then, in most cases, the output signal from the GC will have to go to some external data collection and storage device, usually a recorder. In such a display, the area under the peak of interest is proportional to the amount injected. However, peak height, to a reasonable approximation, can be considered proportional to the amount injected. This approximation is only true for measurements performed under identical conditions, i.e., identical column, identical temperature, etc. Thus if there are substantial changes in ambient temperature such as would be encountered, for example, in going from an indoor to an outdoor location, and the GC is operating at ambient temperature (unheated), then the peak height at each location must be related to a standard run under the same conditions.
b. Calibration
As mentioned before, column selection should be done in the laboratory before going into the field. Calibration should be done in the field. There are two reasons for this. First, the potential differences in ambient temperatures between lab and field could lead to errors as mentioned previously. Second, there is almost always drift in the response of detectors with time. Therefore, calibration should always be done as close to the time of actual use as possible. If experience shows that detector drift is significant over the period of time that samples will be taken, then it will be necessary to make frequent injections of standards during the course of the day. Even in the event that drift does not seem to be a problem, occasional injections of standards throughout the period of use is recommended to make certain that the instrument is operating as expected.
Calibration standards may be of two types: commercially available certified gas mixtures or gas bags prepared from a known gas or liquid aliquot of the desired compound plus a carefully metered volume of diluent, usually air or nitrogen. It is strongly advised that the preparation of standards be done in the laboratory as a practice exercise before attempting it in the field, and that the ability to produce accurate bag standards is verified in the laboratory by comparison with a standard concentration.
If the injection system operates by means of a loop, the calibration curve should be of the form of concentration vs. peak height (obtained from the recorder trace). If a gas-tight syringe is used, then the calibration curve is more conveniently of the form of amount (concentration, w/v, times volume injected) or volume (ppm times injection volume) vs. peak height. The range of standards should encompass the expected ambient concentration at the field site. If this is not known, then preliminary samples should be taken to determine the range of expected concentrations. Duplicate injections of each standard should be made and the average value used for the calibration graph. From the data on the precision of the calibration graph, information about the precision of the measurements and the limits of detection can be obtained using standard statistical methods.
c. Procedure
Once the calibration graph has been generated, the next step is to gather field data. This is done simply by injecting samples of the ambient air periodically into the portable GC and obtaining concentration values by comparison with the calibration graph. The frequency of sample taking is governed by the time necessary for the peak of interest to appear (retention time) plus the time necessary for any other contaminants to elute from the column. As it is seldom possible to know the exact composition of the atmosphere to be sampled beforehand,
