37
Applications Using GCDetection Systems
Purge andTrapApplicationsUsingTandemPID-ELCD:
USEPAMethods 502.2, 601, 602, 8010, 8020, 8021B
EPAmethods for GC analyses of volatile compounds require purge and trap units for con-
centrating the contaminants inwater, soil, or wastewater.While purge and trap concentration
significantly increases sensitivity, relative to other sample introduction techniques, it does
have a downside: early-eluting volatile compounds typically exhibit broad peaks, due to
inefficient sample transfer from the trap to theGC. This distorted peak shape decreases res-
olution between closely eluting compounds, placing demands on the analytical system and
requiring optimizedGC operating conditions.Although cryofocusing improves separations
of early eluting compounds, most environmental laboratories do not use this approach
because it increases costs.
EPAmethods formonitoring volatiles byGC often recommend using a PID and anELCD,
connected in tandem or series. Coelutions of target compounds are allowed, as long as they
are resolved by the detectors.
13
For example, in Figure 39 bromoform and styrene elutewith
the same retention time, but bromoform elicits a response only from the ELCD and styrene
elicits a response only from the PID. Thus, the selective detectors resolve these two com-
pounds. Because it characteristically produces tailing peaks, the ELCD is themore problem-
atic of the two detectors; sensitivity can be increased, but not without a sacrifice in peak
shape. Optimization of anELCDminimizes tailing andmaximizes sensitivity.
Analysis Time:
Several factors contribute to the total analysis time for volatiles separations,
including purge and trap cycle time, sample analysis time, andGC oven cool-down time
(time required for the oven to cool from the final temperature to the initial temperature for
the next analysis). Long purge and trap cycles are a product of long purge times, dry purges,
long desorb times, and long trap bake times. Long oven cycle times result from low initial
oven temperatures (i.e., subambient to 35°C) and slow temperature program rates.A column
that unnecessarily exceeds the length needed to resolve the analytes can increase analysis
time and cost without significantly adding to the data obtained.
AnRtx
®
-VGC primary column pairedwith anRtx
®
-VRX confirmation columnmake a good
combination for analyzing the compounds listed in Figures 39A&B. The target list
includes unregulated but commonly analyzed compounds such asmethyl-
tert
-butyl ether
(MTBE) and Freon
®
113 (1,1,2-trichloro-1,2,2-trifluoroethane).A 35°C starting temperature
is necessary to resolve Freon
®
113 from 1,1-dichloroethane. Figure 39A shows there are no
early-analyte coelution problems on the primary columnwhen using PID/ELCD detectors in
tandem – the gases and the trihalomethanes are separated.
Figures 40A&B show the analysis ofMethod 8021A/502.2 compounds, without Freon
®
113, using anRtx
®
-VGC column and anRtx
®
-502.2 column.A 50°C initial oven tempera-
ture can be used, which greatly reduces the time needed for theGC to complete the oven
cycle and return to the starting temperature (cycle time) and, therefore, increases through-
put.AnAgilent 5890GC ovenwill cool from 205°C to 35°C in 9minutes; this time, added
to the 28-minute analysis time in Figure 39, produces the fastest cycle time for this analysis:
37minutes. In the analysis in Figure 40, the starting temperature is 50°C, the final tempera-
ture is 200°C, and the oven takes 4minutes to cool. The total cycle time, less than 30min-
utes, is significantly faster than for other pairs of columns. For example, anRtx
®
-VRX col-
umn requires a starting temperature of 40°C; this, combinedwith a 28minute analysis time,
means the total cycle time cannot be faster than 35minutes.
13. EPAMethod 8000B, DeterminativeChromatographic Separations; USEPA. U.S. Government
PrintingOffice:Washington, DC, 1996, Rev. 2.
