Single-column ion-chromatographic determination of chromium(VI) in aqueous soil and sludge extracts

Talanta, Vol. 36, No. 9, pp. 889-892, 1989 Printed in Great Britain. All rights reserved

0039-9140/89 $3.00 + 0.00 Copyright 0 1989 Pergamon Press plc

SINGLE-COLUMN ION-CHROMATOGRAPHIC DETERMINATION OF CHROMIUM(V1) IN AQUEOUS SOIL AND SLUDGE EXTRACTS

Department

H. C. MEHRA and W. T. FRANKENBRRGFR, JR.* of Soil and Environmental Sciences, University of California, Riverside, California 92521,

U.S.A. (Received 9 November 1988. Revised 9 February 1989. Accepted 9 May 1989)

Stmrmary-Single-column ion-chromatography (SCIC) was investigated as a routine, rapid, precise and selective analytical method for the determination of chrornium(V1)in aqueous extracts of soil and sewage sludge. Chromatographic parameters were optimized for determination of Cr(VI), NO, and S@-. A low-capacity resin-based column was used for the separation and the anions were determined by conductometric detection. p-Hydroxybenzoic acid (5mM) at pH 8.5 was used as the eluent. The limit of detection, defined as S/N = 3, was 92 pgll. The resolution between Cr(V1)and S@- was 2.8, the precision ranged from 0.9% for NO; to 2.0% for Cr(V1)with a 500-pl injection. The SCIC results for Cr(VI) agreed closely with those obtained by inductively coupled argon-plasma emission and spectrophotometry.

The terrestrial abundance of chromium in igneous and sedimentary rocks typically ranges from 5 to 120 pg/g.’ Chromium(III), which is essential in human nutrition, is less toxic and also less mobile in the soil environment than chromium(VI).2 Chromium finds its way into soils through industrial wastes such as electroplating sludge, tannery wastes, manufacture of corrosion inhibitors, and municipal sewage sludges. The World Health Organization (WHO) has set a 50 pg/l. standard for chromium in drinking water.3 Analytical methods currently used for the trace determination of Cr in environmental samples include visual spectrophotometry,’ atomic-absorption spectrometry (AAS),5*6inductively coupled plasma emission (ICP),’ polarography,* and suppressed ion-chromatography (SIC).9*10Many of these methods involve organic extraction, and/or reduction and are subject to interferences by other ions. lo Furthermore, AAS and ICP are not selective for speciation of chromium. Ion-chromatography (IC) is an invaluable technique that is becoming very popular for determination of anions in aqueous matrices. Its versatility stems from the wide choice of eluent composition, pH, flow-rate and method of detection, and IC can be very sensitive and selective. Suppressed ionchromatography (SIC) has been used for the detection of Cr(VI)9*‘o but the chromatographic peaks obtained were very broad, which made accurate quantification difficult. The objective of this study was to develop a novel single-column ion-chromatographic (SCIC) method for the on-line determination of Cr(V1) together with other important ions such as nitrate and sulphate which are inherently present in soil and sewage sludge *Author for correspondence. 889

samples. This study provides a routine analytical method for the determination of trace levels of Cr(V1) with good accuracy and precision. EXPERIMENTAL. Single-column ion-chromatography apparatur‘

A schematic representation of the SCIC system was shown in a previous paper.” The HPLC analysis was performed on a Beckman HPLC Model 332 liquid chromatograph, equipped with a Model 1lOA pump and a Model 210 sample injector. Conductometric detection was carried out with a Wescan (San Jose, CA) Model 213 detector. A Hewlett-Packard Model 3390A printer-plotter integrator with variable input voltage was used to monitor signal output with a chart speed of 0.5 cm/mm. The analytical column consisted of a low-capacity anion/R (Wescan 269-029) resin-based anion-exchange column (250 x 4.6 mm). A Wescan guard column (40 x 4.6 mm) packed with pellicular anion-exchange material (269-003) was attached before the analytical column, with zero deadvolume fittings. To prevent drift in conductance because of temperature variation, the column was insulated with a column heater (Eldex Laboratories, Menlo Park, CA, Model III) and maintained at 25”. Samule iniection 100~s of 100,500 and 2000 pl volume were used in testing the 6mit.s of detection. The mobile phase consisted of p-hydroxybenxoic acid (PHBA) (Sigma, St. Louis, MO) solutions, 3.0-7.OmM, adjusted to pH 7.5-10.0 with sodium hydroxide. An “ascarite” tube was fixed over the flask containing the mobile phase in order to absorb C02. The flow-rate was 2 ml/min, the column inlet pressure was _ 70 bar (1000 psig) and the detector output was 10 mV. Column conditioning procedures are given elsewhere.” Reagents

