New Selenite Ion-Selective Electrodes Based on 5,10,15,20-Tetrakis-(4-Methoxyphenyl)-21H,23H-Porphyrin-Co(II)

Journal of Hazardous Materials 181 (2010) 857–867

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

New selenite ion-selective electrodes based on 5,10,15,20-tetrakis-(4-methoxyphenyl)-21H,23H-porphyrin-Co(II)

H. Ibrahim, Y.M. Issa ∗, Ola R. Shehab

Chemistry Department, Faculty of Science, Cairo University, Gamma Street, Giza, Cairo 12613, Egypt article info abstract

Article history: New polymeric membrane (PME), modiﬁed carbon paste (MCPE), and coated wire (CWE) selen- Received 11 January 2010 ite ion-selective electrodes based on 5,10,15,20-tetrakis-(4-methoxyphenyl)-21H,23H-porphyrin-Co(II) Received in revised form 29 March 2010 (CoTMeOPP) are reported. The best composition was the electrode containing 2% CoTMeOPP as the active Accepted 19 May 2010 material and 49% TCP as plasticizer. The electrodes reveal a Nernstian behavior over a concentration range Available online 26 May 2010 of 5.5 × 10−5 to 1.1 × 10−2 M for PME, 5.2 × 10−5 to 1.2 × 10−2 M for MCPE and 1.2 × 10−4 to 4.4 × 10−3 M for CWE. The potentiometric response is pH dependent, since selenous acid is a diprotic acid. The slope Keywords: of the selenite PVC electrode was −57.0 mV for the monovalent anion at pH 6.47, and −26.0 mV for the Selenite × −5 × −5 Graphite divalent anion at pH 11.00. The detection limits were 3.4 10 and 4.7 10 M at pH values 6.47 and Membrane 11.00, respectively. The electrodes manifest advantages of low resistance, very short response time (15 s), Coated wires and most importantly good selectivities relative to a wide variety of other anions. In fact, the proposed Porphyrin selenite ion-selective electrodes show a great improvement compared to previously reported electrodes for selenite ion. The electrode was used for the determination of selenite in selenite/selenate mixture, in sodium selenite raw material powder, and in VitaFit Selenium ACE antioxidant tablets with recovery ranges of 90.0–103.3%. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Selenium occurs naturally in a number of inorganic forms, including selenide, selenite and selenate. Selenium is an essential In the recent years, there has been increasing interest in the trace trace element for plants, animals and humans. The biological func- determination of selenium. Apart from natural sources (mainly tion of selenium shows dual characteristics, it can cause disease by metal–sulfur minerals), selenium compounds are widely spread its deficiency but is toxic at level relatively close to those required out from the combustion of fossil fuels, and used in the glass, elec- for health [2]. Elemental selenium and most metallic selenides have tronic industries and in agriculture [1]. The toxicity and essential relatively low toxicities because of their low bioavailability. By con- nature of selenium depends on its chemical form. Selenium nat- trast, selenite and selenate are very toxic and have modes of action urally exists in several oxidation states in inorganic and organic similar to that of arsenic. Speciation of selenium in environmental forms. The inorganic species most frequently found in water and and biological samples is necessary to understand the biochemical 2− 2− soil are selenite (SeO3 ) and selenate (SeO4 ). cycle, mobility and uptake of selenium, as well as its toxicity. The development of reliable techniques to study the speciation Hydride generation coupled to atomic absorption spectrome- of selenium in environmental and biological samples is necessary try (AAS), inductively coupled plasma (ICP) or atomic fluorescence to understand the biological cycle, mobility, and uptake of sele- spectrometry (AFS) are the techniques that most commonly used nium, as well as its toxicity. Many problems in selenium speciation for the speciation of inorganic selenium [3–6]. Detection limits analysis are associated with the low concentration of each species were 0.1–0.4 ␮g/l. Since this reaction is specific for Se(IV), these to be determined. Significant errors may also arise by loses of techniques can be applied for total inorganic selenium (the sum of volatile selenium compounds, instability of its chemical species Se(IV) and Se(VI)) after quantitative reduction of selenate to selen- specially in analytical methods involving many chemical treat- ite. The content of selenate is obtained as the difference between ments [1]. two determinations. The disadvantages of these techniques are the complexity and cost of the instrument. The use of ion-selective electrodes has gained importance because of their ease of handling and their selectivities on spe- ∗ Corresponding author. Tel.: +20 2 35868094; fax: +20 2 35728843. cific ions. There are only a few investigations regarding selenite E-mail address: [email protected] (Y.M. Issa). ion-selective electrodes. Ion-selective electrodes for the determi-

as the outer reference electrode. Millipore Elix S (Automatic Sani- tization Module, ASM) was used for obtaining the deionized water. CARY 50 Probe UV–Visible spectrophotometer, Varian was used for the spectrophotometric measurements. Anhydrous sodium selenite (Na2SeO3) was purchased from Loba Chemie PVT LTD., India. 5,10,15,20-tetrakis-(4-methoxyphenyl)- 21H,23H-porphyrin-Co(II) [C48H36CoN4O4], poly(vinyl chloride) (PVC) of high relative molecular weight, graphite powder, tetrahy- drofuran (THF), AgNO3, tris(hydroxymethyl)-aminomethane and the plasticizers, o-nitrophenyloctyl ether (o-NPOE), dibutyl phthalate (DBP), dioctyl phthalate (DOP), dibutyl-butyl phosphonate (DBBP), and tricresyl phosphate (TCP) were purchased from Aldrich chemical company, USA.

