Straightforward silicon determination in water-in-oil-in-water emulsions used for silicon supplementations in food by high-resolution continuum source flame atomic absorption spectrometry

Spectrochimica Acta Part B 148 (2018) 44–50

Contents lists available at ScienceDirect

Spectrochimica Acta Part B journal homepage: www.elsevier.com/locate/sab

Straightforward silicon determination in water-in-oil-in-water emulsions used for silicon supplementations in food by high-resolution continuum source flame atomic absorption spectrometry☆

T

⁎

Beatriz Gómez-Nietoa, Mª Jesús Gismeraa, , Mª Teresa Sevillaa, Susana Cofradesb, María Freireb, Jesús R. Procopioa a

Departamento de Química Analítica y Análisis Instrumental, Facultad de Ciencias, Avda. Francisco Tomás y Valiente, 7, Universidad Autónoma de Madrid, 28049 Madrid, Spain Instituto de Ciencia y Tecnología de Alimentos y Nutrición (ICTAN-CSIC), C/ José Antonio Novais, 10, 28040 Madrid, Spain

b

A R T I C LE I N FO

A B S T R A C T

Keywords: Silicon determination Double emulsion Straightforward analysis High-resolution continuum source flame AAS Discontinuous sample introduction

Water-in-oil-in-water (W1/O/W2) double emulsions are complex liquid dispersions employed to entrap, protect and control the release of different substances such as minerals in cosmetics, pharmaceuticals, and food. The present paper proposes a simple and fast analytical procedure for silicon determination in double emulsion samples formulated for the supplementation of silicon in different food products. Silicon determination was performed by flame atomic absorption spectrometry (FAAS) at the main analytical line (251.611 nm) using a high-resolution continuum source instrument. The introduction of the sample in the spectrometer in a continuous or discontinuous mode and the influence of double emulsion matrix on silicon absorbance signals were investigated. At the optimized conditions, the double emulsion samples were analysed using silicon standards in ultrapure water for calibration. The limits of detection (LODs) were 0.04 and 0.11 mg L−1 and the upper limits of silicon linear working ranges were up to 23 and 70 mg L−1 for continuous and discontinuous sample introduction modes, respectively. Silicon spiked double emulsion samples were analysed for validation purposes. The good recoveries (within 95–105%) demonstrated the suitability of the proposed method.

1. Introduction Water-in-oil-in-water (W1/O/W2) emulsions are complex multiphase liquid systems in which aqueous droplets are dispersed in larger oil drops that are in turn dispersed within a second aqueous continuous phase. In this way, these compartmentalized systems permit to encapsulate, protect and deliver at a controlled rate water-soluble chemical substances such as drugs, vitamins, active components, nutrients, or minerals initially entrapped in the internal aqueous phase. Due to these characteristics, double emulsions are employed in cosmetics, pharmaceutical and food industries to incorporate these substances and improve the properties of their products [1–6]. Silicon is an element widely used for diverse industrial applications including its use as additive in food and beverages. Silicon is not considered nowadays as an essential element for humans because its exact biochemical function is still unclear. However, it is suspected that this metalloid plays a key role in the development of bones, cartilages, and other connective tissues and its deficiency could produce aging of hair, ☆ ⁎

nails, and skin. In addition, recent findings indicate that silicon may have a modulating effect on the immune and inflammatory response, and it has been associated with mental health. Due to the benefits of this element, silicon supplementations in food, cosmetics, and pharmaceuticals have drawn an increasing interest in the last decade [7–11]. In this regard, double emulsions can be an adequate option to encapsulate bioavailable and water-soluble silicon forms and develop silicon delivery systems. For example, different W1/O/W2 emulsions have recently been investigated to evaluate their possible use as potential ingredients for the development of healthier foods [3]. To study the encapsulation processes and ensure the quality criteria of new products that incorporate silicon in these double emulsions, adequate analytical methods for accurate and precise silicon determination are required. Silicon determination is considered an important challenging task, mainly when samples with complex organic matrices such as oils, pharmaceutical formulations, or food products must be analysed. Analytical methodologies based on atomic techniques, neutron

Selected Paper from the Colloquium Spectroscopicum Internationale XL (CSI 2017), Pisa, Italy, 11–16 June 2017. Corresponding author. E-mail address: [email protected] (M.J. Gismera).

https://doi.org/10.1016/j.sab.2018.06.001 Received 27 October 2017; Received in revised form 29 May 2018; Accepted 4 June 2018 Available online 05 June 2018 0584-8547/ © 2018 Elsevier B.V. All rights reserved.

