Direct determination of arsenite by differential pulse polarography in

---

393

Anal. Chem. 1987, 59, 393-395

Direct Determination of Arsenite by Differential Pulse Polarography in the Presence of Lead(I I) and Thallium(I) Margaret A. Reed' and Richard J. Stohberg* Department of Chemistry, University of Alaska, Fairbanks, Alaska 99775-0520

Interference from Pb(I1) and TI(1) In the differential pulse polarographic determination of aresenlte Is eliminated by chromatography on a chelating ion exchange resin. Strong ligands prevent the removal of Pb, but addltlon of Cu( I I ) before chromatography results in successful analysis by dlssoclatlng the Pb complex. Slnce the Interfering Ions are removed from solutlon, greater than a 1000-fold mass excess of Pb and TI can be tolerated.

Table I. Calibration Curve Parameters for As(II1) after Chromatography CPb,ppm

slopea

SD slope

0 0.1

1.21

0.05 0.05

1.0

1.18

10

1.18 1.19

0.04 0.03

intercepta SD intercept -1 3 7 1

6 6 5 4

Plot of current (nA) vs. vvb As. Differential pulse po!arography (DPP) has been used for direct determination of arsenite in the presence of other arsenic species (I). Detection limits lie in the low parts-perbillion range, but there is significant interference due to overlapping polarographic waves of Pb, Tl, and Sn (2). These interferences can be accounted for in favorable cases by oxidation of As(II1) to polarographically inactive As(V) with Ce(1V) (2,3). When the concentration of As(II1) is very small, uncertainty in base line position is a limiting factor (2). In addition, when the size of the signal due to Pb, T1, and Sn is large compared to the As(II1) signal, precision and accuracy are poor since the arsenic concentration is calculated as a small difference between two large numbers. A method based on removal of P b and T1 by chromatography on a chelating ion exchange resin is presented. It is similar to one described by Hamilton et al. ( 4 ) for determination of As and Sn in copper by anodic stripping voltammetry. We show here that the presence of strong ligands interferes with P b removal, but this interference can be eliminated by addition of Cu(I1) prior to chromatography. Since the method leaves the As(II1) in solution and does not affect the polarographic base line, it can be used when the ratio of interfering ion to As(II1) is very large, with little or no degradation in detection limit.

EXPERIMENTAL SECTION Differential pulse polarograms were obtained with a Princeton Applied Research Model 114A polarographic analyzer, 174-70drop knocker, and dropping mercury electrode in a VC-2 polarographic cell (Bioanalytical Systems). Reference and counter electrodes were Ag/AgC1(3 M Cl-) and platinum, respectively. Polarographic conditions were 5 mV/s scan rate, 1 s drop time, and 50 mV modulation amplitude. Samples were deoxygenated for 10 min with nitrogen which had been passed through a vanadium(I1) scrubber and deionized water. Standard lo00 ppm As(II1) was prepared from reagent grade As203 (Fisher). Ethylenediaminetetraacetate (EDTA, Baker), nitrilotriacetate(NTA,Baker), and diethylenetriaminetetraacetate (DTPA, LaMont) solutions (0.1 M) were prepared from reagent grade acids (NTA, DTPA) or salts (Na2EDTA.2H20) and standarized by titration with Cu(II), using a Cu Selectrode (Radiometer) for end point detection. Copper standards (1000 ppm) were prepared with 99.9% (minimum) Cu foil (MCB) and "OB. All other solutions were prepared from reagent grade salts in distilled, deionized water. Working solutions containing

+

0 1

w

[1:

50

X

P

+

a S

o

25

x 0

0

4

5

6

7

PH

Flgure 3. Effect of pH on dissociation of PbEDTA by Cu. Chromatography was done 0.5 h (0),2 h (X), and 9 h (+) after Cu was added. Pb was measured by DPP peak current at -0.58 V. Initiil C , = 2.5 pM, CEDTA = 5 pM, and CAS= 200 ppb.

