Designation: D3859 08

Standard Test Methods for

Selenium in Water1
This standard is issued under the fixed designation D3859; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon () indicates an editorial change since the last revision or reapproval.

bility of regulatory limitations prior to use. For specific hazard

statements, see 11.12 and 13.14.

1. Scope
1.1 These test methods cover the determination of dissolved
and total recoverable selenium in most waters and wastewaters.
Both test methods utilize atomic absorption procedures, as
follows:

2. Referenced Documents
2.1 ASTM Standards:3
D1129 Terminology Relating to Water
D1193 Specification for Reagent Water
D2777 Practice for Determination of Precision and Bias of
Applicable Test Methods of Committee D19 on Water
D3370 Practices for Sampling Water from Closed Conduits
D3919 Practice for Measuring Trace Elements in Water by
Graphite Furnace Atomic Absorption Spectrophotometry
D4841 Practice for Estimation of Holding Time for Water
Samples Containing Organic and Inorganic Constituents
D5810 Guide for Spiking into Aqueous Samples
D5847 Practice for Writing Quality Control Specifications
for Standard Test Methods for Water Analysis

Sections
Test Method AGaseous Hydride AAS2
Test Method BGraphite Furnace AAS

7 to 16
17 to 26

1.2 These test methods are applicable to both inorganic and

organic forms of dissolved selenium. They are applicable also
to particulate forms of the element, provided that they are
solubilized in the appropriate acid digestion step. However,
certain selenium-containing heavy metallic sediments may not
undergo digestion.
1.3 These test methods are most applicable within the
following ranges:
Test Method AGaseous Hydride
AAS2
Test Method BGraphite Furnace
AAS

1 to 20 g/L

3. Terminology
3.1 Definitions:
3.1.1 For definitions of terms used in these test methods,
refer to Terminology D1129.
3.2 Definitions of Terms Specific to This Standard:
3.2.1 total recoverable seleniuman arbitrary analytical
term relating to the recoverable forms of selenium that are
determinable by the digestion procedures included in these test
methods.

2 to 100 g/L

These ranges may be extended (with a corresponding loss in

precision) by decreasing the sample size or diluting the original
sample, but concentrations much greater than the upper limits
are more conveniently determined by flame atomic absorption
spectrometry.
1.4 The values stated in SI units are to be regarded as
standard. No other units of measurement are included in this
standard.
1.5 This standard does not purport to address all of the
safety concerns, if any, associated with its use. It is the
responsibility of the user of this standard to establish appropriate safety and health practices and determine the applica-

4. Significance and Use

4.1 In most natural waters selenium concentrations seldom
exceed 10 g/L. However, the runoff from certain types of
seleniferous soils at various times of the year can produce
concentrations as high as several hundred micrograms per litre.
Additionally, industrial contamination can be a significant
source of selenium in rivers and streams.
4.2 High concentrations of selenium in drinking water have
been suspected of being toxic to animal life. Selenium is a

1
These test methods are under the jurisdiction of ASTM Committee D19 on
Water and are the direct responsibility of Subcommittee D19.05 on Inorganic
Constituents in Water.
Current edition approved Oct. 1, 2008. Published October 2008. Originally
approved in 1984. Last previous edition approved in 2003 as D3859 03. DOI:
10.1520/D3859-08.
2
Lansford, M., McPherson, E. M., and Fishman, M. J., Atomic Absorption
Newsletter, Vol 13(4), 1974, pp. 103105. Pollack, E. N., and West, S. J., Atomic
Absorption Newsletter , Vol 12(1), 1973, pp. 68.

3
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM
Standards volume information, refer to the standards Document Summary page on
the ASTM website.

Copyright (C) ASTM International. 100 Barr Harbor Drive, P.O. box C-700 West Conshohocken, Pennsylvania 19428-2959, United States

D3859 08
8. Summary of Test Method
8.1 The determination consists of the conversion of selenium in its various forms to gaseous selenium hydride (hydrogen selenide), with the subsequent analysis of the gas by flame
AAS.
8.1.1 The conversion consists of (a) decomposition and
oxidation to selenium (VI), (b) reduction to selenium (IV), and
(c) final reduction to selenium hydride.
8.1.2 The absorbance is determined at 196.0 nm in a
hydrogen-argon (air-entrained) flame.
8.2 Sample concentrations are obtained directly from a
simple concentration versus absorbance calibration curve.
8.3 Total recoverable selenium is determined by treating the
entire sample as the procedure indicates, and the dissolved
selenium is determined by treating the filtrate after the sample
is filtered through a 0.45-m membrane filter.

priority pollutant and all public water agencies are required to

monitor its concentration.
4.3 These test methods determine the dominant species of
selenium reportedly found in most natural and wastewaters,
including selenities, selenates, and organo-selenium compounds.
5. Purity of Reagents
5.1 Reagent grade chemicals shall be used in all tests.
Unless otherwise indicated, it is intended that all reagents shall
conform to the specifications of the Committee on Analytical
Reagents of the American Chemical Society, where such
specifications are available.4 Other grades may be used, provided it is ascertained that the reagent is of sufficiently high
purity to permit its use without lessening the accuracy of the
determination.
5.2 Purity of WaterUnless otherwise indicated, reference
to water shall be understood to mean reagent water conforming
to Specification D1193, Type I. Other reagent water types may
be used provided it is first ascertained that the water is of
sufficiently high purity to permit its use without adversely
affecting the bias and precision of the test method. Type II
water was specified at the time of round robin testing of this
test method.

