Accepted Manuscript Analytical methods Effect of sample preparation procedure for the determination of As, Sb and Se in fruit juices by HG-ICP-OES Maja Welna, Anna Szymczycha-Madeja PII: DOI: Reference:

S0308-8146(14)00437-3 http://dx.doi.org/10.1016/j.foodchem.2014.03.046 FOCH 15575

To appear in:

Food Chemistry

Received Date: Revised Date: Accepted Date:

8 September 2013 4 March 2014 9 March 2014

Please cite this article as: Welna, M., Szymczycha-Madeja, A., Effect of sample preparation procedure for the determination of As, Sb and Se in fruit juices by HG-ICP-OES, Food Chemistry (2014), doi: http://dx.doi.org/ 10.1016/j.foodchem.2014.03.046

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effect of sample preparation procedure for the determination of As, Sb and Se in fruit juices by HG-ICP-OES Maja Welna*, Anna Szymczycha-Madeja

Wroclaw University of Technology, Chemistry Department, Analytical Chemistry Division, Wybrzeże Wyspiańskiego 27, 50-370 Wroclaw, Poland *Corresponding author. Tel.: +48 71 3203232; fax.: +48 71 3202494 E-mail address: [email protected] (M. Welna)

Abstract Various sample preparation procedures for the simultaneous determination of As, Sb and Se in fruit juices by hydride generation inductively coupled plasma optical emission spectrometry (HG-ICP-OES) were examined. Applicability of total wet digestion with HNO3/H2O2, partial decomposition (solubilisation in aqua regia), 1:1 dilution with 2% (v/v) HNO3 and later centrifugation or filtration and direct analysis after centrifugation were evaluated. Hydrides were generated in the reaction of an acidified sample solution with NaBH4 reductant after pre-reduction with KI-ascorbic acid for total As and Sb, and boiling with HCl for total Se. Best results, i.e. limits of detection (LODs) of 0.51-0.73 ng mL-1, precision (RSD) within 1.7-3.6% and recoveries for spiked juice samples between 101-106% were found using auqa regia treatment. Under the optimal condition, the external calibration using standards treated and measured in the same way as samples was applied for the analysis. This procedure simplifying and improving the step of sample preparation prior to measurements of As, Sb and Se in fruit juices by HG-ICP-OES, thus could be adequate for the routine analysis in terms of the quality control of these drinks. Several commercial carton fruit juices were analysed with the developed procedure.

1

Keywords: Fruit juice; Sample preparation; Arsenic; Selenium; Antimony; Hydride generation; ICP-OES

1. Introduction

Today, in the quest to get healthy, more and more people choose fruit juices, being recognized as a source of nutrient compounds, including essential elements. Therefore, the assessment of mineral content of fruit juices is of a great importance, particularly because of habitual consumption of this beverage as well as to monitor their quality. An extended literature review devoted to the subject of the elemental analysis of fruit juices has been presented very recently by Szymczycha-Madeja, Welna, Jedryczko, & Pohl (2014). Elemental analysis of fruit juices is not an easy task due to the complexity of the matrix, resulting in the necessity of proper sample preparation prior to measurements (Juranović-Cindrić, Zeiner, Kröppl, & Stindeger, 2011). On the other hand, analyte concentration range is also the critical factor in choosing the suitable detection technique (Cautela, Santelli, Boscaino, Laratta, Servillo, & Castaldo, 2009). A survey of the literature indicates that the majority of works dealing with analysis of fruit juices are devoted to determination of important to human health macro- and micro-elements (e.g. Abid et al., 2014; Dosumu, Oluwaniyi, Awolola, & Okunola, 2009; Eisele & Drake, 2005; Jalbani, Ahmed, Kazi, Rashid, Munshi, & Kandhro, 2010; Juranović-Cindrić et al. 2011; Konić-Ristić et al., 2011; Krośniak, Gąsoł, Szałkowski, Zagrodzki, & Drewisz, 2010; Nascentes, Arruda, Nogueira, & Nóbrega, 2008; Peuke, 2009; Soylak, Elci, Saracoglu, & Divrikli, 2002). Considering the quality and the safety of the fruit juices not only levels of essential macro-

2

and micro-nutrients, but also the contents of trace elements should be known. In this field, a limited number of studies investigate fruit juices with respect to elements present at very low levels. Among various trace elements, Se is essential to the human health (Navarro-Alarcón & López-Martínez, 2010), while As or Sb are classified as toxic (Kapaj, Peterson, Liber, & Bhattacharya, 2006; Sundar & Chakravarty, 2010). Concentrations of the aforementioned elements in real samples are typically very low, usually below the detection limit of common instrumental techniques used. The exception is ICP-MS, which belongs to the one of the most sensitive analytical tool for the determination of elements at the ppt level (Gutzeit, Winterhalter, & Jerz, 2008). At present, the hydride generation technique (HG) coupled with the atomic spectrometry is a popular method determining for traces of As, Sb and Se (e.g. Koreňvská, 2006; Chen & Chen, 2014; Ventura, Stibilj, do Carmo Freitas, & Pacheco, 2007; Reyes, Cervera, & de la Guardia, 2011). The reduction of hydride-forming elements to volatile hydrides in the acid media by sodium tetrahydroborate (NaBH4) increases remarkably their detection performance as compared to the conventional sample introduction by the pneumatic nebulisation (PN) (Pohl, 2004). In general, analysis of fruit juices on the content of As, Sb and Se is performed using spectrochemical methods such as ICP-OES (Barnes, 1999; Cautela et al., 2009; Tufuor, Bentum, Essumang, & Koranteng-Addo, 2011), ICP-MS (Gutzeit et al., 2008), F-AAS (Ashraf, Jaffar, & Masud, 2000; Maduabuchi, Adigba, Nzegwu, Oragwu, Okonkwo, & Orisakwe, 2007), GF-AAS (Cautela et al., 2009; Orak, 2009), ET-AAS (Liu, Gong, Li, Xu, & Lin, 1996) and HG-AAS (Cautela et al., 2009). In order to reduce matrix effects originating from the organic compounds samples undergo the wet oxidative digestion in pressurized closed vessel microwave assisted (Barnes, 1997; Gutzeit et al., 2008; Orak, 2009) and open vessel systems using hot plates (Ashraf et al., 2000; Liu et al., 1996; Maduabuchi et al., 2007; Tufuor et al., 2011) or heating blocks (Cautela et al., 2009) in the presence of concentrated