2.2. Solutions

Stock solution of 0.1 M sodium selenite was prepared by dissolving 1.73 g Na2SeO3 in 100 ml deionized water. The solution was kept in refrigerator. Acetate buffer solution of pH 5.89 was prepared by mixing 95 ml of 0.1 M sodium acetate solution with 5 ml 0.1 M acetic acid solution. Phosphate buffer solutions of pH values, 5.59, 6.47, and 7.38 Fig. 1. Structure of (CoTMeOPP). were prepared by mixing the appropriate volumes 0.01 M KH2PO4 and 0.01 M Na2HPO4 solutions [14]. Tris–HCl buffer solution pH 8.5 was prepared by the addition of 0.01 M HCl to 50 ml 0.01 M tris nation of selenite were constructed by compressing either HgSeO3 solution [15]. Ammoniacal buffer solution of pH 9.5 was prepared or a mixture of HgSeO3 and Hg2Cl2 [7]. Cai et al. [8] had developed selenite ion-selective electrode using 4,6-dibromopiaselenole as by the addition of the required volume of 0.01 M NH4OH to 50 ml active material, PVC as membrane matrix and dibutyl phthalate as 0.01 M NH4Cl solution then completed to 100 ml using deionized plasticizer. Three different selenite-selective electrodes were fab- water. Sodium chloride–sodium hydroxide solution of pH 10.2 and 11.0 were prepared by the addition of the required volume of 0.01 M ricated, the first was developed using solid salts of Ag2Se and Cu2S [9]. Piaselenol (PIS) was used in the form of 1,2-phenylenediamine NaOH to 50 ml 0.01 M NaCl solution then completed to 100 ml using selen complex as the active material for ion-selective electrode deionized water. To investigate the selectivity of the proposed electrodes, 0.5 M membrane sensitive to selenite ion [10]. Silicon rubber [11] was − − − used as the active membrane matrix for selenite-selective electrode sodium salt solution of each of the following ions, Cl ,Br ,NO3 , 2− consisted of 4% dialkyldimethylammonium chloride, and QAD 86 SO4 , and 0.1 M sodium salt solution of each of the following ions, 2− − 2− 3− 2− S ,F ,SO3 ,PO4 , and 0.1 M potassium salt solution of SeO4 , PI. A self-made ion-selective electrode was made using Ag2Se as − electroactive material [12], and used for the determination of sele- SCN were prepared in deionized water. nium in biological materials. It can be observed that in all papers reported in the literature, there is no mention about seleneous acid 2.3. Preparation of the electrodes species as a function of pH. It was, therefore, felt worthwhile to develop a better sensor 2.3.1. Preparation of PVC-membrane electrodes for selenite ions using 5,10,15,20-tetrakis-(4-methoxyphenyl)- Membranes of different compositions and plasticizers were pre- 21H,23H-porphyrin-Co(II) (CoTMeOPP) (Fig. 1) as an electroactive pared. The membrane was prepared by dissolving the required phase in PVC matrix or carbon paste. amount of PVC and plasticizer in 5 ml THF. The calculated amount of An unusual feature of porphyrin chelate is that the metal ion ionophore was dissolved in THF and mixed with the PVC/plasticizer is bonded to four nitrogen atoms, which are themselves linked solution in Petridish (5.0 cm diameter). The total weight of con- together in a conjugated system. This imparts a high degree of stituents in each batch is fixed at 0.2 g. The membranes were left mesomerism whereby a substitution anywhere in the porphyrin to dry in air (not less than 24 h) to obtain homogeneous and uni- nucleus can relay its electron donating and attracting tendency to form thickness. The membranes formed as a film in the bottom all the four coordinating atoms that makes metalloporphyrins so are taken out carefully and glued to a homemade electrode bod- sensitive to the influence of substituents [13]. ies with an adhesive of PVC dissolved in THF. The electrodes were PVC membranes, coated wires, and modified carbon paste elec- filled with 10−1 M NaCl and 10−3 M sodium selenite solution and trodes based on CoTMeOPP as ionophore for selenite ions were preconditioned by soaking in 10−3 M selenite solution for 24 h. prepared, optimized, and checked at different concentration ranges for selenite. The electrodes were used for the determination of 2.3.2. Preparation of coated wire electrodes selenite in pure solutions, selenite/selenate mixture, in raw mate- Pure silver, graphite, glassy carbon and cupper rods of 1 mm rial of sodium selenite, and in VitaFit Selenium ACE antioxidant diameter and 12 cm in length were insulated leaving 2 cm at one tablets. end for coating and 1 cm at the other end for connection. The pol- ished rod surface of each type was coated with active membrane 2. Experimental by dipping the exposed end into the coating solution, as previously described in membrane preparation, and allowing the film to dry 2.1. Apparatus and reagents in air for about 1 min. The process was repeated until a plastic film of approximately 1 mm thickness was formed (about 10 times). For potential measurements a JENWAY 3010 digital pH/mV The prepared electrodes were preconditioned by soaking for 1 h meter was used. A SENTEK R1/2MM Ag/AgCl electrode was used in 10−3 M selenite solution. H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867 859

2.3.3. Preparation of carbon paste electrodes 2.5. Potentiometric determination A teflon holder (12 cm length) with a hole at one end (7 mm diameter and 3.5 mm depth) for the carbon paste filling served The standard addition method was applied [18]. In this method, as the electrode body. Electrical contact is made through a stain- the proposed electrode was immersed into a sample of 25 ml less steel rod through the center of the holder. This rod can move with unknown concentration (∼10−7 to 10−4 M) sample solution up and down by screw movement to press the paste down when and the equilibrium potential of Eu was recorded, then 0.1 ml of renewal of the electrode surface is needed. The modified paste was 1.0 × 10−3 M of standard selenite solution was added into the test prepared by mixing the appropriate weight of ionophore and high solution and the equilibrium potential (Es) was obtained. From the purity graphite with acetone. The mixture was homogenized, left potential change E =(Eu−Es) one can determine the concentration at room temperature to evaporate acetone, then a weighed amount of the test sample using the equation: of plasticizer was added and a very intimate homogenization was −1 V n E/S V then achieved by careful mixing with agate pestle in agate mortar. C = C s 10 ( ) − x x s V + V V + V The paste was then packed into the hole of the electrode body. The x s s x carbon paste was smoothed onto a paper until it had a shiny appear- where Cx is the concentration to be determined, Vx is the volume of ance and used directly for potentiometric measurements without the original sample solution, Vs and Cs are the volume and concen- preconditioning. tration of the standard solution added to the sample to be analyzed, respectively, E is the change in potential after the addition of cer- 2.4. Selectivity of the sensor tain volume of standard solution, and S is the slope of the calibration graph. Potentiometric selectivity factor was evaluated using the matched potential method [16]. According to this method, the 3. Results and discussion −5 activity of analyte was increased from aA = 1.0 × 10 M (reference solution) to a = 1.0 × 10−3 M, and the change in potential (E) A 3.1. PVC-membrane electrodes corresponding to this increase in activity is measured. Then, 0.1 M solution of an interfering ion is added to a new 1.0 × 10−5 M analyte 3.1.1. Effect of pH on the electrode response reference solution until the same potential change (E) is recorded Calibration graphs (electrode potential vs. −log [selenite]) were [17], the concentration of the added amount is thus a . The selec- B constructed using standard selenite solution adjusted to different tivity coefficient KMPM for each interferent was calculated using the A,B pH values in the range of 5.5–11.0 using acetate, phosphate, and tris following equation: buffers. The slope of the calibration graphs, Fig. 2, Table 1 was found a − a to be in the range of 50–60 mV/concentration decade for selen- KMPM = A A A,B ite samples buffered at pH 5.59–7.38 indicating that the electrode aB responds to the monovalent anion at this pH range. Increasing the

Fig. 2. Calibration graphs for selenite PVC-selective electrode at different pH values (a) 5.59, (b) 6.47, (c) 7.38, (d) 8.50, (e) 9.50, (f) 10.20, and (g) 11.00. 860 H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867

