Electromembrane Extraction: A New Tool for the Analytical Workflow

The fundamental mechanism of electromembrane extraction (EME) involves the use of an electrical field to extract charged analytes from a sample solution through a liquid membrane into an acceptor solution (Figure 1). The liquid membrane (termed membrane in this article) comprises a few micro-liters of organic solvent immobilized within the pores of a porous support membrane. The membrane is hydrophobic and acts as a barrier that facilitates the selective extraction of ionic analytes. The selectivity is controlled by the chemical composition of the membrane, the direction and magnitude of the electrical field, and pH in the sample and acceptor. Due to the membrane and the electrical field, the selectivity is high and EME provides very clean extracts. This is particularly beneficial when dealing with complex matrices such as biological fluids,¹ environmental samples,² and food products.³ The minimal use of organic solvents in EME not only reduces costs but also lessens the environmental impact, adhering to the principles of green sample preparation.^4,5 This positions EME as a sustainable alternative to more conventional extraction techniques like liquid-liquid extraction (LLE) and solid-phase extraction (SPE).

EME's versatility is further highlighted by its broad scope of applications. It has been successfully used in the extraction and quantification of a variety of analytes, such as pharmaceuticals,⁶ endogenous metabolites,⁷,peptides,⁸ environmental pollutants,⁹ and various organic and inorganic ions.^{10, 11}Also, EME has been conducted in microfluidic devices,¹² polymeric tubes,¹³ vials,¹⁴ and 96-well plates.¹⁵ Its adaptability makes it suitable for a wide range of fields, including drug discovery, pharmaceutical development, forensic science, clinical chemistry, environmental science, and food safety. The technique's compatibility with downstream analytical techniques such as high-performance liquid chromatography (HPLC) and mass spectrometry (MS) may facilitate its integration into existing analytical workflows to enhance overall laboratory efficiency and throughput.

In 2020, we reported the first preliminary testing of prototype equipment for electromembrane extraction in the North American issue of this publication.¹⁶ The prototype setup was based on coupling the electrical field through sample and acceptor vials made of conductive polymer (Figure 1). By such, expensive electrodes of platinum were avoided, and replaced with technology suited for single use, automation, and high-throughput operation. Recently, the equipment was launched, and a new technique has been added to the commercial sample preparation toolbox.

The first commercial equipment for EME was launched in 2024 at Analytica in Germany.¹⁷ However, method development in EME is very different from current practice, and therefore a set of generic EME methods has been proposed recently.^18–20 In the current paper, we summarize these efforts, based on experimental data obtained with the new equipment. We present the generic methods, and discuss how they have been developed and optimized

Current

During EME, each sample (aqueous) is in contact with the membrane (oil) and acceptor (aqueous), and an electrical potential is applied across this aqueous-oil-aqueous system. The electrical field strength (V/cm) is low in the aqueous sample and acceptor but is strong in the membrane due to high electrical resistance. In the initial phase of extraction, the membrane is non-conductive and it acts as a capacitor; the membrane is charged (electrons), and a corresponding peak current is observed. However, the transfer of ions into the membrane is fast and increases the conductivity. After five to fifteen seconds in the conductive phase, the membrane starts serving as an electrical resistor. From this point, the current is controlled by the flux of ions across the membrane. These include analyte ions, sample matrix ions, and ions added to control pH in the sample and acceptor. Under optimal membrane conditions, the transfer of analyte is fast, while the transfer of other ions is close to zero. In such cases, current is generally low and decreases during EME. Under sub-optimal conditions, however, changes in the thickness or chemical composition of the membrane may occur and the current is affected. Therefore, the current profile for each individual sample is an important diagnostic tool. A typical current profile is shown in Figure 2. In this case, the current decreased as function of time, and this behavior is typical for a stable EME system operated under optimal membrane conditions. On the other hand, if the current is increasing, this may indicate non-optimal conditions. If the current is increasing gradually, the membrane is leaking or penetrated by water. A sudden increase in current indicates that the membrane has ruptured. Current profiles therefore are stored for documentation purposes.