9. Interferences
9.1 Mercury and arsenic at concentrations greater than 500
g/L and greater than 100 g/L, respectively, may inhibit the
formation of selenium hydride.
10. Apparatus
10.1 An apparatus similar to that depicted in Fig. 1, with the
components specified in 10.2-10.4.8, is recommended for this
test method.5
10.2 Atomic Absorption SpectrophotometerThe instruments shall consist of an atomizer and burner, suitable pressure
and flow regulation devices capable of maintaining constant
diluent and fuel pressure for the duration of the test, a selenium
lamp, an optical system capable of isolating the desired
wavelength, an adjustable slit, a photomultiplier tube or other
photosensitive devices such as a light measuring and amplifying device, and a readout mechanism for indicating the amount
of absorbed radiation. A background corrector may be used, but
is not absolutely essential.
10.2.1 Selenium Electrodeless Discharge LampThe sensitivity of selenium to atomic absorption spectroscopy is
generally improved with this lamp, although some hollowcathode lamps produce equivalent results. The intensity and
stability of the lamp shall be adequate to determine selenium in
the range from 1 to 20 g/L.
10.2.2 Recorder or Digital Readout, or BothAny multirange, variable-speed recorder, or digital readout accessory that
is compatible with the atomic absorption detection system, is
suitable.
10.2.3 The manufacturers instructions are to be followed
for all instrument parameters.
10.3 Gas System:
10.3.1 See 11.14 for materials for the gas system.
10.3.2 Pressure-Reducing ValvesPressure-reducing
valves shall be capable of maintaining argon pressure at 40
psig (275 kPa) and hydrogen pressure at 20 psig (138 kPa).
10.4 Additional Equipment:

6. Sampling
6.1 Collect the samples in accordance with Practices
D3370. Take the samples in acid-washed TFE-fluorocarbon or
glass bottles. Other types of bottles may be used for sampling,
but should be checked for selenium absorption. The holding
time for the samples may be calculated in accordance with
Practice D4841.
6.2 When determining only dissolved selenium, filter the
sample through a 0.45-m membrane filter as soon as possible
after sampling. Add HNO3 to the filtrate to bring the pH to
<2.0.
6.3 When determining total recoverable selenium, add
HNO3 to the unfiltered sample to a pH of <2.0.
TEST METHOD AGASEOUS HYDRIDE AAS
7. Scope
7.1 This test method covers the determination of dissolved
and total recoverable selenium in the range from 1 to 20 g/L.
The range may be extended by decreasing the sample size or
diluting the original sample.
7.2 This test method has been used successfully with
reagent water, natural water, wastewater, and brines. The
information on precision may not apply to waters of other
matrices.

4
Reagent Chemicals, American Chemical Society Specifications, American
Chemical Society, Washington, DC. For Suggestions on the testing of reagents not
listed by the American Chemical Society, see Annual Standards for Laboratory
Chemicals, BDH Ltd., Poole, Dorset, U.K., and the United States Pharmacopeia
and National Formulary, U.S. Pharmacopeial Convention, Inc. (USPC), Rockville,
MD.

5
A static system, such as one using a balloon, has been found satisfactory for this
purpose. See McFarren, E. F., New, Simplified Method for Metal Analysis,
Journal of American Water Works Association, Vol 64, 1972, p. 28.

D3859 08

FIG. 1 Apparatus for Selenium Determination

10.4.1 Flask HeaderThe flask header shall consist of a

three-hole rubber stopper into which is inserted:
10.4.1.1 A sintered-glass aeration tube for the argon sweep
gas,
10.4.1.2 A small gas chromatographic-type septum (5 to 10
mm in diameter), for injection of the borohydride solution, and
10.4.1.3 A glass outlet tube for the reaction gases to exit.

10.4.2 Fittings and AdaptersStainless steel fittings and

adapters shall be used to install the reaction-flask header in
series with the auxiliary oxidant line and the burner. Plastic or
other metals may be substituted if proven acceptable.
10.4.3 TubingAny commercially available plastic tubing
that is not susceptible to attack by hydrochloric acid, selenium
hydride, or other gases from the reaction mixture is acceptable.
Poly(vinyl chloride) tubing has been found acceptable.
10.4.4 Gas-Flow RegulatorA suitable in-line gas-flow
valve shall be used to adjust the flow of argon to the
reaction-flask header.

NOTE 1Instead of the gas chromatographic-type septum, a more

secure seal may be obtained by using a glass tube with a septum cap.
These items are commercially available on an individual basis. A different
header may be used if proven reliable.

D3859 08
11.14 Gases:
11.14.1 Argon (nitrogen may be used in place of argon)
Standard, commercially available argon is the usual diluent.
11.14.2 HydrogenStandard, commercially available hydrogen is the usual fuel.

10.4.5 Water Trap (optional)Any commercially available

glass trap suitable to prevent carryover moisture from going to
the burner is acceptable.
10.4.6 One-Way Gas Check Valve (optional)A one-way
check valve can be installed in series with the water trap and
burner to prevent hydrogen from back flowing to the generating flask whenever samples are changed. However, precautionary measures could generally preclude the use of this device,
since only when the flask header is removed for prolonged
periods would there be significant hydrogen back flow.
10.4.7 Reaction Flasks, 250-mL spoutless beakers, or their
equivalent, with graduations may be used. Conical and restricted neck flasks do not perform as reliably as spoutless
beakers.
10.4.8 Hypodermic Syringe, 2-mL capacity with a 50-mm
needle.

12. Standardization
12.1 Transfer 0.0, 0.5, 1.0, 2.0, 5.0, and 10.0-mL portions of
the standard selenium solution (1.0 mL = 0.10 g Se) to freshly
washed 250-mL reaction flasks. Adjust the volume to 50 mL
with water.
12.2 Proceed as directed in 13.3-13.15.
12.3 Prepare a calibration curve by plotting absorbance (or
recorder scale readings) versus micrograms of selenium on
linear graph paper. Alternatively, if provided with this capability, calibrate the spectrophotometer to output micrograms of
selenium directly.