Analyte solutions were prepared by dissolving sodium chloride (Mallinckrodt, St.-L&is, Md), potas&n nitrate (Aldrich, Milwaukee, WI), potassium sulphate (Baker, Phillipsburg, NJ), and potassium chromate (Mallinckrodt) in HPLC-grade water obtained by filtering demineralized

890

H. C. MEHRAand W. T. FRAN~CENBERGER, JR,

water successively through an HN organic removal resin (Bamstead, Boston, MA), an HN Ultrapure DI exchange column (Bamstead), and a 0.22~pm membrane GS filter (Millipore, Bedford, MA). The Cr(V1) standards were prepared daily from a standard stock solution. All analytes were determined on an elemental basis. Field samples and their preparation The sewage sludge samples were collected from various treatment plants in California. These samples were dried at 65”, ground (to c2 mm) and then stored at room temperature. The surface soil samples were collected from a depth of O-l 5 cm, air-dried and sieved (c 2 mm). Soil and sewage sludge extracts were prepared by shaking 50 ml of demineralized water with 5 g of the air-dried sample for 2 hr, and filtering the suspension (Whatman No. 42 filter paper). One sludge sample used in this study had a relatively high sulphate content, part of which was removed by addition of 5% aqueous barium acetate solution under acidic conditions and filtration; the flltrate was alkalized with 3M sodium hydroxide, heated at 70” for 2 hr, then stirred and filtered to remove any precipitate which might have formed. All sample extracts were slowly passed through a Supelclean LC-Si tube (Supelco Park, Bellefonte, PA) under positive pressure, at a flow-rate of 2 ml/min. This step aids in the removal of organic impurities from the sample. After suitable dilution the extract was passed through a 0.22~pm Millinore GS filter before the SCIC analvsis. Td check the reproducibility of the meihod, solutions of combined standards were injected a minimum of 10 times. Detection limits obtained with various injection volumes were calculated as the concentration equivalent to three times the baseline noise (S/N = 3). Spectrophotometry The absorbances of standard and sample solutions were measured at 540 nm with a Bausch and Lomb Spectronic 20 spectrophotometer. The diphenylcarbazide method’* was used. Inductively coupled argon-plasma emission spectrometry

A Jarrell-Ash Atom Comp 800 ICAP was used to confirm Cr(V1) detection. The wavelength was 267.7 nm and a Fassel type torch was used with a forward power of 1.75 kW. The viewing height was 13 mm above the coil and the flow-rate of the Ar coolant gas was 14.1 ml/min. The preintegration and integration times were each 17 sec. RESULTS AND Influence of eluent pH resolution

7

8

9

IO

II

pH OF MOBILE PHASE

Fig. 1. Effect of eluent pH (5mM PHBA) on the capacity factor (k’). when the pH was increased from 7.5 to 10.0, whereas there was only a slight decline in the k’ values for Cl- and NO,. The resolution, R,, for So’and Cr(V1) was >2.5 in the pH range 8.5-9.5. Increasing the pH of the mobile phase alters the degree of ionization of PHBA, which loses a second proton at pH >/ 8.5. Further studies were made of the concentration of the mobile phase. Figure 2 shows that k’ for the anionic species tested decreased with increasing eluent concentration. With 5-7mM PHBA, R, for Cr(VI)/SO:- was fairly constant (2.8-3.2). The optimum mobile phase chosen for detection of Cr(V1) was 5mM PHBA adjusted to pH 8.5 with sodium hydroxide. was observed

