Analysis of Polycyclic Aromatic Hydrocarbons in Water Samples Using Deep Eutectic Solvent as a Dispersant in Dispersive Liquid–Liquid Microextraction Based on the Solidification of Floating Organic Droplet

Abstract

A novel dispersive liquid–liquid microextraction method based on the solidification of a floating organic droplet was proposed to pre-concentrate 16 polycyclic aromatic hydrocarbons (PAHs) from water samples prior to their determination using gas chromatography–mass spectrometry, in which the effect of hydrophilic deep eutectic solvent (DES) used as a dispersant was investigated. The main extraction parameters were optimized, and the procedure was validated. DES2 synthesized using choline chloride with acetic acid at a molar ratio of 1:2 was selected as the dispersant. Under the optimum extraction conditions, 12 mL of the water sample was injected into the mixed solvent containing 60 μL of 1-dodecanol (extractant) and 316 μL of DES2, ultrasound-mixed for 4 min, and then centrifuged for 5 min to separate the phases. The proposed method showed good linearity in the range of 0.02–5.0 μg/L; the limits of detection were 3.5–14.1 ng/L, the limits of quantification were 11.8–46.9 ng/L, the relative standard deviations were below 6.1%, and the enrichment factors ranged from 142 to 175 for the 16 PAHs. Finally, the proposed method was successfully employed to determine PAHs in real water samples.

1. Introduction

Polycyclic aromatic hydrocarbons (PAHs) are a group of aromatic substances containing two or more benzene rings; they are released into the environment mainly through natural and anthropogenic sources [1,2,3]. PAHs are common environmental pollutants because of their genotoxicity, mutagenicity, and carcinogenicity in humans, and their ubiquitous existence in air, soil, and water [4,5,6]. In addition to air and soil, PAH pollution is a serious issue in water; studies have shown that rivers around the world are polluted by PAHs [7,8,9]. However, assessing PAH pollution in water is not easy, as there are often only trace levels of PAHs in water due to their low hydrophilicity; therefore, sample preparation is required to achieve sufficient sensitivity before determining PAHs in water samples.

Sample preparation is usually necessary when analyzing analytes in water samples; this can pre-concentrate the analytes and reduce the matrix effect [10,11,12]. In recent years, analytical researchers have attempted to develop a variety of methods for the preparation of water samples. In 2006, Rezaee et al. developed the dispersive liquid–liquid microextraction (DLLME) method, whereby the extractant is dispersed into microdroplets in the aqueous solution in the presence of the dispersant, thus significantly increasing the contact area between the two phases and enhancing the extraction efficiency [13]. DLLME overcomes the limitations of the traditional liquid–liquid extraction method, such as being a cumbersome operation, being time consuming, and using a large amount of organic reagents. In order to simplify the collection of microvolumes of the extractant droplets, DLLME based on the solidification of organic droplet (DLLME-SFOD) method was introduced by Leong et al. [14]. In DLLME-SFOD, the extractant has a lower density than water and a melting point near room temperature. After phase separation, the extractant in the upper layer solidifies at a low temperature; it is then collected and melts at room temperature for determination.

In DLLME, a chemical dispersant is often used to facilitate the microdroplet formation of extractant due to its miscibility with the organic solvent and water [12,13,14]. Methanol, ethanol, acetonitrile, and acetone are commonly used as dispersants, but these organic solvents are potentially harmful to human health and the environment. Hence, the use of greener dispersants has attracted the attention of analytical researchers. Deep eutectic solvents (DESs) are synthesized by combining a hydrogen bond acceptor (HBA) and hydrogen bond donor (HBD) into a mixture [15]. DESs have similar properties to ionic liquids and most are liquid at room temperature; due to their properties, such as low toxicity, easy preparation, and biodegradability, DESs have attracted attention as green alternatives to conventional hazardous solvents in microextraction [16,17]. Recently, many hydrophobic DESs have been used as extractants in DLLME for the preconcentration of organic and inorganic analytes from aqueous samples [18,19,20]. On the other hand, some hydrophilic DESs have successfully served as the dispersants in DLLME [21,22,23]. However, the application of DESs as the microextraction dispersant for determining PAHs has scarcely been reported.

