6.4 Conclusion  6.3 Measurements of Chemical Composition  6.3.4 Resonance and Laser-Induced Fluorescence

6.3.5 Ambient Measurements of Chemical Composition - VOCs and NO_x, NO_yMeasurements

Background

Atmospheric chemistry is extremely complex with a vast array of compounds participating in the chemical soup around us and above us. In this section, we examine two classes of compounds, volatile organic compounds (VOCs) and NO_x [nitrogen oxide (NO) and nitrogen dioxide (NO₂)] which are important precursors in the formation of tropospheric ozone in rural and urban areas. A high level of ozone is indicative of a polluted air mass and has been shown to have deleterious effects on human health and plant life.

Measurements of ozone were first made in the mid 1850s using wet chemical techniques. By the early 1900s, the molecular structure of ozone was known. Soon, stratospheric measurements of ozone were being recorded and it was recognized that the majority of total ozone in the atmosphere was found the in the stratosphere. The early researchers believed that the stratosphere was indeed the source for all observed ozone in the troposphere. However, in the 1940s and 1950s observation of high ozone concentrations in the Los Angeles basin discounted this theory (Haagen-Smit, 1952). Further studies conducted in environmental chambers (in which studies are carried out under simulated atmospheric conditions) showed that ozone may be formed from 

$$VOCs + NO_x + O_2 + h\nu \rightarrow\rightarrow O_3 + {\rm other\ products}$$ (6.26)

 Here, $h\nu$ represents light energy required to drive the reaction. Equation (6.26) actually represents a net result of a complex reaction sequence, the details of which do not concern us here. It became clear that tropospheric ozone formation is possible under conditions in which the four ingredients on the left side of (6.26) are simultaneously present. Because of the deleterious effects of ozone, the U.S. Environmental Protection Agency has sought to limit the amount of O₃ produced in urban and rural areas by limiting the emission of anthropogenic VOC precursors. Recently researchers have realized that natural sources of VOCs can significantly complicate the control strategies. These recent events have led scientists to believe that systematic and reliable measurements of NO_x and VOCs are needed in order to determine the best possible ozone control strategies.

Measurement Techniques

   Volatile Organic Compounds

We will restrict our discussion here to an important subset of VOCs, the non-methane hydrocarbons. These include compounds such as ethane, butane, benzene, etc. As indicated in the previous discussion, these compounds play an important role in photochemical smog formation in rural and urban areas. To gain a better understanding of the process leading to ozone formation and hence insight into effective control strategies, it has been recognized that speciated hydrocarbon measurements are essential. This is a difficult task owing to the often very large number of compounds present. The most widely used technique for these measurements is gas chromatographic separation followed by detection. The detector used most commonly is a flame-ionization detector (FID). The FID is a non-specific detector that has shown great utility in hydrocarbon measurements because it responds nearly linearly with carbon number of the compound of interest. Figure 8a shows a typical experimental arrangement for measuring hydrocarbons. Figure 8b shows an example of a chromatogram that is the result of the analysis. The concentration unit in these measurements is typically parts per billion carbon (ppbC). To illustrate the difference between ppb and ppbC, consider n-butane which, in a particular sample, is present at 10 ppb. The calibrated FID response for this compound would be 40 ppbC since each carbon present on the molecule contributes equally to the total signal and there are four carbons present in n-butane. Another commonly used GC detector for non-methane hydrocarbons is mass spectrometry. This method has its greatest utility when used for identification of species. Neither of these detectors is sufficiently sensitive to allow for real time measurements of typical ambient air. To get around this problem, scientists use a technique called enrichment. Here, a volume of air is passed through a device which retains (traps) the material of interest (e.g., NMHCs) while allowing the remainder to pass through. The volume of air that is passed through is then measured. The two major types of trapping are cryogenic trapping on glass beads and solid adsorbent trapping. The analysis begins when the trapped material is desorbed onto a gas chromatographic column where the separation of the complex mixture takes place. This is followed by detection. A typical duty cycle is 30-60 minutes/sample.

Measurements of hydrocarbons may be performed in-situ, i.e., where an instrument is deployed at the site of interest, or with grab samples. If grab samples are taken, they must be transported to a laboratory where the measurement is performed. The most common type of container used for grab samples of hydrocarbons is an electropolished stainless steel canister.

There has been some concern about the reliability of hydrocarbon measurements by scientists throughout the world. There are many potential problems and highly trained personnel are needed for precise measurements. Proper calibration of the gas chromatograph is one of the keys to generation of reliable data. Calibration is achieved by determining the response factor of a detector for analysis of standards of known concentration. A response factor may be determined for each compound of interest by running a standard for each compound. Standard gases may be fabricated by individual scientists or purchased from laboratories providing the service. The National Institute for Standards and Technology (USA) and the National Physical Laboratory (UK) are examples of major sources for gravimetric standards. To prepare a gravimetric standard, compounds are precisely weighed so that the fraction of standard compound to added air or nitrogen is known, thus yielding standard mixtures in the ppb to ppm range. Because the FID detector responds nearly linearly with carbon number, it is possible to calibrate the GC with one or several hydrocarbons and use the response factors generated to estimate the concentrations of all of the hydrocarbons in a given air sample (e.g., ppbC). This is important when one realizes that it is not uncommon for over 100 compounds to be present in an ambient sample.

An ongoing study at NCAR, NOMHICE (Non-Methane Hydrocarbon Intercomparison Experiment) is monitoring the ability of individual scientists around the world to identify and quantify a large range of hydrocarbon species (Apel et al., 1993). The goal of this experiment is to help ensure reliable hydrocarbon measurements throughout the world. Early results from this study suggests that there are difficulties with identification and quantification of mixtures of even a few hydrocarbons.