Electrolysis

Because the sample and acceptor are exposed to an electrical field, electrolysis of water occurs according to the following reactions:

Anode (anodic vial):H₂O(l) →2H⁺(aq) + ½O₂(g) + 2e^- (1)

Cathode (cathodic vial):2H⁺(aq) +2e⁻ →H₂(g) (2)

Excessive electrolysis is problematic because small gas bubbles will be formed at different rates in the sample and acceptor. In a closed extraction system, a pressure difference may develop, and this may challenge the mechanical stability of the membrane. Furthermore, upon excessive electrolysis, pH will decrease in the anodic vial and increase in the cathodic vial, and this may affect the charge of the analytes and their extraction. Electrolysis increases with increasing current but is normally not an issue if the current is less than 50 µA for each sample. If the current is higher, the operator may reduce the extraction potential. Ohm’s law applies to the EME system, and current decreases linearly with decreasing extraction potential. However, if the extraction potential is reduced, the extraction time to reach maximum recovery increases. Alternatively, the operator may increase the hydrophobicity of the membrane (discussed later). In such cases, less sample matrix will transfer through the membrane, and the current will be reduced.

Due to electrolysis, pH will change in the sample and acceptor during EME. Drifting pH is circumvented by maintaining a low current and using solutions of buffer, acid, or base as sample diluent and acceptor, where pH should be at least two to three units below/above pK_a of the target base/acid.

System Stability

The EME setup is an aqueous-oil-aqueous system exposed to an external electrical field. For optimal performance, the system should be stable. The composition of the membrane plays a key role in this respect. Based on experience, organic solvents, modifiers, and additives more soluble in water than 0.5 mg/mL should be avoided, as they tend to leak into the sample and acceptor during extraction. In such cases, the membrane thickness changes with time, and acceptors may get contaminated with organic solvent. The stability of the membrane and the EME system is monitored in the current profile, and operators should be aware that:

• Current should preferably be below 50 µA for each sample to avoid excessive electrolysis
• Current should not increase during EME (except for the first few seconds of extraction)

Another requirement for a stable EME system is to use non-volatile solvents and substances as membrane constituents to avoid evaporation.

Charge and log P

Analytes may transfer through the membrane either as neutral or charged species. As neutral species, they transfer by passive diffusion and are extracted by three-phase liquid-phase microextraction (LPME). ²¹ The analytes are unaffected by the electrical field. As charged species, they transfer mainly by electro-kinetic migration and are extracted by EME. EME is faster than liquid phase microectraction ( LPME) and provides higher recoveries for polar analytes than LPME. For efficient extraction of basic analytes, EME should be performed with pH in the sample and acceptor at least two to three pH units below their pK_a-values. If pH is closer to pK_a, extraction is a mix of EME and LPME, and with pH>pK_a extraction is LPME. With EME, basic analytes are mainly extracted in charge state z=+1 or z=+2, and acids are extracted with z=-1 or z=-2. For acids, pH should be at least two to three units above pK_a.

In addition to charge, mass transfer in EME depends on analyte polarity. This is illustrated in Figure 3a, where 90 different basic pharmaceuticals were extracted with 2-nitrophenyl octyl ether (NPOE) as membrane. The polarity of the analytes was in the range -4.2<log P<8.1, and the analytes were plotted according to recovery and log P in Figure 3a. EME provided highly efficient extraction of analytes in the polarity range 2<log P<6. For polar analytes with log P<2, extraction was suppressed due to poor transfer into the membrane (2-nitrophenyl octyl ether,[NPOE]), while highly non-polar analytes with log P>6 were prone to accumulation in the membrane.¹⁹ For NPOE, the log P-range from 2 to 6 was defined as the extraction window. Replacing this liquid with one of somewhat higher polarity, shifted the extraction window towards lower log P.¹⁹

Solvation

Analytes in their ionized form have very low partition into the organic membrane solvent, unless the electrical field is applied, and the analytes are solvated by the membrane. Multiple types of inter-molecular phenomena are involved in this process, but at least one of the following strong inter-molecular interactions (principal interactions) are required:

• Hydrogen bond interaction
• Cation-π interaction
• Ionic interaction

With membranes of NPOE, protonated bases are prone to hydrogen bond and cation-π interactions with the membrane solvent. NPOE serves as hydrogen bond acceptor and π-electron donor, while the protonated analytes are hydrogen bond donors and π-electron acceptors. Because two different principal interactions are involved, NPOE is an excellent membrane for both mono- and dibasic analytes (z=+1 and z=+2, respectively). However, due to the hydrophobic nature of NPOE (log P=4.9), bases with log P<2 are too polar in their protonated form to enter the membrane. Bases with log P>6 are poorly solvated in the aqueous acceptor even in their protonated form. Followingly, NPOE can be used in the polarity range 2<log P<6, and this is defined as the extraction window for this particular solvent ¹⁹.