The purpose of this study is to introduce a hydrophilic DES as the dispersant in DLLME-SFOD for the preconcentration of 16 PAHs from water samples prior to their determination by gas chromatography–mass spectrometry (GC-MS). The influence of the main experimental parameters during the extraction process was investigated to optimize the proposed method. Under the optimum conditions, the proposed method was validated and then applied to determine PAHs in real water samples.

2. Materials and Methods

2.1. Chemicals and Reagents

The mixed standard solution of 16 PAHs (naphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benz[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, indeno [1,2,3-c,d]pyrene, dibenz[a,h]anthracene, and benzo[g,h,i]perylene) was purchased from AccuStandard (New Haven, CT, USA). The stock solution of the mixed PAHs at a concentration of 10 mg/L (for each PAH) was prepared by diluting the PAH standard in acetonitrile and storing it at −20 °C.

The 1-decanol, 1-undecanol, and 1-dodecanol were supplied by Macklin Biochmical Co., Ltd. (Shanghai, China). Formic acid, acetic acid, propionic acid, and choline chloride (ChCl) were obtained from Energy Chemical Co., Ltd. (Anqing, China). Chromatographically pure methanol, ethanol, acetonitrile, and acetone were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). The ultrapure water was prepared using a Milli-Q water purification system (Millipore, Bedford, MA, USA).

2.2. Instrumentations

The analyses were performed on an Agilent 7890A gas chromatography coupled with an Agilent 5975 mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) equipped with an Agilent HP-5 capillary column (30 m × 0.32 mm × 0.25 μm). The carrier gas was helium at a constant flow rate of 1.3 mL/min. The column temperature was initially held at 50 °C for 1 min, then increased to 200 °C at 15 °C/min, held for 5 min, and finally increased to 290 °C at 10 °C/min and held for 1 min. The ionization voltage of the electron impact ion source was 70 eV, and quantitative analyses of PAHs in water samples were performed in selective ion monitoring mode. The GC-MS parameters for the analysis of the 16 PAHs are recorded in

Table 1. The GC-MS parameters for analysis of the 16 PAHs.

PAH	Retention Time (Min)	Characteristic Ions (m/z)
Naphthalene	5.804	127, 128 *, 129
Acenaphthylene	7.994	151, 152 *, 153
Acenaphthene	8.236	152, 153 *, 154
Fluorene	9.303	165, 166 *, 167
Phenanthrene	12.546	176, 178 *, 179
Anthracene	12.752	152, 176, 178 *, 179
Fluoranthene	16.057	101, 200, 202 *, 203
Pyrene	16.551	101, 200, 202 *, 203
Benz[a]anthracene	18.971	114, 226, 228 *, 229
Chrysene	19.030	113, 226, 228 *, 229
Benzo[b]fluoranthene	20.794	126, 250, 252 *, 253
Benzo[k]fluoranthene	20.845	126, 250, 252 *, 253
Benzo[a]pyrene	21.391	126, 250, 252 *, 253
Indeno[1,2,3-c,d]pyrene	23.950	138, 276 *, 274, 277
Dibenz[a,h]anthracene	24.040	139, 276, 278 *, 279
Benzo[g,h,i]perylene	24.664	138, 274, 276 *, 277

Note(s): * Quantitative ion.

. The chromatogram of Table 1 is shown in Figure S1.

An SB-800DT ultrasonic bath (Ningbo Xinzhi Biological Co., Ltd., Ningbo, China) was used in the extraction procedure. The Fourier transform infrared (FT-IR) spectra of DESs were recorded using a Thermo Scientific Nicolet 6700 FT-IR spectrometer (Thermo, Saint Louis, MA, USA). The proton nuclear magnetic resonance (¹H NMR 400 MHz) spectra of DESs were recorded using a Bruker spectrometer (Bruker, Fällanden, Switzerland). The DESs were diluted with deuterated dimethyl sulfoxide (DMSO-d₆) without further pretreatment before ¹H NMR analysis.

2.3. Sample Preparation

The working solution of PAHs was obtained by diluting the stock solution with acetonitrile. To optimize the DLLME conditions, the spiked water samples were freshly prepared by mixing ultrapure water with the PAH working solution at a certain concentration.

The tap water and river water samples were collected from our laboratory and a local river (Taizhou, China), respectively. The tap water sample was used without any treatment, whereas the river water sample was filtered by a 0.45 μm filter membrane before use.