3

aggressive digestion reagents. The decomposition of samples by microwave-assisted digestion is realized using concentrated (65%) HNO3 alone (10 g of sample; 8 mL of reagent; element (As) measurement directly in the resulting solution) (Orak, 2009) or its mixtures, i.e. 65% HNO3+30% H2O2, 1.5/1 v/v ratio (1 g of sample; 5 mL of reagents; 25 mL of final volume) (Gutzeit et al., 2008) and 65% HNO3+98% H2SO4 followed by addition of a 30% H2O2, 5/1/1.5 v/v ratio (10 g of sample; 15 mL of reagents; 100 mL of final volume) (Barnes, 1997). In case of conventional wet digestion in open-vessel systems, facilitated using hot plates or heating blocks, samples are decomposed in a mixture of concentrated solutions, including 50% HNO3+70% HClO4, 4/1 v/v ratio (100 mL of sample; 25 mL of reagents; element (As) measurement directly in the resulting solution after gentle heating) (Ashraf et al., 2000), 65% HNO3 :98% H2SO4 (5:1 v/v ratio) + H2O, 2/1 v/v ratio (25 g of sample; 30 mL of reagents; 50 mL of final volume) (Cautela et al., 2009), aqua regia alone (10 g of sample; 20 mL of reagent; 25 mL of final volume) (Liu et al., 1996) or aqua regia with the addition of a 30% H2O2 solution, 6/1 v/v ratio (50 mL of sample; 70 mL of reagents; 50 mL of final volume) (Tufuor et al., 2011). Diluted HNO3 solution (1:2) for this type of digestion has also been established to be advantageous (5 mL of sample; 45 mL of reagent; 10mL of final volume) (Maduabuchi et al., 2007). The traditional approach to the sample preparation is the most tedious and timeconsuming step of the whole analysis, requires the use of hazardous reagents and may lead to the sample contamination or the loss of the analytes. Alternative methodologies, involving partial or no previous decomposition, certainly could avoid or minimize all inconveniences related to the sample digestion. According to our best knowledge, there is no work on the evaluation of optimal conditions for the determination of total As, Sb and Se in fruit juices by the HG-ICP-OES with as less as possible sample treatment. So far, El Hadri, Morales-Rubio, & de la Guardia (2007) proposed a simple method for the determination of total As in soft

4

drinks (as apple juices) by HG-AFS. In the method, arsenic hydride was generated directly in the reaction between NaBH4 (3% (m/v)) and 2-5 mL of sample pre-reduced and acidified, before dilution to 50 mL, with KI and HCl (final concentrations of 1% (m/v) and 2 mol L-1, respectively). The proposed procedure maintained satisfactory recoveries for spiked juice samples (94±6%), the precision between 0.1-9% and LOD in the range from 0.01 to 0.03 ng mL-1. Similarly, an alternative approach using dilution (1:1) with 2% (v/v) HNO3 and later centrifugation for the measurements of total concentrations of microelements (including Sb) in fruit juice by ICP-OES with pneumatic nebulisation (PN) has been optimised by Froes, Neto, e Silva, Naveira, Nascentes, & da Silva (2009). However, in case of Sb, it could be reliable for its presence >0.5 mg L-1. Simple dilutions (usually at ratios of 1:1 to 1:20) with diluted (0.2% or 2% (v/v)) HNO3 or with water have also been proposed (e.g. AkpinarBayizit, 2010; Ince & Coskun, 2008; Juranović-Cindrić et al., 2011), but with a special attention into the determination of non-hydride elements. Lastly, we evaluated a simple and fast method for the multi-elemental analysis of 100%, naturally cloudy fruit juices using PN-ICP-OES (Szymczycha-Madeja & Welna, 2013). Although this approach was found to be very efficient for the simultaneous determination of Al, Ba, Ca, Cd, Cr, Cu, Fe, Mg, Mn, Ni, P, Pb, Sr and Zn, it was not sensitive for the quantification of As, Sb and Se. For this reason, the main goal of the present work was to continue our previous investigation by evaluating an undemanding procedure, i.e. with no previous complete digestion, being accurate and precise for the simultaneous determination of total As, Sb and Se within a low concentration range in these fruit juices by HG-ICP-OES. Five preparation procedures of fruit juice samples prior to the determination of As, Sb and Se by HG-ICP-OES were compared and their analytical performance was characterized.

