Table 1. Polymeric tubing used in sampling trace-level organics.
Surface-area-
Cost
to-solution-
per foota
Dimensions (cm)
Length
volume ratio
(cm1)
($)
I.D.
O.D.
wall
(cm)
Flexible polymersb
polyproplyene-based material with plasticizer, (formulation 1)
0.58
0.64
0.95
0.16
20
6.3
polypropylene-based material with plasticizer, (formulation 2)
2.48
0.64
0.95
0.16
20
6.3
polyvinylchloride (PVC)
0.89
0.64
0.95
0.16
20
6.3
thermoplastic elastomerc (TPE)
0.96
0.64
0.95
0.16
20
6.3
linear copolymer of vinylidene fluoride and
hexafluoropropylene P(VDF-HFP)
1.99
0.64
0.80
0.08
20
6.3
polyurethane
0.64
0.64
0.95
0.16
20
6.3
fluoroelastomer
8.70
0.64
0.95
0.16
20
6.3
Rigid polymersd
polyethylene, low density (LDPE)
0.19
0.64
0.95
0.16
20
6.3
polyethylene, cross-linked high density (XLPE)
0.43
0.64
0.95
0.16
20
6.3
polyethylene liner in ethyl vinyl acetate shell
0.57
0.64
0.95
0.16
20
6.3
polyethylene liner cross-linked to ethyl vinyl acetate shell
1.08
0.64
0.95
0.16
20
6.3
co-extruded polyester lining in PVC shell
0.77
0.64
0.95
0.16
20
6.3
polypropylene (PP)
0.27
0.64
0.95
0.16
20
6.3
polyamide (nylon)
0.71
0.71
0.95
0.12
18
5.6
polytetrafluoroethylene (PTFE)
4.27
0.75
0.95
0.10
17
5.3
perfluoroalkoxy (PFA)
5.58
0.64
0.95
0.16
20
6.3
ethylene tetrafluoroethylene (ETFE)
5.50
0.48
0.64
0.08
27
8.4
polyvinylidene fluoride (PVDF)
1.80
0.64
0.95
0.16
20
6.3
fluorinated ethylene-propylene (FEP)
3.90
0.64
0.95
0.16
20
6.3
FEP-lined polyethylene
3.00
0.64
0.80
0.08
20
6.3
a
Cost varies with quantity, dimensions, and supplier.
b
Finger pressure can collapse tubing.
c
Styrene-ethylene-butylene block copolymer modified with silicon oil.
d
Can be stepped on without collapsing the tubing.
loss of the analytes during the filling process, the
eral volumes of deionized water and left to air-dry.
solutions in these vials served as controls and thus
One end of each of the tubings was plugged with a
were used to determine the initial analyte concen-
glass rod whose diameter matched the internal
trations for each sampling time.
diameter of the tubing. The glass rod was inserted
When it was time to sample a tubing, one of the
in the tubing to a depth of 1 cm, and then the out-
plugged ends of the tubing was cut with a special
side of the tubing was clamped with a plastic tub-
cutter for rigid tubings and then a Pasteur pipet
ing clamp. (The length of the glass plugs was taken
was used to transfer an aliquot of the test solution
into account when figuring the surface areas and
to a 1.8-mL HPLC autosampler vial. The control
solution volumes.) For each type of tubing, there
solutions were removed from the refrigerator and
were five sampling times (1, 8, 24, 48, and 72 hours)
allowed to warm before analysis.
and two replicates for each sampling time (i.e., 10
Analytical determinations were performed us-
tubing pieces of each material).
ing RP-HPLC. A modular system was employed
For each sample time, the tubings were filled in
consisting of a Spectra Physics SP8875 autosam-
random order using a glass re-pipettor. The top of
pler with a 100-L injection loop, a Spectra Physics
the tubings was sealed immediately after filling by
SP8810 isocratic pump, a Spectra Physics SP8490
inserting a glass plug, leaving no head space, and
variable wavelength detector set at 215 nm, and a
then clamped as described previously. The tubings
Hewlett Packard 3396 series II digital integrator.
were stored in the dark at room temperature. Dur-
Separations were obtained on a 25-cm 0.46-cm (5
ing this process, three high-performance liquid
m) LC18 column (Supelco) eluted with 65/35
chromatography (HPLC) autosampler vials (1.8
(V/V) methanol/water at a flow rate of 2.0 mL/
mL) were filled with the test solution at the begin-
min. The detector response was obtained from the
ning and at the end of filling each set of tubings
digital integrator operating in the peak height
(i.e., for each time period). The vials were filled so
mode.
there was no headspace, capped with Teflon-lined
For each analyte, a single compound standard
plastic caps, and stored in the dark in a refrigera-
was made by adding the neat (undiluted) com-
tor. Because we anticipated there would be some
5
Previous Page