2.4. DES Preparation

DESs were synthesized by mixing ChCl (as an HBA) with three short-chain acids (as HBDs) at a molar ratio of 1:2, respectively. Three kinds of DESs, including ChCl-formic acid (DES1), ChCl-acetic acid (DES2), and ChCl-propionic acid (DES3), were synthesized in screwed bottles. The mixture of ChCl and different HBDs was heated and stirred at 80 °C in a water bath until a clear liquid was formed. The synthesized DES structures are shown in Figure 1.

2.5. DLLME-SFOD Procedure

A total of 12 mL of the water sample was injected into the test tube containing the mixture of 60 μL of 1-dodecanol and 316 μL of DES2. An emulsion was formed after ultrasound mixing for 4 min. The emulsion was centrifuged at 4000 rpm for 5 min to separate the two phases. The test tube was transferred into a freezer at −20 °C for 5 min to solidify the floating extracted solvent, and then the solidified extracted solvent was collected and melted at room temperature. A total of 1 μL aliquot of extracted solvent was injected into a GC-MS for the PAH analysis.

2.6. Enrichment Factor Calculation

The enrichment factor (EF) was defined as the ratio of analyte concentration in the extracted solvent (C_final) and the initial concentration of analyte in the water sample (C_initial) (Equation (1)).

$$\mathrm{E}\mathrm{F}=\frac{C_{final}}{C_{initial}}$$

(1)

3. Results and Discussion

3.1. Characterization of DESs

The formation of hydrogen bonding between the halide anion of ChCl and different short-chain acids is the main force for DES formation. To confirm the formation of hydrogen bonding, the FT-IR spectra of pure ChCl, pure acetic acid, and DES2 were recorded; the results are shown in Figure 2a. The O-H and C-N stretching vibration peaks of ChCl are positioned at 3217.39 cm⁻¹ and 951.69 cm⁻¹, respectively. Both characteristic peaks of ChCl and acetic acid were observed in the spectra of DES2. Compared with that of acetic acid (3034.93 cm⁻¹), the O-H stretching vibration peak of DES2 shifted to a lower wave number (2961.97 cm⁻¹). This may be due to the transfer of the oxygen atom’s electron cloud to hydrogen bonding, consequently leading to a decrease in the absorption peak [24,25]. Accordingly, the shift of the O-H stretching vibration peak indicated the existence of hydrogen bonding between its components when DES2 was formed. The FT-IR spectra of DES1 and DES3 are reported in the Supplementary Materials (Figure S1).

The chemical structures of DESs were elucidated by ¹H NMR spectra. As shown in Figure 2b, all peaks of DES2 can be assigned to ChCl and acetic acid, and no new peaks were identified, suggesting that no side reaction is present during the synthesis of DES2. Thus, the ¹H NMR spectra indicated that the formation of DES2 occurred between the hydroxyl groups. The ¹H NMR spectra of DES1 and DES3 are reported in the Supplementary Materials (Figure S2).

3.2. Optimization of the Extraction Procedure

In this study, unlike the usual DLLME, the water sample was injected into the DES extraction mixtures and extractant using a syringe due to the high viscosity of the extraction mixtures and the difficulty of their aspiration into the syringe. To optimize the main experimental parameters, the DLLME procedure was first optimized using a one-factor-at-a-time approach followed by the Box–Behnken design (BBD) of the response surface methodology (RSM). The extraction efficiency was assessed in terms of the 16 PAHs’ total peak area.

3.2.1. Selection of Extractant

Three aliphatic alcohols, including 1-decanol, 1-undecanol, and 1-dodecanol, were selected as possible extractants. Based on the results in Figure 3, 1-dodecanol provides the best extraction efficiency among the studied extraction solvents. This can be attributed to an increase in the length of aliphatic alcohol, which enhances the solubility of hydrophobic analytes in the extraction solvent [22]. Therefore, 1-dodecanol was selected for the following studies. The peak area value of each PAH is recorded in the Supplementary Materials (Table S1).

