Direct Determination of Arsenic in Sea Water by Electrothermal. Atomization Atomic Absorption Spectrometry using D2 and Zeeman Background Correction.
Mikrochim. Acta 128, 215-22l (1998)
Mikrochimica Acta 9 Springer-Verlag 1998 Printed in Austria
Direct Determination of Arsenic in Sea Water by Electrothermal Atomization Atomic Absorption Spectrometry using D2 and Zeeman Background Correction Pilar Bermejo-Barrera*, Jorge Moreda-Pifieiro, Antonio Moreda-Pifieiro, and Adela Bermejo-Barrera Department of AnalyticalChemistry,Nutrition and Bromatology,Facultyof Chemistry,Universityof Santiago de Compostela, 15706 Santiago de Compostela,Spain
Abstract. Arsenic in sea water was determined directly by graphite furnace atomic absorption spectrometry (GFAAS) using palladium nitrate as chemical modifier, at an optimum concentration of 15 mg 1-1. Deuterium and Zeeman effect background correction were compared and gave detection limits of 0.6 and 0.8 gg 1-1, respectively. Precisions between 8 and 2%, for both correctors, were obtained with an injection volume of 40 gl. The accuracy obtained with different reference materials: CRM-403 (1.461 gg kg-1), NASS-4 (1.26 4- 0.09~tg1-1) and IAEA/W-4 (2431 gg1-1) was studied for large injection volumes for both background correction systems. Interferences by chloride, sodium, potassium, calcium and silicon were removed by Zeeman correction, whereas deuterium correction was much less effective and was insufficiently accurate for sea water samples. Key words: arsenic, sea water, deuterium and Zeeman effect background correction,ET-AAS. The interest in arsenic in sea water is due to the requirement contained in European Directive 79/923/ CEE [1] relative to the quality of water used to store molluscs. Atomic absorption spectrometry with electrothermal atomization (ET-AAS) has been used successfully for the direct determination of arsenic in sea water [2-4]. Due to the high volatility of arsenic and the important interferences of aluminium, sodium, potassium and sulphate [5, 6], a chemical modifier is required to reach satisfactory stabilization * To whomcorrespondenceshouldbe addressed
of arsenic at high temperatures and to reduce the interferences. The use of palladium nitrate appears to be advantageous [3, 7, 8-14]. To obtain good sensitivity for the determination of arsenic by ET-AAS, a preconcentration step is required to measure the content in water samples. Some authors use extractive procedures for arsenic in sea water by complexation of arsenic with ammonium pyrrolidinedithiocarbamate and extraction of the complex in IBMK [15, 16], toluene [14], carbon tetrachloride and chloroform (1:1) [17], chIoroform and re-extraction in water [18] and with ammonium diethylphosphorodithionate and extraction in chloroform and re-extraction in water [18]. Other authors use co-precipitation methods by using Fe203 [19], ammonium pyrrolidinedithiocarbamate [20], zirconium hydroxide [21] or sodium tetrahydroborate [22] and redissolution of the precipitate formed in sulfuric acid, IMBK and hydrochloric acid. Different cation-exchange resins [7, 23-26] to preconcentrate traces of arsenic in sea water samples were also used. However, these preconcentration methods present some disadvantages, as the procedures are laborious and there is possible loss of anaiyte and sample contamination. In the present work, to achieve a lower detection limit, a large volume of sample was injected into the graphite tube. However, the use of large volumes produces an important background signal that the deuterium background correction cannol: easily correct. Thus, a comparative study between deuterium and Zeeman effect background correction systems has
216
P. Bermejo-Barrera et al.
b e e n investigated for the direct determination of arsenic in sea water samples. I n addition, a study of the interfering behaviour of m a j o r and m i n o r constituents of the sea water was carried out using both b a c k g r o u n d correction systems.
Experimental
Communities (Community Bureau of Reference) were used. Others reference materials: IAEA/W-4 (simulated fresh water, supplied by Imernational Atomic Energy Agency, with a certified value of 24-31 ~tg1-1 As) and SLRS-2 (reverine water, supplied by the National Research Council of Canada, with a certified value of 0.77 • 0.09 ~tgl-~ As) were used also. Argon N-50 purity (99.999%) used as sheath gas for the atomizer and to purge internally was obtained from SEO (Madrid, Spain).
