Microfluidic liquid-phase microextraction based on natural deep eutectic solvents immobilized in agarose membranes

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ “This is an Accepted Manuscript of an article published by Elsevier in: JOURNAL OF CHROMATOGRAPHY A on 2021, available at: https://doi.org/10.1016/j.chroma.2021.462580” 1 Microfluidic liquid-phase microextraction based on natural deep eutectic solvents immobilized in agarose membranes Samira Dowlatshaha,b, Mohammad Sarajib, Stig Pedersen-Bjergaardc,d , María Ramos-Payána,* a Department of Analytical Chemistry, Faculty of Chemistry, University of Seville, c/Prof. García González s/n, 41012, Seville, Spain b Department of Chemistry, Isfahan University of Technology, Isfahan 84156–83111, Iran c Department of Pharmacy, University of Oslo, P.O. Box 1068, Blindern, 0316 Oslo, Norway d Department of Pharmacy, Faculty of Health and Medical Sciences University of Copenhagen, Universitetesparken 2, 2100 Copenhagen Ø, Denmark *Corresponding author: [email protected] 2 Abstract In liquid-phase microextraction (LPME), the sample and the acceptor are separated by a synthetic organic solvent, which is immobilized in a porous polymeric membrane of polypropylene or polyvinylidene fluoride. The organic solvent serves as extraction phase, while the polymeric membrane serves as support membrane. The combination of extraction phase and support membrane is termed supported liquid membrane (SLM). In this paper, we developed for the first time fully green and biodegradable supported SLMs, based on natural deep eutectic solvents as extraction phase and agarose as support membrane. This highly green approach was developed and studied with sulfonamide pharmaceuticals as model analytes, and performance was compared with LPME using conventional SLMs. All experiments were conducted in a microfluidic device. Model analytes were extracted from acidic sample (pH 1.0) and into alkaline acceptor (pH 12.0). Both sample and acceptor were pumped at 1 μL min-1 into the microfluidic device, and the optimal SLM was based on 3 µL of coumarin and thymol (1:2 molar ratio) as the extraction phase. The proposed green microfluidic device was successfully applied for the determination of sulfonamides in urine samples with spiking recoveries in the range of 77-100%. LPME with deep eutectic solvent immobilized in agarose showed similar performance as with conventional SLMs. Thus, the data presented in this paper demonstrate that highly green microextraction systems may be developed in the future, based on natural solvents and biodegradable materials. Keywords: Sample preparation; liquid-phase microextraction; microfluidic device; natural deep eutectic solvents; agarose membrane 3 1. Introduction For several decades, liquid-liquid extractions have been downscaled to liquid-phase microextraction (LPME). Major incentives for this have been to reduce consumption of hazardous organic solvents and to pre-concentrate analytes into a micro liter volume of aqueous or organic phase, which can be injected and analyzed directly by liquid or gas chromatography. LPME has been further downscaled and implemented successfully in microfluidic systems [1–4]. In microfluidic LPME, analytes are extracted from a donor phase to an acceptor phase through a supported liquid membrane. The latter is a micro liter volume of organic solvent (extractant), immobilized by capillary forces in the pores of a polymeric membrane (support membrane). To date, polypropylene and polypropylene or polyvinylidene fluoride have been the common support membranes for LPME. Microfluidic LPME systems have significantly reduced the volume of extractant used as SLM [5-7], increasing mass transfer rate, accelerating extraction and reducing analysis times. Typical extractants are synthetic organic solvents, among others 1-octanol, dihexyl ether, and dodecylacetate. On the other hand, a new generation solvents such as supramolecular solvents (SUPRASs) [8,9], ionic solvents (ILs) [10,11] and deep eutectic solvents (DESs) [12,13] have caught the interest of many researchers for development of greener analytical