11. Reagents and Materials

11.1 Calcium Chloride Solution (30 g/L)Commercially
purchase or dissolve 30 g of calcium chloride (CaCl22H2O) in
water and dilute to 1 L.
11.2 Hydrochloric Acid (sp gr 1.19), concentrated hydrochloric acid (HCl).
11.3 Hydrochloric Acid (1 + 1)Add 1 volume of HCl (sp
gr 1.19) to 1 volume of water. Always add acid to water.
11.4 Hydrochloric Acid (1 + 99)Add 1 volume of HCl (sp
gr 1.19) to 99 volumes of water. Always add acid to water.
11.5 Methyl Orange Indicator Solution (25 mg/100 mL)
Dissolve 25 mg of methyl orange in 100 mL of water.
11.6 Nitric Acid (sp gr 1.42), concentrated nitric acid
(HNO3).
11.7 Nitric Acid (1 + 99)Add 1 volume of HNO3(sp gr
1.42) to 99 volumes of water.
11.8 Potassium Permanganate Solution (0.3 g/L) Dissolve 0.3 g of potassium permanganate (KMnO4) in water and
dilute to 1 L.
11.9 Selenium Solution, Stock (1.00 mL = 1.00 mg
selenium)Accurately weigh 1.000 g of gray elemental selenium and place in a small beaker. Add 5 mL of HNO3 (sp gr
1.42). Warm until the reaction is complete, then cautiously
evaporate to dryness. Redissolve with HCl (1 + 99) and dilute
to 1 L with the same acid solution.
11.9.1 Alternatively, certified selenium stock solutions are
commercially available through chemical supply vendors and
may be used.
11.10 Selenium Solution, Intermediate (1.00 mL = 10 g
selenium)Dilute 5 mL of the selenium stock solution to 500
mL with HCl (1 + 99).
11.11 Selenium Solution, Standard (1.00 mL = 0.10 g
selenium)Dilute 10 mL of the selenium intermediate solution
to 1000 mL with HCl (1 + 99). Prepare fresh daily and store in
a TFE-fluorocarbon or other acceptable container. To minimize
waste only prepare 100 mL of the Selenium Standard Solution.
11.12 Sodium Borohydride Solution (4 g/100 mL) Dissolve 4 g of sodium borohydride (NaBH4) and 2 g of sodium
hydroxide in water and dilute to 100 mL. Prepare fresh weekly.
Warning: Sodium borohydride reacts strongly with acids.
11.13 Sodium Hydroxide Solution (4 g/L)Dissolve 4 g of
sodium hydroxide (NaOH) in water and dilute to 1 L.

13. Procedure
13.1 It is emphasized that careful control of pH, oxidant
concentration, temperature, and time are imperative if accurate
and precise selenium determinations are to be obtained.
13.2 For each sample, transfer 50 mL or less (to contain not
more than 1.0 g selenium) to a freshly washed 250-mL
reaction flask. Make up to 50 mL with water if necessary.
13.3 To each sample, standard, and blank, add a few drops
of methyl orange solution, 0.5 mL of CaCl2 solution and three
or four boiling stones.
13.4 Adjust the pH to the red end point of methyl orange
(pH = 3.1) with HCl (1 + 99) or NaOH solution (4 g/L). Add
0.5 mL of HCl (1 + 99) in excess. A pH meter may be used in
place of the indicator if the sample is sufficiently discolored to
affect the methyl orange end point.
13.5 Add potassium permanganate solution dropwise (about
3 drops) to maintain the purple tint indicating excess KMnO4.
Boil the solution on a hotplate, carefully maintaining the purple
tint until the volume is reduced to about 25 mL. Add 2 mL of
NaOH solution (4 g/L) and concentrate the solutions to
dryness, being careful not to overheat the residue.
13.6 Cool and add 15 mL of concentrated HCl (sp gr 1.19).
Heat on a hot water or steam bath for 20 min. Do not boil. This
step reduces the selenium (VI) to selenium (IV).
13.7 Cool and add HCl (1 + 1) to adjust the volume to 50
mL. Hold these solutions until all samples and standards are
brought to this stage.
13.8 Set the atomic absorption instrument parameters in
accordance with the manufacturers instructions. Typical settings are as follows:
Grating
Wavelength
Burner
Radiation Source
Slit
Flame

ultraviolet
196.0 nm
triple-slot or equivalent
selenium electrodeless discharge lamp or equivalent
2.0 nm
hydrogen-argon (nitrogen may be used in place of
argon)

13.9 If the gas control box is not equipped with separate

controls for argon and hydrogen, simply connect the oxidant
inlet line for the control box to the argon tank regulator and
connect the fuel inlet line for the control box to the hydrogen
tank regulator. The oxidant controls will then control the argon
4

D3859 08
15. Precision and Bias 6
15.1 The overall and single-operator precision of this test
method within its designated range for reagent water and
nonreagent water varies with the quantity being measured in
accordance with Table 1. These values were established for
four laboratories, using six operators over three consecutive
days. The nonreagent waters included natural, waste, and brine
waters.
15.1.1 The overall precision for reagent water varies linearly with the quantity being measured, and it may be
expressed mathematically using Eq 2:

diluent gas and the fuel controls will control the hydrogen gas.
To preclude the possibility of accidentally mixing the hydrogen
fuel with the air oxidant normally used with atomic absorption
spectroscopy, shut off all sources of air oxidant to the system.
Set the tank pressures, the burner control box pressures, and the
flow rates in accordance with the manufacturers instructions
for argon and hydrogen.
13.10 Center the burner about 5 mm below the optical light
path. Ignite the flame. Since the flame does not give off visible
light, optical flame sensors must be bypassed, but the presence
of the low-temperature flame may be verified by aspirating tap
water, which contains soluble salts that impart color to the
flame. Optimize the burner position to give maximum absorbance while aspirating the intermediate selenium standard (1.0
mL = 10 g selenium).
13.11 Interrupt the auxiliary oxidant line at the burner
connection and attach the gas lines, the flask header, and the
associated equipment. Connect in series, in this order, the
auxiliary oxidant line, the in-line gas flow regulator, and the
header aeration tube. Then connect the header outlet tube, the
water trap (optional), the one-way check valve (optional), and
the auxiliary oxidant inlet. Use minimum lengths of tubing to
minimize dilution of the selenium hydride. Attach a reaction
flask containing 50 mL of water to the flask header. With argon
flowing through the system, adjust the in-line flow regulator to
permit a maximum flow of the argon sweep gas to the reaction
flask, with negligible solution carryover into the outlet line.
The set-up is then complete.
13.12 If a recorder is used, adjust the span so that an
absorbance of 0.500 from the spectrophotometer reads full
scale on the recorder.
13.13 Rinse the reaction flask and header with water and
introduce the blank, sample, or standard into the reaction flask.
Replace the header and secure to form a tight seal. Allow the
system to stabilize and prepare to record the peak absorbance
or the total absorbance.
13.14 The precision of this test method is highly dependent
on the use of a consistently reproduced technique in this final
step. Inject the 2-mL hypodermic needle through the septum
and quickly add 2.0 mL NaBH4 solution (4 g/100 mL) to the
sample. The H2Se evolution will peak within a few seconds,
but will trail off for up to 30 s afterward. After the H2Se is
swept from the system, remove the header and rinse well with
water. Warning: Selenium hydride is toxic to certain organs of
the body. Avoid inhalation.
13.15 Treat each succeeding sample, blank, and standard in
a like manner.

St 5 0.146X 1 0.49

(2)

where:
St = overall precision, g/L, and
X = concentration of selenium, g/L.
15.2 The bias of this test method determined from recoveries of known amounts of selenium from selenium dioxide and
selenium triphenylchloride in a series of prepared standards are
given in Table 2.
15.3 The information on precision and bias may not apply to
other wastewaters.
15.4 This section on precision and bias conforms to Practice
D2777 77 which was in place at the time of collaborative
testing. Under the allowances made in 1.4 of Practice
D2777 06, these precision and bias data do meet existing
requirements of interlaboratory studies of Committee D19 test
methods.
16. Quality Control
16.1 In order to be certain that analytical values obtained
using these test methods are valid and accurate within the
confidence limits of the test, the following QC procedures must
be followed when analyzing selenium.
16.2 Calibration and Calibration Verification
16.2.1 Analyze at least six working standards containing
concentrations of selenium that bracket the expected sample
concentration, prior to analysis of samples, to calibrate the
instrument. The calibration correlation coefficient shall be
equal to or greater than 0.990. In addition to the initial
calibration blank, a calibration blank shall be analyzed at the
end of the batch run to ensure contamination was not a problem
during the batch analysis.

6
Supporting data have been filed at ASTM International Headquarters and may
be obtained by requesting Research Report D19-1056.

TABLE 1 Overall (ST) and Single-Operator (SO) Interlaboratory

Precision for Selenium by Gaseous Hydride AAS, Test Method A

14. Calculation
14.1 Determine the weight of selenium in each sample by
referring to 12.3. Calculate the concentration of selenium in the
sample in micrograms per litre, using Eq 1:
Selenium g/L 5 ~1000/V! 3 W

Concentration (X),
g/L
2.79
8.50
17.89

(1)

where:
1000 = 1000 mL / Litre
V
= volume of sample, mL, and
W
= weight of selenium in sample, g.

2.69
8.56
18.35

ST
Reagent Water
0.95
1.62
2.98
Natural Water
0.69
1.70
2.13

SO

0.72
1.44
1.71
0.78
1.63
1.67

D3859 08
TABLE 2 Recovery and Bias Data, Test Method A
(Gaseous Hydride AAS)
Amount
Added,
g/L

Amount
Found,
g/L

3
8
17

2.8
8.5
17.9

Recovery,
%

Bias, %

16.5.1 Analyze a reagent water test blank with each batch.

The concentration of selenium found in the blank should be
less than 0.5 times the lowest calibration standard. If the
concentration of selenium is found above this level, analysis of
samples is halted until the contamination is eliminated, and a
blank shows no contamination at or above this level, or the
results must be qualified with an indication that they do not fall
within the performance criteria of the test method.
16.6 Matrix Spike (MS)
16.6.1 To check for interferences in the specific matrix
being tested, perform a MS on at least one sample from each
batch by spiking an aliquot of the sample with a known
concentration of selenium and taking it through the analytical
method.
16.6.2 The spike concentration plus the background concentration of selenium must not exceed the high calibration
standard. The spike must produce a concentration in the spiked
sample that is 2 to 5 times the analyte concentration in the
unspiked sample, or 10 to 50 times the detection limit of the
test method, whichever is greater.
16.6.3 Calculate the percent recovery of the spike (P) using
the following formula:

Statistically
Significant at 95 %
Confidence Level

Reagent Water (Type II)

93
106
105

7
+6
+5

no
no
no

Nonreagent Water (Natural, Waste, and Brine)