Replacing NPOE with 2-undecanone, the extraction window is changed to 1<log P<4.5 (Figure 3b). 2-Undecanone (log P=3.9) is more polar than NPOE, and therefore the extraction window is shifted towards lower log P-values. 2-Undecanone is non-aromatic, and in terms of principal interactions, this solvent is limited to hydrogen bonding. Therefore, 2-undecanone is suited for mono-basic analytes only, and appears to be inefficient for dibasic analytes.¹⁹

Extraction of analytes with log P<0 is difficult based on hydrogen bond and cation-π interactions only; in such cases an ionic carrier must be added to the membrane solvent (Figure 3c). For basic analytes, di(2-ethylhexyl) phosphate (DEHP) is often used as ionic carrier. Because ionic interactions are strong, even analytes with log P<0 are transferred across the membrane.¹⁸

Generic Methods

Based on the discovery that mass transfer in EME largely can be predicted from analyte charge and polarity, a set of generic methods has been developed recently (Figure 4 and Table 1).^18,,19,,22 All membrane components were selected carefully to provide stable extraction systems. For non-polar bases, generic method B1 is recommended. In B1, NPOE is used as liquid membrane. Accordingly, B1 is intended for mono- and dibasic analytes in the polarity range 2<log P<6. Dilute formic acid is used as sample diluent and acceptor. The acceptor is directly injectable in liquid chromatogrpahy (LC) or liquid chromatography-mass spectrometry (LC–MS). Thus, after EME, there is no solvent evaporation and reconstitution, as done with most of current sample preparation methods. The recommended extraction potential with B1 is 100 V, and the recommended extraction time is 30 min. The two latter parameters are compound dependent and may be further optimized. The recommended agitation rate is somewhat dependent on the sample and acceptor volume, and typically it is set to 1000 rpm. Above this level, repeatability tends to decrease.

Generic method B2 is intended for monobases in the polarity range 1<log P<4.5. B2 is based on 2-undecanone as membrane solvent.¹⁹ Because 2-undecanone is more polar than NPOE, this membrane should be operated at lower voltage (50 V) to avoid excessive current from sample matrix and ions added to control pH. All other parameters with B2 are the same as with B1.

Generic method B3 is intended for mono- and dibases in the polarity range -2<log P<4.¹⁸ The membrane used in B3 is a mixture of 6-methyl-coumarin, thymol, 2-undecanone, and DEHP. 6-Methyl-coumarin and thymol are both solids. They are mixed in 1:2 w/w ratio and heated to form a hydrophobic deep eutectic solvent (DES). The DES is immiscible with water and provides strong hydrogen bond acceptor and donor properties. In addition, the DES is aromatic and serves as π-electron donor for cation-π interactions. DEHP (1% w/w) is added as ionic carrier to the membrane, and by such, analytes are solvated based on a mix of ionic, hydrogen bond, and cation-π interactions. This is the reason why B3 extracts bases with high polarity. To avoid excessive electrolysis, the membrane composition is diluted with 2-undecanone. The latter component decreases current, without affecting the extraction window.

For acidic analytes, currently two different generic methods have been proposed (A1 and A2, Table 1). ²² A1 is for acids in the polarity range 2<log P<6. The membrane is a mixture of dodecyl methyl sulfoxide and thymol. The latter is a strong hydrogen bond donor, and solvation is achieved by hydrogen bonding as the principal interactions. A2 is based on a deep eutectic solvent (DES) with 6-methyl-coumarin and thymol, diluted with NPOE to reduce the current. For the same reason, A2 is operated at lower voltage than A1. Due to the use of DES, A2 extracts more polar acids than A1, and the extraction window of A2 is set to 0.5<log P<6.

Currently, we have not finalized a generic membrane for extraction of acids with log P<0. Polar acids represent a challenge, and we anticipate two reasons for this; polar acids have low affinity for organic solvents immiscible with water, and the negative charge is often delocalized due to resonance. However, we are systematically investigating different membrane systems, and several candidate membranes are in progress.