Spectrochimica Acta Part B 148 (2018) 44–50

B. Gómez-Nieto et al.

activation analysis, nuclear magnetic resonance, and X-ray fluorescence have been applied for silicon determination in this kind of samples. [7, 12]. Among them, atomic spectrometry techniques such as inductively coupled plasma-mass spectrometry (ICP-MS), ICP optical emission spectrometry (ICP-OES) and atomic absorption spectrometry (AAS) with flame or graphite furnace atomizers are the most commonly employed to carry out the analysis after an adequate pre-treatment of the samples [12, 13]. In these atomic spectrometry techniques, the analytical response of silicon is highly dependent on the chemical form of the analyte and the nature of the sample [12, 14, 15]. The determination of metals and metalloids in organic matrices by ICP-MS or ICP-OES can mainly be affected by changes on the plasma characteristics, the formation of carbon deposits and the appearance of spectral interferences [16, 17]. In fact, methods based on ICP-MS are highly affected by spectral interferences due to polyatomic ions produced by the presence of oxygen, carbon and nitrogen (e.g. 12C16O+, 14 N2+, 14N16O+) that can influence the accurate determination of silicon isotopes [18]. Although methods based on AAS with flame or graphite furnace atomizers usually present a higher tolerance to organic matrices, they also show disadvantages in silicon determination, some of them similar to those observed using ICP-MS or ICP-OES methods. For example, an important limitation for the silicon determination in AAS using graphite furnace atomization (GFAAS) is the formation of volatile compounds such as SiO and/or refractory silicon carbides (SiC) so adequate modifiers are required to avoid an inaccurate determination [19–22]. Additional difficulties can be found using conventional line-source AAS instruments due to radiation scattering effects, that are more frequent at the short wavelengths of the silicon resonance lines (251.611/251.432 nm), as well as by spectral interferences [23]. In flame atomic absorption spectrometry (FAAS), the formation and occurrence of diatomic molecules with rotational fine structure such as PO, CS, or NO, and their molecular spectra are due to the atomizer (the flame composition) and the matrix constituents of the sample or standard introduced in the spectrometer [20, 24]. For the silicon determination, a spectral interference due to the CS molecule formed in the flame has been observed in the analysis of samples containing S compounds. The presence of vanadium, cobalt, iron, and/or tungsten at high concentrations can cause spectral interferences since some low sensitivity absorption lines of these elements are located close to the silicon lines at 251.611 nm and 251.432 nm [25]. Moreover, the absorbance signal registered for silicon determination using a conventional line-source AAS spectrometer is typically a combined signal from the silicon lines at 251.611 nm and 251.432 nm [20]. Using a highresolution continuum source FAAS (HR-CS FAAS) spectrometer, these drawbacks can be minimized or eliminated. The excellent resolution provided by this spectrometer permitted to use the silicon absorbance signal at 251.611 nm separated from the silicon signal at 251.432 nm, and ensure the absence of spectral interferences due to the presence of high amounts of other elements such as vanadium or iron [25]. In addition, the spectral interference due to CS molecule was effectively corrected by the least-squares background correction procedure available in the software of this instrument [25, 26]. Regardless of the atomic spectrometry technique used for silicon determination, previous treatments of the sample for total matrix decomposition and minimization of spectral interferences are usually required. These sample treatments present important drawbacks derived from the possible loss of the analyte in highly volatile forms and the risk of contamination due to the ubiquity of this element [7, 12]. Indeed, in an interlaboratory study carried out about silicon determination in biological matrices by spectrometric techniques the main problems arose when the sample digestion pre-treatments were needed prior to the analysis [13]. Due to these problems, methods based on direct analysis of the sample are attractive approaches for silicon determination. Direct solid sampling has been employed for silicon determination by GFAAS in polyamide [22], hardly soluble oxides [27], or plant materials [28]. Because methods based on FAAS are simple, widely

used and relatively low cost, the possibility of carrying out the analysis without a previous sample digestion or only with a very simple sample treatment is a very interesting issue. Analytical methodologies based on FAAS have been used and reported for the quantification of silicon. For instance, we have developed a straightforward sequential method based on HR-CS FAAS for the determination of copper, zinc, manganese, magnesium and silicon in beverages, herbal infusions and dietary supplements [29]. Different approaches to minimize the sample treatment have been proposed for silicon determination in oily samples. Oliveira et al. [30] reported the direct determination of silicon in vegetable oils and biodiesel samples by HR-CS FAAS after the dilution with xylene and using standards prepared in xylene for calibration. As an alternative to the sample dilution in organic solvents, the preparation of oil-in-water emulsions has been successfully used as sample pretreatment to determine magnesium, chromium, copper, lead, nickel, aluminium and silicon in lubricating oils by FAAS [31]. In general, the main advantages attributed to the use of emulsified systems as a sample preparation procedure in analytical chemistry are that they do not require the previous destruction of the sample matrix or the use of large amounts of organic solvents as diluents [32, 33]. Compared to the other sample pre-treatments, the sample handling and the reagent addition are minimized in the preparation of emulsions, decreasing the risk of sample contamination. Moreover, an emulsified system with an adequate stability and a similar viscosity to the observed for aqueous solutions can be achieved optimizing the preparation procedure. These are very important characteristics to analyse samples with complex organic matrices by spectrometric methods since the calibration can be performed using aqueous standards. Although, emulsified systems have been successfully applied to sample preparation of petroleum-based products, foodstuffs and cosmetics for metal determination by atomic spectrometry techniques [32, 33] few works have been devoted to silicon determination [21, 31, 34]. Taking into account the applicability of silicon encapsulation in W1/ O/W2 emulsions and the problems for silicon determination in complex matrices, the aim of this work is to optimize a simple and fast FAAS methodology for the determination of silicon in W1/O/W2 emulsions formulated as potential ingredients to obtain healthier food products. In a previous paper [3], we studied the physicochemical properties and the encapsulation of silicon in these double emulsions. In that work, the silicon encapsulation efficiency in the different prepared double emulsions was evaluated in an indirect mode by measuring the released silicon concentration in the outer aqueous phase. For this purpose, after the preparation of the double emulsions, they were centrifuged to separate the fat globules from the outer aqueous phase. The encapsulation efficiency was estimated from the initial silicon amount added to the inner aqueous phase and the silicon content determined in the outer aqueous phase. In this work, our objective is to design a FAAS method for silicon determination in these double emulsion samples directly or with a minimal sample treatment. To achieve this purpose, the measurement parameters and sample introduction conditions were optimized. In addition, the influence of the double emulsion matrix on the absorbance signals of silicon was investigated. The feasibility to perform the calibration with standards prepared in ultrapure water was evaluated and the principal analytical parameters were calculated. A HR-CS FAAS instrument was employed in this work because the remarkable instrumental and signal processing software advantages of this equipment permit to get excellent analytical performances for silicon determination in complex samples. 2. Experimental 2.1. Instrumentation and measurement conditions A contrAA 700 model high-resolution continuum source atomic absorption spectrometer (Analytik Jena, Germany) with a flame atomizer system was used for silicon measurements. This instrument 45