Table 1 membrane demonstrated Nernstian response and remarkable Slopes, linear ranges, and detection limits of the PVC electrode containing 2% CoT- selectivity for the respective selenite ion over several common MeOPP and 49% TCP at different pH values. inorganic anions. The key ingredient of poly(vinyl chloride) (PVC) pH Slope, mV/decade Linear range, M Detection membrane electrodes is the incorporated ionophore, which defines limit, M the selectivity of the electrode. In plastic membrane electrodes, the 5.59 −64.00 3.50 × 10−5 to 7.90 × 10−3 1.60 × 10−5 amount of this ionophore should be sufficient to obtain reasonable − − − 6.47 −57.00 5.5 0 × 10 5 to 1.00 × 10 2 3.40 × 10 5 response at the gel layer-test solution interface which is responsi- 7.38 −40.00 5.30 × 10−5 to 6.90 × 10−3 3.46 × 10−5 ble for the membrane potential. Also, the amount of the plasticizer 8.50 −33.00 6.60 × 10−5 to 6.10 × 10−3 3.46 × 10−5 9.50 −17.50 1.30 × 10−4 to 1.40 × 10−2 7.41 × 10−5 should be to the extent that produces a membrane of good physical 10.20 −23.50 8.10 × 10−5 to 1.20 × 10−2 5.37 × 10−5 properties and at the same time plays efficiently its role as a solvent 11.00 −26.00 6.60 × 10−5 to 1.20 × 10−2 4.70 × 10−5 mediator for the ionophore. Investigations reveal that, the performance characteristics reported for a given PVC membrane may vary depending on the pH of selenite samples leads to a decrease in the slope of the calibra- membrane composition and the nature of solution to which the tion graphs until they reach to about 23–30 mV at pH 8.50–11.00 electrodes are exposed [20–22]. Thus, several membranes of vary- indicating that the electrodes also respond to the divalent anion. ing nature and ratio of ionophore/PVC/plasticizer were prepared for This in agreement with the fact that seleneous acid is a diprotic the systematic investigation of each membrane composition. The acid with pK = 2.62 and pK = 8.32, thus [19]: 1 2 amount of ionophore was found to seriously affect the sensitivity of + − H2SeO3 H + HSeO3 pK1 = 2.62 the membrane electrodes, experimental trials proved that certain percentage of ionophore was optimum as indicated by the Nern- − + 2− stian behavior of the electrode. However, further increase of the HSeO3 H + SeO3 pK2 = 8.32 ionophore over this percentage resulted in a diminished response As indicated from the distribution curve of seleneous acid slope of the electrode, most probably due to some inhomogenities species as a function of pH, Fig. 3, it is obvious that the monovalent and possible saturation of the membrane [23]. − HSeO3 ions are the predominating species within the pH range The amount of ionophore should be sufficient to obtain a reason- 2− 4.5–6.5 and at pH higher than 10.0, the divalent SeO3 anions are able ionic exchange at the gel layer-test solution interface, which the predominating species. Accordingly, the electrode response has is responsible for the membrane potential. Also, the amount of the to be measured at fixed pH value using an appropriate buffer. plasticizer should be to the extent that produces a membrane of good physical properties and at the same time plays efficiently its 3.1.2. Composition of membrane electrodes role as a solvent mediator for the ionophore. It is evident that the On using an ion-selective electrode in analytical determination, change of composition affects significantly the response of the elec- there are many factors that should be taken into consideration as trode (Table 2). The data collected in Table 2, shows that a PVC they affect the performance of the electrode towards its respective membrane plasticized with TCP and containing 1.0% CoTMeOPP ion. Some of these factors are: composition of the membranes and has a slope of −47.80 mV while the one containing 2.0% CoTMeOPP the amount of ionophore, plasticizer, the effect of soaking time on has a −57.00 mV slope and the 3% CoTMeOPP membrane shows a life span and the effect of operational conditions such as temper- diminished response of −54.00 mV slope. The detection limit was ature, response time, presence of interferents and pH, etc. These 4.70 × 10−5 M for both 1.0% and 3.0% CoTMeOPP membrane compo- studies have been performed according to the recommendation of sition while it was 3.40 × 10−5 M for the 2% CoTMeOPP composition. the IUPAC [18]. The electrode containing 2.0% CoTMeOPP has a linear working The composition of the membrane was varied so as to reach range (5.50 × 10−5 to 1.00 × 10−2 M) which is wider than the other the optimum composition exhibiting the best performance char- two compositions (Table 2). acteristics (slope of calibration graph, linear concentration range and reproducibility of the results). In preliminary experiments, 3.1.3. Effect of plasticizer membranes with and without ionophore were constructed. The Besides the critical role of the nature and the amount of membrane with no ionophore displayed no measurable response ionophore on the characteristics of the sensors, some other impor- towards the studied selenite ion, whereas, in the presence of the tant features such as the nature of the solvent mediator, the proposed ionophore for selenite-selective electrode, the optimized plasticizer/PVC ratio and the nature of any additives used, are known to significantly influence the sensitivity and selectivity of ion-selective electrodes [24–26]. The nature of the plasticizer influences both dielectric constant of the membrane and the mobility of the ionophore [27]. The solvent mediator has a dual function, it acts as liquefying agent, enabling homogeneous solubilization and modifying the distribution constant of the ionophore used. For plasticizer to be adequate for use in sensors, it should gather certain properties and characteristics, such as having high lipophilicity, high molecular weight, low vapor pressure, and high capacity to dissolve the substrate and other additives present in the matrix [28]. The influence of the plasticizer type and its quantity on the characteristics of the studied sensors was investigated using five plasticizers with different polarities including DBP, DOP, DBBP, TCP and NPOE. As shown in Fig. 4(a) and (b), the electrode containing TCP generally shows better potentiometric responses, i.e. sensitivity and linearity range of the calibration plots. This is due to its high dielectric constant and relatively high molecular weight. Fig. 3. Distribution curves of seleneous acid species as a function of pH. The results given in Table 2 indicate that the sensor composed H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867 861

Table 2 Composition, slopes, linear ranges, and detection limits of calibration curves for selenite PVC membranes, chemically modiﬁed carbon paste and coated wires electrodes.

% composition Slope, mV/decade pH 6.67 Detection limit, M Linear range, M

Conventional type

Ionophore PVC Plasticizer

1.0 49.5 49.5 (TCP) −47.80 4.70 × 10−5 5.80 × 10−5 to 2.60 × 10−3 2.0* 49.0 49.0 (TCP) −57.00 3.40 × 10−5 5.50 × 10−5 to 1.00 × 10−2 3.0 48.5 48.5 (TCP) −54.00 4.70 × 10−5 5.80 × 10−5 to 2.90 × 10−3 2.0 49.0 49.0 (DBP) −49.50 6.40 × 10−5 8.50 × 10−5 to 6.30 × 10−3 2.0 49.0 49.0 (NPOE) −65.80 1.40 × 10−5 3.50 × 10−5 to 6.30 × 10−3 2.0 49.0 49.0 (DOP) −52.00 3.90 × 10−5 7.20 × 10−5 to 5.60 × 10−3 2.0 49.0 49.0 (DBBP) −31.90 4.80 × 10−5 6.90 × 10−5 to 7.90 × 10−3

pH 11.00

2.0* 49.0 49.0 (TCP) −26.00 4.70 × 10−5 6.60 × 10−5 to 1.20 × 10−2 2.0 49.0 49.0 (DBP) −27.60 3.10 × 10−4 2.60 × 10−4 to 1.20 × 10−2 2.0 49.0 49.0 (NPOE) −37.40 6.90 × 10−5 3.20 × 10−4 to 7.90 × 10−3 2.0 49.0 49.0 (DOP) −27.00 3.50 × 10−4 1.20 × 10−4 to 1.50 × 10−2 2.0 49.0 49.0 (DBBP) −19.00 6.90 × 10−5 7.70 × 10−4 to 1.50 × 10−2

