Primary analyses were conducted using a
allowing the phases to separate for 15 min, the
Waters Model 600 system controller, Model 610
ACN phase was carefully removed using a Pas-
fluid unit, Model 717 plus auto injector set for a
teur pipette and placed in a 10-mL graduated cyl-
50-L injection, a 486 UV variable wavelength
inder. An additional 1.0-mL aliquot of ACN was
detector set at 245 nm, and a Maxima chromatog-
then added to the 100-mL volumetric flask, and
raphy work station. Confirmation analysis was
the flask was rotated on the turntable for 15 min.
conducted on a Waters LC Module 1 with a 486
Again the phases were allowed to separate for 15
UV variable wavelength detector (245 nm), a 717
min, and the resulting ACN phase was added to
plus auto injector (50 L), and a Maxima 820 chro-
the 10-mL graduated cylinder. The volume of the
matography work station.
resulting extract was measured and diluted 1:1
with reagent-grade water prior to analysis.
Salting-out solvent extraction/nonevaporative
preconcentration procedure
Cartridge solid-phase extraction
A 251.3-g portion of reagent-grade sodium
Prepacked cartridges of Porapak RDX (Sep-Pak,
chloride was added to a 1-L volumetric flask. A
6 cc, 500 mg) were obtained from Waters Corpora-
770-mL sample of water was measured with a 1-
tion. The cartridges were cleaned by placing them
L graduated cylinder and added to the flask. A
on a Visiprep solid-phase extraction manifold (Su-
stir bar was added and the contents stirred at max-
pelco) and passing 15 mL of acetonitrile through
imum rpm until the salt was completely dis-
each using gravity flow. The acetonitrile was then
solved. A 164-mL aliquot of acetonitrile (ACN),
flushed from the cartridges using 30 mL of reagent-
measured with a 250-mL graduated cylinder, was
grade water. Care was taken to ensure that the
added while the solution was being stirred, and
cartridges were never allowed to dry after the ini-
stirring was continued for at least 15 minutes. If
tial cleaning.
the ACN was slow in dissolving due to poor mix-
A connector was placed on the top of each car-
tridge and fitted with a length of 1/8-in.-diameter
ing, a Pasteur pipette was used to withdraw a
portion of the undissolved ACN layer and rein-
Teflon tubing. The other end of the tubing was
ject it into the vortex of the stirring aqueous phase.
placed in a 1-L fleaker containing 500 mL of sam-
After equilibrium has been established, only about
ple. The vacuum was turned on and the flow rate
5 mL of ACN should remain in a separate phase.
through each cartridge was set at about 10 mL/
The stirrer was turned off and the phases allowed
min. If the flow rate declined significantly due to
to separate for 15 min. If no emulsion was present,
partial plugging from suspended material, it was
the ACN phase was removed and placed in a 100-
readjusted. After the sample had been extracted,
mL volumetric flask and 10 mL of fresh ACN was
the top plug containing the fitted tubing was re-
added to the 1-L flask. The 1-L flask was again
moved from each cartridge, and 10 mL of reagent-
stirred for 15 min, followed by 15 min for phase
grade water was passed through the cartridge,
separation. The ACN was removed and combined
using gravity flow unless the cartridges were
sufficiently plugged to require vacuum. A 5-mL
with the initial extract in the 100-mL volumetric
aliquot of acetonitrile was used to elute retained
flask. When emulsions were present, that portion
analytes from the cartridges under gravity flow.
of the sample was removed from the volumetric
The volume of the recovered ACN was measured
flask with a Pasteur pipette, placed in a 20-mL
and diluted 1:1 with reagent-grade water.
scintillation vial and centrifuged for 5 min at 2000
rpm. The supernate was also pipetted into the 100-
mL volumetric flask, the scintillation vial was
Membrane solid-phase extraction
Empore styrene-divinylbenzene membranes
rinsed with 5 mL of acetonitrile, and the acetoni-
(47 mm) were obtained from 3M Corporation. The
trile was added to the 100-mL volumetric flask.
membranes were designated SDB-RPS and were
For the first extraction, the pellet that formed af-
not commercially available at the time the study
ter centrifugation was added back to the 100-mL
was conducted. The styrene-divinylbenzene used
flask, but if it occurred in the second extraction, it
in these membranes had been modified to pro-
was discarded.
vide extra retention for polar organics such as
To reduce the volume of ACN, an 84-mL ali-
HMX.* The membranes were precleaned by cen-
quot of salt water (325 g NaCl per 1000 mL of
water) was then added to the 100-mL volumetric
flask. The flask was placed on a vertical turntable
* Craig Markell, 1995, 3M Corporation, personal communi-
and rotated at about 60 rpm for 15 min. After
cation.
3
Previous Page
