STANDARD OPERATING PROCEDURE FOR THE ANALYSIS OF
ARSENIC AND SELENIUM IN SOILS, SEDIMENTS AND SOLIDS BY GFAA
ASSY0 OR ASSY1 AND SESY0 OR SESY1
December 3, 1990
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
REGION 5 CENTRAL REGIONAL LABORATORY
536 SOUTH CLARK STREET (5SCRL)
CHICAGO, ILLINOIS 60605
Signed John V. Morris 12/3/90
Signed James H. Adams 12/4/90
QUALITY CONTROL COORDINATOR
Signed James H. Adams 12/4/90
Signed Charles T. Elly 12/4/90
STANDARD OPERATING PROCEDURE FOR THE ANALYSIS OF ARSENIC AND SELENIUM BY IN SOILS AND SEDIMENTS BY GFAA
TABLE OF CONTENTS
1. SCOPE AND APPLICATION
2. SAFETY AND WASTE HANDLING
3. SUMMARY OF METHOD
4. SAMPLE HANDLING AND PRESERVATION
10. QUALITY CONTROL
11. PREVENTIVE MAINTENANCE
STANDARD OPERATING PROCEDURE FOR THE ANALYSIS OF ARSENIC AND SELENIUM IN SEDIMENTS AND SOILS BY GFAA
1. SCOPE AND APPLICATION
1.1 The method is applicable to the determination of arsenic and selenium in soils and sediments as requested for all programs. Without dilution, the working range is from 0.4 to 8 mg/kg, while the estimated detection limits are 0.4 mg/kg for both elements.
2. SAFETY AND WASTE HANDLING
2.1 Unused portions of concentrated stock (1000 ug/L) solutions must be poured into a red labeled (acidic metals) waste container.
2.2 Intermediate solutions must be poured into the red labeled waste containers.
2.3 Concentrated nitric acid must be handled with caution. Gloves and lab coat must be worn during all phases of the sample preparation.
3. SUMMARY OF METHOD
3.1 Solid samples are air dried overnight and weighed to correct for moisture. The sample is pulverized in a plastic vial in a ball mill. If there are many rocks, the procedure in Appendix A and the Calculations section should be followed. About one half gram of sample is weighed into the teflon digestion vessel. Nitric acid is added and the sample heated in the microwave oven. The digest is diluted and filtered into a volumetric flask. The volumetrics are filled to volume, then poured into unused polyethylene bottles. The samples are converted to weight basis results at the time of analysis.
3.2 The samples are analyzed by graphite furnace atomic absorption using Zeeman background correction. The samples are to be analyzed by stabilized temperature platform graphite furnace using the standard curve methodology given in the corresponding water methods (references 13.1 and 13.2), and to proceed with standard additions if the stated criteria are not met (references 13.3 and 13.4).
4. SAMPLE HANDLING AND PRESERVATION
4.1 Samples should be collected in polyethylene or glass bottles with solid polypropylene caps. Bakelite caps, even with Teflon liners, are not acceptable. The bottles need not be pretreated if they are virgin polyethylene. If the bottles need cleaning for some specific reason, these bottles must be soaked for twenty four hours with dilute nitric acid before use. The usual contaminant has been found to be lead. At least 100g of sample is requested for accuracy in the percent solids determination.
4.2 Samples should be shipped on ice, and should be kept refrigerated when not in use. This is particularly important when the sample jar is being shared for other tests. Samples can be held for six months before analyses.
4.3 Solid samples are not preserved. The samples are air dried overnight and percent solids determined. The dried sample is pulverized in a plastic vial with a plastic ball in a ball mill shaker. If rocks are a problem, use the procedure in Appendix A. The dried sample is used for the digestion.
4.4 If the sample is hazardous, a warning should accompany the samples and should be sent to the Metals Team Leader. If the samples contain large amounts of oil or volatile materials, the procedure for drying and digestion may be modified to fit the samples.
5.1 There are two major types of spectroscopic interference in atomic absorption.
5.1.1 Spectral interference exists when an emission line of a second element is present within a band pass of the wavelength that the analyte absorbs. Although this is rare, the Zeeman background correction eliminates this problem entirely.
5.1.2 Background absorption occurs when non-specific interferences overlap the atomic absorption wavelength of the analyte. This undesirable apparent absorption is caused by scattered light from undissociated molecules of the matrix over a wide wavelength region. To correct this, the background absorption is measured and subtracted from the total absorption signal.
The Zeeman effect is a splitting and polarizing of the absorption band with strong magnetic field. The magnetic field causes plane polarized light to be selectively absorbed by the atoms. A polarized filter is permanently mounted in the beam path. The strong magnetic field also moves to a different wavelength the elemental polarized absorption passed by the filter. The background absorption is then measured in the absence of the absorption band and subtracted.
