description of this analysis has been published
Table 4. Characteristics of laboratory-treated soils.
elsewhere (Hewitt 1994a, 1994b).
Silt and
Grain
Wt.†
Sand
clay
size*
(m)
Matrix
(%)
(%)
(g)
EXPERIMENTAL
Ottawa sand
100
400
4
Rocky Mountain Arsenal
NA
NA
NA
4
Several reference materials of different solid-
Lebanon landfill
45
55
300
4
waste matrices, along with field-contaminated and
CRREL soil
NA
NA
NA
4
laboratory-treated soils, were analyzed to assess
Tampa Bay sediments
95
5
200
4
Ft. Edwards clay
30
70
30
2
the performance of both the FP and RF/Comp. Kα
normalization.
* 95% cut off
† Weight of soil subsample spiked
NA Not analyzed
Reference materials
Table 3 lists the certified reference materials
Alfa, Johnson Matthey). A complete description
purchased from NIST and the Resource Technol-
of how these soil subsamples were treated has been
ogy Corporation (RTC). Those purchased from
presented elsewhere (Hewitt 1994b). The soil char-
NIST have certified concentrations for the total
acteristics and the weight of the treated subsamples
amount of metal present, whereas the RTC mate-
are shown in Table 4.
rials report certified values based on the USEPA
Briefly, the soils were air dried and thoroughly
SW846, 3000-series metal acid extraction proce-
mixed prior to placing subsamples into 31-mm-
dures (U.S. EPA 1986).
diameter analysis cups. Analyte spikes were made
by pipetting between 0.4 and 0.025 mL quantities
Field samples
of the aqueous standards directly onto the indi-
Six river sediment subsamples from a suspected
vidual soil subsamples, increasing the metal
hazardous-waste site were analyzed. These field
concentration by 1000, 500, 250, 125, or 0 g/g.
subsamples were taken from thoroughly homog-
Only five analytes were applied to a set of five rep-
enized samples that had been previously charac-
licate soil subsamples in order to limit the total
terized by inductively coupled plasma (ICP) analy-
volume of solution added to each subsample (0.2
sis following 3000-series metal acid extraction pro-
cedures (U.S. EPA 1986).
mL/g). Following this protocol, one set of 30 soil
subsamples (5 6) was spiked with Cr, Cu, Zn, As,
and Pb, a second with Ni, Se, Hg, Tl, and Co, and
Treated laboratory soils
Six different soils were spiked with Cr, Cu, Zn,
a third with Ag, Cd, Sn, Sb, and Ba. Analyte addi-
As, Pb, Ni, Se, Hg, Tl, Co, Ag, Cd, Sn, Sb, and Ba
tions were performed so that a given soil sub-
by using concentrated 10,000 mg/L aqueous pure
sample was not treated with the same concentra-
element atomic absorption standards (AESAR/
tion more than once (Table 5). A sixth untreated
subsample of each soil type was also analyzed with
each group and served as the matrix blank.
Table 3. List of certified reference materials.
National Institute of Standards and Technology
RESULTS AND DISCUSSION
SRM 1645--River sediment
Table 6 lists the estimates of detection reported
SRM 1646--Estuarine sediment
SRM 1579a--Powdered lead-based paint
by the manufacturers for FP analysis, along with
SRM 2704--Buffalo River sediment
some values established using the RF/Comp. Kα
SRM 2709--San Joaquin soil
normalization and method detection limit (MDL)
SRM 2710--Montana soil
(Federal Register 1984). This table also includes the
SRM 2711--Montana soil
analyte intensities measured for the RMA soil
treated with 1000 g metal/g. These intensities
Resource Technology Corporation
were included to provide a means of predicting a
CRM012--Incinerated sludge
detection limit, based on the assumption that there
CRM014--Baghouse dust
is a fairly constant inverse relationship between
CRM013--Paint chips
these two parameters. This table indicates that re-
CRM020--Soil (from EPA Superfund site)
CRM021--Soil (from contaminated waste site)
gardless of the method of analysis, all of these
4
Previous Page