The generic methods will be subject to continuous development, and updated information will be available.^2.3

Performance

Recently, several applications have been published using the prototype device installed in a clinical pharmacology routine laboratory at the University Hospital of Trondheim (Norway), and tested in combination with LC–tandem mass spectrometry (MS/MS) for detection of antidepressive pharmaceuticals, opioids, and amphetamine in patient serum, urine, and saliva samples.^24-26 The EME method for antidepressive pharmaceuticals used NPOE as membrane solvent and was almost identical with generic method B1. Validation data for duloxetine, amitriptyline, clomipramine, and doxepin are presented in Table 2. Recoveries ranged from 78% to 95% after 15 min extraction. Recoveries were even higher after 30 min, but 15 min was set as the final extraction time to keep the time consumption as low as possible. The validation data (linearity, precision, accuracy) were all high quality and in compliance with the requirements of EMA and FDA guidelines for bioanalytical validation. Due to efficient cleanup, no matrix effects were observed during LC–MS/MS analysis. Especially, phospholipids were not detected in the acceptor after EME. This is a great advantage, because phospholipids are known to cause ion suppression in LC–MS/MS.²⁷

Table 3 compares the EME method with the reference method of the laboratory based on protein precipitation and phospholipid removal. The consumption of organic solvent was much less with EME, and no nitrogen gas was required. Consumables and time were less with EME, and the workflow was simplified as compared to the reference method.

Method Development

For scientists working with EME for the first time, we recommend starting method development by testing performance using the generic methods. For basic analytes, the generic methods B1, B2, and B3 should all be tested, regardless of log P of the analyte. The reason for this is that the respective extraction windows are overlapping. Similar for acids, A1 and A2 should be tested first. If performance is not sufficient, then extraction time may be optimized, followed by optimization of pH in the sample and acceptor.

We are currently working with generic methods for polar bases (log P<-3) and polar acids (log P<-1). Extraction recoveries for highly polar analytes are not related to charge and polarity only but depend also on other molecular properties. Therefore, we expect two or three different generic methods will be required to cover highly polar bases, and a similar number to cover polar acids. We expect all the methods will involve solvation based on ionic interactions. Until this research has been completed, scientists need to consult the literature for extraction conditions applicable to highly polar substances.

In our testing of new membrane compositions, we always make sure membrane components are less soluble in water than 0.5 mg/mL, non-volatile, commercially available, inexpensive, and non-toxic. For EME of acids, we primarily test membranes with hydrogen bond donor properties, and for bases we are looking for membranes with hydrogen bond acceptor properties. We also test addition of ionic carries, which are lipophilic with either an acid or base functional group. Ionic carriers increase the conductivity of the membrane, and the concentration of ionic carrier should be low (<5%) to avoid excessive current during EME. We always operate systems below 50 µA, and current should not increase with time.

If samples are without buffer capacity, or if sample pH needs to be changed, we add a sample diluent to the sample. pH in the sample and acceptor should be controlled with buffer, acid, or base, with sufficient capacity to maintain stable pH conditions during extraction. For pure EME, pH should be no less than two units below pK_a for bases, and no less than two units above pK_a for acids.

The work with generic methods discussed in this section have been performed with many pharmaceuticals up to 650 Da size. We are also working with EME of peptides. Due to size and polarity, peptides are not extracted with generic method B1, but liquid membranes based on deep eutectic solvents are promising.²⁸ At the moment, we have been able to extract peptides up to 1600 Da. We are currently performing experiments to extend this size limitation, testing new liquid membrane compositions and additives to influence folding and surface charge of the biomolecules.

Conclusions

With electromembrane extraction (EME), exhaustive microextraction and green sample preparation can be completed in 15 to 30 min with minimal use of organic solvents, chemicals, and consumables. The acceptor is aqueous and can be injected directly into LC–MS/MS after extraction. Thus, solvent evaporation and sample reconstitution is eliminated. Due to efficient cleanup, no matrix effects are observed during LC–MS/MS analysis. With plasma and serum samples as an example, proteins and phospholipids are not detected in the acceptor after EME. This is a great advantage, because proteins may clog the LC column, and phospholipids are known to cause ion suppression in LC–MS/MS.

For the first time, commercial equipment is now available for EME. While current research and development of the technique have been conducted with laboratory-built equipment, EME can now be conducted using equipment of industrial standard, and experimental data can be easily reproduced.