Carbon paste Slope, mV/decade Detection limit, M Linear range, M

Ionophore Graphite Plasticizer pH 6.47

1.0 49.5 49.5 (TCP) −55.48 3.50 × 10−5 5.20 × 10−5 to 7.90 × 10−3 2.0* 49.0 49.0 (TCP) −65.00 3.50 × 10−5 5.20 × 10−5 to 1.20 × 10−2 3.0 48.5 48.5 (TCP) −50.33 3.50 × 10−5 5.20 × 10−5 to 1.20 × 10−2 2.0 49.0 49.0 (NPOE) −51.00 5.20 × 10−5 8.30 × 10−5 to 7.90 × 10−3 2.0 49.0 49.0 (DOP) −35.00 5.20 × 10−5 7.40 × 10−5 to 7.90 × 10−3 2.0 49.0 49.0 (DBBP) −30.60 6.30 × 10−5 8.70 × 10−5 to 7.90 × 10−3 pH 11.00 2.0* 49.0 49.0 (TCP) −29.40 7.70 × 10−5 8.70 × 10−5 to 1.20 × 10−2 2.0 49.0 49.0 (NPOE) −26.40 5.40 × 10−5 9.70 × 10−5 to 7.90 × 10−3 2.0 49.0 49.0 (DOP) −15.70 5.00 × 10−5 7.20 × 10−5 to 1.10 × 10−2 2.0 49.0 49.0 (DBBP) −14.40 3.90 × 10−5 6.90 × 10−5 to 3.10 × 10−3

Coated wires (2% ionophore + 49.0% TCP)

Type of bed Slope, mV Detection limit, M Linear range, M

pH 6.47 pH 11.00 pH 6.47 pH 11.00 pH 6.47 pH11.00

Ag −45.90 −32.20 5.80 × 10−5 3.90 × 10−6 9.10 × 10−5 to 3.20 × 10−3 7.90 × 10−6 to 2.60 × 10−3 Ag (1% graphite) −61.80 −39.00 9.10 × 10−5 3.90 × 10−6 1.20 × 10−4 to 4.40 × 10−3 7.90 × 10−6 to 1.30 × 10−3 Graphite −23.60 – 8.30 × 10−5 – 1.90 × 10−4 to 1.50 × 10−3 – Copper −43.00 −19.40 4.70 × 10−4 7.90 × 10−4 8.90 × 10−4 to 8.70 × 10−3 1.20 × 10−3 to 7.90 × 10−3 Glassy carbon −19.60 – 3.20 × 10−5 – 5.00 × 10−4 to 3.10 × 10−3 –

* The optimum composition chosen.

Fig. 4. Effect of plasticizer on potential response of PVC membrane containing 2.0% CoTMeOPP at pH 6.47 (a) and pH11.00 (b). 862 H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867

Table 3 Effect of soaking on selenite PVC electrode at pH 6.47 and 25.0 ± 1.0 ◦C.

Soaking time Slope, mV/decade Linear range, M Response time, s

1h −38.80 8.30 × 10−5 to 2.90 × 10−3 ≤20 1 day −57.00 8.30 × 10−5 to 2.90 × 10−3 ≤20 4 days −49.80 8.30 × 10−5 to 2.90 × 10−3 ≤20 6 days −44.00 8.30 × 10−5 to 2.90 × 10−3 ≤20 of 2.0% ionophore, and TCP as plasticizer gives best sensitivity with near Nernstian slope of −57.00 mV/10-fold at pH 6.47 with detection limit of 3.4 × 10−5 M over a linear range of 5.50 × 10−5 to 1.00 × 10−2 M and −26.00 mV/10-fold at pH 11.00 with detection limit of 4.70 × 10−5 M over a linear range of 6.60 × 10−5 to 1.20 × 10−2 M. Therefore, this composition was used to study vari- ous operation parameters of the electrode. The same composition and plasticizer was taken as the optimum composition of a coating mixture for the CWE.

3.1.4. Effect of soaking on life span of the PVC-membrane electrodes Fig. 5. Calibration graphs of selenite coated Ag wire electrode (a), doped with 1.0% The performance characteristics of the investigated electrodes graphite (b) at pH 11.00 (I) and pH 6.47 (II). were studied as a function of soaking times. For this purpose, the × −3 − electrode was soaked in 1.00 10 M selenite solution for dif- (1.62 ␮ cm 1) of silver. A significant improvement of CWE was ferent intervals and then the effect of soaking on the calibration obtained when transducer layer with electronic conductivity was graph slope, usable concentration range and the response time applied between the substrate and the ion-selective membrane, to were measured (Table 3). As shown in the table, while the slope yield the so-called “all-solid-state ion-selective electrode” [30].In − of the electrode after 1 h was 38.80 mV/decade, it changed to this type of electrodes, electrochemically synthesized conducting − 57.00 mV/decade after 1 day. The membrane losses its sensitivity polymer is electrodeposited directly onto an electric conducting − after 4 days, as the slope decreased to 49.80 mV/decade. This neg- substrate or a conducting polymer is included in the cocktail of ion- ative effect of soaking is attributed to the leaching of the ionophore selective membrane components and the mixture is casted directly and plasticizer to the bathing solution which is related to the dis- on a solid conducting substrate [31–33]. It was found that very fine tribution equilibria and diffusion rates. Another explanation can be pure graphite powder can be used as electrically conducting addi- related to the penetration of the water molecules from the bathing tive added to the membrane matrix instead of using conducting solution to the membrane and consequently slow solvation of the polymer. This type of potentiometric sensors result in a signifi- lipophilic salts in situ, so they are slowly leached out and limit the cant improvement of charge transfer between the substrate and electrode life [29]. However, it was noted that, the electrode in use the membrane phase. As a result, stability of the sensor potential if kept dry in refrigerator showed good preservation of its slope and lowering of detection limits were observed [17]. In this work, a value and response extending to 2 months with refreshing the elec- small mass percentage (1.0% graphite) was added to the plastic ion- trode by immersing into soaking solution for 30 min or after a set selective membrane phase of selenite and coating the silver wire. of measurements. The results show good improvement in the electrode characteristics, slope, and response time. Fig. 5 represents the improvement 3.2. Coated wire electrodes (CWEs) of the slope of the coated silver wire electrode when doped with 1.0% graphite in which the slope has increased from −45.90 to The main advantage of these electrodes towards other types of −61.80 mV for the monovalent selenite ion in acidic medium pH ion-selective electrodes arises from the fact that, there is no need 6.47 over linear range of 1.20 × 10−4 to 4.40 × 10−3M with detec- for an inner electrolyte solution. The electrodes, simply; consist of tion limit 9.10 × 10−5 M. While, for the divalent selenite ion in basic conductive bed coated with a coating mixture containing an elec- medium pH 11.00 the slope increased from −32.20 to −39.00 mV troactive material. The optimum composition of a coating mixture over a linear range of 7.90 × 10−6 to 1.30 × 10−3 M with detection is that which exhibits a Nernstian slope and a wide dynamic range limit 3.90 × 10−6 M(Table 2). of concentration when it was used as PVC membrane. 3.3. Chemically modified carbon paste electrode (CMCPE) 3.2.1. Effect of electrode bed To investigate the effect of the bed nature on the efficiency of CoTMeOPP as a carrier was found to be highly sensitive to selen- coated wire electrodes, the optimized coating mixture was used ite with respect to several other anions. Therefore, we studied in in the preparation of electrodes with different conductive beds, details the performance of the CMCPE containing this ionophore for namely silver, copper, graphite and glassy carbon. After condi- selenite in aqueous solutions. It is well known that the selectivity, tioning, each electrode was examined in the concentration range linear dynamic range and sensitivity obtained for a given CMCPE 1.00 × 10−7 to 1.00 × 10−2 M of selenite solution. The dynamic depends significantly on the paste composition [34] and the nature range of concentration and the limit of detection for the electrodes of the solvent mediator [35,36]. were evaluated according to the IUPAC recommendations [18].As The amount of the ionophore in the paste affects the response can be noticed from Table 2, all wires give unsatisfactory response of the electrode. The results in Table 2 show that the electrode towards selenite except silver wire. Ag wire coated electrode has containing 1.0% CoTMeOPP has a slope of −55.48 mV, while that a slope −45.90 mV at pH 6.47 and a Nernstian behavior with slope containing 2.0% CoTMeOPP has a slope of −65.00 mV, then the slope −32.20 mV at pH 11.00. This is attributed to the lowest resistivity decreased to −50.33 mV when 3.0% of CoTMeOPP was used. The H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867 863

