 µm (Pall Corporation, Ann Arbor, Michigan, USA) prior to microextraction 139 procedure. 140 2.5. Calculations of extraction efficiency and enrichment factor 141 The enrichment factor (EFi) for the analyte i was calculated according to the following equation (1): 142 143 𝐸𝐹𝑖=𝐶𝑓,𝑎,𝑜𝑢𝑡𝑙𝑒𝑡 𝐶𝑖,𝑠,𝑖𝑛𝑙𝑒𝑡 (1) 144 145 where𝐶𝑓,𝑎,𝑜𝑢𝑡𝑙𝑒𝑡 is the concentration of the analyte i at the outlet of the acceptor channel and 𝐶𝑖,𝑠,𝑖𝑛𝑙𝑒𝑡 146 is the initial concentration of the analyte in the sample. 𝐶𝑓,𝑎,𝑜𝑢𝑡𝑙𝑒𝑡was determined by HPLC UV147 detection using external calibration. The enrichment factor is calculated using the same equation either 148 using double-flow or stopped-flow conditions. The extraction efficiency (EE) was defined as the 149 7 fraction of analyte transferred to the acceptor phase from the sample. Using a double-flow working 150 mode, the extraction efficiency (EE %) was calculated according to the following equation (2): 151 𝐸𝐸 (%)= 𝐶𝑓,𝑎,𝑜𝑢𝑡𝑙𝑒𝑡 𝐶𝑖,𝑠,𝑖𝑛𝑙𝑒𝑡 𝑥 𝑣𝑎 𝑣𝑠 𝑥 100= 𝐸𝐹𝑖 𝑥 𝑣𝑎 𝑣𝑠 𝑥 100 (2) 152 Where 𝑣𝑎 𝑎𝑛𝑑 𝑣𝑠, are the acceptor and sample flow rate, respectively. 153 However, under stopped-flow conditions, the extraction efficiency (EE %) was calculated by 154 substituting the parameter “acceptor and sample flow rate” by the “acceptor and sample volume” 155 corresponding to each phase sample. 156 In order to obtain a global EE value for the 9 analytes, the average extraction efficiency index (avEEi) 157 was defined (3): 158 𝐴𝑣𝐸𝐸𝑖=1 − √∑(𝐸𝐸𝑖−100)2 𝑛 𝑖=1 𝑛 (3) 159 160 3. Results and discussion 161 3.1. Principle of the extraction 162 The model analytes corresponded to two different families: non-steroidal antiinflamatories and 163 parabens. The extraction of the analytes is based on a passive diffusion process due to a strong pH 164 gradient difference between the acceptor and the sample solution. Non-steroidal antiinflamatories 165 contain acid groups within a pKa range value of 2.5-5, while the paraben are esters of 166 parahydroxybenzoic acid and contain alcohol group which pKa value are within the range of 5-8.8. A 167 three phases liquid phase microextraction configuration presents two aqueous solutions (acceptor and 168 sample) separated by the support liquid membrane (organic solvent). The analytes of interest were in 169 neutral form in the sample solution and negatively charged in the acceptor solution. This way, a pH 170 value of under 7 (HCl solution) and over 9 (NaOH solution) were used as sample and acceptor solution, 171 respectively.The microfluidic device was tested using two different working modes: double-flow mode 172 and stopped-flow conditions. The membrane was reused for consecutive extractions without observing 173 8 memory effects and the acceptor phase collected was analyzed by HPLC once the extraction was 174 completed. Under stopped-flow conditions, the acceptor phase was pumped continuously for at least 175 2-3 minutes between extractions to clean the SLM avoiding memory effects. 176 3.2. Optimization of the microchip´s geometry 177 In LPME, the extraction of the analytes depends on a transport phenomenon based on passive 178 diffusion. The design of a new geometry was focused on the increasement of the channel volume 179 capacity (compatible with direct injection into HPLC) without decelerating the transport phenomena. 