3
8
17

2.7
8.6
18.4

90
107
108

10
+7
+8

no
no
yes

16.2.2 Verify instrument calibration after standardization by

analyzing a standard at the concentration of one of the
calibration standards. The concentration of a mid-range standard should fall within 615% of the known concentration.
16.2.3 If calibration cannot be verified, recalibrate the
instrument.
16.3 Initial Demonstration of Laboratory Capability
16.3.1 If a laboratory has not performed the test before, or if
there has been a major change in the measurement system, for
example, new analyst, new instrument, and so forth, a precision
and bias study must be performed to demonstrate laboratory
capability.
16.3.2 Analyze seven replicates of a standard solution
prepared from an Independent Reference Material containing a
midrange concentration of selenium. The matrix and chemistry
of the solution should be equivalent to the solution used in the
collaborative study. Each replicate must be taken through the
complete analytical test method including any sample preservation and pretreatment steps.
16.3.3 Calculate the mean and standard deviation of the
seven values and compare to the acceptable ranges of bias in
Table 2. This study should be repeated until the recoveries are
within the limits given in Table 1. If a concentration other than
the recommended concentration is used, refer to Practice
D5847 for information on applying the F test and t test in
evaluating the acceptability of the mean and standard deviation.
16.4 Laboratory Control Sample (LCS)
16.4.1 To ensure that the test method is in control, analyze
a LCS containing a known concentration of selenium with each
batch or ten samples. If large numbers of samples are analyzed
in the batch, analyze the LCS after every ten samples. The
laboratory control samples for a large batch should cover the
analytical range when possible. The LCS must be taken
through all of the steps of the analytical method including
sample preservation and pretreatment. The result obtained for
a mid-range LCS shall fall within 615 % of the known
concentration.
16.4.2 If the result is not within these limits, analysis of
samples is halted until the problem is corrected, and either all
the samples in the batch must be reanalyzed, or the results must
be qualified with an indication that they do not fall within the
performance criteria of the test method.
16.5 Method Blank

P 5 100 [A~Vs 1 V! BVs# / CV

(3)

where
A = analyte concentration (mg/L) in spiked sample,
B = analyte concentration (mg/L) in unspiked sample,
C = concentration (mg/L) of analyte in spiking solution,
Vs = volume (mL) of sample used, and
V = volume (mL) of spiking solution added.
16.6.4 The percent recovery of the spike shall fall within the
limits, based on the analyte concentration, listed in Test
Method D5810, Table 1. If the percent recovery is not within
these limits, a matrix interference may be present in the sample
selected for spiking. Under these circumstances, one of the
following remedies must be employed: the matrix interference
must be removed, all samples in the batch must be analyzed by
a test method not affected by the matrix interference, or the
results must be qualified with an indication that they do not fall
within the performance criteria of the test method.
NOTE 2Acceptable spike recoveries are dependent on the concentration of the component of interest. See Test Method D5810 for additional
information.

16.7 Duplicate
16.7.1 To check the precision of sample analyses, analyze a
sample in duplicate with each batch. If the concentration of the
analyte is less than five times the detection limit for the analyte,
a matrix spike duplicate (MSD) should be used.
16.7.2 Calculate the standard deviation of the duplicate
values and compare to the precision in the collaborative study
using an F test. Refer to 6.4.4 of Practice D5847 for information on applying the F test.
16.7.3 If the result exceeds the precision limit, the batch
must be reanalyzed or the results must be qualified with an
indication that they do not fall within the performance criteria
of the test method.
16.8 Independent Reference Material (IRM)
6

D3859 08
16.8.1 88 In order to verify the quantitative value produced
by the test method, analyze an Independent Reference Material
(IRM) submitted as a regular sample (if practical) to the
laboratory at least once per quarter. The concentration of the
IRM should be D3859 in the concentration mid-range for the
method chosen. The value obtained must fall within the control
limits established by the laboratory.

19.2.1 In the procedure specified, the magnitude of these

interferences is reduced by matrix modification through the
addition of nickel. The nickel also acts to increase the
sensitivity of this test method and to permit the use of higher
charring temperatures.7,9,10
19.2.2 Residual chemical interferences are compensated for
by a method of standard additions, unless such interferences
are known to be absent.
19.3 Background interferences are also frequently encountered in the determination of selenium. Unless known to be
absent, they are eliminated by means of automatic simultaneous background correction or by means of background
correction at an appropriate nonabsorbing spectral line.
19.4 Spectral interferences are not frequently encountered,
but for selenium determined at 196.0 nm, such an interference
is known to be caused by moderate concentrations of iron. The
interference is eliminated by performing the analysis at a
wavelength free of such interference.

TEST METHOD BGRAPHITE FURNACE AAS

17. Scope
17.1 This test method includes the determination of dissolved and total recoverable selenium in the range from 2 to
100 g/L. The range may be extended by decreasing the sample
size or diluting the original sample.
17.2 This test method has been used successfully with
reagent water, waste treatment plant effluent, tap water, well
water, and treated wood plant effluent. The information on
precision may not apply to waters of other matrices.

20. Apparatus
20.1 Atomic Absorption Spectrophotometer, for use at about
206.0 nm. (Other wavelengths will be used if they are
determined to be both necessary and suitable.)
20.1.1 The optical system, photosensitive device and amplifier, and readout mechanism are to be comparable to those
specified in 10.2. A capability for automatic simultaneous
background correction will be highly advantageous.
20.1.2 Selenium Light SourceA single element hollowcathode lamp is satisfactory. Multielement lamps (if they do
not provide for spectral interferences) and electrodeless discharge lamps may also be used.
20.2 Graphite Furnace, compatible with spectrophotometer,
with temperature programmer, capable of reaching temperatures sufficient to atomize selenium. Ideally, the temperature
programmer should have considerable flexibility so that the
analyst can alter the time-at-temperature profile extensively to
provide assurances that the profile is satisfactory for the
analytical requirements and appropriate for the samples.
20.2.1 Graphite Tubes, standard, compatible with furnace
device.

18. Summary of Test Method

18.1 Selenium is determined by a method of graphite
furnace AAS, expanded to include background correction and
matrix modification, with concentrations determined by a
method of standard additions.
18.1.1 The sample, modifier, and standard are placed in the
furnace where they are dried, charred (ashed or pyrolyzed), and
atomizedthe absorbance being determined during atomization. The analysis is repeated for at least two different standard
levels for each determination.
18.1.2 The analytical concentration is calculated as functions of the sample absorbance (unspiked) and the absorbance/
concentration slope established for the analyte by the method
of standard additions.
18.2 A general guide for the application of this procedure is
given in Practice D3919.
18.3 Dissolved selenium is determined on a filtered sample
with no pretreatment. Total recoverable selenium is normally
determined following acid digestion and filtration. The filtration is omitted only if suspended material is not present.