3.2.2. Selection of Dispersant

To assess the effect of dispersant type on extraction efficiency, two hydrophilic DESs (DES1 and DES2) and four commonly used dispersion solvents (methanol, ethanol, acetonitrile, and acetone) were investigated. DES3 was discarded because of its incomplete miscibility with the extractant. The results in Figure 4 show that DES2 synthesized with ChCl and acetic acid provides a higher extraction efficiency. Moreover, it was observed that decomposition of the DES occurred after the water sample injection. DES decomposition can disperse the extractant well, possibly resulting in a salting-out effect created by the ChCl that promotes PAH extraction [23]. Thus, DES2 was selected for further experimentation. The peak area value of each PAH is recorded in the Supplementary Materials (Table S2).

3.2.3. Box–Behnken Design

RSM is a combination of statistical and mathematical techniques, which is useful for processing the experiment’s optimization parameters. In this study, the BBD of RSM was adopted using Design-Expert 12.0.3.0. Some other main parameters that influenced the extraction efficiency of 16 PAHs, including extractant volume, dispersant volume, and ultrasound time, were investigated. The experiments were performed using ultrapure water spiked with a 100 ng/L standard mixture of 16 PAHs. Experimental factors and levels in the BBD matrix are recorded in

Table 2. Experimental factors and levels in the Box–Behnken design matrix.

Factors	Levels
	Low (−1)	Central (0)	High (1)
X1-Volume of 1-dodecanol	60	90	120
X2-Volume of DES2	150	350	550
X3-Ultrasound time	1	3	5
Runs	X1	X2	X3	Total peak area
1	120	550	3	987,898
2	90	350	3	1,147,172
3	60	550	3	1,229,678
4	60	150	3	1,226,731
5	90	150	5	1,074,630
6	120	150	3	624,997
7	90	350	3	1,179,585
8	90	350	3	1,228,592
9	120	350	1	834,673
10	90	550	5	1,086,868
11	90	150	1	823,972
12	90	550	1	1,026,248
13	60	350	1	1,307,096
14	120	350	5	985,991
15	90	350	3	1,199,126
16	90	350	3	1,159,579
17	60	350	5	1,348,729

.

The quadratic model to predict the total peak area of the 16 PAHs in terms of actual factors is shown in Equation (2):

Y = 1,320,960.13 − 8685.67 X₁ + 1833.89 X₂ + 90,517.04 X₃ + 14.99 X_1×2 + 457.02 X₁X₃ − 118.77 X₂X₃ − 27.38 X₁² − 3.52 X₂² − 9760.63 X₃²

(2)

where Y is the total peak area of 16 PAHs, X₁ is the volume of 1-dedecanol, X₂ is the volume of DES2, and X₃ is the ultrasound time.

To examine the significance and fitness of the model, analysis of variance (ANOVA) was applied in the data analysis.

Table 3. ANOVA of the response surface model for the optimization of three experimental parameters of PAHs.

Source	Sum of Squares	Degree of Freedom	Mean Square	F-Value	p-Value
Model	568,267,029,114	9	63,140,781,013	50.20	<0.0001 **
X1-Volume of 1-dodecanol	352,243,719,453	1	352,243,719,453	280.07	<0.0001 **
X2-Volume of DES2	42,102,506,380	1	42,102,506,381	33.48	0.0007 **
X3-Ultrasound time	31,780,860,555	1	31,780,860,555	25.27	0.0015 **
X1X2	32,391,720,529	1	32,391,720,529	25.757	0.0014 **
X1X3	3,007,699,806	1	3,007,699,806	2.39	0.1659
X2X3	9,028,610,361	1	9,028,610,361	7.18	0.0316 *
X12	2,557,585,467	1	2,557,585,467	2.038	0.1969
X22	83,518,149,657	1	83,518,149,657	66.40	<0.0001 **
X32	6,418,184,246	1	6,418,184,246	5.10	0.0584
Residual	8,803,979,294	7	1,257,711,328
Lack of Fit	4,621,628,887	3	1,540,542,962	1.47	0.3486
Pure Error	4,182,350,407	4	1,045,587,602
Core Total	577,071,008,408	16
R2	0.9847
Adjusted R2	0.9651
Predicted R2	0.8605

Note(s): * Significant, p < 0.05. ** Highly significant, p < 0.01.

shows that the results of ANOVA revealed that the model was significant, with a p-value < 0.05 and F-value of 50.20. The lack of fit (LOF) p-value of 0.3486 (LOF p-value > 0.05) was non-significant; this was considered positive for the experiment. The positive relationship between the experimental data and fitted model was confirmed by the statistical parameters (determination coefficient (R²) = 0.9847, adjusted R² = 0.9651, and predicted R² = 0.8605). As a result, the proposed model could be well fitted to the experimental values.

