745
Anal. Chem. 1984, 56. 745-749
to extend the wavelength range by using higher wattage halogen lamps but they were also unsatisfactory because of inadequate emission at lower wavelengths and excessive heat production. A xenon flashlamp provided excellent emission over the entire visible spectrum that was controlled by the computer and produced no detectable heat. To further enhance sensitivity a PDA was chosen with a 432-pm aperature instead of the 25.4-pm aperature originally used (5-7). The Kodak Ektachem 400 is a commercially available system that uses multilayer dry slides but is not adequate for analyzing small quantities of serum samples since it has a pipet dead volume of 30 pL and requires an additional 10 pL for each test. Furthermore, additional serum must be present in the cup to prevent aspiration of air. In contrast, the proposed system uses blood collected and separated in a microhematocrit tube that provides up to 30 pL of serum or plasma. A 10-pL syringe is used to hold and protect samples from evaporation and provides a negligible dead volume. As demonstrated in earlier work ( 1 , 4 , 5 ) variations , in sample volume produce minimal changes in color intensity. Dispensing 2-pL samples from a 10-pL syringe is convenient and slight variations in pipetting volumes do not significantly alter the final results. Not all dry film chemistry slides were suitable because of poor spreading characteristics found in less sophisticated slides such as those produced by Ames that were based on a paper matrix. Placement of small serum quantities on these slides did not form distinct and uniform spots. Eastman Kodak slides work satisfactorly since they were designed to allow the sample to spread smoothly. Dry film slides from Boehringer Mannheim and Fuji Photo Film Co. were not available for evaluation. The principles set forth here open the way to performing multiple simultaneous analyses on a single slide. Performing ten analyses on a slide designed for a single test would be very cost effective and the savings would be significant. Improvements in multilayer dry film technology that result in
thinner slides would reduce the minimal sample requirements to only a fraction of a microliter and would allow a more dense sample application pattern. The ease of operation, small storage space required for slides, low power consumption, and small sample size would make the proposed system suitable for future use in a neonatal or adult intensive care unit and would avoid the costly and time-consuming sample transportation to the laboratory and problematical delivery of results to the ward. At this time ten differnet slide tests have been evaluated and are in the process of extensive clinical trials. Additional tests are becoming available. ACKNOWLEDGMENT The author wishes to thank Alfred Zettner for his invaluable sugestions and Walter Fontana for his expert technical assistance. Registry No. Urea, 57-13-6. LITERATURE CITED Curme, H. G.; Columbus, R. L.; Dappen, G. M.; et ai. Clin. Chem. (Winston-Salem, N . C . ) 1978, 2 4 , 1335-1342. Walter, B. Anal. Chem. 1983, 55, 498A-514A. Ohkubo, A.; Kamel, S.;Yamanaka, M.; et al. Clin. Chem. (WinstonSalem, N . C . ) 1981, 27, 1287-1290. Neeley, W. E.; Zettner, A. Abstract presented at Second International Congress on Pedlatric Laboratory Medicine, Toronto, Ontario, Canada, June 1983. Neeiey, W. E.; Zettner, A. Clln. Chem. (Winston-Salem, N . C . ) 1983, 2 9 , 2103-2105. Neeley, W. E.; Epstein, D.; Zettner, A. Clin. Chem. ( Winston-Salem, N . C . ) 1981, 27, 1665-1668. Neelev, W. E.: Zettner, A. Clin. Chem. (Wlnston-Salem, N . C . ) 1983, 2 9 , 1038-1041. Wllllams, F. C.; Clapper, F. R. J . Opt. SOC.Am. 1953, 43, 595-599. Glucose Test Methodology, HSMD Publication MP2-8, Eastman Kodak Co., Rochester, NY, 1961. BUN/Urea Test Methodology, HSMD publication MP2-9, Eastman Kodak Co., Rochester, NY, 1981.
RECEIVED for review November 4,1983. Accepted December 23, 1983.
Selective Leaching of Trace Metals from Sediments as a Function of pH John H. Trefry* and Simone Metz Department of Oceanography & Ocean Engineering, Florida Institute of Technology, Melbourne, Florida 32901
Trace metals were leached from sediments and suspended particulates by using phthalate buffers at pH values of 2.2-6. Cadmlum, Cu, Fe, Mn, Pb, and Zn were determined In the resulting leachates by flame or flameless atomic absorption spectrometry. The fractlon of total metal removed varied wlth sample composition, flnal pH, and element determined. Analytical preclslon for the leach was generally 9 (21). Marine sediments commonly have pH values of 6-8 (21). In streams where mine 0 1984 American Chemical Society
748
ANALYTICAL CHEMISTRY, VOL. 56, NO. 4, APRIL 1984
Table I. Selected Reagents Used for Leaching Trace Metals from Sediment and Soil reagent EDTA 1 N ammonium chloride, pH 7 1 N magnesium chloride, pH 7 0.1 M potassium pyrophosphate, pH 10 ammonium oxalate, pH 3 sodium dithionite, pH 3, 4.7, 7 30% hydrogen peroxide sodium hypochlorite, pH 8.5 acetic acid-sodium acetate, pH 5.2 25% acetic acid 25% acetic acid-1 M hydroxylamine hydrochloride 0.1 N hydrochloric acid nitric-perchloric acid ( l / l ) aqua regia nitric acid-hydrochloric acid ( 9 / l )
Table 11. Formulations for Preparation of Phthalate Buffersa
ref 1
2 3 4, 5 5, 6 7, 8 9 3 10 11 12
13,14 15 15 16
tailings have been discharged, pH may be as low as 2 (21). Recent concern for acid rain and acid soil has spurred a global focus on pH (22). As pH decreases, metal release to solution increases. This added metal may serve as a needed micronutrient or be toxic to organisms. Thus, within a wide range of environmental pH values, we need to be able to predict the amounts and rates a t which metals are released. Large pH variations are common to invertebrate and vertebrate digestive systems. The pH of the digestive tract has been listed as a key factor in determining biological uptake of metals from ingested particles (18). Invertebrates typically have an internal pH of 5-8 (23),whereas vertebrate digestive systems may have a pH of - 3 (24). More metal is stripped from particles at lower pH values. However, a poor correlation has been observed between gut pH and metal uptake (25). This inconsistency may partially result from inhibition of metal complexation a t low pH. Despite this apparent discrepancy, knowledge of the amount of metal released from ingested particles is a necessary starting point in metal bioavailability studies. Many previous techniques (Table I) and the processes they seek to understand are pH related. Van Valin and Morse (20) show that metal removal by a number of different leaching methods can be related to the pH of the leaching solution. The technique to be dedcribed provides one direct and sensitive measurement of tlie role of pH in metal uptake/release processes.