NOTE 3Metal furnaces have reportedly been used successfully in

place of graphite furnaces. However, no round-robin data are presently
available using these furnaces.

19. Interferences
19.1 For a complete discussion of general interferences,
refer to Practice D3919.
19.2 The determination of selenium by graphite furnace
AAS is especially susceptible to enhancement or suppression
interferences. In the determination of 50 g/L selenium in 5%
HCl, copper, germanium, nickel, antimony, tellurium, titanium,
and tungsten at 10 mg/L elevate the analytical response by 50
to 150 %. Chromium, iron, mercury, molybdenum, vanadium,
and erbium elevate the response by 150 to 300 %. On the other
hand, sulfate and phosphate (and, to a lesser degree, nitrate and
chloride) suppress the evolvement of selenium during atomization.7,8

20.3 Pipettes, microlitre with disposable tips, as required.

20.4 Data Storage and Reduction Devices, Computer- and
Microprocessor-Controlled Devices, or Strip Chart Recorders,
shall be utilized for collection, storage, reduction, and problem
recognition (such as drift, incomplete atomization, changes in
sensitivity, etc.). Strip chart recorders shall have a full scale
deflection time of 0.2 s or less to ensure accuracy.
21. Reagents and Materials
21.1 Hydrogen Peroxide (30%).
9

Atomic Absorption Newsletter, Vol 14, No. 5, SeptemberOctober 1975, p.

127.
7
8

10

Analytical Chemistry, Vol 47, No. 3, March 1975, p. 428.

Atomic Absorption Newsletter, Vol 15, No. 2, MarchApril 1976, p. 29.

109.

Atomic Absorption Newsletter, Vol 14, No. 5, SeptemberOctober 1975, p.

D3859 08
23.2 Measure 100.0 mL of the well-mixed sample into a
125-mL beaker or flask. For total recoverable selenium, add 5
mL of 30% H2O2 and 5 mL of HNO3 (sp gr 1.42). For
dissolved selenium, proceed with 23.5.
23.3 Heat the sample at 95C on a steam bath or hotplate in
a well-ventilated fume hood until the volume has been reduced
to 15 to 20 mL, making sure that the samples do not boil.
23.4 Cool and filter the sample through a suitable filter, such
as fine-textured, acid washed, ashless paper, into a 100-mL
volumetric flask. Wash the filter paper 2 or 3 times with reagent
water and bring to volume. If suspended material is not present,
the filtration may be omitted.
23.5 Inject measured aliquots of sample, nickel nitrate (1.0
mL = 2.0 mg nickel), and reagent water into the furnace, using
the volumes specified in 22.2.1.
23.6 In the absence of automatic simultaneous background
correction, additionally determine the absorbance at an appropriate nonabsorbing line. (The mercury line at 194.2 nm is
usually satisfactory for determinations at 196.0 nm.) Calculate
the background-corrected absorbance as the difference between
the absorbance determined at the analytical line and the
absorbance determined at the nonabsorbing line.
23.7 Carefully observe the absorbance peak assumed to be
due to the analyte. With respect to shape and position, it must
be identical to that derived for the standards, and it must be
completely resolved from other peaks.
23.7.1 If not, modify the time-at-temperature program in
accordance with the manufacturers instructions, to provide for
identical and interference-free behavior of the analyte, and
reestablish the calibration curve.
23.8 Estimate the concentration of selenium in the sample
by reference to the analytical curve. If the apparent concentration is greater than about 40 % of the highest linear concentration, dilute the sample so that the concentration falls near
that value and analyze the sample again.
23.9 Using either the original sample or the diluted sample,
as appropriate, repeat the analysis, but substitute the highest
linear selenium standard in place of the reagent water. This will
provide for the spike verification detailed in Practice D3919.
23.10 For the determination of selenium in many matrices,
analytical verification in this manner will indicate significant
enhancement or suppression effects, which must be compensated for by the method of standard additions. The standard
procedure for this test method (specified in Practice D3919)
should be used unless the abbreviated method (23.10.123.10.5) is shown to be satisfactory.
23.10.1 In some cases, however, the procedure specified in
the general practice may be substantially abbreviated. Certainly for samples where satisfactory relative precision is
indicated by the reproducibility for duplicate analyses, concentrations may be determined directly from the absorbance values
obtained for the original analysis and the spiking verification.
23.10.2 In such cases, the concentration of selenium in the
injected sample may be calculated by the method detailed in
Practice D3919, using the equivalent concentration of the spike
as the concentration of the standard addition.
23.10.3 Alternatively, the concentration may be calculated
using Eq 4:

