A Natural Monoterpene Enol for Dispersive Liquid–Liquid Microextraction Based on Solidification of Floating Organic Droplets for Determination of Benzophenone Compounds in Water Samples

Abstract

In the current study, an effective and simple procedure of extraction for the four benzophenone compounds from water samples was achieved by dispersive liquid–liquid microextraction (DLLME) based on the solidification of floating organic droplets (SFO) with a natural monoterpene enol as the extractant. As a natural, high solidification point, inexpensive, and environmentally friendly hydrophobic solvent, α−terpineol was selected firstly as an extractant for DLLME and could be collected and transferred easily after extraction by solidification at a lower temperature. Several main parameters closely related to extraction efficiencies, such as volume of extractant, extraction time, pH and salt concentration of the sample solution, temperature, and time of the solidification process, were investigated in detail. The results showed that the established method had good extraction performance for benzophenone compounds with enrichment factors in the range of 29–47. Furthermore, the linearities were over the range of 2–2000 µg/L, and the limits of detection were 0.12–0.53 µg/L for four benzophenone compounds. The recoveries ranged from 80.2% to 108.4%, with RSDs (intra- and inter-assay) less than 8.5%. At last, the method applicability was investigated by the determination of the benzophenone compounds in aqueous solutions, and satisfactory recoveries (83.0–107.0%) were acquired. Taken together, α−terpineol, a natural monoterpene enol, was first used as an extractant of DLLME-SFO, which provided an alternative method with simplicity and rapidity for the determination of benzophenone compounds in aqueous samples.

1. Introduction

Nowadays, sample preparation technology has been developing towards high efficiency, rapidity, environmental sustainability, economy, and miniaturization [1,2]. Microextraction technologies, including solid phase microextraction (SPME) [3,4] and liquid phase microextraction (LPME) [5,6,7,8], have obtained great attention and overcome some intrinsic disadvantages such as consumption of many organic solvents, cumbersome steps, and time-consuming and laborious processes in traditional sample preparation techniques. Accordingly, microextraction technology, with its advantages of simplicity and rapid extraction equilibrium, plays a more and more important role in environmental, biological, and medical analysis [2,8,9].

Among various LPME techniques, dispersive liquid–liquid microextraction (DLLME), first proposed by Assadi et al. [10] in 2006, has aroused widespread interest in sample preparation [10,11]. However, in traditional DLLME, the available extractants are very limited, and a small amount of extractant is difficult to collect for the following instrumental analysis. Some high-density extractants (>1 g/cm³ at 20 °C) can be collected stably at the bottom of the centrifuge tube after the centrifugation step [11]. However, the recovery process is relatively difficult for lighter-density extractants than water. Accordingly, solidification of the floating organic droplet (SFO) [12,13,14,15,16,17,18], collection by nanoparticles [5,19], and the upper narrow neck of the vessel [20,21] are usually employed for the recovery of those light extractants. In the SFO procedure, the extractant droplet with a high solidification point can be collected and transferred easily after extraction by solidification at a lower temperature [13]. Currently, only a handful of traditional organic extractants such as 1-undecanol, 1-dodecanol, 1-nonanoic acid, and 1-octanoic acid have been employed in DLLME-SFO to extract some halogenated hydrocarbons and phytosterol compounds from aqueous or oil samples [12,13,16]. Considering the great application potential of DLLME-SFO in sample preparation, some new-style, natural, and environmentally friendly extractants with low densities and appropriate solidification points should be explored.

α−Terpineol, a monoterpene enol essential oil, is widely used in the preparation of daily and edible flavors [22]. It has a high melting point (37 °C) and low vapor pressure (0.4 mmHg), is slightly soluble in water (0.71 g/L at 20 °C), and lighter than water (0.934 g/mL at 20 °C) (shown in

Table 1. Structure and physicochemical properties of benzophenone compounds and extractants in this study.