DISCUSSION

and concentration on anion

The choice of eluent and working pH was important in optimizing separation of Cr(V1) from other

ions inherently associated with soils and sewage sludge. Selection of the eluent was based not only on resolution but also on signal response and analysis time. Among the eluents tested (sodium hydroxide, sodium benzoate, PHBA, potassium hydrogen phthalate, and phthalic acid solutions), PHBA (p-hydroxybenzoic acid) provided the best overall chromatographic conditions. With PHBA the capacity factor (k’) of the most strongly retained ion, Cr(VI), was less than 10. The analysis time was also shorter than that obtained with the other eluents and peak broadening was negligible. Figure 1 illustrates effect of mobile phase pH on the k’ values for Cl-, NO,, SOi- and Cr(V1). A dramatic decrease in retention of Cr(V1) and SOi-

k’

CONCENTRATION OF MOBILE PHASE (mbl)

Fig. 2. Effect of eluent concentration (pH = 8.5) on the capacity factor (k’).

Determination

of chromium(VI) in soil and sludge extracts

Precision

The precision of the proposed method, determined by repeated injections of combined standards, is given in Table 1. The results show that the relative standard deviations (RSD) for all the aualytes tested ranged from 0.9 to 2.0% with a 500~~1 injection. Precision with the 2-ml injections ranged from 3.2 to 7.4%.

Limits of detection (LOD, S/N = 3) for various sample sizes are given in Table 3. Increasing the sample loop size above 2 ml did little to decrease the LOD for Cr(V1) because the peaks became distorted and overlapped. The conductometric detection limit was found to be 92 pg/l. for a 2-ml sample injection. Lowering the pH of the sample solution to less than 7.0 increased the detection limit for Cr(V1). The calibration plot for Cr(V1) peak area against concentration was linear in the range 1.8-27.6 pg/ml. Interferences

One of the major concerns in analysing soil and sludge samples by ion-chromatography is the precise determination of the ion of interest in the presence of other ions which could mask the signal. In previous determinations of Cr(V1) in water samples by SIC, nitrate and sulphate, which are generally present in such samples, were not considered. In our proposed SCIC method relatively high concentrations of nitrate and sulphate did not significantly affect the analysis for Cr(V1) and these ions can also be determined. Resolution of Cr(VI)/Sqwas excellent (R, = 2.8) and very sharp peaks were obtained. Further studies with standards indicated that other ions such as phosphate (k’ = 2.64), arsenate (k’ = 2.70) and selenite (k’ = 6.28) did not affect the determination of Cr(V1). In another SIC study,” the analysis time for Cr(V1) was 45 min and a broad peak was obtained; reduction of the analysis time to 25 min resulted in interference by sulphate. The proposed method determines only the waterextractable Cr(V1) in soils and sludges. Its use for determination of total chromium would be vitiated

Table 1. Precision of the ion-chromatographic method* for determination of CrtVI) and other anions Sample injection volume, ill 500 1000 2000

Ion Parameter Concentration, RSDt, % Concentration, RSD, % Concentration, RSD. %

pgglml pgglml pgglml

Table 2. Cr(VI) content @g/g) in sewage sludge and soil, determined by SCIC, ICP and speetrophotometric analysis of aqueous extracts Sample

SCIC*

Ontario sludge # 315 Napa sludge # 137 Visalia sludge #204-l Spiked, 2.2 pg Spiked, 2.8 pg Soil 6125-3

3.02 3.16 5.20 5.42 N.D.t N.D. 2.05 2.18 2.70 2.84 N.D. N.D.

3.08 5.10 N.D. 2.14 2.68 N.D.

2.45 2.04

2.40 1.90

Spiked, 2.5 fig Spiked, 2.0 pg

Detection limits and linearity

NO;

SO?-

CrfVI)

12.0 0.9 8.0 2.3 4.0 3.2

15.0 1.8 10.0 3.7 6.0 5.5

10.0 2.0 8.0 4.4 5.0 7.4

*Column, Wescan resin-based anion-exchange; mobile phase, 5mM PHBA (pH 8.5); detection, conductometric. tRSD = relative standard deviation from 10 measurements.