21.2 Nickel Solution (1 mL = 2.0 mg nickel)

Commercially purchase or dissolve 9.9 g of nickel nitrate
(Ni(NO3)26H2O) in reagent water and dilute to 1 L.
21.3 Nitric Acid (sp gr 1.42)Concentrated nitric acid
(HNO3).
21.4 Nitric Acid (1 + 1)Add 1 volume of HNO3(sp gr
1.42) to 1 volume of water. Always add acid to water.
21.5 Nitric Acid (1 + 499)Add 1 volume of HNO3(sp gr
1.42) to 499 volumes of water. Always add acid to water.
21.6 Selenium Solution, Stock (1.00 mL = 1.00 mg
selenium)Accurately weigh 1.000 g of gray elemental selenium and place in a small beaker. Add 5 mL of HNO3 (sp gr
1.42). Warm until the reaction is complete, then cautiously
evaporate to dryness. Redissolve with HNO3 (1 + 499) and
dilute to 1 L with the same acid solution.
21.6.1 Alternatively, certified selenium stock solutions are
commercially available through chemical supply vendors and
may be used.
21.7 Selenium Solution, Intermediate (1.00 mL = 10 g
selenium)Dilute 5 mL of the selenium stock solution to 500
mL with HNO3(1 + 499).
21.8 Selenium Solution, Standard (1.00 mL = 0.10 g
selenium)Dilute 10 mL of the selenium intermediate solution
to 1000 mL with HNO3 (1 + 499). Prepare fresh daily and
store in a TFE-fluorocarbon or other acceptable container. To
minimize waste only prepare 100 mL of the Selenium Standard
Solution.
21.9 Argon, standard, welders grade. Nitrogen and hydrogen may also be used if recommended by the instrument
manufacturer.
22. Standardization
22.1 Initially set the instrument in accordance with the
manufacturers specifications. Follow the general instructions
provided in Practice D3919, being sure to utilize automatic
simultaneous background correction if available.
22.2 Prepare a calibration curve using a blank and a series
of standards in the appropriate concentration range in accordance with the general instructions, but with the following
modifications:
22.2.1 For each standard and blank, inject into the furnace a
sample volume which is no greater than one-half the maximum
recommended volume for the furnace. Additionally, inject an
aliquot of nickel solution (1.0 mL = 2.0 mg nickel) and an
aliquot of reagent water, each equal to one-half the sample
volume. The distribution of standard, matrix modifier, and
water will then be identical to the distribution of sample,
matrix modifier, and water or standard addition for the
samples.
22.2.2 Utilize the calibration curve to establish the linear
concentration range for the test method and to estimate the
concentration of the analyte in the samples. Determine the
analytical concentration precisely, however, by the method of
standard additions.
23. Procedure
23.1 Clean all glassware by rinsing first with HNO3 (1 + 1)
and then with reagent water. If possible, soak the glassware
overnight in HNO3 (1 + 1).
8

D3859 08
A 3 C/~B 2 A! 5 g/L Se ~in injected sample!

(4)

St 5 0.1072X 1 0.54

where:
A = absorbance of the unspiked sample,
B = absorbance of the spiked sample, and
C = equivalent concentration of the spike, sample basis,
g/L Se.
23.10.4 The precision of the abbreviated test method may
be substantially enhanced by the inclusion of a third data point,
which may be obtained by repeating the spiking verification
procedure with a selenium standard of a concentration equal to
one-half the highest linear concentration. In this way, the
absorbance value for a second standard addition (equivalent to
one-half the addition previously utilized) may be obtained.
23.10.5 For the three-point extrapolation, the concentration
of selenium in the injected sample may be calculated by the
graphic method in 23.10.2.
23.11 As to a provision for spectral interferences, this is
only necessary for one of a group of samples of generally
similar composition, and it is only possible for samples with
apparent undiluted selenium concentrations near or greater
than the highest linear concentration. For such samples, however, the absence of spectral interferences is established as
follows:
23.11.1 Abiding by the specifications established for the
original analysis, repeat the analysis at an alternative wavelength, using a sample concentration which is appropriate with
respect to the decreased sensitivity at the secondary line.
23.11.2 Calculate the concentration of selenium so determined and the concentration originally determined in accordance to 25.1-25.3. If the values are identical within the limits
of precision at the two wavelengths, the analyses are probably
free of spectral interferences and may be performed at either
line. If the values differ significantly, spectral interference has
contributed to the higher value.
23.11.3 If these considerations indicate that the analytical
wavelength should be altered, do so in accordance with the
manufacturers instructions. Prepare a calibration curve at the
new conditions, and analyze the samples again.

where:
St = overall precision, and
X = concentration of selenium determined, g/L.
25.2 For waste-treatment plant effluent, tap water, well
water, and treated wood plant effluent waters, as established by
five laboratories, the overall precision of this test method
within its designated range may be expressed using Eq 7:
St 5 0.2658X 2 0.03

(7)

where:
St = overall precision, and
X = concentration of selenium determined, g/L.
25.3 Because of the large number of metals analyzed in this
study, the requirement for replicate tests was waived; therefore,
single-operator precision is not available.
25.4 For the same collaborative test data, the bias of this test
method are given in Table 3.
25.5 The information on precision and bias may not apply to
other waters.
25.6 This section on precision and bias conforms to Practice
D2777 77 which was in place at the time of collaborative
testing. Under the allowances made in 1.4 of D2777 06, these
precision and bias data do meet existing requirements of
interlaboratory studies of Committee D19 test methods.
26. Quality Control
26.1 In order to be certain that analytical values obtained
using these test methods are valid and accurate within the
confidence limits of the test, the following QC procedures must
be followed when analyzing selenium.
26.2 Calibration and Calibration Verification
26.2.1 Analyze at least three working standards containing
concentrations of selenium that bracket the expected sample
concentration, prior to analysis of samples, to calibrate the
instrument. The calibration correlation coefficient shall be
equal to or greater than 0.990. In addition to the initial
calibration blank, a calibration blank shall be analyzed at the
end of the batch run to ensure contamination was not a problem
during the batch analysis.
26.2.2 Verify instrument calibration after standardization by
analyzing a standard at the concentration of one of the

24. Calculation
24.1 For the standard procedure, calculate the concentration
of selenium in the injected sample in accordance with Practice
D3919 for the method of standard additions.
24.2 For the abbreviated procedure, the calculations for the
concentration of selenium in the injected sample are given in
23.10.2 and 23.10.3.
24.3 Correct the values so obtained for all sample dilutions
prior to analysis, and report the final value using Eq 5:
Selenium, g/L 5 C 3 D

(6)

TABLE 3 Recovery and Bias Data, Test Method B

(Graphite Furnace AAS)
Amount
Added,
g/L

Amount
Found,
g/L

3.0
6.0
14.0

2.5
5.4
13.2

Recovery,
%

Bias, g/L

Bias, %

Statistically
Significant
at 95 % Confidence Level

15
10
5

no
no
no

24
7
10

yes
no
no

(5)
Reagent Water (Type II)

where:
C = concentration in injected sample and
D = total dilution prior to injection.