180 Different length, wide and depth were tested in order to obtain an adequate and stable laminar flow 181 during the extraction, considering a final channel volume capacity between 7 and 10 µL for its direct 182 analysis by HPLC after stopped-flow conditions. The length was fixed at 23 mm and the wide and 183 depth were tested between 1-3 mm and 50-300 µm, respectively. The depth was the most critical 184 parameter and it was limited to 120 µm since an increased depth significantly decelerated the transport 185 phenomena. In one hand, a less deep channel kept high extraction efficiencies under double-flow 186 conditions (over 90%) but the channel volume capacity was not enough for working under stopped187 flow conditions. On the other hand, a depth over 150 µm decreased the extraction efficiency under 188 double-flow conditions (less than 70%) and the enrichment factor decreased 20 % for all compounds 189 under stopped-flow conditions. Additionally, a wide of 2 mm required a deeper channel in order to 190 increase the volume capacity and it decreased the extraction efficiencies and a wide of 4 mm did not 191 offer good reproducibility and stable flow rate. For this reasons, a compromise between depth, length 192 and width was carried out to increase transport phenomena and channel´s volume but still maintaining 193 miniaturization size and simple handling conditions. The best results and the most reproducible and 194 stable flow were obtained with a channel geometry of 23 mm length, 120 µm depth and 3 mm width. 195 Based on the fundamental basis for LPME, this new geometry presents longer and wider channels for 196 increasing the contact area between the analytes and the support liquid membrane compared to 197 9 previous one made on PMMA. Additionally, the depth was increased to allow a greater volume 198 capacity compared to the only 2 µL volume capacity from the previous µLPME device [41]. 199 3.3. Optimization and evaluation of experimental conditions 200 According to data founded in the bibliography for individual microextraction of parabens or non201 steroidal antiinflamatories, preliminary experiments were performed to determinate the most suitable 202 organic solvent to be used as SLM for the simultaneous extraction of both families. For the 203 optimization of the organic solvent, a pH 3 (HCl) sample solution, pH 11.75 (NaOH) as acceptor 204 solution, and 1 µL min-1 as sample and acceptor flow rate were used. 2-nitrophenil octyl ether (NPOE), 205 1-octanol, 1-heptanol and dihexylether were tested by opening the microfluidic device for membrane 206 exchange after each organic solvent test. As seen in Table 1, the extraction efficiency (based on 3 207 replicate experiments) was very different depending on the analyte. Best avEEi was obtained when 208 dihexylether was used as support liquid membrane, which was consequently used as SLM for the rest 209 of the study. 210 For optimization of sample and acceptor composition, the acceptor and donor phase were tested within 211 the ranges of pH 10-12 (aqueous NaOH solutions) and 1-4 (aqueous HCl solutions), respectively. The 212 donor phase, containing 1 mg mL-1 of each analyte, was tested keeping the acceptor phase fixed at pH 213 11.75. As seen in Figure 2, the highest peak areas were obtained at pH 1.5 after 7 minutes extraction, 214 not observing a significant decrease for the rest of the pH range tested. Then, acceptor phase 215 composition was optimized by keeping the sample solution fixed at pH 1.5 for all experiments. Figure 216 3 shows that the highest peak areas were obtained at pH 12 and pH 11.75 for five non-steroidal 217 antiinflamatories and four parabens, respectively. On the other hand, parabens were not stable over pH 218 12 due to a slight degradation during their extraction, so a compromised pH of 11.75 was selected for 219 the extraction of both families. A relative standard deviation (RSDs %) below 4 % for all analytes 220 resulted based on 3 replicate experiments of each experimental point for Figure 2 and 3. Consequently, 221 16 liquid membranes. Tuning the membrane chemistry to different types of drug substances, J. 370 Chromatogr. A. 1124 (2006) 29–34. 