Fig. 6. Potential–time plot for the response of selenite PVC membrane at pH 6.47 and 11.00.

Fig. 7. Calibration graphs of some inorganic anions using selenite CMCP electrode at pH 6.47 and 11.00. three electrodes have the same detection limit and almost the same ±1 mV) after successive immersion of the electrode in a series linear range. of its respective selenite solution, each having a 10-fold increase in concentration from 1.00 × 10−4 to 1.00 × 10−1 M. The resulting 3.3.1. Effect of plasticizer on CMCPE potential–time plots for PVC electrode is shown in Fig. 6. The PVC- It is noteworthy that the nature of the solvent inﬂuences the membrane electrode yielded steady potentials within 20 s while mobility of ions and provides the appropriate conditions for the the CMCPE reaches the steady state potential within 15 s. The CWE incorporation of selenite ion into the paste prior to its exchange shows better response time shorter than both PVC and CMCP elec- with the soft ionophore. trodes (5 s). The relatively higher response time of the PVC, and Four plasticizers with different polarities including TCP, DBBP, CMCP electrodes may be attributed to the higher noise originated NPOE, and DOP were studied and the optimum composition for from the higher resistance of the liquid-contact membranes, which the electrode was given in Table 2. The use of TCP as a solvent was as a consequence of their large thickness [37]. The potential mediator, as low polarity compound, in this electrode results in readings stay constant, to within ±1 mV, for at least 4 min. This a Nernstian linear plot over a concentration range of 5.20 × 10−5 to indicates the fast response time of the suggested electrodes and 1.20 × 10−2 M, 8.70 × 10−5 to 1.20 × 10−2 M with slopes of −65.00 their high stability after attending the steady state value. and −29.40 mV at pH 6.47 and 11.00 respectively. Whereas, in case of other solvent mediators; the potentiometric response is much 3.5. Selectivity of selenite electrode different from the expected Nernstian value. The selectivity coefﬁcient is the main source of information 3.4. Response time concerning interferences on the electrode response. In analytical applications, the response for the analyte must be as high as pos- The response time [18] of the electrode was tested by measur- sible, i.e. the response for foreign substances must be very small, ing the time required to achieve a steady state potential (within so that the electrode exhibits a Nernstian dependence on the pri- 864 H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867

Table 4 Selectivity coefﬁcient values −log K MPM for selenite-selective electrode. selenite,j

− log K MPM selenite,j

Interferent PVC membrane CMCPE CWE

pH 6.47 pH 11.00 pH 6.47 pH 11.00 pH 6.47 pH 11.00

Cl− 3.60 3.10 3.60 2.70 2.60 2.30 Br− 3.30 3.00 3.50 2.50 2.90 2.50 2− SO4 2.80 2.60 3.70 2.70 2.50 2.30 2− SO3 3.50 2.80 3.60 1.73 2.30 2.00 S2− 2.30 1.10 1.90 2.00 1.82 1.77 2− SeO4 3.10 2.30 2.00 2.70 2.20 2.00 F− 3.00 2.70 3.50 2.50 2.80 2.30 − NO3 2.70 3.10 2.50 2.50 2.30 2.00 SCN− 0.59 0.69 0.20 0.36 0.51 0.33

Table 5 Determination of selenite in pure solution and in real samples using selenite PVC-membrane electrode by applying standard addition method and potentiometric titrations at pH 6.4.

Taken, M Found, M Found, M by ICP-AES RSD, % Recovery, %

Pure selenite solutions 1.00 × 10−5 9.50 × 10−6 9.90 × 10−6 0.32 95.00 1.00 × 10−4 9.70 × 10−5 9.99 × 10−5 0.21 97.00 1.00 × 10−3 1.03 × 10−3 1.01 × 10−3 0.46 103.00 Selenite/Selenate mixture 1.00 × 10−5 9.60 × 10−6 9.89 × 10−6 0.52 96.00 1.00 × 10−4 9.70 × 10−5 1.00 × 10−4 0.56 97.00 1.00 × 10−3 1.01 × 10−3 1.04 × 10−3 0.78 101.00 Sodium selenite raw material powder 1.00 × 10−5 9.50 × 10−6 9.80 × 10−6 0.67 95.00 1.00 × 10−4 9.70 × 10−5 1.09 × 10−4 0.54 97.00 1.00 × 10−3 9.89 × 10−4 9.89 × 10−4 0.48 98.90 Antioxidant VitaFit Selenium tablets 1 tablet in 250 ml 1.70 × 10−6 1.68 × 10−6 1.72 × 10−6 0.35 98.90 3 tablets in 100 ml 1.30 × 10−5 1.25 × 10−5 1.29 × 10−5 0.57 96.10 Potentiometric titrations Taken, mg Found, mg RSD Jump, mV Recovery, % 10.24 10.05 0.45 118.0 98.14 6.40 6.59 0.23 100.0 102.96 3.84 3.96 0.51 80.0 103.12 1.28 1.15 0.67 140.0 90.00

