2. Synthesis and Properties of Deep Eutectic Solvents
2.1. Synthesis of Deep Eutectic Solvents
One of the approaches used to synthesize DESs is by mixing and heating two or more salts to form a homogenous solution. Typically, ambient temperature molten salts have been formed by mixing quaternary ammonium salts, a hydrogen bond acceptor (HBA) with metal salts
[32][33]. Hence, four main different types of DES occur: I. quaternary salt with a metal chloride, II. quaternary salt with a hydrated metal chloride, III. quaternary salt (HBA) with an HBD compound and IV. metal chloride with an HBD. The respective general formulas of the four types are shown at
Table 1. The first two types are used to synthesize hydrophilic DESs, whilst the other two for hydrophobic DESs
[20][34]. Due to the stability in the aquatic solutions which type III and IV provide, they are often utilized in many papers, with choline chloride (ChCl) being the most used HBA. In
Figure 1, some common HBA and HBD are shown.
Figure 1. Structures of some halide salts (HBAs) and hydrogen bond donors (HBDs) used in the formation of deep eutectic solvents.
Table 1. The four types of deep eutectic solvents.
Cat+ = any phosphonium, ammonium or sulfonium cation, X− = a Lewis base, generally a halide anion, MClx = metal chloride, RZ = organic compound.
2.2. Properties of Deep Eutectic Solvents
Various interactions (such as anion exchange, weak non-covalent interactions, π-π and/or hydrogen bonding) take place, amongst an HBD and an HBA in various combinations and molar ratios that contribute to some of the physicochemical properties discussed below
[35]. Those properties are viscosity, density, conductivity, acidity, surface tension, volatility, and melting or freezing point. There is also an intensive study explaining the effect of hydrogen bonding proton transfer mechanism on the physical and chemical properties
[36]. Other properties taken into account before selecting the optimal DES are biodegradability, toxicity, and thermal stability
[37]. Moreover, both ILs and DESs can exhibit different physicochemical properties as binary mixtures, depending on the selected co-solvent, such as water or organic solvents. The interactions that may occur from the solute–solvent combination are defined as solvatochromic properties, which can further aid in the comprehension of chemical reactions through preferential solvation (PS)
[38]. A useful tool to investigate the PS parameters and interactions, like hydrogen bond between molecules and ions, is the molecular dynamics (MD) simulation, which can be used as completion and confirmation of experimental results. For instance, Aryafard et. al. conducted an extensive investigation for three different DESs and their binary mixtures with polyethylene glycol (PEG 400) as co-solvent, through MD simulations analysis, while showing that PEG 400 can make a strong hydrogen bond with DESs
[39].
3. Microextraction Techniques Using DES
3.1. Solid Phase Microextraction
3.1.1. Solid Phase Extraction
Solid phase extraction (SPE) is an efficient technique commonly used to handle aqueous samples and extract semi-volatile or non-volatile analytes, due to it being easy, inexpensive, and possible to be automated. SPE was firstly introduced in the 1950s
[40], and it uses appropriate sorbents so as to isolate the target compounds from the sample. Currently, there is a wide variety of available sorbents to cover almost all the possible interactions with the compounds. Nevertheless, analytes and interferents could coelute from the SPE sorbent because of the limited selectivity of conventional solid sorbents, such as silica-based, carbon-based, and clay-based resins
[41][42]. Furthermore, due to the consumption of significant quantities of sample, organic solvents, and sorbents, the SPE technique is still improving
[41].
There are many modifications to the conventional SPE technique, including magnetic SPE (MSPE), pipette-tip SPE (PT-SPE) and dispersive SPE (dSPE or DSPE)
[43]. In MSPE, magnetic nanoparticles are utilized, such as magnetite (Fe
3O
4) and maghemite (γ-Fe
2O
3), and afterwards, they are embedded in a suitable coating of inorganic substances, mainly in graphene
[44]. PT-SPE is a miniaturized form of SPE that resembles the procedure of conventional SPE in terms of conditioning, sample loading, washing, and elution
[45]. DSPE is based on the dispersion of a solid sorbent in liquid samples for the extraction, isolation, and cleaning of various target compounds from complex matrices
[46].