Spectrochimica Acta Part B 148 (2018) 44–50

B. Gómez-Nieto et al.

prepared daily, by the adequate dilution of the 1000 mg L−1 silicon stock solution in ultrapure water or in double emulsion blanks prepared as it is described in Section 2.3. To ensure homogeneity, silicon standards in ultrapure water or in double emulsions were thoroughly shaken by manual agitation prior to place them into the autosampler plastic tubes for the silicon absorbance measurement. All glassware and plastic containers used to prepare and/or store samples and standards were cleaned with 10% (v/v) HNO3, for at least 24 h, and abundantly rinsed with ultrapure water before their use.

comprises a high-intensity xenon short-arc lamp operating in “hot-spot mode” as the continuum radiation source and a high-resolution double monochromator consisting in a prism pre-monochromator and an echelle grating monochromator. The detector is a charge-couple device (CCD) array with 588 pixels, 200 of which are used for analytical purposes. The spectrometer was operated using the Aspect CS 2.1.2.0 software (Analytik Jena, Germany). The main analytical line for silicon at 251.611 nm was employed for the analysis. High-purity nitrous oxide 99.998% and high-purity acetylene 99.6% (Carburos Metálicos, Spain) were used as oxidant and fuel gases, respectively, in the employed flame for silicon determination. The burner head incorporates a scraper accessory to automatically remove, before each measurement, the carbonaceous residues from the 50 mm burner slot. The instrument is also equipped with an AS 52S autosampler and a SFS 6 injection module (both from Analytik Jena, Germany) for the sample aspiration. With the segmented flow star SFS 6 injection module the flame is more stable since a carrier solution of ultrapure water is introduced avoiding aspiration of air when standards or sample solutions are changed. This accessory also permits the continuous aspiration of the sample or the introduction of discrete sample volumes during the measurement depending on the operating conditions stablished by the software, so steady-stable or transient signals are obtained, respectively. The aspiration rate was fixed at 10.0 mL min−1 and the burner height and flame stoichiometry were optimized. The peak volume selected absorbance (PVSA), obtained from the sum of the individual integrated absorbance values of designated pixels, was used for signal evaluation in the discontinuous sample introduction mode. The wavelength-selected absorbance (WSA), obtained from the sum of the individual absorbance values of designated pixels, was employed for quantification of steadystate signals in the continuous sample introduction mode. All measurements were carried out, at least by triplicate, using three pixels, the central pixel (CP) plus the two adjacent ones (CP ± 1). The selected measurement conditions for silicon determination are included in Table 1.

2.3. Preparation of W1/O/W2 emulsions The double emulsions samples used in this work were specifically designed to be employed as potential ingredients for the development of healthier food products. Three types of W1/O/W2 emulsions with a different composition of the inner aqueous phase (W1) were formulated: a double emulsion with NaCl (denoted as DE-1), a double emulsion with NaCl plus sodium caseinate (DE-2), and a double emulsion with NaCl plus gelatin (DE-3). In order to prevent microbial growth, sodium azide was also incorporated in the inner aqueous phase. Extra Virgin olive oil (Carbonell, SOS Cuétara S.A, Spain) was used as the lipid phase (O). This oil meets the quality requirements and specifications for this type of vegetable oil [35, 36]. A two-stage emulsification procedure was used to prepare the double emulsions. Firstly, primary water-in-oil (W1/O) emulsions were obtained by mixing the different inner aqueous phases with the lipid phase containing olive oil and polyglycerol polyricinoleate as lipophilic surfactant. Then, the primary emulsion was gradually added to the outer aqueous phase (W2) to form the double emulsion (W1/O/W2) [3]. In this way, W1/O/W2 emulsions with a final olive oil content of 30% were prepared. A commercial silicon rich aqueous plant extract obtained from nettles was used as a natural and bioavailable silicon source. In the label of this product, it was indicated that the plant extract contains about 4800 mg of nettles in 40 mL, but the silicon content was not specified. For this reason, the lot of the employed commercial silicon plant extract to prepare the different double emulsion samples of this work was analysed, and a silicon content of 373 ± 9 mg L−1 was found. W1/O/W2 emulsions without silicon (double emulsion blanks) and W1/O/W2 emulsions obtained after the incorporation of the silicon plant extract in the inner aqueous phase for its encapsulation (double emulsion samples), were prepared using the different inner aqueous phase compositions (DE-1, DE-2 and DE-3), previously described. All double emulsions (different blanks and samples) were stored at 3 °C in a refrigerator. Prior to sampling, they were equilibrated at room temperature and homogenized in a vortex mixer. For the analysis of the samples, an adequate volume of each type of sample (DE-1 double emulsion samples, DE-2 double emulsion samples or DE-3 double emulsion samples) was transferred to a volumetric flask, and diluted to 10.0 mL with ultrapure water. These double emulsion samples were analysed by FAAS without further treatment.

2.2. Reagents and standards The reagents and solvents used in this work were at least of analytical grade. Ultrapure water (resistivity 18.2 MΩ cm) obtained from an Ultra Clear™ TWF UV EDI water purification system (Siemens, Germany) was employed to prepare the standard solutions in ultrapure water and dilute the samples. Cold-soluble gelatin (Tradissimo Trades, Spain), NaCl (Panreac Química S.A., Spain), sodium caseinate (Friesland Campina DMV, The Netherlands), sodium azide (SigmaAldrich, Spain), and polyglycerol polyricinoleate (PGPR) (Bavaro Chemicals, Spain) were used to prepare double emulsions. A commercially available TraceCert® 1000 mg L−1 silicon stock solution for AAS in 2% sodium hydroxide, prepared with high purity silicon metal (Sigma-Aldrich, Spain) was used to obtain silicon standards. They were

3. Results and discussion

Table 1 Optimal instrumental parameters for silicon determination by HR-CS FAAS.