EXPERIMENTAL SECTION Preparation of Phthalate Buffers. Phthalate buffers used for these leaching experiments were prepared as outlined in Table 11. Reagents used were high-purity grade KHC8H404(Fisher Scientific Co., Certified A.C.S.), Baker "Instra-Analyzed"HCl for trace metal analysis, and reagent grade NaOH (Mallinckrodt). The pH of each solution was measured before and after sample leaching to *0.01 pH unit. Procedure. Two Gulf of Mexico sediment samples and two Mississippi River suspended particulate samples were used to test the leaching technique. Sediment samples were collected from the Gulf of Mexico in water depths of 33 m (station 9, 28" 53.6' N, 89O 29.9' W) and 106 m (station 7, 28O 37.9' N, 89" 43.0' W) at distances of 8 and 40 km southwest of the mouth of the Mississippi River. Suspended particulate samples were collected at Head of Passes in the lower Mississippi River. Sediment and suspended particulate pH leaches were initially carried out with 0.050.400-g (dry wt) aliquots of sample in 50-mL polypropylene centrifuge tubes. On the basis of the results to be shown, 0.400-g samples were used for most of the tests performed. Twenty milliliters of each pH buffer solution was added to separate sample aliquots prior to shaking for 24 h on a wrist-action shaker. Selected samples were also equilibrated for 0.5,1,3,6,12,and 48 h. After the allotted time period, the solution
pH
volumes of each reagent given in mL 0.1 M 0.1 N 0.1 N KHC,H,O, HC1 NaOH
2.2
50
3.0
50
4.0
50
5.0
50 50
6.0
46 21
0 0
0 0 0
0.4 24 46
a After Clark and Lubs (26). Each buffer is made u p to a final volume of 100 mL with distilled, deionized water.
was centrifuged at 2000 rpm for 10 min and then decanted into 30-mL conventional polyethylene bottles. The final pH of each solution was determined on separate 5-mL aliquots to avoid trace metal contamination. Drying samples prior to pH leaching might change the fraction of metal removed. To check this possibility, the leach series was also carried out on aliquots of wet and freeze-dried suspended particulates. Reagent and procedural blanks were prepared and analyzed for each buffer solution. All glassware, polypropylene centrifuge tubes, and polyethylene bottles used during leaching were acid washed with warm 3 N HN03 and rinsed well with distilled, deionized water (DDW). For sediment analysis, these cleaning procedures are sufficient since they result in glass and plasticware blanks which are below detection limits. Leaching extracts and blanks were analyzed for Cd, Cu, Fe, Mn, Pb, and Zn. Determination of Cd, Cu, and Pb was by flameless atomic absorption spectrometry (AAS) with a PerkinElmer 4000 instrument equipped with an HGA-400 graphite atomizer, an AS-40 autosampler, and deuterium background correction. Fe, Mn, and Zn were determined by flame AAS with a Perkin-Elmer 460 instrument. All metal concentrations are reported on a Mg/g dry weight basis. No deuterium background correction was required during flame atomic absorption analysis. Matrix interferences were monitored for all elements at each pH using standard additions analysis. Total dissolution of the samples was also carried out to establish the fraction of metal removed at each pH. Digestion of 0.400-g samples was performed in Teflon beakers by using concentrated HNO3-HC10,-HF in a three-step process. First, 1mL of HC104, 1mL of "OB, and 3 mL of HF were added, a Teflon watch cover was put in place, and the sample was heated at -50 "C until a moist paste was obtained. The mixture was heated for 3 h at 80 "C with an additional 2 mL of HN03 and 3 mL of HF and then brought to dryness. Finally, 1mL of HN03 and -30 mL of DDW were added to the sample and heated to dissolve perchlorate salts and reduce the volume. The completely dissolved and clear samples were then diluted to 25 mL with DDW. Analysis by flame or flameless AAS followed as described above.
RESULTS AND DISCUSSION Precision, Standard Additions, Blanks. Reproducibility for the pH leach was generally quite good as shown by the standard deviations for triplicate analyses in Table 111. After 12 h of shaking, replicates at each of the three pH values (Table 111) had exactly the same final pH. When precision for metal concentrations is reported as [ (mean/standard deviation) X loo%], most analyses show values