5

2. Experimental

2.1. Samples and reagents Seven samples (N=7) of naturally cloudy 100% carton (1 L) fruit juices of one brand (apple, grape, grapefruit, orange, pineapple, pear and plum) from the local market (Wroclaw, Poland) were analysed. For each sample, one container was bought. The samples were stored at ambient temperature and prior to analysis were thoroughly shaken. All chemicals were of analytical grade and deionised water (18.3 MΏ cm) from an EASYpure system Model D7033 (Barnstead, Thermolyne Corporation, USA) was used throughout. Concentrated 65% (v/v) HNO3 (Merck KGaA, Darmstadt, Germany), 37% (v/v) HCl (POCh S.A., Gliwice, Poland) and 30% (m/v) H2O2 (POCh) solutions were used for sample preparation. Aqua regia was prepared just before use by mixing 3:1 (v/v) concentrated HCl and HNO3 solutions. Stock standard solutions (1000 µg mL-1) of As(III), As(V), Sb(V), Se(IV) and Se(VI) were prepared by dissolving the appropriate amounts of NaAsO2 (SigmaAldrich, St. Louis, MO, USA), Na2HAsO4×7H2O (Sigma-Aldrich), K[Sb(OH)6] (SigmaAldrich), Na2SeO3 (POCh) and Na2SeO4 (POCh) in water, respectively. In case of Sb(III), a Merck ICP standard solution (Merck) was used. All stock standard solutions were stored in a refrigerator at 4 °C. Working solutions were prepared daily by appropriate stepwise dilutions with water of the concentrated stock standard solutions (1000 µg mL-1). A 1.0% (m/v) of reductant was made daily from the powder of NaBH4 (Sigma-Aldrich) dissolved in 0.1 mol L1

NaOH (POCh) and filtered (0.45 µm) before being used. The antifoaming agent used to

avoid foam formation during hydride generation was the Antifoam A aqueous emulsion (30% (m/v), Sigma-Aldrich). Solid KI (POCh), L(+)-ascorbic acid (POCh) and thiourea (POCh) were used as pre-reducing reagents. A mixture of 10% (m/v) KI in 2.0% (m/v) ascorbic acid

6

in addition to single solutions (10% (m/v)) of KI and thiourea were prepared by dissolving respective reagents in water.

2.2. Sample preparation methods

Four sample preparation methods were examined: direct analysis, minimal treatment (acidic dilution), partial decomposition (aqua regia solubilisation) and total decomposition (wet digestion). The hot-plate heating digestion with HNO3/H2O2 mixture (P1), the dilution (1:1) with 2% (v/v) HNO3 and later centrifugation (P2) or filtration (P3) and the direct analysis after centrifugation (P4) proceed as described previously (Szymczycha-Madeja & Welna, 2013). In the procedure with partial decomposition (P5), 5 mL of fruit juice samples were solubilised in 2 mL of aqua regia. The resulting mixtures were sonicated in an ultrasonic bath for 15 min at room temperature, then made up to 20 mL with deionised water and centrifuged (10 min, 12.000 rpm) to remove solid particles. All the sample solutions were analysed in triplicate (n=3). As previously (SzymczychaMadeja & Welna, 2013), optimization studies were carried out with the grapefruit juice. With each set of sample solutions, procedural blanks were prepared to correct final results following the same treatments as in the case of samples. To avoid differences between the composition of the sample and standard solutions, working standard solutions of As (As(III), As(V), Sb (Sb(III), Sb(V)) and Se (Se(IV), Se(VI)) were also prepared and passed through all preparation procedures. Finally, the most suitable sample preparation procedure (in terms of precision, accuracy and detection limit) was applied to prepare remaining fruit juices and analyse them on the content of As, Sb and Se by HG-ICP-OES against the external calibration.

7

2.3. Pre-treatment for total As, Sb and Se determination by HG-ICP-OES

Total As and Sb (as As(III) and Sb(III)) and Se (as Se(IV)) were separately measured in the sample solution aliquots. Prior to the HG measurements, As(V) and Sb(V) were prereduced to As(III) and Sb(III) with 1% KI (in 0.2% ascorbic acid) and 3 mol L-1 HCl, while Se(VI) was converted to Se(IV) by heating with 6 mol L-1 HCl. Accordingly, for As and Sb, an aliquot of 2.5 mL of each sample solution was transferred to a 5-mL volumetric flask, then 0.5 mL of 10% (m/v) KI in 2% (m/v) ascorbic acid and 1.25 mL of concentrated HCl were added and left to react for about 20 min. After this time flask was diluted to the volume with deionised water. For Se, an aliquot of 1.5 mL of each sample solution was transferred to a 10mL tube with 1.5 mL of concentrated HCl. Tube was heated in a water bath at 90 ºC for 30 min. Blank samples and standards were also pre-reduced.

2.4. Hydride generation

As, Sb and Se hydrides were generated in a continuous flow system with a gas-liquid phase separation connected to the ICP-OES spectrometer. The system consisted of a modified cyclonic spray chamber (used to separate gas and liquid phases), a parallel pneumatic nebulizer placed in the tangential tube of the chamber and peristaltic pumps with delivery tubes (Welna, Lasowska, & Zyrnicki, 2011). In the manifold, an acidified sample and the NaBH4 solutions were simultaneously pumped in two separate streams to a Y-junction where they merged and the HG reaction occurred. Then, the reaction mixture was introduced at the bottom of the chamber to a special reaction cavity. Hydrides and other gaseous co-products were separated from the spent liquid and transported to the plasma torch by a stream of carrier Ar introduced through the gas inlet of the nebulizer. The sample inlet of the nebulizer was

8

blocked. Post-reaction solution was drained to waste with the aid of a peristaltic pump. For better plasma stability, the level of liquid in the chamber was controlled and kept at a constant level.