chemistry in this decade. One of the most important aspects of developing green methods is either reducing the volume of organic solvent used in extraction, or employing less toxic organic solvents, such as DESs. In this sense, the microfluidic systems in which our group specializes, is undoubtedly an attractive alternative to achieve a significant reduction in the volumes of organic solvent. DESs are designed solvents made by mixing a proper portion of two (or more) solid components, one with strong hydrogen bond acidity (HBD) and another with strong hydrogen bond basicity (HBA). DESs are considered 4 green solvents, they are highly stable at room temperature, and have unique solvent properties. Therefore, more recently, they have been employed in different microextraction systems and more traditional liquid extraction systems [14] for extraction of among others polycyclic aromatic hydrocarbons in food and beverages [15], phenolic compunds from oils [16], and a variety of contaminants in environmental samples [17]. Recently DESs have also been explored as extractant in microfluidic LPME and in electromembrane extraction (EME) [18,19]. On the other hand, the use of biomaterials and bio-adsorbents has emerged recently in direction of greener analytical chemistry [20]. For example, agarose membranes are simple to modify as a selective extraction material and have allowed widespread applicability [21-26]. In accordance with the concept of green analytical chemistry, it is important to evolve biopolymers-based procedures integrated with miniaturized sample preparation methods to reduce the use and waste of hazardous chemicals and materials. Therefore, for the first time, a microfluidic LPME system based on natural deep eutectic solvent implemented on agarose gel membrane was developed. In this system, both the extractant and the support membrane were green and biodegradable materials, and the total waste of chemicals were less than 1 mg per sample. The green and biodegradable supported liquid membrane was compared with traditional supported liquid membranes, in order to verify performance of the former. 5 2. Experimental 2.1. Chemicals and materials All reagents and chemicals were of analytical grade or above. Sulfadiazine (SDI), sulfamerazine (SMR), sulfamethazine (SMZ), and sulfamethoxazole (SMX) were purchased from Fluka-Sigma- Aldrich (Madrid, Spain). Camphor (Ca), DL-menthol (Me), coumarin (Co), and thymol (Ty) were from Sigma-Aldrich (Madrid, Spain). Formic acid, sodium hydroxide, hydrochloric acid, 2- nitrophenyl octyl ether (NPOE), dihexyl ether (DHE), and 1-octanol were supplied by Fluka- Sigma-Aldrich (St. Louis, Mo, USA). Methanol, acetonitrile, nonanol, decanol, undecanol, and tributyl phosphate (TBP) were obtained from Merck (Darmstadt, Germany). The stock solutions of the sulfonamides were prepared in methanol (100 mg L−1) and preserved at 4°C in a refrigerator. Working solutions were daily provided by dilution of the stock solutions with deionized (DI) water. A Milli-Q Plus water purification system (Millipore, Billerica, MA, USA) was used to prepare all dilutions with ultrapure water. 2.2. Chromatographic conditions Analyses were performed using an Agilent 1100 series liquid chromatograph (Barcelona, Spain) consisting of a G1312A binary pump system, an UV/Vis diode array detector (DAD), and a G1313A auto sampler. A LiChroCART 75-4 Purosphere STAR RP-18e 3 µm (75 mm x 4.0 mm i.d.) (VWR, Spain) proceeded by a guard column Kromasil1 100 Å, C18, 5 µm (20 mm x 4.6 mm i.d.) (Scharlab S.L., Barcelona, Spain) was employed for chromatographic separations. Mobile phase A was 0.1% formic acid (pH 2.6) and mobile phase B was acetonitrile. The flow rate was 0.8 mL min-1. The mobile phase gradient was from 85% A to 70% A in 10 minutes, followed by 3 minutes equilibration time. The injection volume was 7 µL. The wavelength used for DAD was 6 254 nm for all analytes. The chromatogram was completed in 10 min, with retention times of 2.5, 3.3, 4.2, and 8.3 minutes for SDI, SMR, SMT and SMX, respectively. 