To optimize the extraction efficiency, response surface plots for the effect of independent variables on the response values were constructed, which visually demonstrate the between-factor interactions. The experiments were carried out using 60–120 μL of 1-dodecanol to assess the effect of extractant volume on extraction efficiency. The results show that a higher total peak area of the 16 PAHs can be obtained by using 60 μL of 1-dodecanol (Figure 5a,b). Moreover, due to the dilution effect, the total peak area decreases as the extractant volume increases.

The dispersant type and the dispersant volume can affect the extraction efficiency in microextraction. Smaller dispersant volumes may be insufficient to disperse the extractant well, whereas larger dispersant volumes may increase the solubility of the hydrophobic analytes in the aqueous phase due to co-solvency, thus reducing the extraction efficiency [26]. Figure 5a,c demonstrate the DES2 volume’s effect on the extraction efficiency in the 150–550 μL range. The results show that the total peak area of the 16 PAHs gradually increases to 316 μL and then drops.

Ultrasonic waves can facilitate extractant diffusion and quicken mass transfer between two phases during microextraction, thus using less dispersant, shortening the operation time, and enhancing extraction efficiency [27,28]. To assess the effect of ultrasonic waves on extraction efficiency, ultrasound times ranging from 1 to 5 min were investigated. The results show that the total peak area of the 16 PAHs increases as the ultrasound time increases up to 4 min, whereas extensively long sonication results in a slight decrease in the total peak area (Figure 5b,c). This is speculated to be related to the acoustic cavitation in aqueous media, which can lead to the degradation of various organic substances, including PAHs [20,29,30].

According to the response surface plots, the optimum extraction conditions were chosen as follows: 60 μL of 1-dodecanol, 316 μL of DES2, and 4 min of ultrasound time. Under the optimum conditions, the experimental value was 1,421,153 ± 41,497, and the predicted value by BBD was 1,395,701, which indicated that the actual value and predicted value agreed well (1.82% error), further validating the fitted model.

3.3. Method Validation

Validation of the proposed method was performed with a series of water samples at different spiked concentrations. The linear ranges (LRs), limits of detection (LODs), limits of quantitation (LOQs), EFs, and repeatability of the 16 PAHs were investigated under the optimum conditions.

Table 4. Analytical characteristics of the proposed method.

PAH	LR (μg/L)	R2	LOD (ng/L)	LOQ (ng/L)	%RSD (n = 5)	EF
Naphthalene	0.03–5.0	0.9989	6.0	19.9	3.9	164
Acenaphthylene	0.03–5.0	0.9992	5.8	19.2	4.8	157
Acenaphthene	0.02–5.0	0.9991	3.5	11.8	4.3	160
Fluorene	0.03–5.0	0.9992	4.4	14.6	3.1	169
Phenanthrene	0.03–5.0	0.9973	6.3	21.1	4.2	165
Anthracene	0.05–5.0	0.9968	9.3	31.0	5.6	172
Fluoranthene	0.03–5.0	0.9997	4.5	14.9	2.8	152
Pyrene	0.03–5.0	0.9994	3.9	13.1	3.8	175
Benz[a]anthracene	0.03–5.0	0.9985	5.3	17.5	4.9	148
Chrysene	0.03–5.0	0.9984	4.8	16.1	5.4	154
Benzo[b]fluoranthene	0.05–5.0	0.9958	8.7	29.0	4.7	163
Benzo[k]fluoranthene	0.05–5.0	0.9965	9.1	30.5	5.9	146
Benzo[a]pyrene	0.05–5.0	0.9983	9.7	32.4	3.7	158
Indeno[1,2,3-c,d]pyrene	0.08–5.0	0.9936	14.1	46.9	5.5	149
Dibenz[a,h]anthracene	0.08–5.0	0.9943	12.8	42.7	6.1	142
Benzo[g,h,i]perylene	0.08–5.0	0.9977	11.7	38.9	5.2	151

shows the proposed method resulting in a good linearity of the 16 PAHs in the range of 0.02–5.0 μg/L, with R² ≥ 0.9936. The LODs were calculated to be between 3.5 and 14.1 ng/L based on a signal-to-noise ratio (SNR) of 3, and the LOQs (at SNR = 10) were between 11.8 and 46.9 ng/L. The EFs of the 16 PAHs were in the range of 142–175, with the relative standard deviation (RSD) ranging from 2.8% to 6.1%, thus showing good repeatability for the 16 PAHs.