2.5. Apparatus

Measurements of element concentrations were performed using a Jobin Yvon (France) sequential ICP-OES instrument (JY 38S). Working parameters for the HG reaction and the ICP-OES detection are listed in Table 1. After the reaction mixture passed into the chamber, 30 s were necessary to achieve steady As, Sb and Se signals. A JP Selecta (Barcelona, Spain) ultrasonic cleaning bath with heating (UltrasonsH) was used for experiments with aqua regia. A MPW-350 centrifuge (MPW Med. Instruments, Poland) was used to accelerate the separation of liquid phase from solid particles.



3. Results and Discussion

3.1. Pre-reduction and hydride generation conditions

Preliminary experiments were made based on the HG reaction of As, Sb and Se forms with NaBH4 in HCl medium. Typical concentration of NaBH4 was used (1% (m/v)). Higher than 1% (m/v) concentration was avoided, since it made the plasma unstable, likely due to large amounts of H2 produced in the HG reaction. Noteworthy, the concentration of NaBH4 at

9

1.0% (m/v) is the most common in the HG technique hyphenated with ICP-OES (Pohl, 2004). Because As(III), Sb(III) and Se(IV) forms are favourable for the effective HG reaction any As(V), Sb(V) and Se(VI) present were pre-reduced before analysis of real samples. Various pre-reducing reagents were tested: KI, KI-ascorbic acid, thiourea and HCl. It was found that the KI-ascorbic acid mixture and thiourea provided an effective reduction of As(V), while only the use of the KI-ascorbic acid mixture led to a complete conversion of Sb(V) into the Sb(III) form. Unfortunately, Sb(V) was only partially reduced in the presence of thiourea. For Se, only the boiling with HCl was established to be the most appropriate. Mixture of KI and ascorbic acid at different concentration ratios for As and Sb, and HCl at 90-100 ºC for Se has also been commonly suggested in the literature (see e.g. Koreňvská, 2006; Chen & Chen, 2014; Ventura et al., 2007; Reyes et al., 2011). El Hadri et al. (2007) proposed KI alone for the effective reduction of As(V)-As(III) before the determination of total As in apple juice samples by HG-AFS, however, when KI is used as a reductant the addition of ascorbic acid is advisable to prevent the interference of an excess of I3¯ being formed by oxygen (Chen & Chen, 2014; Cava-Montesinos, Cervera, Pastor, & de la Guardia 2004). Such an effect was also observed in the present study. To conclude, quantiative reduction of As(V) and Sb(V) to As(III) and Sb(III) was achieved using KI-ascorbic acid mixture in HCl medium (final concentrations of 1%-0.2% (m/v) and 3 mol L-1, respectively), while complete reduction of Se(VI) to Se(IV) was assured with 6 mol L-1 HCl by heating in a water batch at 90 ºC for 30 min. Independently of the sample preparation procedure used (P1-P5), complete prereductions were obtained.

3.2. Comparison of sample preparation procedures

10

Five preparation procedures of fruit juice samples prior to HG-ICP-OES measurements of As, Sb and Se were compared by evaluating the precision (as %RSD), the accuracy and the limit of detection (LOD). Ought to the lack of certified reference material for fruit juices, the accuracy of the methods was assessed by the recovery studies with the spiked samples. LODs were determined as concentrations corresponding to three times the standard deviation (3×SD) of 10 executive measurements of respective procedural blanks (3σ criterion). For the spiking experiments and the determination of recoveries, grapefruit juice samples were spiked with the known amounts of As, Sb and Se species ranged between 0.050.5 µg mL-1 (final concentrations of 0.025-0.25 µg mL-1 after pre-reductions) and analysed following the same sample preparation procedures and the pre-reductions. These results are listed in Table 2.



As can be seen, total wet digestion (P1) is a suitable method for the fruit juice sample preparation. RSDs and LODs varied between 3.7-5.6% and 0.6-1.3 ng mL-1, respectively. In general, the LOD values are lower than obtained using PN-ICP-OES (Barnes, 1997) and are in the same order of magnitude as obtained with GF-AAS (Barnes, 1997). These limits are also similar to reported results obtained for As by HG-AAS (Cautela et al., 2009) and for As and Se by ICP-MS (Gutzeit et al., 2008). The recovery values were, independently of the concentration and the speciation form of the element, changed from 93.1 to 102.8%. The recoveries of the spiked values obtained are in agreement with the literature devoted to the analysis of juice sample by ICP-OES or ET-AAS (Barnes, 1997; Tufour et al., 2011; Liu et al., 1996) prepared by complete acidic digestion.

