The Reston Chlorofluorocarbon Laboratory

Low-Level VOCs Lab

Water samples are analyzed for VOCs using a purge and trap GC-ECD method similar to that of Busenberg and Plummer (1992) that is used at the USGS Chlorofluorocarbon Laboratory, Reston, VA for CFC analyses.

Analytical Method

The analytical system is calibrated daily for quantitative analysis of CFC-11, CFC-12, and CFC-113 using an air standard calibrated on the SIO-98 scale. A sample volume of approximately 34 cm3 is stripped of VOCs for a period of 4 minutes using ultra-pure N2 that has passed through a mole sieve. The system and analytical procedure is nearly identical to that of Busenberg and Plummer (1992), with these few exceptions: the analysis time was extended from 10 to 30 minutes to detect a larger suite of VOCs; the 3 meter analytical column was packed with Porasil C and heated at 72°C in the GC oven; a constant flow of about 25 cm3/min of ultra-pure N2 carrier gas was used; and the ECD was maintained at 300°C.

Limitations/Advantages

Although the MDLs can be as low as 1 picogram per liter (pg/L, 1x10-6 ug/L, or 1 part per quadrillion) for some halogenated VOCs using purge and trap GC-ECD methods, the ECD does not detect most non-halogenated hydrocarbons. High concentrations (tens of parts per billion, ppb) of some halogenated VOCs can be difficult to quantify by GC-ECD because of non-linear response of ECDs, especially at higher concentrations. The sensitivity of the ECD also varies with the electron affinity of different halogenated VOCs. The identification of compounds in ground-water samples by GC-ECD is based on retention time, and for the compounds identified, concentrations are determined using calibration standards. Many purgeable halogenated VOCs are detected in GC-ECD chromatograms of untreated drinking water, but have not been detected in any of the commercially available standards analyzed and are, therefore, unidentified.

The GC-ECD analytical procedure was not designed for quantitative analysis of contaminated samples. For example, large concentrations (greater than a few ug/L) can be underestimated by GC-ECD analysis due to peak .clipping. during peak integration. Further, large contaminant peaks can obscure detection and quantification of other VOCs on chromatograms.

Because of the limitations of GC-ECD methods for analyzing a wide range of VOCs, mass-spectrometric methods are more commonly used in identifying and quantifying concentrations of VOCs in ground water. With assessment levels on the order of about 0.2 to 0.02 ug/L, the mass-spectrometric method is sufficient to meet USEPA regulatory requirements for drinking water. The primary advantage of purge and trap GC-ECD methods relative to the mass spectrometric analysis is the extremely low MDL, which allows vulnerable waters to be detected more frequently than they would be using GC-MS.

GC-ECD peak identification

Compound retention times in the purge and trap GC-ECD analytical system are determined by (1) analyzing commercial gas mixtures containing concentrations of 100 ppb of specific VOCs in nitrogen, (2) analyzing commercially available single VOC compounds dissolved in methanol that are then diluted in ultra-pure nitrogen, (3) comparing chromatograms of commercial standards containing various combinations of the more common halogenated VOCs, and (4) comparing detections in GC-ECD chromatograms of ground-water samples with the corresponding GC-MS compound identifications and analyses for the same sample. Sixty-three VOCs, with retention times between 2.0 to 28.1 minutes, have been detected by GC-ECD in ground water. Twenty-five of these compounds are identified (Table 1) by retention time. Peak identification by GC-ECD may not be unique for a specific compound, and depending on instrumental conditions and peak area, compound identification can be obscured by overlapping peaks (multiple VOCs with similar retention times).

Table 1: Halogenated VOCs detected, approximate retention time in minutes, and minimum detection level (MDL) in picograms per liter (pg/L).

Compound Retention Time (Minutes) Alternate Name GC-ECD MDL pg/L*
sulfur hexafluoride 1.99 25
dichlorodifluoromethane 2.53 CFC-12 5.5
Halon 1211 3.36 Halon 1211 1.3
dichlorotetrafluoroethane 3.52 CFC-114 22
trichlorofluoromethane 4.34 CFC-11 1.2
chloromethane 4.88 methyl chloride 326
1,1-dichloroethene 6.00 72
bromomethane 6.64 methyl bromide 356
vinyl chloride 6.66 1435
trichlorotrifluoroethane 7.13 CFC-113 9.4
trans-1,2-dichloroethene 8.53 902
methyl iodide 9.45 7.5
carbon tetrachloride 10.17 19
dichloromethane 12.42 methylene chloride 564
chloroethane 12.49 1390
trichloromethane 15.93 chloroform 52
trichloroethene 16.52 TCE 20
1,1,2-trichloroethane 16.54 2522
1,1,2,2-tetrachloroethane 16.72 466
cis-1,2-dichloroethene 18.84 3292
tetrachloroethene 20.35 PCE 9.9
1,2-dichloroethane 20.50 435
bromochloromethane 20.77 37
1,1,1-trichloroethane 23.24 methyl chloroform 161
1,1-Dichloroethane 28.11 25

*For sample volume of 34 cc.

Concentration ranges detected by GC-ECD

The GC-ECD is calibrated daily for CFC-11, CFC-12, and CFC-113 using air standards as described for CFC analyses. Concentrations of these CFCs are well known, and have MDLs of about 1, 5, and 9 pg/L for CFC-11, CFC-12, and CFC-113 in water, respectively. The MDLs for CFCs determined by GC-ECD are more than 4 orders of magnitude lower than the LT-MDL for CFC-11 (80,000 pg/L) and CFC-12 (180,000 pg/L) obtained in GC-MS analysis at the USGS NWQL (Table 1).

The concentrations of most of the VOCs identified in Table 1 (above) were quantified using commercially available halogenated VOC gas standards. The GC-ECD MDL is defined for each VOC as two times the peak area of the smallest peak that could be quantified for that VOC above background. The minimum concentrations detected by GC-ECD are given in Table 1 where they are compared to the LT-MDL from the GC-MS analysis. For many VOCs, there were an insufficient number of detections to permit compound identification or quantification of MDLs. Furthermore, correlation of many VOC peak areas from the GC-ECD with GC-MS analyses could not be made because the concentrations were below the LT-MDL of the GC-MS analysis.