2.3. Preparation of deep eutectic solvent and gel membrane To prepare a DES, a proper portion of two solid components (hydrogen bond donor (HBD) and hydrogen bond acceptor (HBA)) were weighted and adjusted to obtain a specific molar ratio. The two components were then transferred to a vial and capped tightly. After that, the mixture was heated in an oven at 80°C for 15 min. The agarose gel membrane was prepared according to previous work [24]. 0.40 g agarose was weighted, and added to 100 mL DI water. The solution was then shaked and boiled at 100°C for 30 min. A Petri dish (90 mm, i.d.) was filled with 8.0 mL of the warm solution. To cool and gel the solution, the mixture in the dish was kept at room temperature for at least 30 min. Then, the content was dried in the oven at 50°C for 24 h. A membrane of 0.01 thickness and a pore size between 15-20 nm was obtained. The formed gel membrane was cut into 30×0.5 mm pieces for further use. 2.4. Chip device fabrication and procedure A laser ablation cutter (Epilog Mini 24–30 W) was used for the fabrication of the microfluidic device which consisted of two poly(methyl methacrylate) (PMMA) plates assembled through four screws. Figure 1 shows a scheme of the proposed microfluidic device. Fabrication conditions were: writing speed of 40%, power of 24%, a resolution of 1500, and a frequency of 5000. Each layer contained a channel (14 mm length, 70 μm deep and 3 mm wide) and a flat gel membrane was placed covering the entire channel and separating them from each other. The membrane was 7 impregnated with 2 µL of Co:Ty (2:1) and subsequently closed using four screws. The device was reused, opened and closed as many times as necessary. Each channel had two holes: one for inlet solution and another for outlet solution. Both the sample (containing the model analytes) and the acceptor were connected to a micro-syringe pump (Cetoni GmbH, Korbussen, Germany), operating at 1 µL min-1 to introduce the solutions into the microfluidic device. After 5 min for SLM stabilization, the acceptor was collected during 10 minutes in a micro insert tube and then injected into the HPLC-UV system for analysis. 2.5. Calculations of extraction efficiency Extraction efficiency was calculated according to the following equation: 𝐸𝐸 (%) = 𝑛𝑎 𝑛𝑑𝑥100% = 𝑉𝑎 𝐶𝑎𝑓𝑖𝑛𝑎𝑙 𝑉𝑑𝐶𝑑𝑖𝑛𝑖𝑡𝑖𝑎𝑙 𝑥100% (Eq. 1) Here 𝑛𝑑 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 was the initial amount of analyte present in the sample and 𝑛𝑎 𝑓𝑖𝑛𝑎𝑙 was the final amount of analyte in acceptor solution. 𝑉𝑑 was the sample volume and 𝑉 𝑎 was the acceptor volume. Finally, 𝐶𝑎 𝑓𝑖𝑛𝑎𝑙 and 𝐶𝑑 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 were the initial and final concentration of analyte in the acceptor and sample, respectively. 2.6. Real samples Urine samples were collected from a 31-year-old healthy adult female volunteer (prior consent) and undiluted samples were tested by the microfluidic LPME device. The samples were adjusted to pH 1.0 with HCl solution, and spiked with sulfonamides to concentrations within the range 0.5- 3 µg mL-1). Samples were filtered through Pall NylafloTM nylon membrane filter 0.45 µm (Pall Corporation, Ann Arbor, Michigan, USA) and extracted with the microfluidic device. 8 3. Results and discussion 3.1. Agarose membrane composition This study compared support membranes of agarose (biodegradable) with traditional ones of polypropylene in a microfluidic LPME system. This system worked under dynamic conditions with flow in both the sample and acceptor channels, and therefore SLM stability and performance was investigated in details. The concentration of the agarose membrane determines the pore size and fragility of the network. Therefore, initial experiments were carried out to prepare and test support membranes of different agarose percentage (0.3, 0.4, 0.5 and 0.8 w/v %). The support membranes were prepared according to previous work [24], and extraction conditions were adopted from similar work with sulfonamide pharmaceuticals using support membranes of polypropylene (non-biodegradable) [27]: TBP as extraction phase, pH 4.0 in sample (containing model analytes at 2 µg mL−1), pH 12.0 as acceptor, and 1 µL min-1 as flow rate for both