3.1. Optimization of the HR-CS FAAS instrumental parameters

Parameters Wavelength Relative sensitivity C2H2 flow rate N2O flow rate Burner height Pixels for quantification Injection time Measurement time Delay time Wash time a b

The most sensitive analytical line of silicon located at 251.611 nm was selected to develop the methodology. The high temperature nitrous oxide-acetylene flame was employed to generate relatively free Si atoms and avoid the formation and electronic transition of SiO molecules in the flame [20, 25]. Silicon standards in ultrapure water were used to optimize the instrumental parameters. The influence of burner height and flame composition on the absorbance signal was studied and their optimal values are indicated in Table 1. To obtain adequate analytical properties the injection and measurement times for both continuous and discontinuous sample introduction modes were optimized. An injection time of 3 s and a measurement time of 3 s were chosen for the continuous sample introduction analysis. For the

251.611 nm 100% 270 L h−1 376 L h−1 7 mm 3 (CP ± 1) 3.0 sa 1.0 sb 3.0 sa 7.0 sb 12.0 s 20.0 s

Continuous sample introduction mode. Discontinuous sample introduction mode. 46

Spectrochimica Acta Part B 148 (2018) 44–50

B. Gómez-Nieto et al.

0.030

0.08

0.025

0.06

Absorbance

Absorbance

0.020 0.015 0.010

0.04

0.02 0.005 0.000

0.00 0

1

2

3

4

5

6

7

A

8

B

C

D

−1

Fig. 2. Absorbance values for 10.0 mg L silicon standard (n = 3) in ultrapure water (A) and in double emulsions with a different composition of the inner aqueous phase: DE-1 (B), DE-2 (C) and DE-3 (D) using the continuous sample introduction mode at the instrumental conditions shown in Table 1. Error bars represent standard deviation for three replicates.

Time / s −1

Fig. 1. Overlaid transient signals for a 10.0 mg L silicon standard in ultrapure water at 0.5 s (continuous black line), 1.0 s (dashed black line), 2.0 s (dotted black line), 3.0 s (continuous red line) and 4.0 s (dashed red line) injection times using the discontinuous sample introduction mode. Other instrumental conditions according to Table 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

the sample transport to the flame or the atomization process. Spectral interferences can be detected and corrected with the HR-CS AAS spectrophotometer since this instrument allows the simultaneous monitoring of a very close spectral region around the analytical line [23–26]. In the wavelength resolved time-integrated absorbance spectra registered at the vicinity of 251.611 nm for silicon standards prepared in ultrapure water and in double emulsions, spectral background structures and other atomic absorption lines were not found. Therefore, the increase of absorbance in the silicon standards prepared in double emulsions could not be attributed to the overlap of the silicon main line with a molecular spectrum and/or an absorption line of other element. According to this, other factors could be responsible of the observed differences in the absorbance values. Analytical signals in FAAS are influenced by experimental variables such as sample aspiration rate and nebulization efficiency that can be modified by the characteristics of the sample. For example, large amounts of organic substances such as solvents, oils, or surfactants in the samples can cause important changes in viscosity, density, and surface tension, affecting the aspiration rate and the nebulization process and consequently, the amount of sample that reaches the flame. To study the effect of double emulsion matrix on the sample transport, silicon standards prepared in ultrapure water and double emulsions were measured in continuous and discontinuous sample introduction modes. As can be seen in Fig. S1 (Appendix) the absorbance for the standard in ultrapure water was lower than the signal for the silicon standard in double emulsion using the continuous sample introduction mode, whereas an opposite behaviour was observed in the discontinuous mode. This decrease in the signal for the standard in double emulsion using discontinuous sampling could indicate a diminution of sample deliver to the flame although the same injection time was used for both silicon standards. This result could suggest that the double emulsion matrix had a negative influence on the sample transport and the aspiration rate, although the results in the continuous sample introduction mode apparently contradicted this assumption. These facts seemed to indicate that different processes associated to the characteristics of the double emulsion matrix concurred and influenced on the measured signals. To obtain additional information about the effects of double emulsion matrix on these measurements, aspiration rates and absorbance values were measured for silicon standards prepared in double emulsion with different dilution factors, expressed as volume of double emulsion to total volume (sum of double emulsion

discontinuous mode, different injection times from 0.5 s to 4.0 s were evaluated in order to obtain transient absorbance signals with an almost Gaussian peak profile and an adequate precision in replicate measurements. Fig. 1 shows the overlaid transient signals for a 10.0 mg L−1 silicon standard in ultrapure water at different injection times. For injection times from 0.5 s to 2.0 s, both the peak height and peak wide increased. Similar peak heights were observed for injection times longer than 2.0 s, but the peak wide increased, the almost Gaussian peak profile was lost, and the measurement time required to evaluate these transient signals must be increased. In addition, the sample consumption also increased. For example, the sample volume needed to carry out the determination in triplicate using 1.0 s as injection time was about 2–3 times lower than the volume required for an injection time of 4.0 s. According to these results, we selected as optimal conditions 1.0 s and 7.0 s as the injection and measurement times, respectively, in the discontinuous sample introduction mode. The influence of the number of pixels used for the quantification on the sensitivity and limit of detection (LOD) of silicon determination was also studied. For both modes of sample introduction, the best LODs were obtained using 3 pixels (CP ± 1) (Table S1 Appendix). The use of more pixels contributed to moderately increase the sensitivity but, the noise level also increased and the LODs values were deteriorated. For these reasons, 3 pixels (CP ± 1) were chosen for quantification purposes. 3.2. Influence of double emulsion composition on silicon absorbance To evaluate the influence of sample matrix on the silicon absorbance values, standards containing 10.0 mg L−1 of silicon were prepared in the three double emulsions (DE-1, DE-2 or DE-3) studied in this work. As it can be seen in Fig. 2, the absorbance values were similar for standards in the three double emulsions. Therefore, it can be deduced that the different composition of the inner aqueous phase of DE-1, DE-2 and DE-3 double emulsions (NaCl, NaCl plus sodium caseinate, or NaCl plus gelatin, respectively) did not significantly influence on silicon absorbance. However, absorbance values for 10.0 mg L−1 silicon standards prepared in the double emulsion were higher than the observed for the silicon standard prepared in ultrapure water. This fact could indicate that the double emulsion matrix influenced on the measurement process, due to spectral interferences or other factors associated to 47