EME combines the principles of partition and electrophoresis, and methods and method development in EME is very different from traditional sample preparation. Therefore, a set of generic methods has been developed, currently for basic analytes with log P>-2 and for acidic analytes with log P>0.5. Work is in progress with corresponding methods for highly polar substances, to complete the array of methods ready to use.

EME has been developed by many scientists for 20 years, and more than 500 scientific articles have been published. Hopefully, laboratories will implement EME and take advantage of this technology. For routine and contract laboratories, motivation for this may be to make sample preparation greener, or to simplify the experimental workflow. For academic institutions, motivation may be to make fundamental research on EME and to advance the principle. Implementation in microfluidic devices, extraction of large biomolecules, or playing with the selectivity are just a few examples in this direction. We expect much more innovation in the framework of EME in the years to come.

References

1. Gao, T. F.; Dong, Z. Z.; Shen, X. T. et al. Advances in Electromembrane Extraction of Drugs-of-Abuse from Biological Samples. Journal of Instrum. Anal. 2025, 44 (1), 25-33, DOI: 10.12452/j.fxcsxb.241024482

2. Shi, L.; Chen, M.; Zhao, G. et al. Environmental Applications of Electromembrane Extraction: A Review. Membranes 2023, 13 (8). DOI: 10.3390/membranes13080705

3. Khan, W. A.; Varanusupakul, P.; Arain, M. B. et al. Modified and Hyphenated Electromembrane Extraction Techniques and their Applications for Food Analysis. TrAC - Trends Anal. Chem. 2024, 178. DOI: 10.1016/j.trac.2024.117861

4. López-Lorente, Á. I.; Pena-Pereira, F.; Pedersen-Bjergaard, S. et al. The Ten Principles of Green Sample Preparation. TrAC - Trends Anal. Chem. 2022, 148. DOI: 10.1016/j.trac.2022.116530

5. Nowak, P. M.; Wietecha-Posłuszny, R.; Pawliszyn, J. White Analytical Chemistry: An approach to reconcile the principles of Green Analytical Chemistry and functionality. TrAC - Trends Anal. Chem. 2021, 138. DOI: 10.1016/j.trac.2021.116223

6. Luo, Y.; Zheng, M.; Huang, C. et al. Electromembrane Extraction of Pharmaceutical Analytes: Systematic Review and Meta-Analysis. Sep. Purif. Technol. 2024, 338. DOI: 10.1016/j.seppur.2024.126587

7. Drouin, N.; Rudaz, S.; Schappler, J. New Supported Liquid Membrane for Electromembrane Extraction of Polar Basic Endogenous Metabolites. J. Pharm. Biomed. Anal. 2018, 159, 53-59. DOI: 10.1016/j.jpba.2018.06.029

8. Pourahadi, A.; Nojavan, S.; Hosseiny Davarani, S. S. Gel-Electromembrane Extraction of Peptides: Determination of Five Hypothalamic Agents in Human Plasma Samples. Talanta 2020, 217. DOI: 10.1016/j.talanta.2020.121025

9. Safari, M.; Nojavan, S.; Davarani, S. S. H. et al. Speciation of Chromium in Environmental Samples by Dual Electromembrane Extraction System Followed by High Performance Liquid Chromatography. Anal. Chim. Acta 2013, 789, 58-64. DOI: 10.1016/j.aca.2013.06.023

10. Martins, R. O.; de Araújo, G. L.; Simas, R. C. et al. Electromembrane Extraction (EME): Fundamentals and Applications. Talanta Open 2023, 7. DOI: 10.1016/j.talo.2023.100200

11. Shang, Q.; Mei, H.; Huang, C. et al Fundamentals, Operations and Applications of Electromembrane Extraction: An Overview of Reviews. Microchem. J. 2022, 181. DOI: 10.1016/j.microc.2022.107751

12. Javadian, G.; Yamini, Y.; Zarghampour, F. Microfluidic Electromembrane Extraction Followed by Fluorescence Detection Based on Digital Image Analysis of Solid-State Fluorophore: Easy and Sensitive Analytical Procedure. Microchem. J. 2024, 205. DOI: 10.1016/j.microc.2024.111230