5.2 The stabilized temperature platform furnace is a method that eliminates many matrix interferences. Instead of atomizing the analyte from the wall of the graphite tube, the sample is atomized from a platform of pyrolytic graphite. A sample atomized from the wall enters an atmosphere that is not at equilibrium. The platform has limited contact with the tube and lags behind the tube wall in temperature. When the analyte is atomized from the platform, the atmosphere is closer to thermal equilibrium. As a result, many effects which manifest themselves as matrix interferences are eliminated.
5.3 Matrix modification is required for platform analysis of arsenic or selenium. A solution of nickel nitrate is used to control interferences. A 5 uL injection of this solution is equivalent to 10 ug of Ni.
6.1 Digestion Apparatus
6.1.1 Glass, class A, 5-10 mL volumetric pipettes
6.1.2 Mettler PR700 electronic balance
6.1.3 100 mL volumetric flasks
6.1.4 Polyethylene sample bottles 60 mL and 125 mL
6.1.5 Solid polypropylene caps without liners
6.1.6 30 mL plastic cups
6.1.7 Eppendorf 1 mL pipet and tips
6.1.8 Microwave oven, Model MDS 81D from CEM Corp. The oven is designed for laboratory work with a teflon lining, a vented cavity, a turntable and programming capability.
6.1.9 Teflon digestion vessels with pressure release valves
6.1.10 Capping station to tighten the caps of the digestion vessels to 12 lbs, so the pressure relief works properly, and to release the caps after digestion.
6.2 Mixer/Mill, Spex model 8000-115 for pulverizing dried samples
6.2.1 Polystyrene vial, Spex 6133, 3/4" diameter by 2" long, with polyethylene cap
6.2.2 Methacrylate balls, Spex 3112 or 3119, for use in vials
6.3 All digestion glassware, after being run through a dishwasher, is rinsed with distilled water and placed in an aqua regia bath overnight. It is rinsed thoroughly and allowed to dry.
6.4 The Teflon digestion vessels are washed with the other glassware and rinsed in distilled water. If contamination is suspected, the vessels may be cleaned by heating in the microwave with acid in them. It is a good procedure to keep the vessels used for soil digestion separate from those used for low level water digestions. In practice, the vessels numbered from 1-49 are used for sediments, while the higher numbers are reserved for waters.
6.5 The atomic absorption instruments and glassware are given in the appropriate water analysis methods.
7.1 Nitric Acid - Ultrex , Baker Instra-Analyzed or GFS Chemicals redistilled.
7.2 Water - Laboratory reverse osmosis water is passed through a mixed bed resin column before use. The water is called "Super Q", after the column used. All water used for sample preparation and GFAA analysis and is "Super Q".
7.3 Atomic Absorption Standards: 1000 mg/L Fisher, Ventron, or Spex brands have been found to be acceptable. Ventron or Spex standards may deviate from the 1000 mg/L by a few percent, but the true concentration is stated on the label.
7.4 The matrix modifier is described in the appropriate water analysis method, as are the standards.
7.5 Digestion spiking solution: Two milliliters each of stock solutions (1000 mg/L) for arsenic and selenium are added to a 100 mL volumetric flask containing Super Q water and 0.5 mL nitric acid, and brought to volume with water. One milliliter of this solution will be added to the dried sample in the digestion procedure.
8.1 INSTRUMENT PREPARATION
8.1.1 For operation of the atomic absorption spectrophotometers, refer to the corresponding water analysis method (references 13.1 - 13.4).
8.2 Solids Digestion
8.2.1 Fill out prep sheet (see example) with all the samples to be digested. Include a blank, a duplicate and a spike for every ten samples or fraction thereof. A solid reference sample is also digested with every run. Matching the sample matrix is limited by the reference solids available, however consideration should be given to the matrix in the choice of reference sample. File a copy of the prep sheet in the metals lab ring binder for digestions. This binder will also furnish the run number to be put on the page. Field duplicates and spikes are sometimes designated, if not, choose samples for the duplicate and spike.
8.2.2 The sediment or soil is thoroughly mixed in the jar in which it is received. A subsample is withdrawn with a spatula and placed in an evaporating dish. Generally, 10 g of the wet sample is taken, but if the sample has a high moisture content, more can be taken. The sample is placed in the hood to dry overnight. The sample must be weighed as for total solids, as because a different moisture content may be derived by this method. See section 9.1 for details. Note that if large rocks are present, proceed as in Appendix A.
8.2.3 The dried sediment is placed into a labeled polystyrene vial with a methacrylate ball. This vial is mounted into the mixer/mill, and ground for 10 min. Alternatively, the sample may be ground with a mortal and pestle until the entire sample passes through a 10-mesh polypropylene sieve. The sieved sample is then poured into a 60 mL polyethylene bottle, which is capped and labeled.