Spectrochimica Acta Part B 148 (2018) 44–50

0.06

0.14

0.14

10

Discontinuous introduction

0.04 6 0.03 4 0.02 2

0.01

Integrated absorbance / s

8

Continuous introduction

0.10

0.10

0.08

0.08

0.06

0.06

0.04

0.04

0.02

0.02

0.00

0.00 0.5

0.00 0.8

0.6

0.4

0.2

1.0

2.0

2.5

3.0

4.0

3.0

Injection time /s

0 1.0

0.12

Absorbance

0.12

0.05

Aspiration rate / mL min-1

Absorbance or Integrated absorbance /s

B. Gómez-Nieto et al.

0.0

Fig. 4. Absorbance signals for 10 mg L−1 silicon standards (n = 3) in ultrapure water (red), undiluted DE-1 double emulsion (yellow) and DE-1 double emulsion with a 0.8 dilution factor (grey) using the discontinuous sample introduction mode at different injection times and the continuous sample introduction mode at 3 s injection time. Other instrumental conditions according to Table 1. Error bars represent standard deviation for three replicates. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Dilution factor of the DE-1 double emulsion Fig. 3. Aspiration rate (red circles) and absorbance signals of 10.0 mg L−1 silicon standards in ultrapure water (0.0 dilution factor), DE-1 double emulsion (1.0 dilution factor) and different DE-1 double emulsion dilutions in ultrapure water using the continuous (black squares) and discontinuous (blue triangles) sample introduction mode at the instrumental conditions shown in Table 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

In order to obtain additional information about the influence of double emulsion matrix on the silicon signal, standards (10.0 mg L−1) prepared in ultrapure water, undiluted double emulsion and diluted double emulsion with a dilution factor of 0.8 were aspired and measured in the discontinuous sample introduction mode using different injection times. According to that observed in Fig. 1, an increase of the signal was expected when the injection time increased. Moreover, using very long injection times the absorbance signal reached a maximum value due to a pseudo-continuous supply of sample, and the signal behaviour could be resembled to that obtained using the continuous sample introduction mode. In Fig. 4 are presented the integrated absorbance values at different injection times for silicon standards in ultrapure water and in undiluted and diluted double emulsion matrix. For comparison, the obtained signals in the continuous mode for these standards are also included in this figure. In the discontinuous mode, the signals increased as the injection time increased for all the silicon standards. For short injection times, the signals in double emulsion matrixes were lower than those obtained for standards in ultrapure water, being more pronounced the difference for the standards prepared in undiluted double emulsion. However, for injection times of 3.0 s or 4.0 s the signals for the standards in double emulsion were higher than those shown for the standards in ultrapure water. This behaviour was similar to that observed for these standards using the continuous introduction mode.

volume + diluent volume). These silicon standards in diluted double emulsion were prepared by mixing volumes of the DE-1 double emulsion with the adequate amount of concentrated silicon standard in ultrapure water and making up to the total volume with ultrapure water. The absorbance signal values and aspiration rates for dilution factors between 1.0 (undiluted DE-1 double emulsion) to 0.0 (no containing double emulsion) are shown in Fig. 3. As it can be seen in the figure, the aspiration rate gradually increased when the double emulsion dilution factor value decreased. In fact, the aspiration flow rate was approximately reduced to 80% (from 10 mL min−1 to 2 mL min−1) when measurements were carried out using a silicon standard in ultrapure water (dilution factor 0.0) or in undiluted DE-1 double emulsion (dilution factor 1.0). For both sample introduction modes, no significant variations in the signal values were found for prepared silicon standards in double emulsion with dilution factors between 0.4 and 0.0. In the discontinuous introduction mode, the absorbance signal diminished when the dilution factor increased from 0.4 to 1.0. Indeed, the integrated absorbance for a silicon standard in undiluted double emulsion (dilution factor 1.0) was about 40% lower than the value for a silicon standard in diluted double emulsion with a dilution factor between 0.4 and 0.0. A different behaviour was found using the continuous sample introduction mode. In this case, a progressive increase of the absorbance was observed when the silicon standards were prepared in double emulsion with a dilution factor higher than 0.4. These results could indicate that changes in the atomization efficiency were produced when a high amount of double emulsion matrix was introduced into the nebulization chamber and reached the flame. The variations of absorbance signals reported for FAAS measurements carried out in organic solvents have been attributed to modifications of the aspiration rate, the efficiency of nebulization and the atomization process due to changes of the chemical composition of the flame owing to the characteristics of the matrix [37]. Taking into account the high oil content in the double emulsions, the observed changes on the absorbance signals for silicon standards prepared in these media with different dilution factors could be explained by processes induced by the matrix during the measurement. The results for silicon standards in double emulsions with dilution factors between 1.0 and 0.4, seemed to indicate that some of these parameters, the aspiration rate, the nebulization efficiency, and the atomization process, affect mostly to the signals.

3.3. Calibration studies and figures of merit The feasibility of performing the calibration procedure using the external calibration method with silicon standards in ultrapure water was evaluated. For this purpose, the slopes of the obtained calibration curves for silicon standards prepared in ultrapure water and double emulsion with different dilution factors using the continuous and discontinuous sample introduction modes were compared. For both sample introduction modes, the slopes for the obtained calibration curves using standards in ultrapure water (0.00317 ± 0.00005 L mg−1 and 0.00233 ± 0.00004 s L mg−1, for continuous and discontinuous modes, respectively) were not significant different from those obtained for the calibration curves prepared in double emulsion with dilution factors lower or equal than 0.4 (0.00310 ± 0.00004 L mg−1 and 0.00238 ± 0.00009 s L mg−1 in continuous and discontinuous modes, 48

Spectrochimica Acta Part B 148 (2018) 44–50

B. Gómez-Nieto et al.

methodologies developed for the analysis of complex samples the standards used for calibration purposes must be prepared in simulated and similar matrixes to those of the samples for achieving optimal results. For instance, in a method for the silicon determination in vegetal oil and biodiesel samples by HR-CS FAAS the calibration was performed using silicon standards prepared in mineral oil matrixes diluted with xylene [30]. In addition, our method did not require the previous digestion of the sample and can directly be applied with a simple dilution of the sample in ultrapure water, unlike other HR-CS FAAS methods for silicon determination in complex samples that needed to carry out an acid digestion of the samples and use the internal standardization for calibration [26].