Compound	Structure	pKa	LogP
4-Hydroxybenzophenone (HB)		7.95	3.07
2,4-Dihydroxybenzophenone (DB)		7.53	2.96
Benzophenone (BP)		—	3.38
2-Hydroxy-4-methoxybenzophenone (HMB)		7.56	3.52
α-Terpineol		15.09	2.98

), which makes it an excellent hydrophobic extractant for DLLME-SFO. In our previous work, a deep eutectic solvent (DES) was synthesized by mixing α−terpineol with 1-octanol and used for the extraction of phenols in aqueous solutions [23]. As far as we know, no other reports have been published directly using α−terpineol as an extractant. Therefore, it is of significance to extend its application as an extractant in DLLME-SFO.

Benzophenone compounds [24,25,26] are widely applied to the formulation of sunscreen products and industrial products such as plastics and coatings. The structure and properties of four representative benzophenone compounds are shown in Table 1. However, some reports from the International Agency for Research on Cancer (IARC) showed that benzophenones in general had the environmentally critical properties of being highly lipophilic, persistent, bioaccumulative, and toxic and would be harmful to aquatic organisms [27]. Even so, it will be highly desirable to develop an efficient and environmentally friendly method for analyzing benzophenones in water solutions. At present, the methods for detecting such substances in the literature include HPLC-UV [5,28,29,30], CE-UV [31], GC-MS [7,9,32], and LC-MS [6,26,33,34,35,36]. Among them, HPLC-UV was mostly popular in routine monitoring due to its low cost and ease of operation.

Herein, an effective method for analyzing the benzophenone compounds in aqueous solutions was developed by DLLME-SFO combined with HPLC. α−Terpineol was used as the extractant to capture benzophenone compounds through hydrophobic and hydrogen bonding interaction. Several parameters closely related to the extraction efficiencies of benzophenones were investigated in detail. At last, the developed method’s applicability was examined by the determination of the benzophenone compounds in water samples.

2. Experimental

2.1. Chemicals and Reagents

Four benzophenone compounds, including 4-hydroxybenzophenone (HB, 99%), 2, 4-hydroxybenzophenone (DB, 99%), benzophenone (BP, 99%), and 2-hydroxybenzophenone (HMB, 99%), were obtained from Anpel Technologies (Shanghai, China). α−Terpineol (98%) was brought from Aladdin (Shanghai, China). Acetic acid (AcOH), acetonitrile (ACN), and methanol (MeOH) of chromatographic grade were obtained by Sinopharm (Shanghai, China). Purified water was produced by the Millipore apparatus (Bedford, USA).

The benzophenone stock solution (1000 mg/L) was prepared with MeOH, and the working solution was prepared by diluting the above stock solution with water. All prepared solutions are stored in a refrigerator at 4 °C.

2.2. Instrumentation

The LC20 HPLC analysis system was obtained from Shimadzu (Japan) with a diode array detector (DAD). Chromatographic separation was performed using a Symmetry C18 column (4.6 × 250 mm, I.D. 5 μm), with the isocratic elution composed of ACN and water containing 1% AcOH (60:40, v/v). The UV wavelengths were conducted at 290 nm for BP and 254 nm for the other analytes (HB, DB, and HMB), respectively, and the flow rate was set at 1.0 mL/min.

2.3. SFO-DLLME Procedure

An aliquot of the sample (8.0 mL) was added to a 10 mL screwcap glass centrifuge tube, and then 80 µL of melted α−terpineol was added using a long injection needle. The mixture was homogenized by ultrasound at 40 KHz for 10 min to form a cloudy solution. After centrifugation at 5000 rpm for 2 min, the screwcap glass centrifuge tube was put into a container filled with ice water for 5 min, and then the solidified extractant was taken out by spoon and melted in a water bath at 40 ℃. Finally, 20 µL of the extractant was mixed fully with ACN of equal volume before injection.