   Nitrogen Oxide Species

It is well recognized by the atmospheric science community that reactive nitrogen species play a major role in atmospheric chemistry. Reliable measurements of NO, NO_x, (NO + NO₂) and $NO_y\,(NO + NO_2 + NO_3 +$ peroxyacetyl nitrate (PAN) + ...) are crucial to the understanding of the chemistry of the atmosphere leading to photochemical smog production. These compounds are also important participants in acid deposition and reactions leading to the production of the hydroxyl radical. It is clear that there are large anthropogenic sources of these compounds (mostly NO) and the elucidation of biogenic sources is proceeding at a rapid pace but large uncertainties remain. Photochemical reaction of NO and NO₂ with various organic and inorganic species convert them into reservoir species within the NO_y group. Reliable measurement techniques have been developed for NO, NO₂ and NO_y. These are described briefly below. However, the implementation of these techniques in a large scale monitoring network to track ozone precursors has not yet occurred.

   NO

Measurement techniques capable of detecting NO down to 10 ppt or less exist. There are two techniques which are presently used: laser induced fluorescence or LIF (Bradshaw et al., 1985) and the more commonly used chemiluminescence detection (e.g., Ridley and Howlett, 1974; Schiff et al., 1979, Mcfarland et. al. 1979, Ridley and Grahek, 1990). The LIF technique requires sophisticated optical equipment and was developed in part as an intercomparative method for the chemiluminescence detection scheme. In order to fluoresce, NO must first be optically pumped to an excited electronic state. This is achieved through a resonant, two-photon process. NO is first excited from the ground electronic state to the $A^2\Sigma$ state with a UV laser emitting radiation at 266 nm. This intermediate "resonant" state is then excited with a second photon (infrared laser, wavelengths ranging from 1.06 to 1.15 $\mu$m) to the $D^2\Sigma$ level from which the fluorescence occurs. This technique is inherently highly selective and has proven to be very sensitive as well with detection limits in the low pptv range.

The chemiluminescent technique is based upon the reaction of NO with ozone, which is introduced into the instrument in large excess over NO. This reaction forms NO₂ in the excited state. The emission from this excited state species can be measured and, from this, the ambient NO concentration determined. This technique is also highly selective and sensitive, and is linear in response over a wide dynamic range.

   NO₂

Nitrogen dioxide in the atmosphere has been measured by a variety of methods, and some of these have been simultaneously intercompared (Fehsenfeld et al., 1990). The NO detectors described above can be used by first converting a portion of the ambient NO₂ into NO and then measuring the increase in signal. For this method to be accurate a highly specific conversion technique is needed. The best method to date is photolysis of NO₂ by ultraviolet radiation, which is the same photolytic reaction that takes place in the atmosphere. Other conversion techniques, such as the surface-mediated reaction on ferrous sulfate, are not as accurate because nitrogen oxide species other than NO₂ are converted as well (Fehsenfeld et al., 1987). Because of the rapid interchange of NO and NO₂ in the atmosphere, NO₂ detection based on these conversion methods also requires accurate and rapid coincident measurements of NO. Another method used for detection of ambient NO₂ is based on the chemiluminescent reaction between NO₂ and luminol. Instruments based on this reaction are highly sensitive and have rapid response but are also susceptible to interference from O₃ and PAN and have a somewhat nonlinear response. Nevertheless, many scientists use these instruments because they are small, lightweight, and require little power to operate.

Finally, laser systems have been deployed to measure ambient levels of NO₂. Tunable diode lasers (TDLAS) use the infrared absorption features of NO₂ to qualify the level of this compound in the atmosphere. These instruments are very specific and are reasonably sensitive but require highly trained operators to obtain accurate results. Other laser systems use ultraviolet or visible spectral features to detect atmospheric NO₂. These systems are also very specific but they generally require very long path lengths (a few km) because the absorption cross section at these wavelengths is relatively small.

   NO_y

There are two common techniques for measuring NO_y. These involve the conversion of all NO_y species to NO with subsequent detection of NOby the chemiluminescent technique (see for example, Bollinger, 1983; Dickerson, 1985). In one technique, NO_y species are converted to NO using CO as a reducing agent and gold as a catalytic surface (Fahey et al., 1985) upon which the following reaction occurs: 

$$NO_y + CO \;{{\rm Gold}\atop {\rightarrow\rightarrow}}\;NO + CO_2 + {\rm other\products}$$ (6.27)

This type of NO_y detector is used only by research groups and not for routine monitoring. In another technique, a heated molybdenum surface is used (300 - 500C) to aid in the reduction of NO_y to NO. In this case, the molybdenum surface is not a true catalyst but actually takes part in a surface reaction which may be represented (for NO₂) as : 

$$2NO_2 + Mo \rightarrow\rightarrow 2NO + MoO_2$$ (6.28)

 Unlike the gold convertor, both commercial and research molybdenum convertors are in use. The efficiency of a convertor can be measured by introducing a known standard (e.g., NO₂) and measuring the amount of NO produced for a given known amount of standard. Each type of convertor has been known, under certain conditions, to lose efficiency with time. Under such condtions, the convertors must be reconditioned. Each convertor has its own set of advantages and disadvantages which will not be discussed here. The reader is referred to an intercomparison between the two convertors which was carried out in 1987 by Fehsenfeld et al.

In the proper implementation of any of the above techniques, it is imperative that high quality calibrations be performed with reliable standards. 

  
 6.4 Conclusion  6.3 Measurements of Chemical Composition  6.3.4 Resonance and Laser-Induced Fluorescence 
http://www.asp.ucar.edu