Table 2 Figures of merit for silicon determination by HR-CS FAAS. Parameters

a

Sensitivity r2 Upper limit of linear range (mg L−1) LOD (mg L−1) LOQ (mg L−1) Repeatability (RSD, %)b Reproducibility (RSD, %)b

Sample introduction mode Continuous

Discontinuous

0.00317 ± 0.00005 0.9951 23

0.00233 ± 0.00004 0.9959 70

0.04 0.15 < 4.5% < 7.2%

0.11 0.36 < 7.0% < 11%

a L mg−1 for continuous and s L mg−1 for discontinuous sample introduction modes. b (n = 3).

3.4. Analysis of silicon in double emulsion samples The proposed method was applied for the silicon determination in double emulsion samples specifically formulated to be used as potential ingredients for the supplementation of silicon in different food products. Three types of double emulsion samples (three lots for each type of double emulsion sample) with different composition of the inner aqueous phase (DE-1 with NaCl, DE-2 with NaCl plus sodium caseinate and DE-3 with NaCl plus gelatin) and the same amount of the commercial silicon rich plant extract were prepared according to the procedure indicated in the experimental section. Prior to the analysis and being in mind the high expected content of silicon in these samples, the double emulsion samples were diluted with ultrapure water up to a 0.1 dilution factor. In Table 3 are shown the results (mean and standard deviation) for the different types of double emulsion samples analysed using the continuous and discontinuous sample introduction modes. As it can be seen in this table and in spite of the different composition of the inner aqueous phase, similar silicon contents were found for all the analysed samples. These results were in accordance with those reported in a previous work about these double emulsion samples where it was observed that the efficiency of silicon encapsulation was not affected by the composition of the inner aqueous phase [3]. In addition, we did not find significant differences between the obtained values of silicon in this work and the expected silicon concentrations in these samples calculated taking into account the results obtained in that previous work (the efficiency of silicon encapsulation, the preparation procedure of these double emulsions and the silicon amount of the commercial plant extract incorporated in the inner phase of the double emulsion samples) [3]. In order to validate the methodology, the double emulsion samples were spiked at different concentrations of silicon. As it can be seen in Table 3 adequate recoveries with values ranging from 95 to 105% were found for both sample introduction modes. These results suggested that the proposed method can be considered a rapid, adequate, and quasi-direct methodology that can successfully be applied for silicon determination in double emulsion samples.

respectively, for 0.4 dilution factor). These results demonstrated that silicon can directly be quantified in the W1/O/W2 emulsion samples by the adequate dilution of the sample with ultrapure water up to a dilution factor ≤0.4 using the external calibration method with silicon standards prepared in ultrapure water and any of the two sample introduction modes. The analytical parameters for the proposed method such as sensitivity, limit of detection (LOD), limit of quantification (LOQ), linear working range, and precision obtained using silicon standards prepared in ultrapure water for both, the continuous and discontinuous sample introduction modes, are summarized in Table 2. The upper limit of the linear working range was graphically calculated as the concentration at which a deviation of a ± 5% around the ideal linearity plot window was found. The linear working ranges for silicon determination were within 0.15 to 23 mg L−1 for continuous and up to 70 mg L−1 in discontinuous sample introduction modes, respectively. The correlation coefficient (r2) of the linear least square fit of the calibration points was higher than 0.995 for both sample introduction modes. The LOD and LOQ were calculated as three and ten times, respectively, the standard deviation of ten measurements of blank solutions divided by the slope of the straight-line calibration plot. The LOD and LOQ values were better or similar to those previously reported for silicon determination in aqueous and/or organic media using HR-CS FAAS with continuous and/or discontinuous sample introduction modes [25, 29, 30], and higher than the one reported by Raposo et al. [26]. It should be noted that in that work the reported LOD was obtained using an internal standard calibration procedure and tungsten as the internal standard. The precision of the method, expressed as relative standard deviation (RSD), was evaluated using silicon standards prepared in ultrapure water and diluted double emulsion with a 0.4 dilution factor. To estimate repeatability, the absorbance of silicon standard solutions at two different concentrations (2.0 and 10.0 mg L−1) were measured in triplicate (n = 3), whereas measurements of these solutions were carried out in three different days (n = 3) to study reproducibility. Satisfactory RDS values were found employing both sample introduction modes. The best precision was obtained for the continuous sample introduction mode with RSD values lower than 4.5% and 7.2% for repeatability and reproducibility, respectively using silicon standards in ultrapure water, and 6.9% for repeatability and 9.6% for reproducibility with standards prepared in the diluted double emulsion. Using the discontinuous sample introduction mode, RSD values slightly higher were obtained (< 7.0% and < 11% for repeatability and reproducibility, respectively in ultrapure water, and < 8.5% and < 15% for repeatability and reproducibility, respectively with standards prepared in the diluted double emulsion). An important advantage of the here proposed method was the possibility of performing the calibration by the external calibration method using standards prepared in ultrapure water. This fact was a very noticeable characteristic because in the most of the analytical

Table 3 Results (mean ± standard deviation, n = 3) for silicon determination in double emulsion samples. Sample

DE-1 DE-2 DE-2 DE-2 DE-3 DE-3 DE-3

49

Add (mg L−1)