11

Among procedures with no previous digestion, simple dilution with diluted HNO3 (P2,P3) or direct analysis (P4) were justified but only to the As and Sb determination. Quantitative recoveries were obtained for the two considered As and Sb species (As: 98.8103.8%, Sb: 99.9-102.1%). RSD values (1.9-4.0%) after sample dilution (P2,P3) were lower than achieved for the wet digestion procedure (P1), but LODs resulted from the use of diluted acid were slightly higher (by 20-40%). Noteworthy, the precision obtained with the 2-nd procedure (dilution followed by centrifigutation) was better than resulted from the 3-rd one (dilution and the later filtration). The direct analysis (P4) led to lower precision (4.7-7.7%) and detectability as compared to these achieved from the digestion procedure (P1). Detection limits were from 40% (As) up to two times higher (Sb). It must be noticed that during HG reaction of untreated samples or samples treated with diluted acid it was necessary to add an antifoaming agent to the NaBH4 solution prior to the ICP measurements to avoid extensive foam formation and to obtain accurate results. Without the antifoaming reagent a mass of foams were formed, which affected the stability of the emission signals of As, Sb and Se (decrease of sensitivity and precision). The same effect during Se HG in coffee and milk samples diluted only with low concentrated HNO3 was observed (Asfaw & Wibetoe, 2005). Antifoam A reagent was used as the antifoaming reagent, since it has been usually employed (Cava-Montesinos et al., 2004; Chen & Chen, 2014; Ródenas-Torralba, Morales-Rubio, & de la Guardia, 2005; Rasmussen, Quian, & Sloth, 2013). On the other hand, the effect of the Antifoam A presence on the plasma condition was not observed. Unfortunately, these procedures (P2-P4) were found to be ineffective for the Se determination (0% as the recovery) and a possible reason is that the volatile element species could be retained in the solid particles formed when the recommended pre-reduction step with HCl prior to the Se hydride generation was applied. The problem was satisfactorily solved when aqua regia reagent was added. The use of aqua regia (P5) was found to completely

12

solubilise As, Sb and Se species, making them available for the HG reaction. This is in agreement with previous reports (Asfaw & Wibetoe, 2005; Cava-Montesinos et al., 2004). As it can be seen from Table 2, complete recoveries were achieved for the studied elements (101.5-105.9%) and the best precision of their measurements (1.7-3.6% as RSD). LODs of As, Sb and Se were also satisfactory. Accordingly, for As and Sb they were close and for Se even two times lower than obtained for the wet digestion (P1) as well as 2 orders on magnitude lower in comparison to those estimated when applying PN to measure As, Sb and Se by ICP-OES (As: 47 ng mL-1; Sb: 44 ng mL-1; Se: 54 ng mL-1). Samples of grapefruit juice treated with aqua regia were orange-yellow what suggested a partial decomposition of the organic matter. No precipitation was found when the pre-reduction step prior to the total As, Sb and Se determination by HG was carried out. In contrast, for the 2-nd and the 3-rd procedure, low concentrated HNO3 led to simple sample dilution instead of the partial decomposition that can explain the latter difficulties in Se measurements. Previously (Szymczycha-Madeja & Welna, 2013), analysis of fruit juices after 1:1 dilution with 2% (v/v) HNO3 and later centrifugation (P2) was found to be optimal for the determination of 14 non-hydride elements by ICP-OES with conventional pneumatic nebulisation. No sample decomposition was required. In this work, analysis after simple dilution has demonstrated to be accurate only for the As and Sb determination. Satisfactory, both elements could also be determined directly in untreated juice sample (P4), what is in agreement with the results obtained by El-Hadri et al. (2007) in case of total As determination in apple juice samples by HG-AFS. However, it should be commented that, for the direct analysis of juice sample they achieved lower detection limits compared to those attained in this study. On the other hand, according to previous work, this procedure (P4) was proved to be the least efficient and not applicable in practice for the multi-element analysis by PN-ICPOES. Also, both direct analysis (P4) and analysis after dilution (P2) could not be used for the

13

determination of Se for the reason discussed earlier. It could be detected but only if the traditional approach of sample preparation, i.e. previous total decomposition (P1) was applied. Satisfactory, simultaneous determination of As, Sb and Se by ICP-OES (introduced as hydrides) without the need of the complete digestion of sample matrix was allowed after partial decomposition of the sample with aqua regia in an ultrasound water batch. Consequently, this treatment (P5) was chosen for further studies.

3.3. Optimization of chosen procedure

Aqua regia-based procedure (P5) was optimized by evaluation of an optimal L/L ratio (liquid-to-liquid, v/v) assuming 2 mL of aqua regia used for the sample solubilisation. The following ratios were examined: 1:10, 1:4 and 1:2 (v/v), whereas sample volumes of 2, 5 and 10 mL and final solution volume of 20 mL were considered. The resulting mixtures were sonicated at room temperature for 15 minutes, made up to the required volume with water (20 mL) and centrifuged before the analysis. Concentrations of the final solutions were equal to 10, 25 and 50% (v/v), respectively. The results were evaluated as elements recoveries using both the standard addition method and aqueous standard external calibration lines (Table 3). For the external calibration, standard solutions were treated in the same way as samples.



It was observed that releasing of As and Sb into solution was independent of the L/L ratio and calibration method used. Every time, recoveries within 100% were achieved. In this case, standard addition slopes were similar to those found using external standards, therefore, it indicates that the matrix effects on measuring of As and Sb were not observed in practice. In

14

contrast, Se could be totally recovered but only if standard addition calibration was used. Application of external standards led to two times lower recovery of Se for the most concentrated final sample solution (1:2 ratio). It was suspected, because these sample solutions were dense and solid particles have appeared when the pre-reduction with HCl prior to Se measurement was employed. The solubisation was not complete, leading to interference from the sample matrix during Se determination. On the other hand, for the lowest sample concentration (1:10 ratio), the recovery value for Se although high, was not quantitative. The best results were found for the sample solution of 25% (1:4 ratio), thus condition, i.e. 5 mL of juice sample and 20 mL of the final volume was taken as the optimal.