− − 2− 2− − mary ion over a wide concentration range. The selectivity of the For Carbon paste electrode, HSeO3 > SCN >SO4 >S >NO3 > 2− 2− − − 2− − − ionophore of the electrode sensor depends on the selectivity of SO3 > SeO4 >Br >Cl at pH 6.47, and SeO3 > SCN > > − − − 2− 2− 2− the ionophore process at the sensor-test solution interface and the NO3 >Br >Cl >SO4 > SeO4 >SO3 at pH 11.00. mobilities of the respective ions in the matrix of the sensor. The literature revealed that for developing anion selective electrodes that The selectivity sequence significantly differs from the so-called exhibit non-Hofmeister selectivity patterns, it is essential to have Hofmeister selectivity sequence (i.e. that is to say selectivity based selective interaction between the anionic species of interest and the solely on the lipophilicity of anion). Fig. 7 shows that thiocyanate lipophilic ionophore incorporated [38–40]. The use of 5,10,15,20- ions may cause some interference as the slope of its graph is near tetrakis-(4-methoxyphenyl)-21H,23H-porphyrin-Co(II) which can Nernstian but at low concentration it would not cause any inter- induce the non-Hofmeister selectivity pattern via selective axial ference, as selectivity is concentration dependent. ligation of the Co(II) and selenite ions. Two specialized IUPAC committees were held concerning the The influence of some inorganic anions on the behavior of each determination of potentiometric selectivity coefficients [41].Inthe electrode was investigated graphically by plotting the potential first IUPAC committee, held in 1975 [42], the separate solution responses for all different anionic species against their concentra- method (SSM) was recommended only if the electrode exhibits tion. As it is seen from the calibration curves (Fig. 7), except for Nernstian response, but it was considered less desirable compared selenite and thiocyanate ions, the slopes of the obtained curves to the fixed interferences method (FIM), because it does not rep- are much lower than the expected Nernstian slope and plots suf- resent as well the actual conditions under which the electrode is fer from limited linear range. However, the selenite ion resulted used. In 2002, the second IUPAC committee on methods for report- × −5 in a Nernstian response over a concentration range of 5.50 10 ing selectivity coefficients recommended the matched potential × −2 × −5 to 1.00 10 M for PVC-membrane electrode, 5.20 10 to method (MPM) which is independent of Nikolsky–Eisenman equa- × −2 1.20 10 M for CMCP electrode. This most probably due to both tion [16]. The selectivity coefficient values were calculated by the selective behavior of the ionophore towards selenite in com- applying the matched potential method and collected in Table 4. parison to other anions tested and the rapid exchange kinetics of the anion between the aqueous and membrane phase. The selectivity sequence of the studied electrodes for different 3.6. Analytical applications anions follows the order: 3.6.1. Determination of selenite in pure solution and in sodium selenite raw material powder − − − 2− − 2− For PVC membrane, HSeO3 > SCN >NO3 >S >Br >SO4 > The electrode was successfully used for the potentiometric 2− 2− 2 − − − 2− SO3 > SeO4 at pH 6.47, and, SeO3 > SCN >NO3 > SeO4 > determination of selenite in pure solution by applying standard − − 2− 2− Br >Cl >SO4 >SO3 at pH 11.00. addition method (Table 5), the recovery percents range from 95.0% H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867 865

Table 6 Statistical treatment of data obtained for the determination of selenite at pH 11.00 using selenite-selective PVC electrode in comparison with CMCP electrode and CW electrode.

Type of electrode F-test t-Test

PVC vs. CMCP 3.67 5.01 PVC vs. CW 4.23 5.46 CMCP vs. CW 2.24 4.03

Tabulated F value = 6.39. Tabulated t = 8.61.

Table 7 Statistical treatments of data obtained for the determination of selenite using PVC selenite electrode in pure and real samples in comparison with the ofﬁcial ICP method.

Sample Recovery, % F-test t-Test

ICP PVC electrode

Pure solution 97.6 97.0 1.133 4.98 Selenite raw material powder 99.0 98.9 1.012 1.56 Selenite/selenate mixture 100.5 101.0 0.909 5.28 VitaFit antioxidant tablet 99.5 98.9 1.023 2.57

to 103.0%. The electrode was also used for the determination of selenite in raw material powder Na2SeO3 with recovery percents (95.0–98.9%). Fig. 8. Potentiometric titration curves of 8 ml (a), 5 ml (b), 3 ml (c), and 1 ml (d) 0.01 M selenite in 25 ml buffer pH 6.47 with 0.01 M silver nitrate solution. 3.6.2. Determination of selenite in selenite/selenate mixture It is known that selenite and selenate are the most mobile and 3.6.4. Potentiometric titrations biogeochemically important forms of selenium. A solution was Selenite was previously determined argentometrically prepared by mixing definite amounts of selenite and selenate solu- by Masson [43]. The lower solubility of silver selenite salt tions, and the pH was adjusted with buffer of pH 6.47. The electrode (−log Ksp = 14.74–15.58) is sufficient to produce analytically useful under investigation (Table 2), composed of 2.0% CoTMeOPP and TCP titration curves [44–46]. Thus, the newly developed sensor was as plasticizer, is successfully able to determine the concentration of successfully used as indicator electrode for the titration of selenite selenite in the presence of selenate with a percentage of recovery with silver nitrate at pH 6.47. For this, 1, 3, 5, and 8 ml of 0.01 M ranging from 96.0% to 101.0% (Table 5). selenite solution were transferred into the titration cell and completed to 25 ml with buffer of pH 6.47 and titrated against 0.01 M 3.6.3. Determination of selenite in VitaFit Selenium antioxidant silver nitrate using the developed selenite-selective electrode tablets (Fig. 8). Results of the percent recovery and relative standard The tablet weighing 1.3348 g was ground well then dissolved deviation are given in Table 5. As can be seen, the jump for each in deionized water and transferred to a 250-ml volumetric mea- titration was large enough to be used as indication for the end suring flask. The potential of this solution is measured by the new point. The percent recoveries are within 90.0–103.3%. selenite-selective electrode, then standard additions are made and the potential changes are recorded. From these potential values the 3.6.5. Statistical treatment of the results quantity of selenite is calculated. The results are in a good agree- The calculated F values [47] were less than the tabulated F value ment with those obtained by inductive coupled plasma (ICP-AES) (6.39) where v1 = 4 and v2 = 4 at 95% confidence level. t-Test [47] measurements (Table 5). was also done at 99.9% confidence level (tabulated t = 8.61) and

Table 8 General performance characteristics of the previously developed selenite electrodes.