– – 20.0 30.0 – 10.0 20.0

Continuous sample introduction

Discontinuous sample introduction

Found (mg L−1)

Recovery (%)

Found (mg L−1)

Recovery (%)

18.4 ± 0.8 17.6 ± 0.9 37 ± 2 47 ± 2 16.2 ± 0.7 25 ± 1 38 ± 2

– – 98 99 – 95 105

19 18 36 50 17 26 39

– – 95 104 – 96 105

± ± ± ± ± ± ±

1 1 3 3 1 2 3

Spectrochimica Acta Part B 148 (2018) 44–50

B. Gómez-Nieto et al.

4. Conclusion In this work, we present a new method for the determination of silicon in W1/O/W2 emulsions using HR-CS FAAS. The good resolution provided by these spectrometers made possible to ensure the absence of spectral interferences and the use of the silicon absorbance signal at 251.611 nm, separated from the silicon signal at 251.432 nm, for quantification purposes. The corroboration of these findings is difficult to get using conventional line source AAS instruments. The instrumental parameters and measurement conditions for sample introduction in continuous and discontinuous modes were optimized. The figures of merit were calculated for both sample introduction modes. The best LOD was obtained for the continuous sample introduction mode, but the large amount of sample that was aspired, required a more frequent stopping of the sequence of the analysis to clean the system and avoid the clogging of the nebulization system than that needed using the discontinuous sample introduction mode. With the discontinuous sample introduction mode this undesirable effect was minimized, the sample consumption was reduced and the upper limit of the linear range was increased about three times compared to the obtained using the continuous mode. However, the better precision and LOQ were obtained using the continuous sample introduction mode. Therefore, depending on the expected silicon content in the sample and the requirements of the analysis the continuous or discontinuous introduction mode can be selected. The silicon determination in double emulsion samples was carried out at the optimized conditions being only necessary a simple dilution of the sample with ultrapure water. Thus, the accurate and precise quantification of silicon can be achieved in less time and with inferior cost than employing methods that required tedious sample treatments. In our opinion, the successful results obtained in this work indicated that the proposed methodology was a useful and competitive analytical tool for silicon determination in double emulsion systems.

[10]

[11] [12]

[13]

[14]

[15]

[16] [17]

[18] [19] [20] [21]

[22]

[23]

[24]

Acknowledgements

[25]

Beatriz Gómez-Nieto gratefully acknowledges the Spanish Royal Society of Chemistry (RSEQ) for the grant to attend the XL Colloquium Spectroscopicum Internationale (CSI-XL).

[26]

Appendix A. Supplementary data

[27]

Supplementary data to this article can be found online at https:// doi.org/10.1016/j.sab.2018.06.001.

[28]

References

[29]

[1] J. Wang, A. Shi, D. Agyei, Q. Wang, Formulation of water-in-oil-in-water (W/O/W) emulsions containing trans-resveratrol, RSC Adv. 7 (2017) 35917–35927. [2] B. Li, Y. Jiang, F. Liu, Z. Chai, Y. Li, Y. Li, X. Leng, Synergistic effects of whey protein–polysaccharide complexes on the controlled release of lipid-soluble and water-soluble vitamins in W1/O/W2 double emulsion systems, Int. J. Food Sci. Technol. 47 (2012) 248–254. [3] S. Cofrades, R. Bou, B. Gómez-Nieto, J.R. Procopio, A. Errabi, F. JimenezColmenero, Physicochemical properties and encapsulation of silicon in double emulsions for healthier food applications, J. Food Sci. Technol. 53 (2016) 3884–3893. [4] A. Agrawal, S. Kulkarni, S.B. Sharma, Recent advancements and applications of multiple emulsions, Int. J. Adv. Pharm. 4 (2015) 94–102. [5] G. Muschiolik, E. Dickinson, Double emulsions relevant to food systems: preparation, stability, and applications, Compr. Rev. Food Sci. Food Saf. 16 (2017) 532–555. [6] F. Jiménez-Colmenero, Potential applications of multiple emulsions in the development of healthy and functional foods, Food Res. Int. 52 (2013) 64–74. [7] A.V. Hirner, D. Flasbeck, Speciation of silicon (Eds.), in: R. Cornelis, H. Crews, J. Caruso, K.G. Heumann (Eds.), Handbook of Elemental Speciation II: Species in the Environment, Food, Medicine & Occupational Health, John Wiley & Sons, Ltd., 2005, pp. 366–377. [8] K.R. Martin, Silicon: The Health Benefits of a Metalloid, in: A. Sigel, H. Sigel, R.K.O. Sigel (Eds.), Interrelations Between Essential Metal Ions and Human Diseases, Metal Ions in Life Sciences, 13 Springer, 2013, pp. 451–473. [9] A. Barel, M. Calomme, A. Timchenko, K. De Paepe, N. Demeester, V. Rogiers,

[30]

[31] [32] [33]

[34]

[35]

[36]

[37]