3.4. Analysis of Real Water Samples

PAH determination of two real water samples (one tap water and one river water sample) was performed to assess the applicability of the proposed method. The pH and conductivity of the river water sample are 7.41 and 685.3 μs/cm, respectively. Under the optimum conditions, the real water samples with spiked concentrations of 2 and 5 μg/L were assessed. The results show that the relative recovery (RR) values of the spiked water samples ranged from 93.8% to 105.2% with RSDs of 2.6–6.4% (

Table 5. Analysis of real water samples under optimum conditions.

PAH	Spiked (μg/L)	Tap Water			River Water
		Found (μg/L)	%RR a	%RSD (n = 3)	Found (μg/L)	%RR	%RSD (n = 3)
Naphthalene	0	nd b	—	—	0.23	—	—
	2	2.07	103.50	3.8	2.27	102.00	4.2
	5	4.89	97.80	3.6	5.15	98.40	3.8
Acenaphthylene	0	nd	—	—	nd	—	—
	2	1.89	94.50	5.0	1.93	96.50	4.9
	5	4.69	93.80	4.6	4.96	99.20	4.8
Acenaphthene	0	nd	—	—	0.19	—	—
	2	1.99	99.50	4.1	2.14	97.50	3.8
	5	4.76	95.20	3.7	5.03	96.80	4.2
Fluorene	0	nd	—	—	nd	—	—
	2	1.97	98.50	3.1	1.96	98.00	2.6
	5	5.08	101.60	2.7	5.16	103.20	2.9
Phenanthrene	0	nd	—	—	nd	—	—
	2	2.10	105.00	4.3	2.05	102.50	4.5
	5	4.89	97.80	4.0	5.03	100.60	4.1
Anthracene	0	nd	—	—	nd	—	—
	2	1.91	95.50	5.8	2.05	102.50	5.4
	5	4.81	96.20	5.2	4.96	99.20	5.6
Fluoranthene	0	nd	—	—	nd	—	—
	2	1.97	98.50	2.6	1.89	94.50	2.9
	5	5.08	101.60	2.9	4.99	99.80	3.1
Pyrene	0	nd	—	—	nd	—	—
	2	1.98	99.00	3.6	1.97	98.50	3.9
	5	4.79	95.80	3.7	5.08	101.60	4.1
Benz[a]anthracene	0	nd	—	—	nd	—	—
	2	1.94	97.00	4.9	1.91	95.50	4.8
	5	5.06	101.20	4.6	4.83	96.60	5.1
Chrysene	0	nd	—	—	nd	—	—
	2	1.93	96.50	4.7	2.03	101.50	4.9
	5	4.89	97.80	5.2	4.70	94.00	5.5
Benzo[b]fluoranthene	0	nd	—	—	nd	—	—
	2	2.02	101.00	4.9	2.07	103.50	4.6
	5	5.23	104.60	4.6	4.98	99.60	4.3
Benzo[k]fluoranthene	0	nd	—	—	nd	—	—
	2	1.98	99.00	5.6	1.96	98.00	5.8
	5	5.26	105.20	5.9	5.12	102.40	6.2
Benzo[a]pyrene	0	nd	—	—	nd	—	—
	2	2.06	103.00	3.6	1.99	99.50	4.2
	5	4.91	98.20	4.0	5.18	103.60	3.8
Indeno[1,2,3-c,d]pyrene	0	nd	—	—	nd	—	—
	2	1.95	97.50	5.6	2.09	104.50	5.7
	5	5.04	100.80	5.4	4.89	97.80	5.9
Dibenz[a,h]anthracene	0	nd	—	—	nd	—	—
	2	2.04	102.00	5.9	1.95	97.50	6.2
	5	4.92	98.40	6.0	5.16	103.20	6.4
Benzo[g,h,i]perylene	0	nd	—	—	nd	—	—
	2	1.89	94.50	4.9	2.05	102.50	4.9
	5	4.86	97.20	4.7	5.01	100.20	5.1

Note(s): ^a Relative recovery (RR) = (concentration_found − concentration_nonspiked)/concentration_spiked. ^b nd = not detected.

