Vortex-Assisted Ionic Liquid-Based Liquid-Phase Microextraction: a Simple, Low-Cost, and Environmentally Friendly Method for Speciation of Antimony in Water

A sensitive method is proposed to determine antimony species using vortex-assisted ionic liquid-based liquid-phase microextraction and digital imaging colorimetry. Sb III forms a ternary complex with 2-(5-bromo-2-pyridylazo)-5-(diethylamino)phenol, ascorbic acid, and iodide in an acid medium. The acceptor phase consists of the ionic liquid 1-butyl-3-methylimidazole hexafluorophosphate, dispersed in the aqueous solution with a vortex stirrer. After centrifugation, Sb III was determined directly in the rich phase. Sb V can also be measured by converting to reduced species. The method presented a limit of detection of 1.4 and 4.2 μg L −1 and an enrichment factor of 360. The precision (relative standard deviation) of the method is 7.0%. The procedure was applied to the speciation of antimony in river water and mineral water samples. Some results were compared with those obtained by electrothermal atomic absorption spectrometry (ET AAS). The method is an attractive alternative for determining antimony species, considering the aspects of green


Introduction
Antimony is a cumulative toxic element whose compounds are widely used in various industrial applications. 1,2 Some antimony compounds, such as antimony trioxide (Sb 2 O 3 ), are catalysts for synthesizing polyethylene terephthalate (PET). PET containers store drinking water, beverages, and condiments as raw materials in food packaging. Some researchers [1][2][3][4][5] have studied the contamination of water packed in PET bottles evaluating the leaching of Sb from the polymeric material to the water. The factors that influence the Sb content in water have also been studied. Thus, the development of robust and reliable methods for determining antimony species in water and food has been required.
Preconcentration techniques are essential for obtaining low limits of detection in the determination of many substances. Extraction techniques have been widely used for this purpose, such as liquid-phase extraction (LPE), 6,7 solid phase extraction (SPE), 8,9 and cloud point extraction (CPE), 8 among others. The formation of stable complexes between the antimony and some pyridylazo and thiazolylazo compounds has been the basis of procedures involving these techniques. [10][11][12][13][14][15] Liquid-phase microextraction (LPME) has reduced or eliminated some disadvantages of the classical technique (LPE), such as using large amounts of reagents and consequent waste management, operating time, tedious procedures, and risk to the operator, among others. Many approaches to LPME have been described in the literature, such as single-drop microextraction (SDME), 16 hollow-fiber liquid-phase microextraction (HF-LPME), 17 and dispersive liquid-liquid microextraction (DLLME). 18 DLLME methods are based on the dispersion of the extraction solvent into fine microdroplets. The cloudy mixture formed increases the surface area between the extraction solvent and the aqueous phase. A dispersive solvent is necessary. [19][20][21][22] In some LPME techniques, the dispersion of the extraction solvent can be performed in other ways, such as vortex agitation, ultrasound, pressure variation, and controlled temperature, among others, eliminating the use of dispersive solvent. These approaches have given rise to several microextraction techniques, such as pressure variation in-syringe dispersive liquid-liquid microextraction (PV-IS-DLLME), 23 ultrasound-assisted dispersive liquid-phase microextraction (UA-DLPME), 24 air-assisted dispersive liquid-phase microextraction (AA-DLPME), 25 temperature-controlled dispersive liquid-phase microextraction (TC-DLPME), 26 and VA-DLPME (vortex-assisted dispersive liquid-phase microextraction). 27,28 In VA-DLPME procedures, the extraction solvent is dispersed in the aqueous solution with a vortex stirrer, promoting emulsification. The solvent droplets formed in the emulsification quickly extract the component from the donor phase due to the large surface area obtained. The technique has low cost, short extraction time, and ease of operation. [29][30][31] Several methods employing LPME describe the use of ionic liquids (IL). 7,32-35 IL presents several advantages over conventional solvents. Some of the main benefits are the almost non-existent vapor pressure and the ability to solubilize organic, inorganic, and polymer compounds. 36 The search for environmentally friendly solvents and sustainable chemical processes is a priority between the principles of green chemistry and green analytical chemistry, and IL is an alternative to environmentally harmful organic solvents. 7,[37][38][39] Methods based on digital images have been widely used in analytical chemistry, allowing the development of fast, accurate, and low-cost analytical methodologies. [40][41][42][43][44][45] Digital images can be obtained from different imaging devices, including digital cameras, webcams, scanners, microscopes, and mobile phones. 23,[46][47][48][49][50] Sensors present in image capture devices, such as CCD (charge-coupled device) and CMOS (complementary metal-oxidesemiconductor), can convert the intensity of incident radiation into an analog electrical signal. This signal is converted to a digital value stored in bits and pixels arranged in an array, similar to the RGB model. 40 The most common color space on capture devices is RGB (R: red, G: green, and B: blue). RGB has been widely used in quantifying solutions and immobilized substances using digital image processing and specific software that allows extracting colorimetric information by selecting the areas of interest in an image. 43 The photographs are processed and analyzed with image processing software.
This work proposes a method for determining antimony species in water samples using microextraction in the dispersive liquid phase based on vortex-assisted ionic liquid-based liquid-phase microextraction (VA-IL-LPME) and digital imaging spectrometry (DIC). A rapid, low-cost, and environmentally correct approach to antimony speciation is proposed.