8.2.4 About one half gram of dry sample is weighed into the teflon digestion vessel sitting on the balance. The weight is recorded on the digestion sheet.
8.2.5 A duplicate is another weighed aliquot of sample treated as a separate sample. A field designated duplicate may be used or select a sample.
8.2.6 A blank is an empty vessel, to which is added nitric acid, and is treated as a sample.
8.2.7 For the spike, one milliliter of the sediment spike solution (7.5) is added if the final volume is 100 milliliters. If another final volume is used, the spike must be adjusted to correspond.
8.2.8 Ten milliliters of concentrated nitric acid is added to the vessel.
8.2.9 The cap is partially tightened by hand, then placed in the capping station and tightened to the required torque. Six of the digestion vessels are placed in pairs in the twelve position carrousel. The other six positions are occupied by the overflow vessels. The overflow vessels have two openings for the connection tubing. One tubing is used to connect the digestion vessel to the overflow vessel. Another piece of tubing is used to connect the overflow vessel to the central tub for overflow.
8.2.10 The full carrousel is placed in the microwave oven on the central turntable. The main power is turned on. The turntable is turned on and free rotation is checked. The adjustable fan is turned to 10, the highest speed. The keypad is used to enter 1000 for the time, followed by an enter and 100 for the power also followed by an enter. The start key is pressed and the digestion begun.
8.2.11 In ten minutes the digestion is done, but the chamber has lots of NOx and the vessels are hot. The vessels are allowed to stay in the chamber while the filters and bottles are prepared. One half to one quarter of an hour is needed for cooling.
8.2.12 Rinse virgin polyethylene 125 mL bottles and caps with "super Q" water and shake out the large water drops. Label the bottles with the data set, sample ID, sample weight and vessel number.
8.2.13 The folded filter papers are placed in the disposable plastic funnels in a dish drainer over the sink. The funnels are then washed with one fill (of the funnel) of 0.5% nitric or hydrochloric acid. The filters are allowed to drain over the sink and moved to volumetric flasks on the bench.
8.2.14 The carrousel is placed in the hood immediately upon removal from the oven. If the vessels are still hot, they are allowed to cool more. After the vessels are cool enough to handle easily, the digestion vessel and its attached overflow with all tubes are removed from the carrousel.
8.2.15 The tube is removed from the top of the digestion vessel; this leaves two tubes attached to the overflow vessel. A wash bottle with a small diameter tip is used to rinse any condensate from the tubes into the overflow vessel. The tubes are placed in a pile for later rinsing with more distilled water at the tap. Remove the cap from the overflow vessel in the hood. Rinse the cap lightly if necessary. Rinse the sides of the overflow vessel sparingly.
8.2.16 Use the capping station to loosen the cap of the digestion vessel. Remove the cap of the digestion vessel. Placing a finger over the vent fitting in the top of the cap, pour the water from the overflow vessel into the cap. Then pour the liquid from the cap into the digestion vessel. Allow the brown fumes from the digestion vessel to dissipate in the hood while opening another vessel. After the fumes have gone away, sparingly rinse the sides of the digestion vessel.
8.2.17 Pour the solution into a prepared funnel sitting in the prepared volumetric flask. If the funnel is not draining well, lift the funnel slightly; the funnel and the flask sometimes develop an airtight seal. Rinse the digestion vessel once or twice to remove the last traces of solution. Use the fine tipped wash bottle.
8.2.18 Usually enough space remains in the flask to permit rinsing of the filter paper once or twice. There has been no problem with the acid eating through the filter paper with this procedure. The solution from the digestion vessel usually takes up 2/3 to 3/4 of the volume in the filter paper; the wash from the vessel another quarter. The filtered samples are taken up to volume in the flasks. The samples are then poured into labeled, washed plastic bottles and capped with unlined caps.
8.3 The samples are now ready for analysis by the appropriate water method. There is no limitation that digestion quality control need accompany the corresponding samples on the same wheel, but it is a good practice.
9.1 The concentrations derived from the instrument for the digests are in ug/L. These results must be converted to mg/kg dry weight. Because there may be a difference between the total solids derived by drying at 105oC and the air drying procedure described in section 8.2.2, this correction must be applied.
9.1.1 The air dried sample is weighed both before and after drying, in a tared evaporating dish. These weights are used to calculate a modified percent solids, smod.
9.1.2 Using the percent solids derived from oven drying, s, the calculation of the concentration of arsenic or selenium in the dry soil or sediment is as follows:
concentration in mg/kg = (smod/s) C x V / W
where C is the concentration of analyte in the digest in ug/L, after any correction for dilution, V is the volume of the digest in liters, and W is the weight of the dry sample aliquot in grams.
