In this study, we tested two methods of com-
white phosphorus particles have, over the
positing. The first was designed to estimate aver-
years, resulted in a partitioning that favors
age concentration and variance to conform with
adsorption of molecular white phosphorus to
current protocols where a cleanup level might be
sediment surfaces.
set. The second was designed to detect the pres-
White phosphorus concentrations in the fine-
ence of hot spots containing high, acutely toxic
grain fraction of samples will be significantly
levels of white phosphorus. Details as to numbers
higher in samples that contain solid particles
and sizes of samples are given in the Methods sec-
than those without.
tion, and an overview of the rationale, objectives,
If composite samples are formed by taking
assumptions, and approach follows.
several sediment samples from designated
areas and mixing each with water from those
areas, the fine-grained sediment will remain
COMPOSITE SAMPLING METHODS
suspended in the water column longer than
the solid chunks of white phosphorus. The
Stirred composites
The traditional method of homogenizing com-
concentrations of white phosphorus found in
posite soil samples uses drying, grinding, and
the suspended sediment and water following
sieving. Because of the potential hazard and loss
thorough mixing will be correlated with the
of white phosphorus by sublimation and oxida-
total amounts within the sample.
tion when soils are dried, this procedure is not
To meet the above objective, composite sam-
applicable for sediments contaminated by white
pling allows the analysis of fewer samples
phosphorus.
than does discrete sampling.
We have found that white phosphorus con-
The limit of quantitation of the analytical
tamination is generally confined to areas with
method is adequately low to accommodate
standing water. Previous work has demonstrated
possible dilution effects caused by combining
that white phosphorus concentrations in the wa-
some blank samples with one or more con-
ter column above contaminated sites are generally
taminated samples.
with the concentrations found in the sediment
Approach
(Walsh 1995). However, if the sediments are dis-
Collect samples systematically using a two-
turbed by stirring, water column concentrations
dimensional grid with spacing designed to
detect a single hot spot, if one exists, with
sediment concentrations. Our method for com-
positing white-phosphorus-contaminated sam-
than 10%.
ples was tested based on these features. We com-
Composite the samples by thorough mixing
posited sediment samples from areas known to be
and take five subsamples of the composite. To
contaminated, and mixed the sediment with an
sample only the fine-grained fraction, add
equal volume of water from the area. We then
water to the composite, mix thoroughly and
sampled the suspended sediment. Following
sample the water and suspended sediment.
analysis, concentrations were expressed on a dry
Repeat with additional water to see if preci-
weight basis.
sion is increased.
Determine white phosphorus in subsamples
Objective of sampling
of the composite.
We need to estimate mean white phosphorus
concentration to rank contaminated areas by
Sieved composites
If the objective of our sampling strategy is to
in concentration after a remedial action.
determine whether solid pieces of white phosphorus
are available to waterfowl, the large size difference
Assumptions
between typical ERF sediments and milligram-size
Small areas (hot spots) having high concen-
white phosphorus particles presents us with an
trations and containing solid particles of
opportunity to collect numerous point samples,
white phosphorus punctuate much larger
which can be composited and then reduced in vol-
areas containing low or undetectable concen-
ume by sieving through an appropriate size mesh.
trations.
White phosphorus particles, if present, are retained
Slow dissolution and diffusion from solid
on the mesh and can be detected by various
4