3.4. Application

The proposed methodology (the solubilisation in aqua regia, P5) was applied to the analysis of seven fruit juice samples for total As, Sb and Se by HG-ICP-OES. Based on the results, it was obtained that examined fruit juices have not detectable (ND) amounts of As, Sb and Se. Concentrations of these elements were less than their respective LODs, i.e. below 0.6, 0.5 and 0.7 ng mL-1, respectively. Compared the results shown in this study with those reported for fruit juices (including apple, orange, plum, exotic or red fruits) of different origin it can be seen that the concentrations of the above mentioned elements can vary in a wide range, i.e. As: ND-20 ng mL-1 (El-Hadri et al., 2007; Cautela et al., 2009; Tufuor et al., 2011; Barnes, 1997; Orak, 2009; Gutzeit et al., 2008; Maduabuchi et al., 2007), Sb: ND-19 ng mL-1 (Cautela et al., 2009) and Se: ND-246 ng mL-1 (Cautela et al., 2009; Barnes, 1997; Gutzeit et., 2008; Liu et al., 1996). The great variability in element concentrations depends on various factors such as the geographic area, the characteristic of soil, agricultural practices applied during the growth of plant and fruits, kind of fruit as well as technology employed during the

15

production process of juices (Cautela et al., 2009; Froes et al., 2009; Gutzeit et al., 2008; Juranović-Cindrić et al., 2011; Tufuor et al., 2011). It can be concluded that, in general, concentrations of As, Sb and Se in fruit juices are low. However, elevated As levels, i.e. upper safe limit of 1 mg L-1, have also been reported for local fruit juices consumed in Pakistan (Ashraf et al., 2000), being the highest, up to 2.920 mg L-1 in apple juice. In this situation, it is understandable that regular control of fruit juices on the content of trace elements is mandatory.

4. Conclusions

Methodology for the determination of traces of As, Sb and Se in fruit juices by HG-ICPOES, without the need of a previous total digestion of samples, was proposed. Procedure based on the partial decomposition of samples using their solubilisation in aqua regia and ultrasonication at room temperature, followed by the pre-reduction of element species with KI-ascorbic acid (for total As and Sb) in HCl medium and boiling with HCl (for total Se) demonstrates the appropriate reproducibility, precision, accuracy and detectability as well eliminates losses of As, Sb and Se. The degree of the sample decomposition was critical. The partial decomposition was found to be especially essential in case of Se, which could be completely lost during the recommended pre-reduction step of Se(VI) with HCl prior to the HG reaction if other sample preparation procedures were used, e.g. the dilution with low concentrated HNO3. On the other hand, analysis after the minimal treatment (simple dilution and later centrifugation), but also the direct analysis (only centrifugation) followed by the prereduction with KI-ascorbic acid mixture were found to be the simplest methods, however, only for the determination of As and Sb, as demonstrated here.

16

The analysis of various naturally cloudy carton fruit juices indicates that they contain not detectable amounts of As, Sb and Se. Further studies in this area (sample proceeding for determination of hydride and non-hydride elements in different beverages) are undertaken.

Acknowledgements

The work was financed by a statutory activity subsidy from the Polish Ministry of Science and Higher Education for the Faculty of Chemistry of Wrocław University of Technology.

References Abid, M., Jabbar, S., Wua, T., Hashim, M. M., Hua, B., Lei, S., & Zeng, X. (2014). Sonication enhances polyphenolic compounds, sugars, carotenoids and mineral elements of apple juice. Ultrasonics Sonochemistry, 21, 93–97. Akpinar-Bayizit, A. (2010). Analysis of Mineral Content in Pomegranate Juice by ICP-OES. Asian Journal of Chemistry, 22, 6542-6546. Asfaw, A., & Wibetoe, G. (2005 Simultaneous determination of hydride (Se) and nonhydride-forming (Ca, Mg, K, P, S and Zn) elements in various beverages (beer, coffee, and milk), with minimum sample preparation, by ICP-AES and use of a dual-mode sample-introduction system. 173-179. Ashraf, W., Jaffar, M., & Masud, K. (2000). Heavy Trace Metal and Macronutrient Levels in Various Soft Drinks and Juices. Journal of The Chemical Society of Pakistan, 22, 119124.

17

Barnes, K. W. (1997). Trace Metal Determination in Fruit, Juice, and Juice Products Using an Axially Viewed Plasma. Atomic Spectroscopy, 18, 84-100. Cautela, D., Santelli, F., Boscaino, F., Laratta, B., Servillo, L., & Castaldo, D. (2009). Elemental content and nutritional study of blood orange juice. Journal of the Science of Food and Agriculture, 89, 2283-2291. Cava-Montesinos, P., Cervera, M. L., Pastor, A., & de la Guardia, M. (2004). Determination of As, Sb, Se, Te and Bi in milk by slurry sampling hydride generation atomic fluorescence spectrometry. Talanta, 62, 175-184. Chen, G., & Chen, T. (2014). SPE speciation of inorganic arsenic in rice followed by hydridegeneration atomic fluorescence spectrometric quantification. Talanta, 119, 202-206. Dosumu, O. O., Oluwaniyi, O. O., Awolola, G. V., & Okunola, M. O. (2009). Stability studies and mineral concentration of some Nigerian packed fruit juices, concentrate and local beverages. African Journal of Food Science, 3, 082-085. Eisele, T. A., & Drake, S. R. (2005). The partial compositional characteristics of apple juice from 175 apple varieties. Journal of Food Composition and Analysis, 18, 213-221. El-Hadri, F., Morales-Rubio, A., & de la Guardia, M. (2007). Determination of total arsenic in soft drinks by hydride generation atomic fluorescence spectrometry. Food Chemistry, 105, 1195-1200. Froes, R. E. S., Neto, W. B., e Silva, N. O. C., Naveira, R. L. P., Nascentes, C. C., & da Silva, J. B. B. (2009). Multivariate optimization by exploratory analysis applied to the determination of microelements in fruit juice by inductively coupled plasma optical emission spectrometry. Spectrochimica Acta Part B, 64, 619-622. Gutzeit, D., Winterhalter, P., & Jerz, G. (2008). Nutritional Assessment of Processing Effects on Major and Trace Element Content in Sea Buckthorn Juice (Hippophaë rhamnoides L. ssp. rhamnoides). Journal of Food Science, 73, 97-102.