Jeong et al. presented the idea of tailoring and recycling of deep eutectic solvents as a sustainable and efficient extraction media to maximize the extraction efficiencies of several compounds found in ginseng. While experimenting with different solvents and ratios, they found that the combination of three effective DES components of sucrose, L-proline, and glycerol at a molar ratio of 1:4:9 provided the best results. The recovery of this study was determined to be 102.6 ± 4.1% (
n = 5), and the recycled solvents’ extraction efficiencies were 91.9 (±2.9%), 85.4 (±2.3%), and 82.6 (±4.7%) for every recurring recycle, indicating that the DES can be recycled and reused at least three times, so as to achieve a high level of extraction yields
[47].
N. Fu et al. experimented with environmentally friendly and non-polluting solvents, in order to extract polyphenols (such as protocatechuic acid, catechins, caffeic acid and epicatechin) from palm samples by using different ChCl DES. They tried eight different compounds with the ChCl to optimize the conditions of enrichment and purification of SPE, so as to acquire high yields. The optimal DES for the enrichment step was ChCl-FA, with a molar ratio of 1:1, and for the purification step, they utilized a ChCl-Ph DES-modified absorbent for the SPE cartridges and ChCl-Urea-H
2O (1:1:5), ChCl-Gly (1:1), ChCl-FA-H
2O (1:1:5), and H
2O as eluents. The recoveries of the four analytes were 100% on the first cycle of usage of the DES, and they maintained a percentage between 60–117% throughout the five cycles studied in this work
[48].
G. Li et al. tested DES modified molecular imprinted polymers (DES-MIPs) and DES modified non-imprinted polymers (DES-NIPs) for the purification of chlorogenic acid (CA) from honeysuckle. The preparation of these absorbents is explained in their paper. The DES used in this work had a ChCl base and glycerol, with a molar ratio of 1:2. The recoveries of MIPs, NIPs, DES-MIPs, and DES-NIPs were determined by the comparison of the peak areas of CA before and after the usage of SPE and were found to be 69.34%, 60.08%, 72.56%, and 64.79%, respectively
[49].
Wang et al. utilized graphene (G) and graphene oxide (GO) whilst modifying them with DES and supporting them on silica as absorbents for SPE, for the preconcetration and extraction of three chlorophenols (4-chlorophenol, 2,4-dichlorophenol, and 2,4,6-trichlorophenol) from environmental water samples. The DESs used in this work were ChCl with FA, acetic acid, propionic acid, urea, Gly, ethylene glycol, and 1,4-butanodiol at the molar ratios of 1:2
[50].
In another study, G. Li et al. worked with Fe
3O
4/MIPs which were modified by ChCl based DES for the swift purification of alkaloid isomers (theophylline and theobromine) from green tea samples. MSPE was used as the purification method of the two compounds. The washing solution used was a mixture of methanol and water (80:20
v/v), and the eluent consisted of methanol/acetic acid (80:20
v/v) at pH 3 and with a volume of 4 mL. The recoveries of theophylline and theobromine in green tea were 87.51% and 92.27%, respectively. The optimal DES for this study was ChCl-Urea (1:2)
[51].
X. Li et al. applied ternary DES in MIPs for the purification of levofloxacin, a known quinolone. The mixture of DES contained Caffeic Acid, ChCl, and FA at different molar ratios and were prepared by the heating method. The DES that exhibited the highest recovery rate (91.4%) was the one with a molar ratio of 1:3:1.5
[52].
Chen et al. utilized ternary hydrosulphonyl-based DES (THS-DES) modified by magnetic GO for the removal of mercury (Hg
2+) from water, by using MSPE. The THS-DES was consisted of ChCl-itaconic acid-3-mercaptopropionic acid (2:1:1), and the removal efficiency achieved was 99.91% under optimized conditions. The THS-DES@M-GO absorbent retained a high removal efficiency of 90.23% even after seven MSPE cycles
[53].