3. Results and Discussion

3.1. Characterization of α−Terpineol

The vibration of atoms in molecules reveals characteristic functional groups of organic compounds [37,38]. Figure 1 showed the characteristic spectra peaks of α−terpineol at 2966 and 3383 cm⁻¹, which represented the stretching vibrations of C–H and O–H, and the same groups also exists in benzophenone compounds, so hydrophobic and intermolecular hydrogen bonding interaction might be attributed to the extraction mechanism between them.

3.2. Optimization of Extraction Conditions

Several parameters related to the extraction efficiency, such as volume of extractant, extraction time, pH and salt concentration of sample solution, temperature, and time of solidification, were investigated, respectively. Herein, the optimization of the DLLME-SFO procedure was performed in a mixed benzophenone standard solution at 50 µg/L. Three replicates were set for each trial.

3.2.1. Volume of the Extractant

The amount of α−terpineol is of significance to efficiently extract target analytes in DLLME. In the current study, the influences of α−terpineol ranging from 60 μL to 120 μL on the extraction efficiency were investigated (Figure 2a). As the amount of the extractant increased, the response of the four benzophenone compounds gradually decreased. In other words, lower enrichment factors but higher absolute extraction recoveries could be obtained with an increasing amount of the extractant. Of course, a higher enrichment factor could be obtained by using a relatively small amount of extractant. However, when the amount of α−terpineol was less than 80 μL, the extractant was inconvenient to be collected after solidification. Considering comprehensively, 80 μL of α−terpineol was selected as the subsequent experimental.

3.2.2. Extraction Time

In order to accelerate the dispersion of the extractant in the water solution and reach the extraction equilibrium quickly, extraction time was investigated ranging from 5 min to 20 min. Moreover, we defined the extraction time as ultrasonic time after α−terpineol was injected into the sample solution. As shown in Figure 2b, extraction equilibrium could be reached within 10 min, and prolonged extraction time had no effect on the extraction efficiencies. Therefore, the extraction time was set at 10 min to speed up the sample preparation process.

3.2.3. pH

As the status of some benzophenone compounds with phenolic hydroxyl groups was transformed by the pH of the solutions, the pH investigated ranged from 3 to 11. As shown in Figure 2c, the pH had no effect on the extraction efficiency of BP because of the lack of ionizable groups, while the extraction efficiencies of HB, DB, and HMB decreased slightly when the pH was from 3.0 to 7.0 and dropped rapidly when the pH exceeded 7.0. This is attributed to the fact that the pKa of HB, DB, and HMB was 7.95, 7.53, and 7.56, and the benzophenone compounds existed mainly in a neutral state when the pH value was less than two units relative to their pKa value, which was favorable for extraction through hydrophobic and hydrogen bonding. The benzophenone compounds gradually ionized with the increase in pH, which was not conducive to extraction. Therefore, pH 5.0 was the optimal extraction condition.

3.2.4. Salt Concentration

To evaluate the influence of salt concentration on extraction efficiency, 0–100 mM of sodium chloride was added to the aqueous solutions. As shown in Figure 2d, the extraction efficiency increased slowly with increasing salt concentration and then decreased gradually in the range of 0–120 mM of sodium chloride solutions, mainly because the addition of inorganic salt in liquid samples could increase the ionic strength of the solution and decrease the solubility of organic compounds, i.e., the “salting-out effect” [39]. However, excessive salt content could also hinder the mass transfer process. In general, the ionic strength in this range had a limited effect on improving the extraction efficiency. Therefore, no sodium chloride was added for ease of operation.

3.2.5. Temperature and Time of Solidification Process

To investigate the effect of the solidification process of four benzophenone compounds, the condensation system ranged from −20 °C to 20 °C (the solidification time was set at 15 min). The extraction phase did not solidify within 24 h when the solidification temperature was set at 20 °C. As shown in Figure 2e, the specified temperature ranged from −20 °C to 10 °C had little effect on the extraction efficiency. Moreover, α−terpineol solidified slowly at 10 °C (about 9 min), while it only took 5 min to solidify at 0 °C. It also could be observed that, when the temperature was −20 °C, the water phase also began to solidify, which affected the separation of the extractant from the water phase. Taken together, the temperature of solidification was set at 0 °C.