891

ICP

Spectrophotometry

2.48 1.97

*Chromatographic conditions as in Table 1. Based on four 500-~1 injections. tN.D. = not detected.

by the overlap of the Cr(V1) peak by the peaks of the anions of mineral acids that would be necessary for extraction of the total chromium from the samples. Comparative methodr of determination

To investigate the validity of the proposed SCIC method for Cr(V1) in soils and sludges, some samples were spiked with known amounts of Cr(V1) and immediately analysed by SCIC, ICP and spectrophotometry. The results are shown in Table 2, and were similar. The relationships between SCIC (X) and the other two methods of determination were: Y,,, = 1.06&c,, - 0.078; r = 0.997 (P < 0.001)

Yspectrophotomctry= 0.997X,,,,

+ 0.036;

r = 0.995 (P < 0.001) The regression equations indicate excellent agreement with the comparative methods of determination. SCZC analysis for Cr ( VZ) A typical single-column chromatogram for an aqueous sewage sludge extract is shown in Fig. 3. The concentration of Cr(V1) in the dried sludge, calculated on the basis of peak area, was 5.2 pg/g (Table 2). The chromatography time was 16 min. The results obtained in this study show that the proposed SCIC method can be used satisfactorily to determine Cr(V1) in water extracts of soil and slvdge samples. The method requires minimal sample preparation, is rapid and selective, and affords the degree Table 3. Detection limits* of Cr(V1) and other anions, in aqueous solution, by the SCICt method Sample injection volume, PI.

Cr(VI), pgcgll.

NO;, !Jggll.

100 500 2000

1540 320 92

490 105 28

*S/N = 3. tChromatographic

so:-, pggll. 850 172 42

conditions as in Table 1.

H. C. MEHRAand W. T. FRAN~CENBERGER, JR.

892

are grateful to Wescan Instruments analytical column used in this study.

for the gift of the

REFEEENCES

0

4

6

I2

16

RETENTION TIME (MN)

Fig. 3. Chromatogram of an aqueous sludge extract. Column: resin-based anion-exchange; eluent: 5mM PHBA, pH 8.5; detection: conductometric. of accuracy required for determination of trace levels of Cr(V1) in various environmental samples. Acknowledgements-We thank G. Bradford for his assistance with the ICP analysis. This research was partially supported by the U.C. Salinity/Drainage Task Force. We

1. A. Kabata-Pendias and H. Pendias, Trace Elements in Soils and Plants, p. 193. CRC Press, Boca Raton, Florida, 1984. 2. R. J. Bartlett and J. M. Kimble, J. Enoiron. Qual., 1976, 5, 383. 3. World Health Organization (WHO), GuiaMnes for Drinking Water &ality, Vol.. I-~ecommen&ti&s, II. 130. WHO. Geneva. Switzerland. 1984. 4. 7. L. Allen, Anal. Che&., 1958, 30,.447. 5. L. Barnes, ibid., 1966, 38, 1083. 6. J. Obiols, R. Devesa, J. Garcia-Berro and J. Serra, Intern. J. Environ. Anal. Ckem., 1987, 30, 197. 7. S. Greentield, H. McD. McGeachin and P. B. Smith, Talanta, 1975, 22, 553. 8. J. J. Lingane and I. M. Kolthoff, J. Am. Chem. Sot., 1940, 62, 852. 9. S. G. Chen, K. L. Cheng and C. R. Vogt, Mikrockim. Acra, 1983 I, 473. 10. Yu. A. Zolotov, 0. A. Shpigun and L. A. Bubchikova, 2. Anal. Chem., 1983, 316,8. 11. K. F. Nieto and W. T. Frankenberaer. Jr.. Soil Sci. Sot. Am. J., 1985, 49, 587. 12. H. M. Reioenauer, in Methods of Soil Analysis, Part 2, A. L. Page, R. H. Miller and D. R. Keeney (eds.), p. 331. ASA, Madison, Wisconsin, 1982.