85
90
95

0.5
0.6
0.8

Nonreagent Water A
6

25. Precision and Bias

25.1 For reagent water, as established by eight laboratories,
the overall precision of this test method within its designated
range may be expressed using Eq 6:

3.0
6.0
14.0

2.3
5.6
12.6

76
93
90

0.7
0.4
1.4

A
Waste treatment plant effluent, tap water, U-gas process condensate, well
water, and treated wood plant effluent.

D3859 08
26.6.2 The spike concentration plus the background concentration of selenium must not exceed the high calibration
standard. The spike must produce a concentration in the spiked
sample that is 2 to 5 times the analyte concentration in the
unspiked sample, or 10 to 50 times the detection limit of the
test method, whichever is greater.
26.6.3 Calculate the percent recovery of the spike (P) using
the following formula:

calibration standards. The concentration of a mid-range standard should fall within 615% of the known concentration.
26.2.3 If calibration cannot be verified, recalibrate the
instrument.
26.3 Initial Demonstration of Laboratory Capability
26.3.1 If a laboratory has not performed the test before, or if
there has been a major change in the measurement system, for
example, new analyst, new instrument, and so forth, a precision
and bias study must be performed to demonstrate laboratory
capability.
26.3.2 Analyze seven replicates of a standard solution
prepared from an Independent Reference Material containing a
midrange concentration of selenium. The matrix and chemistry
of the solution should be equivalent to the solution used in the
collaborative study. Each replicate must be taken through the
complete analytical test method including any sample preservation and pretreatment steps.
26.3.3 Calculate the mean and standard deviation of the
seven values and compare to the acceptable ranges of bias in
Table 3. This study should be repeated until the recoveries are
within the limits given in Table 3. If a concentration other than
the recommended concentration is used, refer to Practice
D5847 for information on applying the F test and t test in
evaluating the acceptability of the mean and standard deviation.
26.4 Laboratory Control Sample (LCS)
26.4.1 To ensure that the test method is in control, analyze
a LCS containing a known concentration of selenium with each
batch or ten samples. If large numbers of samples are analyzed
in the batch, analyze the LCS after every ten samples. The
laboratory control samples for a large batch should cover the
analytical range when possible. The LCS must be taken
through all of the steps of the analytical method including
sample preservation and pretreatment. The result obtained for
a mid-range LCS shall fall within 615 % of the known
concentration.
26.4.2 If the result is not within these limits, analysis of
samples is halted until the problem is corrected, and either all
the samples in the batch must be reanalyzed, or the results must
be qualified with an indication that they do not fall within the
performance criteria of the test method.
26.5 Method Blank
26.5.1 Analyze a reagent water test blank with each batch.
The concentration of selenium found in the blank should be
less than 0.5 times the lowest calibration standard. If the
concentration of selenium is found above this level, analysis of
samples is halted until the contamination is eliminated, and a
blank shows no contamination at or above this level, or the
results must be qualified with an indication that they do not fall
within the performance criteria of the test method.
26.6 Matrix Spike (MS)
26.6.1 To check for interferences in the specific matrix
being tested, perform a MS on at least one sample from each
batch by spiking an aliquot of the sample with a known
concentration of selenium and taking it through the analytical
method.

P 5 100 [A~Vs 1 V! BVs# / CV

(8)

where
A = analyte concentration (mg/L) in spiked sample,
B = analyte concentration (mg/L) in unspiked sample,
C = concentration (mg/L) of analyte in spiking solution,
Vs = volume (mL) of sample used, and
V = volume (mL) of spiking solution added.
26.6.4 The percent recovery of the spike shall fall within the
limits, based on the analyte concentration, listed in Test
Method D5810, Table 1. If the percent recovery is not within
these limits, a matrix interference may be present in the sample
selected for spiking. Under these circumstances, one of the
following remedies must be employed: the matrix interference
must be removed, all samples in the batch must be analyzed by
a test method not affected by the matrix interference, or the
results must be qualified with an indication that they do not fall
within the performance criteria of the test method.
NOTE 4Acceptable spike recoveries are dependent on the concentration of the component of interest. See Test Method D5810 for additional
info rmation.

26.7 Duplicate
26.7.1 To check the precision of sample analyses, analyze a
sample in duplicate with each batch. If the concentration of the
analyte is less than five times the detection limit for the analyte,
a matrix spike duplicate (MSD) should be used.
26.7.2 Calculate the standard deviation of the duplicate
values and compare to the precision in the collaborative study
using an F test. Refer to 6.4.4 of Practice D5847 for information on applying the F test.
26.7.3 If the result exceeds the precision limit, the batch
must be reanalyzed or the results must be qualified with an
indication that they do not fall within the performance criteria
of the test method.
26.8 Independent Reference Material (IRM)
26.8.1 In order to verify the quantitative value produced by
the test method, analyze an Independent Reference Material
(IRM) submitted as a regular sample (if practical) to the
laboratory at least once per quarter. The concentration of the
IRM should be in the concentration mid-range for the method
chosen. The value obtained must fall within the control limits
established by the laboratory.
27. Keywords
27.1 atomic absorption; furnace (Method B); hydride technique (Method A); selenium; total recoverable selenium; water

10

D3859 08
ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.
This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.
This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or service@astm.org (e-mail); or through the ASTM website
(www.astm.org).

11