9.2 Calculation of spike recoveries follows the standard formula
% recovery = 100 x (Cspike - Csample)/ A
where the concentrations Cspike and Csample have been calculated as in 9.1.2, and the added spike A has been converted to mg/kg by using the same calculation, using 200 ug/L as the expected concentration and the weight of the spiked aliquot for W.
9.3 Sediments are normally reported to two significant figures. Quality control and other special types of samples may be reported to more figures, usually three, for statistical analysis.
10. QUALITY CONTROL
10.1 The analysis must meet all the quality control requirements of the water method, including quality of fit, instrument blank, and AQC. It must also meet requirements for the digestion quality control. This includes the digestion blank, duplicate, spike, and laboratory control standard.
10.2 The digestion blank demonstrates there is no contamination in the reagents or digestion vessels. Values for blanks exceeding either plus or minus the stated detection limit are flagged.
10.3 Duplicates are evaluated by relative percent difference. The expected RPD is + 20% or, near the detection limit, within plus or minus detection limit. Duplicates give an indication of sample homogeneity and consistency of subsampling.
10.4 Spikes are evaluated by percent recovery. The expected percent recovery is 100 + 20%. For a complex sample low spike recovery may indicate interferences.
10.5 The digested laboratory control standard shall be within the stated acceptance window given in the documentation received with the solid LCS. Currently, LCS 0287 from EMSL-Las Vegas is being used, because it has values for many of the elements analyzed by this method.
10.6 Failure to meet the stated limits on a digestion audit is generally a cause for redigestion and reanalysis. Exceptions may be granted in certain situations, such as a positive blank, when no samples were above detection. In any case, the number of reanalyses is generally limited to two, after which the data is flagged. Consult with the Metals Team Leader whenever audits are outside the stated limits.
11. PREVENTIVE MAINTENANCE
11.1 Consult the corresponding instrument operating manual.
12.1 For any data set there is a standard set of documents which accompany the data to the evidence files. The standard set includes:
12.1.1 A copy of the reported values for the samples of the data set
12.1.2 A copy of the digestion sheet which tells what other data sets are in the same digestion group and analysis run
12.1.3 A copy of the raw data from all runs from which results are derived. All calculations for duplicate and spike, and all corrections for dilution shall be included. These may be done on a separate sheet for clarity. Use of spreadsheet software for the calculations is acceptable.
12.2 The original printout of the data report from the instrument is placed in consecutive order in ring binders found in the AA lab for that purpose.
13.1 CRL Method 206.2 DNS(5100), "Standard Operating Procedure for the Determination of Arsenic by Flameless Platform Atomic Absorption", Sept. 1990.
13.2 CRL Method 270.2 DNS(5100), "Standard Operating Procedure for the Determination of Selenium by Flameless Platform Atomic Absorption", Oct. 1990.
13.3 CRL Method 206.2 DNS (AA, Furnace, Standard Addition) "Standard Operating Procedure for the Analysis of Arsenic in Water", Sept. 1987.
13.4 CRL Method 270.2 DNS (AA, Furnace, Standard Addition) "Standard Operating Procedure for the Analysis of Selenium in Water", Sept. 1987.
13.2 "Methods for Chemical Analysis of Water and Wastes," U.S. EPA Publication, EPA 600/4-79-020.
Some sediment samples contain large rocks and, in some cases, have been nearly half pea gravel. In order to achieve homogeneity in the analyzed fractions, the samples shall be sieved using a 10-mesh polypropylene sieve. The material which passes the sieve will be dried for total solids and analyzed for ICP metals and cyanide. Excluding the rocks from the calculations would cause the concentration reported for the soil to be too high. The entire sample is weighed before the separation. The rocks are weighed after separation, so the weight can be mathematically included in the calculations.
The sample mass can be expressed as:
M = W + D + R
where M is mass of the entire sample, W is the mass of the water, D is the mass of the dry soil, and R is the mass of the rocks. This assumes that the water which is on the surface of the rocks after sieving is negligible. After separation, an aliquot of the wet soil is dried for solids determination.
% solids = P = 100xD/(W + D)
The metals determinations are made on the dried fraction, and the cyanide is analyzed from another aliquot of the wet soil.
The determinations which are made on the dried fraction, or are corrected to dry weight using P, the total solids of the soil portion, can be corrected to dry weight of the total sample, rocks and all, by multiplying the result by the factor
D/(D + R).
Since D was not directly measured, but M and R were, it can be found by
D = (M - R)xP/100
Using this value, the true total solids for the sample can be determined
% solids (total sample) = 100x(D + R)/M
A LOTUS spreadsheet can be set up to handle these calculations.
This calculation takes into account the entire sample, and avoids an artificially high result. Separation of the rocks allows better sample homogenization, and more accurate results.