Ma et al. synthesized two types of MIPs based on magnetic chitosan with different DES (Fe
3O
4-CTS@DES-MIPs) and applied them as adsorbents in MSPE for the selective separation of catechins ((+)-catechin, (−)-epicatechin, and (−)-epigallocatechin gallate) in black tea samples. The two DESs were formed by ChCl/methacrylic acid (1:2 molar ratio) and betaine/methacrylic acid/H
2O (1:2:1). The DES which gave the highest recovery rates was the first one, showing outstanding recognition, selectivity, specificity and magnetism. The recovery rates of the three catechins were 95.4%, 92.4%, and 90.3%, respectively
[54].
Another study by Ma et al. that utilized MSPE was about extracting catechins from green tea, by using DES-MIP as a functional monomer and magnetic molybdenum disulfide as a base (Fe
3O
4@MoS
2@DES-MIP). The catechins examined were the three mentioned above plus (−)-epigallocatechin and (−)-epicatechin gallate. The DES consisted of vinyl pyrrolidone and malonic acid (molar ratio 1:1). Both intra-day and inter-day yielded recoveries of the five analytes between 80–98%
[55].
Liu et al. used DES modified graphene for the determination of sulfamerazine in river water samples, with the PT-SPE as the extraction technique. In particular, they tested the use of ChCl as a base and ethylene glycol, Gly, or urea as an HBD (at a 1:2 molar ratio with every solvent combination). Afterwards, they synthesized the DES-graphene and placed it in a 100 μL pipette tip using degrease cotton at both ends as frits. A 1 mL PT was inserted into the 100 μL PT, and the adsorbent was activated with 1 mL methanol and 1 mL of water. The extraction recovery by DES-G was 95.38%, whilst the G presented a 26.98% recovery rate
[56].
Sun et al. utilized DES modified by a heteroatom doped GO for the application of PT-SPE and the extraction of two flavonoids, myricetin and rutin. Two different types of GO were prepared, one doped with nitrogen (N-GO), and the other with boron (B-GO). Subsequently, they prepared the N-GO-DES and B-GO-DES, with the DES comprising of ChCl-ethylene glycol. Finally, N-GO-DES was selected as the optimum adsorbent for this method, due to its permeability and high recoveries (that of myricetin was in the range of 70.61–99.77%, and rutin in between 76.58–98.14%)
[57].
Yousefi et al. utilized dSPE for the ultra-trace analysis of organochlorine pesticides by using hydrophilic DES magnetic bucky gels (MBG) and magnetic multi-walled carbon nanotube nanocomposites (MMWCNT), prior to GC-μECD. The DES acted as a carrier and, at same time, as a disperser of the MMWCNTs and MBG. Thus, they opted for ChCl:Urea, 1:2 molar ratio, which had proper stearical interactions with MMWCNT, and it was miscible with water, so as to improve the dispersion of the sorbent. The LOD of this method was found to be 0.04–0.27 ng/L, and the enhancement factors were between 270–340, which are better than other techniques cited in the study
[58].
Lamei et al. also used dSPE with magnetic GO nanoparticles coated with DES and ultrasound for the preconcentration of methadone in water and biological samples followed by GC-MS and GC-FID. The developed magnetic GO-DES (Fe
3O
4@GO-DES) was dispersed by ultrasound in the mixture of deionized water, NaOH and the sample solution that contained methadone to extract and preconcentrate the analyte. The DES studied in this work were ChCl with TNO (5,5,8,8-tetramethyl-5,6,7,8-tetrahydronaphthalen-2-ol), glycerol, and phenol at different molar ratios. The most efficient of the three was TNO, at molar ratio of 1:2 (ChCl:TNO), due to the high recoveries it provided. The LOD by using this method with GC-FID was 0.8 μg/L, whilst with GC-MS it was 0.03 μg/L. The recoveries of methadone in different samples were as follow: in distilled water, 98.4–101.5%; in urine, 95.3–99.3%; and in plasma, 95.1–96.4%
[59].