sample and acceptor. The different agarose membranes were impregnated with 2 μL of TBP and tested for LPME. As seen in Figure 2, the extraction efficiency increased when increasing the agarose percentage from 0.3 to 0.4% (w/v). A decrease in the extraction efficiency was observed above 0.4 % (w/v), probably due to reduction of the membrane porosity. Therefore, 0.4% (w/v) agarose membranes were used in following experiments. The superior benefits of these method is because of easy preparation of gel membrane enabled to get different shapes and thicknesses. Additionally, the methods require no ion-pairing reagents for extraction of polar analytes. 3.2. Supported liquid membrane (extraction phase and support membrane) 15 [9] M. Moradi, Y. Yamini, N. Feizi, Development and challenges of supramolecular solvents in liquid-based microextraction methods, TrAC - Trends Anal. Chem. 138 (2021) 116231. [10] Y. Tao, J.F. Liu, X.L. Hu, H.C. Li, T. Wang, G. Bin Jiang, Hollow fiber supported ionic liquid membrane microextraction for determination of sulfonamides in environmental water samples by high-performance liquid chromatography, J. Chromatogr. A. 1216 (2009) 6259– 6266. [11] D. Han, B. Tang, Y. Ri Lee, K. Ho Row, Application of ionic liquid in liquid phase microextraction technology, J. Sep. Sci. 35 (2012) 2949–2961. [12] M. Nazraz, Y. Yamini, A.M. Ramezani, Z. Dinmohammadpour, Deep eutectic solvent dependent carbon dioxide switching as a homogeneous extracting solvent in liquid-liquid microextraction, J. Chromatogr. A. 1636 (2021) 461756. [13] W. Tang, Y. An, K.H. Row, Emerging applications of (micro) extraction phase from hydrophilic to hydrophobic deep eutectic solvents: opportunities and trends, TrAC - Trends Anal. Chem. 136 (2021) 116187. [14] S.C. Cunha, J.O. Fernandes, Extraction techniques with deep eutectic solvents, TrAC Trends Anal. Chem. 105 (2018) 225–239. [15] W. Deng, A. Huang, Q. Zheng, L. Yu, X. Li, H. Hu, Y. Xiao, A density-tunable liquidphase microextraction system based on deep eutectic solvents for the determination of polycyclic aromatic hydrocarbons in tea, medicinal herbs and liquid foods, Food Chem. 352 (2021) 129331. [16] Tongnian Gu, Mingliang Zhang, Ting Tan, Jia Chen, Zhan Li, Qinghua Zhang, Hongdeng Qiu, Deep eutectic solvents as novel extraction media for phenolic compounds from model oil, Chem. Commun. 50 (2014) 11749–11752. [17] J. Lee, D. Jung, K. Park, Hydrophobic deep eutectic solvents for the extraction of organic and inorganic analytes from aqueous environments, TrAC Trends Anal. Chem. 118 (2019) 853–868. [18] E. Santigosa, S. Pedersen-Bjergaard, M. Muñoz, M. Ramos-Payán, Green microfluidic liquid-phase microextraction of polar and non-polar acids from urine, Anal. Bioanal. Chem. (2021) 1–7. [19] F.A. Hansen, E. Santigosa-Murillo, M. Ramos-Payán, M. Muñoz, E. Leere Øiestad, S. Pedersen-Bjergaard, Electromembrane extraction using deep eutectic solvents as the liquid membrane, Anal. Chim. Acta. 1143 (2021) 109–116. [20] H. Tabani, M. Alexovič, J. Sabo, M. Ramos Payán, An overview on the recent applications of agarose as a green biopolymer in micro-extraction-based sample preparation techniques, Talanta. 224 (2021) 121892. [21] A.Y.S. Lee, M.W. Lin, Polymeric IgA paraprotein on agarose gel electrophoresis immunofixation identifies a unique subset of IgA myeloma patients, Clin. Chim. Acta. 512 (2021) 112–116. [22] P. Hashemi, M.M. Abolghasemi, Preparation of a novel optical sensor for low pH values using agarose membranes as support, Sensors Actuators, B Chem. 115 (2006) 49–53. 