17
BroadPeaks:
Peak broadening is another problem often experiencedwhen analyzingVOCs
by purge and trapmethods. Broad peaks are caused either by poor sample transfer from the
purge and trap to theGC or by dead volumewithin either system. Broad peaks frequently
result from dead volume in the connection between the purge and trap unit and theGC sys-
tem. (See pages 18-21 for connectionmethods.) Because trap desorption is a relatively slow
process, the sample band transferred to theGC can be verywide. To reduce this bandwidth,
the samplemust be transferred quickly and/or refocused at the inlet of theGC column
through a secondary cold trapping technique such as cryofocusing or subambient cooling.
The transfer time can be reduced by using the desorb preheat feature. During this step, the
trap is heated to 5°C below the desorption temperature, and the valve is positioned so no
flow passes through the trap. This helps the compounds trapped on the adsorbents to rapidly
migrate from the trapwhen backflushing begins.
The desorb flow rate alsowill affect the bandwidth. If the desorb flow is too low
(
<9mL/min.),
the band becomes broad (Figure 12) andmust be refocused at the column inlet. If faster flow
rates are used (>9mL/min.) in conjunctionwith long, thick-film columns, the bandwidth can be
reduced enough so that secondary trapping is not required. Ideally, desorbing at a flow rate of
20-30mL/min. yields a very narrow bandwidth. However, when using narrow bore columns, it
might be necessary to split the flow at the injection port tomaintain column efficiency.
Foaming Samples:
Analysts deal with foaming samples in two primaryways: by dilution or
by addition of an anti-foaming agent. Diluting the sample compromises the detection limit,
but in the endmay save instrument downtime.Anti-foaming agents such as polydimethyl-
siloxane and silicon dioxidemethylcellulose are designed to reduce foaming of surfactants in
a liquidmatrix. These are effective at preventing a sample from foaming, but they generally
produce artifact peaks that can interferewith the target analytes.An anti-foam blankmust be
run prior to samples to determine the contribution of artifact peaks from the anti-foaming
agent. If dilution or anti-foaming agents do not reduce foaming or if samples have not been
screened for surfactants, use a 5 or 10mL sample in a 25mL purge vessel to prevent the bub-
bles from entering the fittings and, ultimately, the trap. If you are running an unattended
autosampler, you can insert a plug of deactivated fused silica or glasswool into the top of
the purge vessel to prevent foam from entering the purge and trap lines. If all else fails con-
sider switching to a fritless sparge tube and increasing the purge time to effectively remove
the volatile analytes.
3.50
min.
4.50
5.50
6.50 7.50
8.50 9.50 10.50 11.50
5,6
4
3
2
1
Figure 12.
A low desorb flow can produce tailing peaks, as in this example, desorbed at 9mL/min.
1. benzene
2.
α
,
α
,
α
-trifluorotoluene (SS)
3. toluene
4. ethylbenzene
5. m-xylene
6. p-xylene
7. o-xylene
EPAMethod 8020, Rtx
®
-5SilMS column, 40m, 0.45mm ID, 1.5µm (cat. #12798)
Carrier gas:
9mL/min.@ constant pressure
GC:
Finnigan 9001
Detector:
FID
Oven temp.:
40°C (hold 2min.)
85°C@ 4°C/min.
(hold 1min.) to
225°C@ 40°C/min.
(hold 2min.)
7
Wastewater samples commonly
contain surfactants and other
material that can contaminate
the concentrator.