Apparatus
A Quimis model Q400AS pH meter (Diadema, Brazil) was used for pH measurements. The dispersion of the solvent in the aqueous phase was performed using a Lab Dancer (Ika, Campinas, Brazil) stirrer vortex. The capture of the digital images was performed in a system consisting of a box (32 × 23 × 8 cm) of medium-density fiberboard of white color to avoid the effects of reflection and the influence of external light. The lighting control inside the box was made using a lamp of LEDs (light-emitting diode) (6W) in white color, positioned at the top of the box, to guarantee uniformity and incidence of light and reproducibility of the image during measurements. The digital images of the system were recorded in a multifunctional portable microscope (Supereyes model B007, Shenzhen, China). A PerkinElmer (Shelton, USA) spectrometer model AAnalyst 400 equipped with an HGA 900 graphite atomizer was used to compare the results. An SP Labor centrifuge (model SP-32 T, Presidente Prudente, Brazil) was used to accelerate the decanting of the enriched phase.

Sample preparation
Mineral water samples were purchased in commercial establishments in Jequié, Bahia, Brazil. Four samples of each brand of mineral water were acquired. River water samples were collected in the Contas River in Jequié, Bahia, Brazil. The samples were filtered through a quantitative porosity filter of 7-12 μm with vacuum filtration. After filtration, the water samples were submitted for pH measurement, and the preconcentration procedure was applied. All samples analyzed in this study presented pH in the range between 6.0 and 6.7.  6 ] were added to the mixture. The system was taken to constant agitation in a vortex stirrer at 2500 rpm and then subjected to centrifugation for 5 min. The aqueous phase was discarded, and the reddish-rich-phase was photographed directly in the test tube. A schematic representation of the detection system is shown in Figure 1. The images were obtained directly in the test tube without the need to remove the enriched phase. The determination of total Sb was performed by promoting Sb III reduction, using 1.0% m v -1 potassium iodide and 10.0% m v -1 ascorbic acid as reducing agents. 51,52 The Sb V content was calculated by the difference between the quantity of total antimony and Sb III .
The images were stored in JPEG format with a resolution of 96 dpi. The image size was 640 × 880 mm, and the bit depth was 32. A square region of interest (ROI) in a fixed position was defined using Chemostat software. 53 This program allows the acquisition of RGB data for all pixels in ROI. This information is organized into a color histogram, and the average value of each color channel, red (R), green (G), and blue (B), is calculated. The absorbance (A) is represented as A = -log (P/P 0 ) versus the concentration of the element, where P is the average value of channel R, G, or B measured by sample or standard solution, and P 0 is the average acquired value of analytical blank.

Extraction solvent
The extraction efficiency is related to the proper choice of solvent. This substance should have some characteristics, such as low solubility in the aqueous phase and high extraction capacity for the components. Chloroform, trichloroethylene, carbon tetrachloride, and [C 4 MIM][PF 6 ] were tested as acceptor phases. As shown in Figure 2, chloroform, trichloroethylene, and [C 4 MIM][PF 6 ] provided similar results. Then, the ionic liquid was selected for further studies due to its potential to meet various environmental, health, and safety characteristics that differentiate it from conventional solvents. These characteristics make the solvent quite suitable for the criteria established by green chemistry.

Amount of extraction solvent
In microextraction methods, the amount of the extraction solvent controls the volume of the rich phase and the extraction efficiency. The effect of the extraction  solvent volume was studied in the range of 50.0 to 170.0 μL. According to Figure 3, the extraction solvent volume of 50.0 μL provided low analytical signals due to an insufficient acceptor phase. Volumes between 70.0 and 150.0 μL provided near average signals, although the smallest volumes in this range resulted in lower accuracy. When volumes above 150.0 μL were used, a decrease in signal was also observed due to excessive dilution. Based on these results, 120.0 μL of [C 4 MIM][PF 6 ] was chosen as the volume of extraction solvent for the subsequent studies.

Stirring time
A vortex agitator was used to promote the dispersion of the extraction solvent in the aqueous solution. This external energy source promotes a high efficiency in the mass transfer of the component. The stirring time required for displacement of the analyte from the aqueous phase to the extraction solvent was studied. The results can be seen in Figure 4. The time range investigated was 10 to 60 s. The analytical signal increases as a more significant amount of Sb III is transferred to the acceptor phase. After 30 s, the absorbance remains constant. So, the 40 s was the period chosen for the stirrer of the system in the later experiments.

Amount of Br-PADAP
The influence of Br-PADAP concentration on the extraction and determination of Sb III was evaluated at 2.15 × 10 -6 to 1.42 × 10 -5 mol L −1 range. According to the results observed in Figure 5, the analytical signal is maximum for concentrations greater than or equal to 7.16 × 10 -6 mol L −1 . Concentrations of ligands lower than this value are not enough to complex all the Sb III present in the medium so that a part of the element remains in the donor phase, impairing the extraction. Therefore, Br-PADAP solutions at a concentration of 7.50 × 10 -6 mol L −1 were used in later experiments.