16 [23] S.H. Loh, M.M. Sanagi, W.A. Wan Ibrahim, M.N. Hasan, Solvent-impregnated agarose gel liquid phase microextraction of polycyclic aromatic hydrocarbons in water, J. Chromatogr. A. 1302 (2013) 14–19. [24] M.M. Sanagi, S.H. Loh, W.A. Wan Ibrahim, M.N. Hasan, Agarose film liquid phase microextraction combined with gas chromatography-mass spectrometry for the determination of polycyclic aromatic hydrocarbons in water, J. Chromatogr. A. 1262 (2012) 43–48. [25] H. Tabani, A. Shokri, S. Tizro, S. Novajan, P. Varanusupakul, M. Alexovič, Evaluation of dispersive liquid–liquid microextraction by coupling with green-based agarose gelelectromembrane extraction: An efficient method to the tandem extraction of basic drugs from biological fluids, Talanta 199 (2019) 329–335. [26] H. Tabani, F. Dorabadi Zare, W. Alahmad, P. Varanusupakul, Determination of Cr(III) and Cr(VI) in water by dual-gel electromembrane extraction and a microfluidic paper-based device, Environ. Chem. Lett. 18 (2020) 187–196. [27] S. Dowlatshah, E. Santigosa, M. Saraji, M. Ramos-Payán, A selective and efficient microfluidic method-based liquid phase microextraction for the determination of sulfonamides in urine samples, J. Chromatogr. A. 1652 (2021) 462344. [28] M. Ramos-Payán, H. Jensen, N.J. Petersen, S.H. Hansen, S. Pedersen-Bjergaard, Liquidphase microextraction in a microfluidic-chip - High enrichment and sample clean-up from small sample volumes based on three-phase extraction, Anal. Chim. Acta. 735 (2012) 46– 53. Table and figure captions 17 Table 1. Extraction efficiencies (RSD %) of the analytes using different organic solvents as SLM on PP membranes. Table 2. Extraction efficiencies (RSD %) of the analytes using different organic solvents as SLM on gel agarose flat membranes. Table 3. Method detection limit (MLOD), method quantitation limit (MLOQ) and extraction efficiencies at optimal conditions. Table 4. Recoveries (average of three determinations ± standard deviation) from spiked urine samples. Fig. 1. Scheme of the microfluidic device. Fig. 2. Agarose concentration optimization. Fig. 3. Optimization of sample composition using PP membranes. SLM: Co:Ty 1:2, acceptor pH: 12, flow rate: 1 μL min-1 (acceptor and sample). Fig. 4. Optimization of sample composition using agarose membranes. SLM: Co:Ty 1:2, acceptor pH: 12, flow rate: 1 μL min-1 (acceptor and sample). Fig. 5. Optimization of acceptor composition using PP membranes. SLM: Co:Ty 1:2, sample pH: 1, flow rate: 1 μL min-1 (acceptor and sample). Fig. 6. Extraction efficiency versus sample flow rate. SLM: Co:Ty 1:2, sample pH: 1, acceptor pH: 12, acceptor flow rate: 1 μL min-1. Fig. 7.Chromatogram of a (I) blank human urine and (II) spiked human urine at 1 µg mL-1. 18 Table 1 Extraction efficiency % (RSD%, n=4) SDI SMR SMT SMZ SLM (Ca:Me 1:2) 17 (9) 33 (9) 36 (7) 30 (6) SLM (Ca:Ty 1:1) 33 (9) 35 (9) 33 (8) 38 (9) SLM (Ca:Ty 2:1) 13 (9) 15 (7) 18 (9) 28 (4) SLM (Co:Ty 1:1) 26 (9) 29 (9) 31 (9) 28 (8) SLM (Co:Ty 2:1) 13 (9) 17 (6) 16 (5) 14 (9) SLM (Co:Ty 1:2) 40 (3) 45 (5) 30 (6) 46 (5) SLM (Ca:Me 1:1) 17 (9) 30 (9) 31 (6) 42 (6) SLM (Ca:Ty 2:1) 21 (8) 22 (6) 22 (9) 18 (7) 19 Table 2. Extraction efficiency % (RSD%, n=4) SDI SMR SMZ SMX Decanol 20 (2) 18 (3) 21 (2) 20 (3) Octanol 21 (3) 22 (4) 21 (2) 24 (4) DHE 20 (6) 21 (7) 20 (6) 22 (6) NPOE 23 (5) 21 (6) 21 (7) 23 (8) TBP 22 (6) 22 (6) 24 (5) 22 (4) SLM (Ca:Me 1:2) 27 (3) 26 (5) 28 (4) 25 (6) SLM (Ca:Ty 1:1) 23 (5) 25 (7) 30 (6) 28 (5) SLM (Ca:Ty 2:1) 18 (7) 35 (3) 28 (6) 24 (4) SLM (Co:Ty 1:1) 25 (7) 19 (4) 31 (5) 28 (6) SLM (Co:Ty 2:1) 16 (7) 17 (6) 16 (5) 14 (4) SLM (Co:Ty 1:2) 27 (1) 42 (5) 30 (4) 28 (6) SLM (Ca:Me 1:1) 19 (7) 19 (3) 29 (6) 32 (6) SLM (Ca:Ty 2:1) 22 (5) 23 (6) 25 (4) 28 (7) 20 Table 3 MLODa (µg mL-1) MLOQa (µg mL-1) EE(%) a,* MLODb (µg mL-1) MLOQb (µg mL-1) EE(%)b, * SDI 0.063 0.21 63 (6) 0.070 0.23 46 (3) SMR 0.077 0.26 65 (4) 0.077 0.26 61 (5) SMT 0.072 0.24 64 (5) 0.083 0.28 47 (6) SMX 0.046 0.15 60 (5) 0.050 0.17 50 (3) *% Extraction efficiency (%RSD, n=4) in water a PP membranes b Agarose membranes 21 Table 4. Analyte Spiked level (µg/mL) Spiking recovery (%)a Spiking recovery (%)b RSD (%)a RSD (%)b SDI 0.5 1 3 100 97 93 96 95 102 7 8 6 5 7 6 SMR 0.5 1 3 96 89 88 77 78 73 7 5 6 6 7 8 SMZ 0.5 1 3 83 79 78 93 88 87 5 7 4 5 3 6 SMX 0.5 1 3 95 91 88 101 100 105 6 7 5 6 4 8 a PP membrane b agarose membrane 22 Fig. 1 Acceptor outlet Acceptor inlet Sample inlet Waste SLM Sample channel Acceptor channel Sample channel Acceptor channel Screws 23 Membrana liquida soportada Canal donador Canal aceptor Placa de metacrilato Placa de metacrilato 24 Fig. 2 Fig. 3 0 10 20 30 40 50 60 70 0,5 1 1,5 2 3 4 5 Extraction efficiency (%) Sample pH SDI SMA SMZ SMX