Formaldehyde is one of the most prevalent of the carbon-based compound gases in urban areas as well as the remote troposphere. Along with its direct effects on flora and fauna, CH2O is a major source of odd hydrogen, which affects the oxidizing capacity of the atmosphere. Over the years, atmospheric chemists have used a variety of techniques to measure formaldehyde, but each has its limitations:
Air samples can be trapped in cartridges and later analyzed. Though portable and inexpensive, this technique is unwieldy for frequent sampling and prone to systematic errors.
In liquid-phase techniques, air can be concentrated into a solution for later analysis. High-resolution results can be obtained, but the process is labor-intensive.
Spectrometers can analyze the absorption patterns of a beam of ultraviolet or infrared light to infer its chemical make-up. This provides quick, highly specific results, but the instruments are larger, costlier, and more physically demanding.
A more compact adaptation, the NCAR tunable diode laser absorption spectrometer (TDLAS), was used as the linchpin of this intercomparison. Developed by ACD's tunable diode laser group (Alan Fried, Scott Sewell, Bruce Henry, and Bryan Wert), it has been used in field studies for three years and made its in-flight debut last year. Field tests of TDLAS show its resolution to be as fine as 40 to 50 parts of formaldehyde per trillion parts of air. It is among the instruments least prone to interference.
The main intercomparison took place in mobile laboratories at the Mesa Lab from 29 May to 3 June 1995, conducted by all four of the AGU presenters. "What's unique about this study is that the formaldehyde standards we used were verified by four independent techniques before the experiment began," says Alan.
The first phase of the study used synthetic air spiked with standard levels of formaldehyde and various potential interferences, with the TDLAS as a reference base. Participants submitted their data in a blind fashion to the NCAR referee team. Ambient air was sampled in the second phase, with TDLAS treated as one of the blind test participants. The three liquid-phase instruments were provided by the University of Rhode Island, Brookhaven National Laboratory, and Texas Tech University. Cartridge-based instruments were provided by Washington State University and ManTech Environmental (Research Triangle Park, North Carolina). The weather lent itself to to the experiment, providing a range of conditions over the study week from sunshine to storminess.
When the spiked-air measurements were arranged linearly for analysis, the regression slopes for the liquid-phase instruments ranged from 68% to 138% of that obtained by TDLAS. According to Alan, "This large scatter resulted from the fact that the three liquid-phase instruments did not employ gas-phase standards to correct for variability in calibration, collection, and recovery efficiencies." After each instrument wa calibrated using the same standards applied to TDLAS, rather than those normally used by each research group, the range of agreement narrowed to 87-104%.
Complications arose with the cartridge-based measurements. One problem with all cartridge-based techniques is that ozone can produce both positive and negative interference. "Before this study," says Alan, "it was commonly believed that only one of the types of cartridge methods we studied required an ozone scrubber. However, our results showed that all of them needed a scrubber. Also, we found that the cartridge methods, like the liquid-phase methods, require gas-phase standards to get accurate results."
According to Jack, the intercomparison shows that a range of techniques can work well as long as they are calibrated directly in the field using accurate gas-phase standards. The similar results eventually obtained from totally different techniques also works to increase confidence in the accuracy of each. "The lessons learned from this study will help significantly in providing high-quality formaldehyde measurements in future atmospheric oxidation studies," says Jack.
An NCAR scientist for 33 years and formerly head of CGD, Warren has served as president of the American Meteorological Society and is a fellow of both the AMS and the American Association for the Advancement of Science. He currently serves on the National Science Board.
Warren's photo is being taken by NCAR's Carlye Calvin. Many of the collection's photos from the early 20th century are by Addison Scurlock, whose Washington-based career as one of America's great photographers spanned more than six decades.