371 [9] S. Pedersen-Bjergaard, K.E. Rasmussen, Electrokinetic migration across artificial liquid 372 membranes: New concept for rapid sample preparation of biological fluids, J. Chromatogr. A. 373 1109 (2006) 183–190. 374 [10] S. Pedersen-Bjergaard, K.E. Rasmussen, Extraction across supported liquid membranes by use 375 of electrical fields, Anal. Bioanal. Chem. 388 (2007) 521–523. 376 [11] S. Pedersen-Bjergaard, K.E. Rasmussen, Electrical potential can drive liquid-liquid extraction 377 for sample preparation in chromatography, TrAC - Trends Anal. Chem. 27 (2008) 934–941. 378 [12] T.M. Middelthon-Bruer, A. Gjelstad, K.E. Rasmussen, S. Pedersen-Bjergaard, Parameters 379 affecting electro membrane extraction of basic drugs, J. Sep. Sci. 31 (2008) 753–759. 380 [13] M.D.R. Payán, B. Li, N.J. Petersen, H. Jensen, S.H. Hansen, S. Pedersen-Bjergaard, Nano381 electromembrane extraction, Anal. Chim. Acta. 785 (2013) 60–66. 382 [14] M. Rezazadeh, Y. Yamini, S. Seidi, B. Ebrahimpour, Electromembrane surrounded solid 383 phase microextraction: A novel approach for efficient extraction from complicated matrices, J. 384 Chromatogr. A. 1280 (2013) 16–22. 385 [15] A. Gjelstad, H. Jensen, K.E. Rasmussen, S. Pedersen-Bjergaard, Kinetic aspects of hollow 386 fiber liquid-phase microextraction and electromembrane extraction, Anal. Chim. Acta. 742 387 (2012) 10–16. 388 [16] L.E.E. Eibak, K.E. Rasmussen, E.L. ??iestad, S. Pedersen-Bjergaard, A. Gjelstad, Parallel 389 electromembrane extraction in the 96-well format, Anal. Chim. Acta. 828 (2014) 48–52. 390 [17] M. Ramos-Payán, R. Fernández-Torres, J.L. Pérez-Bernal, M. Callejón-Mochón, M. Ángel 391 Bello-López, A novel approach for electromembrane extraction based on the use of silver 392 nanometallic-decorated hollow fibers, Anal. Chim. Acta. 849 (2014) 7–11. 393 [18] K.S. Hasheminasab, A.R. Fakhari, A. Shahsavani, H. Ahmar, A new method for the 394 17 enhancement of electromembrane extraction efficiency using carbon nanotube reinforced 395 hollow fiber for the determination of acidic drugs in spiked plasma, urine, breast milk and 396 wastewater samples, J. Chromatogr. A. 1285 (2013) 1–6. 397 [19] S.S.H. Davarani, A. Pourahadi, S. Nojavan, M.H. Banitaba, M. Nasiri-Aghdam, Electro 398 membrane extraction of sodium diclofenac as an acidic compound from wastewater, urine, 399 bovine milk, and plasma samples and quantification by high-performance liquid 400 chromatography, Anal. Chim. Acta. 722 (2012) 55–62. 401 [20] M. Ramos-Payán, M.Á. Bello-López, R.F. Torres, M. Villar-Navarro, M. Callejón-Mochón, 402 Electromembrane extraction (EME) and HPLC determination of non-steroidal anti403 inflammatory drugs (NSAIDs) in wastewater samples, Talanta. 85 (2011) 394–399. 404 [21] C. Lacey, G. McMahon, J. Bones, L. Barron, A. Morrissey, J.M. Tobin, An LC-MS method for 405 the determination of pharmaceutical compounds in wastewater treatment plant influent and 406 effluent samples, Talanta. 75 (2008) 1089–1097. 407 [22] D.S.M. Shukri, M.M. Sanagi, W.A.W. Ibrahim, N.N.Z. Abidin, H.Y. Aboul-Enein, Liquid 408 Chromatographic Determination of NSAIDs in Urine After Dispersive Liquid???Liquid 409 Microextraction Based on Solidification of Floating Organic Droplets, Chromatographia. 