18

Ince, H., & Coskun, N. (2008). Determination of Heavy metals in Fruit Juice by Flame Atomic Absorption Spectrometry. Asian Journal of Chemistry, 20, 3537-3542. Jalbani, N., Ahmed, F., Kazi, T. G., Rashid, U., Munshi, A. B., & Kandhro, A. (2010). Determination of essential elements (Cu, Fe and Zn) in juices of commercially available in Pakistan. Food and Chemical Toxicology, 48, 2737-2740. Juranović-Cindrić, I., Zeiner, M., Kröppl, M., & Stingeder, G. (2011). Comparison of sample preparation methods for the ICP-AES determination of minor and major elements in clarified apple juices. Microchemical Journal, 99, 364-369. Kapaj, S., Peterson, H., Liber, K., & Bhattacharya, P. (2006). Human health effects from chronic arsenic poisoning-a review. Journal of Environmental Science and Health Part A, 41, 2399-2428. Konić-Ristić, A., Šavikin, K., Zdunić, G., Janković, T., Juranic, Z., Menković, N., & Stanković, I. (2011). Biological activity and chemical composition of different berry juices. Food Chemistry, 125, 1412-1417. Koreňvská, M. 2006. Determination of arsenic, antimony, and selenium by FI-HG-AAS in foods consumed in Slovakia. Journal of Food Nutrition Research, 45, 84-88. Krośniak, M., Gąsoł, M., Szałkowski, M., Zagrodzki, P., & Drewisz, M. (2010). Cornelian Cherry (Cornus mas L.) Juices as a Source of Minerals in Human Diet. Journal of Toxicology and Environmental Health, Part A, 73, 1155-1158. Liu, Y., Gong, B., Li, Z., Xu, Y., & Lin, T. (1996). Direct determination of selenium in a wild fruit juice by electrothermal atomic absorption spectrometry. Talanta, 43, 985-989. Maduabuchi, J-M. U., Adigba, E. O., Nzegwu, C. N., Oragwu, C. I., Okonkwo, I. P., & Orisakwe, O. E. (2007). Arsenic and Chromium in Canned and Non-Canned Beverages in Nigeria: A Potential Public Health Concern. International Journal of Environmental Research and Public Health, 4, 28-33.

19

Nascentes, C. C., Arruda, M. A. Z., Nogueira, A. R., & Nóbrega, J. A. (2008). Direct determination of Cu and Zn in fruit juices and bovine milk by thermospray flame furnace atomic absorption spectrometry. Talanta, 64, 912-917. Navarro-Alarcón, M., & López-Martínez, M. C. (2010). Essentiality of selenium in the human body: Relationship with different diseases. Science of the Total Environment, 249, 347371. Orak, H. H. (2009). Evaluation of antioxidant activity, colour and some nutritional characteristics of pomegranate (Punica granatum L.) juice and its sour concentrate processed by conventional evaporation. International Journal of Food Science and Nutrition, 60, 1-11. Peuke, A. D. (2009). Nutrient composition of leaves and fruit juice of grapevine as affected by soil and nitrogen fertilization. Journal of Plant Nutrition and Soil Science, 172, 557564. Pohl, P. (2004). Recent advance in chemical vapour generation through the reaction with sodium tetrahydroborate. Trends in Analytical Chemistry, 23, 1-16. Rasmussen, R. R., Quian, Y., & Sloth, J. J. (2013). SPE HG-AAS method for the determination of inorganic arsenic in rice-results from method validation studies and a survey on rice products. Analytical and Bioanalytical Chemistry, 405, 7851-7857. Ryes, M. N. M., Cervera, M. L., & de la Guardia, M. (2011). Determination of Inorganic species of Sb and Te in Cereals by Hydride Generation Atomic Fluorescence Spectrometry. Journal of Brazilian Chemical Society, 22, 197-203. Ródenas-Torralba, E., Morales-Rubio, Á., & de la Guardia, M. (2005). Multicommutation hydride generation atomic fluorescence determination of inorganic tellerium species in milk. Food Chemistry, 91, 181-189.

20

Soylak, M., Elci, L., Saracoglu, S., & Divrikli, U. (2002). Chemical analysis of Fruit Juice of European Cranberrybush (Viburnum opulus) from Kayseri-Turkey. Asian Journal of Chemistry, 14, 135-138. Sundar, S., & Chakravarty, J. (2010). Antimony Toxicity. International Journal of Environmental Research and Public Health, 7, 4267-4277. Szymczycha-Madeja, A., & Welna, M. (2013). Evaluation of a simple and fast method for the multi-elemental analysis in commercial fruit juice samples using atomic emission spectrometry. Food Chemistry, 141, 3466-3472. Szymczycha-Madeja, A., Welna, M., Jedryczko, D., & Pohl, P. (2014). Developments and strategies in the spectrochemical elemental analysis of fruit juices, Trends in Analytical Chemistry, doi: http://dx.doi.org/10.1016/j.trac.2013.12.005. Tufuor, J. K., Bentum, J. K., Essumang, D. K., & Koranteng-Addo, J. E. (2011). Analysis of heavy metals in citrus juice from the Abura-Asebu-Kwamankese District, Ghana. Journal of Chemical and Pharmaceutical Research, 3, 397-402. Ventura, M. G., Stibilj, V., do Carmo Freitas, M., & Pacheco, A. M. G. (2007). Determination of ultratrace levels of selenium in fruit and vegetable samples grown and consumed in Portugal. Food Chemistry, 105, 1195-1200. Welna, M., Lasowska, J., & Zyrnicki, W. (2011). Determination of some inorganic species of Fe, Mn and Cr by chemical vapor generation hyphenated with inductively coupled plasma atomic emission spectrometry. Journal of Brazilian Chemical Society, 22, 11641169.