Zarei et al. applied DES based magnetic colloidal gel for dSPE of ultra-trace amounts of some nitroaromatic compounds in water samples, followed by GC-μECD. The magnetic colloidal gel consisted of MMWCNTs and a deep eutectic solvent (ChCl:urea, ChCl:phenol, ChCl:acetic acid, or ChCl:gly at a molar ratio of 1:2). The optimal DES that provided the highest extraction efficiency was ChCl:urea. Then, the mixture was sonicated for half an hour to obtain the black paste or gel. The compounds examined in this work were 2-NT (2-nitrotoluene), 3-NT (3-nitrotoluene), 2,6 DNT (2,6-dinitrotoluene), 2,4 DNT (2,4-dinitrotoluene), and TNT (2,4,6-trinitrotoluene). The LODs of each compound analyzed with this method were 12.4 ng/L, 8.9 ng/L, 0.9 ng/L, 1.6 ng/L, and 0.8 ng/L, respectively, and the relative recovery rates were 90–105%, 90–110%, 95–110%, 90–110%, and 95–107%, respectively, for each compound in the different water samples
[60].
Majidi et al. developed an alcohol-based DES as a carrier of SiO
2@Fe
3O
4, a sorbent used in magnetic d-μSPE. The aim of this study was to apply the above for the preconcentration and determination of morin in grape and apple juices, in diluted and acidic extracts of dried onion and in green tea infusion samples. Different molar ratios of tetramethylammonium chloride and EG were tested, and the optimal ratio was 1:3. The addition of 20%
v/v methanol in the DES substantially enhanced the extraction efficiency of morin. At optimal conditions, the LOD of the analyte was observed as 0.91 μg/L, while the recoveries of morin were 82.80–90.56% in apple juice samples, 91.06–95.56% in commercial red grape juice samples, 93.00–98.93% in onion samples, and 94.26–95.30% in green tea samples
[61].
Li G. et al. used a mixture of three DES in MIPs for the separation of two xanthines (theobromine and theophylline), (+)-catechin hydrate and caffeic acid from green tea samples via d-μMSPE. The ternary DES synthesized were ChCl-OA-EG, with molar ratios of 1:1:1, 1:1:2, and 1:1:3, ChCl-OA-Gly (1:1:1) and ChCl-Caffeic Acid-EG (1:1:1). ChCl-OA-Gly (1:1:1) gave the highest percentages of recovery for the four analytes, being 89.72–92.25% for theophylline, 87.15–91.86% for theobromine, 87.16–90.17% for (+)-catechin hydrate, and 86.17–92.32% for caffeic acid
[62].
Lastly, in an additional study, Li G. et al. synthesized hydrophilic MI chitosan (HMICS) based on DESs that were used as a template and as a functional monomer for the enrichment of gallic acid from red ginseng tea samples by SPE. From the three different molar ratios studied in this work, the best was ChCl-Gallic acid (1:2), as it was the only one that had a stable form. The adsorption capacity of gallic acid on HMICS with DES was 10.13 mg/g, 2.36 times higher than that of NICS (non-imprinted chitosan) without DES
[63].
3.1.2. Solid Phase Microextraction
Solid phase microextraction (SPME) is a sample preparation technique developed by Pawliszyn in 1990, where a fiber, with a small diameter, coated with a sorbent, is placed in aqueous solutions, so as to extract the analytes
[64]. There are mainly two approaches: headspace SPME (HS-SPME), when the coated fiber was exposed to the gas phase, absorbing volatile compounds in the headspace of gaseous, gas or liquid samples, and direct immersion SPME (DI-SPME), in which the fiber is immersed in a small volume of the liquid sample, extracting non-volatile and semi-volatile analytes. When the equilibrium between the fiber and the analytes is completed, the SPME fiber is inserted into a GC for thermal desorption or into a desorption solvent to a LC
[65]. Other approaches to this technique are thin-film SPME (TF-SPME), in-tube SPME, hollow fiber SPME (HF-SPME), and in-needle SPME
[43].