Subsequently, the time of solidification was also investigated. The result in Figure 2f showed that the specified time had little effect on the extraction efficiency in the range of 5–25 min. Thus, 5 min of solidification time was sufficient to shorten the sample preparation process.

3.3. Method Performance of SFO-DLLME-HPLC

The feasibility of the developed method was evaluated through relevant parameters under the above optimal parameters, including linearity, the limit of detection (LOD), accuracy, precision, and enrichment factor. As summarized in

Table 2. Linearity range, correlation coefficient, limit of detection, enrichment multiple, accuracy, and precision (RSD) of the developed method.

Analytes	Linearity Range (μg/L)	R2	LOD (μg/L)	Enrichment Factor	Recoveries (RSD, %)
					Intra-Assay (n = 6)			Inter-Assay (n = 3)
					2 µg/L	200 µg/L	2000 µg/L	2 µg/L	200 µg/L	2000 µg/L
HB	2–2000	0.9967	0.12	36	80.9 (6.2)	95.8 (3.4)	98.7 (3.8)	80.2 (4.4)	98.2 (5.7)	99.6 (4.8)
DB	2–2000	0.9986	0.30	38	81.4 (5.8)	100.3 (5.6)	93.7 (2.2)	85.0 (5.9)	96.4 (6.0)	97.9 (5.2)
BP	2–2000	0.9925	0.37	47	80.6 (4.7)	97.6 (4.3)	98.9 (4.5)	81.2 (5.6)	99.5 (4.9)	99.3 (3.8)
HMB	2–2000	0.9915	0.53	29	81.2 (6.4)	95.1 (6.2)	96.5 (5.8)	82.0 (8.5)	108.4 (5.1)	97.9 (4.9)

, good linearities (R2 > 0.9915) were acquired, ranging from 2 to 2000 µg/L for the benzophenone compounds. The LODs for the proposed method, defined as an S/N of 3, ranged from 0.12 to 0.53 µg/L. The accuracy and precision were reflected by the recoveries and relative standard deviations (RSDs) of three different concentration levels, respectively. As shown in Table 2, excellent accuracies were found ranging from 80.2% to 108.4% with RSDs less than 6.4% for intra-assay (n = 6) and 8.5% for inter-assay (n = 3). The enrichment factors of the four benzophenone compounds were between 29 and 47.

3.4. Determination of Real Water Samples

Four samples collected in the surrounding area of Wuhan city (China), including tap water, lake water, domestic wastewater, and wastewater from a paint factory, were analyzed using the above-developed method. As shown in

Table 3. Analytical results of environmental water samples.

Sample	Analytes	Found (μg/L)	Added (μg/L)	Determined (μg/L)	Recoveries (%)	RSDs (%) (n = 3)
The tap water	HB	N.D. a	50	42.6	85.2	4.6
	DB	N.D.	50	47.9	95.8	3.4
	BP	N.D.	50	50.2	100.4	5.7
	HMB	N.D.	50	48.4	96.8	4.1
The lake water	HB	N.D.	50	47.8	95.6	6.3
	DB	N.D.	50	49.1	98.2	4.8
	BP	N.D.	50	43.7	87.4	3.9
	HMB	N.D.	50	48.9	97.8	2.9
Domestic wastewater	HB	N.D.	50	50.7	101.4	2.6
	DB	N.D.	50	50.6	101.2	2.1
	BP	N.D.	50	53.5	107.0	3.7
	HMB	N.D.	50	47.9	95.8	1.4
The paint factory wastewater	HB	54.2	50	95.7	83.0	2.5
	DB	N.D.	50	45.8	91.6	1.5
	BP	N.D.	50	46.7	93.4	3.8
	HMB	N.D.	50	41.8	83.6	0.7

^a Not detected.