21

Table 1. HG-ICP-OES operating parameters. Generator (MHz)

40.68

Rf power (W)

1000

Injector i.d. (mm)

2.5

Observation zone

12 mm above load coil

Ar flow rates (L min-1)

plasma gas: 13 sheath gas: 0.20 carrier gas: 0.25

Solutions uptake (mL min-1)

1.0

(acidified sample, reductant) Replicates

3

Wavelength (nm)

As I 193.7 Sb I 206.8 Se I 196.1

Hydride generation optimum condition NaBH4 concentration (in 0.1 mol L-1 NaOH) (% (m/v)) 1.0 Sample acidity with HCl (mol L-1)

3.0 (As, Sb) and 6.0 (Se)

KI-ascorbic acid concentration (% (m/v))

1.0-0.2

22

Table 2. The analytical characteristic of different fruit juice sample preparation procedures: the hotplate heating digestion with HNO3/H2O2 (P1), the dilution (1:1) with 2% (v/v) HNO3 and later centrifugation (P2) or filtration (P3), the direct analysis after centrifugation (P4) and solubilisation in aqua regia (P5) for the total As, Sb and Se determination by HG-ICP-OES. Procedure

P1

P2

P3

P4

P5

99.3 ± 5.1

101.3 ± 2.8

98.8 ± 4.7

99.1 ± 4.2

102.1 ± 2.2

3.7

2.7

4.0

6.7

2.3

LOD (ng mL-1)

0.66

1.1

0.97

1.0

0.62

Recoverya (%)

102.8 ± 3.1

103.8 ± 3.8

100.5 ± 3.2

102.4 ± 3.6

102.0 ± 1.0

RSDb (%)

5.2

3.6

3.1

7.7

1.9

LODc (ng mL-1)

0.65

1.1

1.0

1.0

0.59

Recoverya (%)

97.8 ± 5.8

101.7 ± 5.8

100.1 ± 7.4

96.9 ± 9.4

102.8 ± 4.1

4.9

2.3

3.7

7.5

3.6

LOD (ng mL-1)

0.59

0.81

0.88

1.2

0.54

Recoverya (%)

99.9 ± 4.9

102.1 ± 0.6

99.9 ± 7.3

98.9 ± 4.6

101.5 ± 1.8

RSDb (%)

4.4

1.9

3.1

4.7

1.7

LODc (ng mL-1)

0.63

0.77

0.75

1.1

0.51

Recoverya (%)

96.3 ± 4.7

RSDb (%)

NDd

NDd

NDd

5.3

As Recoverya (%) As(III) RSDb (%)

As(V)

Sb

Sb(III) RSDb (%)

Sb(V)

Se Se(IV)

105.9 ± 2.1 2.0

23

LOD (ng mL-1)

1.3

0.67

Recoverya (%)

93.1 ± 5.8

101.8 ± 1.1

5.6

3.4

1.3

0.73

Se(VI) RSDb (%) LODc (ng mL-1) a

For addition of 0.05, 0.1, 0.25 and 0.5 µg mL-1; as average (n=4) with SD.

b

From three independent measurements of sample solutions containing 0.1 µg mL-1 of each of the

elements. c

As As(III) or Sb(III) or Se(IV).

d

Not detected (0% as recovery); solid residue after pre-reduction of Se(VI) into Se(IV) with 6 mol L-

1

HCl.

24

Table 3. Recovery values for fruit juice samples spiked with As(III), Sb(III) and Se(IV) after aqua regia treatment and pre-reductions measured by HG-ICP-OES using external and standard addition calibrations. Recoveryb (%) L/La ratio (v/v)

10

4

2

Element

standard

external

addition

calibration

As

102.3 (5.5)

94.9 (5.0)

Sb

95.3 (4.2)

96.0 (4.3)

Se

102.9 (2.8)

82.5 (1.9)

As

102.0 (2.1)

97.3 (1.2)

Sb

99.8 (4.1)

97.7 (3.2)

Se

103.1 (2.9)

99.0 (4.9)

As

106.1 (6.9)

99.4 (8.8)

Sb

106.8 (3.3)

95.2 (7.4)

Se

108.4 (4.2)

44.2 (9.7)

a

Liquid-to-liquid ratio (v/v).

b

Average values (n=3) with relative standard of

deviation (RSD) in brackets. Additions ranged from 0.1 to 0.5 µg mL-1.

25

Highlights

•

Method for As, Sb, Se determination in fruit juice by HG-ICP-OES was proposed

•

Effect of sample decomposition on HG reaction for As, Sb and Sb was investigated

•

As and Sb could me measured directly from untreated sample or after acidic dilution

•

Partial decomposition was essential prior to HG of Se

•

Aqua regia treatment can be used to routine analysis on content of As, Sb and Se

26