Wang et al. utilized the in-tube method of SPME, as well as a poly-DES monolithic column coupled online to HPLC for the determination of anti-inflammatory and non-steroidal drugs (ketoprofen, flurbiprofen and diclofenac). In fact, they prepared a polymer monolithic column based on the DES monomer of ChCl and itaconic acid, (molar ratio 3:2) and connected it on the six-port valve of the HPLC, in order to build up an online SPME-HPLC device. The linearity of the proposed method was proved by a determination coefficient of 0.9997 in the range of 0.05–50 ng/mL for the first two compounds, and 0.25–500 ng/mL for sodium diclofenac. The LODs are 0.01 ng/mL, 0.01 ng/mL, and 0.05 ng/mL, respectively, and the recoveries of these drugs in lake water samples are 91.6–113.7% for ketoprofen, 87.3–109.5% for flurbiprofen, and 91.4–109.2% for sodium diclofenac, while, in human plasma samples, the recovery rates were 84.5–105.5%, 84.7–104.5%, and 87–99.9%, respectively, for each NSAID
[66].
Li Tiemei et al. produced three hydrophobic DES, consisting of ethyl 4-hydroxybenzoate and methyl trioctyl ammonium chloride, with the molar ratios of 2:1, 1:1, and 1:2, which were used as an effective additive of the sol-gel sorbent coating of a PDMS fiber. This additive could create plenty of neat pores in the surface of the PDMS fiber, substantially improving its extraction efficiency. The new PDMS-DES fiber was evaluated by HS-SPME and GC-FID for the determination of VOCs, such as toluene, ethylbenzene, and o-xylene. The optimal molar ratio of the three that were tested was 2:1, as it exhibited the highest extraction efficiency. The linearity ranges of the new fiber were between 10–1000 μg/L, and the LOD was in the range of 0.005–0.025 μg/L
[67].
Recently, Mirzajani et al. fabricated hollow fiber and monolithic fiber based on MOFs-DES/MIPs for the extraction of phthalate esters in yoghurt, bottled water, and soybean oil under termed hollow fibers liquid membrane-protected SPME (HFLMP-SPME), followed by GC-FID. Under optimal conditions, LODs of dimethylphthalate, diethylphthalate, di-isobutylphthalate, and n-dibutylphthalate were 0.008 μg/L, 0.012 μg/L, 0.030 μg/L, and 0.030 μg/L, respectively, while the mean recoveries of the analytes varied from 95.5–100.0%. The DES consisted of ChCl-Gly (2:1), which was then used for the preparation of the MOF-DES/MIPs SPME fibers
[68].
3.1.3. Stir Bar Sorptive Extraction
Stir bar sorptive extraction (SBSE) was originally developed and published by Baltussen et al. in 1999. They presented a novel approach on sample enrichment by using the sorbent PDMS coated onto stir bars with a length of 10 mm and 40 mm (coated with 55 μL and 219 μL of PDMS liquid phase, respectively). The former is more suitable for stirring sample volumes from 10–50 mL, and the latter is suited for volumes up to 250 mL
[69]. When the adsorption process completes, the back-extraction of the analytes takes place, by liquid desorption, into an appropriate organic solvent. Benedé and her colleagues introduced a combination of SBSE and d-μSPE in 2014, which was called stir bar sorptive-dispersive microextraction (SBSDME). They utilized a neodymium-core stirring bar physically coated with a hydrophobic magnetic nanosorbent (MNP). Depending on the stirring speed, the magnetic nanosorbent either acts as a coating to the stir bar, rendering extraction feasible like SBSE, or as a dispersed medium for the extraction of the target analytes, like d-μSPE. To extract the analytes from the sample, the same methodology with SBSE was carried out
[70].
In a study, Zarei et al. prepared a DES magnetic nanofluid, which was coupled with SBSDME as a hyphenated sample technique for the efficient extraction of ultra-trace nitroaromatic explosives in wastewater. Out of the four DESs, the one that was the most compatible for this work was ChCl:Resorcinol (1:2 molar ratio). The extraction solvent used was acetone, and the extract was then injected into the GC instrument for analysis of the compounds. The LOD of the analytes were 4.92 for 2-NT, 3.19 for 3-NT, 0.25 for 2,6-DNT, 0.45 for 2,4-DNT, and 0.22 ng/L for TNT. The related linear ranges were between 20–1500 ng/L, 15–1500 ng/L, 1–1500 ng/L, 2–1500 ng/L, and 1–1500 ng/L, respectively, for each target compound. It is also mentioned that the developed method is both user and environmentally friendly, according to the principles of the green analytical chemistry
[71].