Ion recognition Slope, mV Detection limit, M Linear range, M Response Interfering ions Ref. time, s

4,6-dibromo-piaselenole −23.6 1.00 × 10−6 1.00 × 10−6 to 1.00 × 10−1 –– [8] −5 −5 −1 HgSeO3 or a mixture of HgSeO3 and Hg2Cl2 −29.5 6.00 × 10 6.00 × 10 to 1.00 × 10 –– [7] 1,2-phenylene-diamine selen complex PIS −21.0 1.00 × 10−5 1.00 × 10−5 to 1.00 × 10−1 180 Cl−,Br− [10] −5 −5 −2 2− + 2+ Ag2Se and Cu2S −28.0 1.00 × 10 1.00 × 10 to 1.00 × 10 60–120 S ,Ag ,Cu [9] Dialkyldimethyl ammonium chloride −19.0 1.00 × 10−6 1.00 × 10−6 to 1.00 × 10−1 240–300 – [11] −7 −7 −4 2− 2+ Ag2Se −28.0 4.50 × 10 6.00 × 10 to 1.00 × 10 <90 SO3 ,Ni [12]

CoTMeOPP PVC membrane (pH 6.47) −57.0 3.40 × 10−5 5.50 × 10−5 to 1.00 × 10−2 15 SCN− This work PVC membrane (pH 11.00) −26.0 4.70 × 10−5 6.60 × 10−5 to 1.20 × 10−2 15 SCN− This work Carbon paste (pH 6.47) −65.0 3.50 × 10−5 5.20 × 10−5 to 1.20 × 10−2 15 SCN− This work Carbon paste (pH 11.00) −29.4 7.70 × 10−5 8.70 × 10−5 to 1.20 × 10−2 15 SCN− This work Coated Ag wire (pH 6.47) −61.8 9.10 × 10−5 1.20 × 10−4 to 4.40 × 10−3 10 SCN− This work Coated Ag wire (pH 11.00) −39.0 3.90 × 10−6 7.90 × 10−6 to 1.30 × 10−3 10 SCN− This work 866 H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867 the results are shown in Tables 6 and 7. The method applied with [12] W. Weixiao, L. Jingyin, D. Huimin, H. Jing, L. Shufang, Determination of selenium the use of constructed electrode does not exhibit significant differ- in biological materials using an ion-selective electrode, Mikrochim. Acta 154 (2006) 143–147. ences in comparison to with the official method which reflects the [13] V.K. Gupta, S. Chandra, D.K. Chauhan, R. Mangla, Membranes of 5,10,15,20- 2− accuracy and precision of this method. tetrakis(4-methoxyphenyl) porphyrinatocobalt (TMOPP-Co) (I) as MoO4 – selective sensors, Sensors 2 (2002) 164–173. [14] A.I. Vogel, Quantitative Inorganic Analysis, 18, 1st ed., Longman, London, S.E., 4. Conclusion 1939. [15] R.G. Bates, Determination of pH Theory and Practice, 2nd ed., John Wiley, New York, 1962, p. 161. New selenite-selective electrodes that are highly selective and [16] Y. Umezawa, K. Umezawa, P. Buhlmann, N. Hamada, H. Aoki, J. Nakanishi, M. sensitive to selenite have been developed using CoTMeOPP as elec- Sato, K. Ping Xiao, Y. Nishimura, Potentiometric selectivity coefficients of ion- selective electrodes, Pure Appl. Chem. 74 (2002) 923–994. troactive material. Experiments show that the PVC based electrode [17] H. Ibrahim, Y.M. Issa, H.M. Abu-Shawish, Improving the detection limits of anti- containing 2.0% CoTMeOPP, and TCP as plasticizer exhibits lin- spasmodic drugs electrodes by using modified membrane sensors with inner − solid contact, J. Pharm. Biomed. Anal. 44 (2007) 8–15. ear response for the monovalent selenite ion HSeO3 with slope − [18] E. Lindner, Y. Umezawa, Performance evaluation criteria for preparation and 57.00 mV at pH 6.47 and linear response for the divalent selen- measurement of macro- and microfabricated ion-selective electrodes (IUPAC 2− ite SeO3 with slope −26.00 mV at pH 11.00. The detection limit Technical Report), Pure Appl. Chem. 80 (2008) 85–104. is 3.40 × 10−5, and 4.70 × 10−5 M at pH values 6.47, and 11.00 [19] A.F. Holleman, E. Wiberg, Inorganic Chemistry, Academic Press, San Diego, 2001. respectively and the response time is within 5–15 s. The carbon [20] Y. Masuda, Y. Zhang, C. Yan, B. Li, Studies on the extraction and sepa- paste electrode exhibits linear response for monovalent selenite ration of lanthanide ions with a synergistic extraction system combined of slope −65.00 mV with detection limit 3.50 × 10−5 M at pH 6.47 with 1,4,10,13-tetrathia-7,16-diazacyclooctadecane and lauric acid, Talanta 46 and linear response for divalent selenite of slope −29.40 mV with (1998) 203–213. [21] M. Shamsipur, M. Javanbakht, M.F. Mousavi, M.R. Ganjali, V. Lippolis, A. Garau, L. −5 detection limit 7.70 × 10 M at pH 11.00. Table 8 compares the Tei, Copper(II) selective membrane electrodes based on some recently synthe- slope, response time, detection limit, and major interfering ions of sized aza-thioether crowns containing a 1,10-phenanthroline sub-unit, Talanta the proposed selenite electrodes with those of selenite-selective 55 (2001) 1047–1054. [22] A.R. Fakhari, M.R. Ganjali, M. Shamsipur, PVC-Based hexathia-18-crown-6- electrodes reported in the literature. We can observe that, our tetraone sensor for mercury(II) ions, Anal. Chem. 69 (1997) 3693–3696. electrodes showed larger slope, faster response time than those [23] D. Ammann, E. Pretsch, W. Simon, E. Lindner, A. Bezegh, E. Pungor, Lipophilic reported in the literature [8,10,11], while it has narrower linear salts as membrane additives and their influence on the properties of macro- and micro-electrodes based on neutral carriers, Anal. Chim. Acta 171 (1985) range. 119–129. From the performance characteristics of the studied PVC, MCPE, [24] M.R. Ganjali, M. Yousefi, M. Javanbakht, T. Poursaberi, M. Salavati-Niasari, L.H. − and coated wire electrodes for selenite, it can be concluded that Babaei, E. Latifi, M. Shamsipur, Determination of SCN in urine and saliva of smokers and non-smokers by SCN−-selective polymeric membrane containing they have closely similar characteristics, namely the linear con- a nickel(II)-azamacrocycle complex coated on a graphite electrode, Anal. Sci. centration range, working pH range, response time, life span, etc. 18 (2002) 887–892. The analytical applications involve determination selenite in pure [25] M.R. Ganjali, T. Poursaberi, F. Basiripour, M. Salavati-Niasari, M. Yousefi, M. Shamsipur, Highly selective thiocyanate poly(vinyl chloride) membrane elec- solution, sodium selenite raw material powder, selenite/selenate trode based on a cadmium-Schiff’s base complex, Fresenius J. Anal. Chem. 370 mixture, and VitaFit Selenium antioxidant tablets. The recovery (2001) 1091–1096. ranges from 95.0% to 103.3%. Standard addition method was proved [26] M.R. Ganjali, T. Poursaberi, M. Hosseini, M. Salavati-Niasari, M. Yousefi, M. to be successful for the determination of selenite ions in pure solu- Shamsipur, Highly selective iodide membrane electrode based on a cerium salen, Anal. Sci. 18 (2002) 289–292. tion, and real samples using the proposed electrodes as sensors. [27] S. Yang, N. Kumar, H. Chi, D.B. Hibbert, P.W. Alexander, Lead-selective The electrode was successfully used as indicator electrode for the membrane electrodes based on dithiophenediazacrown ether derivatives, Elec- titration of selenite with silver nitrate at pH 6.47. troanalysis 9 (1997) 549–553. [28] M.A.A. Perez, L.P. Marin, J.C. Quintana, M.Y. Pedram, Influence of different plasticizers on the response of chemical sensors based on polymeric membranes for nitrate ion determination, Sens. Actuators B 89 (2003) 262–268. References [29] U. Fiedler, J. Ruzicka, Selectrode the universal ion-selective electrode: Part VII. A valinomycin-based potassium electrode with nonporous polymer membrane [1] K. Pyrzynska, Analysis of selenium species by capillary electrophoresis, Talanta and solid-state inner reference system, Anal. Chim. Acta 67 (1973) 1791–2193. 55 (2001) 657–667. [30] A. Cadogan, Z. Gao, A. Lewenstam, A. Lvaska, D. Diamond, All-solid-state [2] EPA U.S. Environmental Protection Agency, Aquatic Life Water Quality Criteria sodium-selective electrode based on a calixarene ionophore in a poly(vinyl for Selenium, Office of Science and Technology, Washington, DC, 2004. chloride) membrane with a polypyrrole solid contact, Anal. Chem. 64 (1992) [3] J. Qiu, Q. Wang, Y. Ma, L. Yang, B. Huang, On-line pre-reduction of Se(VI) by 2496–2501. thiourea for selenium speciation by hydride generation, Spectrochim. Acta B [31] J. Bobacka, T. Lindfors, M. McCarrick, A. Lvaska, A. Lewenstam, Single-piece 61 (2006) 803–809. all-solid-state ion-selective electrode, Anal. Chem. 67 (1995) 3819–3823. [4] D.C. Kapsimali, G.A. Zachariadis, Comparison of tetraethylborate and [32] T. Lindfors, P. Sjoberg, J. Bobacka, A. Lewenstam, A. Lvaska, Characterization tetraphenylborate for selenite determination in human urine by gas chro- of a single-piece all-solid-state lithium-selective electrode based on soluble matography mass spectrometry, after headspace solid phase microextraction, conducting polyaniline, Anal. Chim. Acta 385 (1999) 163–173. Talanta 80 (3) (2010) 1311–1317. [33] T. Lindfors, J. Bobacka, A. Lewenstam, A. Lvaska, Study on soluble polypyr- [5] P. Van Dael, J. Lewis, D. Barclay, Stable isotope-enriched selenite and sele- role as a component in all-solid-state ion-sensors, Electrochim. Acta 43 (1998) nate tracers for human metabolic studies: a fast and accurate method for their 3503–3509. preparation from elemental selenium and their identification and quantifica- [34] I. Svancara, K. Vytras, K. Kalcher, A. Walcarius, J. Wang, Carbon paste electrodes tion using hydride generation atomic absorption spectrometry, J. Trace Elem. in facts, numbers, and notes: a review on the occasion of the 50 years Jubilee of Med. Biol. 18 (2004) 75–80. carbon paste in electrochemistry and electroanalysis, Electroanalysis 21 (2008) [6] M. Gallignani, M. Valero, M.R. Brunetto, J.L. Burguera, M. Burguera, Y. Petit de 7–28. Pena,˜ Sequential determination of Se(IV) and Se(VI) by flow injection-hydride [35] K. Kalcher, J.M. Kauffmann, J. Wang, I. Svancara, K. Vytras, C. Neuhold, Z. Yang, generation-atomic absorption spectrometry with HCl/HBr microwave aided Sensors based on carbon paste in electrochemical analysis: a review with par- pre-reduction of Se(VI) to Se(IV), Talanta 52 (2000) 1015–1024. ticular emphasis on the period 1990–1993, Electroanalysis 7 (1995) 5–22. [7] M.M. Zareh, A.S. Amin, A.A. Manar, New polycrystalline solid state responsive [36] I. Svancara, K. Vytras, J. Barek, J. Zima, Carbon paste electrodes in modern electrodes for the determination of the selenite ion, Electroanalysis 7 (1995) electroanalysis, Crit. Rev. Anal. Chem. 31 (2001) 311–346. 587–590. [37] J. Sutter, A. Radu, S. Peper, E. Bakker, E. Pretsch, Solid-contact polymeric mem- [8] Q. Cai, Y. Ji, W. Shi, Y. Li, Preparation and application of selenite ion selective brane electrodes with detection limits in the subnanomolar range, Anal. Chim. electrode, Talanta 39 (1992) 1269–1272. Acta 523 (2004) 53–59. [9] G. Ekmekci, G. Somer, Preparation and properties of solid state selenite ion [38] K. Fukushi, K. Tada, S. Takeda, S. Wakida, A. Yamane, K. Higashi, K. Hiiro, Simul- selective electrodes and their applications, Talanta 49 (1999) 91–98. taneous determination of nitrate and nitrite ions in seawater by capillary zone [10] G. Ekmekci, G. Somer, A new selenite selective membrane electrode and its electrophoresis using artificial seawater as the carrier solution, J. Chromatogr. application, Talanta 49 (1999) 83–89. A 838 (1999) 303–311. [11] G. Ekmekci, G. Somer, Selenite-selective membrane electrodes based on ion [39] J.E. Melanson, C.A. Lucy, Ultra-rapid analysis of nitrate and nitrite by capillary exchangers and application to anodic slime, Anal. Sci. 16 (2000) 307–311. electrophoresis, J. Chromatogr. A 884 (2000) 311–316. H. Ibrahim et al. / Journal of Hazardous Materials 181 (2010) 857–867 867