, 54.2 µg/L of HB was found in the factory wastewater, and no analyte under investigation was detected in the remaining samples. Subsequently, 50 µg/L of benzophenone standard solution was added to the above samples to evaluate the applicability of the developed method, and the results (Table 3) revealed that the recoveries of the benzophenone compounds ranged from 83.0% to 107.0% with RSDs less than 6.3%, which fully indicated that the developed method had good applicability and could be used for the routine determination of the benzophenone compounds in water samples. The representative chromatograms of the wastewater samples from a paint factory are shown in Figure 3.

3.5. Method Comparison

An object comparison with reported literature for the determination of benzophenone compounds in aqueous solutions was performed, as listed in

Table 4. Comparison with previous literature for the determination of four benzophenone compounds.

Matrix	Extraction Technique	Characteristics	LODs	Recovery (%)	Instrumental Analysis	Ref.
Environmental water	DLLME	IL	DB: 0.02 μg/L HMP: 0.016 μg/L	70.5–90.3 79.5–91.3	UPLC-PDA	[5]
Root vegetables	UA-DLLME	ACN	HB: 0.3 μg/L DB: 0.15 μg/L HMP: 0.03 μg/L	70–90 87–92 98–111	LC-MS/MS	[6]
Beach sand	SPME	AcOEt	HMB: 2.5–3.3 ng/g (LOQ)	>85	GC-MS/MS	[9]
Human whole blood, plasma and urine	FPSE	Sol-gel sorbent coated fabric	DB: 30 μg/L BP: 30 μg/L HMB: 30 μg/L	—	HPLC-PDA	[28]
Aqueous solution	AA-DLLME	DES	HB: 0.1 μg/L DB: 0.1 μg/L BP: 0.1 μg/L HMP: 0.05 μg/L	93.5 94.1 96.6 94.5	HPLC-DAD	[29]
Environmental water	MA-DLLME	AC; SIL	DB: 1.21 ng/L HMP:1.19 ng/L	70–116 82–106	LC-MS/MS	[33]
Tap water, lake water, domestic wasterwater, factory wastewater	SFO-DLLME	α-Terpineol	HB: 0.12 μg/L DB: 0.30 μg/L BP: 0.37 μg/L HMP: 0.53 μg/L	78.9–99.6 75.0–100.3 81.2–99.5 71.2–108.4	HPLC-DAD	This work

* UA, Ultrasound-assisted extraction; AA, Air-assisted; MA, microwave-assisted; FPSE, Fabric phase sorptive extraction; SIL, Stable isotope labeling; PDA, photo-diode array; DAD, Diode array detector; IL, Ionic liquid; AcOEt, Ethyl acetate; DES, Deep eutectic solvent.

. On the whole, low LODs (0.12–0.53 µg/L) and easy operation procedures were found in this study. In particular, some organic reagents, such as ACN, ethyl acetate, and acetone, were frequently used in previously published methods, while only a small amount of organic compound (80 µL) was needed in the proposed SFO-DLLME method. Furthermore, α−terpineol, as a natural edible flavor, is efficient, green, and economical, which could conform to the requirements of green chemistry, which expanded further the application range of DLLME. The presented method was very reliable for the determination of the benzophenone compounds in water samples.

4. Conclusions

In this study, α−terpineol, a natural edible flavor monoterpene enol compound, was firstly used as an extractant of DLLME for the determination of the benzophenone compounds in water samples. Because of the high solidification point and hydrophobic properties, it could be easily collected and transferred in an ice water bath after extraction. Furthermore, only a small amount of an extractant was used in the as-proposed method compared with many organic reagents in previous literature. Under optimized conditions, the as-proposed method exhibited good precision (RSDs ≤ 8.5%), good accuracy (80.2–108.4%), and low LOD (0.12–0.53 µg/L). At last, this method was further confirmed to be effective, simple, and reliable through the determination of benzophenone compounds in real samples, which could also be expected for the analysis of analogous hydrophobic compounds to further expand the application of DLLME.