Apparatus
Sample Collection
A Perkin-Elmer 1100B atomic absorption spectrometer equipped with a deuterium lamp for background correction, an HGA-700 graphite furnace atomizer and an AS-70 autosampler, and a Perkin-Elmer Zeeman 4100ZL atomic absorption spectrometer equipped with a THGA graphite furnace and an AS-71 autosampler were used. The source of radiation was an arsenic electrodeless discharge lamp operated at 8 W, which provides a 193.7nm line. The spectral bandwidth used was 0.7rim. An integrated absorbance signal was used throughout unless otherwise indicated. Pyrolytic-graphite-coated graphite tubes and L'vov platforms and THGA graphite tubes with integral platforms were used. The injection volumes were 20 and 40 [xl. The graphite furnace programme is shown in Table 1.
Sea water samples were collected from coastal surface water of Galicia (north-west Spain) in 100-ml glass bottles. The samples were immediately acidified with 100 ~tl of concentrated nitric acid [5], giving pH < 1.6, avoiding the adsorption of arsenic onto the glass bottle walls.
Reagents All the solution were prepared from analytical-reagent grade chemicals using ultrapure water, resistivity 18 Mf~ cm, which was obtained by using a Milli-Q water purification system (Millipore). Arsenic(HI) chloride stock standard solution, 1.000g1-1, supplied by BDH Chemicals Ltd (Poole, U.K.), was used. Palladium stock standard solution, 3.000g1-1, from palladium supplied by Aldrich, was prepared according to the procedure of Welz et al. [27]. Ammonium nitrate stock standard solution, 1.000g1-1, was prepared from ammonium nitrate (Merck Suprapur, Darmstadt, Germany). Two synthetic sea water samples (SSW) were prepared [28-30] with the following components (all Merck): SSWI, 30 g 1-1 sodium chloride and 10gl -I each of magnesium chloride, potassium chloride, calcium chloride and strontium chloride, giving a salinity of 72.8%0, and SSWII, 32g1-1 sodium chloride, 14g1-1 magnesium sulfate heptahydrate and 0.15 g 1-1 sodium hydrogen carbonate, giving a salinity of 34.2%0. Different sea water reference materials (NASS-4 open sea water, salinity 31.3 %o),supplied by the National Research Council of Canada, with a certified value of 1.26 + 0.09gg1-1 As and CRM-403 sea water (salinity 34.6-34.8%0) with an indicated value of 1461 ngkg -1 As, supplied by the Commission of the European
Table 1. Optimized graphite furnace temperature programme Step
T e m p e r a t u r e Ramp (~ (s)
Hold (s)
Argon flow (ml rain-1)
Drying Charring Atomization Cleaning
150 1200 2100" 2500
15 30 3 3
300 300 0 (read) 300
10 20 0 1
* 1800 ~ when Zeeman effect background corrector was used.
Procedure A portion of the sea water sample, 400 ~tl, with appropriate volumes of palladium solution to give concentrations of 15 mg 1-1 Pd, was transferred into the autosampler cup, then it was brought to a volume of 1 ml and stirred before measurement. A volume of 40 ~tl was injected into the atomizer, using a hot injection technique (120~ A sequential drying-ashing-atomization-cleaning programme (Table 1) for the graphite furnace was followed.
Results and Discussion
Optimum Conditions Graphite furnace program. A b l a n k sea water sample (SSWII) spiked with 60 ~g 1-1 of arsenic was used to find the
o p t i m u m graphite
furnace
temperature
program. The injection v o l u m e of the sea water sample, was 20 pl. D e u t e r i u m and Z e e m a n effect b a c k g r o u n d correction were used to optimize the graphite furnace temperature programme. The absorbance profiles of an aqueous standard solution, synthetic sea water of high (72.8 %o) and low (34.2 %0) salinity and a real sea water sample as a f u n c t i o n of the charring temperature were studied (Fig. 1). The atomization temperature was kept at 2200 ~ for both b a c k g r o u n d correctors. For charring temperatures below 1 0 0 0 ~ scatter on the arsenic absorbance signal was obtained, due to insufficient removal of the b u l k salts of the sea water matrix. W h e n the charring temperature was increased, the b a c k g r o u n d signal decreased and the absorbance signal was kept up to a temperature of 1 2 0 0 ~ for Z e e m a n effect and d e u t e r i u m b a c k g r o u n d correction and for all kinds of sea water samples (Fig. 1). The ramp rate and hold times were optimized b y using different c o m b i n a t i o n of ramp rate times ( 5 - 4 5 s) and
Direct Determination of Arsenic in Sea Water
217
0.25
(B).
0.700
o.~
~.._.....'_,.:.,,
:,,...;,,~-,:::...............~ ~ ..... -~> x
0 o? o.15
Q) o C m ..Q e.._
/ j ,, \\\
0
/
,J~