50

P. Clarys, D. Vanden Berghe, Effect of oral intake of choline-stabilized orthosilicic acid on skin, nails and hair in women with photodamaged skin, Arch. Dermatol. Res. 297 (2005) 147–153. J.L. Domingo, M. Gomez, M.T. Colomina, Oral silicon supplementation: an effective therapy for preventing oral aluminum absorption and retention in mammals, Nutr. Res. 69 (2011) 41–51. F.H. Nielsen, Update on the possible nutritional importance of silicon, J. Trace Elem. Med. Biol. 28 (2014) 379–382. F. Chainet, Ch.P. Lienemann, M. Courtiade, J. Ponthus, O.F.X. Donard, Silicon speciation by hyphenated techniques for environmental, biological and industrial issues: a review, J. Anal. At. Spectrom. 26 (2011) 30–51. K.V. Dyck, H. Robberecht, R.V. Cauwenbergh, H. Deelstra, J. Arnaud, L. Willemyns, F. Benijts, J.A. Centeno, H. Taylor, M.E. Soares, M.L. Bastos, M.A. Ferreira, P.C. D'Haese, L.V. Lamberts, M. Hoenig, G. Knapp, S.J. Lugowski, L. Moens, J. Riondato, R.V. Grieken, M. Claes, R. Verheyen, L. Clement, M. Uytterhoeven, Spectrometric determination of silicon in food and biological samples: an interlaboratory trial, J. Anal. At. Spectrom. 15 (2000) 735–741. M.F. Gazulla, M. Rodrigo, M. Orduña, M.J. Ventura, C. Andreu, High precision measurement of silicon in naphthas by ICP-OES using isooctane as diluent, Talanta 164 (2017) 563–569. R. Sanchez, J.L. Todoli, C.P. Lienemann, J.M. Mermet, Effect of silicon chemical form on the emission intensity in inductively coupled plasma atomic emission spectrometry for xylene matrices, J. Anal. At. Spectrom. 24 (2009) 391–401. R. Sanchez, C. Sánchez, C.P. Lienemann, J.L. Todolí, Metal and metalloid determination in biodiesel and bioethanol, J. Anal. At. Spectrom. 30 (2015) 64–101. R. Sánchez, J.L. Todolí, C.P. Lienemann, J.M. Mermet, Determination of trace elements in petroleum products by inductively coupled plasma techniques: a critical review, Spectrochim. Acta B 88 (2013) 104–126. T.W. May, R.H. Wiedmeyer, A table of polyatomic interferences in ICP-MS, At. Spectrosc. 19 (1998) 150–155. C.J. Rademeyer, I. Vermaak, Atomization mechanisms of silicon in a graphite furnace atomizer, J. Anal. At. Spectrom. 7 (1992) 347–351. B. Welz, M. Sperling, Atomic Absorption Spectrometry, third ed., Wiley-VCH, Weinheim, 1999. M.S. Luz, P.V. Oliveira, Niobium carbide as permanent modifier for silicon determination in petrochemical products by emulsion-based sampling GF AAS, Fuel 116 (2014) 255–260. M. Resano, M. Aramedia, A.B. Volynsky, M.A. Belarra, Solid sampling-graphite furnace atomic absorption spectrometry for the direct determination of trace amounts of silicon in polyimide. Comparison of performance of platinum and palladium as chemical modifiers, Spectrochim. Acta B 59 (2004) 523–531. M. Resano, M.R. Florez, E. Garcia-Ruiz, Progress in the determination of metalloids and non-metals by means of high-resolution continuum source atomic absorption spectrometry. A critical review, Anal. Bioanal. Chem. 406 (2014) 2239–2259. B. Welz, S. Morés, E. Carasek, M.G.R. Vale, H. Becker-Ross, High-resolution continuum source atomic and molecular absorption spectrometry—a review, Appl. Spectrosc. Rev. 45 (2010) 327–354. Z. Kowalewska, J. Pilarczyk, L. Gosciniak, Spectral aspects of the determination of Si in organic and aqueous solutions using high-resolution continuum source or line source flame atomic absorption spectrometry, Spectrochim. Acta B 120 (2016) 45–56. J.L. Raposo Jr., S.R. Oliveira, J.A.G. Neto, J.A. Nobrega, B.T. Jones, Determination of silicon in lubricating oil by high-resolution continuum source flame atomic absorption spectrometry using least-square background correction and internal standardization, Anal. Lett. 44 (2011) 2150–2161. H.M. Dong, V. Krivan, Direct solid-sampling electrothermal atomic absorption spectrometry methods for the determination of silicon in oxides of niobium, titanium and zirconium, Spectrochim. Acta B 56 (2001) 1645–1656. W. Boschetti, L.M.G. Dalagnol A, M. Dullius, A.V. Zmozinski, E.M. Becker, M.G.R. Vale, J.B. de Andrade, Determination of silicon in plant materials using direct solid sample analysis with high-resolution continuum source graphite furnace atomic absorption spectrometry, Microchem. J. 124 (2016) 380–385. B. Gómez-Nieto, M.J. Gismera, M.T. Sevilla, J.R. Procopio, Determination of essential elements in beverages, herbal infusions and dietary supplements using a new straightforward sequential approach based on flame atomic absorption spectrometry, Food Chem. 219 (2017) 69–75. L.C.C. de Oliveira, M.A. Vieira, A.S. Ribeiro, M.T. Lisboa, R.A. Gonçalves, R.C. de Campos, Determination of silicon in vegetable oil and biodiesel by high-resolution continuum source flame atomic absorption spectrometry using sample dilution with xylene, Energy Fuel 26 (2012) 7041−7044. I.M. Goncalves, M. Murillo, A.M. González, Determination of metals in used lubricating oils by AAS using emulsified samples, Talanta 47 (1998) 1033–1042. J.L. Burguera, M. Burguera, Analytical applications of emulsions and microemulsions, Talanta 96 (2012) 11–20. C. Viana, D. Bohrer, L.M. de Carvalho, P.C. do Nascimento, M.B. da Rosa, Emulsified systems for metal determination by spectrometric methods, Trends Anal. Chem. 53 (2014) 49–59. M.V. Reboucas, D. Domingos, A.S.O. Santos, L. Sampaio, Determination of trace metals in naphtha by graphite furnace atomic absorption spectrometry: comparison between direct injection and microemulsion pretreatment procedures, Fuel Process. Technol. 91 (2010) 1702–1709. Commission Implementing Regulation (EU) No 1348/2013 of 16 December 2013 amending Regulation (EEC) No 2568/91 on the characteristics of olive oil and olive-residue oil and on the relevant methods of analysis. Food and Agric. Int. Organization of United Nations (FAO) and World Health Organization (WHO) Codex Alimentarius. International Food Standards. Standard for edible fats and oils not covered by individual standards codex stan. 19–1981. K. Szivos, The application of organic solvents in atomic absorption spectrophotometry, Period, Polytech. Chem. Eng. 25 (1981) 121–132.