78 410 (2015) 987–994. 411 [23] J. Zhang, H.K. Lee, Application of dynamic liquid-phase microextraction and injection port 412 derivatization combined with gas chromatography-mass spectrometry to the determination of 413 acidic pharmaceutically active compounds in water samples, J. Chromatogr. A. 1216 (2009) 414 7527–7532. doi:10.1016/j.chroma.2009.03.051. 415 [24] D.G.J. Larsson, C. de Pedro, N. Paxeus, Effluent from drug manufactures contains extremely 416 high levels of pharmaceuticals, J. Hazard. Mater. 148 (2007) 751–755. 417 [25] M.G. Soni, S.L. Taylor, N.A. Greenberg, G.A. Burdock, Evaluation of the health aspects of 418 methyl paraben: a review of the published literature, Food Chem. Toxicol. 40 (2002) 1335– 419 18 1373. 420 [26] M.G. Soni, G.A. Burdock, S.L. Taylor, N.A. Greenberg, Safety assessment of propyl paraben: 421 A review of the published literature, Food Chem. Toxicol. 39 (2001) 513–532. 422 [27] M.G. Soni, I.G. Carabin, G.A. Burdock, Safety assessment of esters of p-hydroxybenzoic acid 423 (parabens), Food Chem. Toxicol. 43 (2005) 985–1015. 424 [28] R. Golden, J. Gandy, G. Vollmer. A review of the endocrine activity of parabens and 425 implications for potential risk to human health. Crit Rev Toxicol. 35 (2005) 435-58. 426 [29] J.M. Brausch, G.M. Rand, A review of personal care products in the aquatic environment: 427 Environmental concentrations and toxicity, Chemosphere. 82 (2011) 1518–1532. 428 [30] M. Balchen, A. Gjelstad, K.E. Rasmussen, S. Pedersen-Bjergaard, Electrokinetic migration of 429 acidic drugs across a supported liquid membrane, J. Chromatogr. A. 1152 (2007) 220–225. 430 [31] A. Esrafili, Y. Yamini, M. Ghambarian, M. Moradi, Analysis of paraben preservatives in 431 cosmetic samples: Comparison of three different dynamic hollow fiber liquid-phase 432 microextraction methods, Chromatographia. 77 (2014) 317–327. 433 [32] A. Prichodko, K. Jonusaite, V. Vickackaite, Hollow fibre liquid phase microextraction of 434 parabens, Cent. Eur. J. Chem. 7 (2009) 285–290. 435 [33] M. Villar-Navarro, M. del C. Moreno-Carballo, R. Fernández-Torres, M. Callejón-Mochón, 436 M.Á. Bello-López, Electromembrane extraction for the determination of parabens in water 437 samples, Anal. Bioanal. Chem. 408 (2016) 1615–1621. 438 [34] N.J. Petersen, J.S. Pedersen, N.N. Poulsen, H. Jensen, C. Skonberg, S.H. Hansen, S. Pedersen439 Bjergaard, On-chip electromembrane extraction for monitoring drug metabolism in real time 440 by electrospray ionization mass spectrometry., Analyst. 137 (2012) 3321–7. 441 [35] Y. Abdossalami Asl, Y. Yamini, S. Seidi, B. Ebrahimpour, A new effective on chip 442 electromembrane extraction coupled with high performance liquid chromatography for 443 enhancement of extraction efficiency, Anal. Chim. Acta. 898 (2015) 42–49. 444 19 [36] B.C. Giordano, D.S. Burgi, S.J. Hart, A. Terray, On-line sample pre-concentration in 445 microfluidic devices: A review, Anal. Chim. Acta. 718 (2012) 11–24. 446 [37] Z.X. Cai, Q. Fang, H.W. Chen, Z.L. Fang, A microfluidic chip based liquid-liquid extraction 447 system with microporous membrane, Anal. Chim. Acta. 556 (2006) 151–156. 448 [38] P. Wägli, Y.C. Chang, A. Homsy, L. Hvozdara, H.P. Herzig, N.F. De Rooij, Microfluidic 449 droplet-based liquid-liquid extraction and on-chip IR spectroscopy detection of cocaine in 450 human saliva, Anal. Chem. 85 (2013) 7558–7565. 451 [39] M.D. Ramos Payán, H. Jensen, N.J. Petersen, S.H. Hansen, S. Pedersen-Bjergaard, Liquid452 phase microextraction in a microfluidic-chip - High enrichment and sample clean-up from 453 small sample volumes based on three-phase extraction, Anal. Chim. Acta. 735 (2012) 46–53. 