13. Šlampová, A.; Kubáň, P. Micro-Electromembrane Extraction Through Volatile Free Liquid Membrane for the Determination of β-lactam Antibiotics in Biological and Environmental Samples. Talanta 2023, 252. DOI: 10.1016/j.talanta.2022.123831

14. Pedersen-Bjergaard, S.; Rasmussen, K. E. Electrokinetic Migration Across Artificial Liquid Membranes: New Concept for Rapid Sample Preparation of Biological Fluids. J. Pharm. Biomed. Anal. 2006, 1109 (2), 183-190. DOI: 10.1016/j.chroma.2006.01.025

15. Drouin, N.; Mandscheff, J. F.; Rudaz, S. et al. Development of a New Extraction Device Based on Parallel-Electromembrane Extraction. Anal. Chem. 2017, 89 (12), 6346-6350. DOI: 10.1021/acs.analchem.7b01284

16. Drobnjak, M.; Hansen, F.; Øiestad, E. L. et al. Electromembrane Extraction with Vials of Conducting Polymer. LCGC North Am. 2020, 38 (8), 435-439.

17. Hroncich, C. 12 New Laboratory Products Showcased at Analytica 2024. Chromatography Online website2024. https://www.chromatographyonline.com/view/new-laboratory-products-analytica-2024 (accessed 2025 November 11).

18. Zhou, C.; Dowlatshah, S.; Hansen, F. A. et al. Generic Conditions for Electromembrane Extraction of Polar Bases. Talanta 2023, 125215. DOI: 10.1016/j.talanta.2023.125215

19. Zhou, C.; Dowlatshah, S.; Hay, A. O. et al. Generic Liquid Membranes for Electromembrane Extraction of Bases with Low or Moderate Hydrophilicity. Anal. Chem. 2023, 95 (23), 8982-8989. DOI: 10.1021/acs.analchem.3c01052

20. Song, C.; Zhou, C.; Hansen, F. A. et al. Optimization of Generic Conditions for Electromembrane Extraction of Basic Substances of Moderate or Low Polarity. J. Sep. Sci. 2024, 47 (3). DOI: 10.1002/jssc.202300801

21. Pedersen-Bjergaard, S.; Rasmussen, K. E. Liquid-Liquid-Liquid Microextraction for Sample Preparation of Biological Fluids Prior to Capillary Electrophoresis. Anal. Chem. 1999, 71 (14), 2650-2656, DOI: 10.1021/ac990055n

22. Song, C.; Dowlatshah, S.; Gaznawi, S. et al. New Robust and Efficient Liquid Membranes for Conductive Vial Electromembrane Extraction of Acids with Low to Moderate Hydrophilicity in Human Plasma. Anal. Bioanal. Chem 2024. DOI: 10.1007/s00216-024-05503-6

23. Extraction Technologies Norway. Is EME Suitable for My Analytes? 2025. https://how-to-eme.com/ (accessed 2025-11-11).

24. Gottenberg Skaalvik, T.; Leere Øiestad, E.; Trones, R. et al. Determination of Psychoactive Drugs in Serum Using Conductive Vial Electromembrane Extraction Combined with UHPLC-MS/MS. J. Chromatogr. B 2021, 1183, 122926. DOI: 10.1016/j.jchromb.2021.122926

25. Skaalvik, T. G.; Zhou, C.; Øiestad, E. L. et al. Conductive Vial Electromembrane Extraction of Opioids from Oral Fluid. Anal. Bioanal. Chem 2023. DOI: 10.1007/s00216-023-04807-3

26. Skaalvik, T. G.; Øiestad, E. L.; Pedersen-Bjergaard, S. et al. Determination of Amphetamine Enantiomers in Urine by Conductive Vial Electromembrane Extraction and Ultra-High Performance Supercritical Fluid Chromatography Tandem Mass Spectrometry. Drug Test Anal 2023, 15 (8), 909-918. DOI: 10.1002/dta.3487

27. Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Strategies for the Assessment of Matrix Effect in Quantitative Bioanalytical Methods Based on HPLC−MS/MS. Anal. Chem. 2003, 75 (13), 3019-3030. DOI: 10.1021/ac020361s

28. Rye, T. K.; Martinovic, G.; Eie, L. V. et al. Electromembrane Extraction of Peptides using Deep Eutectic Solvents as Liquid Membrane. Anal. Chim. Acta 2021, 1175, 338717. DOI: 10.1016/j.aca.2021.338717