Headspace sampling works well for analytes
was obtained from Aldrich Chemical Co. (Milwau-
with a Henry's Law constant, the equilibrium parti-
kee, Wisconsin). The white phosphorus was sup-
tion coefficient for a chemical between water and
plied as a 5-g stick. It came immersed in water and
the headspace above it, above 90 atm-cm3/mole
had a white coating on its surface. Pieces (100300
(Zhang and Pawliszyn 1993a, b). Since Henry's
mg) of white phosphorus were cut from the stick
Law constant for white phosphorus is over 2000
under water. Care was taken to ensure that the sur-
atm-cm3/mole (Spanggord et al. 1985), headspace
faces of each piece of white phosphorus were freshly
SPME is feasible. We were particularly interested
cut, lustrous in appearance, and showed no evi-
in headspace sampling because environmental
dence of white coating. These translucent, colorless
matrices containing white phosphorus are poten-
pieces were transferred under water to a test tube,
tially complex (surface waters, wastewater, sedi-
and the test tube was placed in a beaker filled with
ments), and the SPME fiber would be damaged or
water. The water with white phosphorus pieces was
heated to 54C to melt the white phosphorus, which
fouled by immersion into these matrices. Head-
space sampling relies on the equilibrium partition-
remains as a separate phase. A Gilson Microman
ing of the analyte among the matrix, headspace and
Positive Displacement Pipet was used to obtain 25-
L (45 mg) droplets of white phosphorus, which
SPME phase (Zhang and Pawliszyn 1993a, b). The
extraction is not exhaustive, and for semi-volatiles
were placed in separate test tubes under water. The
only a small portion of the total analyte present
water was cooled to solidify the droplets.
becomes associated with the fiber. If depletion of
A stock solution for calibration standards was
the analyte is negligible, the same aliquot of a sam-
prepared under nitrogen by dissolving 90 mg of
ple can be analyzed first by headspace, then by
white phosphorus in 250 mL of toluene (Aldrich
solvent extraction. Analysis of the same sample
Chemical Co.). Standards over the range 7.2 to 7200
g/L were prepared by diluting the stock solution
aliquot by headspace SPME followed by solvent
extraction allows direct comparison of the two
with methanol. Aqueous standards over the range
of 0.0072 to 0.72 g/L were prepared by adding 25
methods without the confounding effect of heter-
L of a methanol standard to individual 25-mL ali-
ogeneity, and makes SPME a valuable tool for
screening samples for contamination. Because the
quots of water in 40-mL glass vials sealed with a
majority of samples sent to analytical labs for the
septum.
analysis of volatiles or semi-volatiles tend to be
Aqueous solutions of white phosphorus were
blank (devoid of the analytes of interest at analyti-
prepared by placing pieces of white phosphorus
cal detection limits), considerable time and solvent
into an amber jug containing 4 L of reagent-grade
could be saved by screening samples for contam-
water (Type I) (MilliQ, Millipore) with no head-
ination prior to solvent extraction and analysis.
space, and agitating the jug for more than 60 days.
The objectives of this study were to
1. Determine if headspace SPME provides a de-
Matrices
tection capability similar to solvent extraction
Blank matrices used to prepare spiked samples
methods for white phosphorus in water and
were reagent-grade (Type I) water (MilliQ, Milli-
sediment matrices.
pore); groundwater from a domestic well in
2. Test the feasibility of calibrating headspace
Weathersfield, Vermont; surface water from a pond
SPME for the quantitative analysis of water
in Hanover, New Hampshire; surface water from a
samples.
salt marsh in Anchorage, Alaska; Ottawa sand pur-
chased from U.S. Silica (Ottawa, Illinois); a loamy
3. Evaluate the performance of headspace SPME
soil from the U.S. Army Environmental Center (Ab-
with field-contaminated water and sediment
erdeen Proving Ground, Maryland); and a sandy
samples.
silt from Lebanon, New Hampshire. Soil samples
4. Compare concentration estimates obtained
were wetted to 100% moisture (dry weight basis,
via headspace SPME with estimates obtained
i.e., equal parts water and soil) prior to spiking.
by solvent extraction.
Field-contaminated samples were obtained from
5. Evaluate the suitability of headspace SPME
Eagle River Flats, Fort Richardson, Alaska. Water
for use in a field laboratory.
samples were collected in 1-L amber glass bottles
and sediment samples were collected in 120-mL
METHODS
wide-mouth jars filled so that there was no head-
space. Samples were maintained at 4C until ex-
Standards and spiking solutions
An analytical standard for white phosphorus
tracted.
2
Previous Page