454 [40] B. Li, N.J. Petersen, M.D.R. Pay??n, S.H. Hansen, S. Pedersen-Bjergaard, Design and 455 implementation of an automated liquid-phase microextraction-chip system coupled on-line 456 with high performance liquid chromatography, Talanta. 120 (2014) 224–229. 457 [41] M. Ramos-Payán, S. Maspoch, A. Llobera, An effective microfluidic based liquid-phase 458 microextraction device (μLPME) for extraction of non-steroidal anti-inflammatory drugs from 459 biological and environmental samples, Anal. Chim. Acta. 946 (2016) 56–63. 460 [42] M. Ramos-Payán, S. Maspoch, A. Llobera, A simple and fast Double-Flow microfluidic device 461 based liquid-phase microextraction (DF-µLPME) for the determination of parabens in water 462 samples, Talanta. 165 (2017) 496–501. 463 464 465 466 467 468 469 20 470 471 472 473 474 475 476 477 Leyend for the tables and figures captions 478 Figure 1. Schematic of the microchip device based liquid phase microextraction 479 Figure 2. Optimization of the donor phase composition. SLM: dihexylether, flow rate (donor and 480 acceptor phase): 1 µL min-1, acceptor phase composition: pH 11.75 481 Figure 3. Optimization of the acceptor phase composition. SLM: dihexylether, flow rate (donor phase): 482 1 µL min-1, flow rate (acceptor phase): 1 µL min-1, donor phase composition: pH 1.5 483 Figure 4. Extraction efficiency versus sample flow rate. SLM: dihexylether, flow rate (acceptor phase): 484 1 µL min-1, donor phase composition: pH 11.5 and acceptor phase composition: pH 1.5 485 Figure 5. Extraction enrichment versus sample phase flow rate. SLM: dihexylether, flow rate (acceptor 486 phase): 1 µL min-1, donor phase composition: pH 11.75 and acceptor phase composition: pH 1.5 487 Figure 6. Optimization of sample solution flow rate and extraction time for SAC, Et-P, Pr-P, NPX, 488 KTP, iBu-P, Bu-P, DIC and IBU. 489 Figure 7.Chromatogram of a spiked urine sample containing 16 µg L-1 for all compounds except for 490 IBU that was 30 µg L-1. Extraction time: 20 minutes. SLM: diexylether; donor phase composition: pH 491 11.75 and acceptor phase composition: pH 1.5. Sample flow rate: 20 µL min-1. No sample dilution 492 Table 1. Extraction efficiencies (RSD %) of the model substances using different organic solvents as 493 the SLM for µLPME of acid drugs. 494 21 Table 2. µLPME calibration parameters, method detection limit (MLOD), method quantitation limit 495 (MLOQ), extraction efficiencies and enrichment factor for all analytes in a) stopped-flow conditions 496 mode after 20 min extraction and b) in double-flow conditions mode with an extraction time of 7 min. 497 Table 3. SF-µLPME/HPLC recoveries (average of three determinations ± standard deviation) from 498 non-diluted spiked urine samples. 499 Table 4. Comparison of figures of merit of µLPME with other analytical techniques for determination 500 of non-steroidal anti-inflammatories and parabens. 501 502 503 504 505 506 507 508 509 510 Tables 511 Table 1 512 Table 1. Extraction efficiencies (RSD %) of the model substances using different organic solvents as 513 the SLM for µLPME of acid drugs. 