Whole Air Sampling aboard the DC-8 and P-3B Aircraft During
TRACE-P

Donald R. Blake and F. Sherwood Rowland
Department of Chemistry, University of California, Irvine 92697-2025

Canisters will be evacuated in our UCI laboratory after which 10 torr of purified water vapor will be added to each canister. Adding water helps avoid the problem of wall reactions for some gases in some very dry samples. Once the approximately 150 samples from each flight are filled the cans will be ferried back to our lab by research group members returning from the field. The average time between canister filling and sample analysis is one week.

Each analytical apparatus utilizes two gas chromatographs (GCs) and with 4 detectors and one gas chromatograph mass spectrometer (GC-MS). Each whole air sample is cryogenically trapped with liquid nitrogen, warmed, and injected into a hydrogen flow stream. This stream is then split into five, with each stream feeding a separate GC column. One DB-5MS, one PLOT A12O3/KCl, and two DB-1 columns are used. The DB-5MS is plumbed into an electron capture detector (ECD) and separates about 20 C1-C2 halocarbons. The DB-1/ECD combination detects essentially the same halocarbons as the DB-5MS/ECD but the separation order is different enough that gases co-eluting on one column are usually resolved on the other. All ECDs are oxygen-doped which yields a substantial increase in ECD response, especially for many of the hydrohalocarbons. Several halocarbons and alkyl nitrates are quantified on the GC-MS system.  The DB-1 FID combination separates C3-C10 NMHCs. The PLOT column, also plumbed to an FID, is used for separating the C2-C5 NMHCs, some of which are not resolved adequately by the DB-1 column.

The preparation of standards for the halocarbons employs a pressure balancing method using three different sections of a glass vacuum line. Pure gas is introduced into the first section of the line and is ultimately diluted to a mixing ratio that most closely matches the concentration of the gas in the atmosphere. The range of these halocarbon standards falls between 0.5-600 pptv.

Calibration of the NMHC compounds has been achieved by employing Scott calibration gases available in the 1-100 ppmv mixing ratio range. A NIST propane standard with an absolute accuracy of 0.99±0.0l ppmv is also used in the preparation of standards. In order to simulate whole air characteristics, a known amount of a particular calibration gas is added to a known amount of a previously assayed stratospheric sample. These high altitude samples contain less than 50 pptv of ethane and are usually devoid of any other hydrocarbons (<2 pptv), except methane.  After adding the Scott standards containing the various gases, the augmented air sample concentrations are calculated. In this manner, standards have been prepared over a wide range of concentrations which are then used as primary standards for calibrating the 7500 liter pressurized whole air working standards collected at either Niwot Ridge, CO, Scripps Institute of

Oceanography, La Jolla, CA, or White Mountain Research Station, Bishop, CA. During analysis one aliquot from the working standard is assayed every fifth analysis, thereby quickly indicating any malfunction of the analytical apparatus. To determine any drift in the working standard several other whole air pressurized Aculife and SpectraSeal treated Luxfer cylinders and other stainless steel containers are assayed at rates ranging from once per day to once per week. Alkyl nitrate standards will be provided by E. Atlas and F. Flocke of NCAR.

The stainless steel canisters that will be used in this study were fabricated at UCI and prepared for use by a technique developed over the 18 years since the first canisters were made here. The canisters have been used on several previous NASA and NSF airborne missions during which very clean tropospheric or stratospheric air was collected. Samples from those projects have been assayed within a few days after collection and then again several weeks later and the concentrations of the reported gases exhibited no statistically significant differences between the two data sets.

Canister arrays like the one shown in the photo each consist of 24 2-L evacuated canisters.  Before each flight, seven arrays will be loaded onto the DC-and 6 arrays onto the P-3B. Outside air is collected from beyond the laminar boundary layer of the aircraft and is pumped via 1/4" stainless steel tubing through a two-stage metal bellows pump and a gas handling manifold. To fill an individual canister the operator first builds up pressure in the manifold, opens the valve to the canister, then waits for the pressure inside the canister to reach 40 psi before closing the valve.

One sample will be collected approximately every 3-4 minutes during horizontal flight, and one sample every 1.5 to 2 minutes during vertical profile spirals. Each canister takes about 2.5 minutes to fill at an altitude of 12 km but only about 40s in the boundary layer.