). Thus, the proposed method is suitable for the preconcentration of PAHs from real water samples.

3.5. Comparison of the Proposed Method with Other Methods

The analytical parameters of the proposed method were compared with those reported in other studies for the determination of PAHs in aqueous samples. The results are listed in

Table 6. Comparison of the proposed method with other methods used for the determination of PAHs.

Extraction Method	Detection Method	Matrix	Extractant	Dispersant	LOD (μg/L)	%RSD	EF	Reference
SD- DLLME a	GC-MS	River water	Mixed solvent (Methylene chloride:n-hexane = 1:1, molar ratio)	Acetonitrile	0.0021– 0.0136	5.8– 10.9	94.9–103	[8]
USA- DLLME b	HPLC-UV	Effluent	DES (Thymol: ± Camphor = 1:1, molar ratio)	Acetonitrile	0.0039– 0.0098	2.20– 6.09	—	[20]
ELLME- DES c	HPLC-UV	Tap water, industrial wastewater	DES (Chcl:phenol = 1:2, molar ratio)	Tetrahydrofuran	0.09–0.7	—	151–170	[25]
DLLME- SFOD	HPLC-UV	Wastewater, lake water, tap water	1-dodecanol	Methanol	0.045–1.1	1.3–4.4	88–118	[31]
IL- DLLME d	HPLC-FLD	Tea infusions	Ionic liquid ([MOEDEA][FAP]) f	Acetonitrile	0.002–0.004	1.9–4.7	61–94	[32]
AA-LLME-SFDES e	HPLC-UV	Tea infusions	DES (DL-methol:dedecanoic acid = 3:1, molar ratio)	— g	0.16–0.75	0.9–2.3	15–18	[33]
DLLME- SFOD	GC-MS	Tap water, river water	1-dodecanol	DES (ChCl:acetic acid = 1:2, molar ratio)	0.0035– 0.0141	2.8–6.1	142–175	This work

Note(s): ^a SD-DLLME: solvent demulsification dispersive liquid–liquid microextraction. ^b USA-DLLME: ultrasound-assisted dispersive liquid–liquid microextraction. ^c ELLME-DES: emulsification liquid–liquid microextraction based on deep eutectic solvent. ^d IL-DLLME: ionic liquids-based dispersive liquid–liquid microextraction. ^e AA-LLME-SFDES: air-assisted liquid–liquid microextraction based on the solidification of floating deep eutectic solvents. ^f [MOEDEA][FAP]: ethyl-dimethyl-(2-methoxyethyl)ammonium tris(pentafluoroethyl)trifluorophosphate. ^g —: not referred.

.

The proposed method has a lower LOD and higher EF than DLLME-SFOD when using methanol as the dispersant [31]. Except for IL-DLLME [32], the LOD of the proposed method is comparable to or lower than those reported in other studies [20,25,31,33]. Additionally, the EF of the proposed method is comparable to ELLME-DES [25] and higher than other methods [8,31,32,33]. Moreover, in this method, the dispersant (DES2) is less toxic and volatile than the organic solvents used in other methods [8,20,25,31,32], and the low-density extraction solvent can be collected easily after phase separation without using any special device. Thus, the proposed method is simple, sensitive, and environmentally friendly; therefore, this method could be an attractive candidate method when PAHs in aqueous samples are under evaluation.

4. Conclusions

In this study, two hydrophilic DESs were selected as the potential dispersants in DLLME-SFOD for the preconcentration of 16 PAHs from water samples. The results showed that DES2 synthesized by ChCl and acetic acid at a molar ratio of 1:2 could assist the extractant (1-dodecanol) in pre-concentrating the analytes with better extraction efficiency than the commonly used organic dispersants. The method’s validation results demonstrated that the proposed method possessed low LODs, high EFs, and good repeatability for the 16 PAHs. For its simple operation, high sensitivity, and environmental friendliness, the proposed method was successfully used for the determination of trace levels of PAHs in real water samples.