514 % Extraction efficiency (%RSD, n=3) NPOE Dihexylether 1-heptanol 1-octanol Salicylic acid 15 (2) 89 (1) 3 (1) 9 (1) Ethyl 4-hydroxybenzoate 82 (1) 100 (2) 77 (2) 85 (1) Propyl 4-hydroxybenzoate 99 (1) 99 (1) 79 (1) 88 (1) Ketoprofen 94 (4) 98 (2) 12 (3) 86 (3) Naproxen 81 (2) 93 (2) 17 (1) 58 (2) IsoButyl 4-hydroxybenzoate 100 (2) 100 (1) 82 (2) 94 (2) Butyl 4-hydroxybenzoate 70 (2) 98 (2) 81 (3) 97 (1) Diclofenac 54 (1) 88 (2) 10 (1) 33 (1) 22 Ibuprofen 70 (2) 100 (3) 5 (2) 22 (3) AvEEi 64 94 31 51 a Sample: 1 µL min-1 of HCl at pH 3 containing the nine drugs each at 1 µg mL-1; acceptor: 1 µL min-1 of NaOH at pH 515 11.75; extraction time: 7 min. NPOE: 2-Nitrophenyl octyl ether 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 Table 2. µLPME calibration parameters, method detection limit (MLOD), method quantitation limit 533 (MLOQ), extraction efficiencies and enrichment factor for all analytes in a) stopped-flow conditions 534 mode after 20 min extraction and b) in double-flow conditions mode with an extraction time of 7 min. 535 536 Stooped-flow conditions a Double-flow conditionsb,c Double-flow conditionsb,d MLOD (µg L-1) MLOQ (µg L-1) R2 EF EE* EF EE* EF EE* Salicylic acid 2.0 6.7 0.9997 47 81 9 44 - - - - - - - - - 98 (1) Ethyl 4-hydroxybenzoate 1.1 3.7 0.9995 42 73 15 74 100 (1) Propyl 4-hydroxybenzoate 0.7 2.3 0.9992 35 61 17 87 98 (1) Ketoprofen 2.9 9.7 0.9991 34 60 18 89 99 (1) Naproxen 1.8 6.0 0.9994 41 71 13 64 100 (1) IsoButyl 4-hydroxybenzoate 0.9 3.0 0.9990 21 44 14 70 100 (1) Butyl 4-hydroxybenzoate 1.5 5.0 0.9992 16 27 11 55 99 (1) Diclofenac 4.2 14.0 0.9989 19 34 15 76 92 (1) Ibuprofen 8.5 28.3 0.9991 35 61 19 94 99 (1) 23 *% Extraction efficiency (%RSD, n=4) 537 a Extraction time: 20 min and sample flow rate 20 µL min-1 538 b Extraction time: 7 min 539 c Acceptor flow rate of 1 µL min-1 and sample flow rate of 20 µL min-1 540 d Acceptor and sample flow rate of 1 µL min-1 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 Table 3. SF-µLPME/HPLC recoveries (average of three determinations ± standard deviation) from 568 non-diluted spiked urine samples. 569 Spiked level (µg L-1) SAC KTP NAX DIC IBU Urine 1 7 92.5 ± 0.6 N.Q 94.4 ± 0.9 N.Q N.D 18 94.2 ± 1.2 93.3 ± 0.8 92.4 ± 1.1 86.6 ± 0.7 89.1 ± 0.3* 50 95.5 ± 0.6 95.5 ± 0.8 98.2 ± 0.5 85.0 ± 0.5 90.2 ± 0.4 Urine 2 7 90.1 ± 0.5 N.Q 95.2 ± 0.4 N.Q N.D 18 92.4 ± 0.8 93.4 ± 0.9 93.3 ± 0.8 86.1 ± 0.7 87.2 ± 1.5* 50 95.9 ± 1.2 94.0 ± 1.0 100.0 ± 0.8 88.8 ± 0.4 91.7 ± 0.6 *Spiked concentration: 30 µg L-1 570 571 572 573 574 24 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 Table 4. Comparison of figures of merit of µLPME with other analytical techniques for determination 590 of non-steroidal anti-inflammatories and parabens. 591 Analytical method Analyte Matrix Sample Volume (mL) LOQ (µg L-1) EF EE % Extraction time (min) Multiextraction Reference HF(3)-LPME-HPLC/UV SAC, DIC, IBU Urine 50 41-180 70-900 - 15 No 4 HF(3)-LPME-MS/MS SAC, DIC, IBU Waste water 50 0.5-5 - 50-100 15 No 1 SPE-LC-MS/MS SAC, DIC, IBU Waste Water 500 0.1-3 - 70 > 30 No 21 DLLME-SFO-HPLC/UV KTP, DIC Urine 5 4-5 - 95-100 5 No 22 µLPME-HPLC/UV double-flow SAC, KTP, NAX, DIC, IBU Urine 0.007 100-500 - 75-100 5 Yes 41 HF-LPME-GCa MeP, EtP, PrP Water and urine 8 100-300 21-154 - 40 No 32 HF-LPME-b MeP, EtP, PrP, BuP, iPrP iBuP. BzP Water 3.5 0.5 3-16 24-60 30 No 3 25 EME-HPLC/UV EtP, PrP, BuP, iBuP. BzP Water 10 2.4-5 32-49  8 40 No 33 DF-µLPMEHPLC/UV double-flow EtP, PrP, BuP, iBuP Water 0.05 5-12 9-10 84-100 5 yes 42 DF-µLPMEHPLC/UV (1 µL min-1) SAC, KTP, NAX, DIC, IBU EtP, PrP, BuP, iBuP Urine 0.007 55-980 - 92-100 7 Yes This work DF-µLPMEHPLC/UV (20 µL min-1) SAC, KTP, NAX, DIC, IBU EtP, PrP, BuP, iBuP Urine 0.14 4.5-49 9-19 44-94 7 Yes This work SFµLPMEHPLC/UV SAC, KTP, NAX, DIC, IBU EtP, PrP, BuP, iBuP Urine 0.4 2.3-28 21-47 27-81 20 Yes This work 592 a Hollow fiber liquid phase microextraction2 phases 593 b Hollow fiber liquid phase microextraction3 phases 594 595 596 597