Amount of hydrochloric acid
A three-part ionic coordination compound of bluegreenish coloration is formed between Sb III , iodide, and 5-Br-PADAP. 10 In the medium of the extraction solvent, the complex presents reddish staining. The complex is produced in an acidic medium, which contributes significantly to the intensification of the color. In addition, the acid medium contributes to the increase of the reducing effect of ascorbic acid. Hydrochloric and sulfuric acid were tested, and both showed similar results. Next, the concentration of hydrochloric acid in the medium was studied between 0.08 and 0.60 mol L -1 . Figure 6 presents the results obtained. It is observed that acid solutions of concentrations between 0.20 and 0.40 mol L -1 enabled

Amount of ascorbic acid
Ascorbic acid was used to prevent the release of iodine by any oxidizing species present in the medium. Thus, the introduction of ascorbic acid prevents the conversion of Ito I 2 , which would decrease the amount of complex formed, besides affecting the color of the medium. The reducing agent concentration was studied between values 5.0 × 10 -3 and 5.5 × 10 -2 mol L -1 . According to Figure 7, the best results were obtained when acid solutions of 1.5 and 4.5 mol L -1 were used. Then, 2.0 mol L -1 HCl solutions were used in subsequent experiments.

Amount of iodide
The amount of iodide is a critical condition because it directly influences the quantity of the extracted species. The anion should be added to the extraction medium to ensure the production of a significant amount of ternary complex formed between Sb III , iodide, and 5-Br-PADAP. Potassium iodide was the salt used. The effect of iodide concentration was studied in 0.00 to 0.48 mol L -1 . The addition of a 0.24 mol L −1 potassium iodide solution was sufficient to obtain the maximum signal. A 0.25 mol L −1 iodide solution was then selected. The results are shown in Figure 8.

Selectivity
The possibility of interference caused by some ions in the preconcentration was evaluated using 30.0 μg L -1 Sb III solutions. Each species was added to these solutions, and the analytical signal was compared with those corresponding to the solution signal containing only antimony. A species was considered interfering when the analytical signal in its presence varies ± 10% about the absorbance of the element. The results are shown in Table 1. It is observed that the method is not affected by the presence of the species in the tested quantities.

Analytical features
The analytical performance of the method was evaluated under optimized conditions. Linear calibration ranges,    Co II , Hg II 50 Se IV calibration curve equation, limit of detection (LOD), and limit of quantification (LOQ) were calculated. The calibration curves were constructed to determine Sb III at a 2.5 to 50.0 μg L -1 . The red channel was processed to obtain the analytical signals due to higher sensitivity than the green and blue channels. The calibration curve equation was obtained using the linear regression method, and it is presented in Table 2, where A is the absorbance signal, and C is the Sb III concentration in μg L -1 . The limits of detection and quantification were calculated using the expressions LOD = 3δ/s and LOQ = 10δ/s, respectively, where δ is the standard deviation of eleven measurements of analytical blank and s is the slope of the analytical curve. The method's precision was calculated using a relative standard deviation (RSD) of eight 30.0 μg L −1 Sb III solution measurements. The enrichment factor (EF) was obtained by calculating the ratio between the angular coefficient of the analytical curve and that relative to the direct measurement. 55,56 Application The method developed was applied to determine antimony in river water and mineral water samples. The results are shown in Tables 3 and 4. The following equation was used for the calculation of recovery: , where C is the concentration of Sb III found without addition, C 0 is the concentration of Sb III found with addition, and m is the amount of Sb III added. The recovery values for the samples ranged from 92 to 104%, demonstrating the proposed applicability of the method. The t-test showed that the results obtained with the proposed methodology showed no significant difference at 95%. Table 5 compares the results obtained for the determination of total Sb using electrothermal atomic absorption spectrometry (ET AAS) and the proposed method. The values found by the proposed method are in agreement with those found using ET AAS.

Conclusions
The proposed method presented a simple, low-cost, and fast approach for the extraction and determination of VA-IL-LPME: vortex-assisted ionic liquid-based liquid-phase microextraction; A: absorbance; C: concentration. the concentration of Sb III and Sb V . The procedure made it possible to determine trace species in mineral water and river water samples. The method is based on the principles of green chemistry, as it does not employ toxic or persistent solvents. On the contrary, the acceptor phase used is an ionic liquid, a solvent with several characteristics that differentiate it from conventional solvents. Among these properties, low toxicity, volatility, and vapor pressure can be cited as the ability to extract several species. The procedure also has the advantage of using small amounts of reagents and solvents, reducing waste generation. The association of VA-IL-LPME with DIC resulted in a method whose benefits make it an excellent alternative to determine antimony species in water samples. Other approaches to Sb speciation can be tested, such as inserting operations using flow injection or different types of LPME, such as SDME or headspace single-drop microextraction (HS-SDME). Work in this direction is being developed in this laboratory.