Table 111. Analysis of Fortified Tap and Well Water"

sample house tap lab tap well well + Pb + Tlb

concn of As(III), pg/L no Cu added Cu added 65, 54 45,44 62, 65 53,44

54 54, 58 52, 58 58, 55, 52, 58

"53 ppb As(II1) added to each sample. All samples were chromatographed. * 1 ppm Pb and 1.2 ppm T1 added before analysis.

pH and complete removal of Pb requires approximately 2 h of reaction with Cu at pH 5.5 (Figure 3). Similar experiments show a rapid rate of reaction between Cu and PbDTPA (111 min is sufficient) at pH 4.1. Dissociation of the PbDTPA requires 20-40 min when Hg(I1) is added at pH 3.7 and considerably longer at higher pH. Removal of Tl(1) is barely affected by strong ligands. In the presence of 6 pM EDTA or DTPA, 92% of 5 pM Tl(1) is taken up by the Chelex. NTA (6 pM) has no effect. Addition of 50 pM Cu eliminates the small bias caused by EDTA and DTPA. Stannous ion is incompletely removed by Na Chelex. In the pH range 4.7-9.3, 65-80% of 1 ppm Sn(I1) is removed.

395

At pH 3.0, only 25% is removed. Study of the system is difficult due to oxidation to Sn(IV), precipitate formation, and time-dependent behavior. The chromatographic method is particularly well suited for rapid determinations of arsenite in the presence of large concentrations of Pb or T1 since the measurement of arsenic is direct, rather than by difference. Sixteen duplicate samples plus blanks and standards can be analyzed for As(II1) in approximately 6 h, including 1 h of equilibration. The ratelimiting step is deoxygenation and polarography since chromatography requires 3.5 min per sample when two columns are run simultaneously. Fortified tap and well water samples were analyzed successfully (Table III). The presence of 1 ppm P b and 1.2 ppm T1 added to the very hard well water does not adversely affect the analysis. The standard deviation of the method, based on these eight pairs of data, is 4 ppb. By use of Winefordner's suggested method ( I I ) , the detection limit is 12 ppb and the limit of quantification is 40 ppb. The detection limit is comparable to that reported by Thorpe and co-workers (I, 3). Arsenite concentrations in samples containing 2 4 0 ppb As(II1) can be determined confidently, even in the presence of greater than a thousand-fold mass excess of Pb(I1) and T U . Registry No. Pb, 7439-92-1; T1, 7440-28-0; Cu, 7440-50-8; water, 7732-18-5; arsenite, 15502-74-6.

LITERATURE CITED (1) Henry, F. T.; Thorpe, T. M. Anal. Chem. 1880. 52, 80-83. (2) Myers, David J.; Osteryoung, Janet Anal. Chem. 1973, 45, 267-271. (3) Henry F. T.; Klrch, T. 0.; Thorpe, T. M. Anel. Chem. 1979, 57, 2 15-2 18. (4) Hamilton, T. W.; Ellis, J.; Florence, T. M. Anal. Chim. Acta 1980, 719, 225-233. (5) Marteli, Arthur E.: Smith, Robert M. Critical Stabiliry Constants; PI, num: New York, 1974; Volume 1: Amino Acids. (6) Ramette. Richard W. Chemical Equilibrium and Analysis; AddisonWesley: Reading, MA, 1981; Appendix 5. (7) Stumm, Werner: Morgan, James J. Aquatic Chemistry, 2nd ed.;WlleyInterscience: New York, 1981; p 135. (8) Reed, M. A. M.S. Thesis, University of Alaska, Fairbanks, AK. Dec 1985. (9) Howard, A.; Arbab-Zavar, M. AneMst (London) 1981, 706, 213-220. (10) Holm, Thomas A.; Anderson, Marc A.; Iverson, Dennis G.; Stanforth, Robert S. Chemical Modeling In Aqueous Systems; Jenne, Everett A., Ed.; American Chemical Society: Washlngton, DC, 1979; p 711. (11) Long, Gary L.; Winefordner, J. D. Anal. Chem. 1983, 55, 712A724A.

RECEIVED for review June 9,1986. Accepted October>, 1986. A portion of this work was funded by a Grant from the Corvallis Lab, USEPA, IAG No. DW14931442-01-0.