Nemati et al. developed a SBSE method coupled to solidification of floating droplets in DLLME based on DESs for the effective derivatization and simultaneous extraction of acidic pesticides from tomato samples. Initially, the target compounds are adsorbed on a coated stir bar from tomato juice filled in a narrow tube. Then, the stir bar is removed upon extraction and put in a water-miscible DES (ChCl:EG 1:2) to elute the analytes. Finally, a derivatization agent and a water-immiscible DES (ChCl:Butyric acid 1:2) are added to the eluent. The LODs of the five acidic pesticides were 7 ng/L for dalapon, 9 ng/L for MCPA, 14 ng/L for 2,4-dichlorophenoxyacetic acid 7 for fenoxaprop, and 11 ng/L for haloxyfob. Lastly, their mean relative recoveries were 92–95%, 94%, 93–95%, 92–98%, and 94–99% for each compound, respectively
[72].
3.2. Liquid Phase Microextraction
Liquid phase microextraction (LPME) is a well-known term referring to a group of miniaturized techniques, based on the fundamentals of liquid–liquid extraction, while overcoming many of its drawbacks. LPME has been developed at the late 1990s, following the principal introduction to microextraction after the establishment of SPME
[73][74]. Those techniques focus on the extraction and enrichment of analytes during the sample preparation with the use of fewer resources, primarily by reducing the quantities of solvents to less than 100 μL, thereby harmonizing with the principles of green analytical chemistry (GAC). LPME can be classified into three different types: (1) hollow fiber liquid phase microextraction (HF-LPME), (2) single drop microextraction (SDME), and (3) dispersive liquid–liquid microextraction (DLLME). In addition to the tendency of reducing the use of solvents and the produced waste, in recent years, there have been new applications for specific techniques that include green solvents, such as ionic liquids and deep eutectic solvents, in order to substitute toxic organic reagents and minimize the cost
[75][76].
In the recent years, there are many reports for LPME combined with DESs for the extraction of several compounds, with different applications based on the three different categories mentioned previously. One of the first applications of DES with LPME was introduced by Karimi et al. for the extraction, preconcentration and determination of lead and cadmium in edible oils, coupled with electrothermal atomic absorption spectrometry (ETAAS). The selected DES was a mix of ChCl-Urea (4:1 molar ratio) mixed with a 2% solution of nitric acid. The limits of detection resulted to be 0.008 ng g
−1 for Pb and 0.0002 ng g
−1 for Cd, results that were comparable or even better than those of similar methods. Additionally, the method was applied to edible oils, such as olive oil, sunflower oil, sesame oil, and soybean oil, with good recoveries (95–104%) from spiked samples
[77].
Furthermore, the utilization of DES on ultrasound assisted LPME is substantial, with most applications referring to the determination of metals in water and food samples. The ultrasonication effect is used so that the aggregated DES droplets will gradually break into tiny droplets, while the use of an emulsifier for further enhancement of the process (usually with addition of THF) is possible too and is also referred to as emulsification liquid phase microextraction (ELPME). The stage that follows is centrifugation to achieve separation between the aqueous and DES rich phase. The DES phase is isolated, filled with extra volume of a proper solvent, if necessary, and analyzed with the selected method. The main combination chosen as an extraction solvent is ChCl-phenol at molar ratios of 1:3 and 1:4, applied to different samples for the determination of several analytes such as: cobalt Co (II) in a pharmaceutical supplement and tea samples by a microsample injection system, coupled with flame atomic absorption spectrometer (MS-FAAS)
[78], Cr(III), and Cr(VI) in water samples with MS-FAAS
[79], malachite green (MG) in farmed and ornamental aquarium fish water samples by UV-VIS spectrometry
[80], cadmium (Cd) in water and ice tea samples by electrothermal atomic absorption spectrometry (ETAAS)
[81], Patent Blue V analysis in syrup and water samples by UV-VIS
[82], Hg
2+ and CH
3Hg
+ in water samples and freshwater fish samples with ETAAS
[83], and pesticide residues in different traditional Chinese medicines (TCMs) with HPLC-DAD
[84].