EXPERIMENTAL METHODS - SR95

days (Table 1). These qualities made it possible for

from Point Barrow, Alaska (PBA). Clay and organic

us to distribute subsamples fortified with VOCs

carbon concentrations for these soils are shown in

for an interlaboratory study where both extraction

Table 2.

and determinative accuracy were evaluated.

In preparation for the vapor fortification treat-

Sixteen laboratories were sent duplicate sub-

ment, the soils were air-dried overnight, sieved

samples of three different soils, each vapor-

through a 30-mesh screen and mixed thoroughly.

fortified with trans-1,2-dichloroethylene (TDCE),

Forty individual subsamples were prepared for

trichoroethylene (TCE), benzene (Ben) and toluene

each soil type by transferring weighed quantities

(Tol). To demonstrate precision between prepara-

into 1.0-mL glass ampoules using a stainless steel

tion batches, each soil set consisted of subsamples

spatula and plastic funnel. The PBA subsamples

weighed to 1.50 0.01 g, while the RMA soil and

that were independently fortified. Thirteen labora-

TB sediment subsamples weighed 2.00 0.01 g.

tories (App. A) submitted analyte concentration

estimates for these soils, using Method 8240 of the

Different weights were chosen so that the volume

SW-846 (U.S. Environmental Protection Agency

of soil was consistent and filled only the body of

1986). This method of determination is based on

the glass ampoule.

methanol extraction purge-and-trap gas chroma-

Vapor fortification was performed on batches of

tography/mass spectrometry (PT/GC/MS) analy-

20 subsamples of a given soil at a time so as to limit

sis. The results from 12 of these laboratories were

the exposure period while the ampoules were

used to establish consensus analyte concentrations

sealed. Each batch of soil-filled ampoules was

for these soils after a statistical evaluation using

placed in a 5.6-L desiccator with a dish of anhy-

drous CaSO4. Following two days of desiccation,

Youden plots, laboratory ranking and Dixon's test

the CaSO4 was replaced with an open 60-mL glass

(Youden and Steiner 1975, Dixon 1953). This re-

bottle containing 50 mL of a solution of the ana-

port documents the round-robin design, the re-

lytes of interest to start the fortification process.

sults (App. B), the statistical evaluation and our

This solution was prepared by combining 25 mL of

conclusions about the merits of VOC performance

a MeOH stock solution (0.60 g of Tol, 0.59 g of

evaluation soil subsamples.

TCE, 0.50 g of TDCE and 0.35 g of Ben diluted to

100 mL) with 25 mL of tetraethylene glycol dimeth-

EXPERIMENTAL METHODS

yl ether (tetraglyme). This equal-volume combina-

tion of MeOH and tetraglyme was found to be rel-

Participating laboratories

atively insensitive to laboratory temperatures (17

Twenty-two laboratories were asked to partici-

to 24C) and produced the most predictable and

pate in this round-robin study. Sixteen labora-

precise treatment levels for the solvents and sol-

tories agreed to participate, of which 13 returned

vent mixtures tested (Hewitt 1993).

completed report forms after receiving and ana-

After seven days of vapor fortification, the des-

lyzing the VOC-fortified soil subsamples.

iccator was opened and a 5-mm-diameter glass

bead was rapidly placed on top of each ampoule,

Preparation and distribution of test materials

forming a temporary cap. Then, as quickly as pos-

Three soil matrices were used in this study.

sible, each ampoule was put in a metal tension

Two of these soils serve as reference matrices for

clamp and the neck was heat-sealed using a pro-

the U.S. Army Environmental Center: a marine

pane plumber's torch. This operation took approxi-

sediment from Tampa Bay, Florida (TB), and a

mately 10 minutes from the time the desiccator

composite soil from the Rocky Mountain Arsenal

(RMA) in Denver, Colorado. The third soil was

was opened until all the ampoules were sealed. A

sharp-pointed tip was created when sealing the

Table 2. Characteristics of soils.

ampoules to facilitate breaking when preparing for

analysis. Details on the development of this proce-

Characteristic

TB*

RMA

PBA

dure and its performance have been documented

Organic carbon (%)

0.31

0.05

6.7

elsewhere (Hewitt 1993, 1994d,e, in press).

Clay (%)

20.1

Each fortification batch and the sequence in

which the individual ampoules were heat sealed

* TB = marine sediment, Tampa Bay, Flor-

ida.

were recorded. Sixteen sets, each consisting of six

RMA = Rocky Mt. Arsenal composite soil,

subsamples, were prepared. Each set contained

Denver, Colorado.

one sample randomly selected from the first 16

PBA = soil, Point Barrow, Alaska.

samples in each of the six treatment batches. This

NA = not available