[40] M.M.G. Antonisse, B.H.M. Snellink-RuD el, J.F.J. Engbersen, D.N. Reinhoudt, [43] M.R. Masson, Microdetermination of selenium, tellurium, and arsenic in organic Chemically modified field effect transistors with nitrite or fluoride selectivity, compounds, Mikrochim. Acta 65 (1976) 399–411. J. Chem. Soc. [Perkin II] 4 (1998) 773–777. [44] E.E. Chao, K.L. Cheng, Stepwise titration of some anion mixtures and determina- [41] Y. Umezawa, P. Buhlmann, K. Umezawa, K. Tohda, S. Amemiya, Performance tion of Ksp of silver precipitates with silver ion selective electrode, Anal. Chem. evaluation criteria for preparation and measurement of macro- and microfab- 48 (1976) 267–271. ricated ion-selective electrodes (IUPAC Technical Report), Pure Appl. Chem. 72 [45] L.G. Sillen, A.E. Martetl (compilers), Stability Constants of Metal–Ion Complexes, (2000) 1851–2082. Special Publication No. 17. The Chemical Society, London, (1964). [42] G.G. Guilbault, R.A. Drust, M.S. Frant, H. Freiser, E.H. Hansen, T.S. Light, E. Pungor, [46] E. Högfeidt (compiler), Stability Constants of Metal–Ion Complexes. Part A: G.A. Rechnitz, N.M. Rice, T.J. Rohm, W. Simon, J.D.R. Thomas, Recommendations Inorganic Ligands. Pergamon Press, Oxford, (1982). for nomenclature of ion selective electrodes (Recommendations 1975), Pure [47] G.D. Christian, Analytical Chemistry, 5th. ed., John Wiley, USA, 1994. Appl. Chem. 48 (1976) 127–132.